slurm submission log: 2024-05-29 11:13:48.479645
created following sbatch script: 

###############################

#!/bin/bash

#SBATCH --account=nlp
#SBATCH --cpus-per-task=16
#SBATCH --dependency=afterok:7667677
#SBATCH --gres=gpu:2
#SBATCH --job-name=tthrush-job-1958780
#SBATCH --mem=100G
#SBATCH --nodelist=sphinx2
#SBATCH --open-mode=append
#SBATCH --output=/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/llms/pythia-70m_arc_easy_1/train_job_output.txt
#SBATCH --partition=sphinx
#SBATCH --time=14-0

# activate your desired anaconda environment
. /nlp/scr/tthrush/miniconda3/etc/profile.d/conda.sh ; conda activate pretraining-coreset-selection

# cd to working directory
cd .

# launch commands
srun --unbuffered run_as_child_processes 'torchrun --master_port 29522 --nproc_per_node=2 train_llm.py --dataset_id /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/data/arc_easy --output_dir /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/llms/pythia-70m_arc_easy_1 --output_hub_id pythia-70m_arc_easy --model_id EleutherAI/pythia-70m --learning_rate 1e-3 --warmup_ratio=0.1 --gradient_accumulation_steps 2 --per_device_train_batch_size 256 --seed 1 --num_train_epochs 14'

###############################

submission to slurm complete!


###############################
slurm submission output

Submitted batch job 7667678



###############################

slurm submission log: 2024-05-30 08:40:45.380067
created following sbatch script: 

###############################

#!/bin/bash

#SBATCH --account=nlp
#SBATCH --cpus-per-task=16
#SBATCH --gres=gpu:2
#SBATCH --job-name=tthrush-job-86467
#SBATCH --mem=100G
#SBATCH --nodelist=sphinx2
#SBATCH --open-mode=append
#SBATCH --output=/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/llms/pythia-70m_arc_easy_1/train_job_output.txt
#SBATCH --partition=sphinx
#SBATCH --time=14-0

# activate your desired anaconda environment
. /nlp/scr/tthrush/miniconda3/etc/profile.d/conda.sh ; conda activate pretraining-coreset-selection

# cd to working directory
cd .

# launch commands
srun --unbuffered run_as_child_processes 'torchrun --master_port 29504 --nproc_per_node=2 train_llm.py --dataset_id /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/data/arc_easy --output_dir /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/llms/pythia-70m_arc_easy_1 --output_hub_id pythia-70m_arc_easy --model_id EleutherAI/pythia-70m --learning_rate 1e-3 --warmup_ratio=0.1 --gradient_accumulation_steps 2 --per_device_train_batch_size 256 --seed 1 --num_train_epochs 1'

###############################

submission to slurm complete!


###############################
slurm submission output

Submitted batch job 7670593



###############################

###############################
start time: 2024-05-30 14:18:40.533540
machine: sphinx2
conda env: pretraining-coreset-selection
###############################
running following processes

	torchrun --master_port 29504 --nproc_per_node=2 train_llm.py --dataset_id /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/data/arc_easy --output_dir /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/llms/pythia-70m_arc_easy_1 --output_hub_id pythia-70m_arc_easy --model_id EleutherAI/pythia-70m --learning_rate 1e-3 --warmup_ratio=0.1 --gradient_accumulation_steps 2 --per_device_train_batch_size 256 --seed 1 --num_train_epochs 1


###############################
command outputs: 


[2024-05-30 14:18:48,946] torch.distributed.run: [WARNING] 
[2024-05-30 14:18:48,946] torch.distributed.run: [WARNING] *****************************************
[2024-05-30 14:18:48,946] torch.distributed.run: [WARNING] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
[2024-05-30 14:18:48,946] torch.distributed.run: [WARNING] *****************************************
05/30/2024 14:19:17 - INFO - __main__ - Script parameters ScriptArguments(seed=1, dataset_id='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/data/arc_easy', output_dir='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/llms/pythia-70m_arc_easy_1', output_hub_id='pythia-70m_arc_easy', hf_hub_token=True, model_id='EleutherAI/pythia-70m', per_device_train_batch_size=256, num_train_epochs=1.0, learning_rate=0.001, gradient_accumulation_steps=2, from_scratch=True, warmup_ratio=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, weight_decay=0.01, lr_scheduler_type='cosine', local_rank=0, resume_from_checkpoint=False, deepspeed=None, peft=False)
05/30/2024 14:19:17 - INFO - __main__ - Script parameters ScriptArguments(seed=1, dataset_id='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/data/arc_easy', output_dir='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/llms/pythia-70m_arc_easy_1', output_hub_id='pythia-70m_arc_easy', hf_hub_token=True, model_id='EleutherAI/pythia-70m', per_device_train_batch_size=256, num_train_epochs=1.0, learning_rate=0.001, gradient_accumulation_steps=2, from_scratch=True, warmup_ratio=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, weight_decay=0.01, lr_scheduler_type='cosine', local_rank=0, resume_from_checkpoint=False, deepspeed=None, peft=False)
/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/huggingface_hub/file_download.py:1132: FutureWarning: `resume_download` is deprecated and will be removed in version 1.0.0. Downloads always resume when possible. If you want to force a new download, use `force_download=True`.
  warnings.warn(
/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/huggingface_hub/file_download.py:1132: FutureWarning: `resume_download` is deprecated and will be removed in version 1.0.0. Downloads always resume when possible. If you want to force a new download, use `force_download=True`.
  warnings.warn(

  0%|          | 0/10712 [00:00<?, ?it/s][rank1]:[W reducer.cpp:1360] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration,  which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator())
[rank0]:[W reducer.cpp:1360] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration,  which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator())

  0%|          | 1/10712 [00:56<167:04:25, 56.15s/it]
  0%|          | 2/10712 [01:29<127:20:50, 42.81s/it]
  0%|          | 3/10712 [02:00<111:46:44, 37.58s/it]
  0%|          | 4/10712 [02:32<104:29:41, 35.13s/it]
  0%|          | 5/10712 [03:02<99:35:44, 33.49s/it] 
  0%|          | 6/10712 [03:37<100:45:49, 33.88s/it]
  0%|          | 7/10712 [05:04<152:13:06, 51.19s/it]
  0%|          | 8/10712 [06:10<166:47:27, 56.10s/it]
  0%|          | 9/10712 [06:32<135:02:43, 45.42s/it]
  0%|          | 10/10712 [07:05<123:14:45, 41.46s/it]
  0%|          | 11/10712 [07:24<102:43:11, 34.56s/it]
  0%|          | 12/10712 [07:43<88:49:30, 29.89s/it] 
  0%|          | 13/10712 [08:00<77:11:10, 25.97s/it]
  0%|          | 14/10712 [08:16<68:15:01, 22.97s/it]
  0%|          | 15/10712 [08:35<64:52:07, 21.83s/it]
  0%|          | 16/10712 [08:53<61:08:37, 20.58s/it]
  0%|          | 17/10712 [09:07<55:07:04, 18.55s/it]
  0%|          | 18/10712 [09:20<49:55:28, 16.81s/it]
  0%|          | 19/10712 [09:32<46:19:49, 15.60s/it]
  0%|          | 20/10712 [09:44<42:46:53, 14.40s/it]
  0%|          | 21/10712 [09:56<41:03:04, 13.82s/it]
  0%|          | 22/10712 [10:11<41:34:20, 14.00s/it]
  0%|          | 23/10712 [10:25<41:51:18, 14.10s/it]
  0%|          | 24/10712 [10:38<40:24:27, 13.61s/it]
  0%|          | 25/10712 [10:50<38:49:58, 13.08s/it]
                                                     
{'loss': 10.6516, 'grad_norm': 1.4530287981033325, 'learning_rate': 2.332089552238806e-05, 'epoch': 0.0}

  0%|          | 25/10712 [10:50<38:49:58, 13.08s/it]
  0%|          | 26/10712 [11:00<36:09:32, 12.18s/it]
  0%|          | 27/10712 [11:09<33:52:38, 11.41s/it]
  0%|          | 28/10712 [11:19<32:14:52, 10.87s/it]
  0%|          | 29/10712 [11:28<30:42:21, 10.35s/it]
  0%|          | 30/10712 [11:37<29:49:24, 10.05s/it]
  0%|          | 31/10712 [11:46<28:31:19,  9.61s/it]
  0%|          | 32/10712 [11:54<27:14:52,  9.18s/it]
  0%|          | 33/10712 [12:03<27:10:04,  9.16s/it]
  0%|          | 34/10712 [12:11<26:14:14,  8.85s/it]
  0%|          | 35/10712 [12:19<24:57:23,  8.41s/it]
  0%|          | 36/10712 [12:26<23:54:24,  8.06s/it]
  0%|          | 37/10712 [12:33<22:44:07,  7.67s/it]
  0%|          | 38/10712 [12:39<21:47:02,  7.35s/it]
  0%|          | 39/10712 [12:46<21:15:24,  7.17s/it]
  0%|          | 40/10712 [12:53<21:03:29,  7.10s/it]
  0%|          | 41/10712 [12:59<20:13:46,  6.82s/it]
  0%|          | 42/10712 [13:05<18:58:25,  6.40s/it]
  0%|          | 43/10712 [13:11<19:03:27,  6.43s/it]
  0%|          | 44/10712 [13:17<18:35:59,  6.28s/it]
  0%|          | 45/10712 [13:23<17:54:38,  6.04s/it]
  0%|          | 46/10712 [13:28<17:39:03,  5.96s/it]
  0%|          | 47/10712 [13:34<17:34:37,  5.93s/it]
  0%|          | 48/10712 [13:39<16:40:04,  5.63s/it]
  0%|          | 49/10712 [13:44<16:08:57,  5.45s/it]
  0%|          | 50/10712 [13:49<16:02:18,  5.42s/it]{'loss': 9.9144, 'grad_norm': 1.3664460182189941, 'learning_rate': 4.664179104477612e-05, 'epoch': 0.0}

                                                     

  0%|          | 50/10712 [13:49<16:02:18,  5.42s/it]
  0%|          | 51/10712 [13:55<15:49:32,  5.34s/it]
  0%|          | 52/10712 [14:00<15:50:44,  5.35s/it]
  0%|          | 53/10712 [14:06<16:13:27,  5.48s/it]
  1%|          | 54/10712 [14:11<16:10:33,  5.46s/it]
  1%|          | 55/10712 [14:16<15:16:05,  5.16s/it]
  1%|          | 56/10712 [14:20<14:30:37,  4.90s/it]
  1%|          | 57/10712 [14:25<14:55:00,  5.04s/it]
  1%|          | 58/10712 [14:29<14:09:42,  4.79s/it]
  1%|          | 59/10712 [14:34<13:42:27,  4.63s/it]
  1%|          | 60/10712 [14:39<13:56:18,  4.71s/it]
  1%|          | 61/10712 [14:43<13:42:23,  4.63s/it]
  1%|          | 62/10712 [14:47<12:59:27,  4.39s/it]
  1%|          | 63/10712 [14:51<12:58:08,  4.38s/it]
  1%|          | 64/10712 [14:55<12:29:56,  4.23s/it]
  1%|          | 65/10712 [15:00<12:37:28,  4.27s/it]
  1%|          | 66/10712 [15:04<12:53:12,  4.36s/it]
  1%|          | 67/10712 [15:08<12:35:03,  4.26s/it]
  1%|          | 68/10712 [15:12<12:35:07,  4.26s/it]
  1%|          | 69/10712 [15:15<11:25:08,  3.86s/it]
  1%|          | 70/10712 [15:19<11:29:23,  3.89s/it]
  1%|          | 71/10712 [15:23<11:15:44,  3.81s/it]
  1%|          | 72/10712 [15:26<10:33:43,  3.57s/it]
  1%|          | 73/10712 [15:29<10:23:55,  3.52s/it]
  1%|          | 74/10712 [15:33<10:26:27,  3.53s/it]
  1%|          | 75/10712 [15:36<10:12:51,  3.46s/it]{'loss': 9.2157, 'grad_norm': 1.1206352710723877, 'learning_rate': 6.996268656716417e-05, 'epoch': 0.01}
                                                     

  1%|          | 75/10712 [15:36<10:12:51,  3.46s/it]
  1%|          | 76/10712 [15:40<10:09:33,  3.44s/it]
  1%|          | 77/10712 [15:43<10:00:44,  3.39s/it]
  1%|          | 78/10712 [15:46<9:46:03,  3.31s/it] 
  1%|          | 79/10712 [15:50<10:26:20,  3.53s/it]
  1%|          | 80/10712 [15:54<10:28:11,  3.55s/it]
  1%|          | 81/10712 [15:57<10:08:18,  3.43s/it]
  1%|          | 82/10712 [16:00<9:52:07,  3.34s/it] 
  1%|          | 83/10712 [16:03<9:24:18,  3.19s/it]
  1%|          | 84/10712 [16:05<9:02:41,  3.06s/it]
  1%|          | 85/10712 [16:08<8:52:23,  3.01s/it]
  1%|          | 86/10712 [16:11<8:50:16,  2.99s/it]
  1%|          | 87/10712 [16:15<9:03:41,  3.07s/it]
  1%|          | 88/10712 [16:18<9:01:24,  3.06s/it]
  1%|          | 89/10712 [16:20<8:32:31,  2.89s/it]
  1%|          | 90/10712 [16:23<8:31:51,  2.89s/it]
  1%|          | 91/10712 [16:26<8:38:17,  2.93s/it]
  1%|          | 92/10712 [16:28<7:59:08,  2.71s/it]
  1%|          | 93/10712 [16:31<7:41:30,  2.61s/it]
  1%|          | 94/10712 [16:33<7:38:40,  2.59s/it]
  1%|          | 95/10712 [16:36<7:42:18,  2.61s/it]
  1%|          | 96/10712 [16:38<7:39:38,  2.60s/it]
  1%|          | 97/10712 [16:40<7:15:14,  2.46s/it]
  1%|          | 98/10712 [16:42<6:47:59,  2.31s/it]
  1%|          | 99/10712 [16:44<6:31:44,  2.21s/it]
  1%|          | 100/10712 [16:46<6:17:47,  2.14s/it]
                                                     
{'loss': 8.4427, 'grad_norm': 0.795574426651001, 'learning_rate': 9.328358208955224e-05, 'epoch': 0.01}

  1%|          | 100/10712 [16:46<6:17:47,  2.14s/it]
  1%|          | 101/10712 [16:48<6:12:25,  2.11s/it]
  1%|          | 102/10712 [16:50<6:04:05,  2.06s/it]
  1%|          | 103/10712 [16:53<6:12:47,  2.11s/it]
  1%|          | 104/10712 [16:54<5:53:19,  2.00s/it]
  1%|          | 105/10712 [16:56<5:47:09,  1.96s/it]
  1%|          | 106/10712 [16:58<5:43:27,  1.94s/it]
  1%|          | 107/10712 [17:00<5:33:11,  1.89s/it]
  1%|          | 108/10712 [17:02<5:34:34,  1.89s/it]
  1%|          | 109/10712 [17:04<5:28:32,  1.86s/it]
  1%|          | 110/10712 [17:05<5:23:28,  1.83s/it]
  1%|          | 111/10712 [17:07<5:26:25,  1.85s/it]
  1%|          | 112/10712 [17:09<5:17:58,  1.80s/it]
  1%|          | 113/10712 [17:11<5:36:59,  1.91s/it]
  1%|          | 114/10712 [17:13<5:33:07,  1.89s/it]
  1%|          | 115/10712 [17:15<5:30:45,  1.87s/it]
  1%|          | 116/10712 [17:16<5:19:07,  1.81s/it]
  1%|          | 117/10712 [17:18<5:11:18,  1.76s/it]
  1%|          | 118/10712 [17:20<4:58:04,  1.69s/it]
  1%|          | 119/10712 [17:21<4:56:51,  1.68s/it]
  1%|          | 120/10712 [17:23<4:51:35,  1.65s/it]
  1%|          | 121/10712 [17:24<4:48:28,  1.63s/it]
  1%|          | 122/10712 [17:26<4:49:14,  1.64s/it]
  1%|          | 123/10712 [17:28<4:44:50,  1.61s/it]
  1%|          | 124/10712 [17:29<4:34:11,  1.55s/it]
  1%|          | 125/10712 [17:31<4:43:31,  1.61s/it]
                                                     
{'loss': 7.8155, 'grad_norm': 0.4470035433769226, 'learning_rate': 0.00011660447761194031, 'epoch': 0.01}

  1%|          | 125/10712 [17:31<4:43:31,  1.61s/it]
  1%|          | 126/10712 [17:32<4:38:52,  1.58s/it]
  1%|          | 127/10712 [17:34<4:28:15,  1.52s/it]
  1%|          | 128/10712 [17:35<4:23:32,  1.49s/it]
  1%|          | 129/10712 [17:37<4:28:22,  1.52s/it]
  1%|          | 130/10712 [17:38<4:32:12,  1.54s/it]
  1%|          | 131/10712 [17:40<4:22:57,  1.49s/it]
  1%|          | 132/10712 [17:41<4:23:26,  1.49s/it]
  1%|          | 133/10712 [17:43<4:24:17,  1.50s/it]
  1%|▏         | 134/10712 [17:44<4:13:33,  1.44s/it]
  1%|▏         | 135/10712 [17:45<4:05:44,  1.39s/it]
  1%|▏         | 136/10712 [17:47<4:08:19,  1.41s/it]
  1%|▏         | 137/10712 [17:48<4:12:24,  1.43s/it]
  1%|▏         | 138/10712 [17:50<4:14:28,  1.44s/it]
  1%|▏         | 139/10712 [17:51<4:04:07,  1.39s/it]
  1%|▏         | 140/10712 [17:52<4:03:16,  1.38s/it]
  1%|▏         | 141/10712 [17:53<3:52:54,  1.32s/it]
  1%|▏         | 142/10712 [17:55<3:50:11,  1.31s/it]
  1%|▏         | 143/10712 [17:56<3:52:08,  1.32s/it]
  1%|▏         | 144/10712 [17:57<3:49:13,  1.30s/it]
  1%|▏         | 145/10712 [17:59<3:46:37,  1.29s/it]
  1%|▏         | 146/10712 [18:00<3:42:18,  1.26s/it]
  1%|▏         | 147/10712 [18:01<3:43:14,  1.27s/it]
  1%|▏         | 148/10712 [18:02<3:41:51,  1.26s/it]
  1%|▏         | 149/10712 [18:04<3:40:47,  1.25s/it]
  1%|▏         | 150/10712 [18:05<3:37:26,  1.24s/it]{'loss': 7.3687, 'grad_norm': 0.38879531621932983, 'learning_rate': 0.00013992537313432834, 'epoch': 0.01}

                                                     

  1%|▏         | 150/10712 [18:05<3:37:26,  1.24s/it]
  1%|▏         | 151/10712 [18:06<3:33:09,  1.21s/it]
  1%|▏         | 152/10712 [18:07<3:35:56,  1.23s/it]
  1%|▏         | 153/10712 [18:08<3:33:48,  1.21s/it]
  1%|▏         | 154/10712 [18:09<3:30:53,  1.20s/it]
  1%|▏         | 155/10712 [18:11<3:36:39,  1.23s/it]
  1%|▏         | 156/10712 [18:12<3:28:05,  1.18s/it]
  1%|▏         | 157/10712 [18:13<3:30:42,  1.20s/it]
  1%|▏         | 158/10712 [18:14<3:25:14,  1.17s/it]
  1%|▏         | 159/10712 [18:15<3:17:27,  1.12s/it]
  1%|▏         | 160/10712 [18:16<3:15:52,  1.11s/it]
  2%|▏         | 161/10712 [18:17<3:20:32,  1.14s/it]
  2%|▏         | 162/10712 [18:19<3:19:24,  1.13s/it]
  2%|▏         | 163/10712 [18:20<3:30:46,  1.20s/it]
  2%|▏         | 164/10712 [18:21<3:35:47,  1.23s/it]
  2%|▏         | 165/10712 [18:23<4:00:15,  1.37s/it]
  2%|▏         | 166/10712 [18:25<4:10:36,  1.43s/it]
  2%|▏         | 167/10712 [18:26<4:20:21,  1.48s/it]
  2%|▏         | 168/10712 [18:28<4:22:32,  1.49s/it]
  2%|▏         | 169/10712 [18:29<4:11:14,  1.43s/it]
  2%|▏         | 170/10712 [18:30<4:12:59,  1.44s/it]
  2%|▏         | 171/10712 [18:32<3:58:40,  1.36s/it]
  2%|▏         | 172/10712 [18:33<4:05:48,  1.40s/it]
  2%|▏         | 173/10712 [18:34<3:50:38,  1.31s/it]
  2%|▏         | 174/10712 [18:35<3:50:04,  1.31s/it]
  2%|▏         | 175/10712 [18:37<3:46:10,  1.29s/it]
                                                     {'loss': 6.9946, 'grad_norm': 0.40611329674720764, 'learning_rate': 0.00016324626865671642, 'epoch': 0.02}

  2%|▏         | 175/10712 [18:37<3:46:10,  1.29s/it]
  2%|▏         | 176/10712 [18:38<3:48:56,  1.30s/it]
  2%|▏         | 177/10712 [18:39<3:46:18,  1.29s/it]
  2%|▏         | 178/10712 [18:41<3:51:03,  1.32s/it]
  2%|▏         | 179/10712 [18:42<3:46:18,  1.29s/it]
  2%|▏         | 180/10712 [18:43<3:48:11,  1.30s/it]
  2%|▏         | 181/10712 [18:44<3:40:03,  1.25s/it]
  2%|▏         | 182/10712 [18:46<3:37:54,  1.24s/it]
  2%|▏         | 183/10712 [18:47<3:31:46,  1.21s/it]
  2%|▏         | 184/10712 [18:48<3:37:08,  1.24s/it]
  2%|▏         | 185/10712 [18:49<3:36:58,  1.24s/it]
  2%|▏         | 186/10712 [18:50<3:27:48,  1.18s/it]
  2%|▏         | 187/10712 [18:51<3:16:59,  1.12s/it]
  2%|▏         | 188/10712 [18:52<3:15:22,  1.11s/it]
  2%|▏         | 189/10712 [18:53<3:12:44,  1.10s/it]
  2%|▏         | 190/10712 [18:54<3:05:41,  1.06s/it]
  2%|▏         | 191/10712 [18:55<3:02:08,  1.04s/it]
  2%|▏         | 192/10712 [18:56<2:54:43,  1.00it/s]
  2%|▏         | 193/10712 [18:57<2:55:33,  1.00s/it]
  2%|▏         | 194/10712 [18:58<2:53:45,  1.01it/s]
  2%|▏         | 195/10712 [18:59<2:50:53,  1.03it/s]
  2%|▏         | 196/10712 [19:00<2:51:42,  1.02it/s]
  2%|▏         | 197/10712 [19:01<2:50:51,  1.03it/s]
  2%|▏         | 198/10712 [19:02<2:54:25,  1.00it/s]
  2%|▏         | 199/10712 [19:03<2:55:47,  1.00s/it]
  2%|▏         | 200/10712 [19:04<2:49:47,  1.03it/s]
                                                     
{'loss': 6.6888, 'grad_norm': 0.4568512737751007, 'learning_rate': 0.00018656716417910448, 'epoch': 0.02}

  2%|▏         | 200/10712 [19:04<2:49:47,  1.03it/s]
  2%|▏         | 201/10712 [19:05<2:50:30,  1.03it/s]
  2%|▏         | 202/10712 [19:06<2:48:19,  1.04it/s]
  2%|▏         | 203/10712 [19:07<2:47:11,  1.05it/s]
  2%|▏         | 204/10712 [19:08<2:46:47,  1.05it/s]
  2%|▏         | 205/10712 [19:09<2:45:14,  1.06it/s]
  2%|▏         | 206/10712 [19:10<2:44:54,  1.06it/s]
  2%|▏         | 207/10712 [19:11<2:42:14,  1.08it/s]
  2%|▏         | 208/10712 [19:12<2:43:39,  1.07it/s]
  2%|▏         | 209/10712 [19:13<2:40:07,  1.09it/s]
  2%|▏         | 210/10712 [19:13<2:42:12,  1.08it/s]
  2%|▏         | 211/10712 [19:14<2:36:23,  1.12it/s]
  2%|▏         | 212/10712 [19:15<2:34:26,  1.13it/s]
  2%|▏         | 213/10712 [19:16<2:34:27,  1.13it/s]
  2%|▏         | 214/10712 [19:17<2:34:50,  1.13it/s]
  2%|▏         | 215/10712 [19:18<2:31:51,  1.15it/s]
  2%|▏         | 216/10712 [19:19<2:31:49,  1.15it/s]
  2%|▏         | 217/10712 [19:19<2:30:42,  1.16it/s]
  2%|▏         | 218/10712 [19:20<2:31:25,  1.16it/s]
  2%|▏         | 219/10712 [19:21<2:29:52,  1.17it/s]
  2%|▏         | 220/10712 [19:22<2:28:15,  1.18it/s]
  2%|▏         | 221/10712 [19:23<2:30:52,  1.16it/s]
  2%|▏         | 222/10712 [19:24<2:30:48,  1.16it/s]
  2%|▏         | 223/10712 [19:25<2:31:35,  1.15it/s]
  2%|▏         | 224/10712 [19:26<2:31:49,  1.15it/s]
  2%|▏         | 225/10712 [19:26<2:28:40,  1.18it/s]{'loss': 6.4448, 'grad_norm': 0.4776136577129364, 'learning_rate': 0.00020988805970149256, 'epoch': 0.02}

                                                     

  2%|▏         | 225/10712 [19:26<2:28:40,  1.18it/s]
  2%|▏         | 226/10712 [19:27<2:31:35,  1.15it/s]
  2%|▏         | 227/10712 [19:28<2:36:16,  1.12it/s]
  2%|▏         | 228/10712 [19:29<2:31:52,  1.15it/s]
  2%|▏         | 229/10712 [19:30<2:28:11,  1.18it/s]
  2%|▏         | 230/10712 [19:31<2:26:03,  1.20it/s]
  2%|▏         | 231/10712 [19:31<2:26:13,  1.19it/s]
  2%|▏         | 232/10712 [19:32<2:27:31,  1.18it/s]
  2%|▏         | 233/10712 [19:33<2:26:55,  1.19it/s]
  2%|▏         | 234/10712 [19:34<2:28:45,  1.17it/s]
  2%|▏         | 235/10712 [19:35<2:26:53,  1.19it/s]
  2%|▏         | 236/10712 [19:36<2:25:25,  1.20it/s]
  2%|▏         | 237/10712 [19:36<2:26:17,  1.19it/s]
  2%|▏         | 238/10712 [19:37<2:24:26,  1.21it/s]
  2%|▏         | 239/10712 [19:38<2:22:48,  1.22it/s]
  2%|▏         | 240/10712 [19:39<2:21:47,  1.23it/s]
  2%|▏         | 241/10712 [19:40<2:19:39,  1.25it/s]
  2%|▏         | 242/10712 [19:40<2:17:46,  1.27it/s]
  2%|▏         | 243/10712 [19:41<2:15:26,  1.29it/s]
  2%|▏         | 244/10712 [19:42<2:14:21,  1.30it/s]
  2%|▏         | 245/10712 [19:43<2:15:33,  1.29it/s]
  2%|▏         | 246/10712 [19:44<2:19:03,  1.25it/s]
  2%|▏         | 247/10712 [19:44<2:18:03,  1.26it/s]
  2%|▏         | 248/10712 [19:45<2:17:17,  1.27it/s]
  2%|▏         | 249/10712 [19:47<3:35:57,  1.24s/it]
  2%|▏         | 250/10712 [19:51<5:34:15,  1.92s/it]
                                                     
{'loss': 6.2514, 'grad_norm': 0.5097326636314392, 'learning_rate': 0.00023320895522388062, 'epoch': 0.02}

  2%|▏         | 250/10712 [19:51<5:34:15,  1.92s/it]
  2%|▏         | 251/10712 [19:55<7:45:31,  2.67s/it]
  2%|▏         | 252/10712 [19:59<8:57:24,  3.08s/it]
  2%|▏         | 253/10712 [20:03<9:39:19,  3.32s/it]
  2%|▏         | 254/10712 [20:07<9:57:53,  3.43s/it]
  2%|▏         | 255/10712 [20:11<10:54:35,  3.76s/it]
  2%|▏         | 256/10712 [20:15<10:41:30,  3.68s/it]
  2%|▏         | 257/10712 [20:18<10:22:33,  3.57s/it]
  2%|▏         | 258/10712 [20:22<10:38:09,  3.66s/it]
  2%|▏         | 259/10712 [20:26<10:31:16,  3.62s/it]
  2%|▏         | 260/10712 [20:29<10:32:45,  3.63s/it]
  2%|▏         | 261/10712 [20:30<7:57:44,  2.74s/it] 
  2%|▏         | 262/10712 [20:31<6:08:44,  2.12s/it]
  2%|▏         | 263/10712 [20:31<4:53:09,  1.68s/it]
  2%|▏         | 264/10712 [20:32<3:59:39,  1.38s/it]
  2%|▏         | 265/10712 [20:33<3:24:32,  1.17s/it]
  2%|▏         | 266/10712 [20:33<2:56:57,  1.02s/it]
  2%|▏         | 267/10712 [20:34<2:38:50,  1.10it/s]
  3%|▎         | 268/10712 [20:35<2:25:02,  1.20it/s]
  3%|▎         | 269/10712 [20:35<2:12:56,  1.31it/s]
  3%|▎         | 270/10712 [20:39<4:45:01,  1.64s/it]
  3%|▎         | 271/10712 [20:43<6:34:02,  2.26s/it]
  3%|▎         | 272/10712 [20:46<7:41:14,  2.65s/it]
  3%|▎         | 273/10712 [20:50<8:20:45,  2.88s/it]
  3%|▎         | 274/10712 [20:53<8:38:53,  2.98s/it]
  3%|▎         | 275/10712 [20:56<8:56:14,  3.08s/it]
                                                     
{'loss': 6.0902, 'grad_norm': 0.8254302740097046, 'learning_rate': 0.00025652985074626865, 'epoch': 0.03}

  3%|▎         | 275/10712 [20:56<8:56:14,  3.08s/it]
  3%|▎         | 276/10712 [21:00<9:11:56,  3.17s/it]
  3%|▎         | 277/10712 [21:02<8:44:40,  3.02s/it]
  3%|▎         | 278/10712 [21:05<8:45:49,  3.02s/it]
  3%|▎         | 279/10712 [21:09<9:10:35,  3.17s/it]
  3%|▎         | 280/10712 [21:12<9:13:46,  3.19s/it]
  3%|▎         | 281/10712 [21:16<9:42:04,  3.35s/it]
  3%|▎         | 282/10712 [21:19<9:32:09,  3.29s/it]
  3%|▎         | 283/10712 [21:20<7:28:50,  2.58s/it]
  3%|▎         | 284/10712 [21:20<5:45:49,  1.99s/it]
  3%|▎         | 285/10712 [21:21<4:35:53,  1.59s/it]
  3%|▎         | 286/10712 [21:22<3:44:59,  1.29s/it]
  3%|▎         | 287/10712 [21:22<3:11:22,  1.10s/it]
  3%|▎         | 288/10712 [21:23<2:47:43,  1.04it/s]
  3%|▎         | 289/10712 [21:24<2:30:38,  1.15it/s]
  3%|▎         | 290/10712 [21:24<2:18:07,  1.26it/s]
  3%|▎         | 291/10712 [21:25<2:08:44,  1.35it/s]
  3%|▎         | 292/10712 [21:26<2:55:46,  1.01s/it]
  3%|▎         | 293/10712 [21:29<4:27:06,  1.54s/it]
  3%|▎         | 294/10712 [21:32<5:36:05,  1.94s/it]
  3%|▎         | 295/10712 [21:35<6:36:23,  2.28s/it]
  3%|▎         | 296/10712 [21:38<7:22:32,  2.55s/it]
  3%|▎         | 297/10712 [21:41<7:37:10,  2.63s/it]
  3%|▎         | 298/10712 [21:44<7:43:13,  2.67s/it]
  3%|▎         | 299/10712 [21:47<8:04:19,  2.79s/it]
  3%|▎         | 300/10712 [21:50<8:24:19,  2.91s/it]{'loss': 5.9398, 'grad_norm': 0.6166653037071228, 'learning_rate': 0.0002798507462686567, 'epoch': 0.03}
                                                     

  3%|▎         | 300/10712 [21:50<8:24:19,  2.91s/it]
  3%|▎         | 301/10712 [21:53<8:40:45,  3.00s/it]
  3%|▎         | 302/10712 [21:57<8:47:20,  3.04s/it]
  3%|▎         | 303/10712 [21:59<8:35:55,  2.97s/it]
  3%|▎         | 304/10712 [22:02<8:25:46,  2.92s/it]
  3%|▎         | 305/10712 [22:05<7:58:57,  2.76s/it]
  3%|▎         | 306/10712 [22:07<7:59:20,  2.76s/it]
  3%|▎         | 307/10712 [22:10<7:31:10,  2.60s/it]
  3%|▎         | 308/10712 [22:10<5:45:57,  2.00s/it]
  3%|▎         | 309/10712 [22:11<4:33:02,  1.57s/it]
  3%|▎         | 310/10712 [22:11<3:41:42,  1.28s/it]
  3%|▎         | 311/10712 [22:13<4:21:12,  1.51s/it]
  3%|▎         | 312/10712 [22:16<5:18:18,  1.84s/it]
  3%|▎         | 313/10712 [22:19<5:58:22,  2.07s/it]
  3%|▎         | 314/10712 [22:21<6:30:07,  2.25s/it]
  3%|▎         | 315/10712 [22:23<5:57:06,  2.06s/it]
  3%|▎         | 316/10712 [22:23<4:40:14,  1.62s/it]
  3%|▎         | 317/10712 [22:24<3:46:56,  1.31s/it]
  3%|▎         | 318/10712 [22:25<3:13:15,  1.12s/it]
  3%|▎         | 319/10712 [22:25<2:50:13,  1.02it/s]
  3%|▎         | 320/10712 [22:26<2:32:45,  1.13it/s]
  3%|▎         | 321/10712 [22:27<2:21:15,  1.23it/s]
  3%|▎         | 322/10712 [22:28<2:24:26,  1.20it/s]
  3%|▎         | 323/10712 [22:28<2:14:15,  1.29it/s]
  3%|▎         | 324/10712 [22:29<2:09:07,  1.34it/s]
  3%|▎         | 325/10712 [22:30<2:03:30,  1.40it/s]{'loss': 5.8186, 'grad_norm': 0.592656672000885, 'learning_rate': 0.00030317164179104476, 'epoch': 0.03}
                                                     

  3%|▎         | 325/10712 [22:30<2:03:30,  1.40it/s]
  3%|▎         | 326/10712 [22:30<2:00:41,  1.43it/s]
  3%|▎         | 327/10712 [22:31<1:56:13,  1.49it/s]
  3%|▎         | 328/10712 [22:31<1:55:06,  1.50it/s]
  3%|▎         | 329/10712 [22:32<1:51:57,  1.55it/s]
  3%|▎         | 330/10712 [22:33<1:50:43,  1.56it/s]
  3%|▎         | 331/10712 [22:33<1:49:56,  1.57it/s]
  3%|▎         | 332/10712 [22:34<1:48:16,  1.60it/s]
  3%|▎         | 333/10712 [22:35<1:50:07,  1.57it/s]
  3%|▎         | 334/10712 [22:35<1:51:06,  1.56it/s]
  3%|▎         | 335/10712 [22:36<1:52:13,  1.54it/s]
  3%|▎         | 336/10712 [22:37<1:52:10,  1.54it/s]
  3%|▎         | 337/10712 [22:37<1:51:29,  1.55it/s]
  3%|▎         | 338/10712 [22:38<1:51:46,  1.55it/s]
  3%|▎         | 339/10712 [22:38<1:50:15,  1.57it/s]
  3%|▎         | 340/10712 [22:39<1:48:41,  1.59it/s]
  3%|▎         | 341/10712 [22:40<1:48:47,  1.59it/s]
  3%|▎         | 342/10712 [22:40<1:49:28,  1.58it/s]
  3%|▎         | 343/10712 [22:41<1:48:24,  1.59it/s]
  3%|▎         | 344/10712 [22:42<1:47:29,  1.61it/s]
  3%|▎         | 345/10712 [22:42<1:47:06,  1.61it/s]
  3%|▎         | 346/10712 [22:43<1:45:58,  1.63it/s]
  3%|▎         | 347/10712 [22:43<1:45:41,  1.63it/s]
  3%|▎         | 348/10712 [22:44<1:44:42,  1.65it/s]
  3%|▎         | 349/10712 [22:45<1:44:54,  1.65it/s]
  3%|▎         | 350/10712 [22:45<1:46:08,  1.63it/s]
                                                     {'loss': 5.712, 'grad_norm': 0.6223078966140747, 'learning_rate': 0.00032649253731343285, 'epoch': 0.03}


  3%|▎         | 350/10712 [22:45<1:46:08,  1.63it/s]
  3%|▎         | 351/10712 [22:46<1:45:04,  1.64it/s]
  3%|▎         | 352/10712 [22:46<1:44:07,  1.66it/s]
  3%|▎         | 353/10712 [22:47<1:44:54,  1.65it/s]
  3%|▎         | 354/10712 [22:48<1:48:49,  1.59it/s]
  3%|▎         | 355/10712 [22:48<1:47:03,  1.61it/s]
  3%|▎         | 356/10712 [22:49<1:47:05,  1.61it/s]
  3%|▎         | 357/10712 [22:49<1:46:04,  1.63it/s]
  3%|▎         | 358/10712 [22:50<2:01:59,  1.41it/s]
  3%|▎         | 359/10712 [22:51<1:59:10,  1.45it/s]
  3%|▎         | 360/10712 [22:52<1:55:20,  1.50it/s]
  3%|▎         | 361/10712 [22:52<1:54:14,  1.51it/s]
  3%|▎         | 362/10712 [22:53<1:52:20,  1.54it/s]
  3%|▎         | 363/10712 [22:54<1:49:55,  1.57it/s]
  3%|▎         | 364/10712 [22:54<1:52:10,  1.54it/s]
  3%|▎         | 365/10712 [22:55<1:51:48,  1.54it/s]
  3%|▎         | 366/10712 [22:56<1:52:40,  1.53it/s]
  3%|▎         | 367/10712 [22:56<1:52:23,  1.53it/s]
  3%|▎         | 368/10712 [22:57<1:51:04,  1.55it/s]
  3%|▎         | 369/10712 [22:57<1:50:25,  1.56it/s]
  3%|▎         | 370/10712 [22:58<2:00:37,  1.43it/s]
  3%|▎         | 371/10712 [22:59<2:00:28,  1.43it/s]
  3%|▎         | 372/10712 [23:00<2:00:04,  1.44it/s]
  3%|▎         | 373/10712 [23:00<1:58:10,  1.46it/s]
  3%|▎         | 374/10712 [23:01<1:55:36,  1.49it/s]
  4%|▎         | 375/10712 [23:02<1:53:29,  1.52it/s]
                                                     {'loss': 5.6125, 'grad_norm': 0.6318686604499817, 'learning_rate': 0.0003498134328358209, 'epoch': 0.04}


  4%|▎         | 375/10712 [23:02<1:53:29,  1.52it/s]
  4%|▎         | 376/10712 [23:02<2:05:09,  1.38it/s]
  4%|▎         | 377/10712 [23:03<1:58:37,  1.45it/s]
  4%|▎         | 378/10712 [23:04<1:56:53,  1.47it/s]
  4%|▎         | 379/10712 [23:04<1:55:38,  1.49it/s]
  4%|▎         | 380/10712 [23:05<1:52:38,  1.53it/s]
  4%|▎         | 381/10712 [23:06<1:49:18,  1.58it/s]
  4%|▎         | 382/10712 [23:06<1:47:14,  1.61it/s]
  4%|▎         | 383/10712 [23:07<1:46:50,  1.61it/s]
  4%|▎         | 384/10712 [23:07<1:46:49,  1.61it/s]
  4%|▎         | 385/10712 [23:08<1:48:21,  1.59it/s]
  4%|▎         | 386/10712 [23:09<1:57:26,  1.47it/s]
  4%|▎         | 387/10712 [23:09<1:52:29,  1.53it/s]
  4%|▎         | 388/10712 [23:10<2:00:49,  1.42it/s]
  4%|▎         | 389/10712 [23:11<1:55:15,  1.49it/s]
  4%|▎         | 390/10712 [23:12<1:52:32,  1.53it/s]
  4%|▎         | 391/10712 [23:12<2:04:40,  1.38it/s]
  4%|▎         | 392/10712 [23:13<1:59:21,  1.44it/s]
  4%|▎         | 393/10712 [23:14<1:54:57,  1.50it/s]
  4%|▎         | 394/10712 [23:14<1:50:34,  1.56it/s]
  4%|▎         | 395/10712 [23:15<1:48:16,  1.59it/s]
  4%|▎         | 396/10712 [23:15<1:48:38,  1.58it/s]
  4%|▎         | 397/10712 [23:16<1:48:19,  1.59it/s]
  4%|▎         | 398/10712 [23:17<2:00:06,  1.43it/s]
  4%|▎         | 399/10712 [23:18<1:57:12,  1.47it/s]
  4%|▎         | 400/10712 [23:18<1:53:51,  1.51it/s]
                                                     
{'loss': 5.5329, 'grad_norm': 0.6341351270675659, 'learning_rate': 0.00037313432835820896, 'epoch': 0.04}

  4%|▎         | 400/10712 [23:18<1:53:51,  1.51it/s]
  4%|▎         | 401/10712 [23:19<1:51:01,  1.55it/s]
  4%|▍         | 402/10712 [23:19<1:50:51,  1.55it/s]
  4%|▍         | 403/10712 [23:20<1:51:14,  1.54it/s]
  4%|▍         | 404/10712 [23:21<1:50:21,  1.56it/s]
  4%|▍         | 405/10712 [23:21<1:49:19,  1.57it/s]
  4%|▍         | 406/10712 [23:22<1:58:58,  1.44it/s]
  4%|▍         | 407/10712 [23:23<1:57:14,  1.46it/s]
  4%|▍         | 408/10712 [23:23<1:53:25,  1.51it/s]
  4%|▍         | 409/10712 [23:24<2:03:53,  1.39it/s]
  4%|▍         | 410/10712 [23:25<2:00:20,  1.43it/s]
  4%|▍         | 411/10712 [23:26<1:55:45,  1.48it/s]
  4%|▍         | 412/10712 [23:26<1:54:05,  1.50it/s]
  4%|▍         | 413/10712 [23:27<1:48:53,  1.58it/s]
  4%|▍         | 414/10712 [23:27<1:49:53,  1.56it/s]
  4%|▍         | 415/10712 [23:28<1:48:12,  1.59it/s]
  4%|▍         | 416/10712 [23:29<1:47:09,  1.60it/s]
  4%|▍         | 417/10712 [23:29<1:42:42,  1.67it/s]
  4%|▍         | 418/10712 [23:30<1:40:40,  1.70it/s]
  4%|▍         | 419/10712 [23:30<1:41:44,  1.69it/s]
  4%|▍         | 420/10712 [23:31<1:40:34,  1.71it/s]
  4%|▍         | 421/10712 [23:31<1:39:44,  1.72it/s]
  4%|▍         | 422/10712 [23:32<1:38:39,  1.74it/s]
  4%|▍         | 423/10712 [23:33<1:38:13,  1.75it/s]
  4%|▍         | 424/10712 [23:33<1:39:37,  1.72it/s]
  4%|▍         | 425/10712 [23:34<1:38:48,  1.74it/s]{'loss': 5.4562, 'grad_norm': 0.722440779209137, 'learning_rate': 0.00039645522388059704, 'epoch': 0.04}
                                                     

  4%|▍         | 425/10712 [23:34<1:38:48,  1.74it/s]
  4%|▍         | 426/10712 [23:34<1:37:57,  1.75it/s]
  4%|▍         | 427/10712 [23:35<1:38:43,  1.74it/s]
  4%|▍         | 428/10712 [23:36<1:39:31,  1.72it/s]
  4%|▍         | 429/10712 [23:36<1:40:19,  1.71it/s]
  4%|▍         | 430/10712 [23:37<1:38:08,  1.75it/s]
  4%|▍         | 431/10712 [23:37<1:38:53,  1.73it/s]
  4%|▍         | 432/10712 [23:38<1:38:42,  1.74it/s]
  4%|▍         | 433/10712 [23:38<1:38:16,  1.74it/s]
  4%|▍         | 434/10712 [23:39<1:37:22,  1.76it/s]
  4%|▍         | 435/10712 [23:40<1:36:43,  1.77it/s]
  4%|▍         | 436/10712 [23:40<1:36:42,  1.77it/s]
  4%|▍         | 437/10712 [23:41<1:37:01,  1.77it/s]
  4%|▍         | 438/10712 [23:41<1:36:25,  1.78it/s]
  4%|▍         | 439/10712 [23:42<1:36:53,  1.77it/s]
  4%|▍         | 440/10712 [23:42<1:35:47,  1.79it/s]
  4%|▍         | 441/10712 [23:43<1:35:52,  1.79it/s]
  4%|▍         | 442/10712 [23:43<1:35:54,  1.78it/s]
  4%|▍         | 443/10712 [23:44<1:35:49,  1.79it/s]
  4%|▍         | 444/10712 [23:45<1:37:11,  1.76it/s]
  4%|▍         | 445/10712 [23:45<1:37:13,  1.76it/s]
  4%|▍         | 446/10712 [23:46<1:38:36,  1.74it/s]
  4%|▍         | 447/10712 [23:46<1:37:04,  1.76it/s]
  4%|▍         | 448/10712 [23:47<1:35:51,  1.78it/s]
  4%|▍         | 449/10712 [23:47<1:35:45,  1.79it/s]
  4%|▍         | 450/10712 [23:48<1:36:15,  1.78it/s]{'loss': 5.3724, 'grad_norm': 0.5020679235458374, 'learning_rate': 0.0004197761194029851, 'epoch': 0.04}
                                                     

  4%|▍         | 450/10712 [23:48<1:36:15,  1.78it/s]
  4%|▍         | 451/10712 [23:49<1:34:47,  1.80it/s]
  4%|▍         | 452/10712 [23:49<1:35:12,  1.80it/s]
  4%|▍         | 453/10712 [23:50<1:35:19,  1.79it/s]
  4%|▍         | 454/10712 [23:50<1:35:33,  1.79it/s]
  4%|▍         | 455/10712 [23:51<1:34:53,  1.80it/s]
  4%|▍         | 456/10712 [23:51<1:38:20,  1.74it/s]
  4%|▍         | 457/10712 [23:52<1:39:50,  1.71it/s]
  4%|▍         | 458/10712 [23:53<1:41:48,  1.68it/s]
  4%|▍         | 459/10712 [23:53<1:43:40,  1.65it/s]
  4%|▍         | 460/10712 [23:54<1:42:45,  1.66it/s]
  4%|▍         | 461/10712 [23:54<1:40:04,  1.71it/s]
  4%|▍         | 462/10712 [23:55<1:39:08,  1.72it/s]
  4%|▍         | 463/10712 [23:55<1:36:53,  1.76it/s]
  4%|▍         | 464/10712 [23:56<1:36:04,  1.78it/s]
  4%|▍         | 465/10712 [23:57<1:35:21,  1.79it/s]
  4%|▍         | 466/10712 [23:57<1:35:20,  1.79it/s]
  4%|▍         | 467/10712 [23:58<1:34:40,  1.80it/s]
  4%|▍         | 468/10712 [23:58<1:33:56,  1.82it/s]
  4%|▍         | 469/10712 [23:59<1:34:22,  1.81it/s]
  4%|▍         | 470/10712 [23:59<1:35:05,  1.80it/s]
  4%|▍         | 471/10712 [24:00<1:35:12,  1.79it/s]
  4%|▍         | 472/10712 [24:00<1:34:56,  1.80it/s]
  4%|▍         | 473/10712 [24:01<1:34:05,  1.81it/s]
  4%|▍         | 474/10712 [24:02<1:35:54,  1.78it/s]
  4%|▍         | 475/10712 [24:02<1:35:00,  1.80it/s]
                                                     {'loss': 5.3055, 'grad_norm': 0.5739572048187256, 'learning_rate': 0.00044309701492537316, 'epoch': 0.04}


  4%|▍         | 475/10712 [24:02<1:35:00,  1.80it/s]
  4%|▍         | 476/10712 [24:03<1:34:29,  1.81it/s]
  4%|▍         | 477/10712 [24:03<1:34:23,  1.81it/s]
  4%|▍         | 478/10712 [24:04<1:33:05,  1.83it/s]
  4%|▍         | 479/10712 [24:04<1:34:45,  1.80it/s]
  4%|▍         | 480/10712 [24:05<1:34:52,  1.80it/s]
  4%|▍         | 481/10712 [24:05<1:35:27,  1.79it/s]
  4%|▍         | 482/10712 [24:06<1:35:00,  1.79it/s]
  5%|▍         | 483/10712 [24:07<1:34:03,  1.81it/s]
  5%|▍         | 484/10712 [24:07<1:33:42,  1.82it/s]
  5%|▍         | 485/10712 [24:08<1:33:41,  1.82it/s]
  5%|▍         | 486/10712 [24:08<1:33:03,  1.83it/s]
  5%|▍         | 487/10712 [24:09<1:34:01,  1.81it/s]
  5%|▍         | 488/10712 [24:09<1:33:47,  1.82it/s]
  5%|▍         | 489/10712 [24:10<1:34:23,  1.81it/s]
  5%|▍         | 490/10712 [24:10<1:34:58,  1.79it/s]
  5%|▍         | 491/10712 [24:11<1:35:13,  1.79it/s]
  5%|▍         | 492/10712 [24:11<1:33:48,  1.82it/s]
  5%|▍         | 493/10712 [24:12<1:33:40,  1.82it/s]
  5%|▍         | 494/10712 [24:13<1:33:59,  1.81it/s]
  5%|▍         | 495/10712 [24:13<1:33:36,  1.82it/s]
  5%|▍         | 496/10712 [24:14<1:33:20,  1.82it/s]
  5%|▍         | 497/10712 [24:14<1:31:53,  1.85it/s]
  5%|▍         | 498/10712 [24:15<1:32:06,  1.85it/s]
  5%|▍         | 499/10712 [24:15<1:31:45,  1.86it/s]
  5%|▍         | 500/10712 [24:16<1:32:44,  1.84it/s]{'loss': 5.2447, 'grad_norm': 0.6082661151885986, 'learning_rate': 0.00046641791044776124, 'epoch': 0.05}
                                                     

  5%|▍         | 500/10712 [24:16<1:32:44,  1.84it/s]
  5%|▍         | 501/10712 [24:16<1:32:11,  1.85it/s]
  5%|▍         | 502/10712 [24:17<1:31:58,  1.85it/s]
  5%|▍         | 503/10712 [24:17<1:32:46,  1.83it/s]
  5%|▍         | 504/10712 [24:18<1:32:17,  1.84it/s]
  5%|▍         | 505/10712 [24:19<1:31:54,  1.85it/s]
  5%|▍         | 506/10712 [24:19<1:32:17,  1.84it/s]
  5%|▍         | 507/10712 [24:20<1:32:34,  1.84it/s]
  5%|▍         | 508/10712 [24:20<1:32:26,  1.84it/s]
  5%|▍         | 509/10712 [24:21<1:45:28,  1.61it/s]
  5%|▍         | 510/10712 [24:22<1:41:48,  1.67it/s]
  5%|▍         | 511/10712 [24:22<1:42:31,  1.66it/s]
  5%|▍         | 512/10712 [24:23<1:42:41,  1.66it/s]
  5%|▍         | 513/10712 [24:23<1:40:31,  1.69it/s]
  5%|▍         | 514/10712 [24:24<1:39:10,  1.71it/s]
  5%|▍         | 515/10712 [24:24<1:36:44,  1.76it/s]
  5%|▍         | 516/10712 [24:25<1:46:15,  1.60it/s]
  5%|▍         | 517/10712 [24:26<1:55:41,  1.47it/s]
  5%|▍         | 518/10712 [24:27<1:48:33,  1.57it/s]
  5%|▍         | 519/10712 [24:27<1:44:19,  1.63it/s]
  5%|▍         | 520/10712 [24:28<1:42:45,  1.65it/s]
  5%|▍         | 521/10712 [24:28<1:41:51,  1.67it/s]
  5%|▍         | 522/10712 [24:29<1:38:32,  1.72it/s]
  5%|▍         | 523/10712 [24:29<1:37:10,  1.75it/s]
  5%|▍         | 524/10712 [24:30<1:35:46,  1.77it/s]
  5%|▍         | 525/10712 [24:30<1:35:59,  1.77it/s]
                                                     {'loss': 5.184, 'grad_norm': 0.5593669414520264, 'learning_rate': 0.0004897388059701492, 'epoch': 0.05}

  5%|▍         | 525/10712 [24:30<1:35:59,  1.77it/s]
  5%|▍         | 526/10712 [24:31<1:37:37,  1.74it/s]
  5%|▍         | 527/10712 [24:32<1:36:06,  1.77it/s]
  5%|▍         | 528/10712 [24:32<1:35:56,  1.77it/s]
  5%|▍         | 529/10712 [24:33<1:34:26,  1.80it/s]
  5%|▍         | 530/10712 [24:33<1:46:30,  1.59it/s]
  5%|▍         | 531/10712 [24:34<1:42:21,  1.66it/s]
  5%|▍         | 532/10712 [24:35<1:50:32,  1.53it/s]
  5%|▍         | 533/10712 [24:36<1:56:05,  1.46it/s]
  5%|▍         | 534/10712 [24:36<1:52:30,  1.51it/s]
  5%|▍         | 535/10712 [24:37<1:47:58,  1.57it/s]
  5%|▌         | 536/10712 [24:37<1:53:08,  1.50it/s]
  5%|▌         | 537/10712 [24:38<1:46:55,  1.59it/s]
  5%|▌         | 538/10712 [24:39<1:42:33,  1.65it/s]
  5%|▌         | 539/10712 [24:39<1:40:32,  1.69it/s]
  5%|▌         | 540/10712 [24:40<1:39:13,  1.71it/s]
  5%|▌         | 541/10712 [24:40<1:47:20,  1.58it/s]
  5%|▌         | 542/10712 [24:41<1:44:08,  1.63it/s]
  5%|▌         | 543/10712 [24:42<1:39:56,  1.70it/s]
  5%|▌         | 544/10712 [24:42<1:39:00,  1.71it/s]
  5%|▌         | 545/10712 [24:43<1:37:30,  1.74it/s]
  5%|▌         | 546/10712 [24:43<1:38:43,  1.72it/s]
  5%|▌         | 547/10712 [24:44<1:36:58,  1.75it/s]
  5%|▌         | 548/10712 [24:44<1:37:35,  1.74it/s]
  5%|▌         | 549/10712 [24:45<1:36:54,  1.75it/s]
  5%|▌         | 550/10712 [24:46<1:34:58,  1.78it/s]
                                                     
{'loss': 5.1268, 'grad_norm': 0.654487133026123, 'learning_rate': 0.0005130597014925373, 'epoch': 0.05}

  5%|▌         | 550/10712 [24:46<1:34:58,  1.78it/s]
  5%|▌         | 551/10712 [24:47<2:00:07,  1.41it/s]
  5%|▌         | 552/10712 [24:48<2:20:30,  1.21it/s]
  5%|▌         | 553/10712 [24:48<2:08:09,  1.32it/s]
  5%|▌         | 554/10712 [24:49<1:57:54,  1.44it/s]
  5%|▌         | 555/10712 [24:49<1:49:02,  1.55it/s]
  5%|▌         | 556/10712 [24:50<1:42:50,  1.65it/s]
  5%|▌         | 557/10712 [24:50<1:38:42,  1.71it/s]
  5%|▌         | 558/10712 [24:51<1:35:50,  1.77it/s]
  5%|▌         | 559/10712 [24:51<1:33:55,  1.80it/s]
  5%|▌         | 560/10712 [24:52<1:32:02,  1.84it/s]
  5%|▌         | 561/10712 [24:52<1:31:17,  1.85it/s]
  5%|▌         | 562/10712 [24:53<1:30:21,  1.87it/s]
  5%|▌         | 563/10712 [24:54<1:29:23,  1.89it/s]
  5%|▌         | 564/10712 [24:54<1:29:43,  1.88it/s]
  5%|▌         | 565/10712 [24:55<1:29:29,  1.89it/s]
  5%|▌         | 566/10712 [24:55<1:29:35,  1.89it/s]
  5%|▌         | 567/10712 [24:56<1:29:52,  1.88it/s]
  5%|▌         | 568/10712 [24:56<1:30:12,  1.87it/s]
  5%|▌         | 569/10712 [24:57<1:29:53,  1.88it/s]
  5%|▌         | 570/10712 [24:57<1:29:22,  1.89it/s]
  5%|▌         | 571/10712 [24:58<1:30:35,  1.87it/s]
  5%|▌         | 572/10712 [24:58<1:30:11,  1.87it/s]
  5%|▌         | 573/10712 [24:59<1:30:02,  1.88it/s]
  5%|▌         | 574/10712 [24:59<1:31:02,  1.86it/s]
  5%|▌         | 575/10712 [25:00<1:30:26,  1.87it/s]{'loss': 5.0736, 'grad_norm': 0.7381680607795715, 'learning_rate': 0.0005363805970149254, 'epoch': 0.05}

                                                     

  5%|▌         | 575/10712 [25:00<1:30:26,  1.87it/s]
  5%|▌         | 576/10712 [25:00<1:29:49,  1.88it/s]
  5%|▌         | 577/10712 [25:01<1:29:53,  1.88it/s]
  5%|▌         | 578/10712 [25:02<1:29:54,  1.88it/s]
  5%|▌         | 579/10712 [25:02<1:29:18,  1.89it/s]
  5%|▌         | 580/10712 [25:03<1:28:55,  1.90it/s]
  5%|▌         | 581/10712 [25:03<1:29:47,  1.88it/s]
  5%|▌         | 582/10712 [25:04<1:29:01,  1.90it/s]
  5%|▌         | 583/10712 [25:04<1:28:36,  1.91it/s]
  5%|▌         | 584/10712 [25:05<1:29:34,  1.88it/s]
  5%|▌         | 585/10712 [25:05<1:29:25,  1.89it/s]
  5%|▌         | 586/10712 [25:06<1:29:07,  1.89it/s]
  5%|▌         | 587/10712 [25:06<1:28:29,  1.91it/s]
  5%|▌         | 588/10712 [25:07<1:28:26,  1.91it/s]
  5%|▌         | 589/10712 [25:07<1:28:23,  1.91it/s]
  6%|▌         | 590/10712 [25:08<1:28:40,  1.90it/s]
  6%|▌         | 591/10712 [25:08<1:27:49,  1.92it/s]
  6%|▌         | 592/10712 [25:09<1:28:13,  1.91it/s]
  6%|▌         | 593/10712 [25:09<1:29:02,  1.89it/s]
  6%|▌         | 594/10712 [25:10<1:28:46,  1.90it/s]
  6%|▌         | 595/10712 [25:10<1:29:02,  1.89it/s]
  6%|▌         | 596/10712 [25:11<1:28:29,  1.91it/s]
  6%|▌         | 597/10712 [25:12<1:27:58,  1.92it/s]
  6%|▌         | 598/10712 [25:12<1:26:55,  1.94it/s]
  6%|▌         | 599/10712 [25:13<1:27:43,  1.92it/s]
  6%|▌         | 600/10712 [25:13<1:28:00,  1.92it/s]{'loss': 5.0299, 'grad_norm': 0.6007251739501953, 'learning_rate': 0.0005597014925373134, 'epoch': 0.06}

                                                     

  6%|▌         | 600/10712 [25:13<1:28:00,  1.92it/s]
  6%|▌         | 601/10712 [25:14<1:28:08,  1.91it/s]
  6%|▌         | 602/10712 [25:14<1:28:21,  1.91it/s]
  6%|▌         | 603/10712 [25:15<1:28:38,  1.90it/s]
  6%|▌         | 604/10712 [25:15<1:27:51,  1.92it/s]
  6%|▌         | 605/10712 [25:16<1:28:08,  1.91it/s]
  6%|▌         | 606/10712 [25:16<1:28:52,  1.90it/s]
  6%|▌         | 607/10712 [25:17<1:28:01,  1.91it/s]
  6%|▌         | 608/10712 [25:17<1:27:58,  1.91it/s]
  6%|▌         | 609/10712 [25:18<1:28:47,  1.90it/s]
  6%|▌         | 610/10712 [25:18<1:27:58,  1.91it/s]
  6%|▌         | 611/10712 [25:19<1:28:15,  1.91it/s]
  6%|▌         | 612/10712 [25:19<1:28:10,  1.91it/s]
  6%|▌         | 613/10712 [25:20<1:27:32,  1.92it/s]
  6%|▌         | 614/10712 [25:20<1:27:38,  1.92it/s]
  6%|▌         | 615/10712 [25:21<1:27:18,  1.93it/s]
  6%|▌         | 616/10712 [25:21<1:27:46,  1.92it/s]
  6%|▌         | 617/10712 [25:22<1:27:36,  1.92it/s]
  6%|▌         | 618/10712 [25:22<1:27:23,  1.92it/s]
  6%|▌         | 619/10712 [25:23<1:27:29,  1.92it/s]
  6%|▌         | 620/10712 [25:24<1:27:42,  1.92it/s]
  6%|▌         | 621/10712 [25:24<1:27:47,  1.92it/s]
  6%|▌         | 622/10712 [25:25<1:27:39,  1.92it/s]
  6%|▌         | 623/10712 [25:25<1:27:53,  1.91it/s]
  6%|▌         | 624/10712 [25:26<1:27:37,  1.92it/s]
  6%|▌         | 625/10712 [25:26<1:28:06,  1.91it/s]{'loss': 4.9819, 'grad_norm': 0.4367240071296692, 'learning_rate': 0.0005830223880597015, 'epoch': 0.06}

                                                     

  6%|▌         | 625/10712 [25:26<1:28:06,  1.91it/s]
  6%|▌         | 626/10712 [25:27<1:28:10,  1.91it/s]
  6%|▌         | 627/10712 [25:27<1:27:08,  1.93it/s]
  6%|▌         | 628/10712 [25:28<1:27:44,  1.92it/s]
  6%|▌         | 629/10712 [25:28<1:26:35,  1.94it/s]
  6%|▌         | 630/10712 [25:29<1:26:57,  1.93it/s]
  6%|▌         | 631/10712 [25:29<1:26:47,  1.94it/s]
  6%|▌         | 632/10712 [25:30<1:26:34,  1.94it/s]
  6%|▌         | 633/10712 [25:30<1:26:40,  1.94it/s]
  6%|▌         | 634/10712 [25:31<1:27:45,  1.91it/s]
  6%|▌         | 635/10712 [25:31<1:27:13,  1.93it/s]
  6%|▌         | 636/10712 [25:32<1:27:26,  1.92it/s]
  6%|▌         | 637/10712 [25:32<1:27:21,  1.92it/s]
  6%|▌         | 638/10712 [25:33<1:26:45,  1.94it/s]
  6%|▌         | 639/10712 [25:33<1:26:41,  1.94it/s]
  6%|▌         | 640/10712 [25:34<1:27:25,  1.92it/s]
  6%|▌         | 641/10712 [25:34<1:27:35,  1.92it/s]
  6%|▌         | 642/10712 [25:35<1:26:45,  1.93it/s]
  6%|▌         | 643/10712 [25:35<1:26:42,  1.94it/s]
  6%|▌         | 644/10712 [25:36<1:26:33,  1.94it/s]
  6%|▌         | 645/10712 [25:36<1:26:57,  1.93it/s]
  6%|▌         | 646/10712 [25:37<1:27:32,  1.92it/s]
  6%|▌         | 647/10712 [25:38<1:37:39,  1.72it/s]
  6%|▌         | 648/10712 [25:38<1:35:04,  1.76it/s]
  6%|▌         | 649/10712 [25:39<1:33:05,  1.80it/s]
  6%|▌         | 650/10712 [25:39<1:31:31,  1.83it/s]
                                                     {'loss': 4.9424, 'grad_norm': 0.4756833612918854, 'learning_rate': 0.0006063432835820895, 'epoch': 0.06}

  6%|▌         | 650/10712 [25:39<1:31:31,  1.83it/s]
  6%|▌         | 651/10712 [25:40<1:30:34,  1.85it/s]
  6%|▌         | 652/10712 [25:40<1:29:04,  1.88it/s]
  6%|▌         | 653/10712 [25:41<1:28:26,  1.90it/s]
  6%|▌         | 654/10712 [25:41<1:27:53,  1.91it/s]
  6%|▌         | 655/10712 [25:42<1:26:54,  1.93it/s]
  6%|▌         | 656/10712 [25:42<1:26:12,  1.94it/s]
  6%|▌         | 657/10712 [25:43<1:26:17,  1.94it/s]
  6%|▌         | 658/10712 [25:43<1:25:51,  1.95it/s]
  6%|▌         | 659/10712 [25:44<1:26:18,  1.94it/s]
  6%|▌         | 660/10712 [25:44<1:26:06,  1.95it/s]
  6%|▌         | 661/10712 [25:45<1:26:14,  1.94it/s]
  6%|▌         | 662/10712 [25:45<1:25:25,  1.96it/s]
  6%|▌         | 663/10712 [25:46<1:25:48,  1.95it/s]
  6%|▌         | 664/10712 [25:47<1:26:03,  1.95it/s]
  6%|▌         | 665/10712 [25:47<1:26:05,  1.95it/s]
  6%|▌         | 666/10712 [25:48<1:25:55,  1.95it/s]
  6%|▌         | 667/10712 [25:48<1:25:52,  1.95it/s]
  6%|▌         | 668/10712 [25:49<1:25:45,  1.95it/s]
  6%|▌         | 669/10712 [25:49<1:26:27,  1.94it/s]
  6%|▋         | 670/10712 [25:50<1:25:47,  1.95it/s]
  6%|▋         | 671/10712 [25:50<1:26:27,  1.94it/s]
  6%|▋         | 672/10712 [25:51<1:26:09,  1.94it/s]
  6%|▋         | 673/10712 [25:51<1:26:18,  1.94it/s]
  6%|▋         | 674/10712 [25:52<1:25:50,  1.95it/s]
  6%|▋         | 675/10712 [25:52<1:26:17,  1.94it/s]{'loss': 4.9053, 'grad_norm': 0.527163565158844, 'learning_rate': 0.0006296641791044776, 'epoch': 0.06}
                                                     

  6%|▋         | 675/10712 [25:52<1:26:17,  1.94it/s]
  6%|▋         | 676/10712 [25:53<1:26:06,  1.94it/s]
  6%|▋         | 677/10712 [25:53<1:26:41,  1.93it/s]
  6%|▋         | 678/10712 [25:54<1:26:27,  1.93it/s]
  6%|▋         | 679/10712 [25:54<1:26:45,  1.93it/s]
  6%|▋         | 680/10712 [25:55<1:26:54,  1.92it/s]
  6%|▋         | 681/10712 [25:55<1:26:04,  1.94it/s]
  6%|▋         | 682/10712 [25:56<1:26:08,  1.94it/s]
  6%|▋         | 683/10712 [25:56<1:26:01,  1.94it/s]
  6%|▋         | 684/10712 [25:57<1:26:12,  1.94it/s]
  6%|▋         | 685/10712 [25:57<1:26:03,  1.94it/s]
  6%|▋         | 686/10712 [25:58<1:25:47,  1.95it/s]
  6%|▋         | 687/10712 [25:58<1:25:56,  1.94it/s]
  6%|▋         | 688/10712 [25:59<1:26:47,  1.92it/s]
  6%|▋         | 689/10712 [25:59<1:26:24,  1.93it/s]
  6%|▋         | 690/10712 [26:00<1:26:23,  1.93it/s]
  6%|▋         | 691/10712 [26:00<1:26:15,  1.94it/s]
  6%|▋         | 692/10712 [26:01<1:25:21,  1.96it/s]
  6%|▋         | 693/10712 [26:01<1:25:32,  1.95it/s]
  6%|▋         | 694/10712 [26:02<1:25:31,  1.95it/s]
  6%|▋         | 695/10712 [26:02<1:25:34,  1.95it/s]
  6%|▋         | 696/10712 [26:03<1:25:44,  1.95it/s]
  7%|▋         | 697/10712 [26:04<1:25:39,  1.95it/s]
  7%|▋         | 698/10712 [26:04<1:25:06,  1.96it/s]
  7%|▋         | 699/10712 [26:05<1:25:52,  1.94it/s]
  7%|▋         | 700/10712 [26:05<1:25:27,  1.95it/s]{'loss': 4.8685, 'grad_norm': 0.5090237855911255, 'learning_rate': 0.0006529850746268657, 'epoch': 0.07}
                                                     

  7%|▋         | 700/10712 [26:05<1:25:27,  1.95it/s]
  7%|▋         | 701/10712 [26:06<1:25:27,  1.95it/s]
  7%|▋         | 702/10712 [26:06<1:25:31,  1.95it/s]
  7%|▋         | 703/10712 [26:07<1:25:16,  1.96it/s]
  7%|▋         | 704/10712 [26:07<1:25:24,  1.95it/s]
  7%|▋         | 705/10712 [26:08<1:25:07,  1.96it/s]
  7%|▋         | 706/10712 [26:08<1:24:39,  1.97it/s]
  7%|▋         | 707/10712 [26:09<1:24:42,  1.97it/s]
  7%|▋         | 708/10712 [26:09<1:25:01,  1.96it/s]
  7%|▋         | 709/10712 [26:10<1:24:40,  1.97it/s]
  7%|▋         | 710/10712 [26:10<1:24:46,  1.97it/s]
  7%|▋         | 711/10712 [26:11<1:24:15,  1.98it/s]
  7%|▋         | 712/10712 [26:11<1:24:25,  1.97it/s]
  7%|▋         | 713/10712 [26:12<1:24:28,  1.97it/s]
  7%|▋         | 714/10712 [26:12<1:24:47,  1.97it/s]
  7%|▋         | 715/10712 [26:13<1:25:02,  1.96it/s]
  7%|▋         | 716/10712 [26:13<1:24:58,  1.96it/s]
  7%|▋         | 717/10712 [26:14<1:25:03,  1.96it/s]
  7%|▋         | 718/10712 [26:14<1:24:55,  1.96it/s]
  7%|▋         | 719/10712 [26:15<1:24:43,  1.97it/s]
  7%|▋         | 720/10712 [26:15<1:24:35,  1.97it/s]
  7%|▋         | 721/10712 [26:16<1:24:32,  1.97it/s]
  7%|▋         | 722/10712 [26:16<1:24:16,  1.98it/s]
  7%|▋         | 723/10712 [26:17<1:24:21,  1.97it/s]
  7%|▋         | 724/10712 [26:17<1:24:18,  1.97it/s]
  7%|▋         | 725/10712 [26:18<1:24:24,  1.97it/s]{'loss': 4.8344, 'grad_norm': 0.4122297763824463, 'learning_rate': 0.0006763059701492538, 'epoch': 0.07}
                                                     

  7%|▋         | 725/10712 [26:18<1:24:24,  1.97it/s]
  7%|▋         | 726/10712 [26:18<1:25:01,  1.96it/s]
  7%|▋         | 727/10712 [26:19<1:25:07,  1.96it/s]
  7%|▋         | 728/10712 [26:19<1:25:18,  1.95it/s]
  7%|▋         | 729/10712 [26:20<1:24:59,  1.96it/s]
  7%|▋         | 730/10712 [26:20<1:25:36,  1.94it/s]
  7%|▋         | 731/10712 [26:21<1:25:20,  1.95it/s]
  7%|▋         | 732/10712 [26:21<1:24:59,  1.96it/s]
  7%|▋         | 733/10712 [26:22<1:25:18,  1.95it/s]
  7%|▋         | 734/10712 [26:22<1:24:51,  1.96it/s]
  7%|▋         | 735/10712 [26:23<1:25:12,  1.95it/s]
  7%|▋         | 736/10712 [26:23<1:24:47,  1.96it/s]
  7%|▋         | 737/10712 [26:24<1:24:32,  1.97it/s]
  7%|▋         | 738/10712 [26:24<1:24:25,  1.97it/s]
  7%|▋         | 739/10712 [26:25<1:24:07,  1.98it/s]
  7%|▋         | 740/10712 [26:25<1:24:29,  1.97it/s]
  7%|▋         | 741/10712 [26:26<1:23:54,  1.98it/s]
  7%|▋         | 742/10712 [26:26<1:24:25,  1.97it/s]
  7%|▋         | 743/10712 [26:27<1:24:39,  1.96it/s]
  7%|▋         | 744/10712 [26:27<1:24:37,  1.96it/s]
  7%|▋         | 745/10712 [26:28<1:24:22,  1.97it/s]
  7%|▋         | 746/10712 [26:28<1:24:26,  1.97it/s]
  7%|▋         | 747/10712 [26:29<1:24:01,  1.98it/s]
  7%|▋         | 748/10712 [26:29<1:23:49,  1.98it/s]
  7%|▋         | 749/10712 [26:30<1:24:15,  1.97it/s]
  7%|▋         | 750/10712 [26:30<1:24:28,  1.97it/s]{'loss': 4.7969, 'grad_norm': 0.42375779151916504, 'learning_rate': 0.0006996268656716418, 'epoch': 0.07}
                                                     

  7%|▋         | 750/10712 [26:30<1:24:28,  1.97it/s]
  7%|▋         | 751/10712 [26:31<1:24:29,  1.96it/s]
  7%|▋         | 752/10712 [26:32<1:24:29,  1.96it/s]
  7%|▋         | 753/10712 [26:32<1:24:37,  1.96it/s]
  7%|▋         | 754/10712 [26:33<1:23:59,  1.98it/s]
  7%|▋         | 755/10712 [26:33<1:24:01,  1.98it/s]
  7%|▋         | 756/10712 [26:34<1:24:00,  1.98it/s]
  7%|▋         | 757/10712 [26:34<1:23:42,  1.98it/s]
  7%|▋         | 758/10712 [26:35<1:23:50,  1.98it/s]
  7%|▋         | 759/10712 [26:35<1:23:49,  1.98it/s]
  7%|▋         | 760/10712 [26:36<1:23:51,  1.98it/s]
  7%|▋         | 761/10712 [26:36<1:23:54,  1.98it/s]
  7%|▋         | 762/10712 [26:37<1:23:41,  1.98it/s]
  7%|▋         | 763/10712 [26:37<1:23:48,  1.98it/s]
  7%|▋         | 764/10712 [26:38<1:23:58,  1.97it/s]
  7%|▋         | 765/10712 [26:38<1:23:59,  1.97it/s]
  7%|▋         | 766/10712 [26:39<1:24:09,  1.97it/s]
  7%|▋         | 767/10712 [26:39<1:24:04,  1.97it/s]
  7%|▋         | 768/10712 [26:40<1:24:08,  1.97it/s]
  7%|▋         | 769/10712 [26:40<1:24:22,  1.96it/s]
  7%|▋         | 770/10712 [26:41<1:24:12,  1.97it/s]
  7%|▋         | 771/10712 [26:41<1:23:42,  1.98it/s]
  7%|▋         | 772/10712 [26:42<1:24:06,  1.97it/s]
  7%|▋         | 773/10712 [26:42<1:24:09,  1.97it/s]
  7%|▋         | 774/10712 [26:43<1:24:06,  1.97it/s]
  7%|▋         | 775/10712 [26:43<1:23:42,  1.98it/s]
                                                     {'loss': 4.7778, 'grad_norm': 0.4149554371833801, 'learning_rate': 0.0007229477611940298, 'epoch': 0.07}

  7%|▋         | 775/10712 [26:43<1:23:42,  1.98it/s]
  7%|▋         | 776/10712 [26:44<1:23:56,  1.97it/s]
  7%|▋         | 777/10712 [26:44<1:24:08,  1.97it/s]
  7%|▋         | 778/10712 [26:45<1:24:04,  1.97it/s]
  7%|▋         | 779/10712 [26:45<1:23:41,  1.98it/s]
  7%|▋         | 780/10712 [26:46<1:23:27,  1.98it/s]
  7%|▋         | 781/10712 [26:46<1:23:53,  1.97it/s]
  7%|▋         | 782/10712 [26:47<1:23:42,  1.98it/s]
  7%|▋         | 783/10712 [26:47<1:23:57,  1.97it/s]
  7%|▋         | 784/10712 [26:48<1:24:21,  1.96it/s]
  7%|▋         | 785/10712 [26:48<1:24:13,  1.96it/s]
  7%|▋         | 786/10712 [26:49<1:24:28,  1.96it/s]
  7%|▋         | 787/10712 [26:49<1:23:58,  1.97it/s]
  7%|▋         | 788/10712 [26:50<1:24:43,  1.95it/s]
  7%|▋         | 789/10712 [26:50<1:24:29,  1.96it/s]
  7%|▋         | 790/10712 [26:51<1:23:55,  1.97it/s]
  7%|▋         | 791/10712 [26:51<1:23:43,  1.97it/s]
  7%|▋         | 792/10712 [26:52<1:23:16,  1.99it/s]
  7%|▋         | 793/10712 [26:52<1:23:11,  1.99it/s]
  7%|▋         | 794/10712 [26:53<1:23:21,  1.98it/s]
  7%|▋         | 795/10712 [26:53<1:23:23,  1.98it/s]
  7%|▋         | 796/10712 [26:54<1:23:14,  1.99it/s]
  7%|▋         | 797/10712 [26:54<1:23:32,  1.98it/s]
  7%|▋         | 798/10712 [26:55<1:24:18,  1.96it/s]
  7%|▋         | 799/10712 [26:55<1:24:03,  1.97it/s]
  7%|▋         | 800/10712 [26:56<1:24:31,  1.95it/s]{'loss': 4.7516, 'grad_norm': 0.4027431607246399, 'learning_rate': 0.0007462686567164179, 'epoch': 0.07}

                                                     

  7%|▋         | 800/10712 [26:56<1:24:31,  1.95it/s]
  7%|▋         | 801/10712 [26:56<1:25:00,  1.94it/s]
  7%|▋         | 802/10712 [26:57<1:24:44,  1.95it/s]
  7%|▋         | 803/10712 [26:57<1:24:26,  1.96it/s]
  8%|▊         | 804/10712 [26:58<1:23:38,  1.97it/s]
  8%|▊         | 805/10712 [26:58<1:23:25,  1.98it/s]
  8%|▊         | 806/10712 [26:59<1:23:19,  1.98it/s]
  8%|▊         | 807/10712 [26:59<1:23:24,  1.98it/s]
  8%|▊         | 808/10712 [27:00<1:23:01,  1.99it/s]
  8%|▊         | 809/10712 [27:00<1:23:20,  1.98it/s]
  8%|▊         | 810/10712 [27:01<1:23:05,  1.99it/s]
  8%|▊         | 811/10712 [27:01<1:23:06,  1.99it/s]
  8%|▊         | 812/10712 [27:02<1:23:32,  1.98it/s]
  8%|▊         | 813/10712 [27:02<1:23:35,  1.97it/s]
  8%|▊         | 814/10712 [27:03<1:23:40,  1.97it/s]
  8%|▊         | 815/10712 [27:03<1:23:13,  1.98it/s]
  8%|▊         | 816/10712 [27:04<1:23:14,  1.98it/s]
  8%|▊         | 817/10712 [27:04<1:22:59,  1.99it/s]
  8%|▊         | 818/10712 [27:05<1:23:01,  1.99it/s]
  8%|▊         | 819/10712 [27:05<1:22:47,  1.99it/s]
  8%|▊         | 820/10712 [27:06<1:22:34,  2.00it/s]
  8%|▊         | 821/10712 [27:06<1:23:07,  1.98it/s]
  8%|▊         | 822/10712 [27:07<1:23:05,  1.98it/s]
  8%|▊         | 823/10712 [27:07<1:22:49,  1.99it/s]
  8%|▊         | 824/10712 [27:08<1:22:47,  1.99it/s]
  8%|▊         | 825/10712 [27:08<1:23:04,  1.98it/s]{'loss': 4.7104, 'grad_norm': 0.35946279764175415, 'learning_rate': 0.000769589552238806, 'epoch': 0.08}
                                                     

  8%|▊         | 825/10712 [27:08<1:23:04,  1.98it/s]
  8%|▊         | 826/10712 [27:09<1:23:17,  1.98it/s]
  8%|▊         | 827/10712 [27:09<1:22:45,  1.99it/s]
  8%|▊         | 828/10712 [27:10<1:22:50,  1.99it/s]
  8%|▊         | 829/10712 [27:10<1:23:01,  1.98it/s]
  8%|▊         | 830/10712 [27:11<1:23:00,  1.98it/s]
  8%|▊         | 831/10712 [27:12<1:23:43,  1.97it/s]
  8%|▊         | 832/10712 [27:12<1:23:37,  1.97it/s]
  8%|▊         | 833/10712 [27:13<1:23:29,  1.97it/s]
  8%|▊         | 834/10712 [27:13<1:23:06,  1.98it/s]
  8%|▊         | 835/10712 [27:14<1:23:28,  1.97it/s]
  8%|▊         | 836/10712 [27:14<1:23:34,  1.97it/s]
  8%|▊         | 837/10712 [27:15<1:23:31,  1.97it/s]
  8%|▊         | 838/10712 [27:15<1:23:31,  1.97it/s]
  8%|▊         | 839/10712 [27:16<1:23:38,  1.97it/s]
  8%|▊         | 840/10712 [27:16<1:23:23,  1.97it/s]
  8%|▊         | 841/10712 [27:17<1:23:18,  1.97it/s]
  8%|▊         | 842/10712 [27:17<1:23:47,  1.96it/s]
  8%|▊         | 843/10712 [27:18<1:23:55,  1.96it/s]
  8%|▊         | 844/10712 [27:18<1:23:27,  1.97it/s]
  8%|▊         | 845/10712 [27:19<1:23:13,  1.98it/s]
  8%|▊         | 846/10712 [27:19<1:23:22,  1.97it/s]
  8%|▊         | 847/10712 [27:20<1:23:23,  1.97it/s]
  8%|▊         | 848/10712 [27:20<1:23:14,  1.97it/s]
  8%|▊         | 849/10712 [27:21<1:23:39,  1.96it/s]
  8%|▊         | 850/10712 [27:21<1:23:01,  1.98it/s]
                                                     
{'loss': 4.6911, 'grad_norm': 0.4754085838794708, 'learning_rate': 0.0007929104477611941, 'epoch': 0.08}

  8%|▊         | 850/10712 [27:21<1:23:01,  1.98it/s]
  8%|▊         | 851/10712 [27:22<1:23:10,  1.98it/s]
  8%|▊         | 852/10712 [27:22<1:22:44,  1.99it/s]
  8%|▊         | 853/10712 [27:23<1:22:26,  1.99it/s]
  8%|▊         | 854/10712 [27:23<1:22:41,  1.99it/s]
  8%|▊         | 855/10712 [27:24<1:22:45,  1.98it/s]
  8%|▊         | 856/10712 [27:24<1:22:49,  1.98it/s]
  8%|▊         | 857/10712 [27:25<1:22:22,  1.99it/s]
  8%|▊         | 858/10712 [27:25<1:22:11,  2.00it/s]
  8%|▊         | 859/10712 [27:26<1:22:39,  1.99it/s]
  8%|▊         | 860/10712 [27:26<1:22:36,  1.99it/s]
  8%|▊         | 861/10712 [27:27<1:22:59,  1.98it/s]
  8%|▊         | 862/10712 [27:27<1:23:16,  1.97it/s]
  8%|▊         | 863/10712 [27:28<1:23:02,  1.98it/s]
  8%|▊         | 864/10712 [27:28<1:36:31,  1.70it/s]
  8%|▊         | 865/10712 [27:29<1:42:29,  1.60it/s]
  8%|▊         | 866/10712 [27:30<1:45:11,  1.56it/s]
  8%|▊         | 867/10712 [27:31<1:52:12,  1.46it/s]
  8%|▊         | 868/10712 [27:31<1:53:56,  1.44it/s]
  8%|▊         | 869/10712 [27:32<1:50:48,  1.48it/s]
  8%|▊         | 870/10712 [27:33<1:52:49,  1.45it/s]
  8%|▊         | 871/10712 [27:33<1:50:30,  1.48it/s]
  8%|▊         | 872/10712 [27:34<1:49:04,  1.50it/s]
  8%|▊         | 873/10712 [27:35<1:48:13,  1.52it/s]
  8%|▊         | 874/10712 [27:35<1:40:44,  1.63it/s]
  8%|▊         | 875/10712 [27:36<1:35:58,  1.71it/s]
                                                     {'loss': 4.6654, 'grad_norm': 0.4272907078266144, 'learning_rate': 0.0008162313432835821, 'epoch': 0.08}

  8%|▊         | 875/10712 [27:36<1:35:58,  1.71it/s]
  8%|▊         | 876/10712 [27:36<1:32:12,  1.78it/s]
  8%|▊         | 877/10712 [27:37<1:29:39,  1.83it/s]
  8%|▊         | 878/10712 [27:37<1:28:11,  1.86it/s]
  8%|▊         | 879/10712 [27:38<1:32:48,  1.77it/s]
  8%|▊         | 880/10712 [27:38<1:33:58,  1.74it/s]
  8%|▊         | 881/10712 [27:39<1:45:05,  1.56it/s]
  8%|▊         | 882/10712 [27:40<1:45:39,  1.55it/s]
  8%|▊         | 883/10712 [27:40<1:43:00,  1.59it/s]
  8%|▊         | 884/10712 [27:41<1:44:37,  1.57it/s]
  8%|▊         | 885/10712 [27:42<1:44:18,  1.57it/s]
  8%|▊         | 886/10712 [27:42<1:46:19,  1.54it/s]
  8%|▊         | 887/10712 [27:43<1:43:03,  1.59it/s]
  8%|▊         | 888/10712 [27:44<1:43:30,  1.58it/s]
  8%|▊         | 889/10712 [27:44<1:45:02,  1.56it/s]
  8%|▊         | 890/10712 [27:45<1:48:37,  1.51it/s]
  8%|▊         | 891/10712 [27:46<1:49:53,  1.49it/s]
  8%|▊         | 892/10712 [27:46<1:46:25,  1.54it/s]
  8%|▊         | 893/10712 [27:47<1:39:33,  1.64it/s]
  8%|▊         | 894/10712 [27:47<1:34:18,  1.74it/s]
  8%|▊         | 895/10712 [27:48<1:38:22,  1.66it/s]
  8%|▊         | 896/10712 [27:49<1:38:28,  1.66it/s]
  8%|▊         | 897/10712 [27:49<1:41:29,  1.61it/s]
  8%|▊         | 898/10712 [27:50<1:44:37,  1.56it/s]
  8%|▊         | 899/10712 [27:51<1:44:21,  1.57it/s]
  8%|▊         | 900/10712 [27:51<1:42:06,  1.60it/s]
                                                     
{'loss': 4.6403, 'grad_norm': 0.34675133228302, 'learning_rate': 0.0008395522388059703, 'epoch': 0.08}

  8%|▊         | 900/10712 [27:51<1:42:06,  1.60it/s]
  8%|▊         | 901/10712 [27:52<1:45:26,  1.55it/s]
  8%|▊         | 902/10712 [27:53<1:47:03,  1.53it/s]
  8%|▊         | 903/10712 [27:53<1:47:24,  1.52it/s]
  8%|▊         | 904/10712 [27:54<1:40:28,  1.63it/s]
  8%|▊         | 905/10712 [27:54<1:37:48,  1.67it/s]
  8%|▊         | 906/10712 [27:55<1:37:12,  1.68it/s]
  8%|▊         | 907/10712 [27:55<1:32:54,  1.76it/s]
  8%|▊         | 908/10712 [27:56<1:29:40,  1.82it/s]
  8%|▊         | 909/10712 [27:56<1:27:30,  1.87it/s]
  8%|▊         | 910/10712 [27:57<1:26:16,  1.89it/s]
  9%|▊         | 911/10712 [27:57<1:28:12,  1.85it/s]
  9%|▊         | 912/10712 [27:58<1:36:53,  1.69it/s]
  9%|▊         | 913/10712 [27:59<1:37:29,  1.68it/s]
  9%|▊         | 914/10712 [27:59<1:39:24,  1.64it/s]
  9%|▊         | 915/10712 [28:00<1:44:25,  1.56it/s]
  9%|▊         | 916/10712 [28:01<1:50:38,  1.48it/s]
  9%|▊         | 917/10712 [28:02<1:48:22,  1.51it/s]
  9%|▊         | 918/10712 [28:02<1:47:06,  1.52it/s]
  9%|▊         | 919/10712 [28:03<1:46:54,  1.53it/s]
  9%|▊         | 920/10712 [28:03<1:46:37,  1.53it/s]
  9%|▊         | 921/10712 [28:04<1:55:11,  1.42it/s]
  9%|▊         | 922/10712 [28:05<1:53:30,  1.44it/s]
  9%|▊         | 923/10712 [28:06<1:47:38,  1.52it/s]
  9%|▊         | 924/10712 [28:06<1:43:44,  1.57it/s]
  9%|▊         | 925/10712 [28:07<1:37:21,  1.68it/s]{'loss': 4.6221, 'grad_norm': 0.3793068826198578, 'learning_rate': 0.0008628731343283582, 'epoch': 0.09}

                                                     

  9%|▊         | 925/10712 [28:07<1:37:21,  1.68it/s]
  9%|▊         | 926/10712 [28:07<1:33:02,  1.75it/s]
  9%|▊         | 927/10712 [28:08<1:30:05,  1.81it/s]
  9%|▊         | 928/10712 [28:08<1:27:59,  1.85it/s]
  9%|▊         | 929/10712 [28:09<1:26:18,  1.89it/s]
  9%|▊         | 930/10712 [28:09<1:25:33,  1.91it/s]
  9%|▊         | 931/10712 [28:10<1:24:19,  1.93it/s]
  9%|▊         | 932/10712 [28:10<1:23:22,  1.96it/s]
  9%|▊         | 933/10712 [28:11<1:23:41,  1.95it/s]
  9%|▊         | 934/10712 [28:11<1:23:18,  1.96it/s]
  9%|▊         | 935/10712 [28:12<1:23:09,  1.96it/s]
  9%|▊         | 936/10712 [28:12<1:23:01,  1.96it/s]
  9%|▊         | 937/10712 [28:13<1:22:55,  1.96it/s]
  9%|▉         | 938/10712 [28:13<1:22:26,  1.98it/s]
  9%|▉         | 939/10712 [28:14<1:22:14,  1.98it/s]
  9%|▉         | 940/10712 [28:14<1:22:06,  1.98it/s]
  9%|▉         | 941/10712 [28:15<1:22:01,  1.99it/s]
  9%|▉         | 942/10712 [28:15<1:22:28,  1.97it/s]
  9%|▉         | 943/10712 [28:16<1:22:29,  1.97it/s]
  9%|▉         | 944/10712 [28:16<1:22:07,  1.98it/s]
  9%|▉         | 945/10712 [28:17<1:21:48,  1.99it/s]
  9%|▉         | 946/10712 [28:17<1:22:14,  1.98it/s]
  9%|▉         | 947/10712 [28:18<1:23:09,  1.96it/s]
  9%|▉         | 948/10712 [28:18<1:22:40,  1.97it/s]
  9%|▉         | 949/10712 [28:19<1:22:26,  1.97it/s]
  9%|▉         | 950/10712 [28:19<1:22:12,  1.98it/s]{'loss': 4.5979, 'grad_norm': 0.3419727385044098, 'learning_rate': 0.0008861940298507463, 'epoch': 0.09}

                                                     

  9%|▉         | 950/10712 [28:19<1:22:12,  1.98it/s]
  9%|▉         | 951/10712 [28:20<1:21:49,  1.99it/s]
  9%|▉         | 952/10712 [28:20<1:22:00,  1.98it/s]
  9%|▉         | 953/10712 [28:21<1:22:09,  1.98it/s]
  9%|▉         | 954/10712 [28:21<1:21:50,  1.99it/s]
  9%|▉         | 955/10712 [28:22<1:22:00,  1.98it/s]
  9%|▉         | 956/10712 [28:22<1:21:51,  1.99it/s]
  9%|▉         | 957/10712 [28:23<1:21:28,  2.00it/s]
  9%|▉         | 958/10712 [28:23<1:21:32,  1.99it/s]
  9%|▉         | 959/10712 [28:24<1:21:13,  2.00it/s]
  9%|▉         | 960/10712 [28:24<1:21:03,  2.01it/s]
  9%|▉         | 961/10712 [28:25<1:21:07,  2.00it/s]
  9%|▉         | 962/10712 [28:25<1:21:29,  1.99it/s]
  9%|▉         | 963/10712 [28:26<1:21:04,  2.00it/s]
  9%|▉         | 964/10712 [28:26<1:21:12,  2.00it/s]
  9%|▉         | 965/10712 [28:27<1:22:00,  1.98it/s]
  9%|▉         | 966/10712 [28:27<1:22:03,  1.98it/s]
  9%|▉         | 967/10712 [28:28<1:21:42,  1.99it/s]
  9%|▉         | 968/10712 [28:28<1:21:47,  1.99it/s]
  9%|▉         | 969/10712 [28:29<1:21:21,  2.00it/s]
  9%|▉         | 970/10712 [28:29<1:21:24,  1.99it/s]
  9%|▉         | 971/10712 [28:30<1:22:02,  1.98it/s]
  9%|▉         | 972/10712 [28:30<1:21:58,  1.98it/s]
  9%|▉         | 973/10712 [28:31<1:22:08,  1.98it/s]
  9%|▉         | 974/10712 [28:31<1:21:37,  1.99it/s]
  9%|▉         | 975/10712 [28:32<1:22:02,  1.98it/s]{'loss': 4.5751, 'grad_norm': 0.31271716952323914, 'learning_rate': 0.0009095149253731343, 'epoch': 0.09}
                                                     

  9%|▉         | 975/10712 [28:32<1:22:02,  1.98it/s]
  9%|▉         | 976/10712 [28:32<1:21:35,  1.99it/s]
  9%|▉         | 977/10712 [28:33<1:21:46,  1.98it/s]
  9%|▉         | 978/10712 [28:33<1:22:11,  1.97it/s]
  9%|▉         | 979/10712 [28:34<1:21:56,  1.98it/s]
  9%|▉         | 980/10712 [28:34<1:21:24,  1.99it/s]
  9%|▉         | 981/10712 [28:35<1:21:06,  2.00it/s]
  9%|▉         | 982/10712 [28:35<1:21:12,  2.00it/s]
  9%|▉         | 983/10712 [28:36<1:20:59,  2.00it/s]
  9%|▉         | 984/10712 [28:36<1:20:36,  2.01it/s]
  9%|▉         | 985/10712 [28:37<1:20:31,  2.01it/s]
  9%|▉         | 986/10712 [28:37<1:20:37,  2.01it/s]
  9%|▉         | 987/10712 [28:38<1:20:53,  2.00it/s]
  9%|▉         | 988/10712 [28:38<1:21:08,  2.00it/s]
  9%|▉         | 989/10712 [28:39<1:20:52,  2.00it/s]
  9%|▉         | 990/10712 [28:39<1:20:53,  2.00it/s]
  9%|▉         | 991/10712 [28:40<1:20:54,  2.00it/s]
  9%|▉         | 992/10712 [28:40<1:20:46,  2.01it/s]
  9%|▉         | 993/10712 [28:41<1:20:38,  2.01it/s]
  9%|▉         | 994/10712 [28:41<1:20:31,  2.01it/s]
  9%|▉         | 995/10712 [28:42<1:20:38,  2.01it/s]
  9%|▉         | 996/10712 [28:42<1:20:24,  2.01it/s]
  9%|▉         | 997/10712 [28:43<1:20:33,  2.01it/s]
  9%|▉         | 998/10712 [28:43<1:20:34,  2.01it/s]
  9%|▉         | 999/10712 [28:44<1:21:02,  2.00it/s]
  9%|▉         | 1000/10712 [28:44<1:21:16,  1.99it/s]
                                                      
{'loss': 4.5555, 'grad_norm': 0.3481411337852478, 'learning_rate': 0.0009328358208955225, 'epoch': 0.09}

  9%|▉         | 1000/10712 [28:44<1:21:16,  1.99it/s]
  9%|▉         | 1001/10712 [28:45<1:21:32,  1.98it/s]
  9%|▉         | 1002/10712 [28:45<1:21:26,  1.99it/s]
  9%|▉         | 1003/10712 [28:46<1:20:55,  2.00it/s]
  9%|▉         | 1004/10712 [28:46<1:20:48,  2.00it/s]
  9%|▉         | 1005/10712 [28:47<1:20:38,  2.01it/s]
  9%|▉         | 1006/10712 [28:47<1:20:27,  2.01it/s]
  9%|▉         | 1007/10712 [28:48<1:20:24,  2.01it/s]
  9%|▉         | 1008/10712 [28:48<1:20:37,  2.01it/s]
  9%|▉         | 1009/10712 [28:49<1:20:32,  2.01it/s]
  9%|▉         | 1010/10712 [28:49<1:20:26,  2.01it/s]
  9%|▉         | 1011/10712 [28:50<1:20:19,  2.01it/s]
  9%|▉         | 1012/10712 [28:50<1:20:10,  2.02it/s]
  9%|▉         | 1013/10712 [28:51<1:20:04,  2.02it/s]
  9%|▉         | 1014/10712 [28:51<1:20:16,  2.01it/s]
  9%|▉         | 1015/10712 [28:52<1:20:21,  2.01it/s]
  9%|▉         | 1016/10712 [28:52<1:21:07,  1.99it/s]
  9%|▉         | 1017/10712 [28:53<1:20:53,  2.00it/s]
 10%|▉         | 1018/10712 [28:53<1:20:51,  2.00it/s]
 10%|▉         | 1019/10712 [28:54<1:20:39,  2.00it/s]
 10%|▉         | 1020/10712 [28:54<1:20:42,  2.00it/s]
 10%|▉         | 1021/10712 [28:55<1:20:56,  2.00it/s]
 10%|▉         | 1022/10712 [28:55<1:20:30,  2.01it/s]
 10%|▉         | 1023/10712 [28:56<1:20:40,  2.00it/s]
 10%|▉         | 1024/10712 [28:57<1:33:23,  1.73it/s]
 10%|▉         | 1025/10712 [28:57<1:29:44,  1.80it/s]{'loss': 4.5276, 'grad_norm': 0.31863290071487427, 'learning_rate': 0.0009561567164179105, 'epoch': 0.1}

                                                      

 10%|▉         | 1025/10712 [28:57<1:29:44,  1.80it/s]
 10%|▉         | 1026/10712 [28:58<1:26:44,  1.86it/s]
 10%|▉         | 1027/10712 [28:58<1:24:44,  1.90it/s]
 10%|▉         | 1028/10712 [28:59<1:23:22,  1.94it/s]
 10%|▉         | 1029/10712 [28:59<1:22:26,  1.96it/s]
 10%|▉         | 1030/10712 [29:00<1:22:10,  1.96it/s]
 10%|▉         | 1031/10712 [29:00<1:21:38,  1.98it/s]
 10%|▉         | 1032/10712 [29:01<1:21:03,  1.99it/s]
 10%|▉         | 1033/10712 [29:01<1:20:55,  1.99it/s]
 10%|▉         | 1034/10712 [29:02<1:20:45,  2.00it/s]
 10%|▉         | 1035/10712 [29:02<1:20:43,  2.00it/s]
 10%|▉         | 1036/10712 [29:03<1:33:17,  1.73it/s]
 10%|▉         | 1037/10712 [29:03<1:29:04,  1.81it/s]
 10%|▉         | 1038/10712 [29:04<1:26:45,  1.86it/s]
 10%|▉         | 1039/10712 [29:04<1:24:39,  1.90it/s]
 10%|▉         | 1040/10712 [29:05<1:24:07,  1.92it/s]
 10%|▉         | 1041/10712 [29:05<1:23:01,  1.94it/s]
 10%|▉         | 1042/10712 [29:06<1:21:56,  1.97it/s]
 10%|▉         | 1043/10712 [29:06<1:22:03,  1.96it/s]
 10%|▉         | 1044/10712 [29:07<1:21:18,  1.98it/s]
 10%|▉         | 1045/10712 [29:07<1:21:09,  1.99it/s]
 10%|▉         | 1046/10712 [29:08<1:20:48,  1.99it/s]
 10%|▉         | 1047/10712 [29:08<1:20:32,  2.00it/s]
 10%|▉         | 1048/10712 [29:09<1:20:52,  1.99it/s]
 10%|▉         | 1049/10712 [29:09<1:20:44,  1.99it/s]
 10%|▉         | 1050/10712 [29:10<1:20:28,  2.00it/s]
                                                      
{'loss': 4.523, 'grad_norm': 0.33812209963798523, 'learning_rate': 0.0009794776119402984, 'epoch': 0.1}

 10%|▉         | 1050/10712 [29:10<1:20:28,  2.00it/s]
 10%|▉         | 1051/10712 [29:10<1:20:32,  2.00it/s]
 10%|▉         | 1052/10712 [29:11<1:20:40,  2.00it/s]
 10%|▉         | 1053/10712 [29:11<1:20:17,  2.00it/s]
 10%|▉         | 1054/10712 [29:12<1:20:32,  2.00it/s]
 10%|▉         | 1055/10712 [29:12<1:20:24,  2.00it/s]
 10%|▉         | 1056/10712 [29:13<1:20:01,  2.01it/s]
 10%|▉         | 1057/10712 [29:13<1:19:52,  2.01it/s]
 10%|▉         | 1058/10712 [29:14<1:20:34,  2.00it/s]
 10%|▉         | 1059/10712 [29:14<1:20:38,  2.00it/s]
 10%|▉         | 1060/10712 [29:15<1:20:21,  2.00it/s]
 10%|▉         | 1061/10712 [29:15<1:20:09,  2.01it/s]
 10%|▉         | 1062/10712 [29:16<1:20:09,  2.01it/s]
 10%|▉         | 1063/10712 [29:16<1:20:08,  2.01it/s]
 10%|▉         | 1064/10712 [29:17<1:20:35,  2.00it/s]
 10%|▉         | 1065/10712 [29:17<1:20:08,  2.01it/s]
 10%|▉         | 1066/10712 [29:18<1:20:00,  2.01it/s]
 10%|▉         | 1067/10712 [29:18<1:19:59,  2.01it/s]
 10%|▉         | 1068/10712 [29:19<1:19:52,  2.01it/s]
 10%|▉         | 1069/10712 [29:19<1:19:49,  2.01it/s]
 10%|▉         | 1070/10712 [29:20<1:19:48,  2.01it/s]
 10%|▉         | 1071/10712 [29:20<1:19:35,  2.02it/s]
 10%|█         | 1072/10712 [29:21<1:19:33,  2.02it/s]
 10%|█         | 1073/10712 [29:21<1:19:41,  2.02it/s]
 10%|█         | 1074/10712 [29:22<1:19:47,  2.01it/s]
 10%|█         | 1075/10712 [29:22<1:20:00,  2.01it/s]{'loss': 4.4907, 'grad_norm': 0.3150307238101959, 'learning_rate': 0.0009999997610383759, 'epoch': 0.1}

                                                      

 10%|█         | 1075/10712 [29:22<1:20:00,  2.01it/s]
 10%|█         | 1076/10712 [29:23<1:20:03,  2.01it/s]
 10%|█         | 1077/10712 [29:23<1:19:54,  2.01it/s]
 10%|█         | 1078/10712 [29:24<1:20:31,  1.99it/s]
 10%|█         | 1079/10712 [29:24<1:20:28,  1.99it/s]
 10%|█         | 1080/10712 [29:25<1:20:20,  2.00it/s]
 10%|█         | 1081/10712 [29:25<1:20:24,  2.00it/s]
 10%|█         | 1082/10712 [29:26<1:20:26,  2.00it/s]
 10%|█         | 1083/10712 [29:26<1:20:14,  2.00it/s]
 10%|█         | 1084/10712 [29:27<1:21:04,  1.98it/s]
 10%|█         | 1085/10712 [29:27<1:20:26,  1.99it/s]
 10%|█         | 1086/10712 [29:28<1:20:32,  1.99it/s]
 10%|█         | 1087/10712 [29:28<1:20:23,  2.00it/s]
 10%|█         | 1088/10712 [29:29<1:19:54,  2.01it/s]
 10%|█         | 1089/10712 [29:29<1:20:02,  2.00it/s]
 10%|█         | 1090/10712 [29:30<1:20:05,  2.00it/s]
 10%|█         | 1091/10712 [29:30<1:19:58,  2.00it/s]
 10%|█         | 1092/10712 [29:31<1:20:09,  2.00it/s]
 10%|█         | 1093/10712 [29:31<1:19:54,  2.01it/s]
 10%|█         | 1094/10712 [29:32<1:20:21,  1.99it/s]
 10%|█         | 1095/10712 [29:32<1:20:00,  2.00it/s]
 10%|█         | 1096/10712 [29:33<1:19:43,  2.01it/s]
 10%|█         | 1097/10712 [29:33<1:19:57,  2.00it/s]
 10%|█         | 1098/10712 [29:34<1:19:46,  2.01it/s]
 10%|█         | 1099/10712 [29:34<1:19:35,  2.01it/s]
 10%|█         | 1100/10712 [29:35<1:19:51,  2.01it/s]{'loss': 4.4676, 'grad_norm': 0.2963699400424957, 'learning_rate': 0.0009999791839302003, 'epoch': 0.1}
                                                      

 10%|█         | 1100/10712 [29:35<1:19:51,  2.01it/s]
 10%|█         | 1101/10712 [29:35<1:19:45,  2.01it/s]
 10%|█         | 1102/10712 [29:36<1:19:42,  2.01it/s]
 10%|█         | 1103/10712 [29:36<1:20:03,  2.00it/s]
 10%|█         | 1104/10712 [29:37<1:20:23,  1.99it/s]
 10%|█         | 1105/10712 [29:37<1:20:06,  2.00it/s]
 10%|█         | 1106/10712 [29:38<1:20:21,  1.99it/s]
 10%|█         | 1107/10712 [29:38<1:20:09,  2.00it/s]
 10%|█         | 1108/10712 [29:39<1:20:09,  2.00it/s]
 10%|█         | 1109/10712 [29:39<1:19:58,  2.00it/s]
 10%|█         | 1110/10712 [29:40<1:19:55,  2.00it/s]
 10%|█         | 1111/10712 [29:40<1:19:57,  2.00it/s]
 10%|█         | 1112/10712 [29:41<1:19:42,  2.01it/s]
 10%|█         | 1113/10712 [29:41<1:19:56,  2.00it/s]
 10%|█         | 1114/10712 [29:42<1:19:49,  2.00it/s]
 10%|█         | 1115/10712 [29:42<1:20:14,  1.99it/s]
 10%|█         | 1116/10712 [29:43<1:20:07,  2.00it/s]
 10%|█         | 1117/10712 [29:43<1:19:53,  2.00it/s]
 10%|█         | 1118/10712 [29:44<1:19:31,  2.01it/s]
 10%|█         | 1119/10712 [29:44<1:20:03,  2.00it/s]
 10%|█         | 1120/10712 [29:45<1:19:41,  2.01it/s]
 10%|█         | 1121/10712 [29:45<1:20:03,  2.00it/s]
 10%|█         | 1122/10712 [29:46<1:19:54,  2.00it/s]
 10%|█         | 1123/10712 [29:46<1:19:42,  2.01it/s]
 10%|█         | 1124/10712 [29:47<1:19:52,  2.00it/s]
 11%|█         | 1125/10712 [29:47<1:19:33,  2.01it/s]{'loss': 4.4438, 'grad_norm': 0.31959035992622375, 'learning_rate': 0.0009999254192702488, 'epoch': 0.11}
                                                      

 11%|█         | 1125/10712 [29:47<1:19:33,  2.01it/s]
 11%|█         | 1126/10712 [29:48<1:19:33,  2.01it/s]
 11%|█         | 1127/10712 [29:48<1:19:16,  2.02it/s]
 11%|█         | 1128/10712 [29:49<1:19:43,  2.00it/s]
 11%|█         | 1129/10712 [29:49<1:19:32,  2.01it/s]
 11%|█         | 1130/10712 [29:50<1:19:45,  2.00it/s]
 11%|█         | 1131/10712 [29:50<1:19:40,  2.00it/s]
 11%|█         | 1132/10712 [29:51<1:19:28,  2.01it/s]
 11%|█         | 1133/10712 [29:51<1:19:23,  2.01it/s]
 11%|█         | 1134/10712 [29:52<1:19:41,  2.00it/s]
 11%|█         | 1135/10712 [29:52<1:19:24,  2.01it/s]
 11%|█         | 1136/10712 [29:53<1:19:08,  2.02it/s]
 11%|█         | 1137/10712 [29:53<1:19:08,  2.02it/s]
 11%|█         | 1138/10712 [29:54<1:19:06,  2.02it/s]
 11%|█         | 1139/10712 [29:54<1:19:26,  2.01it/s]
 11%|█         | 1140/10712 [29:55<1:19:40,  2.00it/s]
 11%|█         | 1141/10712 [29:55<1:20:08,  1.99it/s]
 11%|█         | 1142/10712 [29:56<1:20:07,  1.99it/s]
 11%|█         | 1143/10712 [29:56<1:19:35,  2.00it/s]
 11%|█         | 1144/10712 [29:57<1:19:12,  2.01it/s]
 11%|█         | 1145/10712 [29:57<1:19:57,  1.99it/s]
 11%|█         | 1146/10712 [29:58<1:19:37,  2.00it/s]
 11%|█         | 1147/10712 [29:58<1:19:19,  2.01it/s]
 11%|█         | 1148/10712 [29:59<1:19:14,  2.01it/s]
 11%|█         | 1149/10712 [29:59<1:19:11,  2.01it/s]
 11%|█         | 1150/10712 [30:00<1:19:11,  2.01it/s]
                                                      
{'loss': 4.4298, 'grad_norm': 0.292477548122406, 'learning_rate': 0.0009998384706273048, 'epoch': 0.11}

 11%|█         | 1150/10712 [30:00<1:19:11,  2.01it/s]
 11%|█         | 1151/10712 [30:00<1:19:10,  2.01it/s]
 11%|█         | 1152/10712 [30:01<1:19:05,  2.01it/s]
 11%|█         | 1153/10712 [30:01<1:19:12,  2.01it/s]
 11%|█         | 1154/10712 [30:02<1:18:54,  2.02it/s]
 11%|█         | 1155/10712 [30:02<1:18:56,  2.02it/s]
 11%|█         | 1156/10712 [30:03<1:19:01,  2.02it/s]
 11%|█         | 1157/10712 [30:03<1:18:51,  2.02it/s]
 11%|█         | 1158/10712 [30:04<1:18:56,  2.02it/s]
 11%|█         | 1159/10712 [30:04<1:19:14,  2.01it/s]
 11%|█         | 1160/10712 [30:05<1:19:05,  2.01it/s]
 11%|█         | 1161/10712 [30:05<1:29:10,  1.79it/s]
 11%|█         | 1162/10712 [30:06<1:26:02,  1.85it/s]
 11%|█         | 1163/10712 [30:06<1:23:42,  1.90it/s]
 11%|█         | 1164/10712 [30:07<1:22:31,  1.93it/s]
 11%|█         | 1165/10712 [30:07<1:21:53,  1.94it/s]
 11%|█         | 1166/10712 [30:08<1:21:21,  1.96it/s]
 11%|█         | 1167/10712 [30:08<1:20:58,  1.96it/s]
 11%|█         | 1168/10712 [30:09<1:20:45,  1.97it/s]
 11%|█         | 1169/10712 [30:09<1:20:49,  1.97it/s]
 11%|█         | 1170/10712 [30:10<1:20:55,  1.97it/s]
 11%|█         | 1171/10712 [30:10<1:20:31,  1.97it/s]
 11%|█         | 1172/10712 [30:11<1:20:17,  1.98it/s]
 11%|█         | 1173/10712 [30:11<1:19:43,  1.99it/s]
 11%|█         | 1174/10712 [30:12<1:19:19,  2.00it/s]
 11%|█         | 1175/10712 [30:12<1:19:05,  2.01it/s]
                                                      {'loss': 4.4052, 'grad_norm': 0.29016032814979553, 'learning_rate': 0.000999718343772834, 'epoch': 0.11}

 11%|█         | 1175/10712 [30:12<1:19:05,  2.01it/s]
 11%|█         | 1176/10712 [30:13<1:19:40,  1.99it/s]
 11%|█         | 1177/10712 [30:13<1:19:40,  1.99it/s]
 11%|█         | 1178/10712 [30:14<1:19:34,  2.00it/s]
 11%|█         | 1179/10712 [30:14<1:19:05,  2.01it/s]
 11%|█         | 1180/10712 [30:15<1:19:19,  2.00it/s]
 11%|█         | 1181/10712 [30:15<1:19:13,  2.00it/s]
 11%|█         | 1182/10712 [30:16<1:19:14,  2.00it/s]
 11%|█         | 1183/10712 [30:16<1:19:05,  2.01it/s]
 11%|█         | 1184/10712 [30:17<1:19:22,  2.00it/s]
 11%|█         | 1185/10712 [30:17<1:19:02,  2.01it/s]
 11%|█         | 1186/10712 [30:18<1:19:18,  2.00it/s]
 11%|█         | 1187/10712 [30:18<1:19:47,  1.99it/s]
 11%|█         | 1188/10712 [30:19<1:20:04,  1.98it/s]
 11%|█         | 1189/10712 [30:19<1:19:38,  1.99it/s]
 11%|█         | 1190/10712 [30:20<1:19:10,  2.00it/s]
 11%|█         | 1191/10712 [30:20<1:18:57,  2.01it/s]
 11%|█         | 1192/10712 [30:21<1:18:46,  2.01it/s]
 11%|█         | 1193/10712 [30:21<1:19:07,  2.01it/s]
 11%|█         | 1194/10712 [30:22<1:19:04,  2.01it/s]
 11%|█         | 1195/10712 [30:22<1:18:54,  2.01it/s]
 11%|█         | 1196/10712 [30:23<1:19:03,  2.01it/s]
 11%|█         | 1197/10712 [30:23<1:19:10,  2.00it/s]
 11%|█         | 1198/10712 [30:24<1:19:02,  2.01it/s]
 11%|█         | 1199/10712 [30:24<1:19:05,  2.00it/s]
 11%|█         | 1200/10712 [30:25<1:18:56,  2.01it/s]
                                                      {'loss': 4.3816, 'grad_norm': 0.29934656620025635, 'learning_rate': 0.0009995650466806004, 'epoch': 0.11}

 11%|█         | 1200/10712 [30:25<1:18:56,  2.01it/s]
 11%|█         | 1201/10712 [30:25<1:18:48,  2.01it/s]
 11%|█         | 1202/10712 [30:26<1:19:06,  2.00it/s]
 11%|█         | 1203/10712 [30:26<1:19:38,  1.99it/s]
 11%|█         | 1204/10712 [30:27<1:19:29,  1.99it/s]
 11%|█         | 1205/10712 [30:27<1:19:23,  2.00it/s]
 11%|█▏        | 1206/10712 [30:28<1:19:11,  2.00it/s]
 11%|█▏        | 1207/10712 [30:28<1:19:15,  2.00it/s]
 11%|█▏        | 1208/10712 [30:29<1:18:53,  2.01it/s]
 11%|█▏        | 1209/10712 [30:29<1:19:05,  2.00it/s]
 11%|█▏        | 1210/10712 [30:30<1:19:03,  2.00it/s]
 11%|█▏        | 1211/10712 [30:30<1:18:48,  2.01it/s]
 11%|█▏        | 1212/10712 [30:31<1:19:19,  2.00it/s]
 11%|█▏        | 1213/10712 [30:31<1:19:00,  2.00it/s]
 11%|█▏        | 1214/10712 [30:32<1:18:44,  2.01it/s]
 11%|█▏        | 1215/10712 [30:32<1:18:36,  2.01it/s]
 11%|█▏        | 1216/10712 [30:33<1:18:53,  2.01it/s]
 11%|█▏        | 1217/10712 [30:33<1:18:50,  2.01it/s]
 11%|█▏        | 1218/10712 [30:34<1:19:02,  2.00it/s]
 11%|█▏        | 1219/10712 [30:34<1:18:49,  2.01it/s]
 11%|█▏        | 1220/10712 [30:35<1:18:35,  2.01it/s]
 11%|█▏        | 1221/10712 [30:35<1:18:25,  2.02it/s]
 11%|█▏        | 1222/10712 [30:36<1:18:52,  2.01it/s]
 11%|█▏        | 1223/10712 [30:36<1:18:44,  2.01it/s]
 11%|█▏        | 1224/10712 [30:37<1:18:23,  2.02it/s]
 11%|█▏        | 1225/10712 [30:37<1:18:48,  2.01it/s]
                                                      {'loss': 4.3688, 'grad_norm': 0.28090035915374756, 'learning_rate': 0.0009993785895261384, 'epoch': 0.11}


 11%|█▏        | 1225/10712 [30:37<1:18:48,  2.01it/s]
 11%|█▏        | 1226/10712 [30:38<1:18:42,  2.01it/s]
 11%|█▏        | 1227/10712 [30:38<1:19:12,  2.00it/s]
 11%|█▏        | 1228/10712 [30:39<1:18:52,  2.00it/s]
 11%|█▏        | 1229/10712 [30:39<1:18:44,  2.01it/s]
 11%|█▏        | 1230/10712 [30:40<1:18:34,  2.01it/s]
 11%|█▏        | 1231/10712 [30:40<1:18:37,  2.01it/s]
 12%|█▏        | 1232/10712 [30:41<1:18:28,  2.01it/s]
 12%|█▏        | 1233/10712 [30:41<1:18:33,  2.01it/s]
 12%|█▏        | 1234/10712 [30:42<1:18:39,  2.01it/s]
 12%|█▏        | 1235/10712 [30:42<1:18:46,  2.00it/s]
 12%|█▏        | 1236/10712 [30:43<1:18:36,  2.01it/s]
 12%|█▏        | 1237/10712 [30:43<1:18:48,  2.00it/s]
 12%|█▏        | 1238/10712 [30:44<1:18:38,  2.01it/s]
 12%|█▏        | 1239/10712 [30:44<1:18:39,  2.01it/s]
 12%|█▏        | 1240/10712 [30:45<1:18:30,  2.01it/s]
 12%|█▏        | 1241/10712 [30:45<1:18:28,  2.01it/s]
 12%|█▏        | 1242/10712 [30:46<1:18:24,  2.01it/s]
 12%|█▏        | 1243/10712 [30:46<1:18:40,  2.01it/s]
 12%|█▏        | 1244/10712 [30:47<1:18:36,  2.01it/s]
 12%|█▏        | 1245/10712 [30:47<1:18:46,  2.00it/s]
 12%|█▏        | 1246/10712 [30:48<1:18:38,  2.01it/s]
 12%|█▏        | 1247/10712 [30:48<1:18:31,  2.01it/s]
 12%|█▏        | 1248/10712 [30:49<1:18:20,  2.01it/s]
 12%|█▏        | 1249/10712 [30:49<1:18:19,  2.01it/s]
 12%|█▏        | 1250/10712 [30:50<1:18:24,  2.01it/s]{'loss': 4.3565, 'grad_norm': 0.3117658197879791, 'learning_rate': 0.000999158984686076, 'epoch': 0.12}
                                                      

 12%|█▏        | 1250/10712 [30:50<1:18:24,  2.01it/s]
 12%|█▏        | 1251/10712 [30:50<1:18:35,  2.01it/s]
 12%|█▏        | 1252/10712 [30:51<1:18:44,  2.00it/s]
 12%|█▏        | 1253/10712 [30:51<1:18:22,  2.01it/s]
 12%|█▏        | 1254/10712 [30:52<1:18:39,  2.00it/s]
 12%|█▏        | 1255/10712 [30:52<1:18:25,  2.01it/s]
 12%|█▏        | 1256/10712 [30:53<1:18:11,  2.02it/s]
 12%|█▏        | 1257/10712 [30:53<1:18:09,  2.02it/s]
 12%|█▏        | 1258/10712 [30:54<1:18:52,  2.00it/s]
 12%|█▏        | 1259/10712 [30:54<1:18:46,  2.00it/s]
 12%|█▏        | 1260/10712 [30:55<1:18:29,  2.01it/s]
 12%|█▏        | 1261/10712 [30:55<1:18:16,  2.01it/s]
 12%|█▏        | 1262/10712 [30:56<1:18:12,  2.01it/s]
 12%|█▏        | 1263/10712 [30:56<1:18:15,  2.01it/s]
 12%|█▏        | 1264/10712 [30:57<1:18:34,  2.00it/s]
 12%|█▏        | 1265/10712 [30:57<1:18:40,  2.00it/s]
 12%|█▏        | 1266/10712 [30:58<1:18:21,  2.01it/s]
 12%|█▏        | 1267/10712 [30:58<1:18:17,  2.01it/s]
 12%|█▏        | 1268/10712 [30:59<1:18:16,  2.01it/s]
 12%|█▏        | 1269/10712 [30:59<1:18:05,  2.02it/s]
 12%|█▏        | 1270/10712 [31:00<1:18:28,  2.01it/s]
 12%|█▏        | 1271/10712 [31:00<1:19:00,  1.99it/s]
 12%|█▏        | 1272/10712 [31:01<1:18:36,  2.00it/s]
 12%|█▏        | 1273/10712 [31:01<1:18:13,  2.01it/s]
 12%|█▏        | 1274/10712 [31:02<1:18:30,  2.00it/s]
 12%|█▏        | 1275/10712 [31:02<1:18:22,  2.01it/s]{'loss': 4.3509, 'grad_norm': 0.2800970673561096, 'learning_rate': 0.000998906246737314, 'epoch': 0.12}
                                                      

 12%|█▏        | 1275/10712 [31:02<1:18:22,  2.01it/s]
 12%|█▏        | 1276/10712 [31:03<1:18:18,  2.01it/s]
 12%|█▏        | 1277/10712 [31:03<1:18:07,  2.01it/s]
 12%|█▏        | 1278/10712 [31:04<1:18:18,  2.01it/s]
 12%|█▏        | 1279/10712 [31:04<1:18:07,  2.01it/s]
 12%|█▏        | 1280/10712 [31:05<1:17:54,  2.02it/s]
 12%|█▏        | 1281/10712 [31:05<1:17:52,  2.02it/s]
 12%|█▏        | 1282/10712 [31:06<1:18:11,  2.01it/s]
 12%|█▏        | 1283/10712 [31:06<1:18:23,  2.00it/s]
 12%|█▏        | 1284/10712 [31:07<1:18:21,  2.01it/s]
 12%|█▏        | 1285/10712 [31:07<1:18:07,  2.01it/s]
 12%|█▏        | 1286/10712 [31:08<1:17:57,  2.02it/s]
 12%|█▏        | 1287/10712 [31:08<1:18:27,  2.00it/s]
 12%|█▏        | 1288/10712 [31:09<1:18:12,  2.01it/s]
 12%|█▏        | 1289/10712 [31:09<1:18:04,  2.01it/s]
 12%|█▏        | 1290/10712 [31:10<1:17:50,  2.02it/s]
 12%|█▏        | 1291/10712 [31:10<1:18:04,  2.01it/s]
 12%|█▏        | 1292/10712 [31:11<1:17:46,  2.02it/s]
 12%|█▏        | 1293/10712 [31:11<1:17:40,  2.02it/s]
 12%|█▏        | 1294/10712 [31:12<1:17:48,  2.02it/s]
 12%|█▏        | 1295/10712 [31:12<1:17:48,  2.02it/s]
 12%|█▏        | 1296/10712 [31:13<1:17:48,  2.02it/s]
 12%|█▏        | 1297/10712 [31:13<1:17:33,  2.02it/s]
 12%|█▏        | 1298/10712 [31:14<1:17:23,  2.03it/s]
 12%|█▏        | 1299/10712 [31:14<1:17:25,  2.03it/s]
 12%|█▏        | 1300/10712 [31:15<1:17:46,  2.02it/s]{'loss': 4.3238, 'grad_norm': 0.2877048850059509, 'learning_rate': 0.0009986203924560587, 'epoch': 0.12}
                                                      

 12%|█▏        | 1300/10712 [31:15<1:17:46,  2.02it/s]
 12%|█▏        | 1301/10712 [31:15<1:18:08,  2.01it/s]
 12%|█▏        | 1302/10712 [31:16<1:17:57,  2.01it/s]
 12%|█▏        | 1303/10712 [31:16<1:17:50,  2.01it/s]
 12%|█▏        | 1304/10712 [31:17<1:17:43,  2.02it/s]
 12%|█▏        | 1305/10712 [31:17<1:17:58,  2.01it/s]
 12%|█▏        | 1306/10712 [31:18<1:17:49,  2.01it/s]
 12%|█▏        | 1307/10712 [31:18<1:17:40,  2.02it/s]
 12%|█▏        | 1308/10712 [31:19<1:18:04,  2.01it/s]
 12%|█▏        | 1309/10712 [31:19<1:17:52,  2.01it/s]
 12%|█▏        | 1310/10712 [31:20<1:17:49,  2.01it/s]
 12%|█▏        | 1311/10712 [31:20<1:17:38,  2.02it/s]
 12%|█▏        | 1312/10712 [31:21<1:17:41,  2.02it/s]
 12%|█▏        | 1313/10712 [31:21<1:17:32,  2.02it/s]
 12%|█▏        | 1314/10712 [31:22<1:17:36,  2.02it/s]
 12%|█▏        | 1315/10712 [31:22<1:17:42,  2.02it/s]
 12%|█▏        | 1316/10712 [31:23<1:17:42,  2.02it/s]
 12%|█▏        | 1317/10712 [31:23<1:17:45,  2.01it/s]
 12%|█▏        | 1318/10712 [31:24<1:17:44,  2.01it/s]
 12%|█▏        | 1319/10712 [31:24<1:17:57,  2.01it/s]
 12%|█▏        | 1320/10712 [31:25<1:17:40,  2.02it/s]
 12%|█▏        | 1321/10712 [31:25<1:17:27,  2.02it/s]
 12%|█▏        | 1322/10712 [31:26<1:17:22,  2.02it/s]
 12%|█▏        | 1323/10712 [31:26<1:17:37,  2.02it/s]
 12%|█▏        | 1324/10712 [31:27<1:17:25,  2.02it/s]
 12%|█▏        | 1325/10712 [31:27<1:17:30,  2.02it/s]{'loss': 4.3188, 'grad_norm': 0.28635334968566895, 'learning_rate': 0.0009983014408167072, 'epoch': 0.12}

                                                      

 12%|█▏        | 1325/10712 [31:27<1:17:30,  2.02it/s]
 12%|█▏        | 1326/10712 [31:28<1:17:47,  2.01it/s]
 12%|█▏        | 1327/10712 [31:28<1:17:30,  2.02it/s]
 12%|█▏        | 1328/10712 [31:29<1:18:01,  2.00it/s]
 12%|█▏        | 1329/10712 [31:29<1:17:44,  2.01it/s]
 12%|█▏        | 1330/10712 [31:30<1:17:41,  2.01it/s]
 12%|█▏        | 1331/10712 [31:30<1:17:23,  2.02it/s]
 12%|█▏        | 1332/10712 [31:31<1:17:39,  2.01it/s]
 12%|█▏        | 1333/10712 [31:31<1:17:32,  2.02it/s]
 12%|█▏        | 1334/10712 [31:32<1:17:19,  2.02it/s]
 12%|█▏        | 1335/10712 [31:32<1:17:13,  2.02it/s]
 12%|█▏        | 1336/10712 [31:33<1:17:11,  2.02it/s]
 12%|█▏        | 1337/10712 [31:33<1:17:18,  2.02it/s]
 12%|█▏        | 1338/10712 [31:34<1:17:17,  2.02it/s]
 12%|█▎        | 1339/10712 [31:34<1:17:10,  2.02it/s]
 13%|█▎        | 1340/10712 [31:35<1:17:19,  2.02it/s]
 13%|█▎        | 1341/10712 [31:35<1:17:18,  2.02it/s]
 13%|█▎        | 1342/10712 [31:36<1:17:31,  2.01it/s]
 13%|█▎        | 1343/10712 [31:36<1:17:24,  2.02it/s]
 13%|█▎        | 1344/10712 [31:37<1:17:39,  2.01it/s]
 13%|█▎        | 1345/10712 [31:37<1:17:48,  2.01it/s]
 13%|█▎        | 1346/10712 [31:38<1:17:38,  2.01it/s]
 13%|█▎        | 1347/10712 [31:38<1:17:26,  2.02it/s]
 13%|█▎        | 1348/10712 [31:38<1:17:26,  2.02it/s]
 13%|█▎        | 1349/10712 [31:39<1:17:23,  2.02it/s]
 13%|█▎        | 1350/10712 [31:39<1:17:15,  2.02it/s]{'loss': 4.3038, 'grad_norm': 0.2710880935192108, 'learning_rate': 0.0009979494129905891, 'epoch': 0.13}

                                                      

 13%|█▎        | 1350/10712 [31:39<1:17:15,  2.02it/s]
 13%|█▎        | 1351/10712 [31:40<1:17:17,  2.02it/s]
 13%|█▎        | 1352/10712 [31:40<1:18:01,  2.00it/s]
 13%|█▎        | 1353/10712 [31:41<1:18:26,  1.99it/s]
 13%|█▎        | 1354/10712 [31:41<1:18:19,  1.99it/s]
 13%|█▎        | 1355/10712 [31:42<1:18:34,  1.98it/s]
 13%|█▎        | 1356/10712 [31:42<1:18:03,  2.00it/s]
 13%|█▎        | 1357/10712 [31:43<1:17:39,  2.01it/s]
 13%|█▎        | 1358/10712 [31:43<1:17:27,  2.01it/s]
 13%|█▎        | 1359/10712 [31:44<1:17:06,  2.02it/s]
 13%|█▎        | 1360/10712 [31:44<1:17:21,  2.01it/s]
 13%|█▎        | 1361/10712 [31:45<1:17:02,  2.02it/s]
 13%|█▎        | 1362/10712 [31:45<1:16:54,  2.03it/s]
 13%|█▎        | 1363/10712 [31:46<1:16:52,  2.03it/s]
 13%|█▎        | 1364/10712 [31:46<1:16:46,  2.03it/s]
 13%|█▎        | 1365/10712 [31:47<1:16:56,  2.02it/s]
 13%|█▎        | 1366/10712 [31:47<1:17:19,  2.01it/s]
 13%|█▎        | 1367/10712 [31:48<1:17:09,  2.02it/s]
 13%|█▎        | 1368/10712 [31:48<1:17:03,  2.02it/s]
 13%|█▎        | 1369/10712 [31:49<1:16:57,  2.02it/s]
 13%|█▎        | 1370/10712 [31:49<1:16:53,  2.03it/s]
 13%|█▎        | 1371/10712 [31:50<1:16:47,  2.03it/s]
 13%|█▎        | 1372/10712 [31:50<1:16:50,  2.03it/s]
 13%|█▎        | 1373/10712 [31:51<1:16:48,  2.03it/s]
 13%|█▎        | 1374/10712 [31:51<1:16:59,  2.02it/s]
 13%|█▎        | 1375/10712 [31:52<1:16:51,  2.02it/s]
                                                      {'loss': 4.288, 'grad_norm': 0.25179368257522583, 'learning_rate': 0.0009975643323445608, 'epoch': 0.13}


 13%|█▎        | 1375/10712 [31:52<1:16:51,  2.02it/s]
 13%|█▎        | 1376/10712 [31:52<1:17:12,  2.02it/s]
 13%|█▎        | 1377/10712 [31:53<1:17:12,  2.02it/s]
 13%|█▎        | 1378/10712 [31:53<1:17:25,  2.01it/s]
 13%|█▎        | 1379/10712 [31:54<1:17:39,  2.00it/s]
 13%|█▎        | 1380/10712 [31:54<1:17:14,  2.01it/s]
 13%|█▎        | 1381/10712 [31:55<1:17:21,  2.01it/s]
 13%|█▎        | 1382/10712 [31:55<1:17:13,  2.01it/s]
 13%|█▎        | 1383/10712 [31:56<1:17:32,  2.01it/s]
 13%|█▎        | 1384/10712 [31:56<1:17:52,  2.00it/s]
 13%|█▎        | 1385/10712 [31:57<1:17:31,  2.01it/s]
 13%|█▎        | 1386/10712 [31:57<1:17:29,  2.01it/s]
 13%|█▎        | 1387/10712 [31:58<1:17:13,  2.01it/s]
 13%|█▎        | 1388/10712 [31:58<1:17:07,  2.01it/s]
 13%|█▎        | 1389/10712 [31:59<1:17:16,  2.01it/s]
 13%|█▎        | 1390/10712 [31:59<1:17:26,  2.01it/s]
 13%|█▎        | 1391/10712 [32:00<1:17:37,  2.00it/s]
 13%|█▎        | 1392/10712 [32:00<1:17:09,  2.01it/s]
 13%|█▎        | 1393/10712 [32:01<1:17:13,  2.01it/s]
 13%|█▎        | 1394/10712 [32:01<1:17:09,  2.01it/s]
 13%|█▎        | 1395/10712 [32:02<1:16:58,  2.02it/s]
 13%|█▎        | 1396/10712 [32:02<1:16:47,  2.02it/s]
 13%|█▎        | 1397/10712 [32:03<1:16:50,  2.02it/s]
 13%|█▎        | 1398/10712 [32:03<1:17:01,  2.02it/s]
 13%|█▎        | 1399/10712 [32:04<1:16:48,  2.02it/s]
 13%|█▎        | 1400/10712 [32:04<1:16:41,  2.02it/s]{'loss': 4.2803, 'grad_norm': 0.25510862469673157, 'learning_rate': 0.0009971462244394538, 'epoch': 0.13}

                                                      

 13%|█▎        | 1400/10712 [32:04<1:16:41,  2.02it/s]
 13%|█▎        | 1401/10712 [32:05<1:16:46,  2.02it/s]
 13%|█▎        | 1402/10712 [32:05<1:16:45,  2.02it/s]
 13%|█▎        | 1403/10712 [32:06<1:16:50,  2.02it/s]
 13%|█▎        | 1404/10712 [32:06<1:16:37,  2.02it/s]
 13%|█▎        | 1405/10712 [32:07<1:16:30,  2.03it/s]
 13%|█▎        | 1406/10712 [32:07<1:16:27,  2.03it/s]
 13%|█▎        | 1407/10712 [32:08<1:16:26,  2.03it/s]
 13%|█▎        | 1408/10712 [32:08<1:16:26,  2.03it/s]
 13%|█▎        | 1409/10712 [32:09<1:16:48,  2.02it/s]
 13%|█▎        | 1410/10712 [32:09<1:16:44,  2.02it/s]
 13%|█▎        | 1411/10712 [32:10<1:16:35,  2.02it/s]
 13%|█▎        | 1412/10712 [32:10<1:16:37,  2.02it/s]
 13%|█▎        | 1413/10712 [32:11<1:16:37,  2.02it/s]
 13%|█▎        | 1414/10712 [32:11<1:16:36,  2.02it/s]
 13%|█▎        | 1415/10712 [32:12<1:16:33,  2.02it/s]
 13%|█▎        | 1416/10712 [32:12<1:16:52,  2.02it/s]
 13%|█▎        | 1417/10712 [32:13<1:17:03,  2.01it/s]
 13%|█▎        | 1418/10712 [32:13<1:17:33,  2.00it/s]
 13%|█▎        | 1419/10712 [32:14<1:17:22,  2.00it/s]
 13%|█▎        | 1420/10712 [32:14<1:17:20,  2.00it/s]
 13%|█▎        | 1421/10712 [32:15<1:17:00,  2.01it/s]
 13%|█▎        | 1422/10712 [32:15<1:17:18,  2.00it/s]
 13%|█▎        | 1423/10712 [32:16<1:16:59,  2.01it/s]
 13%|█▎        | 1424/10712 [32:16<1:16:52,  2.01it/s]
 13%|█▎        | 1425/10712 [32:17<1:16:49,  2.01it/s]{'loss': 4.271, 'grad_norm': 0.25871479511260986, 'learning_rate': 0.0009966951170283793, 'epoch': 0.13}

                                                      

 13%|█▎        | 1425/10712 [32:17<1:16:49,  2.01it/s]
 13%|█▎        | 1426/10712 [32:17<1:17:06,  2.01it/s]
 13%|█▎        | 1427/10712 [32:18<1:16:47,  2.02it/s]
 13%|█▎        | 1428/10712 [32:18<1:16:38,  2.02it/s]
 13%|█▎        | 1429/10712 [32:19<1:16:28,  2.02it/s]
 13%|█▎        | 1430/10712 [32:19<1:16:52,  2.01it/s]
 13%|█▎        | 1431/10712 [32:20<1:17:06,  2.01it/s]
 13%|█▎        | 1432/10712 [32:20<1:16:54,  2.01it/s]
 13%|█▎        | 1433/10712 [32:21<1:16:40,  2.02it/s]
 13%|█▎        | 1434/10712 [32:21<1:16:45,  2.01it/s]
 13%|█▎        | 1435/10712 [32:22<1:16:29,  2.02it/s]
 13%|█▎        | 1436/10712 [32:22<1:16:24,  2.02it/s]
 13%|█▎        | 1437/10712 [32:23<1:16:24,  2.02it/s]
 13%|█▎        | 1438/10712 [32:23<1:16:13,  2.03it/s]
 13%|█▎        | 1439/10712 [32:24<1:16:10,  2.03it/s]
 13%|█▎        | 1440/10712 [32:24<1:16:24,  2.02it/s]
 13%|█▎        | 1441/10712 [32:25<1:16:18,  2.02it/s]
 13%|█▎        | 1442/10712 [32:25<1:16:14,  2.03it/s]
 13%|█▎        | 1443/10712 [32:26<1:16:30,  2.02it/s]
 13%|█▎        | 1444/10712 [32:26<1:16:38,  2.02it/s]
 13%|█▎        | 1445/10712 [32:27<1:16:33,  2.02it/s]
 13%|█▎        | 1446/10712 [32:27<1:16:27,  2.02it/s]
 14%|█▎        | 1447/10712 [32:28<1:16:21,  2.02it/s]
 14%|█▎        | 1448/10712 [32:28<1:16:44,  2.01it/s]
 14%|█▎        | 1449/10712 [32:29<1:16:30,  2.02it/s]
 14%|█▎        | 1450/10712 [32:29<1:17:10,  2.00it/s]{'loss': 4.2567, 'grad_norm': 0.24963261187076569, 'learning_rate': 0.000996211040054885, 'epoch': 0.14}
                                                      

 14%|█▎        | 1450/10712 [32:29<1:17:10,  2.00it/s]
 14%|█▎        | 1451/10712 [32:30<1:16:57,  2.01it/s]
 14%|█▎        | 1452/10712 [32:30<1:16:45,  2.01it/s]
 14%|█▎        | 1453/10712 [32:31<1:16:27,  2.02it/s]
 14%|█▎        | 1454/10712 [32:31<1:16:19,  2.02it/s]
 14%|█▎        | 1455/10712 [32:32<1:16:11,  2.02it/s]
 14%|█▎        | 1456/10712 [32:32<1:16:06,  2.03it/s]
 14%|█▎        | 1457/10712 [32:33<1:16:02,  2.03it/s]
 14%|█▎        | 1458/10712 [32:33<1:15:55,  2.03it/s]
 14%|█▎        | 1459/10712 [32:34<1:16:05,  2.03it/s]
 14%|█▎        | 1460/10712 [32:34<1:15:57,  2.03it/s]
 14%|█▎        | 1461/10712 [32:35<1:15:55,  2.03it/s]
 14%|█▎        | 1462/10712 [32:35<1:16:04,  2.03it/s]
 14%|█▎        | 1463/10712 [32:36<1:16:16,  2.02it/s]
 14%|█▎        | 1464/10712 [32:36<1:16:07,  2.02it/s]
 14%|█▎        | 1465/10712 [32:37<1:16:17,  2.02it/s]
 14%|█▎        | 1466/10712 [32:37<1:16:08,  2.02it/s]
 14%|█▎        | 1467/10712 [32:38<1:16:33,  2.01it/s]
 14%|█▎        | 1468/10712 [32:38<1:16:38,  2.01it/s]
 14%|█▎        | 1469/10712 [32:39<1:17:06,  2.00it/s]
 14%|█▎        | 1470/10712 [32:39<1:16:41,  2.01it/s]
 14%|█▎        | 1471/10712 [32:40<1:16:54,  2.00it/s]
 14%|█▎        | 1472/10712 [32:40<1:16:36,  2.01it/s]
 14%|█▍        | 1473/10712 [32:40<1:16:23,  2.02it/s]
 14%|█▍        | 1474/10712 [32:41<1:16:12,  2.02it/s]
 14%|█▍        | 1475/10712 [32:41<1:16:05,  2.02it/s]
                                                      
{'loss': 4.2442, 'grad_norm': 0.2789158821105957, 'learning_rate': 0.0009956940256509682, 'epoch': 0.14}

 14%|█▍        | 1475/10712 [32:41<1:16:05,  2.02it/s]
 14%|█▍        | 1476/10712 [32:42<1:16:04,  2.02it/s]
 14%|█▍        | 1477/10712 [32:42<1:15:52,  2.03it/s]
 14%|█▍        | 1478/10712 [32:43<1:15:55,  2.03it/s]
 14%|█▍        | 1479/10712 [32:43<1:15:47,  2.03it/s]
 14%|█▍        | 1480/10712 [32:44<1:16:03,  2.02it/s]
 14%|█▍        | 1481/10712 [32:44<1:16:00,  2.02it/s]
 14%|█▍        | 1482/10712 [32:45<1:16:04,  2.02it/s]
 14%|█▍        | 1483/10712 [32:45<1:16:01,  2.02it/s]
 14%|█▍        | 1484/10712 [32:46<1:16:12,  2.02it/s]
 14%|█▍        | 1485/10712 [32:46<1:16:09,  2.02it/s]
 14%|█▍        | 1486/10712 [32:47<1:16:04,  2.02it/s]
 14%|█▍        | 1487/10712 [32:47<1:15:52,  2.03it/s]
 14%|█▍        | 1488/10712 [32:48<1:15:59,  2.02it/s]
 14%|█▍        | 1489/10712 [32:48<1:16:20,  2.01it/s]
 14%|█▍        | 1490/10712 [32:49<1:16:12,  2.02it/s]
 14%|█▍        | 1491/10712 [32:49<1:16:10,  2.02it/s]
 14%|█▍        | 1492/10712 [32:50<1:15:55,  2.02it/s]
 14%|█▍        | 1493/10712 [32:50<1:15:59,  2.02it/s]
 14%|█▍        | 1494/10712 [32:51<1:16:10,  2.02it/s]
 14%|█▍        | 1495/10712 [32:51<1:16:13,  2.02it/s]
 14%|█▍        | 1496/10712 [32:52<1:16:59,  2.00it/s]
 14%|█▍        | 1497/10712 [32:52<1:16:34,  2.01it/s]
 14%|█▍        | 1498/10712 [32:53<1:16:20,  2.01it/s]
 14%|█▍        | 1499/10712 [32:53<1:16:14,  2.01it/s]
 14%|█▍        | 1500/10712 [32:54<1:16:20,  2.01it/s]{'loss': 4.2422, 'grad_norm': 0.2738645076751709, 'learning_rate': 0.0009951441081349414, 'epoch': 0.14}
                                                      

 14%|█▍        | 1500/10712 [32:54<1:16:20,  2.01it/s]
 14%|█▍        | 1501/10712 [32:54<1:16:21,  2.01it/s]
 14%|█▍        | 1502/10712 [32:55<1:16:08,  2.02it/s]
 14%|█▍        | 1503/10712 [32:55<1:16:07,  2.02it/s]
 14%|█▍        | 1504/10712 [32:56<1:15:54,  2.02it/s]
 14%|█▍        | 1505/10712 [32:56<1:15:47,  2.02it/s]
 14%|█▍        | 1506/10712 [32:57<1:15:46,  2.02it/s]
 14%|█▍        | 1507/10712 [32:57<1:15:48,  2.02it/s]
 14%|█▍        | 1508/10712 [32:58<1:16:01,  2.02it/s]
 14%|█▍        | 1509/10712 [32:58<1:16:00,  2.02it/s]
 14%|█▍        | 1510/10712 [32:59<1:15:49,  2.02it/s]
 14%|█▍        | 1511/10712 [32:59<1:15:45,  2.02it/s]
 14%|█▍        | 1512/10712 [33:00<1:15:36,  2.03it/s]
 14%|█▍        | 1513/10712 [33:00<1:15:35,  2.03it/s]
 14%|█▍        | 1514/10712 [33:01<1:15:30,  2.03it/s]
 14%|█▍        | 1515/10712 [33:01<1:15:31,  2.03it/s]
 14%|█▍        | 1516/10712 [33:02<1:15:29,  2.03it/s]
 14%|█▍        | 1517/10712 [33:02<1:15:28,  2.03it/s]
 14%|█▍        | 1518/10712 [33:03<1:15:29,  2.03it/s]
 14%|█▍        | 1519/10712 [33:03<1:15:36,  2.03it/s]
 14%|█▍        | 1520/10712 [33:04<1:15:32,  2.03it/s]
 14%|█▍        | 1521/10712 [33:04<1:15:36,  2.03it/s]
 14%|█▍        | 1522/10712 [33:05<1:15:52,  2.02it/s]
 14%|█▍        | 1523/10712 [33:05<1:15:51,  2.02it/s]
 14%|█▍        | 1524/10712 [33:06<1:15:46,  2.02it/s]
 14%|█▍        | 1525/10712 [33:06<1:15:39,  2.02it/s]
                                                      
{'loss': 4.2236, 'grad_norm': 0.26533204317092896, 'learning_rate': 0.0009945613240091573, 'epoch': 0.14}

 14%|█▍        | 1525/10712 [33:06<1:15:39,  2.02it/s]
 14%|█▍        | 1526/10712 [33:07<1:15:42,  2.02it/s]
 14%|█▍        | 1527/10712 [33:07<1:15:36,  2.02it/s]
 14%|█▍        | 1528/10712 [33:08<1:15:26,  2.03it/s]
 14%|█▍        | 1529/10712 [33:08<1:15:37,  2.02it/s]
 14%|█▍        | 1530/10712 [33:09<1:16:00,  2.01it/s]
 14%|█▍        | 1531/10712 [33:09<1:15:52,  2.02it/s]
 14%|█▍        | 1532/10712 [33:10<1:16:17,  2.01it/s]
 14%|█▍        | 1533/10712 [33:10<1:16:05,  2.01it/s]
 14%|█▍        | 1534/10712 [33:11<1:15:51,  2.02it/s]
 14%|█▍        | 1535/10712 [33:11<1:16:00,  2.01it/s]
 14%|█▍        | 1536/10712 [33:12<1:15:43,  2.02it/s]
 14%|█▍        | 1537/10712 [33:12<1:15:33,  2.02it/s]
 14%|█▍        | 1538/10712 [33:13<1:15:39,  2.02it/s]
 14%|█▍        | 1539/10712 [33:13<1:15:38,  2.02it/s]
 14%|█▍        | 1540/10712 [33:14<1:15:38,  2.02it/s]
 14%|█▍        | 1541/10712 [33:14<1:15:33,  2.02it/s]
 14%|█▍        | 1542/10712 [33:15<1:15:38,  2.02it/s]
 14%|█▍        | 1543/10712 [33:15<1:15:22,  2.03it/s]
 14%|█▍        | 1544/10712 [33:16<1:15:28,  2.02it/s]
 14%|█▍        | 1545/10712 [33:16<1:15:35,  2.02it/s]
 14%|█▍        | 1546/10712 [33:17<1:15:27,  2.02it/s]
 14%|█▍        | 1547/10712 [33:17<1:15:20,  2.03it/s]
 14%|█▍        | 1548/10712 [33:18<1:15:12,  2.03it/s]
 14%|█▍        | 1549/10712 [33:18<1:15:11,  2.03it/s]
 14%|█▍        | 1550/10712 [33:19<1:15:14,  2.03it/s]{'loss': 4.2157, 'grad_norm': 0.2559494972229004, 'learning_rate': 0.0009939457119575823, 'epoch': 0.14}

                                                      

 14%|█▍        | 1550/10712 [33:19<1:15:14,  2.03it/s]
 14%|█▍        | 1551/10712 [33:19<1:15:44,  2.02it/s]
 14%|█▍        | 1552/10712 [33:20<1:17:43,  1.96it/s]
 14%|█▍        | 1553/10712 [33:20<1:16:48,  1.99it/s]
 15%|█▍        | 1554/10712 [33:21<1:17:26,  1.97it/s]
 15%|█▍        | 1555/10712 [33:21<1:21:49,  1.87it/s]
 15%|█▍        | 1556/10712 [33:22<1:20:56,  1.89it/s]
 15%|█▍        | 1557/10712 [33:22<1:19:12,  1.93it/s]
 15%|█▍        | 1558/10712 [33:23<1:17:52,  1.96it/s]
 15%|█▍        | 1559/10712 [33:23<1:17:05,  1.98it/s]
 15%|█▍        | 1560/10712 [33:24<1:20:01,  1.91it/s]
 15%|█▍        | 1561/10712 [33:24<1:18:30,  1.94it/s]
 15%|█▍        | 1562/10712 [33:25<1:18:43,  1.94it/s]
 15%|█▍        | 1563/10712 [33:25<1:18:49,  1.93it/s]
 15%|█▍        | 1564/10712 [33:26<1:20:27,  1.90it/s]
 15%|█▍        | 1565/10712 [33:26<1:22:27,  1.85it/s]
 15%|█▍        | 1566/10712 [33:27<1:21:14,  1.88it/s]
 15%|█▍        | 1567/10712 [33:27<1:20:30,  1.89it/s]
 15%|█▍        | 1568/10712 [33:28<1:18:46,  1.93it/s]
 15%|█▍        | 1569/10712 [33:28<1:18:53,  1.93it/s]
 15%|█▍        | 1570/10712 [33:29<1:21:04,  1.88it/s]
 15%|█▍        | 1571/10712 [33:30<1:25:54,  1.77it/s]
 15%|█▍        | 1572/10712 [33:30<1:23:39,  1.82it/s]
 15%|█▍        | 1573/10712 [33:31<1:22:12,  1.85it/s]
 15%|█▍        | 1574/10712 [33:31<1:22:33,  1.84it/s]
 15%|█▍        | 1575/10712 [33:32<1:20:08,  1.90it/s]
                                                      
{'loss': 4.2093, 'grad_norm': 0.24138763546943665, 'learning_rate': 0.000993297312843232, 'epoch': 0.15}

 15%|█▍        | 1575/10712 [33:32<1:20:08,  1.90it/s]
 15%|█▍        | 1576/10712 [33:32<1:19:55,  1.91it/s]
 15%|█▍        | 1577/10712 [33:33<1:19:32,  1.91it/s]
 15%|█▍        | 1578/10712 [33:33<1:20:35,  1.89it/s]
 15%|█▍        | 1579/10712 [33:34<1:22:30,  1.84it/s]
 15%|█▍        | 1580/10712 [33:34<1:20:09,  1.90it/s]
 15%|█▍        | 1581/10712 [33:35<1:18:32,  1.94it/s]
 15%|█▍        | 1582/10712 [33:35<1:22:32,  1.84it/s]
 15%|█▍        | 1583/10712 [33:36<1:24:46,  1.79it/s]
 15%|█▍        | 1584/10712 [33:37<1:24:16,  1.81it/s]
 15%|█▍        | 1585/10712 [33:37<1:21:25,  1.87it/s]
 15%|█▍        | 1586/10712 [33:38<1:20:35,  1.89it/s]
 15%|█▍        | 1587/10712 [33:38<1:20:05,  1.90it/s]
 15%|█▍        | 1588/10712 [33:39<1:20:57,  1.88it/s]
 15%|█▍        | 1589/10712 [33:39<1:22:34,  1.84it/s]
 15%|█▍        | 1590/10712 [33:40<1:20:08,  1.90it/s]
 15%|█▍        | 1591/10712 [33:40<1:20:40,  1.88it/s]
 15%|█▍        | 1592/10712 [33:41<1:21:23,  1.87it/s]
 15%|█▍        | 1593/10712 [33:41<1:20:37,  1.89it/s]
 15%|█▍        | 1594/10712 [33:42<1:21:42,  1.86it/s]
 15%|█▍        | 1595/10712 [33:42<1:22:07,  1.85it/s]
 15%|█▍        | 1596/10712 [33:43<1:21:05,  1.87it/s]
 15%|█▍        | 1597/10712 [33:44<1:23:53,  1.81it/s]
 15%|█▍        | 1598/10712 [33:44<1:21:05,  1.87it/s]
 15%|█▍        | 1599/10712 [33:45<1:20:23,  1.89it/s]
 15%|█▍        | 1600/10712 [33:45<1:21:14,  1.87it/s]{'loss': 4.203, 'grad_norm': 0.25748422741889954, 'learning_rate': 0.0009926161697054558, 'epoch': 0.15}
                                                      

 15%|█▍        | 1600/10712 [33:45<1:21:14,  1.87it/s]
 15%|█▍        | 1601/10712 [33:46<1:22:02,  1.85it/s]
 15%|█▍        | 1602/10712 [33:46<1:19:46,  1.90it/s]
 15%|█▍        | 1603/10712 [33:47<1:20:31,  1.89it/s]
 15%|█▍        | 1604/10712 [33:47<1:19:53,  1.90it/s]
 15%|█▍        | 1605/10712 [33:48<1:19:26,  1.91it/s]
 15%|█▍        | 1606/10712 [33:48<1:19:19,  1.91it/s]
 15%|█▌        | 1607/10712 [33:49<1:19:08,  1.92it/s]
 15%|█▌        | 1608/10712 [33:49<1:20:00,  1.90it/s]
 15%|█▌        | 1609/10712 [33:50<1:18:27,  1.93it/s]
 15%|█▌        | 1610/10712 [33:50<1:18:24,  1.93it/s]
 15%|█▌        | 1611/10712 [33:51<1:22:00,  1.85it/s]
 15%|█▌        | 1612/10712 [33:51<1:19:42,  1.90it/s]
 15%|█▌        | 1613/10712 [33:52<1:19:21,  1.91it/s]
 15%|█▌        | 1614/10712 [33:53<1:23:48,  1.81it/s]
 15%|█▌        | 1615/10712 [33:53<1:20:55,  1.87it/s]
 15%|█▌        | 1616/10712 [33:54<1:18:57,  1.92it/s]
 15%|█▌        | 1617/10712 [33:54<1:20:52,  1.87it/s]
 15%|█▌        | 1618/10712 [33:55<1:20:20,  1.89it/s]
 15%|█▌        | 1619/10712 [33:55<1:18:34,  1.93it/s]
 15%|█▌        | 1620/10712 [33:56<1:18:43,  1.93it/s]
 15%|█▌        | 1621/10712 [33:56<1:20:19,  1.89it/s]
 15%|█▌        | 1622/10712 [33:57<1:22:03,  1.85it/s]
 15%|█▌        | 1623/10712 [33:57<1:21:34,  1.86it/s]
 15%|█▌        | 1624/10712 [33:58<1:24:09,  1.80it/s]
 15%|█▌        | 1625/10712 [33:58<1:21:12,  1.86it/s]{'loss': 4.1875, 'grad_norm': 0.24013467133045197, 'learning_rate': 0.000991902327757083, 'epoch': 0.15}

                                                      

 15%|█▌        | 1625/10712 [33:58<1:21:12,  1.86it/s]
 15%|█▌        | 1626/10712 [33:59<1:19:15,  1.91it/s]
 15%|█▌        | 1627/10712 [33:59<1:17:51,  1.94it/s]
 15%|█▌        | 1628/10712 [34:00<1:18:33,  1.93it/s]
 15%|█▌        | 1629/10712 [34:01<1:22:16,  1.84it/s]
 15%|█▌        | 1630/10712 [34:01<1:19:48,  1.90it/s]
 15%|█▌        | 1631/10712 [34:01<1:18:08,  1.94it/s]
 15%|█▌        | 1632/10712 [34:02<1:18:25,  1.93it/s]
 15%|█▌        | 1633/10712 [34:03<1:19:08,  1.91it/s]
 15%|█▌        | 1634/10712 [34:03<1:17:41,  1.95it/s]
 15%|█▌        | 1635/10712 [34:04<1:20:46,  1.87it/s]
 15%|█▌        | 1636/10712 [34:04<1:21:33,  1.85it/s]
 15%|█▌        | 1637/10712 [34:05<1:21:52,  1.85it/s]
 15%|█▌        | 1638/10712 [34:05<1:20:57,  1.87it/s]
 15%|█▌        | 1639/10712 [34:06<1:20:43,  1.87it/s]
 15%|█▌        | 1640/10712 [34:06<1:20:16,  1.88it/s]
 15%|█▌        | 1641/10712 [34:07<1:19:03,  1.91it/s]
 15%|█▌        | 1642/10712 [34:07<1:17:35,  1.95it/s]
 15%|█▌        | 1643/10712 [34:08<1:16:58,  1.96it/s]
 15%|█▌        | 1644/10712 [34:08<1:16:04,  1.99it/s]
 15%|█▌        | 1645/10712 [34:09<1:16:03,  1.99it/s]
 15%|█▌        | 1646/10712 [34:09<1:15:35,  2.00it/s]
 15%|█▌        | 1647/10712 [34:10<1:15:18,  2.01it/s]
 15%|█▌        | 1648/10712 [34:10<1:14:58,  2.01it/s]
 15%|█▌        | 1649/10712 [34:11<1:15:14,  2.01it/s]
 15%|█▌        | 1650/10712 [34:11<1:15:40,  2.00it/s]{'loss': 4.1847, 'grad_norm': 0.2325085997581482, 'learning_rate': 0.000991155834381419, 'epoch': 0.15}
                                                      

 15%|█▌        | 1650/10712 [34:11<1:15:40,  2.00it/s]
 15%|█▌        | 1651/10712 [34:12<1:15:23,  2.00it/s]
 15%|█▌        | 1652/10712 [34:12<1:15:02,  2.01it/s]
 15%|█▌        | 1653/10712 [34:13<1:17:21,  1.95it/s]
 15%|█▌        | 1654/10712 [34:13<1:18:37,  1.92it/s]
 15%|█▌        | 1655/10712 [34:14<1:17:15,  1.95it/s]
 15%|█▌        | 1656/10712 [34:14<1:19:22,  1.90it/s]
 15%|█▌        | 1657/10712 [34:15<1:24:57,  1.78it/s]
 15%|█▌        | 1658/10712 [34:16<1:23:30,  1.81it/s]
 15%|█▌        | 1659/10712 [34:16<1:22:36,  1.83it/s]
 15%|█▌        | 1660/10712 [34:17<1:19:59,  1.89it/s]
 16%|█▌        | 1661/10712 [34:17<1:20:04,  1.88it/s]
 16%|█▌        | 1662/10712 [34:18<1:20:05,  1.88it/s]
 16%|█▌        | 1663/10712 [34:18<1:21:05,  1.86it/s]
 16%|█▌        | 1664/10712 [34:19<1:20:19,  1.88it/s]
 16%|█▌        | 1665/10712 [34:19<1:23:02,  1.82it/s]
 16%|█▌        | 1666/10712 [34:20<1:20:25,  1.87it/s]
 16%|█▌        | 1667/10712 [34:20<1:20:13,  1.88it/s]
 16%|█▌        | 1668/10712 [34:21<1:20:12,  1.88it/s]
 16%|█▌        | 1669/10712 [34:21<1:19:43,  1.89it/s]
 16%|█▌        | 1670/10712 [34:22<1:18:05,  1.93it/s]
 16%|█▌        | 1671/10712 [34:22<1:20:39,  1.87it/s]
 16%|█▌        | 1672/10712 [34:23<1:18:34,  1.92it/s]
 16%|█▌        | 1673/10712 [34:23<1:17:13,  1.95it/s]
 16%|█▌        | 1674/10712 [34:24<1:19:53,  1.89it/s]
 16%|█▌        | 1675/10712 [34:25<1:19:49,  1.89it/s]{'loss': 4.1749, 'grad_norm': 0.24749316275119781, 'learning_rate': 0.0009903767391291024, 'epoch': 0.16}
                                                      

 16%|█▌        | 1675/10712 [34:25<1:19:49,  1.89it/s]
 16%|█▌        | 1676/10712 [34:25<1:23:22,  1.81it/s]
 16%|█▌        | 1677/10712 [34:26<1:22:26,  1.83it/s]
 16%|█▌        | 1678/10712 [34:26<1:21:16,  1.85it/s]
 16%|█▌        | 1679/10712 [34:27<1:19:04,  1.90it/s]
 16%|█▌        | 1680/10712 [34:27<1:18:43,  1.91it/s]
 16%|█▌        | 1681/10712 [34:28<1:20:20,  1.87it/s]
 16%|█▌        | 1682/10712 [34:28<1:18:29,  1.92it/s]
 16%|█▌        | 1683/10712 [34:29<1:20:42,  1.86it/s]
 16%|█▌        | 1684/10712 [34:29<1:20:07,  1.88it/s]
 16%|█▌        | 1685/10712 [34:30<1:21:09,  1.85it/s]
 16%|█▌        | 1686/10712 [34:31<1:22:31,  1.82it/s]
 16%|█▌        | 1687/10712 [34:31<1:21:27,  1.85it/s]
 16%|█▌        | 1688/10712 [34:32<1:23:37,  1.80it/s]
 16%|█▌        | 1689/10712 [34:32<1:20:39,  1.86it/s]
 16%|█▌        | 1690/10712 [34:33<1:18:38,  1.91it/s]
 16%|█▌        | 1691/10712 [34:33<1:19:07,  1.90it/s]
 16%|█▌        | 1692/10712 [34:34<1:19:23,  1.89it/s]
 16%|█▌        | 1693/10712 [34:34<1:17:43,  1.93it/s]
 16%|█▌        | 1694/10712 [34:35<1:21:40,  1.84it/s]
 16%|█▌        | 1695/10712 [34:35<1:19:17,  1.90it/s]
 16%|█▌        | 1696/10712 [34:36<1:17:41,  1.93it/s]
 16%|█▌        | 1697/10712 [34:36<1:17:54,  1.93it/s]
 16%|█▌        | 1698/10712 [34:37<1:17:57,  1.93it/s]
 16%|█▌        | 1699/10712 [34:37<1:21:43,  1.84it/s]
 16%|█▌        | 1700/10712 [34:38<1:19:21,  1.89it/s]
                                                      
{'loss': 4.1598, 'grad_norm': 0.23928652703762054, 'learning_rate': 0.0009895650937148137, 'epoch': 0.16}

 16%|█▌        | 1700/10712 [34:38<1:19:21,  1.89it/s]
 16%|█▌        | 1701/10712 [34:38<1:20:32,  1.86it/s]
 16%|█▌        | 1702/10712 [34:39<1:20:00,  1.88it/s]
 16%|█▌        | 1703/10712 [34:39<1:19:28,  1.89it/s]
 16%|█▌        | 1704/10712 [34:40<1:20:54,  1.86it/s]
 16%|█▌        | 1705/10712 [34:41<1:18:45,  1.91it/s]
 16%|█▌        | 1706/10712 [34:41<1:20:23,  1.87it/s]
 16%|█▌        | 1707/10712 [34:42<1:19:43,  1.88it/s]
 16%|█▌        | 1708/10712 [34:42<1:23:30,  1.80it/s]
 16%|█▌        | 1709/10712 [34:43<1:22:13,  1.82it/s]
 16%|█▌        | 1710/10712 [34:43<1:21:27,  1.84it/s]
 16%|█▌        | 1711/10712 [34:44<1:20:31,  1.86it/s]
 16%|█▌        | 1712/10712 [34:44<1:18:29,  1.91it/s]
 16%|█▌        | 1713/10712 [34:45<1:20:25,  1.86it/s]
 16%|█▌        | 1714/10712 [34:45<1:19:46,  1.88it/s]
 16%|█▌        | 1715/10712 [34:46<1:19:21,  1.89it/s]
 16%|█▌        | 1716/10712 [34:46<1:17:47,  1.93it/s]
 16%|█▌        | 1717/10712 [34:47<1:18:00,  1.92it/s]
 16%|█▌        | 1718/10712 [34:47<1:18:06,  1.92it/s]
 16%|█▌        | 1719/10712 [34:48<1:16:49,  1.95it/s]
 16%|█▌        | 1720/10712 [34:48<1:15:48,  1.98it/s]
 16%|█▌        | 1721/10712 [34:49<1:19:56,  1.87it/s]
 16%|█▌        | 1722/10712 [34:50<1:18:01,  1.92it/s]
 16%|█▌        | 1723/10712 [34:50<1:19:55,  1.87it/s]
 16%|█▌        | 1724/10712 [34:51<1:21:09,  1.85it/s]
 16%|█▌        | 1725/10712 [34:51<1:18:50,  1.90it/s]
                                                      
{'loss': 4.1499, 'grad_norm': 0.2302962690591812, 'learning_rate': 0.0009887209520138443, 'epoch': 0.16}

 16%|█▌        | 1725/10712 [34:51<1:18:50,  1.90it/s]
 16%|█▌        | 1726/10712 [34:52<1:17:18,  1.94it/s]
 16%|█▌        | 1727/10712 [34:52<1:16:17,  1.96it/s]
 16%|█▌        | 1728/10712 [34:53<1:15:25,  1.99it/s]
 16%|█▌        | 1729/10712 [34:53<1:16:38,  1.95it/s]
 16%|█▌        | 1730/10712 [34:54<1:19:17,  1.89it/s]
 16%|█▌        | 1731/10712 [34:54<1:20:40,  1.86it/s]
 16%|█▌        | 1732/10712 [34:55<1:35:38,  1.56it/s]
 16%|█▌        | 1733/10712 [34:56<1:30:37,  1.65it/s]
 16%|█▌        | 1734/10712 [34:56<1:26:36,  1.73it/s]
 16%|█▌        | 1735/10712 [34:57<1:22:43,  1.81it/s]
 16%|█▌        | 1736/10712 [34:57<1:23:12,  1.80it/s]
 16%|█▌        | 1737/10712 [34:58<1:22:04,  1.82it/s]
 16%|█▌        | 1738/10712 [34:58<1:21:11,  1.84it/s]
 16%|█▌        | 1739/10712 [34:59<1:21:34,  1.83it/s]
 16%|█▌        | 1740/10712 [34:59<1:22:25,  1.81it/s]
 16%|█▋        | 1741/10712 [35:00<1:21:05,  1.84it/s]
 16%|█▋        | 1742/10712 [35:00<1:19:54,  1.87it/s]
 16%|█▋        | 1743/10712 [35:01<1:17:57,  1.92it/s]
 16%|█▋        | 1744/10712 [35:02<1:18:22,  1.91it/s]
 16%|█▋        | 1745/10712 [35:02<1:36:46,  1.54it/s]
 16%|█▋        | 1746/10712 [35:03<1:29:40,  1.67it/s]
 16%|█▋        | 1747/10712 [35:03<1:26:06,  1.74it/s]
 16%|█▋        | 1748/10712 [35:04<1:23:36,  1.79it/s]
 16%|█▋        | 1749/10712 [35:04<1:20:34,  1.85it/s]
 16%|█▋        | 1750/10712 [35:05<1:21:29,  1.83it/s]{'loss': 4.151, 'grad_norm': 0.2818034589290619, 'learning_rate': 0.0009878443700585202, 'epoch': 0.16}
                                                      

 16%|█▋        | 1750/10712 [35:05<1:21:29,  1.83it/s]
 16%|█▋        | 1751/10712 [35:06<1:22:00,  1.82it/s]
 16%|█▋        | 1752/10712 [35:06<1:20:45,  1.85it/s]
 16%|█▋        | 1753/10712 [35:07<1:21:46,  1.83it/s]
 16%|█▋        | 1754/10712 [35:07<1:19:11,  1.89it/s]
 16%|█▋        | 1755/10712 [35:08<1:18:53,  1.89it/s]
 16%|█▋        | 1756/10712 [35:08<1:21:20,  1.83it/s]
 16%|█▋        | 1757/10712 [35:09<1:21:48,  1.82it/s]
 16%|█▋        | 1758/10712 [35:09<1:19:11,  1.88it/s]
 16%|█▋        | 1759/10712 [35:10<1:17:23,  1.93it/s]
 16%|█▋        | 1760/10712 [35:10<1:16:36,  1.95it/s]
 16%|█▋        | 1761/10712 [35:11<1:15:58,  1.96it/s]
 16%|█▋        | 1762/10712 [35:11<1:15:07,  1.99it/s]
 16%|█▋        | 1763/10712 [35:12<1:14:30,  2.00it/s]
 16%|█▋        | 1764/10712 [35:12<1:14:06,  2.01it/s]
 16%|█▋        | 1765/10712 [35:13<1:14:56,  1.99it/s]
 16%|█▋        | 1766/10712 [35:13<1:15:13,  1.98it/s]
 16%|█▋        | 1767/10712 [35:14<1:16:05,  1.96it/s]
 17%|█▋        | 1768/10712 [35:14<1:15:25,  1.98it/s]
 17%|█▋        | 1769/10712 [35:15<1:15:52,  1.96it/s]
 17%|█▋        | 1770/10712 [35:15<1:15:46,  1.97it/s]
 17%|█▋        | 1771/10712 [35:16<1:14:58,  1.99it/s]
 17%|█▋        | 1772/10712 [35:16<1:15:44,  1.97it/s]
 17%|█▋        | 1773/10712 [35:17<1:14:58,  1.99it/s]
 17%|█▋        | 1774/10712 [35:17<1:16:13,  1.95it/s]
 17%|█▋        | 1775/10712 [35:18<1:18:15,  1.90it/s]
                                                      
{'loss': 4.1425, 'grad_norm': 0.23346833884716034, 'learning_rate': 0.0009869354060344812, 'epoch': 0.17}

 17%|█▋        | 1775/10712 [35:18<1:18:15,  1.90it/s]
 17%|█▋        | 1776/10712 [35:19<1:24:11,  1.77it/s]
 17%|█▋        | 1777/10712 [35:19<1:22:10,  1.81it/s]
 17%|█▋        | 1778/10712 [35:20<1:20:47,  1.84it/s]
 17%|█▋        | 1779/10712 [35:20<1:18:33,  1.90it/s]
 17%|█▋        | 1780/10712 [35:21<1:19:16,  1.88it/s]
 17%|█▋        | 1781/10712 [35:21<1:18:46,  1.89it/s]
 17%|█▋        | 1782/10712 [35:22<1:18:29,  1.90it/s]
 17%|█▋        | 1783/10712 [35:22<1:18:19,  1.90it/s]
 17%|█▋        | 1784/10712 [35:23<1:20:12,  1.86it/s]
 17%|█▋        | 1785/10712 [35:23<1:18:06,  1.91it/s]
 17%|█▋        | 1786/10712 [35:24<1:19:22,  1.87it/s]
 17%|█▋        | 1787/10712 [35:24<1:19:22,  1.87it/s]
 17%|█▋        | 1788/10712 [35:25<1:17:25,  1.92it/s]
 17%|█▋        | 1789/10712 [35:25<1:17:49,  1.91it/s]
 17%|█▋        | 1790/10712 [35:26<1:18:14,  1.90it/s]
 17%|█▋        | 1791/10712 [35:27<1:20:07,  1.86it/s]
 17%|█▋        | 1792/10712 [35:27<1:20:49,  1.84it/s]
 17%|█▋        | 1793/10712 [35:28<1:18:29,  1.89it/s]
 17%|█▋        | 1794/10712 [35:28<1:18:36,  1.89it/s]
 17%|█▋        | 1795/10712 [35:29<1:18:17,  1.90it/s]
 17%|█▋        | 1796/10712 [35:29<1:19:14,  1.88it/s]
 17%|█▋        | 1797/10712 [35:30<1:17:19,  1.92it/s]
 17%|█▋        | 1798/10712 [35:30<1:20:45,  1.84it/s]
 17%|█▋        | 1799/10712 [35:31<1:18:25,  1.89it/s]
 17%|█▋        | 1800/10712 [35:31<1:20:59,  1.83it/s]
                                                      {'loss': 4.1382, 'grad_norm': 0.23294688761234283, 'learning_rate': 0.000985994120276821, 'epoch': 0.17}

 17%|█▋        | 1800/10712 [35:31<1:20:59,  1.83it/s]
 17%|█▋        | 1801/10712 [35:32<1:19:59,  1.86it/s]
 17%|█▋        | 1802/10712 [35:32<1:19:44,  1.86it/s]
 17%|█▋        | 1803/10712 [35:33<1:19:23,  1.87it/s]
 17%|█▋        | 1804/10712 [35:34<1:26:45,  1.71it/s]
 17%|█▋        | 1805/10712 [35:34<1:22:35,  1.80it/s]
 17%|█▋        | 1806/10712 [35:35<1:21:07,  1.83it/s]
 17%|█▋        | 1807/10712 [35:35<1:22:59,  1.79it/s]
 17%|█▋        | 1808/10712 [35:36<1:19:52,  1.86it/s]
 17%|█▋        | 1809/10712 [35:36<1:19:29,  1.87it/s]
 17%|█▋        | 1810/10712 [35:37<1:19:18,  1.87it/s]
 17%|█▋        | 1811/10712 [35:37<1:19:08,  1.87it/s]
 17%|█▋        | 1812/10712 [35:38<1:17:14,  1.92it/s]
 17%|█▋        | 1813/10712 [35:38<1:17:34,  1.91it/s]
 17%|█▋        | 1814/10712 [35:39<1:18:48,  1.88it/s]
 17%|█▋        | 1815/10712 [35:39<1:18:44,  1.88it/s]
 17%|█▋        | 1816/10712 [35:40<1:18:27,  1.89it/s]
 17%|█▋        | 1817/10712 [35:41<1:21:59,  1.81it/s]
 17%|█▋        | 1818/10712 [35:41<1:20:50,  1.83it/s]
 17%|█▋        | 1819/10712 [35:42<1:19:39,  1.86it/s]
 17%|█▋        | 1820/10712 [35:42<1:19:29,  1.86it/s]
 17%|█▋        | 1821/10712 [35:43<1:20:12,  1.85it/s]
 17%|█▋        | 1822/10712 [35:43<1:19:03,  1.87it/s]
 17%|█▋        | 1823/10712 [35:44<1:19:54,  1.85it/s]
 17%|█▋        | 1824/10712 [35:44<1:19:13,  1.87it/s]
 17%|█▋        | 1825/10712 [35:45<1:17:13,  1.92it/s]{'loss': 4.1286, 'grad_norm': 0.24082444608211517, 'learning_rate': 0.0009850205752660803, 'epoch': 0.17}
                                                      

 17%|█▋        | 1825/10712 [35:45<1:17:13,  1.92it/s]
 17%|█▋        | 1826/10712 [35:45<1:17:12,  1.92it/s]
 17%|█▋        | 1827/10712 [35:46<1:15:49,  1.95it/s]
 17%|█▋        | 1828/10712 [35:46<1:14:59,  1.97it/s]
 17%|█▋        | 1829/10712 [35:47<1:14:19,  1.99it/s]
 17%|█▋        | 1830/10712 [35:47<1:15:29,  1.96it/s]
 17%|█▋        | 1831/10712 [35:48<1:16:07,  1.94it/s]
 17%|█▋        | 1832/10712 [35:48<1:17:56,  1.90it/s]
 17%|█▋        | 1833/10712 [35:49<1:17:44,  1.90it/s]
 17%|█▋        | 1834/10712 [35:49<1:16:12,  1.94it/s]
 17%|█▋        | 1835/10712 [35:50<1:17:55,  1.90it/s]
 17%|█▋        | 1836/10712 [35:50<1:18:04,  1.89it/s]
 17%|█▋        | 1837/10712 [35:51<1:16:26,  1.93it/s]
 17%|█▋        | 1838/10712 [35:51<1:15:17,  1.96it/s]
 17%|█▋        | 1839/10712 [35:52<1:14:28,  1.99it/s]
 17%|█▋        | 1840/10712 [35:52<1:13:58,  2.00it/s]
 17%|█▋        | 1841/10712 [35:53<1:13:34,  2.01it/s]
 17%|█▋        | 1842/10712 [35:53<1:15:02,  1.97it/s]
 17%|█▋        | 1843/10712 [35:54<1:19:21,  1.86it/s]
 17%|█▋        | 1844/10712 [35:55<1:17:19,  1.91it/s]
 17%|█▋        | 1845/10712 [35:55<1:15:55,  1.95it/s]
 17%|█▋        | 1846/10712 [35:56<1:16:39,  1.93it/s]
 17%|█▋        | 1847/10712 [35:56<1:15:23,  1.96it/s]
 17%|█▋        | 1848/10712 [35:57<1:15:57,  1.94it/s]
 17%|█▋        | 1849/10712 [35:57<1:16:38,  1.93it/s]
 17%|█▋        | 1850/10712 [35:58<1:15:26,  1.96it/s]{'loss': 4.1237, 'grad_norm': 0.2337760627269745, 'learning_rate': 0.0009840148356241001, 'epoch': 0.17}
                                                      

 17%|█▋        | 1850/10712 [35:58<1:15:26,  1.96it/s]
 17%|█▋        | 1851/10712 [35:58<1:14:36,  1.98it/s]
 17%|█▋        | 1852/10712 [35:59<1:15:19,  1.96it/s]
 17%|█▋        | 1853/10712 [35:59<1:19:05,  1.87it/s]
 17%|█▋        | 1854/10712 [36:00<1:17:06,  1.91it/s]
 17%|█▋        | 1855/10712 [36:00<1:15:40,  1.95it/s]
 17%|█▋        | 1856/10712 [36:01<1:14:37,  1.98it/s]
 17%|█▋        | 1857/10712 [36:01<1:13:56,  2.00it/s]
 17%|█▋        | 1858/10712 [36:02<1:13:33,  2.01it/s]
 17%|█▋        | 1859/10712 [36:02<1:13:09,  2.02it/s]
 17%|█▋        | 1860/10712 [36:03<1:17:09,  1.91it/s]
 17%|█▋        | 1861/10712 [36:03<1:15:45,  1.95it/s]
 17%|█▋        | 1862/10712 [36:04<1:14:48,  1.97it/s]
 17%|█▋        | 1863/10712 [36:04<1:18:59,  1.87it/s]
 17%|█▋        | 1864/10712 [36:05<1:16:58,  1.92it/s]
 17%|█▋        | 1865/10712 [36:05<1:18:21,  1.88it/s]
 17%|█▋        | 1866/10712 [36:06<1:16:38,  1.92it/s]
 17%|█▋        | 1867/10712 [36:06<1:15:20,  1.96it/s]
 17%|█▋        | 1868/10712 [36:07<1:17:51,  1.89it/s]
 17%|█▋        | 1869/10712 [36:07<1:20:12,  1.84it/s]
 17%|█▋        | 1870/10712 [36:08<1:19:36,  1.85it/s]
 17%|█▋        | 1871/10712 [36:09<1:17:31,  1.90it/s]
 17%|█▋        | 1872/10712 [36:09<1:17:32,  1.90it/s]
 17%|█▋        | 1873/10712 [36:10<1:15:55,  1.94it/s]
 17%|█▋        | 1874/10712 [36:10<1:14:46,  1.97it/s]
 18%|█▊        | 1875/10712 [36:11<1:16:53,  1.92it/s]{'loss': 4.1256, 'grad_norm': 0.240702822804451, 'learning_rate': 0.0009829769681097328, 'epoch': 0.18}

                                                      

 18%|█▊        | 1875/10712 [36:11<1:16:53,  1.92it/s]
 18%|█▊        | 1876/10712 [36:11<1:17:19,  1.90it/s]
 18%|█▊        | 1877/10712 [36:12<1:18:34,  1.87it/s]
 18%|█▊        | 1878/10712 [36:12<1:16:41,  1.92it/s]
 18%|█▊        | 1879/10712 [36:13<1:15:19,  1.95it/s]
 18%|█▊        | 1880/10712 [36:13<1:15:52,  1.94it/s]
 18%|█▊        | 1881/10712 [36:14<1:14:43,  1.97it/s]
 18%|█▊        | 1882/10712 [36:14<1:14:37,  1.97it/s]
 18%|█▊        | 1883/10712 [36:15<1:14:38,  1.97it/s]
 18%|█▊        | 1884/10712 [36:15<1:13:54,  1.99it/s]
 18%|█▊        | 1885/10712 [36:16<1:15:04,  1.96it/s]
 18%|█▊        | 1886/10712 [36:16<1:19:33,  1.85it/s]
 18%|█▊        | 1887/10712 [36:17<1:19:08,  1.86it/s]
 18%|█▊        | 1888/10712 [36:17<1:17:03,  1.91it/s]
 18%|█▊        | 1889/10712 [36:18<1:15:41,  1.94it/s]
 18%|█▊        | 1890/10712 [36:18<1:14:39,  1.97it/s]
 18%|█▊        | 1891/10712 [36:19<1:17:02,  1.91it/s]
 18%|█▊        | 1892/10712 [36:19<1:15:41,  1.94it/s]
 18%|█▊        | 1893/10712 [36:20<1:17:16,  1.90it/s]
 18%|█▊        | 1894/10712 [36:20<1:16:52,  1.91it/s]
 18%|█▊        | 1895/10712 [36:21<1:16:47,  1.91it/s]
 18%|█▊        | 1896/10712 [36:21<1:16:41,  1.92it/s]
 18%|█▊        | 1897/10712 [36:22<1:16:49,  1.91it/s]
 18%|█▊        | 1898/10712 [36:23<1:17:11,  1.90it/s]
 18%|█▊        | 1899/10712 [36:23<1:19:16,  1.85it/s]
 18%|█▊        | 1900/10712 [36:24<1:17:13,  1.90it/s]
                                                      
{'loss': 4.1178, 'grad_norm': 0.2508513331413269, 'learning_rate': 0.00098190704161441, 'epoch': 0.18}

 18%|█▊        | 1900/10712 [36:24<1:17:13,  1.90it/s]
 18%|█▊        | 1901/10712 [36:24<1:17:09,  1.90it/s]
 18%|█▊        | 1902/10712 [36:25<1:18:23,  1.87it/s]
 18%|█▊        | 1903/10712 [36:25<1:16:25,  1.92it/s]
 18%|█▊        | 1904/10712 [36:26<1:16:28,  1.92it/s]
 18%|█▊        | 1905/10712 [36:26<1:17:43,  1.89it/s]
 18%|█▊        | 1906/10712 [36:27<1:16:04,  1.93it/s]
 18%|█▊        | 1907/10712 [36:27<1:16:34,  1.92it/s]
 18%|█▊        | 1908/10712 [36:28<1:17:55,  1.88it/s]
 18%|█▊        | 1909/10712 [36:28<1:16:38,  1.91it/s]
 18%|█▊        | 1910/10712 [36:29<1:16:59,  1.91it/s]
 18%|█▊        | 1911/10712 [36:29<1:15:31,  1.94it/s]
 18%|█▊        | 1912/10712 [36:30<1:15:02,  1.95it/s]
 18%|█▊        | 1913/10712 [36:30<1:14:45,  1.96it/s]
 18%|█▊        | 1914/10712 [36:31<1:13:59,  1.98it/s]
 18%|█▊        | 1915/10712 [36:31<1:13:56,  1.98it/s]
 18%|█▊        | 1916/10712 [36:32<1:14:01,  1.98it/s]
 18%|█▊        | 1917/10712 [36:32<1:13:51,  1.98it/s]
 18%|█▊        | 1918/10712 [36:33<1:13:11,  2.00it/s]
 18%|█▊        | 1919/10712 [36:33<1:13:22,  2.00it/s]
 18%|█▊        | 1920/10712 [36:34<1:13:33,  1.99it/s]
 18%|█▊        | 1921/10712 [36:34<1:13:12,  2.00it/s]
 18%|█▊        | 1922/10712 [36:35<1:12:48,  2.01it/s]
 18%|█▊        | 1923/10712 [36:35<1:12:59,  2.01it/s]
 18%|█▊        | 1924/10712 [36:36<1:12:47,  2.01it/s]
 18%|█▊        | 1925/10712 [36:36<1:12:56,  2.01it/s]
                                                      {'loss': 4.1057, 'grad_norm': 0.22793975472450256, 'learning_rate': 0.0009808051271575705, 'epoch': 0.18}

 18%|█▊        | 1925/10712 [36:36<1:12:56,  2.01it/s]
 18%|█▊        | 1926/10712 [36:37<1:12:40,  2.01it/s]
 18%|█▊        | 1927/10712 [36:37<1:12:30,  2.02it/s]
 18%|█▊        | 1928/10712 [36:38<1:12:22,  2.02it/s]
 18%|█▊        | 1929/10712 [36:38<1:12:23,  2.02it/s]
 18%|█▊        | 1930/10712 [36:39<1:12:53,  2.01it/s]
 18%|█▊        | 1931/10712 [36:39<1:12:31,  2.02it/s]
 18%|█▊        | 1932/10712 [36:40<1:12:17,  2.02it/s]
 18%|█▊        | 1933/10712 [36:40<1:12:36,  2.01it/s]
 18%|█▊        | 1934/10712 [36:41<1:12:39,  2.01it/s]
 18%|█▊        | 1935/10712 [36:41<1:13:05,  2.00it/s]
 18%|█▊        | 1936/10712 [36:42<1:13:28,  1.99it/s]
 18%|█▊        | 1937/10712 [36:43<1:22:58,  1.76it/s]
 18%|█▊        | 1938/10712 [36:43<1:19:31,  1.84it/s]
 18%|█▊        | 1939/10712 [36:43<1:17:05,  1.90it/s]
 18%|█▊        | 1940/10712 [36:44<1:15:37,  1.93it/s]
 18%|█▊        | 1941/10712 [36:44<1:14:29,  1.96it/s]
 18%|█▊        | 1942/10712 [36:45<1:14:04,  1.97it/s]
 18%|█▊        | 1943/10712 [36:45<1:13:20,  1.99it/s]
 18%|█▊        | 1944/10712 [36:46<1:12:54,  2.00it/s]
 18%|█▊        | 1945/10712 [36:46<1:12:55,  2.00it/s]
 18%|█▊        | 1946/10712 [36:47<1:12:37,  2.01it/s]
 18%|█▊        | 1947/10712 [36:47<1:12:23,  2.02it/s]
 18%|█▊        | 1948/10712 [36:48<1:12:52,  2.00it/s]
 18%|█▊        | 1949/10712 [36:48<1:12:28,  2.01it/s]
 18%|█▊        | 1950/10712 [36:49<1:12:58,  2.00it/s]{'loss': 4.1006, 'grad_norm': 0.22303961217403412, 'learning_rate': 0.0009796712978819458, 'epoch': 0.18}

                                                      

 18%|█▊        | 1950/10712 [36:49<1:12:58,  2.00it/s]
 18%|█▊        | 1951/10712 [36:49<1:14:27,  1.96it/s]
 18%|█▊        | 1952/10712 [36:50<1:14:20,  1.96it/s]
 18%|█▊        | 1953/10712 [36:50<1:13:29,  1.99it/s]
 18%|█▊        | 1954/10712 [36:51<1:13:25,  1.99it/s]
 18%|█▊        | 1955/10712 [36:51<1:12:50,  2.00it/s]
 18%|█▊        | 1956/10712 [36:52<1:12:54,  2.00it/s]
 18%|█▊        | 1957/10712 [36:52<1:12:33,  2.01it/s]
 18%|█▊        | 1958/10712 [36:53<1:13:00,  2.00it/s]
 18%|█▊        | 1959/10712 [36:53<1:12:36,  2.01it/s]
 18%|█▊        | 1960/10712 [36:54<1:12:32,  2.01it/s]
 18%|█▊        | 1961/10712 [36:54<1:12:11,  2.02it/s]
 18%|█▊        | 1962/10712 [36:55<1:12:04,  2.02it/s]
 18%|█▊        | 1963/10712 [36:55<1:11:57,  2.03it/s]
 18%|█▊        | 1964/10712 [36:56<1:11:51,  2.03it/s]
 18%|█▊        | 1965/10712 [36:56<1:11:47,  2.03it/s]
 18%|█▊        | 1966/10712 [36:57<1:12:16,  2.02it/s]
 18%|█▊        | 1967/10712 [36:57<1:12:05,  2.02it/s]
 18%|█▊        | 1968/10712 [36:58<1:12:21,  2.01it/s]
 18%|█▊        | 1969/10712 [36:58<1:12:44,  2.00it/s]
 18%|█▊        | 1970/10712 [36:59<1:12:41,  2.00it/s]
 18%|█▊        | 1971/10712 [36:59<1:12:18,  2.01it/s]
 18%|█▊        | 1972/10712 [37:00<1:12:01,  2.02it/s]
 18%|█▊        | 1973/10712 [37:00<1:11:53,  2.03it/s]
 18%|█▊        | 1974/10712 [37:01<1:12:09,  2.02it/s]
 18%|█▊        | 1975/10712 [37:01<1:11:58,  2.02it/s]
                                                      
{'loss': 4.097, 'grad_norm': 0.2381926327943802, 'learning_rate': 0.0009785056290487041, 'epoch': 0.18}

 18%|█▊        | 1975/10712 [37:01<1:11:58,  2.02it/s]
 18%|█▊        | 1976/10712 [37:02<1:12:20,  2.01it/s]
 18%|█▊        | 1977/10712 [37:02<1:12:03,  2.02it/s]
 18%|█▊        | 1978/10712 [37:03<1:11:51,  2.03it/s]
 18%|█▊        | 1979/10712 [37:03<1:11:49,  2.03it/s]
 18%|█▊        | 1980/10712 [37:04<1:11:40,  2.03it/s]
 18%|█▊        | 1981/10712 [37:04<1:12:11,  2.02it/s]
 19%|█▊        | 1982/10712 [37:05<1:12:53,  2.00it/s]
 19%|█▊        | 1983/10712 [37:05<1:13:38,  1.98it/s]
 19%|█▊        | 1984/10712 [37:06<1:12:56,  1.99it/s]
 19%|█▊        | 1985/10712 [37:06<1:13:10,  1.99it/s]
 19%|█▊        | 1986/10712 [37:07<1:12:37,  2.00it/s]
 19%|█▊        | 1987/10712 [37:07<1:12:13,  2.01it/s]
 19%|█▊        | 1988/10712 [37:08<1:12:35,  2.00it/s]
 19%|█▊        | 1989/10712 [37:08<1:12:12,  2.01it/s]
 19%|█▊        | 1990/10712 [37:09<1:11:56,  2.02it/s]
 19%|█▊        | 1991/10712 [37:09<1:11:47,  2.02it/s]
 19%|█▊        | 1992/10712 [37:10<1:12:06,  2.02it/s]
 19%|█▊        | 1993/10712 [37:10<1:11:50,  2.02it/s]
 19%|█▊        | 1994/10712 [37:11<1:11:41,  2.03it/s]
 19%|█▊        | 1995/10712 [37:11<1:11:38,  2.03it/s]
 19%|█▊        | 1996/10712 [37:12<1:12:58,  1.99it/s]
 19%|█▊        | 1997/10712 [37:12<1:12:26,  2.01it/s]
 19%|█▊        | 1998/10712 [37:13<1:12:49,  1.99it/s]
 19%|█▊        | 1999/10712 [37:13<1:12:23,  2.01it/s]
 19%|█▊        | 2000/10712 [37:14<1:12:03,  2.02it/s]
                                                      {'loss': 4.083, 'grad_norm': 0.23066164553165436, 'learning_rate': 0.000977308198032457, 'epoch': 0.19}


 19%|█▊        | 2000/10712 [37:14<1:12:03,  2.02it/s]
 19%|█▊        | 2001/10712 [37:14<1:12:27,  2.00it/s]
 19%|█▊        | 2002/10712 [37:15<1:12:31,  2.00it/s]
 19%|█▊        | 2003/10712 [37:15<1:12:47,  1.99it/s]
 19%|█▊        | 2004/10712 [37:16<1:13:01,  1.99it/s]
 19%|█▊        | 2005/10712 [37:16<1:12:40,  2.00it/s]
 19%|█▊        | 2006/10712 [37:17<1:12:21,  2.01it/s]
 19%|█▊        | 2007/10712 [37:17<1:12:53,  1.99it/s]
 19%|█▊        | 2008/10712 [37:18<1:12:23,  2.00it/s]
 19%|█▉        | 2009/10712 [37:18<1:12:30,  2.00it/s]
 19%|█▉        | 2010/10712 [37:19<1:12:07,  2.01it/s]
 19%|█▉        | 2011/10712 [37:19<1:11:53,  2.02it/s]
 19%|█▉        | 2012/10712 [37:20<1:11:39,  2.02it/s]
 19%|█▉        | 2013/10712 [37:20<1:12:30,  2.00it/s]
 19%|█▉        | 2014/10712 [37:21<1:12:39,  1.99it/s]
 19%|█▉        | 2015/10712 [37:21<1:12:13,  2.01it/s]
 19%|█▉        | 2016/10712 [37:22<1:13:16,  1.98it/s]
 19%|█▉        | 2017/10712 [37:22<1:13:03,  1.98it/s]
 19%|█▉        | 2018/10712 [37:23<1:12:33,  2.00it/s]
 19%|█▉        | 2019/10712 [37:23<1:12:10,  2.01it/s]
 19%|█▉        | 2020/10712 [37:24<1:11:52,  2.02it/s]
 19%|█▉        | 2021/10712 [37:24<1:11:40,  2.02it/s]
 19%|█▉        | 2022/10712 [37:25<1:11:29,  2.03it/s]
 19%|█▉        | 2023/10712 [37:25<1:11:56,  2.01it/s]
 19%|█▉        | 2024/10712 [37:26<1:12:07,  2.01it/s]
 19%|█▉        | 2025/10712 [37:26<1:12:26,  2.00it/s]
                                                      
{'loss': 4.0819, 'grad_norm': 0.2415730506181717, 'learning_rate': 0.0009760790843161207, 'epoch': 0.19}

 19%|█▉        | 2025/10712 [37:26<1:12:26,  2.00it/s]
 19%|█▉        | 2026/10712 [37:27<1:12:02,  2.01it/s]
 19%|█▉        | 2027/10712 [37:27<1:11:44,  2.02it/s]
 19%|█▉        | 2028/10712 [37:28<1:11:35,  2.02it/s]
 19%|█▉        | 2029/10712 [37:28<1:11:26,  2.03it/s]
 19%|█▉        | 2030/10712 [37:29<1:11:22,  2.03it/s]
 19%|█▉        | 2031/10712 [37:29<1:11:41,  2.02it/s]
 19%|█▉        | 2032/10712 [37:30<1:11:31,  2.02it/s]
 19%|█▉        | 2033/10712 [37:30<1:11:25,  2.03it/s]
 19%|█▉        | 2034/10712 [37:31<1:11:27,  2.02it/s]
 19%|█▉        | 2035/10712 [37:31<1:11:21,  2.03it/s]
 19%|█▉        | 2036/10712 [37:32<1:11:37,  2.02it/s]
 19%|█▉        | 2037/10712 [37:32<1:12:33,  1.99it/s]
 19%|█▉        | 2038/10712 [37:33<1:12:41,  1.99it/s]
 19%|█▉        | 2039/10712 [37:33<1:12:14,  2.00it/s]
 19%|█▉        | 2040/10712 [37:34<1:12:23,  2.00it/s]
 19%|█▉        | 2041/10712 [37:34<1:11:51,  2.01it/s]
 19%|█▉        | 2042/10712 [37:35<1:12:11,  2.00it/s]
 19%|█▉        | 2043/10712 [37:35<1:11:44,  2.01it/s]
 19%|█▉        | 2044/10712 [37:36<1:11:29,  2.02it/s]
 19%|█▉        | 2045/10712 [37:36<1:11:19,  2.03it/s]
 19%|█▉        | 2046/10712 [37:37<1:11:07,  2.03it/s]
 19%|█▉        | 2047/10712 [37:37<1:11:59,  2.01it/s]
 19%|█▉        | 2048/10712 [37:38<1:11:42,  2.01it/s]
 19%|█▉        | 2049/10712 [37:38<1:11:50,  2.01it/s]
 19%|█▉        | 2050/10712 [37:39<1:12:12,  2.00it/s]
                                                      
{'loss': 4.0775, 'grad_norm': 0.23349151015281677, 'learning_rate': 0.0009748183694856421, 'epoch': 0.19}

 19%|█▉        | 2050/10712 [37:39<1:12:12,  2.00it/s]
 19%|█▉        | 2051/10712 [37:39<1:12:40,  1.99it/s]
 19%|█▉        | 2052/10712 [37:40<1:12:07,  2.00it/s]
 19%|█▉        | 2053/10712 [37:40<1:11:48,  2.01it/s]
 19%|█▉        | 2054/10712 [37:41<1:11:28,  2.02it/s]
 19%|█▉        | 2055/10712 [37:41<1:11:34,  2.02it/s]
 19%|█▉        | 2056/10712 [37:42<1:11:18,  2.02it/s]
 19%|█▉        | 2057/10712 [37:42<1:11:26,  2.02it/s]
 19%|█▉        | 2058/10712 [37:43<1:11:58,  2.00it/s]
 19%|█▉        | 2059/10712 [37:43<1:11:37,  2.01it/s]
 19%|█▉        | 2060/10712 [37:44<1:11:41,  2.01it/s]
 19%|█▉        | 2061/10712 [37:44<1:12:33,  1.99it/s]
 19%|█▉        | 2062/10712 [37:45<1:12:00,  2.00it/s]
 19%|█▉        | 2063/10712 [37:45<1:11:37,  2.01it/s]
 19%|█▉        | 2064/10712 [37:46<1:12:04,  2.00it/s]
 19%|█▉        | 2065/10712 [37:46<1:11:39,  2.01it/s]
 19%|█▉        | 2066/10712 [37:47<1:11:22,  2.02it/s]
 19%|█▉        | 2067/10712 [37:47<1:11:12,  2.02it/s]
 19%|█▉        | 2068/10712 [37:48<1:11:50,  2.01it/s]
 19%|█▉        | 2069/10712 [37:48<1:11:35,  2.01it/s]
 19%|█▉        | 2070/10712 [37:49<1:13:11,  1.97it/s]
 19%|█▉        | 2071/10712 [37:49<1:12:31,  1.99it/s]
 19%|█▉        | 2072/10712 [37:50<1:12:00,  2.00it/s]
 19%|█▉        | 2073/10712 [37:50<1:11:36,  2.01it/s]
 19%|█▉        | 2074/10712 [37:51<1:11:16,  2.02it/s]
 19%|█▉        | 2075/10712 [37:51<1:11:32,  2.01it/s]
                                                      
{'loss': 4.0731, 'grad_norm': 0.21858148276805878, 'learning_rate': 0.0009735261372245824, 'epoch': 0.19}

 19%|█▉        | 2075/10712 [37:51<1:11:32,  2.01it/s]
 19%|█▉        | 2076/10712 [37:52<1:11:21,  2.02it/s]
 19%|█▉        | 2077/10712 [37:52<1:11:46,  2.00it/s]
 19%|█▉        | 2078/10712 [37:53<1:11:25,  2.01it/s]
 19%|█▉        | 2079/10712 [37:53<1:11:10,  2.02it/s]
 19%|█▉        | 2080/10712 [37:54<1:11:39,  2.01it/s]
 19%|█▉        | 2081/10712 [37:54<1:11:23,  2.02it/s]
 19%|█▉        | 2082/10712 [37:55<1:11:08,  2.02it/s]
 19%|█▉        | 2083/10712 [37:55<1:11:22,  2.02it/s]
 19%|█▉        | 2084/10712 [37:56<1:11:14,  2.02it/s]
 19%|█▉        | 2085/10712 [37:56<1:11:06,  2.02it/s]
 19%|█▉        | 2086/10712 [37:57<1:10:57,  2.03it/s]
 19%|█▉        | 2087/10712 [37:57<1:12:06,  1.99it/s]
 19%|█▉        | 2088/10712 [37:58<1:11:38,  2.01it/s]
 20%|█▉        | 2089/10712 [37:58<1:11:22,  2.01it/s]
 20%|█▉        | 2090/10712 [37:59<1:11:52,  2.00it/s]
 20%|█▉        | 2091/10712 [37:59<1:11:23,  2.01it/s]
 20%|█▉        | 2092/10712 [38:00<1:11:10,  2.02it/s]
 20%|█▉        | 2093/10712 [38:00<1:11:01,  2.02it/s]
 20%|█▉        | 2094/10712 [38:01<1:11:03,  2.02it/s]
 20%|█▉        | 2095/10712 [38:01<1:10:57,  2.02it/s]
 20%|█▉        | 2096/10712 [38:02<1:10:49,  2.03it/s]
 20%|█▉        | 2097/10712 [38:02<1:10:45,  2.03it/s]
 20%|█▉        | 2098/10712 [38:03<1:10:40,  2.03it/s]
 20%|█▉        | 2099/10712 [38:03<1:11:16,  2.01it/s]
 20%|█▉        | 2100/10712 [38:04<1:11:22,  2.01it/s]{'loss': 4.0734, 'grad_norm': 0.2326219230890274, 'learning_rate': 0.0009722024733085633, 'epoch': 0.2}

                                                      

 20%|█▉        | 2100/10712 [38:04<1:11:22,  2.01it/s]
 20%|█▉        | 2101/10712 [38:04<1:11:46,  2.00it/s]
 20%|█▉        | 2102/10712 [38:05<1:11:45,  2.00it/s]
 20%|█▉        | 2103/10712 [38:05<1:11:26,  2.01it/s]
 20%|█▉        | 2104/10712 [38:06<1:11:52,  2.00it/s]
 20%|█▉        | 2105/10712 [38:06<1:11:22,  2.01it/s]
 20%|█▉        | 2106/10712 [38:07<1:11:06,  2.02it/s]
 20%|█▉        | 2107/10712 [38:07<1:11:39,  2.00it/s]
 20%|█▉        | 2108/10712 [38:08<1:11:13,  2.01it/s]
 20%|█▉        | 2109/10712 [38:08<1:10:57,  2.02it/s]
 20%|█▉        | 2110/10712 [38:09<1:10:52,  2.02it/s]
 20%|█▉        | 2111/10712 [38:09<1:10:44,  2.03it/s]
 20%|█▉        | 2112/10712 [38:10<1:10:37,  2.03it/s]
 20%|█▉        | 2113/10712 [38:10<1:10:35,  2.03it/s]
 20%|█▉        | 2114/10712 [38:11<1:10:32,  2.03it/s]
 20%|█▉        | 2115/10712 [38:11<1:10:29,  2.03it/s]
 20%|█▉        | 2116/10712 [38:11<1:10:29,  2.03it/s]
 20%|█▉        | 2117/10712 [38:12<1:10:35,  2.03it/s]
 20%|█▉        | 2118/10712 [38:12<1:10:30,  2.03it/s]
 20%|█▉        | 2119/10712 [38:13<1:10:30,  2.03it/s]
 20%|█▉        | 2120/10712 [38:13<1:10:30,  2.03it/s]
 20%|█▉        | 2121/10712 [38:14<1:10:28,  2.03it/s]
 20%|█▉        | 2122/10712 [38:14<1:10:26,  2.03it/s]
 20%|█▉        | 2123/10712 [38:15<1:10:27,  2.03it/s]
 20%|█▉        | 2124/10712 [38:15<1:10:25,  2.03it/s]
 20%|█▉        | 2125/10712 [38:16<1:10:22,  2.03it/s]{'loss': 4.0595, 'grad_norm': 0.2197241187095642, 'learning_rate': 0.0009708474655995716, 'epoch': 0.2}

                                                      

 20%|█▉        | 2125/10712 [38:16<1:10:22,  2.03it/s]
 20%|█▉        | 2126/10712 [38:16<1:10:46,  2.02it/s]
 20%|█▉        | 2127/10712 [38:17<1:11:11,  2.01it/s]
 20%|█▉        | 2128/10712 [38:17<1:10:57,  2.02it/s]
 20%|█▉        | 2129/10712 [38:18<1:10:47,  2.02it/s]
 20%|█▉        | 2130/10712 [38:18<1:10:45,  2.02it/s]
 20%|█▉        | 2131/10712 [38:19<1:10:31,  2.03it/s]
 20%|█▉        | 2132/10712 [38:19<1:10:28,  2.03it/s]
 20%|█▉        | 2133/10712 [38:20<1:10:26,  2.03it/s]
 20%|█▉        | 2134/10712 [38:20<1:10:28,  2.03it/s]
 20%|█▉        | 2135/10712 [38:21<1:10:24,  2.03it/s]
 20%|█▉        | 2136/10712 [38:21<1:10:18,  2.03it/s]
 20%|█▉        | 2137/10712 [38:22<1:10:20,  2.03it/s]
 20%|█▉        | 2138/10712 [38:22<1:10:20,  2.03it/s]
 20%|█▉        | 2139/10712 [38:23<1:10:21,  2.03it/s]
 20%|█▉        | 2140/10712 [38:23<1:10:20,  2.03it/s]
 20%|█▉        | 2141/10712 [38:24<1:10:21,  2.03it/s]
 20%|█▉        | 2142/10712 [38:24<1:10:21,  2.03it/s]
 20%|██        | 2143/10712 [38:25<1:10:19,  2.03it/s]
 20%|██        | 2144/10712 [38:25<1:10:15,  2.03it/s]
 20%|██        | 2145/10712 [38:26<1:10:11,  2.03it/s]
 20%|██        | 2146/10712 [38:26<1:10:11,  2.03it/s]
 20%|██        | 2147/10712 [38:27<1:10:09,  2.03it/s]
 20%|██        | 2148/10712 [38:27<1:10:10,  2.03it/s]
 20%|██        | 2149/10712 [38:28<1:10:15,  2.03it/s]
 20%|██        | 2150/10712 [38:28<1:10:09,  2.03it/s]
                                                      
{'loss': 4.0587, 'grad_norm': 0.2229669839143753, 'learning_rate': 0.0009694612040401292, 'epoch': 0.2}

 20%|██        | 2150/10712 [38:28<1:10:09,  2.03it/s]
 20%|██        | 2151/10712 [38:29<1:10:16,  2.03it/s]
 20%|██        | 2152/10712 [38:29<1:10:10,  2.03it/s]
 20%|██        | 2153/10712 [38:30<1:10:02,  2.04it/s]
 20%|██        | 2154/10712 [38:30<1:10:08,  2.03it/s]
 20%|██        | 2155/10712 [38:31<1:10:10,  2.03it/s]
 20%|██        | 2156/10712 [38:31<1:10:06,  2.03it/s]
 20%|██        | 2157/10712 [38:32<1:10:07,  2.03it/s]
 20%|██        | 2158/10712 [38:32<1:10:05,  2.03it/s]
 20%|██        | 2159/10712 [38:33<1:10:02,  2.04it/s]
 20%|██        | 2160/10712 [38:33<1:10:17,  2.03it/s]
 20%|██        | 2161/10712 [38:34<1:10:09,  2.03it/s]
 20%|██        | 2162/10712 [38:34<1:10:07,  2.03it/s]
 20%|██        | 2163/10712 [38:35<1:10:07,  2.03it/s]
 20%|██        | 2164/10712 [38:35<1:10:05,  2.03it/s]
 20%|██        | 2165/10712 [38:36<1:10:10,  2.03it/s]
 20%|██        | 2166/10712 [38:36<1:10:08,  2.03it/s]
 20%|██        | 2167/10712 [38:37<1:10:03,  2.03it/s]
 20%|██        | 2168/10712 [38:37<1:10:00,  2.03it/s]
 20%|██        | 2169/10712 [38:38<1:10:01,  2.03it/s]
 20%|██        | 2170/10712 [38:38<1:10:03,  2.03it/s]
 20%|██        | 2171/10712 [38:39<1:09:55,  2.04it/s]
 20%|██        | 2172/10712 [38:39<1:09:54,  2.04it/s]
 20%|██        | 2173/10712 [38:40<1:10:05,  2.03it/s]
 20%|██        | 2174/10712 [38:40<1:09:55,  2.03it/s]
 20%|██        | 2175/10712 [38:41<1:09:49,  2.04it/s]
                                                      
{'loss': 4.0598, 'grad_norm': 0.21800638735294342, 'learning_rate': 0.0009680437806473211, 'epoch': 0.2}

 20%|██        | 2175/10712 [38:41<1:09:49,  2.04it/s]
 20%|██        | 2176/10712 [38:41<1:09:58,  2.03it/s]
 20%|██        | 2177/10712 [38:42<1:10:18,  2.02it/s]
 20%|██        | 2178/10712 [38:42<1:10:45,  2.01it/s]
 20%|██        | 2179/10712 [38:43<1:10:31,  2.02it/s]
 20%|██        | 2180/10712 [38:43<1:10:16,  2.02it/s]
 20%|██        | 2181/10712 [38:44<1:10:05,  2.03it/s]
 20%|██        | 2182/10712 [38:44<1:10:06,  2.03it/s]
 20%|██        | 2183/10712 [38:44<1:10:01,  2.03it/s]
 20%|██        | 2184/10712 [38:45<1:09:57,  2.03it/s]
 20%|██        | 2185/10712 [38:45<1:10:01,  2.03it/s]
 20%|██        | 2186/10712 [38:46<1:10:00,  2.03it/s]
 20%|██        | 2187/10712 [38:46<1:10:06,  2.03it/s]
 20%|██        | 2188/10712 [38:47<1:09:57,  2.03it/s]
 20%|██        | 2189/10712 [38:47<1:09:52,  2.03it/s]
 20%|██        | 2190/10712 [38:48<1:09:54,  2.03it/s]
 20%|██        | 2191/10712 [38:48<1:09:46,  2.04it/s]
 20%|██        | 2192/10712 [38:49<1:09:55,  2.03it/s]
 20%|██        | 2193/10712 [38:49<1:09:50,  2.03it/s]
 20%|██        | 2194/10712 [38:50<1:09:44,  2.04it/s]
 20%|██        | 2195/10712 [38:50<1:09:47,  2.03it/s]
 21%|██        | 2196/10712 [38:51<1:09:47,  2.03it/s]
 21%|██        | 2197/10712 [38:51<1:09:42,  2.04it/s]
 21%|██        | 2198/10712 [38:52<1:09:53,  2.03it/s]
 21%|██        | 2199/10712 [38:52<1:09:50,  2.03it/s]
 21%|██        | 2200/10712 [38:53<1:09:52,  2.03it/s]
                                                      
{'loss': 4.0551, 'grad_norm': 0.24935771524906158, 'learning_rate': 0.0009665952895066891, 'epoch': 0.21}

 21%|██        | 2200/10712 [38:53<1:09:52,  2.03it/s]
 21%|██        | 2201/10712 [38:53<1:09:56,  2.03it/s]
 21%|██        | 2202/10712 [38:54<1:10:02,  2.03it/s]
 21%|██        | 2203/10712 [38:54<1:10:05,  2.02it/s]
 21%|██        | 2204/10712 [38:55<1:09:57,  2.03it/s]
 21%|██        | 2205/10712 [38:55<1:09:49,  2.03it/s]
 21%|██        | 2206/10712 [38:56<1:09:45,  2.03it/s]
 21%|██        | 2207/10712 [38:56<1:09:43,  2.03it/s]
 21%|██        | 2208/10712 [38:57<1:09:40,  2.03it/s]
 21%|██        | 2209/10712 [38:57<1:09:44,  2.03it/s]
 21%|██        | 2210/10712 [38:58<1:10:05,  2.02it/s]
 21%|██        | 2211/10712 [38:58<1:09:53,  2.03it/s]
 21%|██        | 2212/10712 [38:59<1:09:49,  2.03it/s]
 21%|██        | 2213/10712 [38:59<1:10:35,  2.01it/s]
 21%|██        | 2214/10712 [39:00<1:10:16,  2.02it/s]
 21%|██        | 2215/10712 [39:00<1:10:27,  2.01it/s]
 21%|██        | 2216/10712 [39:01<1:10:12,  2.02it/s]
 21%|██        | 2217/10712 [39:01<1:10:22,  2.01it/s]
 21%|██        | 2218/10712 [39:02<1:10:08,  2.02it/s]
 21%|██        | 2219/10712 [39:02<1:09:57,  2.02it/s]
 21%|██        | 2220/10712 [39:03<1:10:22,  2.01it/s]
 21%|██        | 2221/10712 [39:03<1:10:09,  2.02it/s]
 21%|██        | 2222/10712 [39:04<1:09:59,  2.02it/s]
 21%|██        | 2223/10712 [39:04<1:10:04,  2.02it/s]
 21%|██        | 2224/10712 [39:05<1:10:28,  2.01it/s]
 21%|██        | 2225/10712 [39:05<1:10:10,  2.02it/s]{'loss': 4.0468, 'grad_norm': 0.2231460064649582, 'learning_rate': 0.0009651158267659854, 'epoch': 0.21}

                                                      

 21%|██        | 2225/10712 [39:05<1:10:10,  2.02it/s]
 21%|██        | 2226/10712 [39:06<1:10:03,  2.02it/s]
 21%|██        | 2227/10712 [39:06<1:09:49,  2.03it/s]
 21%|██        | 2228/10712 [39:07<1:09:57,  2.02it/s]
 21%|██        | 2229/10712 [39:07<1:09:49,  2.03it/s]
 21%|██        | 2230/10712 [39:08<1:09:44,  2.03it/s]
 21%|██        | 2231/10712 [39:08<1:09:34,  2.03it/s]
 21%|██        | 2232/10712 [39:09<1:09:30,  2.03it/s]
 21%|██        | 2233/10712 [39:09<1:09:27,  2.03it/s]
 21%|██        | 2234/10712 [39:10<1:09:21,  2.04it/s]
 21%|██        | 2235/10712 [39:10<1:09:15,  2.04it/s]
 21%|██        | 2236/10712 [39:11<1:09:13,  2.04it/s]
 21%|██        | 2237/10712 [39:11<1:09:14,  2.04it/s]
 21%|██        | 2238/10712 [39:12<1:09:33,  2.03it/s]
 21%|██        | 2239/10712 [39:12<1:09:27,  2.03it/s]
 21%|██        | 2240/10712 [39:13<1:09:27,  2.03it/s]
 21%|██        | 2241/10712 [39:13<1:09:18,  2.04it/s]
 21%|██        | 2242/10712 [39:14<1:09:22,  2.04it/s]
 21%|██        | 2243/10712 [39:14<1:09:40,  2.03it/s]
 21%|██        | 2244/10712 [39:15<1:09:54,  2.02it/s]
 21%|██        | 2245/10712 [39:15<1:09:53,  2.02it/s]
 21%|██        | 2246/10712 [39:16<1:09:44,  2.02it/s]
 21%|██        | 2247/10712 [39:16<1:09:31,  2.03it/s]
 21%|██        | 2248/10712 [39:17<1:09:24,  2.03it/s]
 21%|██        | 2249/10712 [39:17<1:09:17,  2.04it/s]
 21%|██        | 2250/10712 [39:18<1:09:20,  2.03it/s]
                                                      
{'loss': 4.0418, 'grad_norm': 0.22651419043540955, 'learning_rate': 0.0009636054906287911, 'epoch': 0.21}

 21%|██        | 2250/10712 [39:18<1:09:20,  2.03it/s]
 21%|██        | 2251/10712 [39:18<1:09:20,  2.03it/s]
 21%|██        | 2252/10712 [39:19<1:09:15,  2.04it/s]
 21%|██        | 2253/10712 [39:19<1:09:14,  2.04it/s]
 21%|██        | 2254/10712 [39:20<1:09:37,  2.02it/s]
 21%|██        | 2255/10712 [39:20<1:09:32,  2.03it/s]
 21%|██        | 2256/10712 [39:20<1:09:26,  2.03it/s]
 21%|██        | 2257/10712 [39:21<1:09:41,  2.02it/s]
 21%|██        | 2258/10712 [39:21<1:09:31,  2.03it/s]
 21%|██        | 2259/10712 [39:22<1:09:43,  2.02it/s]
 21%|██        | 2260/10712 [39:22<1:09:51,  2.02it/s]
 21%|██        | 2261/10712 [39:23<1:10:01,  2.01it/s]
 21%|██        | 2262/10712 [39:23<1:09:50,  2.02it/s]
 21%|██        | 2263/10712 [39:24<1:09:39,  2.02it/s]
 21%|██        | 2264/10712 [39:24<1:09:22,  2.03it/s]
 21%|██        | 2265/10712 [39:25<1:09:16,  2.03it/s]
 21%|██        | 2266/10712 [39:25<1:09:16,  2.03it/s]
 21%|██        | 2267/10712 [39:26<1:09:22,  2.03it/s]
 21%|██        | 2268/10712 [39:26<1:09:14,  2.03it/s]
 21%|██        | 2269/10712 [39:27<1:09:22,  2.03it/s]
 21%|██        | 2270/10712 [39:27<1:09:11,  2.03it/s]
 21%|██        | 2271/10712 [39:28<1:09:06,  2.04it/s]
 21%|██        | 2272/10712 [39:28<1:09:02,  2.04it/s]
 21%|██        | 2273/10712 [39:29<1:09:05,  2.04it/s]
 21%|██        | 2274/10712 [39:29<1:09:07,  2.03it/s]
 21%|██        | 2275/10712 [39:30<1:09:07,  2.03it/s]
                                                      
{'loss': 4.0431, 'grad_norm': 0.2332756519317627, 'learning_rate': 0.0009620643813479973, 'epoch': 0.21}

 21%|██        | 2275/10712 [39:30<1:09:07,  2.03it/s]
 21%|██        | 2276/10712 [39:30<1:09:07,  2.03it/s]
 21%|██▏       | 2277/10712 [39:31<1:09:04,  2.04it/s]
 21%|██▏       | 2278/10712 [39:31<1:09:18,  2.03it/s]
 21%|██▏       | 2279/10712 [39:32<1:09:18,  2.03it/s]
 21%|██▏       | 2280/10712 [39:32<1:09:14,  2.03it/s]
 21%|██▏       | 2281/10712 [39:33<1:09:08,  2.03it/s]
 21%|██▏       | 2282/10712 [39:33<1:09:05,  2.03it/s]
 21%|██▏       | 2283/10712 [39:34<1:09:25,  2.02it/s]
 21%|██▏       | 2284/10712 [39:34<1:09:14,  2.03it/s]
 21%|██▏       | 2285/10712 [39:35<1:09:09,  2.03it/s]
 21%|██▏       | 2286/10712 [39:35<1:09:03,  2.03it/s]
 21%|██▏       | 2287/10712 [39:36<1:09:02,  2.03it/s]
 21%|██▏       | 2288/10712 [39:36<1:09:01,  2.03it/s]
 21%|██▏       | 2289/10712 [39:37<1:09:01,  2.03it/s]
 21%|██▏       | 2290/10712 [39:37<1:08:58,  2.04it/s]
 21%|██▏       | 2291/10712 [39:38<1:08:56,  2.04it/s]
 21%|██▏       | 2292/10712 [39:38<1:09:07,  2.03it/s]
 21%|██▏       | 2293/10712 [39:39<1:09:04,  2.03it/s]
 21%|██▏       | 2294/10712 [39:39<1:09:03,  2.03it/s]
 21%|██▏       | 2295/10712 [39:40<1:08:58,  2.03it/s]
 21%|██▏       | 2296/10712 [39:40<1:08:55,  2.04it/s]
 21%|██▏       | 2297/10712 [39:41<1:08:58,  2.03it/s]
 21%|██▏       | 2298/10712 [39:41<1:09:02,  2.03it/s]
 21%|██▏       | 2299/10712 [39:42<1:09:00,  2.03it/s]
 21%|██▏       | 2300/10712 [39:42<1:09:02,  2.03it/s]
                                                      
{'loss': 4.0361, 'grad_norm': 0.23637498915195465, 'learning_rate': 0.000960492601219151, 'epoch': 0.21}

 21%|██▏       | 2300/10712 [39:42<1:09:02,  2.03it/s]
 21%|██▏       | 2301/10712 [39:43<1:08:58,  2.03it/s]
 21%|██▏       | 2302/10712 [39:43<1:08:52,  2.04it/s]
 21%|██▏       | 2303/10712 [39:44<1:08:58,  2.03it/s]
 22%|██▏       | 2304/10712 [39:44<1:08:53,  2.03it/s]
 22%|██▏       | 2305/10712 [39:45<1:08:49,  2.04it/s]
 22%|██▏       | 2306/10712 [39:45<1:08:58,  2.03it/s]
 22%|██▏       | 2307/10712 [39:46<1:08:58,  2.03it/s]
 22%|██▏       | 2308/10712 [39:46<1:08:51,  2.03it/s]
 22%|██▏       | 2309/10712 [39:47<1:08:51,  2.03it/s]
 22%|██▏       | 2310/10712 [39:47<1:08:57,  2.03it/s]
 22%|██▏       | 2311/10712 [39:48<1:08:53,  2.03it/s]
 22%|██▏       | 2312/10712 [39:48<1:08:50,  2.03it/s]
 22%|██▏       | 2313/10712 [39:49<1:08:50,  2.03it/s]
 22%|██▏       | 2314/10712 [39:49<1:08:48,  2.03it/s]
 22%|██▏       | 2315/10712 [39:50<1:08:55,  2.03it/s]
 22%|██▏       | 2316/10712 [39:50<1:08:49,  2.03it/s]
 22%|██▏       | 2317/10712 [39:51<1:08:48,  2.03it/s]
 22%|██▏       | 2318/10712 [39:51<1:08:46,  2.03it/s]
 22%|██▏       | 2319/10712 [39:52<1:08:47,  2.03it/s]
 22%|██▏       | 2320/10712 [39:52<1:08:44,  2.03it/s]
 22%|██▏       | 2321/10712 [39:52<1:08:50,  2.03it/s]
 22%|██▏       | 2322/10712 [39:53<1:08:51,  2.03it/s]
 22%|██▏       | 2323/10712 [39:53<1:08:49,  2.03it/s]
 22%|██▏       | 2324/10712 [39:54<1:08:47,  2.03it/s]
 22%|██▏       | 2325/10712 [39:54<1:08:43,  2.03it/s]
                                                      
{'loss': 4.0213, 'grad_norm': 0.22799009084701538, 'learning_rate': 0.0009588902545736645, 'epoch': 0.22}

 22%|██▏       | 2325/10712 [39:54<1:08:43,  2.03it/s]
 22%|██▏       | 2326/10712 [39:55<1:08:45,  2.03it/s]
 22%|██▏       | 2327/10712 [39:55<1:08:42,  2.03it/s]
 22%|██▏       | 2328/10712 [39:56<1:08:48,  2.03it/s]
 22%|██▏       | 2329/10712 [39:56<1:08:46,  2.03it/s]
 22%|██▏       | 2330/10712 [39:57<1:08:40,  2.03it/s]
 22%|██▏       | 2331/10712 [39:57<1:08:38,  2.03it/s]
 22%|██▏       | 2332/10712 [39:58<1:08:39,  2.03it/s]
 22%|██▏       | 2333/10712 [39:58<1:08:34,  2.04it/s]
 22%|██▏       | 2334/10712 [39:59<1:08:37,  2.03it/s]
 22%|██▏       | 2335/10712 [39:59<1:08:38,  2.03it/s]
 22%|██▏       | 2336/10712 [40:00<1:08:30,  2.04it/s]
 22%|██▏       | 2337/10712 [40:00<1:08:29,  2.04it/s]
 22%|██▏       | 2338/10712 [40:01<1:08:32,  2.04it/s]
 22%|██▏       | 2339/10712 [40:01<1:08:30,  2.04it/s]
 22%|██▏       | 2340/10712 [40:02<1:08:27,  2.04it/s]
 22%|██▏       | 2341/10712 [40:02<1:08:27,  2.04it/s]
 22%|██▏       | 2342/10712 [40:03<1:08:29,  2.04it/s]
 22%|██▏       | 2343/10712 [40:03<1:08:21,  2.04it/s]
 22%|██▏       | 2344/10712 [40:04<1:08:23,  2.04it/s]
 22%|██▏       | 2345/10712 [40:04<1:08:26,  2.04it/s]
 22%|██▏       | 2346/10712 [40:05<1:08:23,  2.04it/s]
 22%|██▏       | 2347/10712 [40:05<1:08:21,  2.04it/s]
 22%|██▏       | 2348/10712 [40:06<1:08:27,  2.04it/s]
 22%|██▏       | 2349/10712 [40:06<1:08:27,  2.04it/s]
 22%|██▏       | 2350/10712 [40:07<1:08:28,  2.04it/s]
                                                      
{'loss': 4.0259, 'grad_norm': 0.22181713581085205, 'learning_rate': 0.0009572574477718903, 'epoch': 0.22}

 22%|██▏       | 2350/10712 [40:07<1:08:28,  2.04it/s]
 22%|██▏       | 2351/10712 [40:07<1:08:32,  2.03it/s]
 22%|██▏       | 2352/10712 [40:08<1:08:25,  2.04it/s]
 22%|██▏       | 2353/10712 [40:08<1:08:24,  2.04it/s]
 22%|██▏       | 2354/10712 [40:09<1:08:28,  2.03it/s]
 22%|██▏       | 2355/10712 [40:09<1:08:22,  2.04it/s]
 22%|██▏       | 2356/10712 [40:10<1:08:23,  2.04it/s]
 22%|██▏       | 2357/10712 [40:10<1:08:28,  2.03it/s]
 22%|██▏       | 2358/10712 [40:11<1:08:26,  2.03it/s]
 22%|██▏       | 2359/10712 [40:11<1:08:29,  2.03it/s]
 22%|██▏       | 2360/10712 [40:12<1:08:23,  2.04it/s]
 22%|██▏       | 2361/10712 [40:12<1:08:19,  2.04it/s]
 22%|██▏       | 2362/10712 [40:13<1:08:19,  2.04it/s]
 22%|██▏       | 2363/10712 [40:13<1:08:19,  2.04it/s]
 22%|██▏       | 2364/10712 [40:14<1:08:12,  2.04it/s]
 22%|██▏       | 2365/10712 [40:14<1:08:14,  2.04it/s]
 22%|██▏       | 2366/10712 [40:15<1:08:22,  2.03it/s]
 22%|██▏       | 2367/10712 [40:15<1:08:25,  2.03it/s]
 22%|██▏       | 2368/10712 [40:16<1:08:18,  2.04it/s]
 22%|██▏       | 2369/10712 [40:16<1:08:12,  2.04it/s]
 22%|██▏       | 2370/10712 [40:17<1:08:16,  2.04it/s]
 22%|██▏       | 2371/10712 [40:17<1:08:16,  2.04it/s]
 22%|██▏       | 2372/10712 [40:18<1:08:14,  2.04it/s]
 22%|██▏       | 2373/10712 [40:18<1:08:14,  2.04it/s]
 22%|██▏       | 2374/10712 [40:19<1:08:13,  2.04it/s]
 22%|██▏       | 2375/10712 [40:19<1:08:12,  2.04it/s]
                                                      
{'loss': 4.0219, 'grad_norm': 0.2270253300666809, 'learning_rate': 0.0009555942891960611, 'epoch': 0.22}

 22%|██▏       | 2375/10712 [40:19<1:08:12,  2.04it/s]
 22%|██▏       | 2376/10712 [40:20<1:08:22,  2.03it/s]
 22%|██▏       | 2377/10712 [40:20<1:08:18,  2.03it/s]
 22%|██▏       | 2378/10712 [40:20<1:08:15,  2.03it/s]
 22%|██▏       | 2379/10712 [40:21<1:08:14,  2.04it/s]
 22%|██▏       | 2380/10712 [40:21<1:08:12,  2.04it/s]
 22%|██▏       | 2381/10712 [40:22<1:08:06,  2.04it/s]
 22%|██▏       | 2382/10712 [40:22<1:08:03,  2.04it/s]
 22%|██▏       | 2383/10712 [40:23<1:08:07,  2.04it/s]
 22%|██▏       | 2384/10712 [40:23<1:08:03,  2.04it/s]
 22%|██▏       | 2385/10712 [40:24<1:08:02,  2.04it/s]
 22%|██▏       | 2386/10712 [40:24<1:08:02,  2.04it/s]
 22%|██▏       | 2387/10712 [40:25<1:08:03,  2.04it/s]
 22%|██▏       | 2388/10712 [40:25<1:08:10,  2.03it/s]
 22%|██▏       | 2389/10712 [40:26<1:08:07,  2.04it/s]
 22%|██▏       | 2390/10712 [40:26<1:08:06,  2.04it/s]
 22%|██▏       | 2391/10712 [40:27<1:08:01,  2.04it/s]
 22%|██▏       | 2392/10712 [40:27<1:08:02,  2.04it/s]
 22%|██▏       | 2393/10712 [40:28<1:08:03,  2.04it/s]
 22%|██▏       | 2394/10712 [40:28<1:08:06,  2.04it/s]
 22%|██▏       | 2395/10712 [40:29<1:08:02,  2.04it/s]
 22%|██▏       | 2396/10712 [40:29<1:08:01,  2.04it/s]
 22%|██▏       | 2397/10712 [40:30<1:08:07,  2.03it/s]
 22%|██▏       | 2398/10712 [40:30<1:08:03,  2.04it/s]
 22%|██▏       | 2399/10712 [40:31<1:08:01,  2.04it/s]
 22%|██▏       | 2400/10712 [40:31<1:08:07,  2.03it/s]
                                                      
{'loss': 4.017, 'grad_norm': 0.21072500944137573, 'learning_rate': 0.000953900889243096, 'epoch': 0.22}

 22%|██▏       | 2400/10712 [40:31<1:08:07,  2.03it/s]
 22%|██▏       | 2401/10712 [40:32<1:08:09,  2.03it/s]
 22%|██▏       | 2402/10712 [40:32<1:08:12,  2.03it/s]
 22%|██▏       | 2403/10712 [40:33<1:08:04,  2.03it/s]
 22%|██▏       | 2404/10712 [40:33<1:07:56,  2.04it/s]
 22%|██▏       | 2405/10712 [40:34<1:08:46,  2.01it/s]
 22%|██▏       | 2406/10712 [40:34<1:10:43,  1.96it/s]
 22%|██▏       | 2407/10712 [40:35<1:09:55,  1.98it/s]
 22%|██▏       | 2408/10712 [40:35<1:09:15,  2.00it/s]
 22%|██▏       | 2409/10712 [40:36<1:08:45,  2.01it/s]
 22%|██▏       | 2410/10712 [40:36<1:08:34,  2.02it/s]
 23%|██▎       | 2411/10712 [40:37<1:08:20,  2.02it/s]
 23%|██▎       | 2412/10712 [40:37<1:08:09,  2.03it/s]
 23%|██▎       | 2413/10712 [40:38<1:08:06,  2.03it/s]
 23%|██▎       | 2414/10712 [40:38<1:08:18,  2.02it/s]
 23%|██▎       | 2415/10712 [40:39<1:08:11,  2.03it/s]
 23%|██▎       | 2416/10712 [40:39<1:08:19,  2.02it/s]
 23%|██▎       | 2417/10712 [40:40<1:08:12,  2.03it/s]
 23%|██▎       | 2418/10712 [40:40<1:08:25,  2.02it/s]
 23%|██▎       | 2419/10712 [40:41<1:08:16,  2.02it/s]
 23%|██▎       | 2420/10712 [40:41<1:08:11,  2.03it/s]
 23%|██▎       | 2421/10712 [40:42<1:08:02,  2.03it/s]
 23%|██▎       | 2422/10712 [40:42<1:07:56,  2.03it/s]
 23%|██▎       | 2423/10712 [40:43<1:07:58,  2.03it/s]
 23%|██▎       | 2424/10712 [40:43<1:07:51,  2.04it/s]
 23%|██▎       | 2425/10712 [40:44<1:08:01,  2.03it/s]
                                                      
{'loss': 4.0106, 'grad_norm': 0.22406361997127533, 'learning_rate': 0.0009521773603172715, 'epoch': 0.23}

 23%|██▎       | 2425/10712 [40:44<1:08:01,  2.03it/s]
 23%|██▎       | 2426/10712 [40:44<1:08:01,  2.03it/s]
 23%|██▎       | 2427/10712 [40:45<1:08:21,  2.02it/s]
 23%|██▎       | 2428/10712 [40:45<1:08:34,  2.01it/s]
 23%|██▎       | 2429/10712 [40:46<1:08:20,  2.02it/s]
 23%|██▎       | 2430/10712 [40:46<1:08:08,  2.03it/s]
 23%|██▎       | 2431/10712 [40:47<1:08:23,  2.02it/s]
 23%|██▎       | 2432/10712 [40:47<1:08:10,  2.02it/s]
 23%|██▎       | 2433/10712 [40:48<1:08:08,  2.02it/s]
 23%|██▎       | 2434/10712 [40:48<1:08:02,  2.03it/s]
 23%|██▎       | 2435/10712 [40:49<1:08:42,  2.01it/s]
 23%|██▎       | 2436/10712 [40:49<1:08:23,  2.02it/s]
 23%|██▎       | 2437/10712 [40:50<1:08:33,  2.01it/s]
 23%|██▎       | 2438/10712 [40:50<1:08:20,  2.02it/s]
 23%|██▎       | 2439/10712 [40:51<1:08:03,  2.03it/s]
 23%|██▎       | 2440/10712 [40:51<1:08:02,  2.03it/s]
 23%|██▎       | 2441/10712 [40:52<1:19:39,  1.73it/s]
 23%|██▎       | 2442/10712 [40:52<1:15:59,  1.81it/s]
 23%|██▎       | 2443/10712 [40:53<1:13:32,  1.87it/s]
 23%|██▎       | 2444/10712 [40:53<1:11:45,  1.92it/s]
 23%|██▎       | 2445/10712 [40:54<1:10:29,  1.95it/s]
 23%|██▎       | 2446/10712 [40:54<1:09:35,  1.98it/s]
 23%|██▎       | 2447/10712 [40:55<1:08:59,  2.00it/s]
 23%|██▎       | 2448/10712 [40:55<1:08:40,  2.01it/s]
 23%|██▎       | 2449/10712 [40:56<1:08:26,  2.01it/s]
 23%|██▎       | 2450/10712 [40:56<1:08:12,  2.02it/s]
                                                      
{'loss': 4.007, 'grad_norm': 0.22760091722011566, 'learning_rate': 0.0009504238168227613, 'epoch': 0.23}

 23%|██▎       | 2450/10712 [40:56<1:08:12,  2.02it/s]
 23%|██▎       | 2451/10712 [40:57<1:08:43,  2.00it/s]
 23%|██▎       | 2452/10712 [40:57<1:08:25,  2.01it/s]
 23%|██▎       | 2453/10712 [40:58<1:08:39,  2.01it/s]
 23%|██▎       | 2454/10712 [40:58<1:08:11,  2.02it/s]
 23%|██▎       | 2455/10712 [40:59<1:08:34,  2.01it/s]
 23%|██▎       | 2456/10712 [40:59<1:08:13,  2.02it/s]
 23%|██▎       | 2457/10712 [41:00<1:08:03,  2.02it/s]
 23%|██▎       | 2458/10712 [41:00<1:07:54,  2.03it/s]
 23%|██▎       | 2459/10712 [41:01<1:19:00,  1.74it/s]
 23%|██▎       | 2460/10712 [41:02<1:15:32,  1.82it/s]
 23%|██▎       | 2461/10712 [41:02<1:13:22,  1.87it/s]
 23%|██▎       | 2462/10712 [41:02<1:11:31,  1.92it/s]
 23%|██▎       | 2463/10712 [41:03<1:10:41,  1.94it/s]
 23%|██▎       | 2464/10712 [41:03<1:09:42,  1.97it/s]
 23%|██▎       | 2465/10712 [41:04<1:09:05,  1.99it/s]
 23%|██▎       | 2466/10712 [41:04<1:08:34,  2.00it/s]
 23%|██▎       | 2467/10712 [41:05<1:08:12,  2.01it/s]
 23%|██▎       | 2468/10712 [41:05<1:07:53,  2.02it/s]
 23%|██▎       | 2469/10712 [41:06<1:07:55,  2.02it/s]
 23%|██▎       | 2470/10712 [41:06<1:07:42,  2.03it/s]
 23%|██▎       | 2471/10712 [41:07<1:07:34,  2.03it/s]
 23%|██▎       | 2472/10712 [41:07<1:07:34,  2.03it/s]
 23%|██▎       | 2473/10712 [41:08<1:07:32,  2.03it/s]
 23%|██▎       | 2474/10712 [41:08<1:07:24,  2.04it/s]
 23%|██▎       | 2475/10712 [41:09<1:07:26,  2.04it/s]
                                                      
{'loss': 4.0013, 'grad_norm': 0.2278626710176468, 'learning_rate': 0.0009486403751560427, 'epoch': 0.23}

 23%|██▎       | 2475/10712 [41:09<1:07:26,  2.04it/s]
 23%|██▎       | 2476/10712 [41:09<1:07:36,  2.03it/s]
 23%|██▎       | 2477/10712 [41:10<1:07:32,  2.03it/s]
 23%|██▎       | 2478/10712 [41:10<1:07:31,  2.03it/s]
 23%|██▎       | 2479/10712 [41:11<1:07:25,  2.03it/s]
 23%|██▎       | 2480/10712 [41:11<1:07:19,  2.04it/s]
 23%|██▎       | 2481/10712 [41:12<1:07:18,  2.04it/s]
 23%|██▎       | 2482/10712 [41:12<1:07:16,  2.04it/s]
 23%|██▎       | 2483/10712 [41:13<1:07:38,  2.03it/s]
 23%|██▎       | 2484/10712 [41:13<1:07:31,  2.03it/s]
 23%|██▎       | 2485/10712 [41:14<1:07:29,  2.03it/s]
 23%|██▎       | 2486/10712 [41:14<1:07:33,  2.03it/s]
 23%|██▎       | 2487/10712 [41:15<1:07:27,  2.03it/s]
 23%|██▎       | 2488/10712 [41:15<1:07:20,  2.04it/s]
 23%|██▎       | 2489/10712 [41:16<1:07:28,  2.03it/s]
 23%|██▎       | 2490/10712 [41:16<1:07:26,  2.03it/s]
 23%|██▎       | 2491/10712 [41:17<1:07:21,  2.03it/s]
 23%|██▎       | 2492/10712 [41:17<1:07:42,  2.02it/s]
 23%|██▎       | 2493/10712 [41:18<1:08:00,  2.01it/s]
 23%|██▎       | 2494/10712 [41:18<1:07:49,  2.02it/s]
 23%|██▎       | 2495/10712 [41:19<1:07:39,  2.02it/s]
 23%|██▎       | 2496/10712 [41:19<1:07:28,  2.03it/s]
 23%|██▎       | 2497/10712 [41:20<1:07:16,  2.04it/s]
 23%|██▎       | 2498/10712 [41:20<1:07:35,  2.03it/s]
 23%|██▎       | 2499/10712 [41:21<1:07:24,  2.03it/s]
 23%|██▎       | 2500/10712 [41:21<1:07:22,  2.03it/s]
                                                      
{'loss': 3.9986, 'grad_norm': 0.22059397399425507, 'learning_rate': 0.000946827153698169, 'epoch': 0.23}

 23%|██▎       | 2500/10712 [41:21<1:07:22,  2.03it/s]
 23%|██▎       | 2501/10712 [41:22<1:07:42,  2.02it/s]
 23%|██▎       | 2502/10712 [41:22<1:07:33,  2.03it/s]
 23%|██▎       | 2503/10712 [41:23<1:07:22,  2.03it/s]
 23%|██▎       | 2504/10712 [41:23<1:07:17,  2.03it/s]
 23%|██▎       | 2505/10712 [41:24<1:07:59,  2.01it/s]
 23%|██▎       | 2506/10712 [41:24<1:07:43,  2.02it/s]
 23%|██▎       | 2507/10712 [41:25<1:07:24,  2.03it/s]
 23%|██▎       | 2508/10712 [41:25<1:07:16,  2.03it/s]
 23%|██▎       | 2509/10712 [41:26<1:07:39,  2.02it/s]
 23%|██▎       | 2510/10712 [41:26<1:07:35,  2.02it/s]
 23%|██▎       | 2511/10712 [41:27<1:07:45,  2.02it/s]
 23%|██▎       | 2512/10712 [41:27<1:07:29,  2.02it/s]
 23%|██▎       | 2513/10712 [41:28<1:07:36,  2.02it/s]
 23%|██▎       | 2514/10712 [41:28<1:07:23,  2.03it/s]
 23%|██▎       | 2515/10712 [41:29<1:07:14,  2.03it/s]
 23%|██▎       | 2516/10712 [41:29<1:07:09,  2.03it/s]
 23%|██▎       | 2517/10712 [41:30<1:07:42,  2.02it/s]
 24%|██▎       | 2518/10712 [41:30<1:07:55,  2.01it/s]
 24%|██▎       | 2519/10712 [41:31<1:07:47,  2.01it/s]
 24%|██▎       | 2520/10712 [41:31<1:07:31,  2.02it/s]
 24%|██▎       | 2521/10712 [41:32<1:07:42,  2.02it/s]
 24%|██▎       | 2522/10712 [41:32<1:07:25,  2.02it/s]
 24%|██▎       | 2523/10712 [41:33<1:07:38,  2.02it/s]
 24%|██▎       | 2524/10712 [41:33<1:07:19,  2.03it/s]
 24%|██▎       | 2525/10712 [41:34<1:07:06,  2.03it/s]
                                                      
{'loss': 3.9965, 'grad_norm': 0.2269017994403839, 'learning_rate': 0.0009449842728069133, 'epoch': 0.24}

 24%|██▎       | 2525/10712 [41:34<1:07:06,  2.03it/s]
 24%|██▎       | 2526/10712 [41:34<1:07:09,  2.03it/s]
 24%|██▎       | 2527/10712 [41:35<1:07:06,  2.03it/s]
 24%|██▎       | 2528/10712 [41:35<1:07:35,  2.02it/s]
 24%|██▎       | 2529/10712 [41:36<1:07:40,  2.02it/s]
 24%|██▎       | 2530/10712 [41:36<1:07:29,  2.02it/s]
 24%|██▎       | 2531/10712 [41:37<1:07:15,  2.03it/s]
 24%|██▎       | 2532/10712 [41:37<1:07:18,  2.03it/s]
 24%|██▎       | 2533/10712 [41:38<1:07:11,  2.03it/s]
 24%|██▎       | 2534/10712 [41:38<1:07:01,  2.03it/s]
 24%|██▎       | 2535/10712 [41:38<1:07:00,  2.03it/s]
 24%|██▎       | 2536/10712 [41:39<1:06:56,  2.04it/s]
 24%|██▎       | 2537/10712 [41:39<1:07:16,  2.03it/s]
 24%|██▎       | 2538/10712 [41:40<1:07:07,  2.03it/s]
 24%|██▎       | 2539/10712 [41:40<1:07:03,  2.03it/s]
 24%|██▎       | 2540/10712 [41:41<1:07:17,  2.02it/s]
 24%|██▎       | 2541/10712 [41:41<1:07:08,  2.03it/s]
 24%|██▎       | 2542/10712 [41:42<1:07:03,  2.03it/s]
 24%|██▎       | 2543/10712 [41:42<1:06:57,  2.03it/s]
 24%|██▎       | 2544/10712 [41:43<1:06:55,  2.03it/s]
 24%|██▍       | 2545/10712 [41:43<1:06:46,  2.04it/s]
 24%|██▍       | 2546/10712 [41:44<1:06:45,  2.04it/s]
 24%|██▍       | 2547/10712 [41:44<1:06:49,  2.04it/s]
 24%|██▍       | 2548/10712 [41:45<1:06:46,  2.04it/s]
 24%|██▍       | 2549/10712 [41:45<1:06:51,  2.04it/s]
 24%|██▍       | 2550/10712 [41:46<1:06:51,  2.03it/s]
                                                      
{'loss': 3.9943, 'grad_norm': 0.23407050967216492, 'learning_rate': 0.0009431118548087783, 'epoch': 0.24}

 24%|██▍       | 2550/10712 [41:46<1:06:51,  2.03it/s]
 24%|██▍       | 2551/10712 [41:46<1:06:56,  2.03it/s]
 24%|██▍       | 2552/10712 [41:47<1:06:54,  2.03it/s]
 24%|██▍       | 2553/10712 [41:47<1:06:52,  2.03it/s]
 24%|██▍       | 2554/10712 [41:48<1:06:50,  2.03it/s]
 24%|██▍       | 2555/10712 [41:48<1:06:51,  2.03it/s]
 24%|██▍       | 2556/10712 [41:49<1:06:48,  2.03it/s]
 24%|██▍       | 2557/10712 [41:49<1:06:46,  2.04it/s]
 24%|██▍       | 2558/10712 [41:50<1:06:45,  2.04it/s]
 24%|██▍       | 2559/10712 [41:50<1:06:41,  2.04it/s]
 24%|██▍       | 2560/10712 [41:51<1:06:42,  2.04it/s]
 24%|██▍       | 2561/10712 [41:51<1:06:48,  2.03it/s]
 24%|██▍       | 2562/10712 [41:52<1:06:44,  2.04it/s]
 24%|██▍       | 2563/10712 [41:52<1:06:44,  2.03it/s]
 24%|██▍       | 2564/10712 [41:53<1:06:47,  2.03it/s]
 24%|██▍       | 2565/10712 [41:53<1:06:44,  2.03it/s]
 24%|██▍       | 2566/10712 [41:54<1:06:42,  2.04it/s]
 24%|██▍       | 2567/10712 [41:54<1:06:42,  2.04it/s]
 24%|██▍       | 2568/10712 [41:55<1:07:04,  2.02it/s]
 24%|██▍       | 2569/10712 [41:55<1:07:02,  2.02it/s]
 24%|██▍       | 2570/10712 [41:56<1:07:04,  2.02it/s]
 24%|██▍       | 2571/10712 [41:56<1:06:54,  2.03it/s]
 24%|██▍       | 2572/10712 [41:57<1:06:49,  2.03it/s]
 24%|██▍       | 2573/10712 [41:57<1:07:07,  2.02it/s]
 24%|██▍       | 2574/10712 [41:58<1:07:00,  2.02it/s]
 24%|██▍       | 2575/10712 [41:58<1:06:53,  2.03it/s]
                                                      
{'loss': 3.9894, 'grad_norm': 0.20638512074947357, 'learning_rate': 0.0009412100239908767, 'epoch': 0.24}

 24%|██▍       | 2575/10712 [41:58<1:06:53,  2.03it/s]
 24%|██▍       | 2576/10712 [41:59<1:06:51,  2.03it/s]
 24%|██▍       | 2577/10712 [41:59<1:06:47,  2.03it/s]
 24%|██▍       | 2578/10712 [42:00<1:07:07,  2.02it/s]
 24%|██▍       | 2579/10712 [42:00<1:07:15,  2.02it/s]
 24%|██▍       | 2580/10712 [42:01<1:07:00,  2.02it/s]
 24%|██▍       | 2581/10712 [42:01<1:06:51,  2.03it/s]
 24%|██▍       | 2582/10712 [42:02<1:06:48,  2.03it/s]
 24%|██▍       | 2583/10712 [42:02<1:06:42,  2.03it/s]
 24%|██▍       | 2584/10712 [42:03<1:06:58,  2.02it/s]
 24%|██▍       | 2585/10712 [42:03<1:06:49,  2.03it/s]
 24%|██▍       | 2586/10712 [42:04<1:06:47,  2.03it/s]
 24%|██▍       | 2587/10712 [42:04<1:06:46,  2.03it/s]
 24%|██▍       | 2588/10712 [42:05<1:07:23,  2.01it/s]
 24%|██▍       | 2589/10712 [42:05<1:07:10,  2.02it/s]
 24%|██▍       | 2590/10712 [42:06<1:06:53,  2.02it/s]
 24%|██▍       | 2591/10712 [42:06<1:06:48,  2.03it/s]
 24%|██▍       | 2592/10712 [42:07<1:23:29,  1.62it/s]
 24%|██▍       | 2593/10712 [42:07<1:18:25,  1.73it/s]
 24%|██▍       | 2594/10712 [42:08<1:14:48,  1.81it/s]
 24%|██▍       | 2595/10712 [42:08<1:12:16,  1.87it/s]
 24%|██▍       | 2596/10712 [42:09<1:10:33,  1.92it/s]
 24%|██▍       | 2597/10712 [42:09<1:09:20,  1.95it/s]
 24%|██▍       | 2598/10712 [42:10<1:08:25,  1.98it/s]
 24%|██▍       | 2599/10712 [42:10<1:07:49,  1.99it/s]
 24%|██▍       | 2600/10712 [42:11<1:07:39,  2.00it/s]
                                                      
{'loss': 3.9832, 'grad_norm': 0.22970610857009888, 'learning_rate': 0.0009392789065926817, 'epoch': 0.24}

 24%|██▍       | 2600/10712 [42:11<1:07:39,  2.00it/s]
 24%|██▍       | 2601/10712 [42:11<1:07:23,  2.01it/s]
 24%|██▍       | 2602/10712 [42:12<1:07:05,  2.01it/s]
 24%|██▍       | 2603/10712 [42:12<1:06:52,  2.02it/s]
 24%|██▍       | 2604/10712 [42:13<1:07:22,  2.01it/s]
 24%|██▍       | 2605/10712 [42:13<1:07:28,  2.00it/s]
 24%|██▍       | 2606/10712 [42:14<1:07:07,  2.01it/s]
 24%|██▍       | 2607/10712 [42:14<1:06:50,  2.02it/s]
 24%|██▍       | 2608/10712 [42:15<1:06:43,  2.02it/s]
 24%|██▍       | 2609/10712 [42:15<1:06:39,  2.03it/s]
 24%|██▍       | 2610/10712 [42:16<1:06:30,  2.03it/s]
 24%|██▍       | 2611/10712 [42:16<1:06:48,  2.02it/s]
 24%|██▍       | 2612/10712 [42:17<1:06:38,  2.03it/s]
 24%|██▍       | 2613/10712 [42:17<1:06:31,  2.03it/s]
 24%|██▍       | 2614/10712 [42:18<1:06:26,  2.03it/s]
 24%|██▍       | 2615/10712 [42:18<1:06:30,  2.03it/s]
 24%|██▍       | 2616/10712 [42:19<1:06:25,  2.03it/s]
 24%|██▍       | 2617/10712 [42:19<1:06:53,  2.02it/s]
 24%|██▍       | 2618/10712 [42:20<1:06:44,  2.02it/s]
 24%|██▍       | 2619/10712 [42:20<1:06:34,  2.03it/s]
 24%|██▍       | 2620/10712 [42:21<1:06:28,  2.03it/s]
 24%|██▍       | 2621/10712 [42:21<1:06:24,  2.03it/s]
 24%|██▍       | 2622/10712 [42:22<1:06:17,  2.03it/s]
 24%|██▍       | 2623/10712 [42:22<1:06:14,  2.04it/s]
 24%|██▍       | 2624/10712 [42:23<1:06:14,  2.03it/s]
 25%|██▍       | 2625/10712 [42:23<1:06:12,  2.04it/s]
                                                      
{'loss': 3.9851, 'grad_norm': 0.2275850772857666, 'learning_rate': 0.000937318630797647, 'epoch': 0.25}

 25%|██▍       | 2625/10712 [42:23<1:06:12,  2.04it/s]
 25%|██▍       | 2626/10712 [42:24<1:06:16,  2.03it/s]
 25%|██▍       | 2627/10712 [42:24<1:06:50,  2.02it/s]
 25%|██▍       | 2628/10712 [42:25<1:06:39,  2.02it/s]
 25%|██▍       | 2629/10712 [42:25<1:06:31,  2.02it/s]
 25%|██▍       | 2630/10712 [42:26<1:06:26,  2.03it/s]
 25%|██▍       | 2631/10712 [42:26<1:06:23,  2.03it/s]
 25%|██▍       | 2632/10712 [42:27<1:06:17,  2.03it/s]
 25%|██▍       | 2633/10712 [42:27<1:06:10,  2.03it/s]
 25%|██▍       | 2634/10712 [42:28<1:06:05,  2.04it/s]
 25%|██▍       | 2635/10712 [42:28<1:06:13,  2.03it/s]
 25%|██▍       | 2636/10712 [42:29<1:06:13,  2.03it/s]
 25%|██▍       | 2637/10712 [42:29<1:06:19,  2.03it/s]
 25%|██▍       | 2638/10712 [42:30<1:06:19,  2.03it/s]
 25%|██▍       | 2639/10712 [42:30<1:06:17,  2.03it/s]
 25%|██▍       | 2640/10712 [42:31<1:06:18,  2.03it/s]
 25%|██▍       | 2641/10712 [42:31<1:06:12,  2.03it/s]
 25%|██▍       | 2642/10712 [42:32<1:06:10,  2.03it/s]
 25%|██▍       | 2643/10712 [42:32<1:06:22,  2.03it/s]
 25%|██▍       | 2644/10712 [42:33<1:06:16,  2.03it/s]
 25%|██▍       | 2645/10712 [42:33<1:06:11,  2.03it/s]
 25%|██▍       | 2646/10712 [42:34<1:06:07,  2.03it/s]
 25%|██▍       | 2647/10712 [42:34<1:06:06,  2.03it/s]
 25%|██▍       | 2648/10712 [42:35<1:06:02,  2.03it/s]
 25%|██▍       | 2649/10712 [42:35<1:05:55,  2.04it/s]
 25%|██▍       | 2650/10712 [42:36<1:05:57,  2.04it/s]{'loss': 3.9773, 'grad_norm': 0.2196168601512909, 'learning_rate': 0.0009353293267246988, 'epoch': 0.25}

                                                      

 25%|██▍       | 2650/10712 [42:36<1:05:57,  2.04it/s]
 25%|██▍       | 2651/10712 [42:36<1:06:00,  2.04it/s]
 25%|██▍       | 2652/10712 [42:37<1:05:55,  2.04it/s]
 25%|██▍       | 2653/10712 [42:37<1:05:52,  2.04it/s]
 25%|██▍       | 2654/10712 [42:38<1:05:54,  2.04it/s]
 25%|██▍       | 2655/10712 [42:38<1:05:58,  2.04it/s]
 25%|██▍       | 2656/10712 [42:39<1:05:50,  2.04it/s]
 25%|██▍       | 2657/10712 [42:39<1:05:53,  2.04it/s]
 25%|██▍       | 2658/10712 [42:39<1:05:53,  2.04it/s]
 25%|██▍       | 2659/10712 [42:40<1:05:50,  2.04it/s]
 25%|██▍       | 2660/10712 [42:40<1:05:49,  2.04it/s]
 25%|██▍       | 2661/10712 [42:41<1:05:51,  2.04it/s]
 25%|██▍       | 2662/10712 [42:41<1:05:51,  2.04it/s]
 25%|██▍       | 2663/10712 [42:42<1:06:17,  2.02it/s]
 25%|██▍       | 2664/10712 [42:42<1:06:05,  2.03it/s]
 25%|██▍       | 2665/10712 [42:43<1:05:59,  2.03it/s]
 25%|██▍       | 2666/10712 [42:43<1:05:59,  2.03it/s]
 25%|██▍       | 2667/10712 [42:44<1:05:53,  2.04it/s]
 25%|██▍       | 2668/10712 [42:44<1:05:56,  2.03it/s]
 25%|██▍       | 2669/10712 [42:45<1:05:59,  2.03it/s]
 25%|██▍       | 2670/10712 [42:45<1:05:55,  2.03it/s]
 25%|██▍       | 2671/10712 [42:46<1:05:48,  2.04it/s]
 25%|██▍       | 2672/10712 [42:46<1:05:49,  2.04it/s]
 25%|██▍       | 2673/10712 [42:47<1:06:06,  2.03it/s]
 25%|██▍       | 2674/10712 [42:47<1:06:01,  2.03it/s]
 25%|██▍       | 2675/10712 [42:48<1:05:53,  2.03it/s]
                                                      
{'loss': 3.9793, 'grad_norm': 0.225908100605011, 'learning_rate': 0.0009333111264195981, 'epoch': 0.25}

 25%|██▍       | 2675/10712 [42:48<1:05:53,  2.03it/s]
 25%|██▍       | 2676/10712 [42:48<1:05:55,  2.03it/s]
 25%|██▍       | 2677/10712 [42:49<1:05:51,  2.03it/s]
 25%|██▌       | 2678/10712 [42:49<1:05:48,  2.03it/s]
 25%|██▌       | 2679/10712 [42:50<1:05:41,  2.04it/s]
 25%|██▌       | 2680/10712 [42:50<1:06:14,  2.02it/s]
 25%|██▌       | 2681/10712 [42:51<1:06:03,  2.03it/s]
 25%|██▌       | 2682/10712 [42:51<1:06:01,  2.03it/s]
 25%|██▌       | 2683/10712 [42:52<1:05:49,  2.03it/s]
 25%|██▌       | 2684/10712 [42:52<1:05:43,  2.04it/s]
 25%|██▌       | 2685/10712 [42:53<1:05:46,  2.03it/s]
 25%|██▌       | 2686/10712 [42:53<1:05:41,  2.04it/s]
 25%|██▌       | 2687/10712 [42:54<1:05:37,  2.04it/s]
 25%|██▌       | 2688/10712 [42:54<1:05:37,  2.04it/s]
 25%|██▌       | 2689/10712 [42:55<1:05:40,  2.04it/s]
 25%|██▌       | 2690/10712 [42:55<1:05:38,  2.04it/s]
 25%|██▌       | 2691/10712 [42:56<1:06:22,  2.01it/s]
 25%|██▌       | 2692/10712 [42:56<1:06:09,  2.02it/s]
 25%|██▌       | 2693/10712 [42:57<1:06:01,  2.02it/s]
 25%|██▌       | 2694/10712 [42:57<1:05:50,  2.03it/s]
 25%|██▌       | 2695/10712 [42:58<1:05:51,  2.03it/s]
 25%|██▌       | 2696/10712 [42:58<1:05:42,  2.03it/s]
 25%|██▌       | 2697/10712 [42:59<1:05:40,  2.03it/s]
 25%|██▌       | 2698/10712 [42:59<1:05:41,  2.03it/s]
 25%|██▌       | 2699/10712 [43:00<1:05:41,  2.03it/s]
 25%|██▌       | 2700/10712 [43:00<1:05:46,  2.03it/s]
                                                      
{'loss': 3.974, 'grad_norm': 0.22616444528102875, 'learning_rate': 0.0009312641638461767, 'epoch': 0.25}

 25%|██▌       | 2700/10712 [43:00<1:05:46,  2.03it/s]
 25%|██▌       | 2701/10712 [43:01<1:05:49,  2.03it/s]
 25%|██▌       | 2702/10712 [43:01<1:05:41,  2.03it/s]
 25%|██▌       | 2703/10712 [43:02<1:05:41,  2.03it/s]
 25%|██▌       | 2704/10712 [43:02<1:05:38,  2.03it/s]
 25%|██▌       | 2705/10712 [43:03<1:05:35,  2.03it/s]
 25%|██▌       | 2706/10712 [43:03<1:05:36,  2.03it/s]
 25%|██▌       | 2707/10712 [43:04<1:05:43,  2.03it/s]
 25%|██▌       | 2708/10712 [43:04<1:05:36,  2.03it/s]
 25%|██▌       | 2709/10712 [43:05<1:05:36,  2.03it/s]
 25%|██▌       | 2710/10712 [43:05<1:05:38,  2.03it/s]
 25%|██▌       | 2711/10712 [43:06<1:06:07,  2.02it/s]
 25%|██▌       | 2712/10712 [43:06<1:05:55,  2.02it/s]
 25%|██▌       | 2713/10712 [43:07<1:05:45,  2.03it/s]
 25%|██▌       | 2714/10712 [43:07<1:05:37,  2.03it/s]
 25%|██▌       | 2715/10712 [43:08<1:05:46,  2.03it/s]
 25%|██▌       | 2716/10712 [43:08<1:06:01,  2.02it/s]
 25%|██▌       | 2717/10712 [43:09<1:05:54,  2.02it/s]
 25%|██▌       | 2718/10712 [43:09<1:05:41,  2.03it/s]
 25%|██▌       | 2719/10712 [43:10<1:05:40,  2.03it/s]
 25%|██▌       | 2720/10712 [43:10<1:05:37,  2.03it/s]
 25%|██▌       | 2721/10712 [43:11<1:06:07,  2.01it/s]
 25%|██▌       | 2722/10712 [43:11<1:05:50,  2.02it/s]
 25%|██▌       | 2723/10712 [43:12<1:05:44,  2.03it/s]
 25%|██▌       | 2724/10712 [43:12<1:05:46,  2.02it/s]
 25%|██▌       | 2725/10712 [43:13<1:05:37,  2.03it/s]
                                                      
{'loss': 3.9645, 'grad_norm': 0.2399749457836151, 'learning_rate': 0.0009291885748774436, 'epoch': 0.25}

 25%|██▌       | 2725/10712 [43:13<1:05:37,  2.03it/s]
 25%|██▌       | 2726/10712 [43:13<1:05:41,  2.03it/s]
 25%|██▌       | 2727/10712 [43:13<1:05:34,  2.03it/s]
 25%|██▌       | 2728/10712 [43:14<1:05:25,  2.03it/s]
 25%|██▌       | 2729/10712 [43:14<1:05:23,  2.03it/s]
 25%|██▌       | 2730/10712 [43:15<1:05:22,  2.03it/s]
 25%|██▌       | 2731/10712 [43:15<1:05:17,  2.04it/s]
 26%|██▌       | 2732/10712 [43:16<1:05:15,  2.04it/s]
 26%|██▌       | 2733/10712 [43:16<1:05:19,  2.04it/s]
 26%|██▌       | 2734/10712 [43:17<1:05:16,  2.04it/s]
 26%|██▌       | 2735/10712 [43:17<1:05:14,  2.04it/s]
 26%|██▌       | 2736/10712 [43:18<1:05:16,  2.04it/s]
 26%|██▌       | 2737/10712 [43:18<1:05:15,  2.04it/s]
 26%|██▌       | 2738/10712 [43:19<1:05:13,  2.04it/s]
 26%|██▌       | 2739/10712 [43:19<1:05:11,  2.04it/s]
 26%|██▌       | 2740/10712 [43:20<1:05:14,  2.04it/s]
 26%|██▌       | 2741/10712 [43:20<1:05:11,  2.04it/s]
 26%|██▌       | 2742/10712 [43:21<1:05:08,  2.04it/s]
 26%|██▌       | 2743/10712 [43:21<1:05:12,  2.04it/s]
 26%|██▌       | 2744/10712 [43:22<1:05:09,  2.04it/s]
 26%|██▌       | 2745/10712 [43:22<1:05:08,  2.04it/s]
 26%|██▌       | 2746/10712 [43:23<1:05:30,  2.03it/s]
 26%|██▌       | 2747/10712 [43:23<1:05:16,  2.03it/s]
 26%|██▌       | 2748/10712 [43:24<1:05:14,  2.03it/s]
 26%|██▌       | 2749/10712 [43:24<1:05:12,  2.04it/s]
 26%|██▌       | 2750/10712 [43:25<1:05:05,  2.04it/s]
                                                      
{'loss': 3.9675, 'grad_norm': 0.2208396941423416, 'learning_rate': 0.0009270844972865675, 'epoch': 0.26}

 26%|██▌       | 2750/10712 [43:25<1:05:05,  2.04it/s]
 26%|██▌       | 2751/10712 [43:25<1:05:10,  2.04it/s]
 26%|██▌       | 2752/10712 [43:26<1:05:20,  2.03it/s]
 26%|██▌       | 2753/10712 [43:26<1:05:14,  2.03it/s]
 26%|██▌       | 2754/10712 [43:27<1:05:16,  2.03it/s]
 26%|██▌       | 2755/10712 [43:27<1:05:14,  2.03it/s]
 26%|██▌       | 2756/10712 [43:28<1:05:09,  2.04it/s]
 26%|██▌       | 2757/10712 [43:28<1:05:10,  2.03it/s]
 26%|██▌       | 2758/10712 [43:29<1:05:11,  2.03it/s]
 26%|██▌       | 2759/10712 [43:29<1:05:10,  2.03it/s]
 26%|██▌       | 2760/10712 [43:30<1:05:29,  2.02it/s]
 26%|██▌       | 2761/10712 [43:30<1:05:24,  2.03it/s]
 26%|██▌       | 2762/10712 [43:31<1:05:37,  2.02it/s]
 26%|██▌       | 2763/10712 [43:31<1:05:29,  2.02it/s]
 26%|██▌       | 2764/10712 [43:32<1:05:18,  2.03it/s]
 26%|██▌       | 2765/10712 [43:32<1:05:11,  2.03it/s]
 26%|██▌       | 2766/10712 [43:33<1:05:07,  2.03it/s]
 26%|██▌       | 2767/10712 [43:33<1:05:03,  2.04it/s]
 26%|██▌       | 2768/10712 [43:34<1:04:58,  2.04it/s]
 26%|██▌       | 2769/10712 [43:34<1:04:56,  2.04it/s]
 26%|██▌       | 2770/10712 [43:35<1:05:04,  2.03it/s]
 26%|██▌       | 2771/10712 [43:35<1:05:04,  2.03it/s]
 26%|██▌       | 2772/10712 [43:36<1:05:23,  2.02it/s]
 26%|██▌       | 2773/10712 [43:36<1:05:11,  2.03it/s]
 26%|██▌       | 2774/10712 [43:37<1:05:07,  2.03it/s]
 26%|██▌       | 2775/10712 [43:37<1:05:09,  2.03it/s]
                                                      
{'loss': 3.9645, 'grad_norm': 0.22441968321800232, 'learning_rate': 0.0009249520707377309, 'epoch': 0.26}

 26%|██▌       | 2775/10712 [43:37<1:05:09,  2.03it/s]
 26%|██▌       | 2776/10712 [43:38<1:05:07,  2.03it/s]
 26%|██▌       | 2777/10712 [43:38<1:05:06,  2.03it/s]
 26%|██▌       | 2778/10712 [43:39<1:05:08,  2.03it/s]
 26%|██▌       | 2779/10712 [43:39<1:05:04,  2.03it/s]
 26%|██▌       | 2780/10712 [43:40<1:05:03,  2.03it/s]
 26%|██▌       | 2781/10712 [43:40<1:04:59,  2.03it/s]
 26%|██▌       | 2782/10712 [43:41<1:04:57,  2.03it/s]
 26%|██▌       | 2783/10712 [43:41<1:04:56,  2.03it/s]
 26%|██▌       | 2784/10712 [43:42<1:04:50,  2.04it/s]
 26%|██▌       | 2785/10712 [43:42<1:04:54,  2.04it/s]
 26%|██▌       | 2786/10712 [43:42<1:04:56,  2.03it/s]
 26%|██▌       | 2787/10712 [43:43<1:04:53,  2.04it/s]
 26%|██▌       | 2788/10712 [43:43<1:04:54,  2.03it/s]
 26%|██▌       | 2789/10712 [43:44<1:04:54,  2.03it/s]
 26%|██▌       | 2790/10712 [43:44<1:04:54,  2.03it/s]
 26%|██▌       | 2791/10712 [43:45<1:04:47,  2.04it/s]
 26%|██▌       | 2792/10712 [43:45<1:04:49,  2.04it/s]
 26%|██▌       | 2793/10712 [43:46<1:04:47,  2.04it/s]
 26%|██▌       | 2794/10712 [43:46<1:04:44,  2.04it/s]
 26%|██▌       | 2795/10712 [43:47<1:04:50,  2.03it/s]
 26%|██▌       | 2796/10712 [43:47<1:05:06,  2.03it/s]
 26%|██▌       | 2797/10712 [43:48<1:05:01,  2.03it/s]
 26%|██▌       | 2798/10712 [43:48<1:04:57,  2.03it/s]
 26%|██▌       | 2799/10712 [43:49<1:05:27,  2.01it/s]
 26%|██▌       | 2800/10712 [43:49<1:05:10,  2.02it/s]
                                                      
{'loss': 3.9608, 'grad_norm': 0.23178760707378387, 'learning_rate': 0.0009227914367768598, 'epoch': 0.26}

 26%|██▌       | 2800/10712 [43:49<1:05:10,  2.02it/s]
 26%|██▌       | 2801/10712 [43:50<1:05:06,  2.03it/s]
 26%|██▌       | 2802/10712 [43:50<1:05:09,  2.02it/s]
 26%|██▌       | 2803/10712 [43:51<1:04:58,  2.03it/s]
 26%|██▌       | 2804/10712 [43:51<1:04:55,  2.03it/s]
 26%|██▌       | 2805/10712 [43:52<1:04:59,  2.03it/s]
 26%|██▌       | 2806/10712 [43:52<1:04:52,  2.03it/s]
 26%|██▌       | 2807/10712 [43:53<1:05:12,  2.02it/s]
 26%|██▌       | 2808/10712 [43:53<1:05:03,  2.02it/s]
 26%|██▌       | 2809/10712 [43:54<1:04:57,  2.03it/s]
 26%|██▌       | 2810/10712 [43:54<1:04:50,  2.03it/s]
 26%|██▌       | 2811/10712 [43:55<1:08:27,  1.92it/s]
 26%|██▋       | 2812/10712 [43:55<1:07:28,  1.95it/s]
 26%|██▋       | 2813/10712 [43:56<1:06:42,  1.97it/s]
 26%|██▋       | 2814/10712 [43:56<1:06:07,  1.99it/s]
 26%|██▋       | 2815/10712 [43:57<1:05:44,  2.00it/s]
 26%|██▋       | 2816/10712 [43:57<1:05:19,  2.01it/s]
 26%|██▋       | 2817/10712 [43:58<1:05:08,  2.02it/s]
 26%|██▋       | 2818/10712 [43:58<1:05:00,  2.02it/s]
 26%|██▋       | 2819/10712 [43:59<1:04:53,  2.03it/s]
 26%|██▋       | 2820/10712 [43:59<1:04:43,  2.03it/s]
 26%|██▋       | 2821/10712 [44:00<1:04:43,  2.03it/s]
 26%|██▋       | 2822/10712 [44:00<1:04:40,  2.03it/s]
 26%|██▋       | 2823/10712 [44:01<1:04:38,  2.03it/s]
 26%|██▋       | 2824/10712 [44:01<1:04:33,  2.04it/s]
 26%|██▋       | 2825/10712 [44:02<1:04:33,  2.04it/s]{'loss': 3.9586, 'grad_norm': 0.20776145160198212, 'learning_rate': 0.0009206027388222276, 'epoch': 0.26}

                                                      

 26%|██▋       | 2825/10712 [44:02<1:04:33,  2.04it/s]
 26%|██▋       | 2826/10712 [44:02<1:04:40,  2.03it/s]
 26%|██▋       | 2827/10712 [44:03<1:04:37,  2.03it/s]
 26%|██▋       | 2828/10712 [44:03<1:04:38,  2.03it/s]
 26%|██▋       | 2829/10712 [44:04<1:04:33,  2.03it/s]
 26%|██▋       | 2830/10712 [44:04<1:04:33,  2.03it/s]
 26%|██▋       | 2831/10712 [44:05<1:04:54,  2.02it/s]
 26%|██▋       | 2832/10712 [44:05<1:04:50,  2.03it/s]
 26%|██▋       | 2833/10712 [44:06<1:04:44,  2.03it/s]
 26%|██▋       | 2834/10712 [44:06<1:04:35,  2.03it/s]
 26%|██▋       | 2835/10712 [44:07<1:04:34,  2.03it/s]
 26%|██▋       | 2836/10712 [44:07<1:04:29,  2.04it/s]
 26%|██▋       | 2837/10712 [44:08<1:04:26,  2.04it/s]
 26%|██▋       | 2838/10712 [44:08<1:04:27,  2.04it/s]
 27%|██▋       | 2839/10712 [44:09<1:04:30,  2.03it/s]
 27%|██▋       | 2840/10712 [44:09<1:04:24,  2.04it/s]
 27%|██▋       | 2841/10712 [44:10<1:04:20,  2.04it/s]
 27%|██▋       | 2842/10712 [44:10<1:04:23,  2.04it/s]
 27%|██▋       | 2843/10712 [44:11<1:04:24,  2.04it/s]
 27%|██▋       | 2844/10712 [44:11<1:04:19,  2.04it/s]
 27%|██▋       | 2845/10712 [44:12<1:04:15,  2.04it/s]
 27%|██▋       | 2846/10712 [44:12<1:04:14,  2.04it/s]
 27%|██▋       | 2847/10712 [44:13<1:04:39,  2.03it/s]
 27%|██▋       | 2848/10712 [44:13<1:04:36,  2.03it/s]
 27%|██▋       | 2849/10712 [44:14<1:04:37,  2.03it/s]
 27%|██▋       | 2850/10712 [44:14<1:04:34,  2.03it/s]
                                                      
{'loss': 3.9577, 'grad_norm': 0.21386730670928955, 'learning_rate': 0.0009183861221549363, 'epoch': 0.27}

 27%|██▋       | 2850/10712 [44:14<1:04:34,  2.03it/s]
 27%|██▋       | 2851/10712 [44:15<1:04:39,  2.03it/s]
 27%|██▋       | 2852/10712 [44:15<1:04:59,  2.02it/s]
 27%|██▋       | 2853/10712 [44:16<1:04:47,  2.02it/s]
 27%|██▋       | 2854/10712 [44:16<1:04:56,  2.02it/s]
 27%|██▋       | 2855/10712 [44:17<1:04:42,  2.02it/s]
 27%|██▋       | 2856/10712 [44:17<1:04:41,  2.02it/s]
 27%|██▋       | 2857/10712 [44:18<1:04:28,  2.03it/s]
 27%|██▋       | 2858/10712 [44:18<1:04:25,  2.03it/s]
 27%|██▋       | 2859/10712 [44:19<1:04:28,  2.03it/s]
 27%|██▋       | 2860/10712 [44:19<1:04:45,  2.02it/s]
 27%|██▋       | 2861/10712 [44:20<1:04:36,  2.03it/s]
 27%|██▋       | 2862/10712 [44:20<1:04:29,  2.03it/s]
 27%|██▋       | 2863/10712 [44:21<1:04:26,  2.03it/s]
 27%|██▋       | 2864/10712 [44:21<1:04:21,  2.03it/s]
 27%|██▋       | 2865/10712 [44:21<1:04:16,  2.03it/s]
 27%|██▋       | 2866/10712 [44:22<1:04:17,  2.03it/s]
 27%|██▋       | 2867/10712 [44:22<1:04:14,  2.04it/s]
 27%|██▋       | 2868/10712 [44:23<1:04:35,  2.02it/s]
 27%|██▋       | 2869/10712 [44:23<1:04:30,  2.03it/s]
 27%|██▋       | 2870/10712 [44:24<1:04:25,  2.03it/s]
 27%|██▋       | 2871/10712 [44:24<1:04:36,  2.02it/s]
 27%|██▋       | 2872/10712 [44:25<1:04:31,  2.02it/s]
 27%|██▋       | 2873/10712 [44:25<1:04:27,  2.03it/s]
 27%|██▋       | 2874/10712 [44:26<1:04:24,  2.03it/s]
 27%|██▋       | 2875/10712 [44:26<1:04:21,  2.03it/s]
                                                      
{'loss': 3.9509, 'grad_norm': 0.24383917450904846, 'learning_rate': 0.0009161417339092721, 'epoch': 0.27}

 27%|██▋       | 2875/10712 [44:26<1:04:21,  2.03it/s]
 27%|██▋       | 2876/10712 [44:27<1:04:28,  2.03it/s]
 27%|██▋       | 2877/10712 [44:27<1:04:20,  2.03it/s]
 27%|██▋       | 2878/10712 [44:28<1:04:17,  2.03it/s]
 27%|██▋       | 2879/10712 [44:28<1:04:11,  2.03it/s]
 27%|██▋       | 2880/10712 [44:29<1:04:07,  2.04it/s]
 27%|██▋       | 2881/10712 [44:29<1:04:13,  2.03it/s]
 27%|██▋       | 2882/10712 [44:30<1:04:08,  2.03it/s]
 27%|██▋       | 2883/10712 [44:30<1:04:04,  2.04it/s]
 27%|██▋       | 2884/10712 [44:31<1:04:25,  2.03it/s]
 27%|██▋       | 2885/10712 [44:31<1:04:20,  2.03it/s]
 27%|██▋       | 2886/10712 [44:32<1:04:16,  2.03it/s]
 27%|██▋       | 2887/10712 [44:32<1:04:11,  2.03it/s]
 27%|██▋       | 2888/10712 [44:33<1:04:11,  2.03it/s]
 27%|██▋       | 2889/10712 [44:33<1:04:05,  2.03it/s]
 27%|██▋       | 2890/10712 [44:34<1:04:02,  2.04it/s]
 27%|██▋       | 2891/10712 [44:34<1:04:09,  2.03it/s]
 27%|██▋       | 2892/10712 [44:35<1:04:05,  2.03it/s]
 27%|██▋       | 2893/10712 [44:35<1:04:03,  2.03it/s]
 27%|██▋       | 2894/10712 [44:36<1:04:04,  2.03it/s]
 27%|██▋       | 2895/10712 [44:36<1:04:31,  2.02it/s]
 27%|██▋       | 2896/10712 [44:37<1:04:24,  2.02it/s]
 27%|██▋       | 2897/10712 [44:37<1:04:13,  2.03it/s]
 27%|██▋       | 2898/10712 [44:38<1:04:06,  2.03it/s]
 27%|██▋       | 2899/10712 [44:38<1:04:03,  2.03it/s]
 27%|██▋       | 2900/10712 [44:39<1:04:01,  2.03it/s]
                                                      
{'loss': 3.9503, 'grad_norm': 0.20805272459983826, 'learning_rate': 0.0009138697230629397, 'epoch': 0.27}

 27%|██▋       | 2900/10712 [44:39<1:04:01,  2.03it/s]
 27%|██▋       | 2901/10712 [44:39<1:04:09,  2.03it/s]
 27%|██▋       | 2902/10712 [44:40<1:04:04,  2.03it/s]
 27%|██▋       | 2903/10712 [44:40<1:03:59,  2.03it/s]
 27%|██▋       | 2904/10712 [44:41<1:03:55,  2.04it/s]
 27%|██▋       | 2905/10712 [44:41<1:04:01,  2.03it/s]
 27%|██▋       | 2906/10712 [44:42<1:04:12,  2.03it/s]
 27%|██▋       | 2907/10712 [44:42<1:04:02,  2.03it/s]
 27%|██▋       | 2908/10712 [44:43<1:04:00,  2.03it/s]
 27%|██▋       | 2909/10712 [44:43<1:03:56,  2.03it/s]
 27%|██▋       | 2910/10712 [44:44<1:03:52,  2.04it/s]
 27%|██▋       | 2911/10712 [44:44<1:03:52,  2.04it/s]
 27%|██▋       | 2912/10712 [44:45<1:03:49,  2.04it/s]
 27%|██▋       | 2913/10712 [44:45<1:04:09,  2.03it/s]
 27%|██▋       | 2914/10712 [44:46<1:04:05,  2.03it/s]
 27%|██▋       | 2915/10712 [44:46<1:04:01,  2.03it/s]
 27%|██▋       | 2916/10712 [44:47<1:03:53,  2.03it/s]
 27%|██▋       | 2917/10712 [44:47<1:03:52,  2.03it/s]
 27%|██▋       | 2918/10712 [44:48<1:03:51,  2.03it/s]
 27%|██▋       | 2919/10712 [44:48<1:03:42,  2.04it/s]
 27%|██▋       | 2920/10712 [44:49<1:03:40,  2.04it/s]
 27%|██▋       | 2921/10712 [44:49<1:03:43,  2.04it/s]
 27%|██▋       | 2922/10712 [44:50<1:03:44,  2.04it/s]
 27%|██▋       | 2923/10712 [44:50<1:03:45,  2.04it/s]
 27%|██▋       | 2924/10712 [44:51<1:03:45,  2.04it/s]
 27%|██▋       | 2925/10712 [44:51<1:03:47,  2.03it/s]
                                                      
{'loss': 3.9509, 'grad_norm': 0.2139056771993637, 'learning_rate': 0.0009115702404271726, 'epoch': 0.27}

 27%|██▋       | 2925/10712 [44:51<1:03:47,  2.03it/s]
 27%|██▋       | 2926/10712 [44:52<1:03:51,  2.03it/s]
 27%|██▋       | 2927/10712 [44:52<1:03:55,  2.03it/s]
 27%|██▋       | 2928/10712 [44:52<1:03:50,  2.03it/s]
 27%|██▋       | 2929/10712 [44:53<1:03:45,  2.03it/s]
 27%|██▋       | 2930/10712 [44:53<1:03:45,  2.03it/s]
 27%|██▋       | 2931/10712 [44:54<1:03:47,  2.03it/s]
 27%|██▋       | 2932/10712 [44:54<1:03:43,  2.03it/s]
 27%|██▋       | 2933/10712 [44:55<1:03:47,  2.03it/s]
 27%|██▋       | 2934/10712 [44:55<1:04:08,  2.02it/s]
 27%|██▋       | 2935/10712 [44:56<1:04:07,  2.02it/s]
 27%|██▋       | 2936/10712 [44:56<1:03:55,  2.03it/s]
 27%|██▋       | 2937/10712 [44:57<1:03:54,  2.03it/s]
 27%|██▋       | 2938/10712 [44:57<1:03:50,  2.03it/s]
 27%|██▋       | 2939/10712 [44:58<1:03:50,  2.03it/s]
 27%|██▋       | 2940/10712 [44:58<1:03:46,  2.03it/s]
 27%|██▋       | 2941/10712 [44:59<1:04:05,  2.02it/s]
 27%|██▋       | 2942/10712 [44:59<1:03:59,  2.02it/s]
 27%|██▋       | 2943/10712 [45:00<1:03:48,  2.03it/s]
 27%|██▋       | 2944/10712 [45:00<1:03:40,  2.03it/s]
 27%|██▋       | 2945/10712 [45:01<1:03:43,  2.03it/s]
 28%|██▊       | 2946/10712 [45:01<1:03:36,  2.03it/s]
 28%|██▊       | 2947/10712 [45:02<1:03:32,  2.04it/s]
 28%|██▊       | 2948/10712 [45:02<1:03:38,  2.03it/s]
 28%|██▊       | 2949/10712 [45:03<1:03:32,  2.04it/s]
 28%|██▊       | 2950/10712 [45:03<1:03:33,  2.04it/s]
                                                      
{'loss': 3.9424, 'grad_norm': 0.22478492558002472, 'learning_rate': 0.0009092434386367234, 'epoch': 0.28}

 28%|██▊       | 2950/10712 [45:03<1:03:33,  2.04it/s]
 28%|██▊       | 2951/10712 [45:04<1:03:40,  2.03it/s]
 28%|██▊       | 2952/10712 [45:04<1:03:35,  2.03it/s]
 28%|██▊       | 2953/10712 [45:05<1:03:31,  2.04it/s]
 28%|██▊       | 2954/10712 [45:05<1:03:32,  2.04it/s]
 28%|██▊       | 2955/10712 [45:06<1:03:30,  2.04it/s]
 28%|██▊       | 2956/10712 [45:06<1:03:26,  2.04it/s]
 28%|██▊       | 2957/10712 [45:07<1:03:28,  2.04it/s]
 28%|██▊       | 2958/10712 [45:07<1:03:26,  2.04it/s]
 28%|██▊       | 2959/10712 [45:08<1:03:19,  2.04it/s]
 28%|██▊       | 2960/10712 [45:08<1:03:22,  2.04it/s]
 28%|██▊       | 2961/10712 [45:09<1:03:25,  2.04it/s]
 28%|██▊       | 2962/10712 [45:09<1:03:38,  2.03it/s]
 28%|██▊       | 2963/10712 [45:10<1:03:37,  2.03it/s]
 28%|██▊       | 2964/10712 [45:10<1:03:26,  2.04it/s]
 28%|██▊       | 2965/10712 [45:11<1:03:25,  2.04it/s]
 28%|██▊       | 2966/10712 [45:11<1:03:27,  2.03it/s]
 28%|██▊       | 2967/10712 [45:12<1:03:50,  2.02it/s]
 28%|██▊       | 2968/10712 [45:12<1:03:43,  2.03it/s]
 28%|██▊       | 2969/10712 [45:13<1:03:58,  2.02it/s]
 28%|██▊       | 2970/10712 [45:13<1:03:47,  2.02it/s]
 28%|██▊       | 2971/10712 [45:14<1:03:40,  2.03it/s]
 28%|██▊       | 2972/10712 [45:14<1:03:34,  2.03it/s]
 28%|██▊       | 2973/10712 [45:15<1:03:29,  2.03it/s]
 28%|██▊       | 2974/10712 [45:15<1:03:26,  2.03it/s]
 28%|██▊       | 2975/10712 [45:16<1:03:23,  2.03it/s]
                                                      
{'loss': 3.9381, 'grad_norm': 0.21013349294662476, 'learning_rate': 0.0009068894721397322, 'epoch': 0.28}

 28%|██▊       | 2975/10712 [45:16<1:03:23,  2.03it/s]
 28%|██▊       | 2976/10712 [45:16<1:03:31,  2.03it/s]
 28%|██▊       | 2977/10712 [45:17<1:03:29,  2.03it/s]
 28%|██▊       | 2978/10712 [45:17<1:03:27,  2.03it/s]
 28%|██▊       | 2979/10712 [45:18<1:03:22,  2.03it/s]
 28%|██▊       | 2980/10712 [45:18<1:03:22,  2.03it/s]
 28%|██▊       | 2981/10712 [45:19<1:03:19,  2.03it/s]
 28%|██▊       | 2982/10712 [45:19<1:03:19,  2.03it/s]
 28%|██▊       | 2983/10712 [45:20<1:03:13,  2.04it/s]
 28%|██▊       | 2984/10712 [45:20<1:03:08,  2.04it/s]
 28%|██▊       | 2985/10712 [45:21<1:03:06,  2.04it/s]
 28%|██▊       | 2986/10712 [45:21<1:03:06,  2.04it/s]
 28%|██▊       | 2987/10712 [45:22<1:03:11,  2.04it/s]
 28%|██▊       | 2988/10712 [45:22<1:03:11,  2.04it/s]
 28%|██▊       | 2989/10712 [45:23<1:03:05,  2.04it/s]
 28%|██▊       | 2990/10712 [45:23<1:03:04,  2.04it/s]
 28%|██▊       | 2991/10712 [45:23<1:03:06,  2.04it/s]
 28%|██▊       | 2992/10712 [45:24<1:03:06,  2.04it/s]
 28%|██▊       | 2993/10712 [45:24<1:03:04,  2.04it/s]
 28%|██▊       | 2994/10712 [45:25<1:03:03,  2.04it/s]
 28%|██▊       | 2995/10712 [45:25<1:03:10,  2.04it/s]
 28%|██▊       | 2996/10712 [45:26<1:03:14,  2.03it/s]
 28%|██▊       | 2997/10712 [45:26<1:03:11,  2.04it/s]
 28%|██▊       | 2998/10712 [45:27<1:03:08,  2.04it/s]
 28%|██▊       | 2999/10712 [45:27<1:03:05,  2.04it/s]
 28%|██▊       | 3000/10712 [45:28<1:03:01,  2.04it/s]
                                                      
{'loss': 3.9385, 'grad_norm': 0.22034481167793274, 'learning_rate': 0.0009045084971874737, 'epoch': 0.28}

 28%|██▊       | 3000/10712 [45:28<1:03:01,  2.04it/s]
 28%|██▊       | 3001/10712 [45:28<1:03:14,  2.03it/s]
 28%|██▊       | 3002/10712 [45:29<1:03:06,  2.04it/s]
 28%|██▊       | 3003/10712 [45:29<1:03:07,  2.04it/s]
 28%|██▊       | 3004/10712 [45:30<1:03:34,  2.02it/s]
 28%|██▊       | 3005/10712 [45:30<1:03:29,  2.02it/s]
 28%|██▊       | 3006/10712 [45:31<1:03:20,  2.03it/s]
 28%|██▊       | 3007/10712 [45:31<1:03:13,  2.03it/s]
 28%|██▊       | 3008/10712 [45:32<1:03:07,  2.03it/s]
 28%|██▊       | 3009/10712 [45:32<1:03:04,  2.04it/s]
 28%|██▊       | 3010/10712 [45:33<1:03:03,  2.04it/s]
 28%|██▊       | 3011/10712 [45:33<1:03:05,  2.03it/s]
 28%|██▊       | 3012/10712 [45:34<1:03:03,  2.04it/s]
 28%|██▊       | 3013/10712 [45:34<1:03:00,  2.04it/s]
 28%|██▊       | 3014/10712 [45:35<1:03:04,  2.03it/s]
 28%|██▊       | 3015/10712 [45:35<1:03:00,  2.04it/s]
 28%|██▊       | 3016/10712 [45:36<1:03:22,  2.02it/s]
 28%|██▊       | 3017/10712 [45:36<1:03:11,  2.03it/s]
 28%|██▊       | 3018/10712 [45:37<1:03:04,  2.03it/s]
 28%|██▊       | 3019/10712 [45:37<1:03:03,  2.03it/s]
 28%|██▊       | 3020/10712 [45:38<1:03:00,  2.03it/s]
 28%|██▊       | 3021/10712 [45:38<1:02:55,  2.04it/s]
 28%|██▊       | 3022/10712 [45:39<1:03:29,  2.02it/s]
 28%|██▊       | 3023/10712 [45:39<1:03:20,  2.02it/s]
 28%|██▊       | 3024/10712 [45:40<1:03:13,  2.03it/s]
 28%|██▊       | 3025/10712 [45:40<1:03:01,  2.03it/s]
                                                      
{'loss': 3.9329, 'grad_norm': 0.22035916149616241, 'learning_rate': 0.0009021006718239868, 'epoch': 0.28}

 28%|██▊       | 3025/10712 [45:40<1:03:01,  2.03it/s]
 28%|██▊       | 3026/10712 [45:41<1:03:06,  2.03it/s]
 28%|██▊       | 3027/10712 [45:41<1:02:59,  2.03it/s]
 28%|██▊       | 3028/10712 [45:42<1:03:00,  2.03it/s]
 28%|██▊       | 3029/10712 [45:42<1:02:57,  2.03it/s]
 28%|██▊       | 3030/10712 [45:43<1:02:58,  2.03it/s]
 28%|██▊       | 3031/10712 [45:43<1:02:55,  2.03it/s]
 28%|██▊       | 3032/10712 [45:44<1:02:54,  2.03it/s]
 28%|██▊       | 3033/10712 [45:44<1:02:51,  2.04it/s]
 28%|██▊       | 3034/10712 [45:45<1:02:42,  2.04it/s]
 28%|██▊       | 3035/10712 [45:45<1:02:44,  2.04it/s]
 28%|██▊       | 3036/10712 [45:46<1:02:45,  2.04it/s]
 28%|██▊       | 3037/10712 [45:46<1:02:42,  2.04it/s]
 28%|██▊       | 3038/10712 [45:47<1:02:41,  2.04it/s]
 28%|██▊       | 3039/10712 [45:47<1:03:00,  2.03it/s]
 28%|██▊       | 3040/10712 [45:48<1:02:52,  2.03it/s]
 28%|██▊       | 3041/10712 [45:48<1:02:48,  2.04it/s]
 28%|██▊       | 3042/10712 [45:49<1:02:47,  2.04it/s]
 28%|██▊       | 3043/10712 [45:49<1:02:43,  2.04it/s]
 28%|██▊       | 3044/10712 [45:50<1:02:37,  2.04it/s]
 28%|██▊       | 3045/10712 [45:50<1:02:41,  2.04it/s]
 28%|██▊       | 3046/10712 [45:51<1:02:41,  2.04it/s]
 28%|██▊       | 3047/10712 [45:51<1:02:38,  2.04it/s]
 28%|██▊       | 3048/10712 [45:52<1:02:45,  2.04it/s]
 28%|██▊       | 3049/10712 [45:52<1:03:02,  2.03it/s]
 28%|██▊       | 3050/10712 [45:53<1:02:58,  2.03it/s]
                                                      
{'loss': 3.9242, 'grad_norm': 0.21868225932121277, 'learning_rate': 0.0008996661558755833, 'epoch': 0.28}

 28%|██▊       | 3050/10712 [45:53<1:02:58,  2.03it/s]
 28%|██▊       | 3051/10712 [45:53<1:02:57,  2.03it/s]
 28%|██▊       | 3052/10712 [45:53<1:02:47,  2.03it/s]
 29%|██▊       | 3053/10712 [45:54<1:02:47,  2.03it/s]
 29%|██▊       | 3054/10712 [45:54<1:02:44,  2.03it/s]
 29%|██▊       | 3055/10712 [45:55<1:02:41,  2.04it/s]
 29%|██▊       | 3056/10712 [45:55<1:03:05,  2.02it/s]
 29%|██▊       | 3057/10712 [45:56<1:02:58,  2.03it/s]
 29%|██▊       | 3058/10712 [45:56<1:02:53,  2.03it/s]
 29%|██▊       | 3059/10712 [45:57<1:02:47,  2.03it/s]
 29%|██▊       | 3060/10712 [45:57<1:02:41,  2.03it/s]
 29%|██▊       | 3061/10712 [45:58<1:02:40,  2.03it/s]
 29%|██▊       | 3062/10712 [45:58<1:02:36,  2.04it/s]
 29%|██▊       | 3063/10712 [45:59<1:02:36,  2.04it/s]
 29%|██▊       | 3064/10712 [45:59<1:02:35,  2.04it/s]
 29%|██▊       | 3065/10712 [46:00<1:02:33,  2.04it/s]
 29%|██▊       | 3066/10712 [46:00<1:02:30,  2.04it/s]
 29%|██▊       | 3067/10712 [46:01<1:02:34,  2.04it/s]
 29%|██▊       | 3068/10712 [46:01<1:02:36,  2.04it/s]
 29%|██▊       | 3069/10712 [46:02<1:02:31,  2.04it/s]
 29%|██▊       | 3070/10712 [46:02<1:02:25,  2.04it/s]
 29%|██▊       | 3071/10712 [46:03<1:02:23,  2.04it/s]
 29%|██▊       | 3072/10712 [46:03<1:02:52,  2.03it/s]
 29%|██▊       | 3073/10712 [46:04<1:02:44,  2.03it/s]
 29%|██▊       | 3074/10712 [46:04<1:02:39,  2.03it/s]
 29%|██▊       | 3075/10712 [46:05<1:02:55,  2.02it/s]
                                                      
{'loss': 3.9237, 'grad_norm': 0.2126615345478058, 'learning_rate': 0.000897205110940239, 'epoch': 0.29}

 29%|██▊       | 3075/10712 [46:05<1:02:55,  2.02it/s]
 29%|██▊       | 3076/10712 [46:05<1:02:52,  2.02it/s]
 29%|██▊       | 3077/10712 [46:06<1:02:46,  2.03it/s]
 29%|██▊       | 3078/10712 [46:06<1:02:38,  2.03it/s]
 29%|██▊       | 3079/10712 [46:07<1:02:35,  2.03it/s]
 29%|██▉       | 3080/10712 [46:07<1:02:34,  2.03it/s]
 29%|██▉       | 3081/10712 [46:08<1:02:28,  2.04it/s]
 29%|██▉       | 3082/10712 [46:08<1:02:28,  2.04it/s]
 29%|██▉       | 3083/10712 [46:09<1:02:26,  2.04it/s]
 29%|██▉       | 3084/10712 [46:09<1:02:26,  2.04it/s]
 29%|██▉       | 3085/10712 [46:10<1:02:25,  2.04it/s]
 29%|██▉       | 3086/10712 [46:10<1:02:33,  2.03it/s]
 29%|██▉       | 3087/10712 [46:11<1:02:35,  2.03it/s]
 29%|██▉       | 3088/10712 [46:11<1:02:31,  2.03it/s]
 29%|██▉       | 3089/10712 [46:12<1:02:25,  2.04it/s]
 29%|██▉       | 3090/10712 [46:12<1:02:21,  2.04it/s]
 29%|██▉       | 3091/10712 [46:13<1:02:20,  2.04it/s]
 29%|██▉       | 3092/10712 [46:13<1:02:23,  2.04it/s]
 29%|██▉       | 3093/10712 [46:14<1:02:19,  2.04it/s]
 29%|██▉       | 3094/10712 [46:14<1:02:55,  2.02it/s]
 29%|██▉       | 3095/10712 [46:15<1:02:45,  2.02it/s]
 29%|██▉       | 3096/10712 [46:15<1:02:38,  2.03it/s]
 29%|██▉       | 3097/10712 [46:16<1:02:26,  2.03it/s]
 29%|██▉       | 3098/10712 [46:16<1:02:20,  2.04it/s]
 29%|██▉       | 3099/10712 [46:17<1:02:20,  2.04it/s]
 29%|██▉       | 3100/10712 [46:17<1:02:21,  2.03it/s]{'loss': 3.9277, 'grad_norm': 0.20981773734092712, 'learning_rate': 0.0008947177003768671, 'epoch': 0.29}
                                                      

 29%|██▉       | 3100/10712 [46:17<1:02:21,  2.03it/s]
 29%|██▉       | 3101/10712 [46:18<1:02:29,  2.03it/s]
 29%|██▉       | 3102/10712 [46:18<1:02:34,  2.03it/s]
 29%|██▉       | 3103/10712 [46:19<1:02:33,  2.03it/s]
 29%|██▉       | 3104/10712 [46:19<1:02:28,  2.03it/s]
 29%|██▉       | 3105/10712 [46:20<1:02:20,  2.03it/s]
 29%|██▉       | 3106/10712 [46:20<1:02:19,  2.03it/s]
 29%|██▉       | 3107/10712 [46:21<1:02:18,  2.03it/s]
 29%|██▉       | 3108/10712 [46:21<1:03:09,  2.01it/s]
 29%|██▉       | 3109/10712 [46:22<1:02:50,  2.02it/s]
 29%|██▉       | 3110/10712 [46:22<1:02:38,  2.02it/s]
 29%|██▉       | 3111/10712 [46:23<1:02:31,  2.03it/s]
 29%|██▉       | 3112/10712 [46:23<1:02:32,  2.03it/s]
 29%|██▉       | 3113/10712 [46:24<1:02:26,  2.03it/s]
 29%|██▉       | 3114/10712 [46:24<1:02:25,  2.03it/s]
 29%|██▉       | 3115/10712 [46:25<1:02:24,  2.03it/s]
 29%|██▉       | 3116/10712 [46:25<1:02:20,  2.03it/s]
 29%|██▉       | 3117/10712 [46:25<1:02:26,  2.03it/s]
 29%|██▉       | 3118/10712 [46:26<1:02:23,  2.03it/s]
 29%|██▉       | 3119/10712 [46:26<1:02:13,  2.03it/s]
 29%|██▉       | 3120/10712 [46:27<1:02:11,  2.03it/s]
 29%|██▉       | 3121/10712 [46:27<1:02:28,  2.03it/s]
 29%|██▉       | 3122/10712 [46:28<1:02:20,  2.03it/s]
 29%|██▉       | 3123/10712 [46:28<1:02:11,  2.03it/s]
 29%|██▉       | 3124/10712 [46:29<1:02:10,  2.03it/s]
 29%|██▉       | 3125/10712 [46:29<1:02:08,  2.03it/s]
                                                      
{'loss': 3.9221, 'grad_norm': 0.22389867901802063, 'learning_rate': 0.0008922040892944755, 'epoch': 0.29}

 29%|██▉       | 3125/10712 [46:29<1:02:08,  2.03it/s]
 29%|██▉       | 3126/10712 [46:30<1:02:08,  2.03it/s]
 29%|██▉       | 3127/10712 [46:30<1:02:10,  2.03it/s]
 29%|██▉       | 3128/10712 [46:31<1:02:42,  2.02it/s]
 29%|██▉       | 3129/10712 [46:31<1:02:50,  2.01it/s]
 29%|██▉       | 3130/10712 [46:32<1:02:33,  2.02it/s]
 29%|██▉       | 3131/10712 [46:32<1:02:53,  2.01it/s]
 29%|██▉       | 3132/10712 [46:33<1:02:40,  2.02it/s]
 29%|██▉       | 3133/10712 [46:33<1:02:40,  2.02it/s]
 29%|██▉       | 3134/10712 [46:34<1:02:22,  2.02it/s]
 29%|██▉       | 3135/10712 [46:34<1:02:19,  2.03it/s]
 29%|██▉       | 3136/10712 [46:35<1:02:09,  2.03it/s]
 29%|██▉       | 3137/10712 [46:35<1:02:00,  2.04it/s]
 29%|██▉       | 3138/10712 [46:36<1:01:59,  2.04it/s]
 29%|██▉       | 3139/10712 [46:36<1:02:00,  2.04it/s]
 29%|██▉       | 3140/10712 [46:37<1:01:55,  2.04it/s]
 29%|██▉       | 3141/10712 [46:37<1:02:29,  2.02it/s]
 29%|██▉       | 3142/10712 [46:38<1:02:16,  2.03it/s]
 29%|██▉       | 3143/10712 [46:38<1:02:13,  2.03it/s]
 29%|██▉       | 3144/10712 [46:39<1:02:02,  2.03it/s]
 29%|██▉       | 3145/10712 [46:39<1:01:57,  2.04it/s]
 29%|██▉       | 3146/10712 [46:40<1:01:56,  2.04it/s]
 29%|██▉       | 3147/10712 [46:40<1:01:56,  2.04it/s]
 29%|██▉       | 3148/10712 [46:41<1:01:52,  2.04it/s]
 29%|██▉       | 3149/10712 [46:42<1:12:04,  1.75it/s]
 29%|██▉       | 3150/10712 [46:42<1:08:57,  1.83it/s]
                                                      
{'loss': 3.9317, 'grad_norm': 0.2149304747581482, 'learning_rate': 0.0008896644445412061, 'epoch': 0.29}

 29%|██▉       | 3150/10712 [46:42<1:08:57,  1.83it/s]
 29%|██▉       | 3151/10712 [46:43<1:07:03,  1.88it/s]
 29%|██▉       | 3152/10712 [46:43<1:05:26,  1.93it/s]
 29%|██▉       | 3153/10712 [46:43<1:04:20,  1.96it/s]
 29%|██▉       | 3154/10712 [46:44<1:03:35,  1.98it/s]
 29%|██▉       | 3155/10712 [46:44<1:03:03,  2.00it/s]
 29%|██▉       | 3156/10712 [46:45<1:02:40,  2.01it/s]
 29%|██▉       | 3157/10712 [46:45<1:02:24,  2.02it/s]
 29%|██▉       | 3158/10712 [46:46<1:02:20,  2.02it/s]
 29%|██▉       | 3159/10712 [46:46<1:02:07,  2.03it/s]
 29%|██▉       | 3160/10712 [46:47<1:02:02,  2.03it/s]
 30%|██▉       | 3161/10712 [46:47<1:01:58,  2.03it/s]
 30%|██▉       | 3162/10712 [46:48<1:01:54,  2.03it/s]
 30%|██▉       | 3163/10712 [46:48<1:01:50,  2.03it/s]
 30%|██▉       | 3164/10712 [46:49<1:01:45,  2.04it/s]
 30%|██▉       | 3165/10712 [46:49<1:01:41,  2.04it/s]
 30%|██▉       | 3166/10712 [46:50<1:01:41,  2.04it/s]
 30%|██▉       | 3167/10712 [46:50<1:01:42,  2.04it/s]
 30%|██▉       | 3168/10712 [46:51<1:11:56,  1.75it/s]
 30%|██▉       | 3169/10712 [46:52<1:08:52,  1.83it/s]
 30%|██▉       | 3170/10712 [46:52<1:06:39,  1.89it/s]
 30%|██▉       | 3171/10712 [46:53<1:05:08,  1.93it/s]
 30%|██▉       | 3172/10712 [46:53<1:04:08,  1.96it/s]
 30%|██▉       | 3173/10712 [46:54<1:03:48,  1.97it/s]
 30%|██▉       | 3174/10712 [46:54<1:03:04,  1.99it/s]
 30%|██▉       | 3175/10712 [46:55<1:02:35,  2.01it/s]{'loss': 3.9161, 'grad_norm': 0.20543763041496277, 'learning_rate': 0.0008870989346932607, 'epoch': 0.3}
                                                      

 30%|██▉       | 3175/10712 [46:55<1:02:35,  2.01it/s]
 30%|██▉       | 3176/10712 [46:55<1:02:23,  2.01it/s]
 30%|██▉       | 3177/10712 [46:56<1:02:09,  2.02it/s]
 30%|██▉       | 3178/10712 [46:56<1:01:55,  2.03it/s]
 30%|██▉       | 3179/10712 [46:57<1:01:52,  2.03it/s]
 30%|██▉       | 3180/10712 [46:57<1:01:49,  2.03it/s]
 30%|██▉       | 3181/10712 [46:58<1:01:40,  2.03it/s]
 30%|██▉       | 3182/10712 [46:58<1:01:40,  2.03it/s]
 30%|██▉       | 3183/10712 [46:59<1:01:42,  2.03it/s]
 30%|██▉       | 3184/10712 [46:59<1:01:38,  2.04it/s]
 30%|██▉       | 3185/10712 [46:59<1:01:39,  2.03it/s]
 30%|██▉       | 3186/10712 [47:00<1:01:41,  2.03it/s]
 30%|██▉       | 3187/10712 [47:00<1:01:39,  2.03it/s]
 30%|██▉       | 3188/10712 [47:01<1:01:36,  2.04it/s]
 30%|██▉       | 3189/10712 [47:01<1:01:35,  2.04it/s]
 30%|██▉       | 3190/10712 [47:02<1:01:31,  2.04it/s]
 30%|██▉       | 3191/10712 [47:02<1:01:27,  2.04it/s]
 30%|██▉       | 3192/10712 [47:03<1:01:29,  2.04it/s]
 30%|██▉       | 3193/10712 [47:03<1:01:32,  2.04it/s]
 30%|██▉       | 3194/10712 [47:04<1:01:25,  2.04it/s]
 30%|██▉       | 3195/10712 [47:04<1:01:26,  2.04it/s]
 30%|██▉       | 3196/10712 [47:05<1:01:28,  2.04it/s]
 30%|██▉       | 3197/10712 [47:05<1:01:27,  2.04it/s]
 30%|██▉       | 3198/10712 [47:06<1:01:23,  2.04it/s]
 30%|██▉       | 3199/10712 [47:06<1:01:22,  2.04it/s]
 30%|██▉       | 3200/10712 [47:07<1:01:23,  2.04it/s]
                                                      {'loss': 3.9168, 'grad_norm': 0.2275894582271576, 'learning_rate': 0.0008845077300437107, 'epoch': 0.3}

 30%|██▉       | 3200/10712 [47:07<1:01:23,  2.04it/s]
 30%|██▉       | 3201/10712 [47:07<1:01:26,  2.04it/s]
 30%|██▉       | 3202/10712 [47:08<1:01:25,  2.04it/s]
 30%|██▉       | 3203/10712 [47:08<1:01:28,  2.04it/s]
 30%|██▉       | 3204/10712 [47:09<1:01:25,  2.04it/s]
 30%|██▉       | 3205/10712 [47:09<1:01:22,  2.04it/s]
 30%|██▉       | 3206/10712 [47:10<1:01:43,  2.03it/s]
 30%|██▉       | 3207/10712 [47:10<1:01:33,  2.03it/s]
 30%|██▉       | 3208/10712 [47:11<1:01:28,  2.03it/s]
 30%|██▉       | 3209/10712 [47:11<1:01:25,  2.04it/s]
 30%|██▉       | 3210/10712 [47:12<1:01:21,  2.04it/s]
 30%|██▉       | 3211/10712 [47:12<1:01:17,  2.04it/s]
 30%|██▉       | 3212/10712 [47:13<1:01:20,  2.04it/s]
 30%|██▉       | 3213/10712 [47:13<1:01:20,  2.04it/s]
 30%|███       | 3214/10712 [47:14<1:01:17,  2.04it/s]
 30%|███       | 3215/10712 [47:14<1:01:18,  2.04it/s]
 30%|███       | 3216/10712 [47:15<1:01:21,  2.04it/s]
 30%|███       | 3217/10712 [47:15<1:01:17,  2.04it/s]
 30%|███       | 3218/10712 [47:16<1:01:16,  2.04it/s]
 30%|███       | 3219/10712 [47:16<1:01:37,  2.03it/s]
 30%|███       | 3220/10712 [47:17<1:01:32,  2.03it/s]
 30%|███       | 3221/10712 [47:17<1:01:28,  2.03it/s]
 30%|███       | 3222/10712 [47:18<1:01:24,  2.03it/s]
 30%|███       | 3223/10712 [47:18<1:01:17,  2.04it/s]
 30%|███       | 3224/10712 [47:19<1:01:15,  2.04it/s]
 30%|███       | 3225/10712 [47:19<1:01:13,  2.04it/s]{'loss': 3.9151, 'grad_norm': 0.21064886450767517, 'learning_rate': 0.0008818910025911937, 'epoch': 0.3}

                                                      

 30%|███       | 3225/10712 [47:19<1:01:13,  2.04it/s]
 30%|███       | 3226/10712 [47:20<1:01:11,  2.04it/s]
 30%|███       | 3227/10712 [47:20<1:01:11,  2.04it/s]
 30%|███       | 3228/10712 [47:21<1:01:12,  2.04it/s]
 30%|███       | 3229/10712 [47:21<1:01:13,  2.04it/s]
 30%|███       | 3230/10712 [47:22<1:01:14,  2.04it/s]
 30%|███       | 3231/10712 [47:22<1:01:16,  2.03it/s]
 30%|███       | 3232/10712 [47:23<1:01:12,  2.04it/s]
 30%|███       | 3233/10712 [47:23<1:01:09,  2.04it/s]
 30%|███       | 3234/10712 [47:24<1:01:08,  2.04it/s]
 30%|███       | 3235/10712 [47:24<1:01:08,  2.04it/s]
 30%|███       | 3236/10712 [47:25<1:01:02,  2.04it/s]
 30%|███       | 3237/10712 [47:25<1:01:06,  2.04it/s]
 30%|███       | 3238/10712 [47:26<1:01:07,  2.04it/s]
 30%|███       | 3239/10712 [47:26<1:01:05,  2.04it/s]
 30%|███       | 3240/10712 [47:26<1:01:03,  2.04it/s]
 30%|███       | 3241/10712 [47:27<1:01:05,  2.04it/s]
 30%|███       | 3242/10712 [47:27<1:01:05,  2.04it/s]
 30%|███       | 3243/10712 [47:28<1:00:58,  2.04it/s]
 30%|███       | 3244/10712 [47:28<1:01:27,  2.03it/s]
 30%|███       | 3245/10712 [47:29<1:01:20,  2.03it/s]
 30%|███       | 3246/10712 [47:29<1:01:16,  2.03it/s]
 30%|███       | 3247/10712 [47:30<1:01:12,  2.03it/s]
 30%|███       | 3248/10712 [47:30<1:01:03,  2.04it/s]
 30%|███       | 3249/10712 [47:31<1:01:02,  2.04it/s]
 30%|███       | 3250/10712 [47:31<1:01:03,  2.04it/s]{'loss': 3.9049, 'grad_norm': 0.21734307706356049, 'learning_rate': 0.0008792489260284965, 'epoch': 0.3}

                                                      

 30%|███       | 3250/10712 [47:31<1:01:03,  2.04it/s]
 30%|███       | 3251/10712 [47:32<1:01:04,  2.04it/s]
 30%|███       | 3252/10712 [47:32<1:00:59,  2.04it/s]
 30%|███       | 3253/10712 [47:33<1:00:57,  2.04it/s]
 30%|███       | 3254/10712 [47:33<1:01:00,  2.04it/s]
 30%|███       | 3255/10712 [47:34<1:01:02,  2.04it/s]
 30%|███       | 3256/10712 [47:34<1:01:00,  2.04it/s]
 30%|███       | 3257/10712 [47:35<1:00:58,  2.04it/s]
 30%|███       | 3258/10712 [47:35<1:01:00,  2.04it/s]
 30%|███       | 3259/10712 [47:36<1:00:59,  2.04it/s]
 30%|███       | 3260/10712 [47:36<1:01:39,  2.01it/s]
 30%|███       | 3261/10712 [47:37<1:01:20,  2.02it/s]
 30%|███       | 3262/10712 [47:37<1:01:11,  2.03it/s]
 30%|███       | 3263/10712 [47:38<1:01:08,  2.03it/s]
 30%|███       | 3264/10712 [47:38<1:01:06,  2.03it/s]
 30%|███       | 3265/10712 [47:39<1:00:55,  2.04it/s]
 30%|███       | 3266/10712 [47:39<1:00:56,  2.04it/s]
 30%|███       | 3267/10712 [47:40<1:00:56,  2.04it/s]
 31%|███       | 3268/10712 [47:40<1:00:53,  2.04it/s]
 31%|███       | 3269/10712 [47:41<1:00:51,  2.04it/s]
 31%|███       | 3270/10712 [47:41<1:00:52,  2.04it/s]
 31%|███       | 3271/10712 [47:42<1:00:54,  2.04it/s]
 31%|███       | 3272/10712 [47:42<1:00:48,  2.04it/s]
 31%|███       | 3273/10712 [47:43<1:00:45,  2.04it/s]
 31%|███       | 3274/10712 [47:43<1:00:52,  2.04it/s]
 31%|███       | 3275/10712 [47:44<1:00:50,  2.04it/s]{'loss': 3.9068, 'grad_norm': 0.2171017974615097, 'learning_rate': 0.0008765816757310259, 'epoch': 0.31}

                                                      

 31%|███       | 3275/10712 [47:44<1:00:50,  2.04it/s]
 31%|███       | 3276/10712 [47:44<1:00:50,  2.04it/s]
 31%|███       | 3277/10712 [47:45<1:00:48,  2.04it/s]
 31%|███       | 3278/10712 [47:45<1:00:48,  2.04it/s]
 31%|███       | 3279/10712 [47:46<1:00:43,  2.04it/s]
 31%|███       | 3280/10712 [47:46<1:00:40,  2.04it/s]
 31%|███       | 3281/10712 [47:47<1:00:44,  2.04it/s]
 31%|███       | 3282/10712 [47:47<1:00:43,  2.04it/s]
 31%|███       | 3283/10712 [47:48<1:00:36,  2.04it/s]
 31%|███       | 3284/10712 [47:48<1:00:39,  2.04it/s]
 31%|███       | 3285/10712 [47:49<1:00:41,  2.04it/s]
 31%|███       | 3286/10712 [47:49<1:00:40,  2.04it/s]
 31%|███       | 3287/10712 [47:50<1:00:37,  2.04it/s]
 31%|███       | 3288/10712 [47:50<1:00:37,  2.04it/s]
 31%|███       | 3289/10712 [47:51<1:00:39,  2.04it/s]
 31%|███       | 3290/10712 [47:51<1:00:35,  2.04it/s]
 31%|███       | 3291/10712 [47:52<1:00:32,  2.04it/s]
 31%|███       | 3292/10712 [47:52<1:00:35,  2.04it/s]
 31%|███       | 3293/10712 [47:53<1:00:33,  2.04it/s]
 31%|███       | 3294/10712 [47:53<1:00:34,  2.04it/s]
 31%|███       | 3295/10712 [47:53<1:00:33,  2.04it/s]
 31%|███       | 3296/10712 [47:54<1:00:39,  2.04it/s]
 31%|███       | 3297/10712 [47:54<1:00:36,  2.04it/s]
 31%|███       | 3298/10712 [47:55<1:00:34,  2.04it/s]
 31%|███       | 3299/10712 [47:55<1:00:41,  2.04it/s]
 31%|███       | 3300/10712 [47:56<1:00:40,  2.04it/s]{'loss': 3.9042, 'grad_norm': 0.23881828784942627, 'learning_rate': 0.0008738894287451674, 'epoch': 0.31}
                                                      

 31%|███       | 3300/10712 [47:56<1:00:40,  2.04it/s]
 31%|███       | 3301/10712 [47:56<1:00:41,  2.03it/s]
 31%|███       | 3302/10712 [47:57<1:00:43,  2.03it/s]
 31%|███       | 3303/10712 [47:57<1:00:39,  2.04it/s]
 31%|███       | 3304/10712 [47:58<1:00:41,  2.03it/s]
 31%|███       | 3305/10712 [47:58<1:00:41,  2.03it/s]
 31%|███       | 3306/10712 [47:59<1:00:38,  2.04it/s]
 31%|███       | 3307/10712 [47:59<1:00:34,  2.04it/s]
 31%|███       | 3308/10712 [48:00<1:00:35,  2.04it/s]
 31%|███       | 3309/10712 [48:00<1:00:33,  2.04it/s]
 31%|███       | 3310/10712 [48:01<1:00:28,  2.04it/s]
 31%|███       | 3311/10712 [48:01<1:00:47,  2.03it/s]
 31%|███       | 3312/10712 [48:02<1:00:38,  2.03it/s]
 31%|███       | 3313/10712 [48:02<1:00:34,  2.04it/s]
 31%|███       | 3314/10712 [48:03<1:00:33,  2.04it/s]
 31%|███       | 3315/10712 [48:03<1:00:30,  2.04it/s]
 31%|███       | 3316/10712 [48:04<1:00:26,  2.04it/s]
 31%|███       | 3317/10712 [48:04<1:00:26,  2.04it/s]
 31%|███       | 3318/10712 [48:05<1:00:25,  2.04it/s]
 31%|███       | 3319/10712 [48:05<1:00:22,  2.04it/s]
 31%|███       | 3320/10712 [48:06<1:00:20,  2.04it/s]
 31%|███       | 3321/10712 [48:06<1:00:25,  2.04it/s]
 31%|███       | 3322/10712 [48:07<1:00:23,  2.04it/s]
 31%|███       | 3323/10712 [48:07<1:00:22,  2.04it/s]
 31%|███       | 3324/10712 [48:08<1:00:22,  2.04it/s]
 31%|███       | 3325/10712 [48:08<1:00:21,  2.04it/s]{'loss': 3.9094, 'grad_norm': 0.20929791033267975, 'learning_rate': 0.0008711723637765334, 'epoch': 0.31}

                                                      

 31%|███       | 3325/10712 [48:08<1:00:21,  2.04it/s]
 31%|███       | 3326/10712 [48:09<1:00:24,  2.04it/s]
 31%|███       | 3327/10712 [48:09<1:00:21,  2.04it/s]
 31%|███       | 3328/10712 [48:10<1:00:22,  2.04it/s]
 31%|███       | 3329/10712 [48:10<1:00:56,  2.02it/s]
 31%|███       | 3330/10712 [48:11<1:00:42,  2.03it/s]
 31%|███       | 3331/10712 [48:11<1:00:33,  2.03it/s]
 31%|███       | 3332/10712 [48:12<1:00:28,  2.03it/s]
 31%|███       | 3333/10712 [48:12<1:00:26,  2.03it/s]
 31%|███       | 3334/10712 [48:13<1:00:25,  2.04it/s]
 31%|███       | 3335/10712 [48:13<1:00:17,  2.04it/s]
 31%|███       | 3336/10712 [48:14<1:00:23,  2.04it/s]
 31%|███       | 3337/10712 [48:14<1:00:23,  2.04it/s]
 31%|███       | 3338/10712 [48:15<1:00:19,  2.04it/s]
 31%|███       | 3339/10712 [48:15<1:00:19,  2.04it/s]
 31%|███       | 3340/10712 [48:16<1:00:21,  2.04it/s]
 31%|███       | 3341/10712 [48:16<1:00:17,  2.04it/s]
 31%|███       | 3342/10712 [48:17<1:00:13,  2.04it/s]
 31%|███       | 3343/10712 [48:17<1:00:10,  2.04it/s]
 31%|███       | 3344/10712 [48:18<1:00:12,  2.04it/s]
 31%|███       | 3345/10712 [48:18<1:00:10,  2.04it/s]
 31%|███       | 3346/10712 [48:19<1:00:10,  2.04it/s]
 31%|███       | 3347/10712 [48:19<1:00:14,  2.04it/s]
 31%|███▏      | 3348/10712 [48:20<1:00:14,  2.04it/s]
 31%|███▏      | 3349/10712 [48:20<1:00:09,  2.04it/s]
 31%|███▏      | 3350/10712 [48:20<1:00:10,  2.04it/s]
                                                      
{'loss': 3.898, 'grad_norm': 0.20542757213115692, 'learning_rate': 0.0008684306611781009, 'epoch': 0.31}

 31%|███▏      | 3350/10712 [48:20<1:00:10,  2.04it/s]
 31%|███▏      | 3351/10712 [48:21<1:00:13,  2.04it/s]
 31%|███▏      | 3352/10712 [48:21<1:00:10,  2.04it/s]
 31%|███▏      | 3353/10712 [48:22<1:00:08,  2.04it/s]
 31%|███▏      | 3354/10712 [48:22<1:00:10,  2.04it/s]
 31%|███▏      | 3355/10712 [48:23<1:00:10,  2.04it/s]
 31%|███▏      | 3356/10712 [48:23<1:00:09,  2.04it/s]
 31%|███▏      | 3357/10712 [48:24<1:00:13,  2.04it/s]
 31%|███▏      | 3358/10712 [48:24<1:00:11,  2.04it/s]
 31%|███▏      | 3359/10712 [48:25<1:00:05,  2.04it/s]
 31%|███▏      | 3360/10712 [48:25<1:00:03,  2.04it/s]
 31%|███▏      | 3361/10712 [48:26<1:00:05,  2.04it/s]
 31%|███▏      | 3362/10712 [48:26<1:00:04,  2.04it/s]
 31%|███▏      | 3363/10712 [48:27<1:00:01,  2.04it/s]
 31%|███▏      | 3364/10712 [48:27<1:00:01,  2.04it/s]
 31%|███▏      | 3365/10712 [48:28<1:00:01,  2.04it/s]
 31%|███▏      | 3366/10712 [48:28<1:00:01,  2.04it/s]
 31%|███▏      | 3367/10712 [48:29<1:00:06,  2.04it/s]
 31%|███▏      | 3368/10712 [48:29<1:00:04,  2.04it/s]
 31%|███▏      | 3369/10712 [48:30<59:59,  2.04it/s]  
 31%|███▏      | 3370/10712 [48:30<59:54,  2.04it/s]
 31%|███▏      | 3371/10712 [48:31<59:58,  2.04it/s]
 31%|███▏      | 3372/10712 [48:31<1:00:01,  2.04it/s]
 31%|███▏      | 3373/10712 [48:32<59:56,  2.04it/s]  
 31%|███▏      | 3374/10712 [48:32<59:56,  2.04it/s]
 32%|███▏      | 3375/10712 [48:33<59:59,  2.04it/s]
                                                    {'loss': 3.8945, 'grad_norm': 0.22066636383533478, 'learning_rate': 0.0008656645029382404, 'epoch': 0.32}


 32%|███▏      | 3375/10712 [48:33<59:59,  2.04it/s]
 32%|███▏      | 3376/10712 [48:33<1:00:05,  2.03it/s]
 32%|███▏      | 3377/10712 [48:34<1:00:01,  2.04it/s]
 32%|███▏      | 3378/10712 [48:34<1:00:03,  2.04it/s]
 32%|███▏      | 3379/10712 [48:35<59:57,  2.04it/s]  
 32%|███▏      | 3380/10712 [48:35<59:53,  2.04it/s]
 32%|███▏      | 3381/10712 [48:36<59:55,  2.04it/s]
 32%|███▏      | 3382/10712 [48:36<59:55,  2.04it/s]
 32%|███▏      | 3383/10712 [48:37<59:50,  2.04it/s]
 32%|███▏      | 3384/10712 [48:37<59:47,  2.04it/s]
 32%|███▏      | 3385/10712 [48:38<59:49,  2.04it/s]
 32%|███▏      | 3386/10712 [48:38<59:51,  2.04it/s]
 32%|███▏      | 3387/10712 [48:39<59:49,  2.04it/s]
 32%|███▏      | 3388/10712 [48:39<59:48,  2.04it/s]
 32%|███▏      | 3389/10712 [48:40<59:50,  2.04it/s]
 32%|███▏      | 3390/10712 [48:40<1:00:08,  2.03it/s]
 32%|███▏      | 3391/10712 [48:41<1:00:01,  2.03it/s]
 32%|███▏      | 3392/10712 [48:41<59:57,  2.03it/s]  
 32%|███▏      | 3393/10712 [48:42<59:54,  2.04it/s]
 32%|███▏      | 3394/10712 [48:42<59:54,  2.04it/s]
 32%|███▏      | 3395/10712 [48:43<59:47,  2.04it/s]
 32%|███▏      | 3396/10712 [48:43<59:48,  2.04it/s]
 32%|███▏      | 3397/10712 [48:44<59:50,  2.04it/s]
 32%|███▏      | 3398/10712 [48:44<59:47,  2.04it/s]
 32%|███▏      | 3399/10712 [48:45<59:44,  2.04it/s]
 32%|███▏      | 3400/10712 [48:45<59:42,  2.04it/s]{'loss': 3.8944, 'grad_norm': 0.2134931981563568, 'learning_rate': 0.0008628740726686357, 'epoch': 0.32}

                                                    

 32%|███▏      | 3400/10712 [48:45<59:42,  2.04it/s]
 32%|███▏      | 3401/10712 [48:46<59:46,  2.04it/s]
 32%|███▏      | 3402/10712 [48:46<59:47,  2.04it/s]
 32%|███▏      | 3403/10712 [48:47<1:00:19,  2.02it/s]
 32%|███▏      | 3404/10712 [48:47<1:00:07,  2.03it/s]
 32%|███▏      | 3405/10712 [48:47<59:58,  2.03it/s]  
 32%|███▏      | 3406/10712 [48:48<59:53,  2.03it/s]
 32%|███▏      | 3407/10712 [48:48<59:52,  2.03it/s]
 32%|███▏      | 3408/10712 [48:49<59:46,  2.04it/s]
 32%|███▏      | 3409/10712 [48:49<59:43,  2.04it/s]
 32%|███▏      | 3410/10712 [48:50<59:42,  2.04it/s]
 32%|███▏      | 3411/10712 [48:50<1:00:10,  2.02it/s]
 32%|███▏      | 3412/10712 [48:51<1:00:17,  2.02it/s]
 32%|███▏      | 3413/10712 [48:51<1:00:12,  2.02it/s]
 32%|███▏      | 3414/10712 [48:52<1:00:00,  2.03it/s]
 32%|███▏      | 3415/10712 [48:52<59:52,  2.03it/s]  
 32%|███▏      | 3416/10712 [48:53<59:51,  2.03it/s]
 32%|███▏      | 3417/10712 [48:53<59:45,  2.03it/s]
 32%|███▏      | 3418/10712 [48:54<59:38,  2.04it/s]
 32%|███▏      | 3419/10712 [48:54<59:39,  2.04it/s]
 32%|███▏      | 3420/10712 [48:55<59:37,  2.04it/s]
 32%|███▏      | 3421/10712 [48:55<59:36,  2.04it/s]
 32%|███▏      | 3422/10712 [48:56<59:32,  2.04it/s]
 32%|███▏      | 3423/10712 [48:56<59:37,  2.04it/s]
 32%|███▏      | 3424/10712 [48:57<59:35,  2.04it/s]
 32%|███▏      | 3425/10712 [48:57<59:30,  2.04it/s]{'loss': 3.8917, 'grad_norm': 0.22930023074150085, 'learning_rate': 0.000860059555592096, 'epoch': 0.32}

                                                    

 32%|███▏      | 3425/10712 [48:57<59:30,  2.04it/s]
 32%|███▏      | 3426/10712 [48:58<59:56,  2.03it/s]
 32%|███▏      | 3427/10712 [48:58<59:42,  2.03it/s]
 32%|███▏      | 3428/10712 [48:59<59:41,  2.03it/s]
 32%|███▏      | 3429/10712 [48:59<59:42,  2.03it/s]
 32%|███▏      | 3430/10712 [49:00<59:36,  2.04it/s]
 32%|███▏      | 3431/10712 [49:00<59:32,  2.04it/s]
 32%|███▏      | 3432/10712 [49:01<59:35,  2.04it/s]
 32%|███▏      | 3433/10712 [49:01<59:34,  2.04it/s]
 32%|███▏      | 3434/10712 [49:02<59:33,  2.04it/s]
 32%|███▏      | 3435/10712 [49:02<59:34,  2.04it/s]
 32%|███▏      | 3436/10712 [49:03<59:32,  2.04it/s]
 32%|███▏      | 3437/10712 [49:03<59:29,  2.04it/s]
 32%|███▏      | 3438/10712 [49:04<59:31,  2.04it/s]
 32%|███▏      | 3439/10712 [49:04<59:31,  2.04it/s]
 32%|███▏      | 3440/10712 [49:05<59:27,  2.04it/s]
 32%|███▏      | 3441/10712 [49:05<59:29,  2.04it/s]
 32%|███▏      | 3442/10712 [49:06<59:29,  2.04it/s]
 32%|███▏      | 3443/10712 [49:06<59:22,  2.04it/s]
 32%|███▏      | 3444/10712 [49:07<59:22,  2.04it/s]
 32%|███▏      | 3445/10712 [49:07<59:24,  2.04it/s]
 32%|███▏      | 3446/10712 [49:08<59:21,  2.04it/s]
 32%|███▏      | 3447/10712 [49:08<59:18,  2.04it/s]
 32%|███▏      | 3448/10712 [49:09<59:19,  2.04it/s]
 32%|███▏      | 3449/10712 [49:09<59:21,  2.04it/s]
 32%|███▏      | 3450/10712 [49:10<59:33,  2.03it/s]{'loss': 3.8868, 'grad_norm': 0.2013440728187561, 'learning_rate': 0.0008572211385302617, 'epoch': 0.32}

                                                    

 32%|███▏      | 3450/10712 [49:10<59:33,  2.03it/s]
 32%|███▏      | 3451/10712 [49:10<59:33,  2.03it/s]
 32%|███▏      | 3452/10712 [49:11<59:24,  2.04it/s]
 32%|███▏      | 3453/10712 [49:11<59:18,  2.04it/s]
 32%|███▏      | 3454/10712 [49:12<59:21,  2.04it/s]
 32%|███▏      | 3455/10712 [49:12<59:21,  2.04it/s]
 32%|███▏      | 3456/10712 [49:13<59:20,  2.04it/s]
 32%|███▏      | 3457/10712 [49:13<59:18,  2.04it/s]
 32%|███▏      | 3458/10712 [49:14<59:22,  2.04it/s]
 32%|███▏      | 3459/10712 [49:14<59:20,  2.04it/s]
 32%|███▏      | 3460/10712 [49:14<59:17,  2.04it/s]
 32%|███▏      | 3461/10712 [49:15<59:21,  2.04it/s]
 32%|███▏      | 3462/10712 [49:15<59:16,  2.04it/s]
 32%|███▏      | 3463/10712 [49:16<59:14,  2.04it/s]
 32%|███▏      | 3464/10712 [49:16<59:16,  2.04it/s]
 32%|███▏      | 3465/10712 [49:17<59:17,  2.04it/s]
 32%|███▏      | 3466/10712 [49:17<59:15,  2.04it/s]
 32%|███▏      | 3467/10712 [49:18<59:20,  2.04it/s]
 32%|███▏      | 3468/10712 [49:18<59:22,  2.03it/s]
 32%|███▏      | 3469/10712 [49:19<59:15,  2.04it/s]
 32%|███▏      | 3470/10712 [49:19<59:11,  2.04it/s]
 32%|███▏      | 3471/10712 [49:20<59:09,  2.04it/s]
 32%|███▏      | 3472/10712 [49:20<59:10,  2.04it/s]
 32%|███▏      | 3473/10712 [49:21<59:06,  2.04it/s]
 32%|███▏      | 3474/10712 [49:21<59:01,  2.04it/s]
 32%|███▏      | 3475/10712 [49:22<59:03,  2.04it/s]{'loss': 3.8894, 'grad_norm': 0.21955867111682892, 'learning_rate': 0.0008543590098912028, 'epoch': 0.32}

                                                    

 32%|███▏      | 3475/10712 [49:22<59:03,  2.04it/s]
 32%|███▏      | 3476/10712 [49:22<59:08,  2.04it/s]
 32%|███▏      | 3477/10712 [49:23<59:07,  2.04it/s]
 32%|███▏      | 3478/10712 [49:23<59:08,  2.04it/s]
 32%|███▏      | 3479/10712 [49:24<59:09,  2.04it/s]
 32%|███▏      | 3480/10712 [49:24<59:08,  2.04it/s]
 32%|███▏      | 3481/10712 [49:25<59:04,  2.04it/s]
 33%|███▎      | 3482/10712 [49:25<59:04,  2.04it/s]
 33%|███▎      | 3483/10712 [49:26<59:08,  2.04it/s]
 33%|███▎      | 3484/10712 [49:26<59:08,  2.04it/s]
 33%|███▎      | 3485/10712 [49:27<59:02,  2.04it/s]
 33%|███▎      | 3486/10712 [49:27<59:06,  2.04it/s]
 33%|███▎      | 3487/10712 [49:28<59:03,  2.04it/s]
 33%|███▎      | 3488/10712 [49:28<59:03,  2.04it/s]
 33%|███▎      | 3489/10712 [49:29<59:02,  2.04it/s]
 33%|███▎      | 3490/10712 [49:29<59:04,  2.04it/s]
 33%|███▎      | 3491/10712 [49:30<59:01,  2.04it/s]
 33%|███▎      | 3492/10712 [49:30<59:02,  2.04it/s]
 33%|███▎      | 3493/10712 [49:31<59:02,  2.04it/s]
 33%|███▎      | 3494/10712 [49:31<58:57,  2.04it/s]
 33%|███▎      | 3495/10712 [49:32<58:55,  2.04it/s]
 33%|███▎      | 3496/10712 [49:32<59:09,  2.03it/s]
 33%|███▎      | 3497/10712 [49:33<59:05,  2.03it/s]
 33%|███▎      | 3498/10712 [49:33<59:05,  2.03it/s]
 33%|███▎      | 3499/10712 [49:34<59:04,  2.03it/s]
 33%|███▎      | 3500/10712 [49:34<59:02,  2.04it/s]{'loss': 3.8812, 'grad_norm': 0.20271483063697815, 'learning_rate': 0.0008514733596569138, 'epoch': 0.33}

                                                    

 33%|███▎      | 3500/10712 [49:34<59:02,  2.04it/s]
 33%|███▎      | 3501/10712 [49:35<59:07,  2.03it/s]
 33%|███▎      | 3502/10712 [49:35<59:07,  2.03it/s]
 33%|███▎      | 3503/10712 [49:36<59:01,  2.04it/s]
 33%|███▎      | 3504/10712 [49:36<58:54,  2.04it/s]
 33%|███▎      | 3505/10712 [49:37<59:00,  2.04it/s]
 33%|███▎      | 3506/10712 [49:37<58:58,  2.04it/s]
 33%|███▎      | 3507/10712 [49:38<58:51,  2.04it/s]
 33%|███▎      | 3508/10712 [49:38<58:52,  2.04it/s]
 33%|███▎      | 3509/10712 [49:39<58:54,  2.04it/s]
 33%|███▎      | 3510/10712 [49:39<58:51,  2.04it/s]
 33%|███▎      | 3511/10712 [49:40<59:09,  2.03it/s]
 33%|███▎      | 3512/10712 [49:40<59:03,  2.03it/s]
 33%|███▎      | 3513/10712 [49:41<59:03,  2.03it/s]
 33%|███▎      | 3514/10712 [49:41<59:02,  2.03it/s]
 33%|███▎      | 3515/10712 [49:41<58:56,  2.04it/s]
 33%|███▎      | 3516/10712 [49:42<58:51,  2.04it/s]
 33%|███▎      | 3517/10712 [49:42<58:52,  2.04it/s]
 33%|███▎      | 3518/10712 [49:43<58:50,  2.04it/s]
 33%|███▎      | 3519/10712 [49:43<58:46,  2.04it/s]
 33%|███▎      | 3520/10712 [49:44<58:44,  2.04it/s]
 33%|███▎      | 3521/10712 [49:44<58:51,  2.04it/s]
 33%|███▎      | 3522/10712 [49:45<58:48,  2.04it/s]
 33%|███▎      | 3523/10712 [49:45<58:48,  2.04it/s]
 33%|███▎      | 3524/10712 [49:46<58:50,  2.04it/s]
 33%|███▎      | 3525/10712 [49:46<58:51,  2.04it/s]
                                                    
{'loss': 3.8855, 'grad_norm': 0.1971137523651123, 'learning_rate': 0.0008485643793707016, 'epoch': 0.33}

 33%|███▎      | 3525/10712 [49:46<58:51,  2.04it/s]
 33%|███▎      | 3526/10712 [49:47<58:52,  2.03it/s]
 33%|███▎      | 3527/10712 [49:47<58:48,  2.04it/s]
 33%|███▎      | 3528/10712 [49:48<58:51,  2.03it/s]
 33%|███▎      | 3529/10712 [49:48<58:50,  2.03it/s]
 33%|███▎      | 3530/10712 [49:49<58:46,  2.04it/s]
 33%|███▎      | 3531/10712 [49:49<58:43,  2.04it/s]
 33%|███▎      | 3532/10712 [49:50<58:46,  2.04it/s]
 33%|███▎      | 3533/10712 [49:50<58:42,  2.04it/s]
 33%|███▎      | 3534/10712 [49:51<58:37,  2.04it/s]
 33%|███▎      | 3535/10712 [49:51<58:33,  2.04it/s]
 33%|███▎      | 3536/10712 [49:52<58:39,  2.04it/s]
 33%|███▎      | 3537/10712 [49:52<58:38,  2.04it/s]
 33%|███▎      | 3538/10712 [49:53<58:35,  2.04it/s]
 33%|███▎      | 3539/10712 [49:53<58:32,  2.04it/s]
 33%|███▎      | 3540/10712 [49:54<58:31,  2.04it/s]
 33%|███▎      | 3541/10712 [49:54<58:37,  2.04it/s]
 33%|███▎      | 3542/10712 [49:55<58:34,  2.04it/s]
 33%|███▎      | 3543/10712 [49:55<58:33,  2.04it/s]
 33%|███▎      | 3544/10712 [49:56<58:35,  2.04it/s]
 33%|███▎      | 3545/10712 [49:56<58:37,  2.04it/s]
 33%|███▎      | 3546/10712 [49:57<58:36,  2.04it/s]
 33%|███▎      | 3547/10712 [49:57<58:36,  2.04it/s]
 33%|███▎      | 3548/10712 [49:58<59:52,  1.99it/s]
 33%|███▎      | 3549/10712 [49:58<59:26,  2.01it/s]
 33%|███▎      | 3550/10712 [49:59<59:07,  2.02it/s]
                                                    
{'loss': 3.8801, 'grad_norm': 0.21767692267894745, 'learning_rate': 0.0008456322621244734, 'epoch': 0.33}

 33%|███▎      | 3550/10712 [49:59<59:07,  2.02it/s]
 33%|███▎      | 3551/10712 [49:59<58:59,  2.02it/s]
 33%|███▎      | 3552/10712 [50:00<58:49,  2.03it/s]
 33%|███▎      | 3553/10712 [50:00<58:44,  2.03it/s]
 33%|███▎      | 3554/10712 [50:01<1:00:26,  1.97it/s]
 33%|███▎      | 3555/10712 [50:01<59:47,  1.99it/s]  
 33%|███▎      | 3556/10712 [50:02<59:23,  2.01it/s]
 33%|███▎      | 3557/10712 [50:02<59:04,  2.02it/s]
 33%|███▎      | 3558/10712 [50:03<58:53,  2.02it/s]
 33%|███▎      | 3559/10712 [50:03<58:43,  2.03it/s]
 33%|███▎      | 3560/10712 [50:04<58:35,  2.03it/s]
 33%|███▎      | 3561/10712 [50:04<58:31,  2.04it/s]
 33%|███▎      | 3562/10712 [50:05<58:31,  2.04it/s]
 33%|███▎      | 3563/10712 [50:05<1:00:05,  1.98it/s]
 33%|███▎      | 3564/10712 [50:06<59:37,  2.00it/s]  
 33%|███▎      | 3565/10712 [50:06<59:14,  2.01it/s]
 33%|███▎      | 3566/10712 [50:07<58:57,  2.02it/s]
 33%|███▎      | 3567/10712 [50:07<58:56,  2.02it/s]
 33%|███▎      | 3568/10712 [50:08<58:44,  2.03it/s]
 33%|███▎      | 3569/10712 [50:08<58:35,  2.03it/s]
 33%|███▎      | 3570/10712 [50:09<58:33,  2.03it/s]
 33%|███▎      | 3571/10712 [50:09<58:33,  2.03it/s]
 33%|███▎      | 3572/10712 [50:10<58:26,  2.04it/s]
 33%|███▎      | 3573/10712 [50:10<59:16,  2.01it/s]
 33%|███▎      | 3574/10712 [50:11<58:59,  2.02it/s]
 33%|███▎      | 3575/10712 [50:11<58:48,  2.02it/s]{'loss': 3.8826, 'grad_norm': 0.20449194312095642, 'learning_rate': 0.0008426772025459173, 'epoch': 0.33}

                                                    

 33%|███▎      | 3575/10712 [50:11<58:48,  2.02it/s]
 33%|███▎      | 3576/10712 [50:12<58:47,  2.02it/s]
 33%|███▎      | 3577/10712 [50:12<58:35,  2.03it/s]
 33%|███▎      | 3578/10712 [50:13<58:30,  2.03it/s]
 33%|███▎      | 3579/10712 [50:13<58:29,  2.03it/s]
 33%|███▎      | 3580/10712 [50:14<58:28,  2.03it/s]
 33%|███▎      | 3581/10712 [50:14<58:42,  2.02it/s]
 33%|███▎      | 3582/10712 [50:15<58:32,  2.03it/s]
 33%|███▎      | 3583/10712 [50:15<58:26,  2.03it/s]
 33%|███▎      | 3584/10712 [50:16<58:26,  2.03it/s]
 33%|███▎      | 3585/10712 [50:16<58:21,  2.04it/s]
 33%|███▎      | 3586/10712 [50:16<58:18,  2.04it/s]
 33%|███▎      | 3587/10712 [50:17<58:17,  2.04it/s]
 33%|███▎      | 3588/10712 [50:17<58:17,  2.04it/s]
 34%|███▎      | 3589/10712 [50:18<58:15,  2.04it/s]
 34%|███▎      | 3590/10712 [50:18<58:17,  2.04it/s]
 34%|███▎      | 3591/10712 [50:19<58:14,  2.04it/s]
 34%|███▎      | 3592/10712 [50:19<58:11,  2.04it/s]
 34%|███▎      | 3593/10712 [50:20<58:12,  2.04it/s]
 34%|███▎      | 3594/10712 [50:20<58:09,  2.04it/s]
 34%|███▎      | 3595/10712 [50:21<58:09,  2.04it/s]
 34%|███▎      | 3596/10712 [50:21<58:10,  2.04it/s]
 34%|███▎      | 3597/10712 [50:22<58:16,  2.03it/s]
 34%|███▎      | 3598/10712 [50:22<58:12,  2.04it/s]
 34%|███▎      | 3599/10712 [50:23<58:09,  2.04it/s]
 34%|███▎      | 3600/10712 [50:23<58:12,  2.04it/s]
                                                    
{'loss': 3.8767, 'grad_norm': 0.20523901283740997, 'learning_rate': 0.0008396993967855852, 'epoch': 0.34}

 34%|███▎      | 3600/10712 [50:23<58:12,  2.04it/s]
 34%|███▎      | 3601/10712 [50:24<58:12,  2.04it/s]
 34%|███▎      | 3602/10712 [50:24<58:10,  2.04it/s]
 34%|███▎      | 3603/10712 [50:25<58:10,  2.04it/s]
 34%|███▎      | 3604/10712 [50:25<58:09,  2.04it/s]
 34%|███▎      | 3605/10712 [50:26<58:05,  2.04it/s]
 34%|███▎      | 3606/10712 [50:26<58:11,  2.04it/s]
 34%|███▎      | 3607/10712 [50:27<58:11,  2.04it/s]
 34%|███▎      | 3608/10712 [50:27<58:08,  2.04it/s]
 34%|███▎      | 3609/10712 [50:28<58:08,  2.04it/s]
 34%|███▎      | 3610/10712 [50:28<58:10,  2.03it/s]
 34%|███▎      | 3611/10712 [50:29<58:08,  2.04it/s]
 34%|███▎      | 3612/10712 [50:29<58:07,  2.04it/s]
 34%|███▎      | 3613/10712 [50:30<58:09,  2.03it/s]
 34%|███▎      | 3614/10712 [50:30<58:04,  2.04it/s]
 34%|███▎      | 3615/10712 [50:31<58:01,  2.04it/s]
 34%|███▍      | 3616/10712 [50:31<58:01,  2.04it/s]
 34%|███▍      | 3617/10712 [50:32<58:03,  2.04it/s]
 34%|███▍      | 3618/10712 [50:32<57:57,  2.04it/s]
 34%|███▍      | 3619/10712 [50:33<57:54,  2.04it/s]
 34%|███▍      | 3620/10712 [50:33<57:59,  2.04it/s]
 34%|███▍      | 3621/10712 [50:34<58:02,  2.04it/s]
 34%|███▍      | 3622/10712 [50:34<57:55,  2.04it/s]
 34%|███▍      | 3623/10712 [50:35<57:56,  2.04it/s]
 34%|███▍      | 3624/10712 [50:35<57:57,  2.04it/s]
 34%|███▍      | 3625/10712 [50:36<57:57,  2.04it/s]
                                                    
{'loss': 3.877, 'grad_norm': 0.21334771811962128, 'learning_rate': 0.000836699042503872, 'epoch': 0.34}

 34%|███▍      | 3625/10712 [50:36<57:57,  2.04it/s]
 34%|███▍      | 3626/10712 [50:36<58:01,  2.04it/s]
 34%|███▍      | 3627/10712 [50:37<58:01,  2.03it/s]
 34%|███▍      | 3628/10712 [50:37<57:59,  2.04it/s]
 34%|███▍      | 3629/10712 [50:38<58:00,  2.04it/s]
 34%|███▍      | 3630/10712 [50:38<57:58,  2.04it/s]
 34%|███▍      | 3631/10712 [50:39<57:56,  2.04it/s]
 34%|███▍      | 3632/10712 [50:39<57:49,  2.04it/s]
 34%|███▍      | 3633/10712 [50:40<57:50,  2.04it/s]
 34%|███▍      | 3634/10712 [50:40<57:51,  2.04it/s]
 34%|███▍      | 3635/10712 [50:41<57:48,  2.04it/s]
 34%|███▍      | 3636/10712 [50:41<57:42,  2.04it/s]
 34%|███▍      | 3637/10712 [50:42<58:07,  2.03it/s]
 34%|███▍      | 3638/10712 [50:42<58:00,  2.03it/s]
 34%|███▍      | 3639/10712 [50:43<57:58,  2.03it/s]
 34%|███▍      | 3640/10712 [50:43<57:55,  2.03it/s]
 34%|███▍      | 3641/10712 [50:43<57:50,  2.04it/s]
 34%|███▍      | 3642/10712 [50:44<57:49,  2.04it/s]
 34%|███▍      | 3643/10712 [50:44<57:50,  2.04it/s]
 34%|███▍      | 3644/10712 [50:45<57:49,  2.04it/s]
 34%|███▍      | 3645/10712 [50:45<57:48,  2.04it/s]
 34%|███▍      | 3646/10712 [50:46<57:49,  2.04it/s]
 34%|███▍      | 3647/10712 [50:46<57:50,  2.04it/s]
 34%|███▍      | 3648/10712 [50:47<57:48,  2.04it/s]
 34%|███▍      | 3649/10712 [50:47<58:35,  2.01it/s]
 34%|███▍      | 3650/10712 [50:48<58:20,  2.02it/s]
                                                    
{'loss': 3.871, 'grad_norm': 0.2115827202796936, 'learning_rate': 0.000833676338857895, 'epoch': 0.34}

 34%|███▍      | 3650/10712 [50:48<58:20,  2.02it/s]
 34%|███▍      | 3651/10712 [50:48<58:16,  2.02it/s]
 34%|███▍      | 3652/10712 [50:49<58:09,  2.02it/s]
 34%|███▍      | 3653/10712 [50:49<57:57,  2.03it/s]
 34%|███▍      | 3654/10712 [50:50<57:52,  2.03it/s]
 34%|███▍      | 3655/10712 [50:50<57:55,  2.03it/s]
 34%|███▍      | 3656/10712 [50:51<57:49,  2.03it/s]
 34%|███▍      | 3657/10712 [50:51<57:44,  2.04it/s]
 34%|███▍      | 3658/10712 [50:52<57:44,  2.04it/s]
 34%|███▍      | 3659/10712 [50:52<57:42,  2.04it/s]
 34%|███▍      | 3660/10712 [50:53<57:37,  2.04it/s]
 34%|███▍      | 3661/10712 [50:53<58:06,  2.02it/s]
 34%|███▍      | 3662/10712 [50:54<57:56,  2.03it/s]
 34%|███▍      | 3663/10712 [50:54<57:48,  2.03it/s]
 34%|███▍      | 3664/10712 [50:55<57:45,  2.03it/s]
 34%|███▍      | 3665/10712 [50:55<57:44,  2.03it/s]
 34%|███▍      | 3666/10712 [50:56<57:43,  2.03it/s]
 34%|███▍      | 3667/10712 [50:56<57:37,  2.04it/s]
 34%|███▍      | 3668/10712 [50:57<57:37,  2.04it/s]
 34%|███▍      | 3669/10712 [50:57<57:37,  2.04it/s]
 34%|███▍      | 3670/10712 [50:58<57:36,  2.04it/s]
 34%|███▍      | 3671/10712 [50:58<57:32,  2.04it/s]
 34%|███▍      | 3672/10712 [50:59<57:33,  2.04it/s]
 34%|███▍      | 3673/10712 [50:59<57:33,  2.04it/s]
 34%|███▍      | 3674/10712 [51:00<57:32,  2.04it/s]
 34%|███▍      | 3675/10712 [51:00<57:32,  2.04it/s]
                                                    {'loss': 3.8753, 'grad_norm': 0.21438553929328918, 'learning_rate': 0.0008306314864882747, 'epoch': 0.34}


 34%|███▍      | 3675/10712 [51:00<57:32,  2.04it/s]
 34%|███▍      | 3676/10712 [51:01<57:38,  2.03it/s]
 34%|███▍      | 3677/10712 [51:01<57:34,  2.04it/s]
 34%|███▍      | 3678/10712 [51:02<57:30,  2.04it/s]
 34%|███▍      | 3679/10712 [51:02<57:26,  2.04it/s]
 34%|███▍      | 3680/10712 [51:03<57:28,  2.04it/s]
 34%|███▍      | 3681/10712 [51:03<57:30,  2.04it/s]
 34%|███▍      | 3682/10712 [51:04<57:28,  2.04it/s]
 34%|███▍      | 3683/10712 [51:04<57:32,  2.04it/s]
 34%|███▍      | 3684/10712 [51:05<57:29,  2.04it/s]
 34%|███▍      | 3685/10712 [51:05<57:26,  2.04it/s]
 34%|███▍      | 3686/10712 [51:06<57:24,  2.04it/s]
 34%|███▍      | 3687/10712 [51:06<57:25,  2.04it/s]
 34%|███▍      | 3688/10712 [51:07<57:24,  2.04it/s]
 34%|███▍      | 3689/10712 [51:07<57:20,  2.04it/s]
 34%|███▍      | 3690/10712 [51:08<57:22,  2.04it/s]
 34%|███▍      | 3691/10712 [51:08<57:21,  2.04it/s]
 34%|███▍      | 3692/10712 [51:09<57:19,  2.04it/s]
 34%|███▍      | 3693/10712 [51:09<57:16,  2.04it/s]
 34%|███▍      | 3694/10712 [51:10<57:17,  2.04it/s]
 34%|███▍      | 3695/10712 [51:10<57:19,  2.04it/s]
 35%|███▍      | 3696/10712 [51:11<57:17,  2.04it/s]
 35%|███▍      | 3697/10712 [51:11<57:18,  2.04it/s]
 35%|███▍      | 3698/10712 [51:11<57:23,  2.04it/s]
 35%|███▍      | 3699/10712 [51:12<57:18,  2.04it/s]
 35%|███▍      | 3700/10712 [51:12<57:15,  2.04it/s]{'loss': 3.88, 'grad_norm': 0.20986370742321014, 'learning_rate': 0.0008275646875058169, 'epoch': 0.35}

                                                    

 35%|███▍      | 3700/10712 [51:12<57:15,  2.04it/s]
 35%|███▍      | 3701/10712 [51:13<57:25,  2.03it/s]
 35%|███▍      | 3702/10712 [51:13<57:21,  2.04it/s]
 35%|███▍      | 3703/10712 [51:14<57:17,  2.04it/s]
 35%|███▍      | 3704/10712 [51:14<57:20,  2.04it/s]
 35%|███▍      | 3705/10712 [51:15<57:19,  2.04it/s]
 35%|███▍      | 3706/10712 [51:15<57:14,  2.04it/s]
 35%|███▍      | 3707/10712 [51:16<57:15,  2.04it/s]
 35%|███▍      | 3708/10712 [51:16<57:15,  2.04it/s]
 35%|███▍      | 3709/10712 [51:17<57:15,  2.04it/s]
 35%|███▍      | 3710/10712 [51:17<57:13,  2.04it/s]
 35%|███▍      | 3711/10712 [51:18<57:16,  2.04it/s]
 35%|███▍      | 3712/10712 [51:18<57:13,  2.04it/s]
 35%|███▍      | 3713/10712 [51:19<57:36,  2.03it/s]
 35%|███▍      | 3714/10712 [51:19<57:28,  2.03it/s]
 35%|███▍      | 3715/10712 [51:20<57:18,  2.03it/s]
 35%|███▍      | 3716/10712 [51:20<57:14,  2.04it/s]
 35%|███▍      | 3717/10712 [51:21<57:15,  2.04it/s]
 35%|███▍      | 3718/10712 [51:21<57:12,  2.04it/s]
 35%|███▍      | 3719/10712 [51:22<57:08,  2.04it/s]
 35%|███▍      | 3720/10712 [51:22<57:11,  2.04it/s]
 35%|███▍      | 3721/10712 [51:23<57:09,  2.04it/s]
 35%|███▍      | 3722/10712 [51:23<57:13,  2.04it/s]
 35%|███▍      | 3723/10712 [51:24<57:13,  2.04it/s]
 35%|███▍      | 3724/10712 [51:24<57:11,  2.04it/s]
 35%|███▍      | 3725/10712 [51:25<57:06,  2.04it/s]{'loss': 3.8632, 'grad_norm': 0.2376428097486496, 'learning_rate': 0.0008244761454780962, 'epoch': 0.35}

                                                    

 35%|███▍      | 3725/10712 [51:25<57:06,  2.04it/s]
 35%|███▍      | 3726/10712 [51:25<57:13,  2.03it/s]
 35%|███▍      | 3727/10712 [51:26<57:16,  2.03it/s]
 35%|███▍      | 3728/10712 [51:26<57:12,  2.03it/s]
 35%|███▍      | 3729/10712 [51:27<57:12,  2.03it/s]
 35%|███▍      | 3730/10712 [51:27<57:08,  2.04it/s]
 35%|███▍      | 3731/10712 [51:28<57:02,  2.04it/s]
 35%|███▍      | 3732/10712 [51:28<57:35,  2.02it/s]
 35%|███▍      | 3733/10712 [51:29<57:28,  2.02it/s]
 35%|███▍      | 3734/10712 [51:29<57:19,  2.03it/s]
 35%|███▍      | 3735/10712 [51:30<57:13,  2.03it/s]
 35%|███▍      | 3736/10712 [51:30<57:09,  2.03it/s]
 35%|███▍      | 3737/10712 [51:31<57:07,  2.04it/s]
 35%|███▍      | 3738/10712 [51:31<57:34,  2.02it/s]
 35%|███▍      | 3739/10712 [51:32<57:20,  2.03it/s]
 35%|███▍      | 3740/10712 [51:32<57:16,  2.03it/s]
 35%|███▍      | 3741/10712 [51:33<57:10,  2.03it/s]
 35%|███▍      | 3742/10712 [51:33<57:09,  2.03it/s]
 35%|███▍      | 3743/10712 [51:34<57:02,  2.04it/s]
 35%|███▍      | 3744/10712 [51:34<57:02,  2.04it/s]
 35%|███▍      | 3745/10712 [51:35<57:04,  2.03it/s]
 35%|███▍      | 3746/10712 [51:35<56:59,  2.04it/s]
 35%|███▍      | 3747/10712 [51:36<57:00,  2.04it/s]
 35%|███▍      | 3748/10712 [51:36<57:03,  2.03it/s]
 35%|███▍      | 3749/10712 [51:37<57:00,  2.04it/s]
 35%|███▌      | 3750/10712 [51:37<56:57,  2.04it/s]{'loss': 3.8658, 'grad_norm': 0.21243515610694885, 'learning_rate': 0.0008213660654159454, 'epoch': 0.35}

                                                    

 35%|███▌      | 3750/10712 [51:37<56:57,  2.04it/s]
 35%|███▌      | 3751/10712 [51:38<57:04,  2.03it/s]
 35%|███▌      | 3752/10712 [51:38<56:59,  2.04it/s]
 35%|███▌      | 3753/10712 [51:39<56:49,  2.04it/s]
 35%|███▌      | 3754/10712 [51:39<56:53,  2.04it/s]
 35%|███▌      | 3755/10712 [51:39<56:54,  2.04it/s]
 35%|███▌      | 3756/10712 [51:40<56:51,  2.04it/s]
 35%|███▌      | 3757/10712 [51:40<56:52,  2.04it/s]
 35%|███▌      | 3758/10712 [51:41<56:51,  2.04it/s]
 35%|███▌      | 3759/10712 [51:41<56:52,  2.04it/s]
 35%|███▌      | 3760/10712 [51:42<56:47,  2.04it/s]
 35%|███▌      | 3761/10712 [51:42<56:49,  2.04it/s]
 35%|███▌      | 3762/10712 [51:43<56:52,  2.04it/s]
 35%|███▌      | 3763/10712 [51:43<56:48,  2.04it/s]
 35%|███▌      | 3764/10712 [51:44<56:48,  2.04it/s]
 35%|███▌      | 3765/10712 [51:44<56:50,  2.04it/s]
 35%|███▌      | 3766/10712 [51:45<56:49,  2.04it/s]
 35%|███▌      | 3767/10712 [51:45<56:48,  2.04it/s]
 35%|███▌      | 3768/10712 [51:46<56:44,  2.04it/s]
 35%|███▌      | 3769/10712 [51:46<56:47,  2.04it/s]
 35%|███▌      | 3770/10712 [51:47<56:45,  2.04it/s]
 35%|███▌      | 3771/10712 [51:47<56:44,  2.04it/s]
 35%|███▌      | 3772/10712 [51:48<56:51,  2.03it/s]
 35%|███▌      | 3773/10712 [51:48<56:48,  2.04it/s]
 35%|███▌      | 3774/10712 [51:49<56:48,  2.04it/s]
 35%|███▌      | 3775/10712 [51:49<56:49,  2.03it/s]{'loss': 3.8659, 'grad_norm': 0.2505894899368286, 'learning_rate': 0.0008182346537598444, 'epoch': 0.35}

                                                    

 35%|███▌      | 3775/10712 [51:49<56:49,  2.03it/s]
 35%|███▌      | 3776/10712 [51:50<56:51,  2.03it/s]
 35%|███▌      | 3777/10712 [51:50<56:48,  2.03it/s]
 35%|███▌      | 3778/10712 [51:51<56:48,  2.03it/s]
 35%|███▌      | 3779/10712 [51:51<56:45,  2.04it/s]
 35%|███▌      | 3780/10712 [51:52<56:42,  2.04it/s]
 35%|███▌      | 3781/10712 [51:52<56:43,  2.04it/s]
 35%|███▌      | 3782/10712 [51:53<56:41,  2.04it/s]
 35%|███▌      | 3783/10712 [51:53<56:38,  2.04it/s]
 35%|███▌      | 3784/10712 [51:54<56:38,  2.04it/s]
 35%|███▌      | 3785/10712 [51:54<56:38,  2.04it/s]
 35%|███▌      | 3786/10712 [51:55<56:35,  2.04it/s]
 35%|███▌      | 3787/10712 [51:55<56:37,  2.04it/s]
 35%|███▌      | 3788/10712 [51:56<56:39,  2.04it/s]
 35%|███▌      | 3789/10712 [51:56<56:39,  2.04it/s]
 35%|███▌      | 3790/10712 [51:57<56:37,  2.04it/s]
 35%|███▌      | 3791/10712 [51:57<56:40,  2.04it/s]
 35%|███▌      | 3792/10712 [51:58<56:38,  2.04it/s]
 35%|███▌      | 3793/10712 [51:58<56:36,  2.04it/s]
 35%|███▌      | 3794/10712 [51:59<56:36,  2.04it/s]
 35%|███▌      | 3795/10712 [51:59<56:35,  2.04it/s]
 35%|███▌      | 3796/10712 [52:00<56:33,  2.04it/s]
 35%|███▌      | 3797/10712 [52:00<56:36,  2.04it/s]
 35%|███▌      | 3798/10712 [52:01<56:37,  2.04it/s]
 35%|███▌      | 3799/10712 [52:01<56:35,  2.04it/s]
 35%|███▌      | 3800/10712 [52:02<56:32,  2.04it/s]
                                                    {'loss': 3.8585, 'grad_norm': 0.2098558247089386, 'learning_rate': 0.0008150821183662201, 'epoch': 0.35}


 35%|███▌      | 3800/10712 [52:02<56:32,  2.04it/s]
 35%|███▌      | 3801/10712 [52:02<56:38,  2.03it/s]
 35%|███▌      | 3802/10712 [52:03<56:37,  2.03it/s]
 36%|███▌      | 3803/10712 [52:03<56:33,  2.04it/s]
 36%|███▌      | 3804/10712 [52:04<56:29,  2.04it/s]
 36%|███▌      | 3805/10712 [52:04<56:34,  2.03it/s]
 36%|███▌      | 3806/10712 [52:05<56:32,  2.04it/s]
 36%|███▌      | 3807/10712 [52:05<56:29,  2.04it/s]
 36%|███▌      | 3808/10712 [52:06<56:28,  2.04it/s]
 36%|███▌      | 3809/10712 [52:06<56:28,  2.04it/s]
 36%|███▌      | 3810/10712 [52:07<56:31,  2.04it/s]
 36%|███▌      | 3811/10712 [52:07<56:29,  2.04it/s]
 36%|███▌      | 3812/10712 [52:07<56:25,  2.04it/s]
 36%|███▌      | 3813/10712 [52:08<56:26,  2.04it/s]
 36%|███▌      | 3814/10712 [52:08<56:28,  2.04it/s]
 36%|███▌      | 3815/10712 [52:09<56:26,  2.04it/s]
 36%|███▌      | 3816/10712 [52:09<56:27,  2.04it/s]
 36%|███▌      | 3817/10712 [52:10<56:24,  2.04it/s]
 36%|███▌      | 3818/10712 [52:10<56:22,  2.04it/s]
 36%|███▌      | 3819/10712 [52:11<56:20,  2.04it/s]
 36%|███▌      | 3820/10712 [52:11<56:24,  2.04it/s]
 36%|███▌      | 3821/10712 [52:12<56:22,  2.04it/s]
 36%|███▌      | 3822/10712 [52:12<56:21,  2.04it/s]
 36%|███▌      | 3823/10712 [52:13<56:19,  2.04it/s]
 36%|███▌      | 3824/10712 [52:13<56:18,  2.04it/s]
 36%|███▌      | 3825/10712 [52:14<56:16,  2.04it/s]{'loss': 3.8721, 'grad_norm': 0.211765319108963, 'learning_rate': 0.0008119086684936474, 'epoch': 0.36}

                                                    

 36%|███▌      | 3825/10712 [52:14<56:16,  2.04it/s]
 36%|███▌      | 3826/10712 [52:14<56:17,  2.04it/s]
 36%|███▌      | 3827/10712 [52:15<56:16,  2.04it/s]
 36%|███▌      | 3828/10712 [52:15<56:17,  2.04it/s]
 36%|███▌      | 3829/10712 [52:16<56:15,  2.04it/s]
 36%|███▌      | 3830/10712 [52:16<56:19,  2.04it/s]
 36%|███▌      | 3831/10712 [52:17<56:33,  2.03it/s]
 36%|███▌      | 3832/10712 [52:17<56:30,  2.03it/s]
 36%|███▌      | 3833/10712 [52:18<56:24,  2.03it/s]
 36%|███▌      | 3834/10712 [52:18<56:20,  2.03it/s]
 36%|███▌      | 3835/10712 [52:19<56:18,  2.04it/s]
 36%|███▌      | 3836/10712 [52:19<56:14,  2.04it/s]
 36%|███▌      | 3837/10712 [52:20<56:13,  2.04it/s]
 36%|███▌      | 3838/10712 [52:20<56:09,  2.04it/s]
 36%|███▌      | 3839/10712 [52:21<56:12,  2.04it/s]
 36%|███▌      | 3840/10712 [52:21<56:10,  2.04it/s]
 36%|███▌      | 3841/10712 [52:22<56:07,  2.04it/s]
 36%|███▌      | 3842/10712 [52:22<56:23,  2.03it/s]
 36%|███▌      | 3843/10712 [52:23<56:14,  2.04it/s]
 36%|███▌      | 3844/10712 [52:23<56:12,  2.04it/s]
 36%|███▌      | 3845/10712 [52:24<56:18,  2.03it/s]
 36%|███▌      | 3846/10712 [52:24<56:11,  2.04it/s]
 36%|███▌      | 3847/10712 [52:25<56:12,  2.04it/s]
 36%|███▌      | 3848/10712 [52:25<56:12,  2.04it/s]
 36%|███▌      | 3849/10712 [52:26<56:09,  2.04it/s]
 36%|███▌      | 3850/10712 [52:26<56:08,  2.04it/s]{'loss': 3.8558, 'grad_norm': 0.21458415687084198, 'learning_rate': 0.00080871451478896, 'epoch': 0.36}

                                                    

 36%|███▌      | 3850/10712 [52:26<56:08,  2.04it/s]
 36%|███▌      | 3851/10712 [52:27<56:15,  2.03it/s]
 36%|███▌      | 3852/10712 [52:27<56:08,  2.04it/s]
 36%|███▌      | 3853/10712 [52:28<56:05,  2.04it/s]
 36%|███▌      | 3854/10712 [52:28<56:11,  2.03it/s]
 36%|███▌      | 3855/10712 [52:29<56:08,  2.04it/s]
 36%|███▌      | 3856/10712 [52:29<56:06,  2.04it/s]
 36%|███▌      | 3857/10712 [52:30<56:08,  2.04it/s]
 36%|███▌      | 3858/10712 [52:30<1:05:13,  1.75it/s]
 36%|███▌      | 3859/10712 [52:31<1:02:29,  1.83it/s]
 36%|███▌      | 3860/10712 [52:31<1:00:33,  1.89it/s]
 36%|███▌      | 3861/10712 [52:32<59:07,  1.93it/s]  
 36%|███▌      | 3862/10712 [52:32<58:08,  1.96it/s]
 36%|███▌      | 3863/10712 [52:33<57:34,  1.98it/s]
 36%|███▌      | 3864/10712 [52:33<57:03,  2.00it/s]
 36%|███▌      | 3865/10712 [52:34<56:44,  2.01it/s]
 36%|███▌      | 3866/10712 [52:34<56:34,  2.02it/s]
 36%|███▌      | 3867/10712 [52:35<56:21,  2.02it/s]
 36%|███▌      | 3868/10712 [52:35<56:10,  2.03it/s]
 36%|███▌      | 3869/10712 [52:36<56:11,  2.03it/s]
 36%|███▌      | 3870/10712 [52:36<56:07,  2.03it/s]
 36%|███▌      | 3871/10712 [52:37<55:58,  2.04it/s]
 36%|███▌      | 3872/10712 [52:37<55:58,  2.04it/s]
 36%|███▌      | 3873/10712 [52:38<55:59,  2.04it/s]
 36%|███▌      | 3874/10712 [52:38<55:55,  2.04it/s]
 36%|███▌      | 3875/10712 [52:39<55:54,  2.04it/s]{'loss': 3.8561, 'grad_norm': 0.2359841912984848, 'learning_rate': 0.0008054998692732676, 'epoch': 0.36}

                                                    

 36%|███▌      | 3875/10712 [52:39<55:54,  2.04it/s]
 36%|███▌      | 3876/10712 [52:39<55:57,  2.04it/s]
 36%|███▌      | 3877/10712 [52:40<55:52,  2.04it/s]
 36%|███▌      | 3878/10712 [52:40<55:51,  2.04it/s]
 36%|███▌      | 3879/10712 [52:41<55:51,  2.04it/s]
 36%|███▌      | 3880/10712 [52:41<55:54,  2.04it/s]
 36%|███▌      | 3881/10712 [52:42<55:49,  2.04it/s]
 36%|███▌      | 3882/10712 [52:42<1:04:55,  1.75it/s]
 36%|███▌      | 3883/10712 [52:43<1:02:12,  1.83it/s]
 36%|███▋      | 3884/10712 [52:43<1:00:14,  1.89it/s]
 36%|███▋      | 3885/10712 [52:44<58:49,  1.93it/s]  
 36%|███▋      | 3886/10712 [52:44<57:54,  1.96it/s]
 36%|███▋      | 3887/10712 [52:45<57:18,  1.99it/s]
 36%|███▋      | 3888/10712 [52:45<56:50,  2.00it/s]
 36%|███▋      | 3889/10712 [52:46<56:27,  2.01it/s]
 36%|███▋      | 3890/10712 [52:46<56:20,  2.02it/s]
 36%|███▋      | 3891/10712 [52:47<56:13,  2.02it/s]
 36%|███▋      | 3892/10712 [52:47<56:06,  2.03it/s]
 36%|███▋      | 3893/10712 [52:48<56:04,  2.03it/s]
 36%|███▋      | 3894/10712 [52:48<55:56,  2.03it/s]
 36%|███▋      | 3895/10712 [52:49<55:49,  2.03it/s]
 36%|███▋      | 3896/10712 [52:49<55:47,  2.04it/s]
 36%|███▋      | 3897/10712 [52:50<55:45,  2.04it/s]
 36%|███▋      | 3898/10712 [52:50<55:44,  2.04it/s]
 36%|███▋      | 3899/10712 [52:51<55:41,  2.04it/s]
 36%|███▋      | 3900/10712 [52:51<55:39,  2.04it/s]{'loss': 3.8548, 'grad_norm': 0.23634569346904755, 'learning_rate': 0.0008022649453278827, 'epoch': 0.36}

                                                    

 36%|███▋      | 3900/10712 [52:51<55:39,  2.04it/s]
 36%|███▋      | 3901/10712 [52:52<55:45,  2.04it/s]
 36%|███▋      | 3902/10712 [52:52<55:42,  2.04it/s]
 36%|███▋      | 3903/10712 [52:53<55:41,  2.04it/s]
 36%|███▋      | 3904/10712 [52:53<55:43,  2.04it/s]
 36%|███▋      | 3905/10712 [52:54<55:41,  2.04it/s]
 36%|███▋      | 3906/10712 [52:54<55:41,  2.04it/s]
 36%|███▋      | 3907/10712 [52:55<55:41,  2.04it/s]
 36%|███▋      | 3908/10712 [52:55<55:38,  2.04it/s]
 36%|███▋      | 3909/10712 [52:56<55:37,  2.04it/s]
 37%|███▋      | 3910/10712 [52:56<55:33,  2.04it/s]
 37%|███▋      | 3911/10712 [52:57<55:36,  2.04it/s]
 37%|███▋      | 3912/10712 [52:57<55:36,  2.04it/s]
 37%|███▋      | 3913/10712 [52:58<55:32,  2.04it/s]
 37%|███▋      | 3914/10712 [52:58<55:35,  2.04it/s]
 37%|███▋      | 3915/10712 [52:59<55:34,  2.04it/s]
 37%|███▋      | 3916/10712 [52:59<55:33,  2.04it/s]
 37%|███▋      | 3917/10712 [53:00<55:33,  2.04it/s]
 37%|███▋      | 3918/10712 [53:00<55:35,  2.04it/s]
 37%|███▋      | 3919/10712 [53:01<55:34,  2.04it/s]
 37%|███▋      | 3920/10712 [53:01<55:28,  2.04it/s]
 37%|███▋      | 3921/10712 [53:02<55:29,  2.04it/s]
 37%|███▋      | 3922/10712 [53:02<55:27,  2.04it/s]
 37%|███▋      | 3923/10712 [53:02<55:28,  2.04it/s]
 37%|███▋      | 3924/10712 [53:03<55:28,  2.04it/s]
 37%|███▋      | 3925/10712 [53:03<55:28,  2.04it/s]{'loss': 3.8526, 'grad_norm': 0.2304372489452362, 'learning_rate': 0.0007990099576801568, 'epoch': 0.37}

                                                    

 37%|███▋      | 3925/10712 [53:03<55:28,  2.04it/s]
 37%|███▋      | 3926/10712 [53:04<55:31,  2.04it/s]
 37%|███▋      | 3927/10712 [53:04<55:27,  2.04it/s]
 37%|███▋      | 3928/10712 [53:05<55:28,  2.04it/s]
 37%|███▋      | 3929/10712 [53:05<55:24,  2.04it/s]
 37%|███▋      | 3930/10712 [53:06<55:23,  2.04it/s]
 37%|███▋      | 3931/10712 [53:06<55:22,  2.04it/s]
 37%|███▋      | 3932/10712 [53:07<55:27,  2.04it/s]
 37%|███▋      | 3933/10712 [53:07<55:27,  2.04it/s]
 37%|███▋      | 3934/10712 [53:08<55:23,  2.04it/s]
 37%|███▋      | 3935/10712 [53:08<55:23,  2.04it/s]
 37%|███▋      | 3936/10712 [53:09<55:25,  2.04it/s]
 37%|███▋      | 3937/10712 [53:09<55:22,  2.04it/s]
 37%|███▋      | 3938/10712 [53:10<55:21,  2.04it/s]
 37%|███▋      | 3939/10712 [53:10<55:23,  2.04it/s]
 37%|███▋      | 3940/10712 [53:11<55:28,  2.03it/s]
 37%|███▋      | 3941/10712 [53:11<55:24,  2.04it/s]
 37%|███▋      | 3942/10712 [53:12<55:20,  2.04it/s]
 37%|███▋      | 3943/10712 [53:12<55:20,  2.04it/s]
 37%|███▋      | 3944/10712 [53:13<55:23,  2.04it/s]
 37%|███▋      | 3945/10712 [53:13<55:22,  2.04it/s]
 37%|███▋      | 3946/10712 [53:14<55:21,  2.04it/s]
 37%|███▋      | 3947/10712 [53:14<55:19,  2.04it/s]
 37%|███▋      | 3948/10712 [53:15<55:26,  2.03it/s]
 37%|███▋      | 3949/10712 [53:15<55:24,  2.03it/s]
 37%|███▋      | 3950/10712 [53:16<55:18,  2.04it/s]
                                                    {'loss': 3.8583, 'grad_norm': 0.22167952358722687, 'learning_rate': 0.0007957351223892271, 'epoch': 0.37}


 37%|███▋      | 3950/10712 [53:16<55:18,  2.04it/s]
 37%|███▋      | 3951/10712 [53:16<55:22,  2.03it/s]
 37%|███▋      | 3952/10712 [53:17<55:22,  2.03it/s]
 37%|███▋      | 3953/10712 [53:17<55:17,  2.04it/s]
 37%|███▋      | 3954/10712 [53:18<55:12,  2.04it/s]
 37%|███▋      | 3955/10712 [53:18<55:13,  2.04it/s]
 37%|███▋      | 3956/10712 [53:19<55:16,  2.04it/s]
 37%|███▋      | 3957/10712 [53:19<55:16,  2.04it/s]
 37%|███▋      | 3958/10712 [53:20<55:17,  2.04it/s]
 37%|███▋      | 3959/10712 [53:20<55:15,  2.04it/s]
 37%|███▋      | 3960/10712 [53:21<55:13,  2.04it/s]
 37%|███▋      | 3961/10712 [53:21<55:15,  2.04it/s]
 37%|███▋      | 3962/10712 [53:22<55:16,  2.04it/s]
 37%|███▋      | 3963/10712 [53:22<55:10,  2.04it/s]
 37%|███▋      | 3964/10712 [53:23<55:13,  2.04it/s]
 37%|███▋      | 3965/10712 [53:23<55:13,  2.04it/s]
 37%|███▋      | 3966/10712 [53:24<55:12,  2.04it/s]
 37%|███▋      | 3967/10712 [53:24<55:11,  2.04it/s]
 37%|███▋      | 3968/10712 [53:25<55:13,  2.04it/s]
 37%|███▋      | 3969/10712 [53:25<55:14,  2.03it/s]
 37%|███▋      | 3970/10712 [53:26<55:12,  2.04it/s]
 37%|███▋      | 3971/10712 [53:26<55:10,  2.04it/s]
 37%|███▋      | 3972/10712 [53:27<55:05,  2.04it/s]
 37%|███▋      | 3973/10712 [53:27<55:07,  2.04it/s]
 37%|███▋      | 3974/10712 [53:28<55:08,  2.04it/s]
 37%|███▋      | 3975/10712 [53:28<55:07,  2.04it/s]{'loss': 3.8559, 'grad_norm': 0.21915723383426666, 'learning_rate': 0.0007924406568316749, 'epoch': 0.37}

                                                    

 37%|███▋      | 3975/10712 [53:28<55:07,  2.04it/s]
 37%|███▋      | 3976/10712 [53:29<55:12,  2.03it/s]
 37%|███▋      | 3977/10712 [53:29<55:10,  2.03it/s]
 37%|███▋      | 3978/10712 [53:29<55:08,  2.04it/s]
 37%|███▋      | 3979/10712 [53:30<55:06,  2.04it/s]
 37%|███▋      | 3980/10712 [53:30<55:10,  2.03it/s]
 37%|███▋      | 3981/10712 [53:31<55:07,  2.04it/s]
 37%|███▋      | 3982/10712 [53:31<55:04,  2.04it/s]
 37%|███▋      | 3983/10712 [53:32<55:07,  2.03it/s]
 37%|███▋      | 3984/10712 [53:32<55:04,  2.04it/s]
 37%|███▋      | 3985/10712 [53:33<55:03,  2.04it/s]
 37%|███▋      | 3986/10712 [53:33<55:02,  2.04it/s]
 37%|███▋      | 3987/10712 [53:34<54:58,  2.04it/s]
 37%|███▋      | 3988/10712 [53:34<54:59,  2.04it/s]
 37%|███▋      | 3989/10712 [53:35<55:04,  2.03it/s]
 37%|███▋      | 3990/10712 [53:35<55:03,  2.03it/s]
 37%|███▋      | 3991/10712 [53:36<55:02,  2.04it/s]
 37%|███▋      | 3992/10712 [53:36<54:58,  2.04it/s]
 37%|███▋      | 3993/10712 [53:37<54:59,  2.04it/s]
 37%|███▋      | 3994/10712 [53:37<54:58,  2.04it/s]
 37%|███▋      | 3995/10712 [53:38<54:56,  2.04it/s]
 37%|███▋      | 3996/10712 [53:38<54:56,  2.04it/s]
 37%|███▋      | 3997/10712 [53:39<54:55,  2.04it/s]
 37%|███▋      | 3998/10712 [53:39<55:11,  2.03it/s]
 37%|███▋      | 3999/10712 [53:40<55:04,  2.03it/s]
 37%|███▋      | 4000/10712 [53:40<55:01,  2.03it/s]{'loss': 3.8446, 'grad_norm': 0.21080200374126434, 'learning_rate': 0.0007891267796870969, 'epoch': 0.37}

                                                    

 37%|███▋      | 4000/10712 [53:40<55:01,  2.03it/s]
 37%|███▋      | 4001/10712 [53:41<55:00,  2.03it/s]
 37%|███▋      | 4002/10712 [53:41<54:55,  2.04it/s]
 37%|███▋      | 4003/10712 [53:42<54:55,  2.04it/s]
 37%|███▋      | 4004/10712 [53:42<54:54,  2.04it/s]
 37%|███▋      | 4005/10712 [53:43<54:49,  2.04it/s]
 37%|███▋      | 4006/10712 [53:43<54:51,  2.04it/s]
 37%|███▋      | 4007/10712 [53:44<54:51,  2.04it/s]
 37%|███▋      | 4008/10712 [53:44<54:50,  2.04it/s]
 37%|███▋      | 4009/10712 [53:45<54:48,  2.04it/s]
 37%|███▋      | 4010/10712 [53:45<54:53,  2.03it/s]
 37%|███▋      | 4011/10712 [53:46<54:48,  2.04it/s]
 37%|███▋      | 4012/10712 [53:46<54:44,  2.04it/s]
 37%|███▋      | 4013/10712 [53:47<54:47,  2.04it/s]
 37%|███▋      | 4014/10712 [53:47<54:46,  2.04it/s]
 37%|███▋      | 4015/10712 [53:48<54:46,  2.04it/s]
 37%|███▋      | 4016/10712 [53:48<54:50,  2.03it/s]
 38%|███▊      | 4017/10712 [53:49<54:47,  2.04it/s]
 38%|███▊      | 4018/10712 [53:49<54:45,  2.04it/s]
 38%|███▊      | 4019/10712 [53:50<54:45,  2.04it/s]
 38%|███▊      | 4020/10712 [53:50<54:45,  2.04it/s]
 38%|███▊      | 4021/10712 [53:51<54:42,  2.04it/s]
 38%|███▊      | 4022/10712 [53:51<54:39,  2.04it/s]
 38%|███▊      | 4023/10712 [53:52<54:43,  2.04it/s]
 38%|███▊      | 4024/10712 [53:52<54:42,  2.04it/s]
 38%|███▊      | 4025/10712 [53:53<54:40,  2.04it/s]
                                                    
{'loss': 3.8437, 'grad_norm': 0.21651799976825714, 'learning_rate': 0.0007857937109235893, 'epoch': 0.38}

 38%|███▊      | 4025/10712 [53:53<54:40,  2.04it/s]
 38%|███▊      | 4026/10712 [53:53<54:47,  2.03it/s]
 38%|███▊      | 4027/10712 [53:54<54:44,  2.04it/s]
 38%|███▊      | 4028/10712 [53:54<54:43,  2.04it/s]
 38%|███▊      | 4029/10712 [53:55<54:39,  2.04it/s]
 38%|███▊      | 4030/10712 [53:55<54:42,  2.04it/s]
 38%|███▊      | 4031/10712 [53:56<54:41,  2.04it/s]
 38%|███▊      | 4032/10712 [53:56<54:42,  2.04it/s]
 38%|███▊      | 4033/10712 [53:57<54:42,  2.03it/s]
 38%|███▊      | 4034/10712 [53:57<54:37,  2.04it/s]
 38%|███▊      | 4035/10712 [53:57<54:36,  2.04it/s]
 38%|███▊      | 4036/10712 [53:58<54:36,  2.04it/s]
 38%|███▊      | 4037/10712 [53:58<54:33,  2.04it/s]
 38%|███▊      | 4038/10712 [53:59<54:33,  2.04it/s]
 38%|███▊      | 4039/10712 [53:59<54:30,  2.04it/s]
 38%|███▊      | 4040/10712 [54:00<54:34,  2.04it/s]
 38%|███▊      | 4041/10712 [54:00<54:33,  2.04it/s]
 38%|███▊      | 4042/10712 [54:01<54:35,  2.04it/s]
 38%|███▊      | 4043/10712 [54:01<54:34,  2.04it/s]
 38%|███▊      | 4044/10712 [54:02<54:30,  2.04it/s]
 38%|███▊      | 4045/10712 [54:02<54:27,  2.04it/s]
 38%|███▊      | 4046/10712 [54:03<54:31,  2.04it/s]
 38%|███▊      | 4047/10712 [54:03<54:29,  2.04it/s]
 38%|███▊      | 4048/10712 [54:04<54:27,  2.04it/s]
 38%|███▊      | 4049/10712 [54:04<54:26,  2.04it/s]
 38%|███▊      | 4050/10712 [54:05<54:28,  2.04it/s]
                                                    {'loss': 3.8443, 'grad_norm': 0.22546787559986115, 'learning_rate': 0.0007824416717831475, 'epoch': 0.38}

 38%|███▊      | 4050/10712 [54:05<54:28,  2.04it/s]
 38%|███▊      | 4051/10712 [54:05<54:30,  2.04it/s]
 38%|███▊      | 4052/10712 [54:06<54:31,  2.04it/s]
 38%|███▊      | 4053/10712 [54:06<54:31,  2.04it/s]
 38%|███▊      | 4054/10712 [54:07<54:27,  2.04it/s]
 38%|███▊      | 4055/10712 [54:07<54:28,  2.04it/s]
 38%|███▊      | 4056/10712 [54:08<54:31,  2.03it/s]
 38%|███▊      | 4057/10712 [54:08<54:26,  2.04it/s]
 38%|███▊      | 4058/10712 [54:09<54:26,  2.04it/s]
 38%|███▊      | 4059/10712 [54:09<54:25,  2.04it/s]
 38%|███▊      | 4060/10712 [54:10<54:27,  2.04it/s]
 38%|███▊      | 4061/10712 [54:10<54:25,  2.04it/s]
 38%|███▊      | 4062/10712 [54:11<54:24,  2.04it/s]
 38%|███▊      | 4063/10712 [54:11<54:26,  2.04it/s]
 38%|███▊      | 4064/10712 [54:12<54:21,  2.04it/s]
 38%|███▊      | 4065/10712 [54:12<54:23,  2.04it/s]
 38%|███▊      | 4066/10712 [54:13<54:26,  2.03it/s]
 38%|███▊      | 4067/10712 [54:13<54:22,  2.04it/s]
 38%|███▊      | 4068/10712 [54:14<54:22,  2.04it/s]
 38%|███▊      | 4069/10712 [54:14<54:25,  2.03it/s]
 38%|███▊      | 4070/10712 [54:15<54:22,  2.04it/s]
 38%|███▊      | 4071/10712 [54:15<54:22,  2.04it/s]
 38%|███▊      | 4072/10712 [54:16<54:18,  2.04it/s]
 38%|███▊      | 4073/10712 [54:16<54:19,  2.04it/s]
 38%|███▊      | 4074/10712 [54:17<54:48,  2.02it/s]
 38%|███▊      | 4075/10712 [54:17<54:38,  2.02it/s]{'loss': 3.8406, 'grad_norm': 0.223909392952919, 'learning_rate': 0.0007790708847669796, 'epoch': 0.38}

                                                    

 38%|███▊      | 4075/10712 [54:17<54:38,  2.02it/s]
 38%|███▊      | 4076/10712 [54:18<54:32,  2.03it/s]
 38%|███▊      | 4077/10712 [54:18<54:32,  2.03it/s]
 38%|███▊      | 4078/10712 [54:19<54:27,  2.03it/s]
 38%|███▊      | 4079/10712 [54:19<54:20,  2.03it/s]
 38%|███▊      | 4080/10712 [54:20<54:19,  2.03it/s]
 38%|███▊      | 4081/10712 [54:20<54:24,  2.03it/s]
 38%|███▊      | 4082/10712 [54:21<54:20,  2.03it/s]
 38%|███▊      | 4083/10712 [54:21<54:17,  2.03it/s]
 38%|███▊      | 4084/10712 [54:22<54:18,  2.03it/s]
 38%|███▊      | 4085/10712 [54:22<54:18,  2.03it/s]
 38%|███▊      | 4086/10712 [54:23<54:17,  2.03it/s]
 38%|███▊      | 4087/10712 [54:23<54:15,  2.04it/s]
 38%|███▊      | 4088/10712 [54:24<54:15,  2.03it/s]
 38%|███▊      | 4089/10712 [54:24<54:15,  2.03it/s]
 38%|███▊      | 4090/10712 [54:25<54:14,  2.03it/s]
 38%|███▊      | 4091/10712 [54:25<54:17,  2.03it/s]
 38%|███▊      | 4092/10712 [54:25<54:15,  2.03it/s]
 38%|███▊      | 4093/10712 [54:26<54:11,  2.04it/s]
 38%|███▊      | 4094/10712 [54:26<54:11,  2.04it/s]
 38%|███▊      | 4095/10712 [54:27<54:08,  2.04it/s]
 38%|███▊      | 4096/10712 [54:27<54:05,  2.04it/s]
 38%|███▊      | 4097/10712 [54:28<54:09,  2.04it/s]
 38%|███▊      | 4098/10712 [54:28<54:09,  2.04it/s]
 38%|███▊      | 4099/10712 [54:29<54:05,  2.04it/s]
 38%|███▊      | 4100/10712 [54:29<54:04,  2.04it/s]{'loss': 3.8358, 'grad_norm': 0.21768060326576233, 'learning_rate': 0.0007756815736207379, 'epoch': 0.38}

                                                    

 38%|███▊      | 4100/10712 [54:29<54:04,  2.04it/s]
 38%|███▊      | 4101/10712 [54:30<54:09,  2.03it/s]
 38%|███▊      | 4102/10712 [54:30<54:06,  2.04it/s]
 38%|███▊      | 4103/10712 [54:31<54:06,  2.04it/s]
 38%|███▊      | 4104/10712 [54:31<54:05,  2.04it/s]
 38%|███▊      | 4105/10712 [54:32<54:03,  2.04it/s]
 38%|███▊      | 4106/10712 [54:32<54:06,  2.03it/s]
 38%|███▊      | 4107/10712 [54:33<54:04,  2.04it/s]
 38%|███▊      | 4108/10712 [54:33<54:03,  2.04it/s]
 38%|███▊      | 4109/10712 [54:34<54:05,  2.03it/s]
 38%|███▊      | 4110/10712 [54:34<54:04,  2.04it/s]
 38%|███▊      | 4111/10712 [54:35<53:59,  2.04it/s]
 38%|███▊      | 4112/10712 [54:35<54:02,  2.04it/s]
 38%|███▊      | 4113/10712 [54:36<54:01,  2.04it/s]
 38%|███▊      | 4114/10712 [54:36<53:57,  2.04it/s]
 38%|███▊      | 4115/10712 [54:37<54:01,  2.04it/s]
 38%|███▊      | 4116/10712 [54:37<53:58,  2.04it/s]
 38%|███▊      | 4117/10712 [54:38<53:54,  2.04it/s]
 38%|███▊      | 4118/10712 [54:38<54:00,  2.03it/s]
 38%|███▊      | 4119/10712 [54:39<54:22,  2.02it/s]
 38%|███▊      | 4120/10712 [54:39<54:13,  2.03it/s]
 38%|███▊      | 4121/10712 [54:40<54:07,  2.03it/s]
 38%|███▊      | 4122/10712 [54:40<54:06,  2.03it/s]
 38%|███▊      | 4123/10712 [54:41<54:02,  2.03it/s]
 38%|███▊      | 4124/10712 [54:41<53:57,  2.03it/s]
 39%|███▊      | 4125/10712 [54:42<53:55,  2.04it/s]
                                                    {'loss': 3.8331, 'grad_norm': 0.22223131358623505, 'learning_rate': 0.000772273963319667, 'epoch': 0.39}


 39%|███▊      | 4125/10712 [54:42<53:55,  2.04it/s]
 39%|███▊      | 4126/10712 [54:42<54:02,  2.03it/s]
 39%|███▊      | 4127/10712 [54:43<53:58,  2.03it/s]
 39%|███▊      | 4128/10712 [54:43<53:57,  2.03it/s]
 39%|███▊      | 4129/10712 [54:44<53:54,  2.04it/s]
 39%|███▊      | 4130/10712 [54:44<53:52,  2.04it/s]
 39%|███▊      | 4131/10712 [54:45<53:58,  2.03it/s]
 39%|███▊      | 4132/10712 [54:45<53:53,  2.04it/s]
 39%|███▊      | 4133/10712 [54:46<53:53,  2.03it/s]
 39%|███▊      | 4134/10712 [54:46<53:55,  2.03it/s]
 39%|███▊      | 4135/10712 [54:47<53:53,  2.03it/s]
 39%|███▊      | 4136/10712 [54:47<54:10,  2.02it/s]
 39%|███▊      | 4137/10712 [54:48<54:03,  2.03it/s]
 39%|███▊      | 4138/10712 [54:48<54:00,  2.03it/s]
 39%|███▊      | 4139/10712 [54:49<53:55,  2.03it/s]
 39%|███▊      | 4140/10712 [54:49<53:51,  2.03it/s]
 39%|███▊      | 4141/10712 [54:50<53:54,  2.03it/s]
 39%|███▊      | 4142/10712 [54:50<53:50,  2.03it/s]
 39%|███▊      | 4143/10712 [54:51<53:44,  2.04it/s]
 39%|███▊      | 4144/10712 [54:51<53:46,  2.04it/s]
 39%|███▊      | 4145/10712 [54:52<53:44,  2.04it/s]
 39%|███▊      | 4146/10712 [54:52<53:44,  2.04it/s]
 39%|███▊      | 4147/10712 [54:53<53:38,  2.04it/s]
 39%|███▊      | 4148/10712 [54:53<53:41,  2.04it/s]
 39%|███▊      | 4149/10712 [54:54<53:39,  2.04it/s]
 39%|███▊      | 4150/10712 [54:54<53:37,  2.04it/s]{'loss': 3.8345, 'grad_norm': 0.26229310035705566, 'learning_rate': 0.0007688482800536699, 'epoch': 0.39}
                                                    

 39%|███▊      | 4150/10712 [54:54<53:37,  2.04it/s]
 39%|███▉      | 4151/10712 [54:54<53:41,  2.04it/s]
 39%|███▉      | 4152/10712 [54:55<53:41,  2.04it/s]
 39%|███▉      | 4153/10712 [54:55<53:39,  2.04it/s]
 39%|███▉      | 4154/10712 [54:56<53:35,  2.04it/s]
 39%|███▉      | 4155/10712 [54:56<53:39,  2.04it/s]
 39%|███▉      | 4156/10712 [54:57<53:37,  2.04it/s]
 39%|███▉      | 4157/10712 [54:57<53:35,  2.04it/s]
 39%|███▉      | 4158/10712 [54:58<53:35,  2.04it/s]
 39%|███▉      | 4159/10712 [54:58<53:35,  2.04it/s]
 39%|███▉      | 4160/10712 [54:59<53:33,  2.04it/s]
 39%|███▉      | 4161/10712 [54:59<53:33,  2.04it/s]
 39%|███▉      | 4162/10712 [55:00<53:32,  2.04it/s]
 39%|███▉      | 4163/10712 [55:00<53:31,  2.04it/s]
 39%|███▉      | 4164/10712 [55:01<53:33,  2.04it/s]
 39%|███▉      | 4165/10712 [55:01<53:31,  2.04it/s]
 39%|███▉      | 4166/10712 [55:02<53:32,  2.04it/s]
 39%|███▉      | 4167/10712 [55:02<53:30,  2.04it/s]
 39%|███▉      | 4168/10712 [55:03<53:28,  2.04it/s]
 39%|███▉      | 4169/10712 [55:03<53:31,  2.04it/s]
 39%|███▉      | 4170/10712 [55:04<53:27,  2.04it/s]
 39%|███▉      | 4171/10712 [55:04<53:26,  2.04it/s]
 39%|███▉      | 4172/10712 [55:05<53:26,  2.04it/s]
 39%|███▉      | 4173/10712 [55:05<53:31,  2.04it/s]
 39%|███▉      | 4174/10712 [55:06<53:27,  2.04it/s]
 39%|███▉      | 4175/10712 [55:06<53:28,  2.04it/s]{'loss': 3.8363, 'grad_norm': 0.21218466758728027, 'learning_rate': 0.000765404751212295, 'epoch': 0.39}
                                                    

 39%|███▉      | 4175/10712 [55:06<53:28,  2.04it/s]
 39%|███▉      | 4176/10712 [55:07<53:32,  2.03it/s]
 39%|███▉      | 4177/10712 [55:07<53:30,  2.04it/s]
 39%|███▉      | 4178/10712 [55:08<53:25,  2.04it/s]
 39%|███▉      | 4179/10712 [55:08<53:23,  2.04it/s]
 39%|███▉      | 4180/10712 [55:09<53:26,  2.04it/s]
 39%|███▉      | 4181/10712 [55:09<53:23,  2.04it/s]
 39%|███▉      | 4182/10712 [55:10<53:21,  2.04it/s]
 39%|███▉      | 4183/10712 [55:10<53:22,  2.04it/s]
 39%|███▉      | 4184/10712 [55:11<53:26,  2.04it/s]
 39%|███▉      | 4185/10712 [55:11<53:24,  2.04it/s]
 39%|███▉      | 4186/10712 [55:12<53:27,  2.03it/s]
 39%|███▉      | 4187/10712 [55:12<53:25,  2.04it/s]
 39%|███▉      | 4188/10712 [55:13<53:22,  2.04it/s]
 39%|███▉      | 4189/10712 [55:13<53:25,  2.04it/s]
 39%|███▉      | 4190/10712 [55:14<53:22,  2.04it/s]
 39%|███▉      | 4191/10712 [55:14<53:21,  2.04it/s]
 39%|███▉      | 4192/10712 [55:15<53:24,  2.03it/s]
 39%|███▉      | 4193/10712 [55:15<53:21,  2.04it/s]
 39%|███▉      | 4194/10712 [55:16<53:16,  2.04it/s]
 39%|███▉      | 4195/10712 [55:16<53:15,  2.04it/s]
 39%|███▉      | 4196/10712 [55:17<53:17,  2.04it/s]
 39%|███▉      | 4197/10712 [55:17<53:15,  2.04it/s]
 39%|███▉      | 4198/10712 [55:18<53:12,  2.04it/s]
 39%|███▉      | 4199/10712 [55:18<53:13,  2.04it/s]
 39%|███▉      | 4200/10712 [55:19<53:15,  2.04it/s]{'loss': 3.8255, 'grad_norm': 0.1989186704158783, 'learning_rate': 0.0007619436053696421, 'epoch': 0.39}

                                                    

 39%|███▉      | 4200/10712 [55:19<53:15,  2.04it/s]
 39%|███▉      | 4201/10712 [55:19<53:16,  2.04it/s]
 39%|███▉      | 4202/10712 [55:20<53:18,  2.04it/s]
 39%|███▉      | 4203/10712 [55:20<53:15,  2.04it/s]
 39%|███▉      | 4204/10712 [55:20<53:10,  2.04it/s]
 39%|███▉      | 4205/10712 [55:21<53:10,  2.04it/s]
 39%|███▉      | 4206/10712 [55:21<53:09,  2.04it/s]
 39%|███▉      | 4207/10712 [55:22<53:09,  2.04it/s]
 39%|███▉      | 4208/10712 [55:22<53:06,  2.04it/s]
 39%|███▉      | 4209/10712 [55:23<53:05,  2.04it/s]
 39%|███▉      | 4210/10712 [55:23<53:12,  2.04it/s]
 39%|███▉      | 4211/10712 [55:24<53:10,  2.04it/s]
 39%|███▉      | 4212/10712 [55:24<53:10,  2.04it/s]
 39%|███▉      | 4213/10712 [55:25<53:13,  2.03it/s]
 39%|███▉      | 4214/10712 [55:25<53:14,  2.03it/s]
 39%|███▉      | 4215/10712 [55:26<53:12,  2.04it/s]
 39%|███▉      | 4216/10712 [55:26<53:09,  2.04it/s]
 39%|███▉      | 4217/10712 [55:27<53:13,  2.03it/s]
 39%|███▉      | 4218/10712 [55:27<53:10,  2.04it/s]
 39%|███▉      | 4219/10712 [55:28<53:08,  2.04it/s]
 39%|███▉      | 4220/10712 [55:28<53:06,  2.04it/s]
 39%|███▉      | 4221/10712 [55:29<53:10,  2.03it/s]
 39%|███▉      | 4222/10712 [55:29<53:08,  2.04it/s]
 39%|███▉      | 4223/10712 [55:30<53:04,  2.04it/s]
 39%|███▉      | 4224/10712 [55:30<53:07,  2.04it/s]
 39%|███▉      | 4225/10712 [55:31<53:07,  2.04it/s]{'loss': 3.8261, 'grad_norm': 0.20653612911701202, 'learning_rate': 0.0007584650722691891, 'epoch': 0.39}
                                                    

 39%|███▉      | 4225/10712 [55:31<53:07,  2.04it/s]
 39%|███▉      | 4226/10712 [55:31<53:06,  2.04it/s]
 39%|███▉      | 4227/10712 [55:32<53:10,  2.03it/s]
 39%|███▉      | 4228/10712 [55:32<53:07,  2.03it/s]
 39%|███▉      | 4229/10712 [55:33<53:01,  2.04it/s]
 39%|███▉      | 4230/10712 [55:33<53:02,  2.04it/s]
 39%|███▉      | 4231/10712 [55:34<52:58,  2.04it/s]
 40%|███▉      | 4232/10712 [55:34<52:56,  2.04it/s]
 40%|███▉      | 4233/10712 [55:35<52:54,  2.04it/s]
 40%|███▉      | 4234/10712 [55:35<52:56,  2.04it/s]
 40%|███▉      | 4235/10712 [55:36<52:56,  2.04it/s]
 40%|███▉      | 4236/10712 [55:36<52:53,  2.04it/s]
 40%|███▉      | 4237/10712 [55:37<52:53,  2.04it/s]
 40%|███▉      | 4238/10712 [55:37<52:54,  2.04it/s]
 40%|███▉      | 4239/10712 [55:38<52:54,  2.04it/s]
 40%|███▉      | 4240/10712 [55:38<52:51,  2.04it/s]
 40%|███▉      | 4241/10712 [55:39<52:54,  2.04it/s]
 40%|███▉      | 4242/10712 [55:39<52:53,  2.04it/s]
 40%|███▉      | 4243/10712 [55:40<52:50,  2.04it/s]
 40%|███▉      | 4244/10712 [55:40<52:48,  2.04it/s]
 40%|███▉      | 4245/10712 [55:41<52:51,  2.04it/s]
 40%|███▉      | 4246/10712 [55:41<52:50,  2.04it/s]
 40%|███▉      | 4247/10712 [55:42<52:49,  2.04it/s]
 40%|███▉      | 4248/10712 [55:42<52:53,  2.04it/s]
 40%|███▉      | 4249/10712 [55:43<52:52,  2.04it/s]
 40%|███▉      | 4250/10712 [55:43<52:54,  2.04it/s]{'loss': 3.8322, 'grad_norm': 0.20374652743339539, 'learning_rate': 0.0007549693828085437, 'epoch': 0.4}

                                                    

 40%|███▉      | 4250/10712 [55:43<52:54,  2.04it/s]
 40%|███▉      | 4251/10712 [55:44<52:58,  2.03it/s]
 40%|███▉      | 4252/10712 [55:44<52:55,  2.03it/s]
 40%|███▉      | 4253/10712 [55:45<52:51,  2.04it/s]
 40%|███▉      | 4254/10712 [55:45<52:51,  2.04it/s]
 40%|███▉      | 4255/10712 [55:46<52:50,  2.04it/s]
 40%|███▉      | 4256/10712 [55:46<52:47,  2.04it/s]
 40%|███▉      | 4257/10712 [55:47<52:52,  2.03it/s]
 40%|███▉      | 4258/10712 [55:47<52:51,  2.03it/s]
 40%|███▉      | 4259/10712 [55:47<52:48,  2.04it/s]
 40%|███▉      | 4260/10712 [55:48<52:51,  2.03it/s]
 40%|███▉      | 4261/10712 [55:48<52:49,  2.04it/s]
 40%|███▉      | 4262/10712 [55:49<52:45,  2.04it/s]
 40%|███▉      | 4263/10712 [55:49<52:47,  2.04it/s]
 40%|███▉      | 4264/10712 [55:50<52:43,  2.04it/s]
 40%|███▉      | 4265/10712 [55:50<52:42,  2.04it/s]
 40%|███▉      | 4266/10712 [55:51<52:43,  2.04it/s]
 40%|███▉      | 4267/10712 [55:51<52:44,  2.04it/s]
 40%|███▉      | 4268/10712 [55:52<52:41,  2.04it/s]
 40%|███▉      | 4269/10712 [55:52<52:44,  2.04it/s]
 40%|███▉      | 4270/10712 [55:53<52:40,  2.04it/s]
 40%|███▉      | 4271/10712 [55:53<52:39,  2.04it/s]
 40%|███▉      | 4272/10712 [55:54<52:37,  2.04it/s]
 40%|███▉      | 4273/10712 [55:54<52:40,  2.04it/s]
 40%|███▉      | 4274/10712 [55:55<52:36,  2.04it/s]
 40%|███▉      | 4275/10712 [55:55<52:37,  2.04it/s]
                                                    
{'loss': 3.8247, 'grad_norm': 0.2012351155281067, 'learning_rate': 0.0007514567690241163, 'epoch': 0.4}

 40%|███▉      | 4275/10712 [55:55<52:37,  2.04it/s]
 40%|███▉      | 4276/10712 [55:56<52:39,  2.04it/s]
 40%|███▉      | 4277/10712 [55:56<52:37,  2.04it/s]
 40%|███▉      | 4278/10712 [55:57<52:36,  2.04it/s]
 40%|███▉      | 4279/10712 [55:57<52:34,  2.04it/s]
 40%|███▉      | 4280/10712 [55:58<52:37,  2.04it/s]
 40%|███▉      | 4281/10712 [55:58<52:33,  2.04it/s]
 40%|███▉      | 4282/10712 [55:59<52:37,  2.04it/s]
 40%|███▉      | 4283/10712 [55:59<52:40,  2.03it/s]
 40%|███▉      | 4284/10712 [56:00<52:35,  2.04it/s]
 40%|████      | 4285/10712 [56:00<52:32,  2.04it/s]
 40%|████      | 4286/10712 [56:01<52:34,  2.04it/s]
 40%|████      | 4287/10712 [56:01<52:33,  2.04it/s]
 40%|████      | 4288/10712 [56:02<52:32,  2.04it/s]
 40%|████      | 4289/10712 [56:02<52:29,  2.04it/s]
 40%|████      | 4290/10712 [56:03<52:32,  2.04it/s]
 40%|████      | 4291/10712 [56:03<52:28,  2.04it/s]
 40%|████      | 4292/10712 [56:04<52:28,  2.04it/s]
 40%|████      | 4293/10712 [56:04<52:28,  2.04it/s]
 40%|████      | 4294/10712 [56:05<52:27,  2.04it/s]
 40%|████      | 4295/10712 [56:05<52:22,  2.04it/s]
 40%|████      | 4296/10712 [56:06<52:21,  2.04it/s]
 40%|████      | 4297/10712 [56:06<52:21,  2.04it/s]
 40%|████      | 4298/10712 [56:07<52:27,  2.04it/s]
 40%|████      | 4299/10712 [56:07<52:23,  2.04it/s]
 40%|████      | 4300/10712 [56:08<52:26,  2.04it/s]{'loss': 3.826, 'grad_norm': 0.20974954962730408, 'learning_rate': 0.0007479274640757172, 'epoch': 0.4}

                                                    

 40%|████      | 4300/10712 [56:08<52:26,  2.04it/s]
 40%|████      | 4301/10712 [56:08<52:28,  2.04it/s]
 40%|████      | 4302/10712 [56:09<52:27,  2.04it/s]
 40%|████      | 4303/10712 [56:09<52:26,  2.04it/s]
 40%|████      | 4304/10712 [56:10<52:25,  2.04it/s]
 40%|████      | 4305/10712 [56:10<52:23,  2.04it/s]
 40%|████      | 4306/10712 [56:11<52:18,  2.04it/s]
 40%|████      | 4307/10712 [56:11<52:23,  2.04it/s]
 40%|████      | 4308/10712 [56:12<52:24,  2.04it/s]
 40%|████      | 4309/10712 [56:12<52:22,  2.04it/s]
 40%|████      | 4310/10712 [56:13<52:24,  2.04it/s]
 40%|████      | 4311/10712 [56:13<52:24,  2.04it/s]
 40%|████      | 4312/10712 [56:13<52:22,  2.04it/s]
 40%|████      | 4313/10712 [56:14<52:20,  2.04it/s]
 40%|████      | 4314/10712 [56:14<52:24,  2.03it/s]
 40%|████      | 4315/10712 [56:15<52:23,  2.03it/s]
 40%|████      | 4316/10712 [56:15<52:21,  2.04it/s]
 40%|████      | 4317/10712 [56:16<52:25,  2.03it/s]
 40%|████      | 4318/10712 [56:16<52:22,  2.03it/s]
 40%|████      | 4319/10712 [56:17<52:16,  2.04it/s]
 40%|████      | 4320/10712 [56:17<52:15,  2.04it/s]
 40%|████      | 4321/10712 [56:18<52:11,  2.04it/s]
 40%|████      | 4322/10712 [56:18<52:12,  2.04it/s]
 40%|████      | 4323/10712 [56:19<52:11,  2.04it/s]
 40%|████      | 4324/10712 [56:19<52:12,  2.04it/s]
 40%|████      | 4325/10712 [56:20<52:15,  2.04it/s]{'loss': 3.8217, 'grad_norm': 0.1973145455121994, 'learning_rate': 0.0007443817022310808, 'epoch': 0.4}
                                                    

 40%|████      | 4325/10712 [56:20<52:15,  2.04it/s]
 40%|████      | 4326/10712 [56:20<52:14,  2.04it/s]
 40%|████      | 4327/10712 [56:21<52:13,  2.04it/s]
 40%|████      | 4328/10712 [56:21<52:15,  2.04it/s]
 40%|████      | 4329/10712 [56:22<52:09,  2.04it/s]
 40%|████      | 4330/10712 [56:22<52:08,  2.04it/s]
 40%|████      | 4331/10712 [56:23<52:07,  2.04it/s]
 40%|████      | 4332/10712 [56:23<52:09,  2.04it/s]
 40%|████      | 4333/10712 [56:24<52:04,  2.04it/s]
 40%|████      | 4334/10712 [56:24<52:01,  2.04it/s]
 40%|████      | 4335/10712 [56:25<52:08,  2.04it/s]
 40%|████      | 4336/10712 [56:25<52:04,  2.04it/s]
 40%|████      | 4337/10712 [56:26<52:01,  2.04it/s]
 40%|████      | 4338/10712 [56:26<52:01,  2.04it/s]
 41%|████      | 4339/10712 [56:27<52:04,  2.04it/s]
 41%|████      | 4340/10712 [56:27<52:04,  2.04it/s]
 41%|████      | 4341/10712 [56:28<52:00,  2.04it/s]
 41%|████      | 4342/10712 [56:28<52:04,  2.04it/s]
 41%|████      | 4343/10712 [56:29<52:05,  2.04it/s]
 41%|████      | 4344/10712 [56:29<52:00,  2.04it/s]
 41%|████      | 4345/10712 [56:30<51:58,  2.04it/s]
 41%|████      | 4346/10712 [56:30<51:57,  2.04it/s]
 41%|████      | 4347/10712 [56:31<52:01,  2.04it/s]
 41%|████      | 4348/10712 [56:31<51:59,  2.04it/s]
 41%|████      | 4349/10712 [56:32<52:00,  2.04it/s]
 41%|████      | 4350/10712 [56:32<52:02,  2.04it/s]
                                                    
{'loss': 3.8259, 'grad_norm': 0.2033911794424057, 'learning_rate': 0.0007408197188503156, 'epoch': 0.41}

 41%|████      | 4350/10712 [56:32<52:02,  2.04it/s]
 41%|████      | 4351/10712 [56:33<52:06,  2.03it/s]
 41%|████      | 4352/10712 [56:33<52:04,  2.04it/s]
 41%|████      | 4353/10712 [56:34<52:01,  2.04it/s]
 41%|████      | 4354/10712 [56:34<52:01,  2.04it/s]
 41%|████      | 4355/10712 [56:35<52:02,  2.04it/s]
 41%|████      | 4356/10712 [56:35<51:58,  2.04it/s]
 41%|████      | 4357/10712 [56:36<52:02,  2.04it/s]
 41%|████      | 4358/10712 [56:36<52:04,  2.03it/s]
 41%|████      | 4359/10712 [56:37<52:01,  2.04it/s]
 41%|████      | 4360/10712 [56:37<51:56,  2.04it/s]
 41%|████      | 4361/10712 [56:38<51:57,  2.04it/s]
 41%|████      | 4362/10712 [56:38<51:59,  2.04it/s]
 41%|████      | 4363/10712 [56:39<51:52,  2.04it/s]
 41%|████      | 4364/10712 [56:39<51:51,  2.04it/s]
 41%|████      | 4365/10712 [56:39<51:54,  2.04it/s]
 41%|████      | 4366/10712 [56:40<51:54,  2.04it/s]
 41%|████      | 4367/10712 [56:40<51:53,  2.04it/s]
 41%|████      | 4368/10712 [56:41<51:58,  2.03it/s]
 41%|████      | 4369/10712 [56:41<51:57,  2.03it/s]
 41%|████      | 4370/10712 [56:42<51:52,  2.04it/s]
 41%|████      | 4371/10712 [56:42<51:51,  2.04it/s]
 41%|████      | 4372/10712 [56:43<51:50,  2.04it/s]
 41%|████      | 4373/10712 [56:43<51:47,  2.04it/s]
 41%|████      | 4374/10712 [56:44<51:46,  2.04it/s]
 41%|████      | 4375/10712 [56:44<51:51,  2.04it/s]
                                                    
{'loss': 3.8162, 'grad_norm': 0.19839434325695038, 'learning_rate': 0.0007372417503702805, 'epoch': 0.41}

 41%|████      | 4375/10712 [56:44<51:51,  2.04it/s]
 41%|████      | 4376/10712 [56:45<51:53,  2.04it/s]
 41%|████      | 4377/10712 [56:45<51:53,  2.03it/s]
 41%|████      | 4378/10712 [56:46<51:55,  2.03it/s]
 41%|████      | 4379/10712 [56:46<51:49,  2.04it/s]
 41%|████      | 4380/10712 [56:47<51:47,  2.04it/s]
 41%|████      | 4381/10712 [56:47<51:50,  2.04it/s]
 41%|████      | 4382/10712 [56:48<51:49,  2.04it/s]
 41%|████      | 4383/10712 [56:48<51:51,  2.03it/s]
 41%|████      | 4384/10712 [56:49<51:51,  2.03it/s]
 41%|████      | 4385/10712 [56:49<51:48,  2.04it/s]
 41%|████      | 4386/10712 [56:50<51:49,  2.03it/s]
 41%|████      | 4387/10712 [56:50<51:46,  2.04it/s]
 41%|████      | 4388/10712 [56:51<51:42,  2.04it/s]
 41%|████      | 4389/10712 [56:51<51:39,  2.04it/s]
 41%|████      | 4390/10712 [56:52<51:43,  2.04it/s]
 41%|████      | 4391/10712 [56:52<51:40,  2.04it/s]
 41%|████      | 4392/10712 [56:53<51:38,  2.04it/s]
 41%|████      | 4393/10712 [56:53<51:39,  2.04it/s]
 41%|████      | 4394/10712 [56:54<51:40,  2.04it/s]
 41%|████      | 4395/10712 [56:54<51:36,  2.04it/s]
 41%|████      | 4396/10712 [56:55<51:34,  2.04it/s]
 41%|████      | 4397/10712 [56:55<51:37,  2.04it/s]
 41%|████      | 4398/10712 [56:56<51:35,  2.04it/s]
 41%|████      | 4399/10712 [56:56<51:34,  2.04it/s]
 41%|████      | 4400/10712 [56:57<51:39,  2.04it/s]
                                                    
{'loss': 3.8204, 'grad_norm': 0.22522036731243134, 'learning_rate': 0.0007336480342888914, 'epoch': 0.41}

 41%|████      | 4400/10712 [56:57<51:39,  2.04it/s]
 41%|████      | 4401/10712 [56:57<51:43,  2.03it/s]
 41%|████      | 4402/10712 [56:58<51:38,  2.04it/s]
 41%|████      | 4403/10712 [56:58<51:38,  2.04it/s]
 41%|████      | 4404/10712 [56:59<51:34,  2.04it/s]
 41%|████      | 4405/10712 [56:59<51:35,  2.04it/s]
 41%|████      | 4406/10712 [57:00<51:35,  2.04it/s]
 41%|████      | 4407/10712 [57:00<51:33,  2.04it/s]
 41%|████      | 4408/10712 [57:01<51:32,  2.04it/s]
 41%|████      | 4409/10712 [57:01<51:35,  2.04it/s]
 41%|████      | 4410/10712 [57:02<51:34,  2.04it/s]
 41%|████      | 4411/10712 [57:02<51:33,  2.04it/s]
 41%|████      | 4412/10712 [57:03<51:35,  2.04it/s]
 41%|████      | 4413/10712 [57:03<51:34,  2.04it/s]
 41%|████      | 4414/10712 [57:04<51:33,  2.04it/s]
 41%|████      | 4415/10712 [57:04<51:38,  2.03it/s]
 41%|████      | 4416/10712 [57:05<51:33,  2.04it/s]
 41%|████      | 4417/10712 [57:05<51:38,  2.03it/s]
 41%|████      | 4418/10712 [57:06<51:36,  2.03it/s]
 41%|████▏     | 4419/10712 [57:06<51:29,  2.04it/s]
 41%|████▏     | 4420/10712 [57:07<51:26,  2.04it/s]
 41%|████▏     | 4421/10712 [57:07<51:24,  2.04it/s]
 41%|████▏     | 4422/10712 [57:07<51:25,  2.04it/s]
 41%|████▏     | 4423/10712 [57:08<51:22,  2.04it/s]
 41%|████▏     | 4424/10712 [57:08<51:23,  2.04it/s]
 41%|████▏     | 4425/10712 [57:09<51:28,  2.04it/s]
                                                    {'loss': 3.8183, 'grad_norm': 0.19121618568897247, 'learning_rate': 0.0007300388091493566, 'epoch': 0.41}


 41%|████▏     | 4425/10712 [57:09<51:28,  2.04it/s]
 41%|████▏     | 4426/10712 [57:09<51:27,  2.04it/s]
 41%|████▏     | 4427/10712 [57:10<51:25,  2.04it/s]
 41%|████▏     | 4428/10712 [57:10<51:27,  2.04it/s]
 41%|████▏     | 4429/10712 [57:11<51:22,  2.04it/s]
 41%|████▏     | 4430/10712 [57:11<51:19,  2.04it/s]
 41%|████▏     | 4431/10712 [57:12<51:19,  2.04it/s]
 41%|████▏     | 4432/10712 [57:12<51:21,  2.04it/s]
 41%|████▏     | 4433/10712 [57:13<51:19,  2.04it/s]
 41%|████▏     | 4434/10712 [57:13<51:17,  2.04it/s]
 41%|████▏     | 4435/10712 [57:14<51:20,  2.04it/s]
 41%|████▏     | 4436/10712 [57:14<51:20,  2.04it/s]
 41%|████▏     | 4437/10712 [57:15<51:20,  2.04it/s]
 41%|████▏     | 4438/10712 [57:15<51:16,  2.04it/s]
 41%|████▏     | 4439/10712 [57:16<51:18,  2.04it/s]
 41%|████▏     | 4440/10712 [57:16<51:17,  2.04it/s]
 41%|████▏     | 4441/10712 [57:17<51:13,  2.04it/s]
 41%|████▏     | 4442/10712 [57:17<51:17,  2.04it/s]
 41%|████▏     | 4443/10712 [57:18<51:17,  2.04it/s]
 41%|████▏     | 4444/10712 [57:18<51:15,  2.04it/s]
 41%|████▏     | 4445/10712 [57:19<51:13,  2.04it/s]
 42%|████▏     | 4446/10712 [57:19<51:16,  2.04it/s]
 42%|████▏     | 4447/10712 [57:20<51:15,  2.04it/s]
 42%|████▏     | 4448/10712 [57:20<51:10,  2.04it/s]
 42%|████▏     | 4449/10712 [57:21<51:14,  2.04it/s]
 42%|████▏     | 4450/10712 [57:21<51:14,  2.04it/s]
                                                    
{'loss': 3.8134, 'grad_norm': 0.21344850957393646, 'learning_rate': 0.0007264143145243422, 'epoch': 0.42}

 42%|████▏     | 4450/10712 [57:21<51:14,  2.04it/s]
 42%|████▏     | 4451/10712 [57:22<51:14,  2.04it/s]
 42%|████▏     | 4452/10712 [57:22<51:16,  2.04it/s]
 42%|████▏     | 4453/10712 [57:23<51:14,  2.04it/s]
 42%|████▏     | 4454/10712 [57:23<51:09,  2.04it/s]
 42%|████▏     | 4455/10712 [57:24<51:05,  2.04it/s]
 42%|████▏     | 4456/10712 [57:24<51:09,  2.04it/s]
 42%|████▏     | 4457/10712 [57:25<51:09,  2.04it/s]
 42%|████▏     | 4458/10712 [57:25<51:06,  2.04it/s]
 42%|████▏     | 4459/10712 [57:26<51:07,  2.04it/s]
 42%|████▏     | 4460/10712 [57:26<51:08,  2.04it/s]
 42%|████▏     | 4461/10712 [57:27<51:06,  2.04it/s]
 42%|████▏     | 4462/10712 [57:27<51:02,  2.04it/s]
 42%|████▏     | 4463/10712 [57:28<51:06,  2.04it/s]
 42%|████▏     | 4464/10712 [57:28<51:05,  2.04it/s]
 42%|████▏     | 4465/10712 [57:29<51:01,  2.04it/s]
 42%|████▏     | 4466/10712 [57:29<51:04,  2.04it/s]
 42%|████▏     | 4467/10712 [57:30<51:02,  2.04it/s]
 42%|████▏     | 4468/10712 [57:30<51:00,  2.04it/s]
 42%|████▏     | 4469/10712 [57:31<51:01,  2.04it/s]
 42%|████▏     | 4470/10712 [57:31<50:59,  2.04it/s]
 42%|████▏     | 4471/10712 [57:32<50:58,  2.04it/s]
 42%|████▏     | 4472/10712 [57:32<50:55,  2.04it/s]
 42%|████▏     | 4473/10712 [57:33<50:55,  2.04it/s]
 42%|████▏     | 4474/10712 [57:33<51:00,  2.04it/s]
 42%|████▏     | 4475/10712 [57:33<50:58,  2.04it/s]{'loss': 3.8163, 'grad_norm': 0.21075329184532166, 'learning_rate': 0.0007227747910000704, 'epoch': 0.42}

                                                    

 42%|████▏     | 4475/10712 [57:33<50:58,  2.04it/s]
 42%|████▏     | 4476/10712 [57:34<51:01,  2.04it/s]
 42%|████▏     | 4477/10712 [57:34<51:02,  2.04it/s]
 42%|████▏     | 4478/10712 [57:35<51:01,  2.04it/s]
 42%|████▏     | 4479/10712 [57:35<50:56,  2.04it/s]
 42%|████▏     | 4480/10712 [57:36<50:56,  2.04it/s]
 42%|████▏     | 4481/10712 [57:36<50:59,  2.04it/s]
 42%|████▏     | 4482/10712 [57:37<50:59,  2.04it/s]
 42%|████▏     | 4483/10712 [57:37<50:58,  2.04it/s]
 42%|████▏     | 4484/10712 [57:38<51:00,  2.04it/s]
 42%|████▏     | 4485/10712 [57:38<50:56,  2.04it/s]
 42%|████▏     | 4486/10712 [57:39<50:55,  2.04it/s]
 42%|████▏     | 4487/10712 [57:39<50:54,  2.04it/s]
 42%|████▏     | 4488/10712 [57:40<50:55,  2.04it/s]
 42%|████▏     | 4489/10712 [57:40<50:54,  2.04it/s]
 42%|████▏     | 4490/10712 [57:41<50:50,  2.04it/s]
 42%|████▏     | 4491/10712 [57:41<50:51,  2.04it/s]
 42%|████▏     | 4492/10712 [57:42<50:53,  2.04it/s]
 42%|████▏     | 4493/10712 [57:42<50:52,  2.04it/s]
 42%|████▏     | 4494/10712 [57:43<50:50,  2.04it/s]
 42%|████▏     | 4495/10712 [57:43<50:50,  2.04it/s]
 42%|████▏     | 4496/10712 [57:44<50:49,  2.04it/s]
 42%|████▏     | 4497/10712 [57:44<50:51,  2.04it/s]
 42%|████▏     | 4498/10712 [57:45<50:46,  2.04it/s]
 42%|████▏     | 4499/10712 [57:45<50:50,  2.04it/s]
 42%|████▏     | 4500/10712 [57:46<50:53,  2.03it/s]
                                                    
{'loss': 3.8112, 'grad_norm': 0.19861102104187012, 'learning_rate': 0.0007191204801603497, 'epoch': 0.42}

 42%|████▏     | 4500/10712 [57:46<50:53,  2.03it/s]
 42%|████▏     | 4501/10712 [57:46<50:55,  2.03it/s]
 42%|████▏     | 4502/10712 [57:47<50:50,  2.04it/s]
 42%|████▏     | 4503/10712 [57:47<50:49,  2.04it/s]
 42%|████▏     | 4504/10712 [57:48<50:49,  2.04it/s]
 42%|████▏     | 4505/10712 [57:48<50:46,  2.04it/s]
 42%|████▏     | 4506/10712 [57:49<50:47,  2.04it/s]
 42%|████▏     | 4507/10712 [57:49<50:47,  2.04it/s]
 42%|████▏     | 4508/10712 [57:50<50:44,  2.04it/s]
 42%|████▏     | 4509/10712 [57:50<50:48,  2.03it/s]
 42%|████▏     | 4510/10712 [57:51<50:45,  2.04it/s]
 42%|████▏     | 4511/10712 [57:51<50:45,  2.04it/s]
 42%|████▏     | 4512/10712 [57:52<50:46,  2.03it/s]
 42%|████▏     | 4513/10712 [57:52<50:43,  2.04it/s]
 42%|████▏     | 4514/10712 [57:53<50:40,  2.04it/s]
 42%|████▏     | 4515/10712 [57:53<50:41,  2.04it/s]
 42%|████▏     | 4516/10712 [57:54<50:41,  2.04it/s]
 42%|████▏     | 4517/10712 [57:54<50:41,  2.04it/s]
 42%|████▏     | 4518/10712 [57:55<50:41,  2.04it/s]
 42%|████▏     | 4519/10712 [57:55<50:43,  2.04it/s]
 42%|████▏     | 4520/10712 [57:56<50:38,  2.04it/s]
 42%|████▏     | 4521/10712 [57:56<50:39,  2.04it/s]
 42%|████▏     | 4522/10712 [57:57<50:40,  2.04it/s]
 42%|████▏     | 4523/10712 [57:57<50:38,  2.04it/s]
 42%|████▏     | 4524/10712 [57:58<50:37,  2.04it/s]
 42%|████▏     | 4525/10712 [57:58<50:41,  2.03it/s]{'loss': 3.8087, 'grad_norm': 0.20870919525623322, 'learning_rate': 0.000715451624570539, 'epoch': 0.42}
                                                    

 42%|████▏     | 4525/10712 [57:58<50:41,  2.03it/s]
 42%|████▏     | 4526/10712 [57:59<50:41,  2.03it/s]
 42%|████▏     | 4527/10712 [57:59<50:39,  2.03it/s]
 42%|████▏     | 4528/10712 [58:00<50:41,  2.03it/s]
 42%|████▏     | 4529/10712 [58:00<50:35,  2.04it/s]
 42%|████▏     | 4530/10712 [58:00<50:31,  2.04it/s]
 42%|████▏     | 4531/10712 [58:01<50:36,  2.04it/s]
 42%|████▏     | 4532/10712 [58:01<50:33,  2.04it/s]
 42%|████▏     | 4533/10712 [58:02<50:27,  2.04it/s]
 42%|████▏     | 4534/10712 [58:02<50:31,  2.04it/s]
 42%|████▏     | 4535/10712 [58:03<50:29,  2.04it/s]
 42%|████▏     | 4536/10712 [58:03<50:27,  2.04it/s]
 42%|████▏     | 4537/10712 [58:04<50:26,  2.04it/s]
 42%|████▏     | 4538/10712 [58:04<50:32,  2.04it/s]
 42%|████▏     | 4539/10712 [58:05<50:27,  2.04it/s]
 42%|████▏     | 4540/10712 [58:05<50:23,  2.04it/s]
 42%|████▏     | 4541/10712 [58:06<50:27,  2.04it/s]
 42%|████▏     | 4542/10712 [58:06<50:26,  2.04it/s]
 42%|████▏     | 4543/10712 [58:07<50:22,  2.04it/s]
 42%|████▏     | 4544/10712 [58:07<50:23,  2.04it/s]
 42%|████▏     | 4545/10712 [58:08<50:24,  2.04it/s]
 42%|████▏     | 4546/10712 [58:08<50:25,  2.04it/s]
 42%|████▏     | 4547/10712 [58:09<50:21,  2.04it/s]
 42%|████▏     | 4548/10712 [58:09<50:23,  2.04it/s]
 42%|████▏     | 4549/10712 [58:10<50:24,  2.04it/s]
 42%|████▏     | 4550/10712 [58:10<50:22,  2.04it/s]
                                                    
{'loss': 3.8055, 'grad_norm': 0.20465131103992462, 'learning_rate': 0.0007117684677614468, 'epoch': 0.42}

 42%|████▏     | 4550/10712 [58:10<50:22,  2.04it/s]
 42%|████▏     | 4551/10712 [58:11<50:25,  2.04it/s]
 42%|████▏     | 4552/10712 [58:11<50:23,  2.04it/s]
 43%|████▎     | 4553/10712 [58:12<50:20,  2.04it/s]
 43%|████▎     | 4554/10712 [58:12<50:19,  2.04it/s]
 43%|████▎     | 4555/10712 [58:13<50:20,  2.04it/s]
 43%|████▎     | 4556/10712 [58:13<50:21,  2.04it/s]
 43%|████▎     | 4557/10712 [58:14<50:21,  2.04it/s]
 43%|████▎     | 4558/10712 [58:14<50:23,  2.04it/s]
 43%|████▎     | 4559/10712 [58:15<50:21,  2.04it/s]
 43%|████▎     | 4560/10712 [58:15<50:16,  2.04it/s]
 43%|████▎     | 4561/10712 [58:16<50:16,  2.04it/s]
 43%|████▎     | 4562/10712 [58:16<50:18,  2.04it/s]
 43%|████▎     | 4563/10712 [58:17<50:14,  2.04it/s]
 43%|████▎     | 4564/10712 [58:17<50:12,  2.04it/s]
 43%|████▎     | 4565/10712 [58:18<50:15,  2.04it/s]
 43%|████▎     | 4566/10712 [58:18<58:32,  1.75it/s]
 43%|████▎     | 4567/10712 [58:19<56:03,  1.83it/s]
 43%|████▎     | 4568/10712 [58:19<54:15,  1.89it/s]
 43%|████▎     | 4569/10712 [58:20<52:59,  1.93it/s]
 43%|████▎     | 4570/10712 [58:20<52:11,  1.96it/s]
 43%|████▎     | 4571/10712 [58:21<51:39,  1.98it/s]
 43%|████▎     | 4572/10712 [58:21<51:13,  2.00it/s]
 43%|████▎     | 4573/10712 [58:22<50:56,  2.01it/s]
 43%|████▎     | 4574/10712 [58:22<50:41,  2.02it/s]
 43%|████▎     | 4575/10712 [58:23<50:29,  2.03it/s]
                                                    {'loss': 3.8063, 'grad_norm': 0.22927306592464447, 'learning_rate': 0.0007080712542131656, 'epoch': 0.43}

 43%|████▎     | 4575/10712 [58:23<50:29,  2.03it/s]
 43%|████▎     | 4576/10712 [58:23<50:25,  2.03it/s]
 43%|████▎     | 4577/10712 [58:24<50:20,  2.03it/s]
 43%|████▎     | 4578/10712 [58:24<50:15,  2.03it/s]
 43%|████▎     | 4579/10712 [58:25<50:09,  2.04it/s]
 43%|████▎     | 4580/10712 [58:25<50:08,  2.04it/s]
 43%|████▎     | 4581/10712 [58:26<50:10,  2.04it/s]
 43%|████▎     | 4582/10712 [58:26<50:10,  2.04it/s]
 43%|████▎     | 4583/10712 [58:27<50:05,  2.04it/s]
 43%|████▎     | 4584/10712 [58:27<50:07,  2.04it/s]
 43%|████▎     | 4585/10712 [58:28<50:05,  2.04it/s]
 43%|████▎     | 4586/10712 [58:28<50:01,  2.04it/s]
 43%|████▎     | 4587/10712 [58:29<50:00,  2.04it/s]
 43%|████▎     | 4588/10712 [58:29<50:02,  2.04it/s]
 43%|████▎     | 4589/10712 [58:30<50:02,  2.04it/s]
 43%|████▎     | 4590/10712 [58:30<49:59,  2.04it/s]
 43%|████▎     | 4591/10712 [58:31<58:20,  1.75it/s]
 43%|████▎     | 4592/10712 [58:31<55:49,  1.83it/s]
 43%|████▎     | 4593/10712 [58:32<54:10,  1.88it/s]
 43%|████▎     | 4594/10712 [58:32<52:54,  1.93it/s]
 43%|████▎     | 4595/10712 [58:33<51:59,  1.96it/s]
 43%|████▎     | 4596/10712 [58:33<51:24,  1.98it/s]
 43%|████▎     | 4597/10712 [58:34<50:58,  2.00it/s]
 43%|████▎     | 4598/10712 [58:34<50:38,  2.01it/s]
 43%|████▎     | 4599/10712 [58:35<50:29,  2.02it/s]
 43%|████▎     | 4600/10712 [58:35<50:21,  2.02it/s]
                                                    {'loss': 3.8075, 'grad_norm': 0.21127530932426453, 'learning_rate': 0.000704360229338845, 'epoch': 0.43}

 43%|████▎     | 4600/10712 [58:35<50:21,  2.02it/s]
 43%|████▎     | 4601/10712 [58:36<50:16,  2.03it/s]
 43%|████▎     | 4602/10712 [58:36<50:11,  2.03it/s]
 43%|████▎     | 4603/10712 [58:37<50:09,  2.03it/s]
 43%|████▎     | 4604/10712 [58:37<50:02,  2.03it/s]
 43%|████▎     | 4605/10712 [58:38<49:59,  2.04it/s]
 43%|████▎     | 4606/10712 [58:38<50:00,  2.04it/s]
 43%|████▎     | 4607/10712 [58:39<49:55,  2.04it/s]
 43%|████▎     | 4608/10712 [58:39<49:57,  2.04it/s]
 43%|████▎     | 4609/10712 [58:40<49:58,  2.04it/s]
 43%|████▎     | 4610/10712 [58:40<49:55,  2.04it/s]
 43%|████▎     | 4611/10712 [58:41<49:50,  2.04it/s]
 43%|████▎     | 4612/10712 [58:41<49:49,  2.04it/s]
 43%|████▎     | 4613/10712 [58:42<49:50,  2.04it/s]
 43%|████▎     | 4614/10712 [58:42<49:49,  2.04it/s]
 43%|████▎     | 4615/10712 [58:43<49:48,  2.04it/s]
 43%|████▎     | 4616/10712 [58:43<49:50,  2.04it/s]
 43%|████▎     | 4617/10712 [58:44<49:48,  2.04it/s]
 43%|████▎     | 4618/10712 [58:44<49:47,  2.04it/s]
 43%|████▎     | 4619/10712 [58:45<49:48,  2.04it/s]
 43%|████▎     | 4620/10712 [58:45<49:49,  2.04it/s]
 43%|████▎     | 4621/10712 [58:46<49:50,  2.04it/s]
 43%|████▎     | 4622/10712 [58:46<49:50,  2.04it/s]
 43%|████▎     | 4623/10712 [58:47<49:44,  2.04it/s]
 43%|████▎     | 4624/10712 [58:47<49:40,  2.04it/s]
 43%|████▎     | 4625/10712 [58:48<49:40,  2.04it/s]
                                                    {'loss': 3.8045, 'grad_norm': 0.2073400318622589, 'learning_rate': 0.0007006356394684002, 'epoch': 0.43}


 43%|████▎     | 4625/10712 [58:48<49:40,  2.04it/s]
 43%|████▎     | 4626/10712 [58:48<49:46,  2.04it/s]
 43%|████▎     | 4627/10712 [58:49<49:44,  2.04it/s]
 43%|████▎     | 4628/10712 [58:49<49:39,  2.04it/s]
 43%|████▎     | 4629/10712 [58:50<49:37,  2.04it/s]
 43%|████▎     | 4630/10712 [58:50<49:36,  2.04it/s]
 43%|████▎     | 4631/10712 [58:51<49:39,  2.04it/s]
 43%|████▎     | 4632/10712 [58:51<49:40,  2.04it/s]
 43%|████▎     | 4633/10712 [58:52<49:38,  2.04it/s]
 43%|████▎     | 4634/10712 [58:52<49:36,  2.04it/s]
 43%|████▎     | 4635/10712 [58:53<49:38,  2.04it/s]
 43%|████▎     | 4636/10712 [58:53<49:40,  2.04it/s]
 43%|████▎     | 4637/10712 [58:54<49:40,  2.04it/s]
 43%|████▎     | 4638/10712 [58:54<49:38,  2.04it/s]
 43%|████▎     | 4639/10712 [58:54<49:38,  2.04it/s]
 43%|████▎     | 4640/10712 [58:55<49:38,  2.04it/s]
 43%|████▎     | 4641/10712 [58:55<49:37,  2.04it/s]
 43%|████▎     | 4642/10712 [58:56<49:40,  2.04it/s]
 43%|████▎     | 4643/10712 [58:56<49:38,  2.04it/s]
 43%|████▎     | 4644/10712 [58:57<49:38,  2.04it/s]
 43%|████▎     | 4645/10712 [58:57<49:37,  2.04it/s]
 43%|████▎     | 4646/10712 [58:58<49:35,  2.04it/s]
 43%|████▎     | 4647/10712 [58:58<49:36,  2.04it/s]
 43%|████▎     | 4648/10712 [58:59<49:34,  2.04it/s]
 43%|████▎     | 4649/10712 [58:59<49:36,  2.04it/s]
 43%|████▎     | 4650/10712 [59:00<49:36,  2.04it/s]{'loss': 3.7992, 'grad_norm': 0.20953071117401123, 'learning_rate': 0.0006968977318321625, 'epoch': 0.43}

                                                    

 43%|████▎     | 4650/10712 [59:00<49:36,  2.04it/s]
 43%|████▎     | 4651/10712 [59:00<49:41,  2.03it/s]
 43%|████▎     | 4652/10712 [59:01<49:37,  2.04it/s]
 43%|████▎     | 4653/10712 [59:01<49:33,  2.04it/s]
 43%|████▎     | 4654/10712 [59:02<49:32,  2.04it/s]
 43%|████▎     | 4655/10712 [59:02<49:31,  2.04it/s]
 43%|████▎     | 4656/10712 [59:03<49:30,  2.04it/s]
 43%|████▎     | 4657/10712 [59:03<49:27,  2.04it/s]
 43%|████▎     | 4658/10712 [59:04<49:30,  2.04it/s]
 43%|████▎     | 4659/10712 [59:04<49:27,  2.04it/s]
 44%|████▎     | 4660/10712 [59:05<49:25,  2.04it/s]
 44%|████▎     | 4661/10712 [59:05<49:22,  2.04it/s]
 44%|████▎     | 4662/10712 [59:06<49:23,  2.04it/s]
 44%|████▎     | 4663/10712 [59:06<49:22,  2.04it/s]
 44%|████▎     | 4664/10712 [59:07<49:18,  2.04it/s]
 44%|████▎     | 4665/10712 [59:07<49:16,  2.05it/s]
 44%|████▎     | 4666/10712 [59:08<49:22,  2.04it/s]
 44%|████▎     | 4667/10712 [59:08<49:23,  2.04it/s]
 44%|████▎     | 4668/10712 [59:09<49:17,  2.04it/s]
 44%|████▎     | 4669/10712 [59:09<49:21,  2.04it/s]
 44%|████▎     | 4670/10712 [59:10<49:24,  2.04it/s]
 44%|████▎     | 4671/10712 [59:10<49:20,  2.04it/s]
 44%|████▎     | 4672/10712 [59:11<49:15,  2.04it/s]
 44%|████▎     | 4673/10712 [59:11<49:14,  2.04it/s]
 44%|████▎     | 4674/10712 [59:12<49:16,  2.04it/s]
 44%|████▎     | 4675/10712 [59:12<49:16,  2.04it/s]{'loss': 3.7994, 'grad_norm': 0.19907726347446442, 'learning_rate': 0.000693146754544468, 'epoch': 0.44}

                                                    

 44%|████▎     | 4675/10712 [59:12<49:16,  2.04it/s]
 44%|████▎     | 4676/10712 [59:13<49:23,  2.04it/s]
 44%|████▎     | 4677/10712 [59:13<49:22,  2.04it/s]
 44%|████▎     | 4678/10712 [59:14<49:18,  2.04it/s]
 44%|████▎     | 4679/10712 [59:14<49:15,  2.04it/s]
 44%|████▎     | 4680/10712 [59:15<49:18,  2.04it/s]
 44%|████▎     | 4681/10712 [59:15<49:17,  2.04it/s]
 44%|████▎     | 4682/10712 [59:16<49:16,  2.04it/s]
 44%|████▎     | 4683/10712 [59:16<49:13,  2.04it/s]
 44%|████▎     | 4684/10712 [59:17<49:17,  2.04it/s]
 44%|████▎     | 4685/10712 [59:17<49:16,  2.04it/s]
 44%|████▎     | 4686/10712 [59:18<49:13,  2.04it/s]
 44%|████▍     | 4687/10712 [59:18<49:13,  2.04it/s]
 44%|████▍     | 4688/10712 [59:19<49:13,  2.04it/s]
 44%|████▍     | 4689/10712 [59:19<49:12,  2.04it/s]
 44%|████▍     | 4690/10712 [59:19<49:08,  2.04it/s]
 44%|████▍     | 4691/10712 [59:20<49:09,  2.04it/s]
 44%|████▍     | 4692/10712 [59:20<49:13,  2.04it/s]
 44%|████▍     | 4693/10712 [59:21<49:11,  2.04it/s]
 44%|████▍     | 4694/10712 [59:21<49:10,  2.04it/s]
 44%|████▍     | 4695/10712 [59:22<49:13,  2.04it/s]
 44%|████▍     | 4696/10712 [59:22<49:12,  2.04it/s]
 44%|████▍     | 4697/10712 [59:23<49:08,  2.04it/s]
 44%|████▍     | 4698/10712 [59:23<49:08,  2.04it/s]
 44%|████▍     | 4699/10712 [59:24<49:09,  2.04it/s]
 44%|████▍     | 4700/10712 [59:24<49:09,  2.04it/s]
                                                    {'loss': 3.7919, 'grad_norm': 0.2191115766763687, 'learning_rate': 0.0006893829565871878, 'epoch': 0.44}


 44%|████▍     | 4700/10712 [59:24<49:09,  2.04it/s]
 44%|████▍     | 4701/10712 [59:25<49:12,  2.04it/s]
 44%|████▍     | 4702/10712 [59:25<49:14,  2.03it/s]
 44%|████▍     | 4703/10712 [59:26<49:11,  2.04it/s]
 44%|████▍     | 4704/10712 [59:26<49:09,  2.04it/s]
 44%|████▍     | 4705/10712 [59:27<49:11,  2.04it/s]
 44%|████▍     | 4706/10712 [59:27<49:06,  2.04it/s]
 44%|████▍     | 4707/10712 [59:28<49:07,  2.04it/s]
 44%|████▍     | 4708/10712 [59:28<49:10,  2.04it/s]
 44%|████▍     | 4709/10712 [59:29<49:09,  2.04it/s]
 44%|████▍     | 4710/10712 [59:29<49:09,  2.04it/s]
 44%|████▍     | 4711/10712 [59:30<49:10,  2.03it/s]
 44%|████▍     | 4712/10712 [59:30<49:08,  2.04it/s]
 44%|████▍     | 4713/10712 [59:31<49:04,  2.04it/s]
 44%|████▍     | 4714/10712 [59:31<49:00,  2.04it/s]
 44%|████▍     | 4715/10712 [59:32<49:03,  2.04it/s]
 44%|████▍     | 4716/10712 [59:32<49:00,  2.04it/s]
 44%|████▍     | 4717/10712 [59:33<48:59,  2.04it/s]
 44%|████▍     | 4718/10712 [59:33<48:59,  2.04it/s]
 44%|████▍     | 4719/10712 [59:34<49:01,  2.04it/s]
 44%|████▍     | 4720/10712 [59:34<48:59,  2.04it/s]
 44%|████▍     | 4721/10712 [59:35<48:55,  2.04it/s]
 44%|████▍     | 4722/10712 [59:35<48:53,  2.04it/s]
 44%|████▍     | 4723/10712 [59:36<48:55,  2.04it/s]
 44%|████▍     | 4724/10712 [59:36<48:53,  2.04it/s]
 44%|████▍     | 4725/10712 [59:37<48:52,  2.04it/s]
                                                    
{'loss': 3.7948, 'grad_norm': 0.22968147695064545, 'learning_rate': 0.0006856065877932025, 'epoch': 0.44}

 44%|████▍     | 4725/10712 [59:37<48:52,  2.04it/s]
 44%|████▍     | 4726/10712 [59:37<48:53,  2.04it/s]
 44%|████▍     | 4727/10712 [59:38<48:54,  2.04it/s]
 44%|████▍     | 4728/10712 [59:38<48:52,  2.04it/s]
 44%|████▍     | 4729/10712 [59:39<48:52,  2.04it/s]
 44%|████▍     | 4730/10712 [59:39<48:51,  2.04it/s]
 44%|████▍     | 4731/10712 [59:40<48:49,  2.04it/s]
 44%|████▍     | 4732/10712 [59:40<48:47,  2.04it/s]
 44%|████▍     | 4733/10712 [59:41<48:54,  2.04it/s]
 44%|████▍     | 4734/10712 [59:41<48:54,  2.04it/s]
 44%|████▍     | 4735/10712 [59:42<48:50,  2.04it/s]
 44%|████▍     | 4736/10712 [59:42<48:49,  2.04it/s]
 44%|████▍     | 4737/10712 [59:43<48:51,  2.04it/s]
 44%|████▍     | 4738/10712 [59:43<48:50,  2.04it/s]
 44%|████▍     | 4739/10712 [59:44<48:45,  2.04it/s]
 44%|████▍     | 4740/10712 [59:44<48:46,  2.04it/s]
 44%|████▍     | 4741/10712 [59:45<48:48,  2.04it/s]
 44%|████▍     | 4742/10712 [59:45<48:45,  2.04it/s]
 44%|████▍     | 4743/10712 [59:45<48:49,  2.04it/s]
 44%|████▍     | 4744/10712 [59:46<48:50,  2.04it/s]
 44%|████▍     | 4745/10712 [59:46<48:47,  2.04it/s]
 44%|████▍     | 4746/10712 [59:47<48:44,  2.04it/s]
 44%|████▍     | 4747/10712 [59:47<48:44,  2.04it/s]
 44%|████▍     | 4748/10712 [59:48<48:45,  2.04it/s]
 44%|████▍     | 4749/10712 [59:48<48:43,  2.04it/s]
 44%|████▍     | 4750/10712 [59:49<48:41,  2.04it/s]{'loss': 3.7935, 'grad_norm': 0.20975317060947418, 'learning_rate': 0.0006818178988298176, 'epoch': 0.44}
                                                    

 44%|████▍     | 4750/10712 [59:49<48:41,  2.04it/s]
 44%|████▍     | 4751/10712 [59:49<48:47,  2.04it/s]
 44%|████▍     | 4752/10712 [59:50<48:45,  2.04it/s]
 44%|████▍     | 4753/10712 [59:50<48:41,  2.04it/s]
 44%|████▍     | 4754/10712 [59:51<48:43,  2.04it/s]
 44%|████▍     | 4755/10712 [59:51<48:43,  2.04it/s]
 44%|████▍     | 4756/10712 [59:52<48:39,  2.04it/s]
 44%|████▍     | 4757/10712 [59:52<48:37,  2.04it/s]
 44%|████▍     | 4758/10712 [59:53<48:37,  2.04it/s]
 44%|████▍     | 4759/10712 [59:53<48:39,  2.04it/s]
 44%|████▍     | 4760/10712 [59:54<48:39,  2.04it/s]
 44%|████▍     | 4761/10712 [59:54<48:35,  2.04it/s]
 44%|████▍     | 4762/10712 [59:55<48:34,  2.04it/s]
 44%|████▍     | 4763/10712 [59:55<48:34,  2.04it/s]
 44%|████▍     | 4764/10712 [59:56<48:34,  2.04it/s]
 44%|████▍     | 4765/10712 [59:56<48:35,  2.04it/s]
 44%|████▍     | 4766/10712 [59:57<48:36,  2.04it/s]
 45%|████▍     | 4767/10712 [59:57<48:37,  2.04it/s]
 45%|████▍     | 4768/10712 [59:58<48:38,  2.04it/s]
 45%|████▍     | 4769/10712 [59:58<48:36,  2.04it/s]
 45%|████▍     | 4770/10712 [59:59<48:32,  2.04it/s]
 45%|████▍     | 4771/10712 [59:59<48:31,  2.04it/s]
 45%|████▍     | 4772/10712 [1:00:00<48:31,  2.04it/s]
 45%|████▍     | 4773/10712 [1:00:00<48:32,  2.04it/s]
 45%|████▍     | 4774/10712 [1:00:01<48:32,  2.04it/s]
 45%|████▍     | 4775/10712 [1:00:01<48:31,  2.04it/s]{'loss': 3.7906, 'grad_norm': 0.22519637644290924, 'learning_rate': 0.000678017141182125, 'epoch': 0.45}

                                                      

 45%|████▍     | 4775/10712 [1:00:01<48:31,  2.04it/s]
 45%|████▍     | 4776/10712 [1:00:02<48:35,  2.04it/s]
 45%|████▍     | 4777/10712 [1:00:02<48:32,  2.04it/s]
 45%|████▍     | 4778/10712 [1:00:03<48:28,  2.04it/s]
 45%|████▍     | 4779/10712 [1:00:03<48:31,  2.04it/s]
 45%|████▍     | 4780/10712 [1:00:04<48:28,  2.04it/s]
 45%|████▍     | 4781/10712 [1:00:04<48:23,  2.04it/s]
 45%|████▍     | 4782/10712 [1:00:05<48:25,  2.04it/s]
 45%|████▍     | 4783/10712 [1:00:05<48:27,  2.04it/s]
 45%|████▍     | 4784/10712 [1:00:06<48:25,  2.04it/s]
 45%|████▍     | 4785/10712 [1:00:06<48:26,  2.04it/s]
 45%|████▍     | 4786/10712 [1:00:07<48:25,  2.04it/s]
 45%|████▍     | 4787/10712 [1:00:07<48:24,  2.04it/s]
 45%|████▍     | 4788/10712 [1:00:08<48:23,  2.04it/s]
 45%|████▍     | 4789/10712 [1:00:08<48:19,  2.04it/s]
 45%|████▍     | 4790/10712 [1:00:09<48:20,  2.04it/s]
 45%|████▍     | 4791/10712 [1:00:09<48:23,  2.04it/s]
 45%|████▍     | 4792/10712 [1:00:10<48:21,  2.04it/s]
 45%|████▍     | 4793/10712 [1:00:10<48:21,  2.04it/s]
 45%|████▍     | 4794/10712 [1:00:11<48:23,  2.04it/s]
 45%|████▍     | 4795/10712 [1:00:11<48:19,  2.04it/s]
 45%|████▍     | 4796/10712 [1:00:11<48:18,  2.04it/s]
 45%|████▍     | 4797/10712 [1:00:12<48:19,  2.04it/s]
 45%|████▍     | 4798/10712 [1:00:12<48:20,  2.04it/s]
 45%|████▍     | 4799/10712 [1:00:13<48:20,  2.04it/s]
 45%|████▍     | 4800/10712 [1:00:13<48:25,  2.03it/s]{'loss': 3.7925, 'grad_norm': 0.19681546092033386, 'learning_rate': 0.0006742045671363106, 'epoch': 0.45}

                                                      

 45%|████▍     | 4800/10712 [1:00:13<48:25,  2.03it/s]
 45%|████▍     | 4801/10712 [1:00:14<48:27,  2.03it/s]
 45%|████▍     | 4802/10712 [1:00:14<48:56,  2.01it/s]
 45%|████▍     | 4803/10712 [1:00:15<48:46,  2.02it/s]
 45%|████▍     | 4804/10712 [1:00:15<48:38,  2.02it/s]
 45%|████▍     | 4805/10712 [1:00:16<48:32,  2.03it/s]
 45%|████▍     | 4806/10712 [1:00:16<48:28,  2.03it/s]
 45%|████▍     | 4807/10712 [1:00:17<48:23,  2.03it/s]
 45%|████▍     | 4808/10712 [1:00:17<48:22,  2.03it/s]
 45%|████▍     | 4809/10712 [1:00:18<48:25,  2.03it/s]
 45%|████▍     | 4810/10712 [1:00:18<48:20,  2.03it/s]
 45%|████▍     | 4811/10712 [1:00:19<48:18,  2.04it/s]
 45%|████▍     | 4812/10712 [1:00:19<48:19,  2.03it/s]
 45%|████▍     | 4813/10712 [1:00:20<48:17,  2.04it/s]
 45%|████▍     | 4814/10712 [1:00:20<48:13,  2.04it/s]
 45%|████▍     | 4815/10712 [1:00:21<48:13,  2.04it/s]
 45%|████▍     | 4816/10712 [1:00:21<48:17,  2.04it/s]
 45%|████▍     | 4817/10712 [1:00:22<48:13,  2.04it/s]
 45%|████▍     | 4818/10712 [1:00:22<48:10,  2.04it/s]
 45%|████▍     | 4819/10712 [1:00:23<48:11,  2.04it/s]
 45%|████▍     | 4820/10712 [1:00:23<48:10,  2.04it/s]
 45%|████▌     | 4821/10712 [1:00:24<48:09,  2.04it/s]
 45%|████▌     | 4822/10712 [1:00:24<48:11,  2.04it/s]
 45%|████▌     | 4823/10712 [1:00:25<48:10,  2.04it/s]
 45%|████▌     | 4824/10712 [1:00:25<48:06,  2.04it/s]
 45%|████▌     | 4825/10712 [1:00:26<48:09,  2.04it/s]{'loss': 3.7828, 'grad_norm': 0.23035882413387299, 'learning_rate': 0.0006703804297629074, 'epoch': 0.45}
                                                      

 45%|████▌     | 4825/10712 [1:00:26<48:09,  2.04it/s]
 45%|████▌     | 4826/10712 [1:00:26<48:13,  2.03it/s]
 45%|████▌     | 4827/10712 [1:00:27<48:09,  2.04it/s]
 45%|████▌     | 4828/10712 [1:00:27<48:07,  2.04it/s]
 45%|████▌     | 4829/10712 [1:00:28<48:10,  2.04it/s]
 45%|████▌     | 4830/10712 [1:00:28<48:07,  2.04it/s]
 45%|████▌     | 4831/10712 [1:00:29<48:04,  2.04it/s]
 45%|████▌     | 4832/10712 [1:00:29<48:07,  2.04it/s]
 45%|████▌     | 4833/10712 [1:00:30<48:10,  2.03it/s]
 45%|████▌     | 4834/10712 [1:00:30<48:05,  2.04it/s]
 45%|████▌     | 4835/10712 [1:00:31<48:09,  2.03it/s]
 45%|████▌     | 4836/10712 [1:00:31<48:04,  2.04it/s]
 45%|████▌     | 4837/10712 [1:00:32<48:01,  2.04it/s]
 45%|████▌     | 4838/10712 [1:00:32<48:02,  2.04it/s]
 45%|████▌     | 4839/10712 [1:00:33<48:04,  2.04it/s]
 45%|████▌     | 4840/10712 [1:00:33<48:10,  2.03it/s]
 45%|████▌     | 4841/10712 [1:00:34<48:11,  2.03it/s]
 45%|████▌     | 4842/10712 [1:00:34<48:07,  2.03it/s]
 45%|████▌     | 4843/10712 [1:00:35<48:08,  2.03it/s]
 45%|████▌     | 4844/10712 [1:00:35<48:09,  2.03it/s]
 45%|████▌     | 4845/10712 [1:00:36<48:02,  2.04it/s]
 45%|████▌     | 4846/10712 [1:00:36<48:02,  2.03it/s]
 45%|████▌     | 4847/10712 [1:00:37<48:03,  2.03it/s]
 45%|████▌     | 4848/10712 [1:00:37<47:59,  2.04it/s]
 45%|████▌     | 4849/10712 [1:00:38<47:56,  2.04it/s]
 45%|████▌     | 4850/10712 [1:00:38<47:57,  2.04it/s]{'loss': 3.786, 'grad_norm': 0.22728563845157623, 'learning_rate': 0.0006665449828999971, 'epoch': 0.45}

                                                      

 45%|████▌     | 4850/10712 [1:00:38<47:57,  2.04it/s]
 45%|████▌     | 4851/10712 [1:00:39<47:58,  2.04it/s]
 45%|████▌     | 4852/10712 [1:00:39<47:55,  2.04it/s]
 45%|████▌     | 4853/10712 [1:00:39<47:55,  2.04it/s]
 45%|████▌     | 4854/10712 [1:00:40<47:53,  2.04it/s]
 45%|████▌     | 4855/10712 [1:00:40<47:53,  2.04it/s]
 45%|████▌     | 4856/10712 [1:00:41<47:51,  2.04it/s]
 45%|████▌     | 4857/10712 [1:00:41<47:50,  2.04it/s]
 45%|████▌     | 4858/10712 [1:00:42<47:53,  2.04it/s]
 45%|████▌     | 4859/10712 [1:00:42<47:53,  2.04it/s]
 45%|████▌     | 4860/10712 [1:00:43<47:53,  2.04it/s]
 45%|████▌     | 4861/10712 [1:00:43<47:57,  2.03it/s]
 45%|████▌     | 4862/10712 [1:00:44<47:53,  2.04it/s]
 45%|████▌     | 4863/10712 [1:00:44<47:52,  2.04it/s]
 45%|████▌     | 4864/10712 [1:00:45<47:52,  2.04it/s]
 45%|████▌     | 4865/10712 [1:00:45<47:52,  2.04it/s]
 45%|████▌     | 4866/10712 [1:00:46<47:52,  2.04it/s]
 45%|████▌     | 4867/10712 [1:00:46<47:52,  2.03it/s]
 45%|████▌     | 4868/10712 [1:00:47<47:51,  2.04it/s]
 45%|████▌     | 4869/10712 [1:00:47<47:48,  2.04it/s]
 45%|████▌     | 4870/10712 [1:00:48<47:50,  2.03it/s]
 45%|████▌     | 4871/10712 [1:00:48<47:50,  2.03it/s]
 45%|████▌     | 4872/10712 [1:00:49<47:47,  2.04it/s]
 45%|████▌     | 4873/10712 [1:00:49<47:46,  2.04it/s]
 46%|████▌     | 4874/10712 [1:00:50<47:44,  2.04it/s]
 46%|████▌     | 4875/10712 [1:00:50<47:40,  2.04it/s]{'loss': 3.7842, 'grad_norm': 0.21419645845890045, 'learning_rate': 0.0006626984811363616, 'epoch': 0.46}
                                                      

 46%|████▌     | 4875/10712 [1:00:50<47:40,  2.04it/s]
 46%|████▌     | 4876/10712 [1:00:51<47:48,  2.03it/s]
 46%|████▌     | 4877/10712 [1:00:51<47:44,  2.04it/s]
 46%|████▌     | 4878/10712 [1:00:52<47:40,  2.04it/s]
 46%|████▌     | 4879/10712 [1:00:52<47:41,  2.04it/s]
 46%|████▌     | 4880/10712 [1:00:53<47:42,  2.04it/s]
 46%|████▌     | 4881/10712 [1:00:53<47:38,  2.04it/s]
 46%|████▌     | 4882/10712 [1:00:54<47:36,  2.04it/s]
 46%|████▌     | 4883/10712 [1:00:54<47:42,  2.04it/s]
 46%|████▌     | 4884/10712 [1:00:55<47:43,  2.04it/s]
 46%|████▌     | 4885/10712 [1:00:55<47:43,  2.03it/s]
 46%|████▌     | 4886/10712 [1:00:56<47:44,  2.03it/s]
 46%|████▌     | 4887/10712 [1:00:56<47:39,  2.04it/s]
 46%|████▌     | 4888/10712 [1:00:57<47:38,  2.04it/s]
 46%|████▌     | 4889/10712 [1:00:57<47:38,  2.04it/s]
 46%|████▌     | 4890/10712 [1:00:58<47:38,  2.04it/s]
 46%|████▌     | 4891/10712 [1:00:58<47:35,  2.04it/s]
 46%|████▌     | 4892/10712 [1:00:59<47:36,  2.04it/s]
 46%|████▌     | 4893/10712 [1:00:59<47:38,  2.04it/s]
 46%|████▌     | 4894/10712 [1:01:00<47:36,  2.04it/s]
 46%|████▌     | 4895/10712 [1:01:00<47:32,  2.04it/s]
 46%|████▌     | 4896/10712 [1:01:01<47:33,  2.04it/s]
 46%|████▌     | 4897/10712 [1:01:01<47:32,  2.04it/s]
 46%|████▌     | 4898/10712 [1:01:02<47:31,  2.04it/s]
 46%|████▌     | 4899/10712 [1:01:02<47:30,  2.04it/s]
 46%|████▌     | 4900/10712 [1:01:03<47:30,  2.04it/s]
                                                      {'loss': 3.7813, 'grad_norm': 0.21029923856258392, 'learning_rate': 0.0006588411797945836, 'epoch': 0.46}


 46%|████▌     | 4900/10712 [1:01:03<47:30,  2.04it/s]
 46%|████▌     | 4901/10712 [1:01:03<47:36,  2.03it/s]
 46%|████▌     | 4902/10712 [1:01:04<47:34,  2.04it/s]
 46%|████▌     | 4903/10712 [1:01:04<47:31,  2.04it/s]
 46%|████▌     | 4904/10712 [1:01:05<47:28,  2.04it/s]
 46%|████▌     | 4905/10712 [1:01:05<47:30,  2.04it/s]
 46%|████▌     | 4906/10712 [1:01:06<47:29,  2.04it/s]
 46%|████▌     | 4907/10712 [1:01:06<47:25,  2.04it/s]
 46%|████▌     | 4908/10712 [1:01:06<47:25,  2.04it/s]
 46%|████▌     | 4909/10712 [1:01:07<47:28,  2.04it/s]
 46%|████▌     | 4910/10712 [1:01:07<47:27,  2.04it/s]
 46%|████▌     | 4911/10712 [1:01:08<47:25,  2.04it/s]
 46%|████▌     | 4912/10712 [1:01:08<47:27,  2.04it/s]
 46%|████▌     | 4913/10712 [1:01:09<47:28,  2.04it/s]
 46%|████▌     | 4914/10712 [1:01:09<47:23,  2.04it/s]
 46%|████▌     | 4915/10712 [1:01:10<47:20,  2.04it/s]
 46%|████▌     | 4916/10712 [1:01:10<47:20,  2.04it/s]
 46%|████▌     | 4917/10712 [1:01:11<47:19,  2.04it/s]
 46%|████▌     | 4918/10712 [1:01:11<47:20,  2.04it/s]
 46%|████▌     | 4919/10712 [1:01:12<47:21,  2.04it/s]
 46%|████▌     | 4920/10712 [1:01:12<47:20,  2.04it/s]
 46%|████▌     | 4921/10712 [1:01:13<47:17,  2.04it/s]
 46%|████▌     | 4922/10712 [1:01:13<47:16,  2.04it/s]
 46%|████▌     | 4923/10712 [1:01:14<47:20,  2.04it/s]
 46%|████▌     | 4924/10712 [1:01:14<47:19,  2.04it/s]
 46%|████▌     | 4925/10712 [1:01:15<47:17,  2.04it/s]{'loss': 3.7887, 'grad_norm': 0.2143683135509491, 'learning_rate': 0.0006549733349140982, 'epoch': 0.46}

                                                      

 46%|████▌     | 4925/10712 [1:01:15<47:17,  2.04it/s]
 46%|████▌     | 4926/10712 [1:01:15<47:25,  2.03it/s]
 46%|████▌     | 4927/10712 [1:01:16<47:21,  2.04it/s]
 46%|████▌     | 4928/10712 [1:01:16<47:20,  2.04it/s]
 46%|████▌     | 4929/10712 [1:01:17<47:19,  2.04it/s]
 46%|████▌     | 4930/10712 [1:01:17<47:19,  2.04it/s]
 46%|████▌     | 4931/10712 [1:01:18<47:16,  2.04it/s]
 46%|████▌     | 4932/10712 [1:01:18<47:17,  2.04it/s]
 46%|████▌     | 4933/10712 [1:01:19<47:15,  2.04it/s]
 46%|████▌     | 4934/10712 [1:01:19<47:16,  2.04it/s]
 46%|████▌     | 4935/10712 [1:01:20<47:18,  2.03it/s]
 46%|████▌     | 4936/10712 [1:01:20<47:16,  2.04it/s]
 46%|████▌     | 4937/10712 [1:01:21<47:10,  2.04it/s]
 46%|████▌     | 4938/10712 [1:01:21<47:14,  2.04it/s]
 46%|████▌     | 4939/10712 [1:01:22<47:15,  2.04it/s]
 46%|████▌     | 4940/10712 [1:01:22<47:11,  2.04it/s]
 46%|████▌     | 4941/10712 [1:01:23<47:15,  2.04it/s]
 46%|████▌     | 4942/10712 [1:01:23<47:14,  2.04it/s]
 46%|████▌     | 4943/10712 [1:01:24<47:09,  2.04it/s]
 46%|████▌     | 4944/10712 [1:01:24<47:08,  2.04it/s]
 46%|████▌     | 4945/10712 [1:01:25<47:09,  2.04it/s]
 46%|████▌     | 4946/10712 [1:01:25<47:06,  2.04it/s]
 46%|████▌     | 4947/10712 [1:01:26<47:04,  2.04it/s]
 46%|████▌     | 4948/10712 [1:01:26<47:06,  2.04it/s]
 46%|████▌     | 4949/10712 [1:01:27<47:05,  2.04it/s]
 46%|████▌     | 4950/10712 [1:01:27<47:06,  2.04it/s]{'loss': 3.7786, 'grad_norm': 0.21597060561180115, 'learning_rate': 0.0006510952032341988, 'epoch': 0.46}
                                                      

 46%|████▌     | 4950/10712 [1:01:27<47:06,  2.04it/s]
 46%|████▌     | 4951/10712 [1:01:28<47:10,  2.04it/s]
 46%|████▌     | 4952/10712 [1:01:28<47:08,  2.04it/s]
 46%|████▌     | 4953/10712 [1:01:29<47:07,  2.04it/s]
 46%|████▌     | 4954/10712 [1:01:29<47:06,  2.04it/s]
 46%|████▋     | 4955/10712 [1:01:30<47:06,  2.04it/s]
 46%|████▋     | 4956/10712 [1:01:30<47:03,  2.04it/s]
 46%|████▋     | 4957/10712 [1:01:31<47:03,  2.04it/s]
 46%|████▋     | 4958/10712 [1:01:31<47:04,  2.04it/s]
 46%|████▋     | 4959/10712 [1:01:32<47:05,  2.04it/s]
 46%|████▋     | 4960/10712 [1:01:32<47:06,  2.04it/s]
 46%|████▋     | 4961/10712 [1:01:33<47:05,  2.04it/s]
 46%|████▋     | 4962/10712 [1:01:33<47:02,  2.04it/s]
 46%|████▋     | 4963/10712 [1:01:33<46:59,  2.04it/s]
 46%|████▋     | 4964/10712 [1:01:34<46:59,  2.04it/s]
 46%|████▋     | 4965/10712 [1:01:34<47:00,  2.04it/s]
 46%|████▋     | 4966/10712 [1:01:35<46:57,  2.04it/s]
 46%|████▋     | 4967/10712 [1:01:35<46:59,  2.04it/s]
 46%|████▋     | 4968/10712 [1:01:36<46:59,  2.04it/s]
 46%|████▋     | 4969/10712 [1:01:36<46:57,  2.04it/s]
 46%|████▋     | 4970/10712 [1:01:37<47:15,  2.02it/s]
 46%|████▋     | 4971/10712 [1:01:37<47:08,  2.03it/s]
 46%|████▋     | 4972/10712 [1:01:38<47:05,  2.03it/s]
 46%|████▋     | 4973/10712 [1:01:38<47:00,  2.03it/s]
 46%|████▋     | 4974/10712 [1:01:39<46:55,  2.04it/s]
 46%|████▋     | 4975/10712 [1:01:39<46:59,  2.03it/s]{'loss': 3.7807, 'grad_norm': 0.20121075212955475, 'learning_rate': 0.000647207042176994, 'epoch': 0.46}
                                                      

 46%|████▋     | 4975/10712 [1:01:39<46:59,  2.03it/s]
 46%|████▋     | 4976/10712 [1:01:40<47:00,  2.03it/s]
 46%|████▋     | 4977/10712 [1:01:40<46:56,  2.04it/s]
 46%|████▋     | 4978/10712 [1:01:41<46:57,  2.04it/s]
 46%|████▋     | 4979/10712 [1:01:41<46:56,  2.04it/s]
 46%|████▋     | 4980/10712 [1:01:42<46:55,  2.04it/s]
 46%|████▋     | 4981/10712 [1:01:42<46:51,  2.04it/s]
 47%|████▋     | 4982/10712 [1:01:43<46:51,  2.04it/s]
 47%|████▋     | 4983/10712 [1:01:43<46:52,  2.04it/s]
 47%|████▋     | 4984/10712 [1:01:44<46:50,  2.04it/s]
 47%|████▋     | 4985/10712 [1:01:44<46:53,  2.04it/s]
 47%|████▋     | 4986/10712 [1:01:45<46:53,  2.04it/s]
 47%|████▋     | 4987/10712 [1:01:45<46:51,  2.04it/s]
 47%|████▋     | 4988/10712 [1:01:46<46:53,  2.03it/s]
 47%|████▋     | 4989/10712 [1:01:46<46:51,  2.04it/s]
 47%|████▋     | 4990/10712 [1:01:47<46:48,  2.04it/s]
 47%|████▋     | 4991/10712 [1:01:47<46:47,  2.04it/s]
 47%|████▋     | 4992/10712 [1:01:48<46:48,  2.04it/s]
 47%|████▋     | 4993/10712 [1:01:48<46:46,  2.04it/s]
 47%|████▋     | 4994/10712 [1:01:49<46:45,  2.04it/s]
 47%|████▋     | 4995/10712 [1:01:49<46:47,  2.04it/s]
 47%|████▋     | 4996/10712 [1:01:50<46:45,  2.04it/s]
 47%|████▋     | 4997/10712 [1:01:50<46:42,  2.04it/s]
 47%|████▋     | 4998/10712 [1:01:51<46:44,  2.04it/s]
 47%|████▋     | 4999/10712 [1:01:51<46:43,  2.04it/s]
 47%|████▋     | 5000/10712 [1:01:52<46:42,  2.04it/s]{'loss': 3.7762, 'grad_norm': 0.241677924990654, 'learning_rate': 0.0006433091098303216, 'epoch': 0.47}

                                                      

 47%|████▋     | 5000/10712 [1:01:52<46:42,  2.04it/s]
 47%|████▋     | 5001/10712 [1:01:52<46:45,  2.04it/s]
 47%|████▋     | 5002/10712 [1:01:53<46:47,  2.03it/s]
 47%|████▋     | 5003/10712 [1:01:53<46:44,  2.04it/s]
 47%|████▋     | 5004/10712 [1:01:54<46:42,  2.04it/s]
 47%|████▋     | 5005/10712 [1:01:54<46:42,  2.04it/s]
 47%|████▋     | 5006/10712 [1:01:55<46:41,  2.04it/s]
 47%|████▋     | 5007/10712 [1:01:55<46:36,  2.04it/s]
 47%|████▋     | 5008/10712 [1:01:56<46:37,  2.04it/s]
 47%|████▋     | 5009/10712 [1:01:56<46:37,  2.04it/s]
 47%|████▋     | 5010/10712 [1:01:57<46:35,  2.04it/s]
 47%|████▋     | 5011/10712 [1:01:57<46:40,  2.04it/s]
 47%|████▋     | 5012/10712 [1:01:58<46:41,  2.03it/s]
 47%|████▋     | 5013/10712 [1:01:58<46:36,  2.04it/s]
 47%|████▋     | 5014/10712 [1:01:59<46:36,  2.04it/s]
 47%|████▋     | 5015/10712 [1:01:59<46:38,  2.04it/s]
 47%|████▋     | 5016/10712 [1:02:00<46:35,  2.04it/s]
 47%|████▋     | 5017/10712 [1:02:00<46:33,  2.04it/s]
 47%|████▋     | 5018/10712 [1:02:00<46:37,  2.04it/s]
 47%|████▋     | 5019/10712 [1:02:01<46:38,  2.03it/s]
 47%|████▋     | 5020/10712 [1:02:01<46:34,  2.04it/s]
 47%|████▋     | 5021/10712 [1:02:02<46:37,  2.03it/s]
 47%|████▋     | 5022/10712 [1:02:02<46:35,  2.04it/s]
 47%|████▋     | 5023/10712 [1:02:03<46:34,  2.04it/s]
 47%|████▋     | 5024/10712 [1:02:03<46:33,  2.04it/s]
 47%|████▋     | 5025/10712 [1:02:04<46:31,  2.04it/s]{'loss': 3.7758, 'grad_norm': 0.22356069087982178, 'learning_rate': 0.0006394016649306164, 'epoch': 0.47}

                                                      

 47%|████▋     | 5025/10712 [1:02:04<46:31,  2.04it/s]
 47%|████▋     | 5026/10712 [1:02:04<46:35,  2.03it/s]
 47%|████▋     | 5027/10712 [1:02:05<46:37,  2.03it/s]
 47%|████▋     | 5028/10712 [1:02:05<46:32,  2.04it/s]
 47%|████▋     | 5029/10712 [1:02:06<46:33,  2.03it/s]
 47%|████▋     | 5030/10712 [1:02:06<46:32,  2.03it/s]
 47%|████▋     | 5031/10712 [1:02:07<46:29,  2.04it/s]
 47%|████▋     | 5032/10712 [1:02:07<46:27,  2.04it/s]
 47%|████▋     | 5033/10712 [1:02:08<46:26,  2.04it/s]
 47%|████▋     | 5034/10712 [1:02:08<46:26,  2.04it/s]
 47%|████▋     | 5035/10712 [1:02:09<46:25,  2.04it/s]
 47%|████▋     | 5036/10712 [1:02:09<46:25,  2.04it/s]
 47%|████▋     | 5037/10712 [1:02:10<46:21,  2.04it/s]
 47%|████▋     | 5038/10712 [1:02:10<46:23,  2.04it/s]
 47%|████▋     | 5039/10712 [1:02:11<46:23,  2.04it/s]
 47%|████▋     | 5040/10712 [1:02:11<46:20,  2.04it/s]
 47%|████▋     | 5041/10712 [1:02:12<46:17,  2.04it/s]
 47%|████▋     | 5042/10712 [1:02:12<46:16,  2.04it/s]
 47%|████▋     | 5043/10712 [1:02:13<46:17,  2.04it/s]
 47%|████▋     | 5044/10712 [1:02:13<46:22,  2.04it/s]
 47%|████▋     | 5045/10712 [1:02:14<46:19,  2.04it/s]
 47%|████▋     | 5046/10712 [1:02:14<46:16,  2.04it/s]
 47%|████▋     | 5047/10712 [1:02:15<46:18,  2.04it/s]
 47%|████▋     | 5048/10712 [1:02:15<46:17,  2.04it/s]
 47%|████▋     | 5049/10712 [1:02:16<46:15,  2.04it/s]
 47%|████▋     | 5050/10712 [1:02:16<46:15,  2.04it/s]{'loss': 3.773, 'grad_norm': 0.20654357969760895, 'learning_rate': 0.0006354849668457362, 'epoch': 0.47}

                                                      

 47%|████▋     | 5050/10712 [1:02:16<46:15,  2.04it/s]
 47%|████▋     | 5051/10712 [1:02:17<46:18,  2.04it/s]
 47%|████▋     | 5052/10712 [1:02:17<46:19,  2.04it/s]
 47%|████▋     | 5053/10712 [1:02:18<46:17,  2.04it/s]
 47%|████▋     | 5054/10712 [1:02:18<46:19,  2.04it/s]
 47%|████▋     | 5055/10712 [1:02:19<46:17,  2.04it/s]
 47%|████▋     | 5056/10712 [1:02:19<46:13,  2.04it/s]
 47%|████▋     | 5057/10712 [1:02:20<46:15,  2.04it/s]
 47%|████▋     | 5058/10712 [1:02:20<46:14,  2.04it/s]
 47%|████▋     | 5059/10712 [1:02:21<46:12,  2.04it/s]
 47%|████▋     | 5060/10712 [1:02:21<46:13,  2.04it/s]
 47%|████▋     | 5061/10712 [1:02:22<46:15,  2.04it/s]
 47%|████▋     | 5062/10712 [1:02:22<46:13,  2.04it/s]
 47%|████▋     | 5063/10712 [1:02:23<46:12,  2.04it/s]
 47%|████▋     | 5064/10712 [1:02:23<46:13,  2.04it/s]
 47%|████▋     | 5065/10712 [1:02:24<46:11,  2.04it/s]
 47%|████▋     | 5066/10712 [1:02:24<46:11,  2.04it/s]
 47%|████▋     | 5067/10712 [1:02:25<46:14,  2.03it/s]
 47%|████▋     | 5068/10712 [1:02:25<46:09,  2.04it/s]
 47%|████▋     | 5069/10712 [1:02:26<46:06,  2.04it/s]
 47%|████▋     | 5070/10712 [1:02:26<46:07,  2.04it/s]
 47%|████▋     | 5071/10712 [1:02:26<46:07,  2.04it/s]
 47%|████▋     | 5072/10712 [1:02:27<46:07,  2.04it/s]
 47%|████▋     | 5073/10712 [1:02:27<46:08,  2.04it/s]
 47%|████▋     | 5074/10712 [1:02:28<46:08,  2.04it/s]
 47%|████▋     | 5075/10712 [1:02:28<46:04,  2.04it/s]{'loss': 3.7717, 'grad_norm': 0.2268342524766922, 'learning_rate': 0.0006315592755577455, 'epoch': 0.47}

                                                      

 47%|████▋     | 5075/10712 [1:02:28<46:04,  2.04it/s]
 47%|████▋     | 5076/10712 [1:02:29<46:10,  2.03it/s]
 47%|████▋     | 5077/10712 [1:02:29<46:09,  2.03it/s]
 47%|████▋     | 5078/10712 [1:02:30<46:05,  2.04it/s]
 47%|████▋     | 5079/10712 [1:02:30<46:05,  2.04it/s]
 47%|████▋     | 5080/10712 [1:02:31<46:07,  2.03it/s]
 47%|████▋     | 5081/10712 [1:02:31<46:04,  2.04it/s]
 47%|████▋     | 5082/10712 [1:02:32<46:02,  2.04it/s]
 47%|████▋     | 5083/10712 [1:02:32<46:04,  2.04it/s]
 47%|████▋     | 5084/10712 [1:02:33<46:03,  2.04it/s]
 47%|████▋     | 5085/10712 [1:02:33<46:00,  2.04it/s]
 47%|████▋     | 5086/10712 [1:02:34<46:01,  2.04it/s]
 47%|████▋     | 5087/10712 [1:02:34<46:00,  2.04it/s]
 47%|████▋     | 5088/10712 [1:02:35<45:55,  2.04it/s]
 48%|████▊     | 5089/10712 [1:02:35<45:59,  2.04it/s]
 48%|████▊     | 5090/10712 [1:02:36<45:57,  2.04it/s]
 48%|████▊     | 5091/10712 [1:02:36<45:55,  2.04it/s]
 48%|████▊     | 5092/10712 [1:02:37<45:54,  2.04it/s]
 48%|████▊     | 5093/10712 [1:02:37<45:57,  2.04it/s]
 48%|████▊     | 5094/10712 [1:02:38<45:56,  2.04it/s]
 48%|████▊     | 5095/10712 [1:02:38<45:53,  2.04it/s]
 48%|████▊     | 5096/10712 [1:02:39<45:53,  2.04it/s]
 48%|████▊     | 5097/10712 [1:02:39<45:53,  2.04it/s]
 48%|████▊     | 5098/10712 [1:02:40<45:52,  2.04it/s]
 48%|████▊     | 5099/10712 [1:02:40<45:49,  2.04it/s]
 48%|████▊     | 5100/10712 [1:02:41<45:52,  2.04it/s]
                                                      {'loss': 3.766, 'grad_norm': 0.2130841612815857, 'learning_rate': 0.0006276248516456589, 'epoch': 0.48}


 48%|████▊     | 5100/10712 [1:02:41<45:52,  2.04it/s]
 48%|████▊     | 5101/10712 [1:02:41<45:53,  2.04it/s]
 48%|████▊     | 5102/10712 [1:02:42<45:53,  2.04it/s]
 48%|████▊     | 5103/10712 [1:02:42<45:55,  2.04it/s]
 48%|████▊     | 5104/10712 [1:02:43<45:52,  2.04it/s]
 48%|████▊     | 5105/10712 [1:02:43<45:49,  2.04it/s]
 48%|████▊     | 5106/10712 [1:02:44<45:52,  2.04it/s]
 48%|████▊     | 5107/10712 [1:02:44<45:51,  2.04it/s]
 48%|████▊     | 5108/10712 [1:02:45<45:48,  2.04it/s]
 48%|████▊     | 5109/10712 [1:02:45<45:45,  2.04it/s]
 48%|████▊     | 5110/10712 [1:02:46<45:48,  2.04it/s]
 48%|████▊     | 5111/10712 [1:02:46<45:49,  2.04it/s]
 48%|████▊     | 5112/10712 [1:02:47<45:46,  2.04it/s]
 48%|████▊     | 5113/10712 [1:02:47<45:44,  2.04it/s]
 48%|████▊     | 5114/10712 [1:02:48<45:48,  2.04it/s]
 48%|████▊     | 5115/10712 [1:02:48<45:47,  2.04it/s]
 48%|████▊     | 5116/10712 [1:02:49<45:45,  2.04it/s]
 48%|████▊     | 5117/10712 [1:02:49<45:48,  2.04it/s]
 48%|████▊     | 5118/10712 [1:02:50<45:45,  2.04it/s]
 48%|████▊     | 5119/10712 [1:02:50<45:44,  2.04it/s]
 48%|████▊     | 5120/10712 [1:02:51<45:43,  2.04it/s]
 48%|████▊     | 5121/10712 [1:02:51<45:46,  2.04it/s]
 48%|████▊     | 5122/10712 [1:02:52<45:42,  2.04it/s]
 48%|████▊     | 5123/10712 [1:02:52<45:38,  2.04it/s]
 48%|████▊     | 5124/10712 [1:02:53<45:38,  2.04it/s]
 48%|████▊     | 5125/10712 [1:02:53<45:42,  2.04it/s]{'loss': 3.7727, 'grad_norm': 0.21181483566761017, 'learning_rate': 0.0006236819562681432, 'epoch': 0.48}
                                                      

 48%|████▊     | 5125/10712 [1:02:53<45:42,  2.04it/s]
 48%|████▊     | 5126/10712 [1:02:53<45:41,  2.04it/s]
 48%|████▊     | 5127/10712 [1:02:54<45:48,  2.03it/s]
 48%|████▊     | 5128/10712 [1:02:54<45:44,  2.03it/s]
 48%|████▊     | 5129/10712 [1:02:55<45:41,  2.04it/s]
 48%|████▊     | 5130/10712 [1:02:55<45:43,  2.03it/s]
 48%|████▊     | 5131/10712 [1:02:56<45:40,  2.04it/s]
 48%|████▊     | 5132/10712 [1:02:56<45:38,  2.04it/s]
 48%|████▊     | 5133/10712 [1:02:57<45:37,  2.04it/s]
 48%|████▊     | 5134/10712 [1:02:57<45:38,  2.04it/s]
 48%|████▊     | 5135/10712 [1:02:58<45:38,  2.04it/s]
 48%|████▊     | 5136/10712 [1:02:58<45:34,  2.04it/s]
 48%|████▊     | 5137/10712 [1:02:59<45:38,  2.04it/s]
 48%|████▊     | 5138/10712 [1:02:59<45:37,  2.04it/s]
 48%|████▊     | 5139/10712 [1:03:00<45:35,  2.04it/s]
 48%|████▊     | 5140/10712 [1:03:00<45:34,  2.04it/s]
 48%|████▊     | 5141/10712 [1:03:01<45:35,  2.04it/s]
 48%|████▊     | 5142/10712 [1:03:01<45:36,  2.04it/s]
 48%|████▊     | 5143/10712 [1:03:02<45:34,  2.04it/s]
 48%|████▊     | 5144/10712 [1:03:02<45:36,  2.03it/s]
 48%|████▊     | 5145/10712 [1:03:03<45:35,  2.03it/s]
 48%|████▊     | 5146/10712 [1:03:03<45:36,  2.03it/s]
 48%|████▊     | 5147/10712 [1:03:04<45:35,  2.03it/s]
 48%|████▊     | 5148/10712 [1:03:04<45:32,  2.04it/s]
 48%|████▊     | 5149/10712 [1:03:05<45:30,  2.04it/s]
 48%|████▊     | 5150/10712 [1:03:05<45:31,  2.04it/s]{'loss': 3.7764, 'grad_norm': 0.21644580364227295, 'learning_rate': 0.0006197308511461836, 'epoch': 0.48}

                                                      

 48%|████▊     | 5150/10712 [1:03:05<45:31,  2.04it/s]
 48%|████▊     | 5151/10712 [1:03:06<45:36,  2.03it/s]
 48%|████▊     | 5152/10712 [1:03:06<45:37,  2.03it/s]
 48%|████▊     | 5153/10712 [1:03:07<45:34,  2.03it/s]
 48%|████▊     | 5154/10712 [1:03:07<45:30,  2.04it/s]
 48%|████▊     | 5155/10712 [1:03:08<45:29,  2.04it/s]
 48%|████▊     | 5156/10712 [1:03:08<45:28,  2.04it/s]
 48%|████▊     | 5157/10712 [1:03:09<45:26,  2.04it/s]
 48%|████▊     | 5158/10712 [1:03:09<45:23,  2.04it/s]
 48%|████▊     | 5159/10712 [1:03:10<45:28,  2.04it/s]
 48%|████▊     | 5160/10712 [1:03:10<45:26,  2.04it/s]
 48%|████▊     | 5161/10712 [1:03:11<45:26,  2.04it/s]
 48%|████▊     | 5162/10712 [1:03:11<45:28,  2.03it/s]
 48%|████▊     | 5163/10712 [1:03:12<45:24,  2.04it/s]
 48%|████▊     | 5164/10712 [1:03:12<45:21,  2.04it/s]
 48%|████▊     | 5165/10712 [1:03:13<45:22,  2.04it/s]
 48%|████▊     | 5166/10712 [1:03:13<45:23,  2.04it/s]
 48%|████▊     | 5167/10712 [1:03:14<45:22,  2.04it/s]
 48%|████▊     | 5168/10712 [1:03:14<45:18,  2.04it/s]
 48%|████▊     | 5169/10712 [1:03:15<45:20,  2.04it/s]
 48%|████▊     | 5170/10712 [1:03:15<45:19,  2.04it/s]
 48%|████▊     | 5171/10712 [1:03:16<45:17,  2.04it/s]
 48%|████▊     | 5172/10712 [1:03:16<45:17,  2.04it/s]
 48%|████▊     | 5173/10712 [1:03:17<45:18,  2.04it/s]
 48%|████▊     | 5174/10712 [1:03:17<45:20,  2.04it/s]
 48%|████▊     | 5175/10712 [1:03:18<45:19,  2.04it/s]{'loss': 3.7771, 'grad_norm': 0.20509040355682373, 'learning_rate': 0.0006157717985457101, 'epoch': 0.48}

                                                      

 48%|████▊     | 5175/10712 [1:03:18<45:19,  2.04it/s]
 48%|████▊     | 5176/10712 [1:03:18<45:21,  2.03it/s]
 48%|████▊     | 5177/10712 [1:03:19<45:24,  2.03it/s]
 48%|████▊     | 5178/10712 [1:03:19<45:20,  2.03it/s]
 48%|████▊     | 5179/10712 [1:03:20<45:19,  2.03it/s]
 48%|████▊     | 5180/10712 [1:03:20<45:20,  2.03it/s]
 48%|████▊     | 5181/10712 [1:03:21<45:18,  2.03it/s]
 48%|████▊     | 5182/10712 [1:03:21<45:15,  2.04it/s]
 48%|████▊     | 5183/10712 [1:03:21<45:11,  2.04it/s]
 48%|████▊     | 5184/10712 [1:03:22<45:14,  2.04it/s]
 48%|████▊     | 5185/10712 [1:03:22<45:16,  2.03it/s]
 48%|████▊     | 5186/10712 [1:03:23<45:10,  2.04it/s]
 48%|████▊     | 5187/10712 [1:03:23<45:11,  2.04it/s]
 48%|████▊     | 5188/10712 [1:03:24<45:11,  2.04it/s]
 48%|████▊     | 5189/10712 [1:03:24<45:10,  2.04it/s]
 48%|████▊     | 5190/10712 [1:03:25<45:06,  2.04it/s]
 48%|████▊     | 5191/10712 [1:03:25<45:04,  2.04it/s]
 48%|████▊     | 5192/10712 [1:03:26<45:08,  2.04it/s]
 48%|████▊     | 5193/10712 [1:03:26<45:06,  2.04it/s]
 48%|████▊     | 5194/10712 [1:03:27<45:07,  2.04it/s]
 48%|████▊     | 5195/10712 [1:03:27<45:10,  2.04it/s]
 49%|████▊     | 5196/10712 [1:03:28<45:07,  2.04it/s]
 49%|████▊     | 5197/10712 [1:03:28<45:08,  2.04it/s]
 49%|████▊     | 5198/10712 [1:03:29<45:09,  2.04it/s]
 49%|████▊     | 5199/10712 [1:03:29<45:06,  2.04it/s]
 49%|████▊     | 5200/10712 [1:03:30<45:02,  2.04it/s]{'loss': 3.7687, 'grad_norm': 0.21554921567440033, 'learning_rate': 0.00061180506126019, 'epoch': 0.49}
                                                      

 49%|████▊     | 5200/10712 [1:03:30<45:02,  2.04it/s]
 49%|████▊     | 5201/10712 [1:03:30<45:10,  2.03it/s]
 49%|████▊     | 5202/10712 [1:03:31<45:09,  2.03it/s]
 49%|████▊     | 5203/10712 [1:03:31<45:02,  2.04it/s]
 49%|████▊     | 5204/10712 [1:03:32<45:04,  2.04it/s]
 49%|████▊     | 5205/10712 [1:03:32<45:04,  2.04it/s]
 49%|████▊     | 5206/10712 [1:03:33<45:00,  2.04it/s]
 49%|████▊     | 5207/10712 [1:03:33<44:59,  2.04it/s]
 49%|████▊     | 5208/10712 [1:03:34<45:00,  2.04it/s]
 49%|████▊     | 5209/10712 [1:03:34<45:00,  2.04it/s]
 49%|████▊     | 5210/10712 [1:03:35<44:59,  2.04it/s]
 49%|████▊     | 5211/10712 [1:03:35<45:00,  2.04it/s]
 49%|████▊     | 5212/10712 [1:03:36<44:58,  2.04it/s]
 49%|████▊     | 5213/10712 [1:03:36<44:55,  2.04it/s]
 49%|████▊     | 5214/10712 [1:03:37<44:59,  2.04it/s]
 49%|████▊     | 5215/10712 [1:03:37<44:56,  2.04it/s]
 49%|████▊     | 5216/10712 [1:03:38<44:51,  2.04it/s]
 49%|████▊     | 5217/10712 [1:03:38<44:51,  2.04it/s]
 49%|████▊     | 5218/10712 [1:03:39<44:50,  2.04it/s]
 49%|████▊     | 5219/10712 [1:03:39<44:51,  2.04it/s]
 49%|████▊     | 5220/10712 [1:03:40<44:49,  2.04it/s]
 49%|████▊     | 5221/10712 [1:03:40<44:50,  2.04it/s]
 49%|████▊     | 5222/10712 [1:03:41<44:54,  2.04it/s]
 49%|████▉     | 5223/10712 [1:03:41<44:52,  2.04it/s]
 49%|████▉     | 5224/10712 [1:03:42<44:50,  2.04it/s]
 49%|████▉     | 5225/10712 [1:03:42<44:53,  2.04it/s]{'loss': 3.7662, 'grad_norm': 0.20557540655136108, 'learning_rate': 0.0006078309025931831, 'epoch': 0.49}

                                                      

 49%|████▉     | 5225/10712 [1:03:42<44:53,  2.04it/s]
 49%|████▉     | 5226/10712 [1:03:43<44:55,  2.04it/s]
 49%|████▉     | 5227/10712 [1:03:43<44:54,  2.04it/s]
 49%|████▉     | 5228/10712 [1:03:44<44:54,  2.03it/s]
 49%|████▉     | 5229/10712 [1:03:44<44:52,  2.04it/s]
 49%|████▉     | 5230/10712 [1:03:45<44:50,  2.04it/s]
 49%|████▉     | 5231/10712 [1:03:45<44:52,  2.04it/s]
 49%|████▉     | 5232/10712 [1:03:46<44:50,  2.04it/s]
 49%|████▉     | 5233/10712 [1:03:46<44:48,  2.04it/s]
 49%|████▉     | 5234/10712 [1:03:47<44:49,  2.04it/s]
 49%|████▉     | 5235/10712 [1:03:47<44:49,  2.04it/s]
 49%|████▉     | 5236/10712 [1:03:47<44:47,  2.04it/s]
 49%|████▉     | 5237/10712 [1:03:48<44:47,  2.04it/s]
 49%|████▉     | 5238/10712 [1:03:48<44:45,  2.04it/s]
 49%|████▉     | 5239/10712 [1:03:49<44:46,  2.04it/s]
 49%|████▉     | 5240/10712 [1:03:49<44:45,  2.04it/s]
 49%|████▉     | 5241/10712 [1:03:50<44:46,  2.04it/s]
 49%|████▉     | 5242/10712 [1:03:50<44:42,  2.04it/s]
 49%|████▉     | 5243/10712 [1:03:51<44:39,  2.04it/s]
 49%|████▉     | 5244/10712 [1:03:51<44:43,  2.04it/s]
 49%|████▉     | 5245/10712 [1:03:52<44:43,  2.04it/s]
 49%|████▉     | 5246/10712 [1:03:52<44:41,  2.04it/s]
 49%|████▉     | 5247/10712 [1:03:53<44:42,  2.04it/s]
 49%|████▉     | 5248/10712 [1:03:53<44:43,  2.04it/s]
 49%|████▉     | 5249/10712 [1:03:54<44:42,  2.04it/s]
 49%|████▉     | 5250/10712 [1:03:54<44:41,  2.04it/s]{'loss': 3.7582, 'grad_norm': 0.2203502058982849, 'learning_rate': 0.0006038495863408649, 'epoch': 0.49}

                                                      

 49%|████▉     | 5250/10712 [1:03:54<44:41,  2.04it/s]
 49%|████▉     | 5251/10712 [1:03:55<44:45,  2.03it/s]
 49%|████▉     | 5252/10712 [1:03:55<44:42,  2.04it/s]
 49%|████▉     | 5253/10712 [1:03:56<44:39,  2.04it/s]
 49%|████▉     | 5254/10712 [1:03:56<44:41,  2.04it/s]
 49%|████▉     | 5255/10712 [1:03:57<44:39,  2.04it/s]
 49%|████▉     | 5256/10712 [1:03:57<44:36,  2.04it/s]
 49%|████▉     | 5257/10712 [1:03:58<44:34,  2.04it/s]
 49%|████▉     | 5258/10712 [1:03:58<44:35,  2.04it/s]
 49%|████▉     | 5259/10712 [1:03:59<44:34,  2.04it/s]
 49%|████▉     | 5260/10712 [1:03:59<44:32,  2.04it/s]
 49%|████▉     | 5261/10712 [1:04:00<44:31,  2.04it/s]
 49%|████▉     | 5262/10712 [1:04:00<44:33,  2.04it/s]
 49%|████▉     | 5263/10712 [1:04:01<44:33,  2.04it/s]
 49%|████▉     | 5264/10712 [1:04:01<44:34,  2.04it/s]
 49%|████▉     | 5265/10712 [1:04:02<44:34,  2.04it/s]
 49%|████▉     | 5266/10712 [1:04:02<44:33,  2.04it/s]
 49%|████▉     | 5267/10712 [1:04:03<44:29,  2.04it/s]
 49%|████▉     | 5268/10712 [1:04:03<44:25,  2.04it/s]
 49%|████▉     | 5269/10712 [1:04:04<44:32,  2.04it/s]
 49%|████▉     | 5270/10712 [1:04:04<44:32,  2.04it/s]
 49%|████▉     | 5271/10712 [1:04:05<44:30,  2.04it/s]
 49%|████▉     | 5272/10712 [1:04:05<44:29,  2.04it/s]
 49%|████▉     | 5273/10712 [1:04:06<44:28,  2.04it/s]
 49%|████▉     | 5274/10712 [1:04:06<44:24,  2.04it/s]
 49%|████▉     | 5275/10712 [1:04:07<52:10,  1.74it/s]{'loss': 3.7597, 'grad_norm': 0.21822427213191986, 'learning_rate': 0.0005998613767745162, 'epoch': 0.49}

                                                      

 49%|████▉     | 5275/10712 [1:04:07<52:10,  1.74it/s]
 49%|████▉     | 5276/10712 [1:04:07<49:53,  1.82it/s]
 49%|████▉     | 5277/10712 [1:04:08<48:13,  1.88it/s]
 49%|████▉     | 5278/10712 [1:04:08<47:05,  1.92it/s]
 49%|████▉     | 5279/10712 [1:04:09<46:19,  1.95it/s]
 49%|████▉     | 5280/10712 [1:04:09<45:44,  1.98it/s]
 49%|████▉     | 5281/10712 [1:04:10<45:19,  2.00it/s]
 49%|████▉     | 5282/10712 [1:04:10<45:02,  2.01it/s]
 49%|████▉     | 5283/10712 [1:04:11<44:51,  2.02it/s]
 49%|████▉     | 5284/10712 [1:04:11<44:39,  2.03it/s]
 49%|████▉     | 5285/10712 [1:04:12<44:32,  2.03it/s]
 49%|████▉     | 5286/10712 [1:04:12<44:32,  2.03it/s]
 49%|████▉     | 5287/10712 [1:04:13<44:28,  2.03it/s]
 49%|████▉     | 5288/10712 [1:04:13<44:27,  2.03it/s]
 49%|████▉     | 5289/10712 [1:04:14<44:27,  2.03it/s]
 49%|████▉     | 5290/10712 [1:04:14<44:24,  2.03it/s]
 49%|████▉     | 5291/10712 [1:04:15<44:18,  2.04it/s]
 49%|████▉     | 5292/10712 [1:04:15<44:19,  2.04it/s]
 49%|████▉     | 5293/10712 [1:04:16<44:21,  2.04it/s]
 49%|████▉     | 5294/10712 [1:04:16<44:20,  2.04it/s]
 49%|████▉     | 5295/10712 [1:04:17<44:20,  2.04it/s]
 49%|████▉     | 5296/10712 [1:04:17<44:19,  2.04it/s]
 49%|████▉     | 5297/10712 [1:04:18<44:13,  2.04it/s]
 49%|████▉     | 5298/10712 [1:04:18<44:13,  2.04it/s]
 49%|████▉     | 5299/10712 [1:04:19<44:12,  2.04it/s]
 49%|████▉     | 5300/10712 [1:04:19<44:12,  2.04it/s]
                                                      {'loss': 3.7635, 'grad_norm': 0.2259405106306076, 'learning_rate': 0.0005958665386229811, 'epoch': 0.49}


 49%|████▉     | 5300/10712 [1:04:19<44:12,  2.04it/s]
 49%|████▉     | 5301/10712 [1:04:20<44:13,  2.04it/s]
 49%|████▉     | 5302/10712 [1:04:20<44:14,  2.04it/s]
 50%|████▉     | 5303/10712 [1:04:21<44:15,  2.04it/s]
 50%|████▉     | 5304/10712 [1:04:21<44:12,  2.04it/s]
 50%|████▉     | 5305/10712 [1:04:22<51:18,  1.76it/s]
 50%|████▉     | 5306/10712 [1:04:22<49:10,  1.83it/s]
 50%|████▉     | 5307/10712 [1:04:23<47:38,  1.89it/s]
 50%|████▉     | 5308/10712 [1:04:23<46:38,  1.93it/s]
 50%|████▉     | 5309/10712 [1:04:24<45:51,  1.96it/s]
 50%|████▉     | 5310/10712 [1:04:24<45:19,  1.99it/s]
 50%|████▉     | 5311/10712 [1:04:25<45:00,  2.00it/s]
 50%|████▉     | 5312/10712 [1:04:25<44:42,  2.01it/s]
 50%|████▉     | 5313/10712 [1:04:26<44:36,  2.02it/s]
 50%|████▉     | 5314/10712 [1:04:26<44:29,  2.02it/s]
 50%|████▉     | 5315/10712 [1:04:27<44:25,  2.03it/s]
 50%|████▉     | 5316/10712 [1:04:27<44:17,  2.03it/s]
 50%|████▉     | 5317/10712 [1:04:28<44:11,  2.03it/s]
 50%|████▉     | 5318/10712 [1:04:28<44:12,  2.03it/s]
 50%|████▉     | 5319/10712 [1:04:29<44:09,  2.04it/s]
 50%|████▉     | 5320/10712 [1:04:29<44:05,  2.04it/s]
 50%|████▉     | 5321/10712 [1:04:30<44:05,  2.04it/s]
 50%|████▉     | 5322/10712 [1:04:30<44:07,  2.04it/s]
 50%|████▉     | 5323/10712 [1:04:31<44:06,  2.04it/s]
 50%|████▉     | 5324/10712 [1:04:31<44:00,  2.04it/s]
 50%|████▉     | 5325/10712 [1:04:32<43:57,  2.04it/s]{'loss': 3.769, 'grad_norm': 0.20543403923511505, 'learning_rate': 0.0005918653370550951, 'epoch': 0.5}

                                                      

 50%|████▉     | 5325/10712 [1:04:32<43:57,  2.04it/s]
 50%|████▉     | 5326/10712 [1:04:32<44:05,  2.04it/s]
 50%|████▉     | 5327/10712 [1:04:33<44:03,  2.04it/s]
 50%|████▉     | 5328/10712 [1:04:33<44:02,  2.04it/s]
 50%|████▉     | 5329/10712 [1:04:34<44:00,  2.04it/s]
 50%|████▉     | 5330/10712 [1:04:34<44:00,  2.04it/s]
 50%|████▉     | 5331/10712 [1:04:35<43:58,  2.04it/s]
 50%|████▉     | 5332/10712 [1:04:35<43:57,  2.04it/s]
 50%|████▉     | 5333/10712 [1:04:36<43:59,  2.04it/s]
 50%|████▉     | 5334/10712 [1:04:36<43:55,  2.04it/s]
 50%|████▉     | 5335/10712 [1:04:37<43:53,  2.04it/s]
 50%|████▉     | 5336/10712 [1:04:37<43:53,  2.04it/s]
 50%|████▉     | 5337/10712 [1:04:38<43:54,  2.04it/s]
 50%|████▉     | 5338/10712 [1:04:38<43:52,  2.04it/s]
 50%|████▉     | 5339/10712 [1:04:39<43:49,  2.04it/s]
 50%|████▉     | 5340/10712 [1:04:39<43:51,  2.04it/s]
 50%|████▉     | 5341/10712 [1:04:40<43:52,  2.04it/s]
 50%|████▉     | 5342/10712 [1:04:40<43:51,  2.04it/s]
 50%|████▉     | 5343/10712 [1:04:41<43:49,  2.04it/s]
 50%|████▉     | 5344/10712 [1:04:41<43:50,  2.04it/s]
 50%|████▉     | 5345/10712 [1:04:42<43:50,  2.04it/s]
 50%|████▉     | 5346/10712 [1:04:42<43:47,  2.04it/s]
 50%|████▉     | 5347/10712 [1:04:42<43:45,  2.04it/s]
 50%|████▉     | 5348/10712 [1:04:43<43:46,  2.04it/s]
 50%|████▉     | 5349/10712 [1:04:43<43:47,  2.04it/s]
 50%|████▉     | 5350/10712 [1:04:44<43:47,  2.04it/s]
                                                      {'loss': 3.7526, 'grad_norm': 0.206803560256958, 'learning_rate': 0.0005878580376620838, 'epoch': 0.5}


 50%|████▉     | 5350/10712 [1:04:44<43:47,  2.04it/s]
 50%|████▉     | 5351/10712 [1:04:44<43:49,  2.04it/s]
 50%|████▉     | 5352/10712 [1:04:45<43:51,  2.04it/s]
 50%|████▉     | 5353/10712 [1:04:45<43:50,  2.04it/s]
 50%|████▉     | 5354/10712 [1:04:46<43:51,  2.04it/s]
 50%|████▉     | 5355/10712 [1:04:46<43:50,  2.04it/s]
 50%|█████     | 5356/10712 [1:04:47<43:46,  2.04it/s]
 50%|█████     | 5357/10712 [1:04:47<43:45,  2.04it/s]
 50%|█████     | 5358/10712 [1:04:48<43:47,  2.04it/s]
 50%|█████     | 5359/10712 [1:04:48<43:46,  2.04it/s]
 50%|█████     | 5360/10712 [1:04:49<43:44,  2.04it/s]
 50%|█████     | 5361/10712 [1:04:49<43:43,  2.04it/s]
 50%|█████     | 5362/10712 [1:04:50<43:44,  2.04it/s]
 50%|█████     | 5363/10712 [1:04:50<43:41,  2.04it/s]
 50%|█████     | 5364/10712 [1:04:51<43:41,  2.04it/s]
 50%|█████     | 5365/10712 [1:04:51<43:41,  2.04it/s]
 50%|█████     | 5366/10712 [1:04:52<43:41,  2.04it/s]
 50%|█████     | 5367/10712 [1:04:52<43:38,  2.04it/s]
 50%|█████     | 5368/10712 [1:04:53<43:40,  2.04it/s]
 50%|█████     | 5369/10712 [1:04:53<43:39,  2.04it/s]
 50%|█████     | 5370/10712 [1:04:54<43:40,  2.04it/s]
 50%|█████     | 5371/10712 [1:04:54<43:37,  2.04it/s]
 50%|█████     | 5372/10712 [1:04:55<43:39,  2.04it/s]
 50%|█████     | 5373/10712 [1:04:55<43:40,  2.04it/s]
 50%|█████     | 5374/10712 [1:04:56<43:41,  2.04it/s]
 50%|█████     | 5375/10712 [1:04:56<43:40,  2.04it/s]{'loss': 3.7473, 'grad_norm': 0.21501114964485168, 'learning_rate': 0.0005838449064399331, 'epoch': 0.5}
                                                      

 50%|█████     | 5375/10712 [1:04:56<43:40,  2.04it/s]
 50%|█████     | 5376/10712 [1:04:57<43:43,  2.03it/s]
 50%|█████     | 5377/10712 [1:04:57<43:42,  2.03it/s]
 50%|█████     | 5378/10712 [1:04:58<43:42,  2.03it/s]
 50%|█████     | 5379/10712 [1:04:58<43:40,  2.03it/s]
 50%|█████     | 5380/10712 [1:04:59<43:37,  2.04it/s]
 50%|█████     | 5381/10712 [1:04:59<43:39,  2.04it/s]
 50%|█████     | 5382/10712 [1:05:00<43:39,  2.03it/s]
 50%|█████     | 5383/10712 [1:05:00<43:36,  2.04it/s]
 50%|█████     | 5384/10712 [1:05:01<43:36,  2.04it/s]
 50%|█████     | 5385/10712 [1:05:01<43:35,  2.04it/s]
 50%|█████     | 5386/10712 [1:05:02<43:32,  2.04it/s]
 50%|█████     | 5387/10712 [1:05:02<43:30,  2.04it/s]
 50%|█████     | 5388/10712 [1:05:03<43:31,  2.04it/s]
 50%|█████     | 5389/10712 [1:05:03<43:30,  2.04it/s]
 50%|█████     | 5390/10712 [1:05:04<43:28,  2.04it/s]
 50%|█████     | 5391/10712 [1:05:04<43:26,  2.04it/s]
 50%|█████     | 5392/10712 [1:05:05<43:33,  2.04it/s]
 50%|█████     | 5393/10712 [1:05:05<43:29,  2.04it/s]
 50%|█████     | 5394/10712 [1:05:06<43:27,  2.04it/s]
 50%|█████     | 5395/10712 [1:05:06<43:31,  2.04it/s]
 50%|█████     | 5396/10712 [1:05:07<43:29,  2.04it/s]
 50%|█████     | 5397/10712 [1:05:07<43:24,  2.04it/s]
 50%|█████     | 5398/10712 [1:05:08<43:22,  2.04it/s]
 50%|█████     | 5399/10712 [1:05:08<43:23,  2.04it/s]
 50%|█████     | 5400/10712 [1:05:08<43:23,  2.04it/s]{'loss': 3.7579, 'grad_norm': 0.2210680991411209, 'learning_rate': 0.0005798262097717339, 'epoch': 0.5}
                                                      

 50%|█████     | 5400/10712 [1:05:08<43:23,  2.04it/s]
 50%|█████     | 5401/10712 [1:05:09<43:24,  2.04it/s]
 50%|█████     | 5402/10712 [1:05:09<43:24,  2.04it/s]
 50%|█████     | 5403/10712 [1:05:10<43:23,  2.04it/s]
 50%|█████     | 5404/10712 [1:05:10<43:23,  2.04it/s]
 50%|█████     | 5405/10712 [1:05:11<43:18,  2.04it/s]
 50%|█████     | 5406/10712 [1:05:11<43:19,  2.04it/s]
 50%|█████     | 5407/10712 [1:05:12<43:22,  2.04it/s]
 50%|█████     | 5408/10712 [1:05:12<43:20,  2.04it/s]
 50%|█████     | 5409/10712 [1:05:13<43:18,  2.04it/s]
 51%|█████     | 5410/10712 [1:05:13<43:16,  2.04it/s]
 51%|█████     | 5411/10712 [1:05:14<43:18,  2.04it/s]
 51%|█████     | 5412/10712 [1:05:14<43:17,  2.04it/s]
 51%|█████     | 5413/10712 [1:05:15<43:14,  2.04it/s]
 51%|█████     | 5414/10712 [1:05:15<43:16,  2.04it/s]
 51%|█████     | 5415/10712 [1:05:16<43:16,  2.04it/s]
 51%|█████     | 5416/10712 [1:05:16<43:14,  2.04it/s]
 51%|█████     | 5417/10712 [1:05:17<43:11,  2.04it/s]
 51%|█████     | 5418/10712 [1:05:17<43:11,  2.04it/s]
 51%|█████     | 5419/10712 [1:05:18<43:12,  2.04it/s]
 51%|█████     | 5420/10712 [1:05:18<43:11,  2.04it/s]
 51%|█████     | 5421/10712 [1:05:19<43:10,  2.04it/s]
 51%|█████     | 5422/10712 [1:05:19<43:12,  2.04it/s]
 51%|█████     | 5423/10712 [1:05:20<43:12,  2.04it/s]
 51%|█████     | 5424/10712 [1:05:20<43:10,  2.04it/s]
 51%|█████     | 5425/10712 [1:05:21<43:11,  2.04it/s]{'loss': 3.7549, 'grad_norm': 0.20999301970005035, 'learning_rate': 0.0005758022144099991, 'epoch': 0.51}

                                                      

 51%|█████     | 5425/10712 [1:05:21<43:11,  2.04it/s]
 51%|█████     | 5426/10712 [1:05:21<43:15,  2.04it/s]
 51%|█████     | 5427/10712 [1:05:22<43:10,  2.04it/s]
 51%|█████     | 5428/10712 [1:05:22<43:10,  2.04it/s]
 51%|█████     | 5429/10712 [1:05:23<43:13,  2.04it/s]
 51%|█████     | 5430/10712 [1:05:23<43:11,  2.04it/s]
 51%|█████     | 5431/10712 [1:05:24<43:08,  2.04it/s]
 51%|█████     | 5432/10712 [1:05:24<43:06,  2.04it/s]
 51%|█████     | 5433/10712 [1:05:25<43:07,  2.04it/s]
 51%|█████     | 5434/10712 [1:05:25<43:06,  2.04it/s]
 51%|█████     | 5435/10712 [1:05:26<43:03,  2.04it/s]
 51%|█████     | 5436/10712 [1:05:26<43:07,  2.04it/s]
 51%|█████     | 5437/10712 [1:05:27<43:05,  2.04it/s]
 51%|█████     | 5438/10712 [1:05:27<43:02,  2.04it/s]
 51%|█████     | 5439/10712 [1:05:28<42:59,  2.04it/s]
 51%|█████     | 5440/10712 [1:05:28<43:00,  2.04it/s]
 51%|█████     | 5441/10712 [1:05:29<43:04,  2.04it/s]
 51%|█████     | 5442/10712 [1:05:29<43:00,  2.04it/s]
 51%|█████     | 5443/10712 [1:05:30<42:56,  2.04it/s]
 51%|█████     | 5444/10712 [1:05:30<42:57,  2.04it/s]
 51%|█████     | 5445/10712 [1:05:31<42:59,  2.04it/s]
 51%|█████     | 5446/10712 [1:05:31<43:02,  2.04it/s]
 51%|█████     | 5447/10712 [1:05:32<43:00,  2.04it/s]
 51%|█████     | 5448/10712 [1:05:32<43:43,  2.01it/s]
 51%|█████     | 5449/10712 [1:05:33<43:31,  2.02it/s]
 51%|█████     | 5450/10712 [1:05:33<43:21,  2.02it/s]
                                                      
{'loss': 3.7581, 'grad_norm': 0.20953498780727386, 'learning_rate': 0.0005717731874589576, 'epoch': 0.51}

 51%|█████     | 5450/10712 [1:05:33<43:21,  2.02it/s]
 51%|█████     | 5451/10712 [1:05:34<43:17,  2.03it/s]
 51%|█████     | 5452/10712 [1:05:34<43:13,  2.03it/s]
 51%|█████     | 5453/10712 [1:05:34<43:08,  2.03it/s]
 51%|█████     | 5454/10712 [1:05:35<43:04,  2.03it/s]
 51%|█████     | 5455/10712 [1:05:35<42:58,  2.04it/s]
 51%|█████     | 5456/10712 [1:05:36<42:57,  2.04it/s]
 51%|█████     | 5457/10712 [1:05:36<43:00,  2.04it/s]
 51%|█████     | 5458/10712 [1:05:37<42:57,  2.04it/s]
 51%|█████     | 5459/10712 [1:05:37<42:52,  2.04it/s]
 51%|█████     | 5460/10712 [1:05:38<42:46,  2.05it/s]
 51%|█████     | 5461/10712 [1:05:38<42:47,  2.04it/s]
 51%|█████     | 5462/10712 [1:05:39<42:52,  2.04it/s]
 51%|█████     | 5463/10712 [1:05:39<42:51,  2.04it/s]
 51%|█████     | 5464/10712 [1:05:40<42:49,  2.04it/s]
 51%|█████     | 5465/10712 [1:05:40<42:53,  2.04it/s]
 51%|█████     | 5466/10712 [1:05:41<42:51,  2.04it/s]
 51%|█████     | 5467/10712 [1:05:41<42:47,  2.04it/s]
 51%|█████     | 5468/10712 [1:05:42<42:47,  2.04it/s]
 51%|█████     | 5469/10712 [1:05:42<42:49,  2.04it/s]
 51%|█████     | 5470/10712 [1:05:43<42:49,  2.04it/s]
 51%|█████     | 5471/10712 [1:05:43<42:47,  2.04it/s]
 51%|█████     | 5472/10712 [1:05:44<42:49,  2.04it/s]
 51%|█████     | 5473/10712 [1:05:44<42:48,  2.04it/s]
 51%|█████     | 5474/10712 [1:05:45<42:45,  2.04it/s]
 51%|█████     | 5475/10712 [1:05:45<42:43,  2.04it/s]
                                                      {'loss': 3.7521, 'grad_norm': 0.20608583092689514, 'learning_rate': 0.0005677393963568244, 'epoch': 0.51}


 51%|█████     | 5475/10712 [1:05:45<42:43,  2.04it/s]
 51%|█████     | 5476/10712 [1:05:46<42:48,  2.04it/s]
 51%|█████     | 5477/10712 [1:05:46<42:49,  2.04it/s]
 51%|█████     | 5478/10712 [1:05:47<42:47,  2.04it/s]
 51%|█████     | 5479/10712 [1:05:47<42:48,  2.04it/s]
 51%|█████     | 5480/10712 [1:05:48<42:46,  2.04it/s]
 51%|█████     | 5481/10712 [1:05:48<42:42,  2.04it/s]
 51%|█████     | 5482/10712 [1:05:49<42:40,  2.04it/s]
 51%|█████     | 5483/10712 [1:05:49<42:42,  2.04it/s]
 51%|█████     | 5484/10712 [1:05:50<42:39,  2.04it/s]
 51%|█████     | 5485/10712 [1:05:50<42:36,  2.04it/s]
 51%|█████     | 5486/10712 [1:05:51<42:39,  2.04it/s]
 51%|█████     | 5487/10712 [1:05:51<42:42,  2.04it/s]
 51%|█████     | 5488/10712 [1:05:52<42:39,  2.04it/s]
 51%|█████     | 5489/10712 [1:05:52<42:39,  2.04it/s]
 51%|█████▏    | 5490/10712 [1:05:53<42:40,  2.04it/s]
 51%|█████▏    | 5491/10712 [1:05:53<42:39,  2.04it/s]
 51%|█████▏    | 5492/10712 [1:05:54<42:37,  2.04it/s]
 51%|█████▏    | 5493/10712 [1:05:54<42:37,  2.04it/s]
 51%|█████▏    | 5494/10712 [1:05:55<42:39,  2.04it/s]
 51%|█████▏    | 5495/10712 [1:05:55<42:36,  2.04it/s]
 51%|█████▏    | 5496/10712 [1:05:56<42:35,  2.04it/s]
 51%|█████▏    | 5497/10712 [1:05:56<42:37,  2.04it/s]
 51%|█████▏    | 5498/10712 [1:05:57<42:35,  2.04it/s]
 51%|█████▏    | 5499/10712 [1:05:57<42:32,  2.04it/s]
 51%|█████▏    | 5500/10712 [1:05:58<42:32,  2.04it/s]{'loss': 3.7464, 'grad_norm': 0.19928264617919922, 'learning_rate': 0.0005637011088580488, 'epoch': 0.51}
                                                      

 51%|█████▏    | 5500/10712 [1:05:58<42:32,  2.04it/s]
 51%|█████▏    | 5501/10712 [1:05:58<42:35,  2.04it/s]
 51%|█████▏    | 5502/10712 [1:05:59<42:32,  2.04it/s]
 51%|█████▏    | 5503/10712 [1:05:59<42:30,  2.04it/s]
 51%|█████▏    | 5504/10712 [1:05:59<42:33,  2.04it/s]
 51%|█████▏    | 5505/10712 [1:06:00<42:31,  2.04it/s]
 51%|█████▏    | 5506/10712 [1:06:00<42:30,  2.04it/s]
 51%|█████▏    | 5507/10712 [1:06:01<42:28,  2.04it/s]
 51%|█████▏    | 5508/10712 [1:06:01<42:29,  2.04it/s]
 51%|█████▏    | 5509/10712 [1:06:02<42:29,  2.04it/s]
 51%|█████▏    | 5510/10712 [1:06:02<42:27,  2.04it/s]
 51%|█████▏    | 5511/10712 [1:06:03<42:29,  2.04it/s]
 51%|█████▏    | 5512/10712 [1:06:03<42:30,  2.04it/s]
 51%|█████▏    | 5513/10712 [1:06:04<42:29,  2.04it/s]
 51%|█████▏    | 5514/10712 [1:06:04<42:27,  2.04it/s]
 51%|█████▏    | 5515/10712 [1:06:05<42:27,  2.04it/s]
 51%|█████▏    | 5516/10712 [1:06:05<42:29,  2.04it/s]
 52%|█████▏    | 5517/10712 [1:06:06<42:25,  2.04it/s]
 52%|█████▏    | 5518/10712 [1:06:06<42:21,  2.04it/s]
 52%|█████▏    | 5519/10712 [1:06:07<42:23,  2.04it/s]
 52%|█████▏    | 5520/10712 [1:06:07<42:23,  2.04it/s]
 52%|█████▏    | 5521/10712 [1:06:08<42:24,  2.04it/s]
 52%|█████▏    | 5522/10712 [1:06:08<42:23,  2.04it/s]
 52%|█████▏    | 5523/10712 [1:06:09<42:24,  2.04it/s]
 52%|█████▏    | 5524/10712 [1:06:09<42:22,  2.04it/s]
 52%|█████▏    | 5525/10712 [1:06:10<42:20,  2.04it/s]{'loss': 3.7455, 'grad_norm': 0.2146102786064148, 'learning_rate': 0.0005596585930155414, 'epoch': 0.52}

                                                      

 52%|█████▏    | 5525/10712 [1:06:10<42:20,  2.04it/s]
 52%|█████▏    | 5526/10712 [1:06:10<42:23,  2.04it/s]
 52%|█████▏    | 5527/10712 [1:06:11<42:23,  2.04it/s]
 52%|█████▏    | 5528/10712 [1:06:11<42:24,  2.04it/s]
 52%|█████▏    | 5529/10712 [1:06:12<42:21,  2.04it/s]
 52%|█████▏    | 5530/10712 [1:06:12<42:22,  2.04it/s]
 52%|█████▏    | 5531/10712 [1:06:13<42:22,  2.04it/s]
 52%|█████▏    | 5532/10712 [1:06:13<42:19,  2.04it/s]
 52%|█████▏    | 5533/10712 [1:06:14<42:16,  2.04it/s]
 52%|█████▏    | 5534/10712 [1:06:14<42:19,  2.04it/s]
 52%|█████▏    | 5535/10712 [1:06:15<42:16,  2.04it/s]
 52%|█████▏    | 5536/10712 [1:06:15<42:14,  2.04it/s]
 52%|█████▏    | 5537/10712 [1:06:16<42:14,  2.04it/s]
 52%|█████▏    | 5538/10712 [1:06:16<42:17,  2.04it/s]
 52%|█████▏    | 5539/10712 [1:06:17<42:15,  2.04it/s]
 52%|█████▏    | 5540/10712 [1:06:17<42:13,  2.04it/s]
 52%|█████▏    | 5541/10712 [1:06:18<42:13,  2.04it/s]
 52%|█████▏    | 5542/10712 [1:06:18<42:13,  2.04it/s]
 52%|█████▏    | 5543/10712 [1:06:19<42:12,  2.04it/s]
 52%|█████▏    | 5544/10712 [1:06:19<42:12,  2.04it/s]
 52%|█████▏    | 5545/10712 [1:06:20<42:09,  2.04it/s]
 52%|█████▏    | 5546/10712 [1:06:20<42:13,  2.04it/s]
 52%|█████▏    | 5547/10712 [1:06:21<42:12,  2.04it/s]
 52%|█████▏    | 5548/10712 [1:06:21<42:12,  2.04it/s]
 52%|█████▏    | 5549/10712 [1:06:22<42:14,  2.04it/s]
 52%|█████▏    | 5550/10712 [1:06:22<42:13,  2.04it/s]{'loss': 3.7522, 'grad_norm': 0.21689048409461975, 'learning_rate': 0.0005556121171628811, 'epoch': 0.52}

                                                      

 52%|█████▏    | 5550/10712 [1:06:22<42:13,  2.04it/s]
 52%|█████▏    | 5551/10712 [1:06:23<42:16,  2.03it/s]
 52%|█████▏    | 5552/10712 [1:06:23<42:15,  2.04it/s]
 52%|█████▏    | 5553/10712 [1:06:24<42:12,  2.04it/s]
 52%|█████▏    | 5554/10712 [1:06:24<42:09,  2.04it/s]
 52%|█████▏    | 5555/10712 [1:06:24<42:06,  2.04it/s]
 52%|█████▏    | 5556/10712 [1:06:25<42:06,  2.04it/s]
 52%|█████▏    | 5557/10712 [1:06:25<42:05,  2.04it/s]
 52%|█████▏    | 5558/10712 [1:06:26<42:04,  2.04it/s]
 52%|█████▏    | 5559/10712 [1:06:26<42:06,  2.04it/s]
 52%|█████▏    | 5560/10712 [1:06:27<42:08,  2.04it/s]
 52%|█████▏    | 5561/10712 [1:06:27<42:07,  2.04it/s]
 52%|█████▏    | 5562/10712 [1:06:28<42:05,  2.04it/s]
 52%|█████▏    | 5563/10712 [1:06:28<42:06,  2.04it/s]
 52%|█████▏    | 5564/10712 [1:06:29<42:06,  2.04it/s]
 52%|█████▏    | 5565/10712 [1:06:29<42:05,  2.04it/s]
 52%|█████▏    | 5566/10712 [1:06:30<42:08,  2.04it/s]
 52%|█████▏    | 5567/10712 [1:06:30<42:04,  2.04it/s]
 52%|█████▏    | 5568/10712 [1:06:31<42:04,  2.04it/s]
 52%|█████▏    | 5569/10712 [1:06:31<42:07,  2.04it/s]
 52%|█████▏    | 5570/10712 [1:06:32<42:04,  2.04it/s]
 52%|█████▏    | 5571/10712 [1:06:32<42:03,  2.04it/s]
 52%|█████▏    | 5572/10712 [1:06:33<42:05,  2.04it/s]
 52%|█████▏    | 5573/10712 [1:06:33<42:00,  2.04it/s]
 52%|█████▏    | 5574/10712 [1:06:34<41:59,  2.04it/s]
 52%|█████▏    | 5575/10712 [1:06:34<42:00,  2.04it/s]{'loss': 3.7439, 'grad_norm': 0.20269644260406494, 'learning_rate': 0.000551561949896504, 'epoch': 0.52}

                                                      

 52%|█████▏    | 5575/10712 [1:06:34<42:00,  2.04it/s]
 52%|█████▏    | 5576/10712 [1:06:35<42:04,  2.03it/s]
 52%|█████▏    | 5577/10712 [1:06:35<42:00,  2.04it/s]
 52%|█████▏    | 5578/10712 [1:06:36<42:01,  2.04it/s]
 52%|█████▏    | 5579/10712 [1:06:36<42:02,  2.03it/s]
 52%|█████▏    | 5580/10712 [1:06:37<42:02,  2.03it/s]
 52%|█████▏    | 5581/10712 [1:06:37<41:58,  2.04it/s]
 52%|█████▏    | 5582/10712 [1:06:38<41:58,  2.04it/s]
 52%|█████▏    | 5583/10712 [1:06:38<42:01,  2.03it/s]
 52%|█████▏    | 5584/10712 [1:06:39<41:58,  2.04it/s]
 52%|█████▏    | 5585/10712 [1:06:39<41:55,  2.04it/s]
 52%|█████▏    | 5586/10712 [1:06:40<41:59,  2.03it/s]
 52%|█████▏    | 5587/10712 [1:06:40<41:55,  2.04it/s]
 52%|█████▏    | 5588/10712 [1:06:41<41:52,  2.04it/s]
 52%|█████▏    | 5589/10712 [1:06:41<41:51,  2.04it/s]
 52%|█████▏    | 5590/10712 [1:06:42<41:51,  2.04it/s]
 52%|█████▏    | 5591/10712 [1:06:42<41:53,  2.04it/s]
 52%|█████▏    | 5592/10712 [1:06:43<41:50,  2.04it/s]
 52%|█████▏    | 5593/10712 [1:06:43<41:50,  2.04it/s]
 52%|█████▏    | 5594/10712 [1:06:44<41:50,  2.04it/s]
 52%|█████▏    | 5595/10712 [1:06:44<41:50,  2.04it/s]
 52%|█████▏    | 5596/10712 [1:06:45<41:49,  2.04it/s]
 52%|█████▏    | 5597/10712 [1:06:45<41:48,  2.04it/s]
 52%|█████▏    | 5598/10712 [1:06:46<41:50,  2.04it/s]
 52%|█████▏    | 5599/10712 [1:06:46<41:46,  2.04it/s]
 52%|█████▏    | 5600/10712 [1:06:47<41:44,  2.04it/s]
                                                      {'loss': 3.7464, 'grad_norm': 0.2129647433757782, 'learning_rate': 0.0005475083600578742, 'epoch': 0.52}

 52%|█████▏    | 5600/10712 [1:06:47<41:44,  2.04it/s]
 52%|█████▏    | 5601/10712 [1:06:47<41:46,  2.04it/s]
 52%|█████▏    | 5602/10712 [1:06:48<41:48,  2.04it/s]
 52%|█████▏    | 5603/10712 [1:06:48<41:48,  2.04it/s]
 52%|█████▏    | 5604/10712 [1:06:49<41:48,  2.04it/s]
 52%|█████▏    | 5605/10712 [1:06:49<41:48,  2.04it/s]
 52%|█████▏    | 5606/10712 [1:06:50<41:44,  2.04it/s]
 52%|█████▏    | 5607/10712 [1:06:50<41:44,  2.04it/s]
 52%|█████▏    | 5608/10712 [1:06:50<41:45,  2.04it/s]
 52%|█████▏    | 5609/10712 [1:06:51<41:42,  2.04it/s]
 52%|█████▏    | 5610/10712 [1:06:51<41:39,  2.04it/s]
 52%|█████▏    | 5611/10712 [1:06:52<41:40,  2.04it/s]
 52%|█████▏    | 5612/10712 [1:06:52<41:40,  2.04it/s]
 52%|█████▏    | 5613/10712 [1:06:53<41:40,  2.04it/s]
 52%|█████▏    | 5614/10712 [1:06:53<41:37,  2.04it/s]
 52%|█████▏    | 5615/10712 [1:06:54<41:39,  2.04it/s]
 52%|█████▏    | 5616/10712 [1:06:54<41:39,  2.04it/s]
 52%|█████▏    | 5617/10712 [1:06:55<41:35,  2.04it/s]
 52%|█████▏    | 5618/10712 [1:06:55<41:34,  2.04it/s]
 52%|█████▏    | 5619/10712 [1:06:56<41:38,  2.04it/s]
 52%|█████▏    | 5620/10712 [1:06:56<41:37,  2.04it/s]
 52%|█████▏    | 5621/10712 [1:06:57<41:35,  2.04it/s]
 52%|█████▏    | 5622/10712 [1:06:57<41:36,  2.04it/s]
 52%|█████▏    | 5623/10712 [1:06:58<41:34,  2.04it/s]
 53%|█████▎    | 5624/10712 [1:06:58<41:31,  2.04it/s]
 53%|█████▎    | 5625/10712 [1:06:59<41:32,  2.04it/s]{'loss': 3.7341, 'grad_norm': 0.20345637202262878, 'learning_rate': 0.0005434516167156389, 'epoch': 0.53}

                                                      

 53%|█████▎    | 5625/10712 [1:06:59<41:32,  2.04it/s]
 53%|█████▎    | 5626/10712 [1:06:59<41:36,  2.04it/s]
 53%|█████▎    | 5627/10712 [1:07:00<41:33,  2.04it/s]
 53%|█████▎    | 5628/10712 [1:07:00<41:31,  2.04it/s]
 53%|█████▎    | 5629/10712 [1:07:01<41:32,  2.04it/s]
 53%|█████▎    | 5630/10712 [1:07:01<41:33,  2.04it/s]
 53%|█████▎    | 5631/10712 [1:07:02<41:30,  2.04it/s]
 53%|█████▎    | 5632/10712 [1:07:02<41:31,  2.04it/s]
 53%|█████▎    | 5633/10712 [1:07:03<41:32,  2.04it/s]
 53%|█████▎    | 5634/10712 [1:07:03<41:31,  2.04it/s]
 53%|█████▎    | 5635/10712 [1:07:04<41:29,  2.04it/s]
 53%|█████▎    | 5636/10712 [1:07:04<41:31,  2.04it/s]
 53%|█████▎    | 5637/10712 [1:07:05<41:28,  2.04it/s]
 53%|█████▎    | 5638/10712 [1:07:05<41:24,  2.04it/s]
 53%|█████▎    | 5639/10712 [1:07:06<41:26,  2.04it/s]
 53%|█████▎    | 5640/10712 [1:07:06<41:27,  2.04it/s]
 53%|█████▎    | 5641/10712 [1:07:07<41:24,  2.04it/s]
 53%|█████▎    | 5642/10712 [1:07:07<41:26,  2.04it/s]
 53%|█████▎    | 5643/10712 [1:07:08<41:27,  2.04it/s]
 53%|█████▎    | 5644/10712 [1:07:08<41:24,  2.04it/s]
 53%|█████▎    | 5645/10712 [1:07:09<41:22,  2.04it/s]
 53%|█████▎    | 5646/10712 [1:07:09<41:24,  2.04it/s]
 53%|█████▎    | 5647/10712 [1:07:10<41:22,  2.04it/s]
 53%|█████▎    | 5648/10712 [1:07:10<41:22,  2.04it/s]
 53%|█████▎    | 5649/10712 [1:07:11<41:21,  2.04it/s]
 53%|█████▎    | 5650/10712 [1:07:11<41:23,  2.04it/s]{'loss': 3.7409, 'grad_norm': 0.1993490755558014, 'learning_rate': 0.0005393919891477681, 'epoch': 0.53}
                                                      

 53%|█████▎    | 5650/10712 [1:07:11<41:23,  2.04it/s]
 53%|█████▎    | 5651/10712 [1:07:12<41:27,  2.03it/s]
 53%|█████▎    | 5652/10712 [1:07:12<41:24,  2.04it/s]
 53%|█████▎    | 5653/10712 [1:07:13<41:25,  2.04it/s]
 53%|█████▎    | 5654/10712 [1:07:13<41:22,  2.04it/s]
 53%|█████▎    | 5655/10712 [1:07:14<41:20,  2.04it/s]
 53%|█████▎    | 5656/10712 [1:07:14<41:20,  2.04it/s]
 53%|█████▎    | 5657/10712 [1:07:15<41:21,  2.04it/s]
 53%|█████▎    | 5658/10712 [1:07:15<41:16,  2.04it/s]
 53%|█████▎    | 5659/10712 [1:07:16<41:16,  2.04it/s]
 53%|█████▎    | 5660/10712 [1:07:16<41:17,  2.04it/s]
 53%|█████▎    | 5661/10712 [1:07:16<41:17,  2.04it/s]
 53%|█████▎    | 5662/10712 [1:07:17<41:14,  2.04it/s]
 53%|█████▎    | 5663/10712 [1:07:17<41:17,  2.04it/s]
 53%|█████▎    | 5664/10712 [1:07:18<41:17,  2.04it/s]
 53%|█████▎    | 5665/10712 [1:07:18<41:13,  2.04it/s]
 53%|█████▎    | 5666/10712 [1:07:19<41:09,  2.04it/s]
 53%|█████▎    | 5667/10712 [1:07:19<41:12,  2.04it/s]
 53%|█████▎    | 5668/10712 [1:07:20<41:13,  2.04it/s]
 53%|█████▎    | 5669/10712 [1:07:20<41:11,  2.04it/s]
 53%|█████▎    | 5670/10712 [1:07:21<41:10,  2.04it/s]
 53%|█████▎    | 5671/10712 [1:07:21<41:12,  2.04it/s]
 53%|█████▎    | 5672/10712 [1:07:22<41:12,  2.04it/s]
 53%|█████▎    | 5673/10712 [1:07:22<41:09,  2.04it/s]
 53%|█████▎    | 5674/10712 [1:07:23<41:05,  2.04it/s]
 53%|█████▎    | 5675/10712 [1:07:23<41:09,  2.04it/s]{'loss': 3.7396, 'grad_norm': 0.2331283539533615, 'learning_rate': 0.0005353297468236806, 'epoch': 0.53}
                                                      

 53%|█████▎    | 5675/10712 [1:07:23<41:09,  2.04it/s]
 53%|█████▎    | 5676/10712 [1:07:24<41:11,  2.04it/s]
 53%|█████▎    | 5677/10712 [1:07:24<41:07,  2.04it/s]
 53%|█████▎    | 5678/10712 [1:07:25<41:06,  2.04it/s]
 53%|█████▎    | 5679/10712 [1:07:25<41:08,  2.04it/s]
 53%|█████▎    | 5680/10712 [1:07:26<41:10,  2.04it/s]
 53%|█████▎    | 5681/10712 [1:07:26<41:06,  2.04it/s]
 53%|█████▎    | 5682/10712 [1:07:27<41:06,  2.04it/s]
 53%|█████▎    | 5683/10712 [1:07:27<41:07,  2.04it/s]
 53%|█████▎    | 5684/10712 [1:07:28<41:04,  2.04it/s]
 53%|█████▎    | 5685/10712 [1:07:28<41:03,  2.04it/s]
 53%|█████▎    | 5686/10712 [1:07:29<41:01,  2.04it/s]
 53%|█████▎    | 5687/10712 [1:07:29<41:03,  2.04it/s]
 53%|█████▎    | 5688/10712 [1:07:30<41:02,  2.04it/s]
 53%|█████▎    | 5689/10712 [1:07:30<41:02,  2.04it/s]
 53%|█████▎    | 5690/10712 [1:07:31<41:04,  2.04it/s]
 53%|█████▎    | 5691/10712 [1:07:31<41:02,  2.04it/s]
 53%|█████▎    | 5692/10712 [1:07:32<41:00,  2.04it/s]
 53%|█████▎    | 5693/10712 [1:07:32<41:01,  2.04it/s]
 53%|█████▎    | 5694/10712 [1:07:33<40:59,  2.04it/s]
 53%|█████▎    | 5695/10712 [1:07:33<40:59,  2.04it/s]
 53%|█████▎    | 5696/10712 [1:07:34<40:59,  2.04it/s]
 53%|█████▎    | 5697/10712 [1:07:34<41:01,  2.04it/s]
 53%|█████▎    | 5698/10712 [1:07:35<40:58,  2.04it/s]
 53%|█████▎    | 5699/10712 [1:07:35<40:57,  2.04it/s]
 53%|█████▎    | 5700/10712 [1:07:36<41:02,  2.04it/s]{'loss': 3.7384, 'grad_norm': 0.22199752926826477, 'learning_rate': 0.0005312651593863571, 'epoch': 0.53}
                                                      

 53%|█████▎    | 5700/10712 [1:07:36<41:02,  2.04it/s]
 53%|█████▎    | 5701/10712 [1:07:36<40:59,  2.04it/s]
 53%|█████▎    | 5702/10712 [1:07:37<40:55,  2.04it/s]
 53%|█████▎    | 5703/10712 [1:07:37<40:57,  2.04it/s]
 53%|█████▎    | 5704/10712 [1:07:38<40:54,  2.04it/s]
 53%|█████▎    | 5705/10712 [1:07:38<40:56,  2.04it/s]
 53%|█████▎    | 5706/10712 [1:07:39<40:55,  2.04it/s]
 53%|█████▎    | 5707/10712 [1:07:39<40:55,  2.04it/s]
 53%|█████▎    | 5708/10712 [1:07:40<40:53,  2.04it/s]
 53%|█████▎    | 5709/10712 [1:07:40<40:50,  2.04it/s]
 53%|█████▎    | 5710/10712 [1:07:41<40:50,  2.04it/s]
 53%|█████▎    | 5711/10712 [1:07:41<40:50,  2.04it/s]
 53%|█████▎    | 5712/10712 [1:07:41<40:50,  2.04it/s]
 53%|█████▎    | 5713/10712 [1:07:42<40:48,  2.04it/s]
 53%|█████▎    | 5714/10712 [1:07:42<40:51,  2.04it/s]
 53%|█████▎    | 5715/10712 [1:07:43<40:53,  2.04it/s]
 53%|█████▎    | 5716/10712 [1:07:43<40:50,  2.04it/s]
 53%|█████▎    | 5717/10712 [1:07:44<40:50,  2.04it/s]
 53%|█████▎    | 5718/10712 [1:07:44<40:50,  2.04it/s]
 53%|█████▎    | 5719/10712 [1:07:45<40:48,  2.04it/s]
 53%|█████▎    | 5720/10712 [1:07:45<40:47,  2.04it/s]
 53%|█████▎    | 5721/10712 [1:07:46<40:45,  2.04it/s]
 53%|█████▎    | 5722/10712 [1:07:46<40:44,  2.04it/s]
 53%|█████▎    | 5723/10712 [1:07:47<40:47,  2.04it/s]
 53%|█████▎    | 5724/10712 [1:07:47<40:48,  2.04it/s]
 53%|█████▎    | 5725/10712 [1:07:48<40:45,  2.04it/s]{'loss': 3.738, 'grad_norm': 0.214952290058136, 'learning_rate': 0.0005271984966344421, 'epoch': 0.53}
                                                      

 53%|█████▎    | 5725/10712 [1:07:48<40:45,  2.04it/s]
 53%|█████▎    | 5726/10712 [1:07:48<40:45,  2.04it/s]
 53%|█████▎    | 5727/10712 [1:07:49<40:44,  2.04it/s]
 53%|█████▎    | 5728/10712 [1:07:49<40:44,  2.04it/s]
 53%|█████▎    | 5729/10712 [1:07:50<40:42,  2.04it/s]
 53%|█████▎    | 5730/10712 [1:07:50<40:41,  2.04it/s]
 54%|█████▎    | 5731/10712 [1:07:51<40:40,  2.04it/s]
 54%|█████▎    | 5732/10712 [1:07:51<40:40,  2.04it/s]
 54%|█████▎    | 5733/10712 [1:07:52<40:41,  2.04it/s]
 54%|█████▎    | 5734/10712 [1:07:52<40:41,  2.04it/s]
 54%|█████▎    | 5735/10712 [1:07:53<40:39,  2.04it/s]
 54%|█████▎    | 5736/10712 [1:07:53<40:39,  2.04it/s]
 54%|█████▎    | 5737/10712 [1:07:54<40:39,  2.04it/s]
 54%|█████▎    | 5738/10712 [1:07:54<40:39,  2.04it/s]
 54%|█████▎    | 5739/10712 [1:07:55<40:39,  2.04it/s]
 54%|█████▎    | 5740/10712 [1:07:55<40:41,  2.04it/s]
 54%|█████▎    | 5741/10712 [1:07:56<40:38,  2.04it/s]
 54%|█████▎    | 5742/10712 [1:07:56<40:38,  2.04it/s]
 54%|█████▎    | 5743/10712 [1:07:57<40:41,  2.04it/s]
 54%|█████▎    | 5744/10712 [1:07:57<40:38,  2.04it/s]
 54%|█████▎    | 5745/10712 [1:07:58<40:34,  2.04it/s]
 54%|█████▎    | 5746/10712 [1:07:58<40:33,  2.04it/s]
 54%|█████▎    | 5747/10712 [1:07:59<40:35,  2.04it/s]
 54%|█████▎    | 5748/10712 [1:07:59<40:34,  2.04it/s]
 54%|█████▎    | 5749/10712 [1:08:00<40:36,  2.04it/s]
 54%|█████▎    | 5750/10712 [1:08:00<40:35,  2.04it/s]{'loss': 3.7388, 'grad_norm': 0.21511632204055786, 'learning_rate': 0.0005231300285043346, 'epoch': 0.54}

                                                      

 54%|█████▎    | 5750/10712 [1:08:00<40:35,  2.04it/s]
 54%|█████▎    | 5751/10712 [1:08:01<40:34,  2.04it/s]
 54%|█████▎    | 5752/10712 [1:08:01<40:33,  2.04it/s]
 54%|█████▎    | 5753/10712 [1:08:02<40:34,  2.04it/s]
 54%|█████▎    | 5754/10712 [1:08:02<40:31,  2.04it/s]
 54%|█████▎    | 5755/10712 [1:08:03<40:28,  2.04it/s]
 54%|█████▎    | 5756/10712 [1:08:03<40:28,  2.04it/s]
 54%|█████▎    | 5757/10712 [1:08:04<40:29,  2.04it/s]
 54%|█████▍    | 5758/10712 [1:08:04<40:27,  2.04it/s]
 54%|█████▍    | 5759/10712 [1:08:05<40:28,  2.04it/s]
 54%|█████▍    | 5760/10712 [1:08:05<40:28,  2.04it/s]
 54%|█████▍    | 5761/10712 [1:08:06<40:27,  2.04it/s]
 54%|█████▍    | 5762/10712 [1:08:06<40:24,  2.04it/s]
 54%|█████▍    | 5763/10712 [1:08:07<40:28,  2.04it/s]
 54%|█████▍    | 5764/10712 [1:08:07<40:26,  2.04it/s]
 54%|█████▍    | 5765/10712 [1:08:07<40:23,  2.04it/s]
 54%|█████▍    | 5766/10712 [1:08:08<40:23,  2.04it/s]
 54%|█████▍    | 5767/10712 [1:08:08<40:24,  2.04it/s]
 54%|█████▍    | 5768/10712 [1:08:09<40:23,  2.04it/s]
 54%|█████▍    | 5769/10712 [1:08:09<40:19,  2.04it/s]
 54%|█████▍    | 5770/10712 [1:08:10<40:19,  2.04it/s]
 54%|█████▍    | 5771/10712 [1:08:10<40:21,  2.04it/s]
 54%|█████▍    | 5772/10712 [1:08:11<40:19,  2.04it/s]
 54%|█████▍    | 5773/10712 [1:08:11<40:16,  2.04it/s]
 54%|█████▍    | 5774/10712 [1:08:12<40:19,  2.04it/s]
 54%|█████▍    | 5775/10712 [1:08:12<40:19,  2.04it/s]{'loss': 3.7357, 'grad_norm': 0.20462240278720856, 'learning_rate': 0.0005190600250522716, 'epoch': 0.54}
                                                      

 54%|█████▍    | 5775/10712 [1:08:12<40:19,  2.04it/s]
 54%|█████▍    | 5776/10712 [1:08:13<40:19,  2.04it/s]
 54%|█████▍    | 5777/10712 [1:08:13<40:18,  2.04it/s]
 54%|█████▍    | 5778/10712 [1:08:14<40:19,  2.04it/s]
 54%|█████▍    | 5779/10712 [1:08:14<40:18,  2.04it/s]
 54%|█████▍    | 5780/10712 [1:08:15<40:15,  2.04it/s]
 54%|█████▍    | 5781/10712 [1:08:15<40:17,  2.04it/s]
 54%|█████▍    | 5782/10712 [1:08:16<40:16,  2.04it/s]
 54%|█████▍    | 5783/10712 [1:08:16<40:14,  2.04it/s]
 54%|█████▍    | 5784/10712 [1:08:17<40:14,  2.04it/s]
 54%|█████▍    | 5785/10712 [1:08:17<40:14,  2.04it/s]
 54%|█████▍    | 5786/10712 [1:08:18<40:12,  2.04it/s]
 54%|█████▍    | 5787/10712 [1:08:18<40:11,  2.04it/s]
 54%|█████▍    | 5788/10712 [1:08:19<40:12,  2.04it/s]
 54%|█████▍    | 5789/10712 [1:08:19<40:14,  2.04it/s]
 54%|█████▍    | 5790/10712 [1:08:20<40:11,  2.04it/s]
 54%|█████▍    | 5791/10712 [1:08:20<40:10,  2.04it/s]
 54%|█████▍    | 5792/10712 [1:08:21<40:09,  2.04it/s]
 54%|█████▍    | 5793/10712 [1:08:21<40:09,  2.04it/s]
 54%|█████▍    | 5794/10712 [1:08:22<40:08,  2.04it/s]
 54%|█████▍    | 5795/10712 [1:08:22<40:08,  2.04it/s]
 54%|█████▍    | 5796/10712 [1:08:23<40:09,  2.04it/s]
 54%|█████▍    | 5797/10712 [1:08:23<40:09,  2.04it/s]
 54%|█████▍    | 5798/10712 [1:08:24<40:31,  2.02it/s]
 54%|█████▍    | 5799/10712 [1:08:24<40:23,  2.03it/s]
 54%|█████▍    | 5800/10712 [1:08:25<40:17,  2.03it/s]{'loss': 3.7319, 'grad_norm': 0.21889856457710266, 'learning_rate': 0.0005149887564364004, 'epoch': 0.54}

                                                      

 54%|█████▍    | 5800/10712 [1:08:25<40:17,  2.03it/s]
 54%|█████▍    | 5801/10712 [1:08:25<40:18,  2.03it/s]
 54%|█████▍    | 5802/10712 [1:08:26<40:13,  2.03it/s]
 54%|█████▍    | 5803/10712 [1:08:26<40:09,  2.04it/s]
 54%|█████▍    | 5804/10712 [1:08:27<40:09,  2.04it/s]
 54%|█████▍    | 5805/10712 [1:08:27<40:10,  2.04it/s]
 54%|█████▍    | 5806/10712 [1:08:28<40:06,  2.04it/s]
 54%|█████▍    | 5807/10712 [1:08:28<40:04,  2.04it/s]
 54%|█████▍    | 5808/10712 [1:08:29<40:05,  2.04it/s]
 54%|█████▍    | 5809/10712 [1:08:29<40:05,  2.04it/s]
 54%|█████▍    | 5810/10712 [1:08:30<40:02,  2.04it/s]
 54%|█████▍    | 5811/10712 [1:08:30<39:59,  2.04it/s]
 54%|█████▍    | 5812/10712 [1:08:31<40:01,  2.04it/s]
 54%|█████▍    | 5813/10712 [1:08:31<40:02,  2.04it/s]
 54%|█████▍    | 5814/10712 [1:08:32<40:01,  2.04it/s]
 54%|█████▍    | 5815/10712 [1:08:32<40:02,  2.04it/s]
 54%|█████▍    | 5816/10712 [1:08:32<40:00,  2.04it/s]
 54%|█████▍    | 5817/10712 [1:08:33<40:02,  2.04it/s]
 54%|█████▍    | 5818/10712 [1:08:33<40:02,  2.04it/s]
 54%|█████▍    | 5819/10712 [1:08:34<40:00,  2.04it/s]
 54%|█████▍    | 5820/10712 [1:08:34<40:01,  2.04it/s]
 54%|█████▍    | 5821/10712 [1:08:35<39:59,  2.04it/s]
 54%|█████▍    | 5822/10712 [1:08:35<39:59,  2.04it/s]
 54%|█████▍    | 5823/10712 [1:08:36<39:59,  2.04it/s]
 54%|█████▍    | 5824/10712 [1:08:36<39:56,  2.04it/s]
 54%|█████▍    | 5825/10712 [1:08:37<39:55,  2.04it/s]
                                                      {'loss': 3.7214, 'grad_norm': 0.2180820256471634, 'learning_rate': 0.0005109164928988476, 'epoch': 0.54}


 54%|█████▍    | 5825/10712 [1:08:37<39:55,  2.04it/s]
 54%|█████▍    | 5826/10712 [1:08:37<39:58,  2.04it/s]
 54%|█████▍    | 5827/10712 [1:08:38<39:57,  2.04it/s]
 54%|█████▍    | 5828/10712 [1:08:38<39:52,  2.04it/s]
 54%|█████▍    | 5829/10712 [1:08:39<39:52,  2.04it/s]
 54%|█████▍    | 5830/10712 [1:08:39<39:56,  2.04it/s]
 54%|█████▍    | 5831/10712 [1:08:40<39:54,  2.04it/s]
 54%|█████▍    | 5832/10712 [1:08:40<39:51,  2.04it/s]
 54%|█████▍    | 5833/10712 [1:08:41<39:49,  2.04it/s]
 54%|█████▍    | 5834/10712 [1:08:41<39:51,  2.04it/s]
 54%|█████▍    | 5835/10712 [1:08:42<39:49,  2.04it/s]
 54%|█████▍    | 5836/10712 [1:08:42<39:47,  2.04it/s]
 54%|█████▍    | 5837/10712 [1:08:43<39:49,  2.04it/s]
 54%|█████▍    | 5838/10712 [1:08:43<39:48,  2.04it/s]
 55%|█████▍    | 5839/10712 [1:08:44<39:48,  2.04it/s]
 55%|█████▍    | 5840/10712 [1:08:44<39:48,  2.04it/s]
 55%|█████▍    | 5841/10712 [1:08:45<39:49,  2.04it/s]
 55%|█████▍    | 5842/10712 [1:08:45<39:48,  2.04it/s]
 55%|█████▍    | 5843/10712 [1:08:46<39:44,  2.04it/s]
 55%|█████▍    | 5844/10712 [1:08:46<39:43,  2.04it/s]
 55%|█████▍    | 5845/10712 [1:08:47<39:43,  2.04it/s]
 55%|█████▍    | 5846/10712 [1:08:47<39:43,  2.04it/s]
 55%|█████▍    | 5847/10712 [1:08:48<39:42,  2.04it/s]
 55%|█████▍    | 5848/10712 [1:08:48<39:45,  2.04it/s]
 55%|█████▍    | 5849/10712 [1:08:49<39:46,  2.04it/s]
 55%|█████▍    | 5850/10712 [1:08:49<39:42,  2.04it/s]{'loss': 3.7369, 'grad_norm': 0.2043040245771408, 'learning_rate': 0.0005068435047477813, 'epoch': 0.55}

                                                      

 55%|█████▍    | 5850/10712 [1:08:49<39:42,  2.04it/s]
 55%|█████▍    | 5851/10712 [1:08:50<39:44,  2.04it/s]
 55%|█████▍    | 5852/10712 [1:08:50<39:44,  2.04it/s]
 55%|█████▍    | 5853/10712 [1:08:51<39:42,  2.04it/s]
 55%|█████▍    | 5854/10712 [1:08:51<39:42,  2.04it/s]
 55%|█████▍    | 5855/10712 [1:08:52<39:39,  2.04it/s]
 55%|█████▍    | 5856/10712 [1:08:52<39:38,  2.04it/s]
 55%|█████▍    | 5857/10712 [1:08:53<39:40,  2.04it/s]
 55%|█████▍    | 5858/10712 [1:08:53<39:38,  2.04it/s]
 55%|█████▍    | 5859/10712 [1:08:54<39:36,  2.04it/s]
 55%|█████▍    | 5860/10712 [1:08:54<39:35,  2.04it/s]
 55%|█████▍    | 5861/10712 [1:08:55<39:35,  2.04it/s]
 55%|█████▍    | 5862/10712 [1:08:55<39:34,  2.04it/s]
 55%|█████▍    | 5863/10712 [1:08:56<39:35,  2.04it/s]
 55%|█████▍    | 5864/10712 [1:08:56<39:36,  2.04it/s]
 55%|█████▍    | 5865/10712 [1:08:57<39:33,  2.04it/s]
 55%|█████▍    | 5866/10712 [1:08:57<39:33,  2.04it/s]
 55%|█████▍    | 5867/10712 [1:08:57<39:35,  2.04it/s]
 55%|█████▍    | 5868/10712 [1:08:58<39:35,  2.04it/s]
 55%|█████▍    | 5869/10712 [1:08:58<39:33,  2.04it/s]
 55%|█████▍    | 5870/10712 [1:08:59<39:32,  2.04it/s]
 55%|█████▍    | 5871/10712 [1:08:59<39:34,  2.04it/s]
 55%|█████▍    | 5872/10712 [1:09:00<39:34,  2.04it/s]
 55%|█████▍    | 5873/10712 [1:09:00<39:30,  2.04it/s]
 55%|█████▍    | 5874/10712 [1:09:01<39:29,  2.04it/s]
 55%|█████▍    | 5875/10712 [1:09:01<39:30,  2.04it/s]{'loss': 3.7266, 'grad_norm': 0.20852674543857574, 'learning_rate': 0.000502770062339467, 'epoch': 0.55}

                                                      

 55%|█████▍    | 5875/10712 [1:09:01<39:30,  2.04it/s]
 55%|█████▍    | 5876/10712 [1:09:02<39:34,  2.04it/s]
 55%|█████▍    | 5877/10712 [1:09:02<39:31,  2.04it/s]
 55%|█████▍    | 5878/10712 [1:09:03<39:30,  2.04it/s]
 55%|█████▍    | 5879/10712 [1:09:03<39:29,  2.04it/s]
 55%|█████▍    | 5880/10712 [1:09:04<39:27,  2.04it/s]
 55%|█████▍    | 5881/10712 [1:09:04<39:29,  2.04it/s]
 55%|█████▍    | 5882/10712 [1:09:05<39:31,  2.04it/s]
 55%|█████▍    | 5883/10712 [1:09:05<39:29,  2.04it/s]
 55%|█████▍    | 5884/10712 [1:09:06<39:30,  2.04it/s]
 55%|█████▍    | 5885/10712 [1:09:06<39:31,  2.04it/s]
 55%|█████▍    | 5886/10712 [1:09:07<39:29,  2.04it/s]
 55%|█████▍    | 5887/10712 [1:09:07<39:24,  2.04it/s]
 55%|█████▍    | 5888/10712 [1:09:08<39:25,  2.04it/s]
 55%|█████▍    | 5889/10712 [1:09:08<39:25,  2.04it/s]
 55%|█████▍    | 5890/10712 [1:09:09<39:23,  2.04it/s]
 55%|█████▍    | 5891/10712 [1:09:09<39:22,  2.04it/s]
 55%|█████▌    | 5892/10712 [1:09:10<39:22,  2.04it/s]
 55%|█████▌    | 5893/10712 [1:09:10<39:23,  2.04it/s]
 55%|█████▌    | 5894/10712 [1:09:11<39:20,  2.04it/s]
 55%|█████▌    | 5895/10712 [1:09:11<39:20,  2.04it/s]
 55%|█████▌    | 5896/10712 [1:09:12<39:22,  2.04it/s]
 55%|█████▌    | 5897/10712 [1:09:12<39:20,  2.04it/s]
 55%|█████▌    | 5898/10712 [1:09:13<39:19,  2.04it/s]
 55%|█████▌    | 5899/10712 [1:09:13<39:20,  2.04it/s]
 55%|█████▌    | 5900/10712 [1:09:14<39:18,  2.04it/s]
                                                      
{'loss': 3.7232, 'grad_norm': 0.219155952334404, 'learning_rate': 0.0004986964360603231, 'epoch': 0.55}

 55%|█████▌    | 5900/10712 [1:09:14<39:18,  2.04it/s]
 55%|█████▌    | 5901/10712 [1:09:14<39:21,  2.04it/s]
 55%|█████▌    | 5902/10712 [1:09:15<39:21,  2.04it/s]
 55%|█████▌    | 5903/10712 [1:09:15<39:23,  2.04it/s]
 55%|█████▌    | 5904/10712 [1:09:16<39:21,  2.04it/s]
 55%|█████▌    | 5905/10712 [1:09:16<39:19,  2.04it/s]
 55%|█████▌    | 5906/10712 [1:09:17<39:19,  2.04it/s]
 55%|█████▌    | 5907/10712 [1:09:17<39:17,  2.04it/s]
 55%|█████▌    | 5908/10712 [1:09:18<39:18,  2.04it/s]
 55%|█████▌    | 5909/10712 [1:09:18<39:18,  2.04it/s]
 55%|█████▌    | 5910/10712 [1:09:19<39:15,  2.04it/s]
 55%|█████▌    | 5911/10712 [1:09:19<39:14,  2.04it/s]
 55%|█████▌    | 5912/10712 [1:09:20<39:14,  2.04it/s]
 55%|█████▌    | 5913/10712 [1:09:20<39:14,  2.04it/s]
 55%|█████▌    | 5914/10712 [1:09:21<39:11,  2.04it/s]
 55%|█████▌    | 5915/10712 [1:09:21<39:13,  2.04it/s]
 55%|█████▌    | 5916/10712 [1:09:22<39:14,  2.04it/s]
 55%|█████▌    | 5917/10712 [1:09:22<39:11,  2.04it/s]
 55%|█████▌    | 5918/10712 [1:09:23<39:10,  2.04it/s]
 55%|█████▌    | 5919/10712 [1:09:23<39:10,  2.04it/s]
 55%|█████▌    | 5920/10712 [1:09:23<39:09,  2.04it/s]
 55%|█████▌    | 5921/10712 [1:09:24<39:08,  2.04it/s]
 55%|█████▌    | 5922/10712 [1:09:24<39:06,  2.04it/s]
 55%|█████▌    | 5923/10712 [1:09:25<39:09,  2.04it/s]
 55%|█████▌    | 5924/10712 [1:09:25<39:05,  2.04it/s]
 55%|█████▌    | 5925/10712 [1:09:26<39:04,  2.04it/s]{'loss': 3.7245, 'grad_norm': 0.20829491317272186, 'learning_rate': 0.000494622896308973, 'epoch': 0.55}
                                                      

 55%|█████▌    | 5925/10712 [1:09:26<39:04,  2.04it/s]
 55%|█████▌    | 5926/10712 [1:09:26<39:07,  2.04it/s]
 55%|█████▌    | 5927/10712 [1:09:27<39:07,  2.04it/s]
 55%|█████▌    | 5928/10712 [1:09:27<39:04,  2.04it/s]
 55%|█████▌    | 5929/10712 [1:09:28<39:04,  2.04it/s]
 55%|█████▌    | 5930/10712 [1:09:28<39:06,  2.04it/s]
 55%|█████▌    | 5931/10712 [1:09:29<39:03,  2.04it/s]
 55%|█████▌    | 5932/10712 [1:09:29<39:02,  2.04it/s]
 55%|█████▌    | 5933/10712 [1:09:30<39:04,  2.04it/s]
 55%|█████▌    | 5934/10712 [1:09:30<39:02,  2.04it/s]
 55%|█████▌    | 5935/10712 [1:09:31<39:00,  2.04it/s]
 55%|█████▌    | 5936/10712 [1:09:31<38:57,  2.04it/s]
 55%|█████▌    | 5937/10712 [1:09:32<38:58,  2.04it/s]
 55%|█████▌    | 5938/10712 [1:09:32<38:59,  2.04it/s]
 55%|█████▌    | 5939/10712 [1:09:33<38:58,  2.04it/s]
 55%|█████▌    | 5940/10712 [1:09:33<38:58,  2.04it/s]
 55%|█████▌    | 5941/10712 [1:09:34<38:59,  2.04it/s]
 55%|█████▌    | 5942/10712 [1:09:34<39:28,  2.01it/s]
 55%|█████▌    | 5943/10712 [1:09:35<39:16,  2.02it/s]
 55%|█████▌    | 5944/10712 [1:09:35<39:12,  2.03it/s]
 55%|█████▌    | 5945/10712 [1:09:36<39:07,  2.03it/s]
 56%|█████▌    | 5946/10712 [1:09:36<39:02,  2.03it/s]
 56%|█████▌    | 5947/10712 [1:09:37<38:59,  2.04it/s]
 56%|█████▌    | 5948/10712 [1:09:37<38:57,  2.04it/s]
 56%|█████▌    | 5949/10712 [1:09:38<38:56,  2.04it/s]
 56%|█████▌    | 5950/10712 [1:09:38<38:56,  2.04it/s]
                                                      
{'loss': 3.7306, 'grad_norm': 0.21391883492469788, 'learning_rate': 0.0004905497134782966, 'epoch': 0.56}

 56%|█████▌    | 5950/10712 [1:09:38<38:56,  2.04it/s]
 56%|█████▌    | 5951/10712 [1:09:39<38:56,  2.04it/s]
 56%|█████▌    | 5952/10712 [1:09:39<38:56,  2.04it/s]
 56%|█████▌    | 5953/10712 [1:09:40<38:54,  2.04it/s]
 56%|█████▌    | 5954/10712 [1:09:40<38:51,  2.04it/s]
 56%|█████▌    | 5955/10712 [1:09:41<38:47,  2.04it/s]
 56%|█████▌    | 5956/10712 [1:09:41<38:50,  2.04it/s]
 56%|█████▌    | 5957/10712 [1:09:42<38:52,  2.04it/s]
 56%|█████▌    | 5958/10712 [1:09:42<38:49,  2.04it/s]
 56%|█████▌    | 5959/10712 [1:09:43<38:50,  2.04it/s]
 56%|█████▌    | 5960/10712 [1:09:43<38:50,  2.04it/s]
 56%|█████▌    | 5961/10712 [1:09:44<38:49,  2.04it/s]
 56%|█████▌    | 5962/10712 [1:09:44<38:45,  2.04it/s]
 56%|█████▌    | 5963/10712 [1:09:45<38:44,  2.04it/s]
 56%|█████▌    | 5964/10712 [1:09:45<38:45,  2.04it/s]
 56%|█████▌    | 5965/10712 [1:09:46<39:16,  2.01it/s]
 56%|█████▌    | 5966/10712 [1:09:46<39:04,  2.02it/s]
 56%|█████▌    | 5967/10712 [1:09:47<38:59,  2.03it/s]
 56%|█████▌    | 5968/10712 [1:09:47<38:56,  2.03it/s]
 56%|█████▌    | 5969/10712 [1:09:48<38:50,  2.04it/s]
 56%|█████▌    | 5970/10712 [1:09:48<38:46,  2.04it/s]
 56%|█████▌    | 5971/10712 [1:09:49<38:44,  2.04it/s]
 56%|█████▌    | 5972/10712 [1:09:49<38:44,  2.04it/s]
 56%|█████▌    | 5973/10712 [1:09:50<38:41,  2.04it/s]
 56%|█████▌    | 5974/10712 [1:09:50<39:08,  2.02it/s]
 56%|█████▌    | 5975/10712 [1:09:51<39:00,  2.02it/s]{'loss': 3.7271, 'grad_norm': 0.20910651981830597, 'learning_rate': 0.00048647715793748213, 'epoch': 0.56}
                                                      

 56%|█████▌    | 5975/10712 [1:09:51<39:00,  2.02it/s]
 56%|█████▌    | 5976/10712 [1:09:51<38:57,  2.03it/s]
 56%|█████▌    | 5977/10712 [1:09:51<38:50,  2.03it/s]
 56%|█████▌    | 5978/10712 [1:09:52<38:46,  2.03it/s]
 56%|█████▌    | 5979/10712 [1:09:52<38:46,  2.03it/s]
 56%|█████▌    | 5980/10712 [1:09:53<38:44,  2.04it/s]
 56%|█████▌    | 5981/10712 [1:09:53<38:39,  2.04it/s]
 56%|█████▌    | 5982/10712 [1:09:54<38:36,  2.04it/s]
 56%|█████▌    | 5983/10712 [1:09:55<45:07,  1.75it/s]
 56%|█████▌    | 5984/10712 [1:09:55<43:11,  1.82it/s]
 56%|█████▌    | 5985/10712 [1:09:56<41:49,  1.88it/s]
 56%|█████▌    | 5986/10712 [1:09:56<40:50,  1.93it/s]
 56%|█████▌    | 5987/10712 [1:09:57<40:07,  1.96it/s]
 56%|█████▌    | 5988/10712 [1:09:57<39:40,  1.98it/s]
 56%|█████▌    | 5989/10712 [1:09:58<39:21,  2.00it/s]
 56%|█████▌    | 5990/10712 [1:09:58<39:07,  2.01it/s]
 56%|█████▌    | 5991/10712 [1:09:59<38:58,  2.02it/s]
 56%|█████▌    | 5992/10712 [1:09:59<38:51,  2.02it/s]
 56%|█████▌    | 5993/10712 [1:10:00<38:47,  2.03it/s]
 56%|█████▌    | 5994/10712 [1:10:00<38:41,  2.03it/s]
 56%|█████▌    | 5995/10712 [1:10:01<38:36,  2.04it/s]
 56%|█████▌    | 5996/10712 [1:10:01<38:32,  2.04it/s]
 56%|█████▌    | 5997/10712 [1:10:02<38:32,  2.04it/s]
 56%|█████▌    | 5998/10712 [1:10:02<38:32,  2.04it/s]
 56%|█████▌    | 5999/10712 [1:10:03<38:30,  2.04it/s]
 56%|█████▌    | 6000/10712 [1:10:03<38:27,  2.04it/s]{'loss': 3.7214, 'grad_norm': 0.20185405015945435, 'learning_rate': 0.00048240550001407955, 'epoch': 0.56}

                                                      

 56%|█████▌    | 6000/10712 [1:10:03<38:27,  2.04it/s]
 56%|█████▌    | 6001/10712 [1:10:04<38:31,  2.04it/s]
 56%|█████▌    | 6002/10712 [1:10:04<38:32,  2.04it/s]
 56%|█████▌    | 6003/10712 [1:10:05<38:30,  2.04it/s]
 56%|█████▌    | 6004/10712 [1:10:05<38:29,  2.04it/s]
 56%|█████▌    | 6005/10712 [1:10:05<38:30,  2.04it/s]
 56%|█████▌    | 6006/10712 [1:10:06<38:31,  2.04it/s]
 56%|█████▌    | 6007/10712 [1:10:06<38:28,  2.04it/s]
 56%|█████▌    | 6008/10712 [1:10:07<38:28,  2.04it/s]
 56%|█████▌    | 6009/10712 [1:10:07<38:29,  2.04it/s]
 56%|█████▌    | 6010/10712 [1:10:08<38:27,  2.04it/s]
 56%|█████▌    | 6011/10712 [1:10:08<38:23,  2.04it/s]
 56%|█████▌    | 6012/10712 [1:10:09<38:23,  2.04it/s]
 56%|█████▌    | 6013/10712 [1:10:09<38:23,  2.04it/s]
 56%|█████▌    | 6014/10712 [1:10:10<44:33,  1.76it/s]
 56%|█████▌    | 6015/10712 [1:10:11<42:45,  1.83it/s]
 56%|█████▌    | 6016/10712 [1:10:11<41:27,  1.89it/s]
 56%|█████▌    | 6017/10712 [1:10:12<40:29,  1.93it/s]
 56%|█████▌    | 6018/10712 [1:10:12<39:48,  1.97it/s]
 56%|█████▌    | 6019/10712 [1:10:13<39:26,  1.98it/s]
 56%|█████▌    | 6020/10712 [1:10:13<39:05,  2.00it/s]
 56%|█████▌    | 6021/10712 [1:10:14<38:47,  2.02it/s]
 56%|█████▌    | 6022/10712 [1:10:14<38:35,  2.03it/s]
 56%|█████▌    | 6023/10712 [1:10:15<38:29,  2.03it/s]
 56%|█████▌    | 6024/10712 [1:10:15<38:24,  2.03it/s]
 56%|█████▌    | 6025/10712 [1:10:16<38:19,  2.04it/s]{'loss': 3.7204, 'grad_norm': 0.21575534343719482, 'learning_rate': 0.00047833500997605707, 'epoch': 0.56}
                                                      

 56%|█████▌    | 6025/10712 [1:10:16<38:19,  2.04it/s]
 56%|█████▋    | 6026/10712 [1:10:16<38:20,  2.04it/s]
 56%|█████▋    | 6027/10712 [1:10:17<38:20,  2.04it/s]
 56%|█████▋    | 6028/10712 [1:10:17<38:20,  2.04it/s]
 56%|█████▋    | 6029/10712 [1:10:18<38:16,  2.04it/s]
 56%|█████▋    | 6030/10712 [1:10:18<38:16,  2.04it/s]
 56%|█████▋    | 6031/10712 [1:10:18<38:16,  2.04it/s]
 56%|█████▋    | 6032/10712 [1:10:19<38:13,  2.04it/s]
 56%|█████▋    | 6033/10712 [1:10:19<38:11,  2.04it/s]
 56%|█████▋    | 6034/10712 [1:10:20<38:12,  2.04it/s]
 56%|█████▋    | 6035/10712 [1:10:20<38:11,  2.04it/s]
 56%|█████▋    | 6036/10712 [1:10:21<38:08,  2.04it/s]
 56%|█████▋    | 6037/10712 [1:10:21<38:08,  2.04it/s]
 56%|█████▋    | 6038/10712 [1:10:22<38:09,  2.04it/s]
 56%|█████▋    | 6039/10712 [1:10:22<38:08,  2.04it/s]
 56%|█████▋    | 6040/10712 [1:10:23<38:06,  2.04it/s]
 56%|█████▋    | 6041/10712 [1:10:23<38:06,  2.04it/s]
 56%|█████▋    | 6042/10712 [1:10:24<38:07,  2.04it/s]
 56%|█████▋    | 6043/10712 [1:10:24<38:09,  2.04it/s]
 56%|█████▋    | 6044/10712 [1:10:25<38:05,  2.04it/s]
 56%|█████▋    | 6045/10712 [1:10:25<38:06,  2.04it/s]
 56%|█████▋    | 6046/10712 [1:10:26<38:08,  2.04it/s]
 56%|█████▋    | 6047/10712 [1:10:26<38:07,  2.04it/s]
 56%|█████▋    | 6048/10712 [1:10:27<38:06,  2.04it/s]
 56%|█████▋    | 6049/10712 [1:10:27<38:05,  2.04it/s]
 56%|█████▋    | 6050/10712 [1:10:28<38:05,  2.04it/s]{'loss': 3.7123, 'grad_norm': 0.2116067260503769, 'learning_rate': 0.0004742659580138609, 'epoch': 0.56}
                                                      

 56%|█████▋    | 6050/10712 [1:10:28<38:05,  2.04it/s]
 56%|█████▋    | 6051/10712 [1:10:28<38:07,  2.04it/s]
 56%|█████▋    | 6052/10712 [1:10:29<38:03,  2.04it/s]
 57%|█████▋    | 6053/10712 [1:10:29<38:06,  2.04it/s]
 57%|█████▋    | 6054/10712 [1:10:30<38:04,  2.04it/s]
 57%|█████▋    | 6055/10712 [1:10:30<38:02,  2.04it/s]
 57%|█████▋    | 6056/10712 [1:10:31<37:59,  2.04it/s]
 57%|█████▋    | 6057/10712 [1:10:31<38:00,  2.04it/s]
 57%|█████▋    | 6058/10712 [1:10:32<37:59,  2.04it/s]
 57%|█████▋    | 6059/10712 [1:10:32<38:01,  2.04it/s]
 57%|█████▋    | 6060/10712 [1:10:33<38:00,  2.04it/s]
 57%|█████▋    | 6061/10712 [1:10:33<37:59,  2.04it/s]
 57%|█████▋    | 6062/10712 [1:10:34<37:57,  2.04it/s]
 57%|█████▋    | 6063/10712 [1:10:34<37:57,  2.04it/s]
 57%|█████▋    | 6064/10712 [1:10:35<37:58,  2.04it/s]
 57%|█████▋    | 6065/10712 [1:10:35<38:00,  2.04it/s]
 57%|█████▋    | 6066/10712 [1:10:36<37:56,  2.04it/s]
 57%|█████▋    | 6067/10712 [1:10:36<37:56,  2.04it/s]
 57%|█████▋    | 6068/10712 [1:10:37<37:58,  2.04it/s]
 57%|█████▋    | 6069/10712 [1:10:37<37:57,  2.04it/s]
 57%|█████▋    | 6070/10712 [1:10:38<37:54,  2.04it/s]
 57%|█████▋    | 6071/10712 [1:10:38<37:54,  2.04it/s]
 57%|█████▋    | 6072/10712 [1:10:39<37:55,  2.04it/s]
 57%|█████▋    | 6073/10712 [1:10:39<37:55,  2.04it/s]
 57%|█████▋    | 6074/10712 [1:10:40<37:52,  2.04it/s]
 57%|█████▋    | 6075/10712 [1:10:40<37:54,  2.04it/s]{'loss': 3.7123, 'grad_norm': 0.20598895847797394, 'learning_rate': 0.0004701986142224813, 'epoch': 0.57}
                                                      

 57%|█████▋    | 6075/10712 [1:10:40<37:54,  2.04it/s]
 57%|█████▋    | 6076/10712 [1:10:41<37:54,  2.04it/s]
 57%|█████▋    | 6077/10712 [1:10:41<37:51,  2.04it/s]
 57%|█████▋    | 6078/10712 [1:10:42<37:49,  2.04it/s]
 57%|█████▋    | 6079/10712 [1:10:42<37:51,  2.04it/s]
 57%|█████▋    | 6080/10712 [1:10:43<37:50,  2.04it/s]
 57%|█████▋    | 6081/10712 [1:10:43<37:48,  2.04it/s]
 57%|█████▋    | 6082/10712 [1:10:43<37:48,  2.04it/s]
 57%|█████▋    | 6083/10712 [1:10:44<37:49,  2.04it/s]
 57%|█████▋    | 6084/10712 [1:10:44<37:48,  2.04it/s]
 57%|█████▋    | 6085/10712 [1:10:45<37:46,  2.04it/s]
 57%|█████▋    | 6086/10712 [1:10:45<37:49,  2.04it/s]
 57%|█████▋    | 6087/10712 [1:10:46<37:48,  2.04it/s]
 57%|█████▋    | 6088/10712 [1:10:46<37:45,  2.04it/s]
 57%|█████▋    | 6089/10712 [1:10:47<37:45,  2.04it/s]
 57%|█████▋    | 6090/10712 [1:10:47<37:44,  2.04it/s]
 57%|█████▋    | 6091/10712 [1:10:48<37:46,  2.04it/s]
 57%|█████▋    | 6092/10712 [1:10:48<37:45,  2.04it/s]
 57%|█████▋    | 6093/10712 [1:10:49<37:45,  2.04it/s]
 57%|█████▋    | 6094/10712 [1:10:49<37:47,  2.04it/s]
 57%|█████▋    | 6095/10712 [1:10:50<37:45,  2.04it/s]
 57%|█████▋    | 6096/10712 [1:10:50<37:43,  2.04it/s]
 57%|█████▋    | 6097/10712 [1:10:51<37:43,  2.04it/s]
 57%|█████▋    | 6098/10712 [1:10:51<37:45,  2.04it/s]
 57%|█████▋    | 6099/10712 [1:10:52<37:41,  2.04it/s]
 57%|█████▋    | 6100/10712 [1:10:52<37:40,  2.04it/s]{'loss': 3.716, 'grad_norm': 0.20318546891212463, 'learning_rate': 0.0004661332485835234, 'epoch': 0.57}

                                                      

 57%|█████▋    | 6100/10712 [1:10:52<37:40,  2.04it/s]
 57%|█████▋    | 6101/10712 [1:10:53<37:43,  2.04it/s]
 57%|█████▋    | 6102/10712 [1:10:53<37:41,  2.04it/s]
 57%|█████▋    | 6103/10712 [1:10:54<37:39,  2.04it/s]
 57%|█████▋    | 6104/10712 [1:10:54<37:38,  2.04it/s]
 57%|█████▋    | 6105/10712 [1:10:55<37:44,  2.03it/s]
 57%|█████▋    | 6106/10712 [1:10:55<37:40,  2.04it/s]
 57%|█████▋    | 6107/10712 [1:10:56<37:38,  2.04it/s]
 57%|█████▋    | 6108/10712 [1:10:56<37:40,  2.04it/s]
 57%|█████▋    | 6109/10712 [1:10:57<37:38,  2.04it/s]
 57%|█████▋    | 6110/10712 [1:10:57<37:37,  2.04it/s]
 57%|█████▋    | 6111/10712 [1:10:58<37:40,  2.04it/s]
 57%|█████▋    | 6112/10712 [1:10:58<37:36,  2.04it/s]
 57%|█████▋    | 6113/10712 [1:10:59<37:34,  2.04it/s]
 57%|█████▋    | 6114/10712 [1:10:59<37:34,  2.04it/s]
 57%|█████▋    | 6115/10712 [1:11:00<37:34,  2.04it/s]
 57%|█████▋    | 6116/10712 [1:11:00<37:32,  2.04it/s]
 57%|█████▋    | 6117/10712 [1:11:01<37:30,  2.04it/s]
 57%|█████▋    | 6118/10712 [1:11:01<37:31,  2.04it/s]
 57%|█████▋    | 6119/10712 [1:11:02<37:31,  2.04it/s]
 57%|█████▋    | 6120/10712 [1:11:02<37:29,  2.04it/s]
 57%|█████▋    | 6121/10712 [1:11:03<37:27,  2.04it/s]
 57%|█████▋    | 6122/10712 [1:11:03<37:29,  2.04it/s]
 57%|█████▋    | 6123/10712 [1:11:04<37:29,  2.04it/s]
 57%|█████▋    | 6124/10712 [1:11:04<37:28,  2.04it/s]
 57%|█████▋    | 6125/10712 [1:11:05<37:28,  2.04it/s]{'loss': 3.7169, 'grad_norm': 0.21083594858646393, 'learning_rate': 0.0004620701309472869, 'epoch': 0.57}

                                                      

 57%|█████▋    | 6125/10712 [1:11:05<37:28,  2.04it/s]
 57%|█████▋    | 6126/10712 [1:11:05<37:33,  2.03it/s]
 57%|█████▋    | 6127/10712 [1:11:06<37:30,  2.04it/s]
 57%|█████▋    | 6128/10712 [1:11:06<37:28,  2.04it/s]
 57%|█████▋    | 6129/10712 [1:11:07<37:28,  2.04it/s]
 57%|█████▋    | 6130/10712 [1:11:07<37:26,  2.04it/s]
 57%|█████▋    | 6131/10712 [1:11:08<37:25,  2.04it/s]
 57%|█████▋    | 6132/10712 [1:11:08<37:22,  2.04it/s]
 57%|█████▋    | 6133/10712 [1:11:08<37:24,  2.04it/s]
 57%|█████▋    | 6134/10712 [1:11:09<37:23,  2.04it/s]
 57%|█████▋    | 6135/10712 [1:11:09<37:25,  2.04it/s]
 57%|█████▋    | 6136/10712 [1:11:10<37:23,  2.04it/s]
 57%|█████▋    | 6137/10712 [1:11:10<37:22,  2.04it/s]
 57%|█████▋    | 6138/10712 [1:11:11<37:19,  2.04it/s]
 57%|█████▋    | 6139/10712 [1:11:11<37:20,  2.04it/s]
 57%|█████▋    | 6140/10712 [1:11:12<37:22,  2.04it/s]
 57%|█████▋    | 6141/10712 [1:11:12<37:19,  2.04it/s]
 57%|█████▋    | 6142/10712 [1:11:13<37:19,  2.04it/s]
 57%|█████▋    | 6143/10712 [1:11:13<37:20,  2.04it/s]
 57%|█████▋    | 6144/10712 [1:11:14<37:19,  2.04it/s]
 57%|█████▋    | 6145/10712 [1:11:14<37:19,  2.04it/s]
 57%|█████▋    | 6146/10712 [1:11:15<37:15,  2.04it/s]
 57%|█████▋    | 6147/10712 [1:11:15<37:16,  2.04it/s]
 57%|█████▋    | 6148/10712 [1:11:16<37:17,  2.04it/s]
 57%|█████▋    | 6149/10712 [1:11:16<37:14,  2.04it/s]
 57%|█████▋    | 6150/10712 [1:11:17<37:13,  2.04it/s]{'loss': 3.7135, 'grad_norm': 0.21317414939403534, 'learning_rate': 0.0004580095310148539, 'epoch': 0.57}

                                                      

 57%|█████▋    | 6150/10712 [1:11:17<37:13,  2.04it/s]
 57%|█████▋    | 6151/10712 [1:11:17<37:18,  2.04it/s]
 57%|█████▋    | 6152/10712 [1:11:18<37:16,  2.04it/s]
 57%|█████▋    | 6153/10712 [1:11:18<37:14,  2.04it/s]
 57%|█████▋    | 6154/10712 [1:11:19<37:14,  2.04it/s]
 57%|█████▋    | 6155/10712 [1:11:19<37:15,  2.04it/s]
 57%|█████▋    | 6156/10712 [1:11:20<37:12,  2.04it/s]
 57%|█████▋    | 6157/10712 [1:11:20<37:09,  2.04it/s]
 57%|█████▋    | 6158/10712 [1:11:21<37:11,  2.04it/s]
 57%|█████▋    | 6159/10712 [1:11:21<37:12,  2.04it/s]
 58%|█████▊    | 6160/10712 [1:11:22<37:10,  2.04it/s]
 58%|█████▊    | 6161/10712 [1:11:22<37:11,  2.04it/s]
 58%|█████▊    | 6162/10712 [1:11:23<37:13,  2.04it/s]
 58%|█████▊    | 6163/10712 [1:11:23<37:11,  2.04it/s]
 58%|█████▊    | 6164/10712 [1:11:24<37:10,  2.04it/s]
 58%|█████▊    | 6165/10712 [1:11:24<37:09,  2.04it/s]
 58%|█████▊    | 6166/10712 [1:11:25<37:08,  2.04it/s]
 58%|█████▊    | 6167/10712 [1:11:25<37:05,  2.04it/s]
 58%|█████▊    | 6168/10712 [1:11:26<37:06,  2.04it/s]
 58%|█████▊    | 6169/10712 [1:11:26<37:07,  2.04it/s]
 58%|█████▊    | 6170/10712 [1:11:27<37:07,  2.04it/s]
 58%|█████▊    | 6171/10712 [1:11:27<37:04,  2.04it/s]
 58%|█████▊    | 6172/10712 [1:11:28<37:04,  2.04it/s]
 58%|█████▊    | 6173/10712 [1:11:28<37:07,  2.04it/s]
 58%|█████▊    | 6174/10712 [1:11:29<37:02,  2.04it/s]
 58%|█████▊    | 6175/10712 [1:11:29<36:59,  2.04it/s]{'loss': 3.7203, 'grad_norm': 0.21442659199237823, 'learning_rate': 0.00045395171832018655, 'epoch': 0.58}
                                                      

 58%|█████▊    | 6175/10712 [1:11:29<36:59,  2.04it/s]
 58%|█████▊    | 6176/10712 [1:11:30<37:02,  2.04it/s]
 58%|█████▊    | 6177/10712 [1:11:30<37:03,  2.04it/s]
 58%|█████▊    | 6178/10712 [1:11:31<37:02,  2.04it/s]
 58%|█████▊    | 6179/10712 [1:11:31<37:01,  2.04it/s]
 58%|█████▊    | 6180/10712 [1:11:32<37:02,  2.04it/s]
 58%|█████▊    | 6181/10712 [1:11:32<36:59,  2.04it/s]
 58%|█████▊    | 6182/10712 [1:11:33<36:57,  2.04it/s]
 58%|█████▊    | 6183/10712 [1:11:33<36:54,  2.05it/s]
 58%|█████▊    | 6184/10712 [1:11:33<36:56,  2.04it/s]
 58%|█████▊    | 6185/10712 [1:11:34<36:57,  2.04it/s]
 58%|█████▊    | 6186/10712 [1:11:34<36:54,  2.04it/s]
 58%|█████▊    | 6187/10712 [1:11:35<36:56,  2.04it/s]
 58%|█████▊    | 6188/10712 [1:11:35<36:56,  2.04it/s]
 58%|█████▊    | 6189/10712 [1:11:36<36:55,  2.04it/s]
 58%|█████▊    | 6190/10712 [1:11:36<36:54,  2.04it/s]
 58%|█████▊    | 6191/10712 [1:11:37<36:53,  2.04it/s]
 58%|█████▊    | 6192/10712 [1:11:37<36:55,  2.04it/s]
 58%|█████▊    | 6193/10712 [1:11:38<36:53,  2.04it/s]
 58%|█████▊    | 6194/10712 [1:11:38<36:52,  2.04it/s]
 58%|█████▊    | 6195/10712 [1:11:39<36:53,  2.04it/s]
 58%|█████▊    | 6196/10712 [1:11:39<36:53,  2.04it/s]
 58%|█████▊    | 6197/10712 [1:11:40<36:51,  2.04it/s]
 58%|█████▊    | 6198/10712 [1:11:40<36:52,  2.04it/s]
 58%|█████▊    | 6199/10712 [1:11:41<36:52,  2.04it/s]
 58%|█████▊    | 6200/10712 [1:11:41<36:52,  2.04it/s]
                                                      
{'loss': 3.7083, 'grad_norm': 0.20749349892139435, 'learning_rate': 0.00044989696221223635, 'epoch': 0.58}

 58%|█████▊    | 6200/10712 [1:11:41<36:52,  2.04it/s]
 58%|█████▊    | 6201/10712 [1:11:42<36:52,  2.04it/s]
 58%|█████▊    | 6202/10712 [1:11:42<36:54,  2.04it/s]
 58%|█████▊    | 6203/10712 [1:11:43<36:52,  2.04it/s]
 58%|█████▊    | 6204/10712 [1:11:43<36:50,  2.04it/s]
 58%|█████▊    | 6205/10712 [1:11:44<36:48,  2.04it/s]
 58%|█████▊    | 6206/10712 [1:11:44<36:51,  2.04it/s]
 58%|█████▊    | 6207/10712 [1:11:45<36:50,  2.04it/s]
 58%|█████▊    | 6208/10712 [1:11:45<36:45,  2.04it/s]
 58%|█████▊    | 6209/10712 [1:11:46<36:48,  2.04it/s]
 58%|█████▊    | 6210/10712 [1:11:46<36:48,  2.04it/s]
 58%|█████▊    | 6211/10712 [1:11:47<36:48,  2.04it/s]
 58%|█████▊    | 6212/10712 [1:11:47<36:46,  2.04it/s]
 58%|█████▊    | 6213/10712 [1:11:48<36:45,  2.04it/s]
 58%|█████▊    | 6214/10712 [1:11:48<36:45,  2.04it/s]
 58%|█████▊    | 6215/10712 [1:11:49<36:43,  2.04it/s]
 58%|█████▊    | 6216/10712 [1:11:49<38:11,  1.96it/s]
 58%|█████▊    | 6217/10712 [1:11:50<37:46,  1.98it/s]
 58%|█████▊    | 6218/10712 [1:11:50<37:26,  2.00it/s]
 58%|█████▊    | 6219/10712 [1:11:51<37:15,  2.01it/s]
 58%|█████▊    | 6220/10712 [1:11:51<37:05,  2.02it/s]
 58%|█████▊    | 6221/10712 [1:11:52<36:57,  2.02it/s]
 58%|█████▊    | 6222/10712 [1:11:52<36:50,  2.03it/s]
 58%|█████▊    | 6223/10712 [1:11:53<36:46,  2.03it/s]
 58%|█████▊    | 6224/10712 [1:11:53<36:46,  2.03it/s]
 58%|█████▊    | 6225/10712 [1:11:54<36:43,  2.04it/s]{'loss': 3.7188, 'grad_norm': 0.2100040763616562, 'learning_rate': 0.000445845531837065, 'epoch': 0.58}
                                                      

 58%|█████▊    | 6225/10712 [1:11:54<36:43,  2.04it/s]
 58%|█████▊    | 6226/10712 [1:11:54<36:42,  2.04it/s]
 58%|█████▊    | 6227/10712 [1:11:55<36:42,  2.04it/s]
 58%|█████▊    | 6228/10712 [1:11:55<36:41,  2.04it/s]
 58%|█████▊    | 6229/10712 [1:11:56<36:38,  2.04it/s]
 58%|█████▊    | 6230/10712 [1:11:56<36:38,  2.04it/s]
 58%|█████▊    | 6231/10712 [1:11:57<36:40,  2.04it/s]
 58%|█████▊    | 6232/10712 [1:11:57<36:36,  2.04it/s]
 58%|█████▊    | 6233/10712 [1:11:58<36:35,  2.04it/s]
 58%|█████▊    | 6234/10712 [1:11:58<36:36,  2.04it/s]
 58%|█████▊    | 6235/10712 [1:11:59<36:34,  2.04it/s]
 58%|█████▊    | 6236/10712 [1:11:59<36:34,  2.04it/s]
 58%|█████▊    | 6237/10712 [1:12:00<36:33,  2.04it/s]
 58%|█████▊    | 6238/10712 [1:12:00<36:34,  2.04it/s]
 58%|█████▊    | 6239/10712 [1:12:01<36:33,  2.04it/s]
 58%|█████▊    | 6240/10712 [1:12:01<36:31,  2.04it/s]
 58%|█████▊    | 6241/10712 [1:12:02<36:30,  2.04it/s]
 58%|█████▊    | 6242/10712 [1:12:02<36:30,  2.04it/s]
 58%|█████▊    | 6243/10712 [1:12:02<36:30,  2.04it/s]
 58%|█████▊    | 6244/10712 [1:12:03<36:30,  2.04it/s]
 58%|█████▊    | 6245/10712 [1:12:03<36:29,  2.04it/s]
 58%|█████▊    | 6246/10712 [1:12:04<36:31,  2.04it/s]
 58%|█████▊    | 6247/10712 [1:12:04<36:27,  2.04it/s]
 58%|█████▊    | 6248/10712 [1:12:05<36:24,  2.04it/s]
 58%|█████▊    | 6249/10712 [1:12:05<36:26,  2.04it/s]
 58%|█████▊    | 6250/10712 [1:12:06<36:26,  2.04it/s]{'loss': 3.7027, 'grad_norm': 0.21989786624908447, 'learning_rate': 0.000441797696119979, 'epoch': 0.58}

                                                      

 58%|█████▊    | 6250/10712 [1:12:06<36:26,  2.04it/s]
 58%|█████▊    | 6251/10712 [1:12:06<36:27,  2.04it/s]
 58%|█████▊    | 6252/10712 [1:12:07<36:28,  2.04it/s]
 58%|█████▊    | 6253/10712 [1:12:07<36:28,  2.04it/s]
 58%|█████▊    | 6254/10712 [1:12:08<36:25,  2.04it/s]
 58%|█████▊    | 6255/10712 [1:12:08<36:23,  2.04it/s]
 58%|█████▊    | 6256/10712 [1:12:09<36:22,  2.04it/s]
 58%|█████▊    | 6257/10712 [1:12:09<36:22,  2.04it/s]
 58%|█████▊    | 6258/10712 [1:12:10<36:20,  2.04it/s]
 58%|█████▊    | 6259/10712 [1:12:10<36:17,  2.04it/s]
 58%|█████▊    | 6260/10712 [1:12:11<36:18,  2.04it/s]
 58%|█████▊    | 6261/10712 [1:12:11<36:21,  2.04it/s]
 58%|█████▊    | 6262/10712 [1:12:12<36:19,  2.04it/s]
 58%|█████▊    | 6263/10712 [1:12:12<36:19,  2.04it/s]
 58%|█████▊    | 6264/10712 [1:12:13<36:20,  2.04it/s]
 58%|█████▊    | 6265/10712 [1:12:13<36:23,  2.04it/s]
 58%|█████▊    | 6266/10712 [1:12:14<36:22,  2.04it/s]
 59%|█████▊    | 6267/10712 [1:12:14<36:19,  2.04it/s]
 59%|█████▊    | 6268/10712 [1:12:15<36:19,  2.04it/s]
 59%|█████▊    | 6269/10712 [1:12:15<36:18,  2.04it/s]
 59%|█████▊    | 6270/10712 [1:12:16<36:18,  2.04it/s]
 59%|█████▊    | 6271/10712 [1:12:16<36:17,  2.04it/s]
 59%|█████▊    | 6272/10712 [1:12:17<36:14,  2.04it/s]
 59%|█████▊    | 6273/10712 [1:12:17<36:16,  2.04it/s]
 59%|█████▊    | 6274/10712 [1:12:18<36:17,  2.04it/s]
 59%|█████▊    | 6275/10712 [1:12:18<36:15,  2.04it/s]
                                                      
{'loss': 3.7139, 'grad_norm': 0.21113233268260956, 'learning_rate': 0.0004377537237476793, 'epoch': 0.59}

 59%|█████▊    | 6275/10712 [1:12:18<36:15,  2.04it/s]
 59%|█████▊    | 6276/10712 [1:12:19<36:18,  2.04it/s]
 59%|█████▊    | 6277/10712 [1:12:19<36:17,  2.04it/s]
 59%|█████▊    | 6278/10712 [1:12:20<36:17,  2.04it/s]
 59%|█████▊    | 6279/10712 [1:12:20<36:13,  2.04it/s]
 59%|█████▊    | 6280/10712 [1:12:21<36:11,  2.04it/s]
 59%|█████▊    | 6281/10712 [1:12:21<36:11,  2.04it/s]
 59%|█████▊    | 6282/10712 [1:12:22<36:12,  2.04it/s]
 59%|█████▊    | 6283/10712 [1:12:22<36:10,  2.04it/s]
 59%|█████▊    | 6284/10712 [1:12:23<36:08,  2.04it/s]
 59%|█████▊    | 6285/10712 [1:12:23<36:08,  2.04it/s]
 59%|█████▊    | 6286/10712 [1:12:24<36:08,  2.04it/s]
 59%|█████▊    | 6287/10712 [1:12:24<36:08,  2.04it/s]
 59%|█████▊    | 6288/10712 [1:12:25<36:07,  2.04it/s]
 59%|█████▊    | 6289/10712 [1:12:25<36:08,  2.04it/s]
 59%|█████▊    | 6290/10712 [1:12:26<36:10,  2.04it/s]
 59%|█████▊    | 6291/10712 [1:12:26<36:07,  2.04it/s]
 59%|█████▊    | 6292/10712 [1:12:27<36:08,  2.04it/s]
 59%|█████▊    | 6293/10712 [1:12:27<36:08,  2.04it/s]
 59%|█████▉    | 6294/10712 [1:12:27<36:04,  2.04it/s]
 59%|█████▉    | 6295/10712 [1:12:28<36:06,  2.04it/s]
 59%|█████▉    | 6296/10712 [1:12:28<36:07,  2.04it/s]
 59%|█████▉    | 6297/10712 [1:12:29<36:05,  2.04it/s]
 59%|█████▉    | 6298/10712 [1:12:29<36:02,  2.04it/s]
 59%|█████▉    | 6299/10712 [1:12:30<35:59,  2.04it/s]
 59%|█████▉    | 6300/10712 [1:12:30<36:01,  2.04it/s]{'loss': 3.7153, 'grad_norm': 0.20461583137512207, 'learning_rate': 0.00043371388315042623, 'epoch': 0.59}

                                                      

 59%|█████▉    | 6300/10712 [1:12:30<36:01,  2.04it/s]
 59%|█████▉    | 6301/10712 [1:12:31<36:02,  2.04it/s]
 59%|█████▉    | 6302/10712 [1:12:31<36:00,  2.04it/s]
 59%|█████▉    | 6303/10712 [1:12:32<36:03,  2.04it/s]
 59%|█████▉    | 6304/10712 [1:12:32<36:01,  2.04it/s]
 59%|█████▉    | 6305/10712 [1:12:33<35:58,  2.04it/s]
 59%|█████▉    | 6306/10712 [1:12:33<36:25,  2.02it/s]
 59%|█████▉    | 6307/10712 [1:12:34<36:19,  2.02it/s]
 59%|█████▉    | 6308/10712 [1:12:34<36:13,  2.03it/s]
 59%|█████▉    | 6309/10712 [1:12:35<36:05,  2.03it/s]
 59%|█████▉    | 6310/10712 [1:12:35<36:02,  2.04it/s]
 59%|█████▉    | 6311/10712 [1:12:36<36:01,  2.04it/s]
 59%|█████▉    | 6312/10712 [1:12:36<35:59,  2.04it/s]
 59%|█████▉    | 6313/10712 [1:12:37<35:57,  2.04it/s]
 59%|█████▉    | 6314/10712 [1:12:37<35:58,  2.04it/s]
 59%|█████▉    | 6315/10712 [1:12:38<35:56,  2.04it/s]
 59%|█████▉    | 6316/10712 [1:12:38<35:53,  2.04it/s]
 59%|█████▉    | 6317/10712 [1:12:39<35:52,  2.04it/s]
 59%|█████▉    | 6318/10712 [1:12:39<35:51,  2.04it/s]
 59%|█████▉    | 6319/10712 [1:12:40<35:50,  2.04it/s]
 59%|█████▉    | 6320/10712 [1:12:40<35:50,  2.04it/s]
 59%|█████▉    | 6321/10712 [1:12:41<35:48,  2.04it/s]
 59%|█████▉    | 6322/10712 [1:12:41<35:50,  2.04it/s]
 59%|█████▉    | 6323/10712 [1:12:42<35:49,  2.04it/s]
 59%|█████▉    | 6324/10712 [1:12:42<35:47,  2.04it/s]
 59%|█████▉    | 6325/10712 [1:12:43<35:46,  2.04it/s]
                                                      
{'loss': 3.709, 'grad_norm': 0.21113188564777374, 'learning_rate': 0.00042967844248422153, 'epoch': 0.59}

 59%|█████▉    | 6325/10712 [1:12:43<35:46,  2.04it/s]
 59%|█████▉    | 6326/10712 [1:12:43<35:50,  2.04it/s]
 59%|█████▉    | 6327/10712 [1:12:44<35:48,  2.04it/s]
 59%|█████▉    | 6328/10712 [1:12:44<35:46,  2.04it/s]
 59%|█████▉    | 6329/10712 [1:12:45<35:51,  2.04it/s]
 59%|█████▉    | 6330/10712 [1:12:45<35:50,  2.04it/s]
 59%|█████▉    | 6331/10712 [1:12:46<35:47,  2.04it/s]
 59%|█████▉    | 6332/10712 [1:12:46<35:46,  2.04it/s]
 59%|█████▉    | 6333/10712 [1:12:47<35:46,  2.04it/s]
 59%|█████▉    | 6334/10712 [1:12:47<35:45,  2.04it/s]
 59%|█████▉    | 6335/10712 [1:12:48<35:42,  2.04it/s]
 59%|█████▉    | 6336/10712 [1:12:48<35:45,  2.04it/s]
 59%|█████▉    | 6337/10712 [1:12:49<35:46,  2.04it/s]
 59%|█████▉    | 6338/10712 [1:12:49<35:44,  2.04it/s]
 59%|█████▉    | 6339/10712 [1:12:50<35:45,  2.04it/s]
 59%|█████▉    | 6340/10712 [1:12:50<35:44,  2.04it/s]
 59%|█████▉    | 6341/10712 [1:12:51<35:43,  2.04it/s]
 59%|█████▉    | 6342/10712 [1:12:51<35:42,  2.04it/s]
 59%|█████▉    | 6343/10712 [1:12:52<35:41,  2.04it/s]
 59%|█████▉    | 6344/10712 [1:12:52<35:41,  2.04it/s]
 59%|█████▉    | 6345/10712 [1:12:52<35:39,  2.04it/s]
 59%|█████▉    | 6346/10712 [1:12:53<35:37,  2.04it/s]
 59%|█████▉    | 6347/10712 [1:12:53<35:39,  2.04it/s]
 59%|█████▉    | 6348/10712 [1:12:54<35:37,  2.04it/s]
 59%|█████▉    | 6349/10712 [1:12:54<35:36,  2.04it/s]
 59%|█████▉    | 6350/10712 [1:12:55<35:36,  2.04it/s]{'loss': 3.7062, 'grad_norm': 0.21465139091014862, 'learning_rate': 0.00042564766961300916, 'epoch': 0.59}

                                                      

 59%|█████▉    | 6350/10712 [1:12:55<35:36,  2.04it/s]
 59%|█████▉    | 6351/10712 [1:12:55<35:40,  2.04it/s]
 59%|█████▉    | 6352/10712 [1:12:56<35:39,  2.04it/s]
 59%|█████▉    | 6353/10712 [1:12:56<35:36,  2.04it/s]
 59%|█████▉    | 6354/10712 [1:12:57<35:34,  2.04it/s]
 59%|█████▉    | 6355/10712 [1:12:57<35:36,  2.04it/s]
 59%|█████▉    | 6356/10712 [1:12:58<35:36,  2.04it/s]
 59%|█████▉    | 6357/10712 [1:12:58<35:35,  2.04it/s]
 59%|█████▉    | 6358/10712 [1:12:59<35:36,  2.04it/s]
 59%|█████▉    | 6359/10712 [1:12:59<35:35,  2.04it/s]
 59%|█████▉    | 6360/10712 [1:13:00<35:33,  2.04it/s]
 59%|█████▉    | 6361/10712 [1:13:00<35:32,  2.04it/s]
 59%|█████▉    | 6362/10712 [1:13:01<35:32,  2.04it/s]
 59%|█████▉    | 6363/10712 [1:13:01<35:32,  2.04it/s]
 59%|█████▉    | 6364/10712 [1:13:02<35:31,  2.04it/s]
 59%|█████▉    | 6365/10712 [1:13:02<35:29,  2.04it/s]
 59%|█████▉    | 6366/10712 [1:13:03<35:30,  2.04it/s]
 59%|█████▉    | 6367/10712 [1:13:03<35:29,  2.04it/s]
 59%|█████▉    | 6368/10712 [1:13:04<35:30,  2.04it/s]
 59%|█████▉    | 6369/10712 [1:13:04<35:28,  2.04it/s]
 59%|█████▉    | 6370/10712 [1:13:05<35:28,  2.04it/s]
 59%|█████▉    | 6371/10712 [1:13:05<35:28,  2.04it/s]
 59%|█████▉    | 6372/10712 [1:13:06<35:24,  2.04it/s]
 59%|█████▉    | 6373/10712 [1:13:06<35:24,  2.04it/s]
 60%|█████▉    | 6374/10712 [1:13:07<35:25,  2.04it/s]
 60%|█████▉    | 6375/10712 [1:13:07<35:26,  2.04it/s]{'loss': 3.7134, 'grad_norm': 0.2094343602657318, 'learning_rate': 0.00042162183209089473, 'epoch': 0.6}

                                                      

 60%|█████▉    | 6375/10712 [1:13:07<35:26,  2.04it/s]
 60%|█████▉    | 6376/10712 [1:13:08<35:25,  2.04it/s]
 60%|█████▉    | 6377/10712 [1:13:08<35:24,  2.04it/s]
 60%|█████▉    | 6378/10712 [1:13:09<35:23,  2.04it/s]
 60%|█████▉    | 6379/10712 [1:13:09<35:23,  2.04it/s]
 60%|█████▉    | 6380/10712 [1:13:10<35:19,  2.04it/s]
 60%|█████▉    | 6381/10712 [1:13:10<35:21,  2.04it/s]
 60%|█████▉    | 6382/10712 [1:13:11<35:22,  2.04it/s]
 60%|█████▉    | 6383/10712 [1:13:11<35:20,  2.04it/s]
 60%|█████▉    | 6384/10712 [1:13:12<35:18,  2.04it/s]
 60%|█████▉    | 6385/10712 [1:13:12<35:18,  2.04it/s]
 60%|█████▉    | 6386/10712 [1:13:13<35:19,  2.04it/s]
 60%|█████▉    | 6387/10712 [1:13:13<35:16,  2.04it/s]
 60%|█████▉    | 6388/10712 [1:13:14<35:17,  2.04it/s]
 60%|█████▉    | 6389/10712 [1:13:14<35:19,  2.04it/s]
 60%|█████▉    | 6390/10712 [1:13:15<35:19,  2.04it/s]
 60%|█████▉    | 6391/10712 [1:13:15<35:16,  2.04it/s]
 60%|█████▉    | 6392/10712 [1:13:16<35:14,  2.04it/s]
 60%|█████▉    | 6393/10712 [1:13:16<35:14,  2.04it/s]
 60%|█████▉    | 6394/10712 [1:13:17<35:14,  2.04it/s]
 60%|█████▉    | 6395/10712 [1:13:17<35:14,  2.04it/s]
 60%|█████▉    | 6396/10712 [1:13:17<35:12,  2.04it/s]
 60%|█████▉    | 6397/10712 [1:13:18<35:12,  2.04it/s]
 60%|█████▉    | 6398/10712 [1:13:18<35:15,  2.04it/s]
 60%|█████▉    | 6399/10712 [1:13:19<35:12,  2.04it/s]
 60%|█████▉    | 6400/10712 [1:13:19<35:10,  2.04it/s]{'loss': 3.7072, 'grad_norm': 0.21603794395923615, 'learning_rate': 0.00041760119714438575, 'epoch': 0.6}

                                                      

 60%|█████▉    | 6400/10712 [1:13:19<35:10,  2.04it/s]
 60%|█████▉    | 6401/10712 [1:13:20<35:13,  2.04it/s]
 60%|█████▉    | 6402/10712 [1:13:20<35:11,  2.04it/s]
 60%|█████▉    | 6403/10712 [1:13:21<35:13,  2.04it/s]
 60%|█████▉    | 6404/10712 [1:13:21<35:09,  2.04it/s]
 60%|█████▉    | 6405/10712 [1:13:22<35:09,  2.04it/s]
 60%|█████▉    | 6406/10712 [1:13:22<35:07,  2.04it/s]
 60%|█████▉    | 6407/10712 [1:13:23<35:08,  2.04it/s]
 60%|█████▉    | 6408/10712 [1:13:23<35:10,  2.04it/s]
 60%|█████▉    | 6409/10712 [1:13:24<35:08,  2.04it/s]
 60%|█████▉    | 6410/10712 [1:13:24<35:07,  2.04it/s]
 60%|█████▉    | 6411/10712 [1:13:25<35:07,  2.04it/s]
 60%|█████▉    | 6412/10712 [1:13:25<35:07,  2.04it/s]
 60%|█████▉    | 6413/10712 [1:13:26<35:08,  2.04it/s]
 60%|█████▉    | 6414/10712 [1:13:26<35:05,  2.04it/s]
 60%|█████▉    | 6415/10712 [1:13:27<35:03,  2.04it/s]
 60%|█████▉    | 6416/10712 [1:13:27<35:02,  2.04it/s]
 60%|█████▉    | 6417/10712 [1:13:28<35:30,  2.02it/s]
 60%|█████▉    | 6418/10712 [1:13:28<35:21,  2.02it/s]
 60%|█████▉    | 6419/10712 [1:13:29<35:13,  2.03it/s]
 60%|█████▉    | 6420/10712 [1:13:29<35:11,  2.03it/s]
 60%|█████▉    | 6421/10712 [1:13:30<35:10,  2.03it/s]
 60%|█████▉    | 6422/10712 [1:13:30<35:06,  2.04it/s]
 60%|█████▉    | 6423/10712 [1:13:31<35:03,  2.04it/s]
 60%|█████▉    | 6424/10712 [1:13:31<35:02,  2.04it/s]
 60%|█████▉    | 6425/10712 [1:13:32<35:03,  2.04it/s]
                                                      
{'loss': 3.7007, 'grad_norm': 0.20505093038082123, 'learning_rate': 0.0004135860316546541, 'epoch': 0.6}

 60%|█████▉    | 6425/10712 [1:13:32<35:03,  2.04it/s]
 60%|█████▉    | 6426/10712 [1:13:32<35:03,  2.04it/s]
 60%|█████▉    | 6427/10712 [1:13:33<34:58,  2.04it/s]
 60%|██████    | 6428/10712 [1:13:33<35:00,  2.04it/s]
 60%|██████    | 6429/10712 [1:13:34<34:59,  2.04it/s]
 60%|██████    | 6430/10712 [1:13:34<34:58,  2.04it/s]
 60%|██████    | 6431/10712 [1:13:35<35:01,  2.04it/s]
 60%|██████    | 6432/10712 [1:13:35<35:01,  2.04it/s]
 60%|██████    | 6433/10712 [1:13:36<34:57,  2.04it/s]
 60%|██████    | 6434/10712 [1:13:36<34:55,  2.04it/s]
 60%|██████    | 6435/10712 [1:13:37<34:54,  2.04it/s]
 60%|██████    | 6436/10712 [1:13:37<34:55,  2.04it/s]
 60%|██████    | 6437/10712 [1:13:38<34:53,  2.04it/s]
 60%|██████    | 6438/10712 [1:13:38<34:51,  2.04it/s]
 60%|██████    | 6439/10712 [1:13:39<34:50,  2.04it/s]
 60%|██████    | 6440/10712 [1:13:39<34:52,  2.04it/s]
 60%|██████    | 6441/10712 [1:13:40<34:51,  2.04it/s]
 60%|██████    | 6442/10712 [1:13:40<34:49,  2.04it/s]
 60%|██████    | 6443/10712 [1:13:41<34:49,  2.04it/s]
 60%|██████    | 6444/10712 [1:13:41<34:50,  2.04it/s]
 60%|██████    | 6445/10712 [1:13:42<34:49,  2.04it/s]
 60%|██████    | 6446/10712 [1:13:42<34:48,  2.04it/s]
 60%|██████    | 6447/10712 [1:13:42<34:50,  2.04it/s]
 60%|██████    | 6448/10712 [1:13:43<34:50,  2.04it/s]
 60%|██████    | 6449/10712 [1:13:43<34:47,  2.04it/s]
 60%|██████    | 6450/10712 [1:13:44<34:47,  2.04it/s]
                                                      {'loss': 3.6971, 'grad_norm': 0.2065543681383133, 'learning_rate': 0.0004095766021398206, 'epoch': 0.6}


 60%|██████    | 6450/10712 [1:13:44<34:47,  2.04it/s]
 60%|██████    | 6451/10712 [1:13:44<34:50,  2.04it/s]
 60%|██████    | 6452/10712 [1:13:45<34:50,  2.04it/s]
 60%|██████    | 6453/10712 [1:13:45<34:47,  2.04it/s]
 60%|██████    | 6454/10712 [1:13:46<34:47,  2.04it/s]
 60%|██████    | 6455/10712 [1:13:46<34:48,  2.04it/s]
 60%|██████    | 6456/10712 [1:13:47<34:45,  2.04it/s]
 60%|██████    | 6457/10712 [1:13:47<34:46,  2.04it/s]
 60%|██████    | 6458/10712 [1:13:48<34:45,  2.04it/s]
 60%|██████    | 6459/10712 [1:13:48<34:43,  2.04it/s]
 60%|██████    | 6460/10712 [1:13:49<34:44,  2.04it/s]
 60%|██████    | 6461/10712 [1:13:49<34:45,  2.04it/s]
 60%|██████    | 6462/10712 [1:13:50<34:44,  2.04it/s]
 60%|██████    | 6463/10712 [1:13:50<34:41,  2.04it/s]
 60%|██████    | 6464/10712 [1:13:51<34:43,  2.04it/s]
 60%|██████    | 6465/10712 [1:13:51<34:43,  2.04it/s]
 60%|██████    | 6466/10712 [1:13:52<34:39,  2.04it/s]
 60%|██████    | 6467/10712 [1:13:52<34:38,  2.04it/s]
 60%|██████    | 6468/10712 [1:13:53<34:39,  2.04it/s]
 60%|██████    | 6469/10712 [1:13:53<34:39,  2.04it/s]
 60%|██████    | 6470/10712 [1:13:54<34:37,  2.04it/s]
 60%|██████    | 6471/10712 [1:13:54<34:35,  2.04it/s]
 60%|██████    | 6472/10712 [1:13:55<34:37,  2.04it/s]
 60%|██████    | 6473/10712 [1:13:55<34:37,  2.04it/s]
 60%|██████    | 6474/10712 [1:13:56<35:05,  2.01it/s]
 60%|██████    | 6475/10712 [1:13:56<34:58,  2.02it/s]
                                                      {'loss': 3.6995, 'grad_norm': 0.21211493015289307, 'learning_rate': 0.00040557317473726437, 'epoch': 0.6}


 60%|██████    | 6475/10712 [1:13:56<34:58,  2.02it/s]
 60%|██████    | 6476/10712 [1:13:57<34:53,  2.02it/s]
 60%|██████    | 6477/10712 [1:13:57<34:46,  2.03it/s]
 60%|██████    | 6478/10712 [1:13:58<34:43,  2.03it/s]
 60%|██████    | 6479/10712 [1:13:58<35:06,  2.01it/s]
 60%|██████    | 6480/10712 [1:13:59<34:56,  2.02it/s]
 61%|██████    | 6481/10712 [1:13:59<34:48,  2.03it/s]
 61%|██████    | 6482/10712 [1:14:00<34:45,  2.03it/s]
 61%|██████    | 6483/10712 [1:14:00<34:42,  2.03it/s]
 61%|██████    | 6484/10712 [1:14:01<34:37,  2.04it/s]
 61%|██████    | 6485/10712 [1:14:01<34:34,  2.04it/s]
 61%|██████    | 6486/10712 [1:14:02<34:36,  2.04it/s]
 61%|██████    | 6487/10712 [1:14:02<34:33,  2.04it/s]
 61%|██████    | 6488/10712 [1:14:03<34:32,  2.04it/s]
 61%|██████    | 6489/10712 [1:14:03<34:32,  2.04it/s]
 61%|██████    | 6490/10712 [1:14:04<34:33,  2.04it/s]
 61%|██████    | 6491/10712 [1:14:04<34:30,  2.04it/s]
 61%|██████    | 6492/10712 [1:14:05<34:28,  2.04it/s]
 61%|██████    | 6493/10712 [1:14:05<34:27,  2.04it/s]
 61%|██████    | 6494/10712 [1:14:06<34:26,  2.04it/s]
 61%|██████    | 6495/10712 [1:14:06<34:26,  2.04it/s]
 61%|██████    | 6496/10712 [1:14:07<34:25,  2.04it/s]
 61%|██████    | 6497/10712 [1:14:07<34:23,  2.04it/s]
 61%|██████    | 6498/10712 [1:14:08<34:25,  2.04it/s]
 61%|██████    | 6499/10712 [1:14:08<34:25,  2.04it/s]
 61%|██████    | 6500/10712 [1:14:09<34:23,  2.04it/s]
                                                      
{'loss': 3.7048, 'grad_norm': 0.21032579243183136, 'learning_rate': 0.000401576015185957, 'epoch': 0.61}

 61%|██████    | 6500/10712 [1:14:09<34:23,  2.04it/s]
 61%|██████    | 6501/10712 [1:14:09<34:25,  2.04it/s]
 61%|██████    | 6502/10712 [1:14:10<34:25,  2.04it/s]
 61%|██████    | 6503/10712 [1:14:10<34:23,  2.04it/s]
 61%|██████    | 6504/10712 [1:14:10<34:22,  2.04it/s]
 61%|██████    | 6505/10712 [1:14:11<34:23,  2.04it/s]
 61%|██████    | 6506/10712 [1:14:11<34:23,  2.04it/s]
 61%|██████    | 6507/10712 [1:14:12<34:21,  2.04it/s]
 61%|██████    | 6508/10712 [1:14:12<34:19,  2.04it/s]
 61%|██████    | 6509/10712 [1:14:13<34:21,  2.04it/s]
 61%|██████    | 6510/10712 [1:14:13<34:20,  2.04it/s]
 61%|██████    | 6511/10712 [1:14:14<34:19,  2.04it/s]
 61%|██████    | 6512/10712 [1:14:14<34:19,  2.04it/s]
 61%|██████    | 6513/10712 [1:14:15<34:20,  2.04it/s]
 61%|██████    | 6514/10712 [1:14:15<34:18,  2.04it/s]
 61%|██████    | 6515/10712 [1:14:16<34:18,  2.04it/s]
 61%|██████    | 6516/10712 [1:14:16<34:17,  2.04it/s]
 61%|██████    | 6517/10712 [1:14:17<34:15,  2.04it/s]
 61%|██████    | 6518/10712 [1:14:17<34:13,  2.04it/s]
 61%|██████    | 6519/10712 [1:14:18<34:13,  2.04it/s]
 61%|██████    | 6520/10712 [1:14:18<34:15,  2.04it/s]
 61%|██████    | 6521/10712 [1:14:19<34:14,  2.04it/s]
 61%|██████    | 6522/10712 [1:14:19<34:11,  2.04it/s]
 61%|██████    | 6523/10712 [1:14:20<34:12,  2.04it/s]
 61%|██████    | 6524/10712 [1:14:20<34:14,  2.04it/s]
 61%|██████    | 6525/10712 [1:14:21<34:13,  2.04it/s]{'loss': 3.6954, 'grad_norm': 0.19803152978420258, 'learning_rate': 0.0003975853888088237, 'epoch': 0.61}

                                                      

 61%|██████    | 6525/10712 [1:14:21<34:13,  2.04it/s]
 61%|██████    | 6526/10712 [1:14:21<34:13,  2.04it/s]
 61%|██████    | 6527/10712 [1:14:22<34:15,  2.04it/s]
 61%|██████    | 6528/10712 [1:14:22<34:10,  2.04it/s]
 61%|██████    | 6529/10712 [1:14:23<34:09,  2.04it/s]
 61%|██████    | 6530/10712 [1:14:23<34:09,  2.04it/s]
 61%|██████    | 6531/10712 [1:14:24<34:09,  2.04it/s]
 61%|██████    | 6532/10712 [1:14:24<34:08,  2.04it/s]
 61%|██████    | 6533/10712 [1:14:25<34:05,  2.04it/s]
 61%|██████    | 6534/10712 [1:14:25<34:05,  2.04it/s]
 61%|██████    | 6535/10712 [1:14:26<34:09,  2.04it/s]
 61%|██████    | 6536/10712 [1:14:26<34:04,  2.04it/s]
 61%|██████    | 6537/10712 [1:14:27<34:03,  2.04it/s]
 61%|██████    | 6538/10712 [1:14:27<34:03,  2.04it/s]
 61%|██████    | 6539/10712 [1:14:28<34:03,  2.04it/s]
 61%|██████    | 6540/10712 [1:14:28<34:05,  2.04it/s]
 61%|██████    | 6541/10712 [1:14:29<34:05,  2.04it/s]
 61%|██████    | 6542/10712 [1:14:29<34:03,  2.04it/s]
 61%|██████    | 6543/10712 [1:14:30<34:04,  2.04it/s]
 61%|██████    | 6544/10712 [1:14:30<34:06,  2.04it/s]
 61%|██████    | 6545/10712 [1:14:31<34:03,  2.04it/s]
 61%|██████    | 6546/10712 [1:14:31<34:02,  2.04it/s]
 61%|██████    | 6547/10712 [1:14:32<33:59,  2.04it/s]
 61%|██████    | 6548/10712 [1:14:32<34:00,  2.04it/s]
 61%|██████    | 6549/10712 [1:14:33<34:30,  2.01it/s]
 61%|██████    | 6550/10712 [1:14:33<34:19,  2.02it/s]{'loss': 3.6906, 'grad_norm': 0.21223019063472748, 'learning_rate': 0.0003936015604951313, 'epoch': 0.61}

                                                      

 61%|██████    | 6550/10712 [1:14:33<34:19,  2.02it/s]
 61%|██████    | 6551/10712 [1:14:34<34:13,  2.03it/s]
 61%|██████    | 6552/10712 [1:14:34<34:09,  2.03it/s]
 61%|██████    | 6553/10712 [1:14:35<34:05,  2.03it/s]
 61%|██████    | 6554/10712 [1:14:35<34:03,  2.04it/s]
 61%|██████    | 6555/10712 [1:14:36<33:59,  2.04it/s]
 61%|██████    | 6556/10712 [1:14:36<33:58,  2.04it/s]
 61%|██████    | 6557/10712 [1:14:36<33:58,  2.04it/s]
 61%|██████    | 6558/10712 [1:14:37<33:56,  2.04it/s]
 61%|██████    | 6559/10712 [1:14:37<33:53,  2.04it/s]
 61%|██████    | 6560/10712 [1:14:38<33:52,  2.04it/s]
 61%|██████    | 6561/10712 [1:14:38<33:54,  2.04it/s]
 61%|██████▏   | 6562/10712 [1:14:39<33:53,  2.04it/s]
 61%|██████▏   | 6563/10712 [1:14:39<33:50,  2.04it/s]
 61%|██████▏   | 6564/10712 [1:14:40<33:49,  2.04it/s]
 61%|██████▏   | 6565/10712 [1:14:40<34:20,  2.01it/s]
 61%|██████▏   | 6566/10712 [1:14:41<34:09,  2.02it/s]
 61%|██████▏   | 6567/10712 [1:14:41<34:01,  2.03it/s]
 61%|██████▏   | 6568/10712 [1:14:42<33:57,  2.03it/s]
 61%|██████▏   | 6569/10712 [1:14:42<33:55,  2.03it/s]
 61%|██████▏   | 6570/10712 [1:14:43<33:52,  2.04it/s]
 61%|██████▏   | 6571/10712 [1:14:43<33:49,  2.04it/s]
 61%|██████▏   | 6572/10712 [1:14:44<33:49,  2.04it/s]
 61%|██████▏   | 6573/10712 [1:14:44<33:49,  2.04it/s]
 61%|██████▏   | 6574/10712 [1:14:45<33:48,  2.04it/s]
 61%|██████▏   | 6575/10712 [1:14:45<33:45,  2.04it/s]
                                                      {'loss': 3.6855, 'grad_norm': 0.2034957855939865, 'learning_rate': 0.00038962479468290583, 'epoch': 0.61}


 61%|██████▏   | 6575/10712 [1:14:45<33:45,  2.04it/s]
 61%|██████▏   | 6576/10712 [1:14:46<33:48,  2.04it/s]
 61%|██████▏   | 6577/10712 [1:14:46<33:48,  2.04it/s]
 61%|██████▏   | 6578/10712 [1:14:47<33:45,  2.04it/s]
 61%|██████▏   | 6579/10712 [1:14:47<33:44,  2.04it/s]
 61%|██████▏   | 6580/10712 [1:14:48<33:46,  2.04it/s]
 61%|██████▏   | 6581/10712 [1:14:48<33:44,  2.04it/s]
 61%|██████▏   | 6582/10712 [1:14:49<33:44,  2.04it/s]
 61%|██████▏   | 6583/10712 [1:14:49<33:43,  2.04it/s]
 61%|██████▏   | 6584/10712 [1:14:50<33:44,  2.04it/s]
 61%|██████▏   | 6585/10712 [1:14:50<33:42,  2.04it/s]
 61%|██████▏   | 6586/10712 [1:14:51<33:43,  2.04it/s]
 61%|██████▏   | 6587/10712 [1:14:51<33:44,  2.04it/s]
 62%|██████▏   | 6588/10712 [1:14:52<33:42,  2.04it/s]
 62%|██████▏   | 6589/10712 [1:14:52<33:40,  2.04it/s]
 62%|██████▏   | 6590/10712 [1:14:53<33:37,  2.04it/s]
 62%|██████▏   | 6591/10712 [1:14:53<33:39,  2.04it/s]
 62%|██████▏   | 6592/10712 [1:14:54<33:38,  2.04it/s]
 62%|██████▏   | 6593/10712 [1:14:54<33:36,  2.04it/s]
 62%|██████▏   | 6594/10712 [1:14:55<33:35,  2.04it/s]
 62%|██████▏   | 6595/10712 [1:14:55<33:35,  2.04it/s]
 62%|██████▏   | 6596/10712 [1:14:56<33:36,  2.04it/s]
 62%|██████▏   | 6597/10712 [1:14:56<33:34,  2.04it/s]
 62%|██████▏   | 6598/10712 [1:14:57<33:34,  2.04it/s]
 62%|██████▏   | 6599/10712 [1:14:57<33:37,  2.04it/s]
 62%|██████▏   | 6600/10712 [1:14:58<33:35,  2.04it/s]{'loss': 3.6995, 'grad_norm': 0.20717182755470276, 'learning_rate': 0.0003856553553413795, 'epoch': 0.62}

                                                      

 62%|██████▏   | 6600/10712 [1:14:58<33:35,  2.04it/s]
 62%|██████▏   | 6601/10712 [1:14:58<33:40,  2.03it/s]
 62%|██████▏   | 6602/10712 [1:14:59<33:39,  2.03it/s]
 62%|██████▏   | 6603/10712 [1:14:59<33:37,  2.04it/s]
 62%|██████▏   | 6604/10712 [1:15:00<33:35,  2.04it/s]
 62%|██████▏   | 6605/10712 [1:15:00<33:34,  2.04it/s]
 62%|██████▏   | 6606/10712 [1:15:01<33:34,  2.04it/s]
 62%|██████▏   | 6607/10712 [1:15:01<33:33,  2.04it/s]
 62%|██████▏   | 6608/10712 [1:15:01<33:33,  2.04it/s]
 62%|██████▏   | 6609/10712 [1:15:02<33:31,  2.04it/s]
 62%|██████▏   | 6610/10712 [1:15:02<33:30,  2.04it/s]
 62%|██████▏   | 6611/10712 [1:15:03<33:31,  2.04it/s]
 62%|██████▏   | 6612/10712 [1:15:03<33:30,  2.04it/s]
 62%|██████▏   | 6613/10712 [1:15:04<33:28,  2.04it/s]
 62%|██████▏   | 6614/10712 [1:15:04<33:27,  2.04it/s]
 62%|██████▏   | 6615/10712 [1:15:05<33:26,  2.04it/s]
 62%|██████▏   | 6616/10712 [1:15:05<33:28,  2.04it/s]
 62%|██████▏   | 6617/10712 [1:15:06<33:26,  2.04it/s]
 62%|██████▏   | 6618/10712 [1:15:06<33:25,  2.04it/s]
 62%|██████▏   | 6619/10712 [1:15:07<33:23,  2.04it/s]
 62%|██████▏   | 6620/10712 [1:15:07<33:23,  2.04it/s]
 62%|██████▏   | 6621/10712 [1:15:08<33:25,  2.04it/s]
 62%|██████▏   | 6622/10712 [1:15:08<33:21,  2.04it/s]
 62%|██████▏   | 6623/10712 [1:15:09<33:20,  2.04it/s]
 62%|██████▏   | 6624/10712 [1:15:09<33:22,  2.04it/s]
 62%|██████▏   | 6625/10712 [1:15:10<33:22,  2.04it/s]
                                                      
{'loss': 3.699, 'grad_norm': 0.21413753926753998, 'learning_rate': 0.00038169350595346896, 'epoch': 0.62}

 62%|██████▏   | 6625/10712 [1:15:10<33:22,  2.04it/s]
 62%|██████▏   | 6626/10712 [1:15:10<33:23,  2.04it/s]
 62%|██████▏   | 6627/10712 [1:15:11<33:22,  2.04it/s]
 62%|██████▏   | 6628/10712 [1:15:11<33:20,  2.04it/s]
 62%|██████▏   | 6629/10712 [1:15:12<33:20,  2.04it/s]
 62%|██████▏   | 6630/10712 [1:15:12<33:18,  2.04it/s]
 62%|██████▏   | 6631/10712 [1:15:13<33:19,  2.04it/s]
 62%|██████▏   | 6632/10712 [1:15:13<33:20,  2.04it/s]
 62%|██████▏   | 6633/10712 [1:15:14<33:21,  2.04it/s]
 62%|██████▏   | 6634/10712 [1:15:14<33:18,  2.04it/s]
 62%|██████▏   | 6635/10712 [1:15:15<33:16,  2.04it/s]
 62%|██████▏   | 6636/10712 [1:15:15<33:18,  2.04it/s]
 62%|██████▏   | 6637/10712 [1:15:16<33:16,  2.04it/s]
 62%|██████▏   | 6638/10712 [1:15:16<33:15,  2.04it/s]
 62%|██████▏   | 6639/10712 [1:15:17<33:14,  2.04it/s]
 62%|██████▏   | 6640/10712 [1:15:17<33:15,  2.04it/s]
 62%|██████▏   | 6641/10712 [1:15:18<33:16,  2.04it/s]
 62%|██████▏   | 6642/10712 [1:15:18<33:13,  2.04it/s]
 62%|██████▏   | 6643/10712 [1:15:19<33:12,  2.04it/s]
 62%|██████▏   | 6644/10712 [1:15:19<33:13,  2.04it/s]
 62%|██████▏   | 6645/10712 [1:15:20<33:12,  2.04it/s]
 62%|██████▏   | 6646/10712 [1:15:20<33:12,  2.04it/s]
 62%|██████▏   | 6647/10712 [1:15:21<33:11,  2.04it/s]
 62%|██████▏   | 6648/10712 [1:15:21<33:14,  2.04it/s]
 62%|██████▏   | 6649/10712 [1:15:22<33:12,  2.04it/s]
 62%|██████▏   | 6650/10712 [1:15:22<33:10,  2.04it/s]{'loss': 3.6905, 'grad_norm': 0.2276364415884018, 'learning_rate': 0.00037773950949828583, 'epoch': 0.62}
                                                      

 62%|██████▏   | 6650/10712 [1:15:22<33:10,  2.04it/s]
 62%|██████▏   | 6651/10712 [1:15:23<33:15,  2.04it/s]
 62%|██████▏   | 6652/10712 [1:15:23<33:15,  2.03it/s]
 62%|██████▏   | 6653/10712 [1:15:24<33:13,  2.04it/s]
 62%|██████▏   | 6654/10712 [1:15:24<33:11,  2.04it/s]
 62%|██████▏   | 6655/10712 [1:15:25<33:11,  2.04it/s]
 62%|██████▏   | 6656/10712 [1:15:25<33:09,  2.04it/s]
 62%|██████▏   | 6657/10712 [1:15:26<33:08,  2.04it/s]
 62%|██████▏   | 6658/10712 [1:15:26<33:10,  2.04it/s]
 62%|██████▏   | 6659/10712 [1:15:26<33:07,  2.04it/s]
 62%|██████▏   | 6660/10712 [1:15:27<33:08,  2.04it/s]
 62%|██████▏   | 6661/10712 [1:15:27<33:05,  2.04it/s]
 62%|██████▏   | 6662/10712 [1:15:28<33:04,  2.04it/s]
 62%|██████▏   | 6663/10712 [1:15:28<33:02,  2.04it/s]
 62%|██████▏   | 6664/10712 [1:15:29<33:03,  2.04it/s]
 62%|██████▏   | 6665/10712 [1:15:29<33:04,  2.04it/s]
 62%|██████▏   | 6666/10712 [1:15:30<33:01,  2.04it/s]
 62%|██████▏   | 6667/10712 [1:15:30<33:00,  2.04it/s]
 62%|██████▏   | 6668/10712 [1:15:31<33:00,  2.04it/s]
 62%|██████▏   | 6669/10712 [1:15:31<33:00,  2.04it/s]
 62%|██████▏   | 6670/10712 [1:15:32<32:59,  2.04it/s]
 62%|██████▏   | 6671/10712 [1:15:32<32:56,  2.04it/s]
 62%|██████▏   | 6672/10712 [1:15:33<32:57,  2.04it/s]
 62%|██████▏   | 6673/10712 [1:15:33<32:59,  2.04it/s]
 62%|██████▏   | 6674/10712 [1:15:34<32:57,  2.04it/s]
 62%|██████▏   | 6675/10712 [1:15:34<32:57,  2.04it/s]{'loss': 3.6938, 'grad_norm': 0.20666924118995667, 'learning_rate': 0.00037379362843368126, 'epoch': 0.62}

                                                      

 62%|██████▏   | 6675/10712 [1:15:34<32:57,  2.04it/s]
 62%|██████▏   | 6676/10712 [1:15:35<33:01,  2.04it/s]
 62%|██████▏   | 6677/10712 [1:15:35<33:02,  2.04it/s]
 62%|██████▏   | 6678/10712 [1:15:36<33:00,  2.04it/s]
 62%|██████▏   | 6679/10712 [1:15:36<32:57,  2.04it/s]
 62%|██████▏   | 6680/10712 [1:15:37<32:56,  2.04it/s]
 62%|██████▏   | 6681/10712 [1:15:37<32:55,  2.04it/s]
 62%|██████▏   | 6682/10712 [1:15:38<32:56,  2.04it/s]
 62%|██████▏   | 6683/10712 [1:15:38<32:54,  2.04it/s]
 62%|██████▏   | 6684/10712 [1:15:39<32:53,  2.04it/s]
 62%|██████▏   | 6685/10712 [1:15:39<32:52,  2.04it/s]
 62%|██████▏   | 6686/10712 [1:15:40<32:51,  2.04it/s]
 62%|██████▏   | 6687/10712 [1:15:40<32:52,  2.04it/s]
 62%|██████▏   | 6688/10712 [1:15:41<32:51,  2.04it/s]
 62%|██████▏   | 6689/10712 [1:15:41<32:48,  2.04it/s]
 62%|██████▏   | 6690/10712 [1:15:42<32:47,  2.04it/s]
 62%|██████▏   | 6691/10712 [1:15:42<32:46,  2.04it/s]
 62%|██████▏   | 6692/10712 [1:15:43<38:12,  1.75it/s]
 62%|██████▏   | 6693/10712 [1:15:43<36:33,  1.83it/s]
 62%|██████▏   | 6694/10712 [1:15:44<35:27,  1.89it/s]
 62%|██████▎   | 6695/10712 [1:15:44<34:39,  1.93it/s]
 63%|██████▎   | 6696/10712 [1:15:45<34:04,  1.96it/s]
 63%|██████▎   | 6697/10712 [1:15:45<33:41,  1.99it/s]
 63%|██████▎   | 6698/10712 [1:15:46<33:26,  2.00it/s]
 63%|██████▎   | 6699/10712 [1:15:46<33:13,  2.01it/s]
 63%|██████▎   | 6700/10712 [1:15:47<33:02,  2.02it/s]{'loss': 3.6886, 'grad_norm': 0.20668712258338928, 'learning_rate': 0.0003698561246788237, 'epoch': 0.63}
                                                      

 63%|██████▎   | 6700/10712 [1:15:47<33:02,  2.02it/s]
 63%|██████▎   | 6701/10712 [1:15:47<32:59,  2.03it/s]
 63%|██████▎   | 6702/10712 [1:15:48<32:55,  2.03it/s]
 63%|██████▎   | 6703/10712 [1:15:48<32:50,  2.03it/s]
 63%|██████▎   | 6704/10712 [1:15:49<32:46,  2.04it/s]
 63%|██████▎   | 6705/10712 [1:15:49<32:46,  2.04it/s]
 63%|██████▎   | 6706/10712 [1:15:50<32:46,  2.04it/s]
 63%|██████▎   | 6707/10712 [1:15:50<32:43,  2.04it/s]
 63%|██████▎   | 6708/10712 [1:15:51<32:42,  2.04it/s]
 63%|██████▎   | 6709/10712 [1:15:51<32:43,  2.04it/s]
 63%|██████▎   | 6710/10712 [1:15:52<32:43,  2.04it/s]
 63%|██████▎   | 6711/10712 [1:15:52<32:41,  2.04it/s]
 63%|██████▎   | 6712/10712 [1:15:53<32:38,  2.04it/s]
 63%|██████▎   | 6713/10712 [1:15:53<32:41,  2.04it/s]
 63%|██████▎   | 6714/10712 [1:15:54<32:38,  2.04it/s]
 63%|██████▎   | 6715/10712 [1:15:54<32:38,  2.04it/s]
 63%|██████▎   | 6716/10712 [1:15:55<32:38,  2.04it/s]
 63%|██████▎   | 6717/10712 [1:15:55<32:38,  2.04it/s]
 63%|██████▎   | 6718/10712 [1:15:56<32:36,  2.04it/s]
 63%|██████▎   | 6719/10712 [1:15:56<32:35,  2.04it/s]
 63%|██████▎   | 6720/10712 [1:15:57<32:36,  2.04it/s]
 63%|██████▎   | 6721/10712 [1:15:57<32:35,  2.04it/s]
 63%|██████▎   | 6722/10712 [1:15:58<32:34,  2.04it/s]
 63%|██████▎   | 6723/10712 [1:15:58<32:34,  2.04it/s]
 63%|██████▎   | 6724/10712 [1:15:59<32:33,  2.04it/s]
 63%|██████▎   | 6725/10712 [1:15:59<32:33,  2.04it/s]
                                                      
{'loss': 3.684, 'grad_norm': 0.2053668349981308, 'learning_rate': 0.0003659272595968135, 'epoch': 0.63}

 63%|██████▎   | 6725/10712 [1:15:59<32:33,  2.04it/s]
 63%|██████▎   | 6726/10712 [1:16:00<32:33,  2.04it/s]
 63%|██████▎   | 6727/10712 [1:16:00<32:34,  2.04it/s]
 63%|██████▎   | 6728/10712 [1:16:01<37:50,  1.76it/s]
 63%|██████▎   | 6729/10712 [1:16:01<36:13,  1.83it/s]
 63%|██████▎   | 6730/10712 [1:16:02<35:07,  1.89it/s]
 63%|██████▎   | 6731/10712 [1:16:02<34:19,  1.93it/s]
 63%|██████▎   | 6732/10712 [1:16:03<33:45,  1.97it/s]
 63%|██████▎   | 6733/10712 [1:16:03<33:20,  1.99it/s]
 63%|██████▎   | 6734/10712 [1:16:04<33:10,  2.00it/s]
 63%|██████▎   | 6735/10712 [1:16:04<32:57,  2.01it/s]
 63%|██████▎   | 6736/10712 [1:16:05<32:47,  2.02it/s]
 63%|██████▎   | 6737/10712 [1:16:05<32:41,  2.03it/s]
 63%|██████▎   | 6738/10712 [1:16:06<32:37,  2.03it/s]
 63%|██████▎   | 6739/10712 [1:16:06<32:34,  2.03it/s]
 63%|██████▎   | 6740/10712 [1:16:07<32:27,  2.04it/s]
 63%|██████▎   | 6741/10712 [1:16:07<32:25,  2.04it/s]
 63%|██████▎   | 6742/10712 [1:16:08<32:26,  2.04it/s]
 63%|██████▎   | 6743/10712 [1:16:08<32:25,  2.04it/s]
 63%|██████▎   | 6744/10712 [1:16:09<32:24,  2.04it/s]
 63%|██████▎   | 6745/10712 [1:16:09<32:22,  2.04it/s]
 63%|██████▎   | 6746/10712 [1:16:10<32:22,  2.04it/s]
 63%|██████▎   | 6747/10712 [1:16:10<32:21,  2.04it/s]
 63%|██████▎   | 6748/10712 [1:16:11<32:21,  2.04it/s]
 63%|██████▎   | 6749/10712 [1:16:11<32:21,  2.04it/s]
 63%|██████▎   | 6750/10712 [1:16:12<32:21,  2.04it/s]
                                                      
{'loss': 3.6871, 'grad_norm': 0.21559420228004456, 'learning_rate': 0.0003620072939773347, 'epoch': 0.63}

 63%|██████▎   | 6750/10712 [1:16:12<32:21,  2.04it/s]
 63%|██████▎   | 6751/10712 [1:16:12<32:23,  2.04it/s]
 63%|██████▎   | 6752/10712 [1:16:13<32:21,  2.04it/s]
 63%|██████▎   | 6753/10712 [1:16:13<32:22,  2.04it/s]
 63%|██████▎   | 6754/10712 [1:16:14<32:19,  2.04it/s]
 63%|██████▎   | 6755/10712 [1:16:14<32:17,  2.04it/s]
 63%|██████▎   | 6756/10712 [1:16:15<32:23,  2.04it/s]
 63%|██████▎   | 6757/10712 [1:16:15<32:20,  2.04it/s]
 63%|██████▎   | 6758/10712 [1:16:16<32:17,  2.04it/s]
 63%|██████▎   | 6759/10712 [1:16:16<32:14,  2.04it/s]
 63%|██████▎   | 6760/10712 [1:16:17<32:15,  2.04it/s]
 63%|██████▎   | 6761/10712 [1:16:17<32:16,  2.04it/s]
 63%|██████▎   | 6762/10712 [1:16:18<32:14,  2.04it/s]
 63%|██████▎   | 6763/10712 [1:16:18<32:13,  2.04it/s]
 63%|██████▎   | 6764/10712 [1:16:18<32:12,  2.04it/s]
 63%|██████▎   | 6765/10712 [1:16:19<32:15,  2.04it/s]
 63%|██████▎   | 6766/10712 [1:16:19<32:15,  2.04it/s]
 63%|██████▎   | 6767/10712 [1:16:20<32:14,  2.04it/s]
 63%|██████▎   | 6768/10712 [1:16:20<32:15,  2.04it/s]
 63%|██████▎   | 6769/10712 [1:16:21<32:14,  2.04it/s]
 63%|██████▎   | 6770/10712 [1:16:21<32:13,  2.04it/s]
 63%|██████▎   | 6771/10712 [1:16:22<32:11,  2.04it/s]
 63%|██████▎   | 6772/10712 [1:16:22<32:10,  2.04it/s]
 63%|██████▎   | 6773/10712 [1:16:23<32:11,  2.04it/s]
 63%|██████▎   | 6774/10712 [1:16:23<32:08,  2.04it/s]
 63%|██████▎   | 6775/10712 [1:16:24<32:10,  2.04it/s]{'loss': 3.6853, 'grad_norm': 0.2021356076002121, 'learning_rate': 0.0003580964880193439, 'epoch': 0.63}
                                                      

 63%|██████▎   | 6775/10712 [1:16:24<32:10,  2.04it/s]
 63%|██████▎   | 6776/10712 [1:16:24<32:12,  2.04it/s]
 63%|██████▎   | 6777/10712 [1:16:25<32:10,  2.04it/s]
 63%|██████▎   | 6778/10712 [1:16:25<32:07,  2.04it/s]
 63%|██████▎   | 6779/10712 [1:16:26<32:08,  2.04it/s]
 63%|██████▎   | 6780/10712 [1:16:26<32:08,  2.04it/s]
 63%|██████▎   | 6781/10712 [1:16:27<32:05,  2.04it/s]
 63%|██████▎   | 6782/10712 [1:16:27<32:04,  2.04it/s]
 63%|██████▎   | 6783/10712 [1:16:28<32:03,  2.04it/s]
 63%|██████▎   | 6784/10712 [1:16:28<32:05,  2.04it/s]
 63%|██████▎   | 6785/10712 [1:16:29<32:04,  2.04it/s]
 63%|██████▎   | 6786/10712 [1:16:29<32:24,  2.02it/s]
 63%|██████▎   | 6787/10712 [1:16:30<32:18,  2.02it/s]
 63%|██████▎   | 6788/10712 [1:16:30<32:13,  2.03it/s]
 63%|██████▎   | 6789/10712 [1:16:31<32:07,  2.04it/s]
 63%|██████▎   | 6790/10712 [1:16:31<32:06,  2.04it/s]
 63%|██████▎   | 6791/10712 [1:16:32<32:05,  2.04it/s]
 63%|██████▎   | 6792/10712 [1:16:32<32:05,  2.04it/s]
 63%|██████▎   | 6793/10712 [1:16:33<32:01,  2.04it/s]
 63%|██████▎   | 6794/10712 [1:16:33<31:58,  2.04it/s]
 63%|██████▎   | 6795/10712 [1:16:34<31:59,  2.04it/s]
 63%|██████▎   | 6796/10712 [1:16:34<31:59,  2.04it/s]
 63%|██████▎   | 6797/10712 [1:16:35<31:57,  2.04it/s]
 63%|██████▎   | 6798/10712 [1:16:35<31:58,  2.04it/s]
 63%|██████▎   | 6799/10712 [1:16:36<31:59,  2.04it/s]
 63%|██████▎   | 6800/10712 [1:16:36<31:58,  2.04it/s]{'loss': 3.6845, 'grad_norm': 0.2116902619600296, 'learning_rate': 0.00035419510131379873, 'epoch': 0.63}

                                                      

 63%|██████▎   | 6800/10712 [1:16:36<31:58,  2.04it/s]
 63%|██████▎   | 6801/10712 [1:16:37<31:58,  2.04it/s]
 63%|██████▎   | 6802/10712 [1:16:37<31:58,  2.04it/s]
 64%|██████▎   | 6803/10712 [1:16:38<31:58,  2.04it/s]
 64%|██████▎   | 6804/10712 [1:16:38<31:57,  2.04it/s]
 64%|██████▎   | 6805/10712 [1:16:39<31:54,  2.04it/s]
 64%|██████▎   | 6806/10712 [1:16:39<31:54,  2.04it/s]
 64%|██████▎   | 6807/10712 [1:16:40<31:54,  2.04it/s]
 64%|██████▎   | 6808/10712 [1:16:40<31:50,  2.04it/s]
 64%|██████▎   | 6809/10712 [1:16:41<31:49,  2.04it/s]
 64%|██████▎   | 6810/10712 [1:16:41<31:49,  2.04it/s]
 64%|██████▎   | 6811/10712 [1:16:42<31:49,  2.04it/s]
 64%|██████▎   | 6812/10712 [1:16:42<31:48,  2.04it/s]
 64%|██████▎   | 6813/10712 [1:16:43<31:46,  2.04it/s]
 64%|██████▎   | 6814/10712 [1:16:43<31:45,  2.05it/s]
 64%|██████▎   | 6815/10712 [1:16:43<31:46,  2.04it/s]
 64%|██████▎   | 6816/10712 [1:16:44<31:45,  2.04it/s]
 64%|██████▎   | 6817/10712 [1:16:44<31:48,  2.04it/s]
 64%|██████▎   | 6818/10712 [1:16:45<31:47,  2.04it/s]
 64%|██████▎   | 6819/10712 [1:16:45<31:46,  2.04it/s]
 64%|██████▎   | 6820/10712 [1:16:46<31:45,  2.04it/s]
 64%|██████▎   | 6821/10712 [1:16:46<31:45,  2.04it/s]
 64%|██████▎   | 6822/10712 [1:16:47<31:45,  2.04it/s]
 64%|██████▎   | 6823/10712 [1:16:47<31:42,  2.04it/s]
 64%|██████▎   | 6824/10712 [1:16:48<31:41,  2.04it/s]
 64%|██████▎   | 6825/10712 [1:16:48<31:41,  2.04it/s]
                                                      
{'loss': 3.6874, 'grad_norm': 0.20325392484664917, 'learning_rate': 0.0003503033928264272, 'epoch': 0.64}

 64%|██████▎   | 6825/10712 [1:16:48<31:41,  2.04it/s]
 64%|██████▎   | 6826/10712 [1:16:49<31:46,  2.04it/s]
 64%|██████▎   | 6827/10712 [1:16:49<31:46,  2.04it/s]
 64%|██████▎   | 6828/10712 [1:16:50<31:42,  2.04it/s]
 64%|██████▍   | 6829/10712 [1:16:50<31:41,  2.04it/s]
 64%|██████▍   | 6830/10712 [1:16:51<31:42,  2.04it/s]
 64%|██████▍   | 6831/10712 [1:16:51<31:41,  2.04it/s]
 64%|██████▍   | 6832/10712 [1:16:52<31:40,  2.04it/s]
 64%|██████▍   | 6833/10712 [1:16:52<31:38,  2.04it/s]
 64%|██████▍   | 6834/10712 [1:16:53<31:39,  2.04it/s]
 64%|██████▍   | 6835/10712 [1:16:53<31:40,  2.04it/s]
 64%|██████▍   | 6836/10712 [1:16:54<31:39,  2.04it/s]
 64%|██████▍   | 6837/10712 [1:16:54<31:38,  2.04it/s]
 64%|██████▍   | 6838/10712 [1:16:55<31:39,  2.04it/s]
 64%|██████▍   | 6839/10712 [1:16:55<31:39,  2.04it/s]
 64%|██████▍   | 6840/10712 [1:16:56<31:35,  2.04it/s]
 64%|██████▍   | 6841/10712 [1:16:56<31:36,  2.04it/s]
 64%|██████▍   | 6842/10712 [1:16:57<31:37,  2.04it/s]
 64%|██████▍   | 6843/10712 [1:16:57<31:36,  2.04it/s]
 64%|██████▍   | 6844/10712 [1:16:58<31:35,  2.04it/s]
 64%|██████▍   | 6845/10712 [1:16:58<31:34,  2.04it/s]
 64%|██████▍   | 6846/10712 [1:16:59<31:35,  2.04it/s]
 64%|██████▍   | 6847/10712 [1:16:59<31:33,  2.04it/s]
 64%|██████▍   | 6848/10712 [1:17:00<31:33,  2.04it/s]
 64%|██████▍   | 6849/10712 [1:17:00<31:34,  2.04it/s]
 64%|██████▍   | 6850/10712 [1:17:01<31:32,  2.04it/s]
                                                      {'loss': 3.6863, 'grad_norm': 0.21490880846977234, 'learning_rate': 0.00034642162088053754, 'epoch': 0.64}

 64%|██████▍   | 6850/10712 [1:17:01<31:32,  2.04it/s]
 64%|██████▍   | 6851/10712 [1:17:01<31:32,  2.04it/s]
 64%|██████▍   | 6852/10712 [1:17:02<31:31,  2.04it/s]
 64%|██████▍   | 6853/10712 [1:17:02<31:31,  2.04it/s]
 64%|██████▍   | 6854/10712 [1:17:03<31:29,  2.04it/s]
 64%|██████▍   | 6855/10712 [1:17:03<31:27,  2.04it/s]
 64%|██████▍   | 6856/10712 [1:17:04<31:27,  2.04it/s]
 64%|██████▍   | 6857/10712 [1:17:04<31:28,  2.04it/s]
 64%|██████▍   | 6858/10712 [1:17:05<31:27,  2.04it/s]
 64%|██████▍   | 6859/10712 [1:17:05<31:26,  2.04it/s]
 64%|██████▍   | 6860/10712 [1:17:06<31:25,  2.04it/s]
 64%|██████▍   | 6861/10712 [1:17:06<31:27,  2.04it/s]
 64%|██████▍   | 6862/10712 [1:17:07<31:26,  2.04it/s]
 64%|██████▍   | 6863/10712 [1:17:07<31:25,  2.04it/s]
 64%|██████▍   | 6864/10712 [1:17:07<31:26,  2.04it/s]
 64%|██████▍   | 6865/10712 [1:17:08<31:25,  2.04it/s]
 64%|██████▍   | 6866/10712 [1:17:08<31:22,  2.04it/s]
 64%|██████▍   | 6867/10712 [1:17:09<31:24,  2.04it/s]
 64%|██████▍   | 6868/10712 [1:17:09<31:26,  2.04it/s]
 64%|██████▍   | 6869/10712 [1:17:10<31:25,  2.04it/s]
 64%|██████▍   | 6870/10712 [1:17:10<31:24,  2.04it/s]
 64%|██████▍   | 6871/10712 [1:17:11<31:24,  2.04it/s]
 64%|██████▍   | 6872/10712 [1:17:11<31:23,  2.04it/s]
 64%|██████▍   | 6873/10712 [1:17:12<31:20,  2.04it/s]
 64%|██████▍   | 6874/10712 [1:17:12<31:19,  2.04it/s]
 64%|██████▍   | 6875/10712 [1:17:13<31:19,  2.04it/s]{'loss': 3.6785, 'grad_norm': 0.20367369055747986, 'learning_rate': 0.0003425500431398719, 'epoch': 0.64}

                                                      

 64%|██████▍   | 6875/10712 [1:17:13<31:19,  2.04it/s]
 64%|██████▍   | 6876/10712 [1:17:13<31:19,  2.04it/s]
 64%|██████▍   | 6877/10712 [1:17:14<31:17,  2.04it/s]
 64%|██████▍   | 6878/10712 [1:17:14<31:19,  2.04it/s]
 64%|██████▍   | 6879/10712 [1:17:15<31:19,  2.04it/s]
 64%|██████▍   | 6880/10712 [1:17:15<31:17,  2.04it/s]
 64%|██████▍   | 6881/10712 [1:17:16<31:15,  2.04it/s]
 64%|██████▍   | 6882/10712 [1:17:16<31:14,  2.04it/s]
 64%|██████▍   | 6883/10712 [1:17:17<31:16,  2.04it/s]
 64%|██████▍   | 6884/10712 [1:17:17<31:14,  2.04it/s]
 64%|██████▍   | 6885/10712 [1:17:18<31:13,  2.04it/s]
 64%|██████▍   | 6886/10712 [1:17:18<31:16,  2.04it/s]
 64%|██████▍   | 6887/10712 [1:17:19<31:13,  2.04it/s]
 64%|██████▍   | 6888/10712 [1:17:19<31:11,  2.04it/s]
 64%|██████▍   | 6889/10712 [1:17:20<31:11,  2.04it/s]
 64%|██████▍   | 6890/10712 [1:17:20<31:12,  2.04it/s]
 64%|██████▍   | 6891/10712 [1:17:21<31:11,  2.04it/s]
 64%|██████▍   | 6892/10712 [1:17:21<31:10,  2.04it/s]
 64%|██████▍   | 6893/10712 [1:17:22<31:10,  2.04it/s]
 64%|██████▍   | 6894/10712 [1:17:22<31:11,  2.04it/s]
 64%|██████▍   | 6895/10712 [1:17:23<31:11,  2.04it/s]
 64%|██████▍   | 6896/10712 [1:17:23<31:09,  2.04it/s]
 64%|██████▍   | 6897/10712 [1:17:24<31:09,  2.04it/s]
 64%|██████▍   | 6898/10712 [1:17:24<31:10,  2.04it/s]
 64%|██████▍   | 6899/10712 [1:17:25<31:10,  2.04it/s]
 64%|██████▍   | 6900/10712 [1:17:25<31:06,  2.04it/s]{'loss': 3.6722, 'grad_norm': 0.20835915207862854, 'learning_rate': 0.0003386889165915025, 'epoch': 0.64}
                                                      

 64%|██████▍   | 6900/10712 [1:17:25<31:06,  2.04it/s]
 64%|██████▍   | 6901/10712 [1:17:26<31:09,  2.04it/s]
 64%|██████▍   | 6902/10712 [1:17:26<31:09,  2.04it/s]
 64%|██████▍   | 6903/10712 [1:17:27<31:07,  2.04it/s]
 64%|██████▍   | 6904/10712 [1:17:27<31:08,  2.04it/s]
 64%|██████▍   | 6905/10712 [1:17:28<31:09,  2.04it/s]
 64%|██████▍   | 6906/10712 [1:17:28<31:07,  2.04it/s]
 64%|██████▍   | 6907/10712 [1:17:29<31:03,  2.04it/s]
 64%|██████▍   | 6908/10712 [1:17:29<31:03,  2.04it/s]
 64%|██████▍   | 6909/10712 [1:17:30<31:04,  2.04it/s]
 65%|██████▍   | 6910/10712 [1:17:30<31:02,  2.04it/s]
 65%|██████▍   | 6911/10712 [1:17:31<31:03,  2.04it/s]
 65%|██████▍   | 6912/10712 [1:17:31<31:02,  2.04it/s]
 65%|██████▍   | 6913/10712 [1:17:32<31:02,  2.04it/s]
 65%|██████▍   | 6914/10712 [1:17:32<30:59,  2.04it/s]
 65%|██████▍   | 6915/10712 [1:17:32<30:57,  2.04it/s]
 65%|██████▍   | 6916/10712 [1:17:33<30:58,  2.04it/s]
 65%|██████▍   | 6917/10712 [1:17:33<30:59,  2.04it/s]
 65%|██████▍   | 6918/10712 [1:17:34<30:59,  2.04it/s]
 65%|██████▍   | 6919/10712 [1:17:34<30:56,  2.04it/s]
 65%|██████▍   | 6920/10712 [1:17:35<30:54,  2.04it/s]
 65%|██████▍   | 6921/10712 [1:17:35<30:56,  2.04it/s]
 65%|██████▍   | 6922/10712 [1:17:36<30:54,  2.04it/s]
 65%|██████▍   | 6923/10712 [1:17:36<30:52,  2.04it/s]
 65%|██████▍   | 6924/10712 [1:17:37<30:53,  2.04it/s]
 65%|██████▍   | 6925/10712 [1:17:37<30:56,  2.04it/s]{'loss': 3.6807, 'grad_norm': 0.23411470651626587, 'learning_rate': 0.00033483849752877394, 'epoch': 0.65}

                                                      

 65%|██████▍   | 6925/10712 [1:17:37<30:56,  2.04it/s]
 65%|██████▍   | 6926/10712 [1:17:38<30:56,  2.04it/s]
 65%|██████▍   | 6927/10712 [1:17:38<30:52,  2.04it/s]
 65%|██████▍   | 6928/10712 [1:17:39<30:53,  2.04it/s]
 65%|██████▍   | 6929/10712 [1:17:39<30:53,  2.04it/s]
 65%|██████▍   | 6930/10712 [1:17:40<30:51,  2.04it/s]
 65%|██████▍   | 6931/10712 [1:17:40<30:49,  2.04it/s]
 65%|██████▍   | 6932/10712 [1:17:41<30:49,  2.04it/s]
 65%|██████▍   | 6933/10712 [1:17:41<30:51,  2.04it/s]
 65%|██████▍   | 6934/10712 [1:17:42<30:51,  2.04it/s]
 65%|██████▍   | 6935/10712 [1:17:42<30:48,  2.04it/s]
 65%|██████▍   | 6936/10712 [1:17:43<30:47,  2.04it/s]
 65%|██████▍   | 6937/10712 [1:17:43<30:48,  2.04it/s]
 65%|██████▍   | 6938/10712 [1:17:44<30:48,  2.04it/s]
 65%|██████▍   | 6939/10712 [1:17:44<30:46,  2.04it/s]
 65%|██████▍   | 6940/10712 [1:17:45<30:47,  2.04it/s]
 65%|██████▍   | 6941/10712 [1:17:45<30:48,  2.04it/s]
 65%|██████▍   | 6942/10712 [1:17:46<30:46,  2.04it/s]
 65%|██████▍   | 6943/10712 [1:17:46<30:44,  2.04it/s]
 65%|██████▍   | 6944/10712 [1:17:47<30:45,  2.04it/s]
 65%|██████▍   | 6945/10712 [1:17:47<30:45,  2.04it/s]
 65%|██████▍   | 6946/10712 [1:17:48<30:43,  2.04it/s]
 65%|██████▍   | 6947/10712 [1:17:48<31:08,  2.01it/s]
 65%|██████▍   | 6948/10712 [1:17:49<31:01,  2.02it/s]
 65%|██████▍   | 6949/10712 [1:17:49<30:56,  2.03it/s]
 65%|██████▍   | 6950/10712 [1:17:50<30:52,  2.03it/s]{'loss': 3.6854, 'grad_norm': 0.20820121467113495, 'learning_rate': 0.0003309990415342904, 'epoch': 0.65}
                                                      

 65%|██████▍   | 6950/10712 [1:17:50<30:52,  2.03it/s]
 65%|██████▍   | 6951/10712 [1:17:50<30:52,  2.03it/s]
 65%|██████▍   | 6952/10712 [1:17:51<30:48,  2.03it/s]
 65%|██████▍   | 6953/10712 [1:17:51<30:47,  2.03it/s]
 65%|██████▍   | 6954/10712 [1:17:52<30:43,  2.04it/s]
 65%|██████▍   | 6955/10712 [1:17:52<30:41,  2.04it/s]
 65%|██████▍   | 6956/10712 [1:17:53<30:40,  2.04it/s]
 65%|██████▍   | 6957/10712 [1:17:53<30:41,  2.04it/s]
 65%|██████▍   | 6958/10712 [1:17:54<30:40,  2.04it/s]
 65%|██████▍   | 6959/10712 [1:17:54<30:38,  2.04it/s]
 65%|██████▍   | 6960/10712 [1:17:55<30:38,  2.04it/s]
 65%|██████▍   | 6961/10712 [1:17:55<30:38,  2.04it/s]
 65%|██████▍   | 6962/10712 [1:17:56<30:39,  2.04it/s]
 65%|██████▌   | 6963/10712 [1:17:56<30:38,  2.04it/s]
 65%|██████▌   | 6964/10712 [1:17:57<30:35,  2.04it/s]
 65%|██████▌   | 6965/10712 [1:17:57<30:36,  2.04it/s]
 65%|██████▌   | 6966/10712 [1:17:57<30:36,  2.04it/s]
 65%|██████▌   | 6967/10712 [1:17:58<30:35,  2.04it/s]
 65%|██████▌   | 6968/10712 [1:17:58<30:33,  2.04it/s]
 65%|██████▌   | 6969/10712 [1:17:59<30:31,  2.04it/s]
 65%|██████▌   | 6970/10712 [1:17:59<30:33,  2.04it/s]
 65%|██████▌   | 6971/10712 [1:18:00<30:33,  2.04it/s]
 65%|██████▌   | 6972/10712 [1:18:00<30:30,  2.04it/s]
 65%|██████▌   | 6973/10712 [1:18:01<30:29,  2.04it/s]
 65%|██████▌   | 6974/10712 [1:18:01<30:30,  2.04it/s]
 65%|██████▌   | 6975/10712 [1:18:02<30:29,  2.04it/s]{'loss': 3.6798, 'grad_norm': 0.20154866576194763, 'learning_rate': 0.0003271708034629518, 'epoch': 0.65}
                                                      

 65%|██████▌   | 6975/10712 [1:18:02<30:29,  2.04it/s]
 65%|██████▌   | 6976/10712 [1:18:02<30:30,  2.04it/s]
 65%|██████▌   | 6977/10712 [1:18:03<30:28,  2.04it/s]
 65%|██████▌   | 6978/10712 [1:18:03<30:29,  2.04it/s]
 65%|██████▌   | 6979/10712 [1:18:04<30:29,  2.04it/s]
 65%|██████▌   | 6980/10712 [1:18:04<30:27,  2.04it/s]
 65%|██████▌   | 6981/10712 [1:18:05<30:28,  2.04it/s]
 65%|██████▌   | 6982/10712 [1:18:05<30:28,  2.04it/s]
 65%|██████▌   | 6983/10712 [1:18:06<30:27,  2.04it/s]
 65%|██████▌   | 6984/10712 [1:18:06<30:26,  2.04it/s]
 65%|██████▌   | 6985/10712 [1:18:07<30:25,  2.04it/s]
 65%|██████▌   | 6986/10712 [1:18:07<30:25,  2.04it/s]
 65%|██████▌   | 6987/10712 [1:18:08<30:24,  2.04it/s]
 65%|██████▌   | 6988/10712 [1:18:08<30:24,  2.04it/s]
 65%|██████▌   | 6989/10712 [1:18:09<30:24,  2.04it/s]
 65%|██████▌   | 6990/10712 [1:18:09<30:23,  2.04it/s]
 65%|██████▌   | 6991/10712 [1:18:10<30:21,  2.04it/s]
 65%|██████▌   | 6992/10712 [1:18:10<30:22,  2.04it/s]
 65%|██████▌   | 6993/10712 [1:18:11<30:21,  2.04it/s]
 65%|██████▌   | 6994/10712 [1:18:11<30:21,  2.04it/s]
 65%|██████▌   | 6995/10712 [1:18:12<30:19,  2.04it/s]
 65%|██████▌   | 6996/10712 [1:18:12<30:18,  2.04it/s]
 65%|██████▌   | 6997/10712 [1:18:13<30:18,  2.04it/s]
 65%|██████▌   | 6998/10712 [1:18:13<30:19,  2.04it/s]
 65%|██████▌   | 6999/10712 [1:18:14<30:17,  2.04it/s]
 65%|██████▌   | 7000/10712 [1:18:14<30:15,  2.05it/s]{'loss': 3.6785, 'grad_norm': 0.211797297000885, 'learning_rate': 0.00032335403742503533, 'epoch': 0.65}

                                                      

 65%|██████▌   | 7000/10712 [1:18:14<30:15,  2.05it/s]
 65%|██████▌   | 7001/10712 [1:18:15<30:19,  2.04it/s]
 65%|██████▌   | 7002/10712 [1:18:15<30:16,  2.04it/s]
 65%|██████▌   | 7003/10712 [1:18:16<30:15,  2.04it/s]
 65%|██████▌   | 7004/10712 [1:18:16<30:18,  2.04it/s]
 65%|██████▌   | 7005/10712 [1:18:17<30:17,  2.04it/s]
 65%|██████▌   | 7006/10712 [1:18:17<30:15,  2.04it/s]
 65%|██████▌   | 7007/10712 [1:18:18<30:14,  2.04it/s]
 65%|██████▌   | 7008/10712 [1:18:18<30:15,  2.04it/s]
 65%|██████▌   | 7009/10712 [1:18:19<30:15,  2.04it/s]
 65%|██████▌   | 7010/10712 [1:18:19<30:14,  2.04it/s]
 65%|██████▌   | 7011/10712 [1:18:20<30:14,  2.04it/s]
 65%|██████▌   | 7012/10712 [1:18:20<30:13,  2.04it/s]
 65%|██████▌   | 7013/10712 [1:18:21<30:12,  2.04it/s]
 65%|██████▌   | 7014/10712 [1:18:21<30:10,  2.04it/s]
 65%|██████▌   | 7015/10712 [1:18:21<30:08,  2.04it/s]
 65%|██████▌   | 7016/10712 [1:18:22<30:10,  2.04it/s]
 66%|██████▌   | 7017/10712 [1:18:22<30:09,  2.04it/s]
 66%|██████▌   | 7018/10712 [1:18:23<30:08,  2.04it/s]
 66%|██████▌   | 7019/10712 [1:18:23<30:10,  2.04it/s]
 66%|██████▌   | 7020/10712 [1:18:24<30:10,  2.04it/s]
 66%|██████▌   | 7021/10712 [1:18:24<30:10,  2.04it/s]
 66%|██████▌   | 7022/10712 [1:18:25<30:09,  2.04it/s]
 66%|██████▌   | 7023/10712 [1:18:25<30:09,  2.04it/s]
 66%|██████▌   | 7024/10712 [1:18:26<30:07,  2.04it/s]
 66%|██████▌   | 7025/10712 [1:18:26<30:07,  2.04it/s]{'loss': 3.6714, 'grad_norm': 0.21976439654827118, 'learning_rate': 0.0003195489967693295, 'epoch': 0.66}
                                                      

 66%|██████▌   | 7025/10712 [1:18:26<30:07,  2.04it/s]
 66%|██████▌   | 7026/10712 [1:18:27<30:08,  2.04it/s]
 66%|██████▌   | 7027/10712 [1:18:27<30:07,  2.04it/s]
 66%|██████▌   | 7028/10712 [1:18:28<30:32,  2.01it/s]
 66%|██████▌   | 7029/10712 [1:18:28<30:26,  2.02it/s]
 66%|██████▌   | 7030/10712 [1:18:29<30:19,  2.02it/s]
 66%|██████▌   | 7031/10712 [1:18:29<30:13,  2.03it/s]
 66%|██████▌   | 7032/10712 [1:18:30<30:12,  2.03it/s]
 66%|██████▌   | 7033/10712 [1:18:30<30:07,  2.03it/s]
 66%|██████▌   | 7034/10712 [1:18:31<30:06,  2.04it/s]
 66%|██████▌   | 7035/10712 [1:18:31<30:02,  2.04it/s]
 66%|██████▌   | 7036/10712 [1:18:32<30:02,  2.04it/s]
 66%|██████▌   | 7037/10712 [1:18:32<30:02,  2.04it/s]
 66%|██████▌   | 7038/10712 [1:18:33<29:59,  2.04it/s]
 66%|██████▌   | 7039/10712 [1:18:33<29:59,  2.04it/s]
 66%|██████▌   | 7040/10712 [1:18:34<29:58,  2.04it/s]
 66%|██████▌   | 7041/10712 [1:18:34<29:57,  2.04it/s]
 66%|██████▌   | 7042/10712 [1:18:35<29:57,  2.04it/s]
 66%|██████▌   | 7043/10712 [1:18:35<29:56,  2.04it/s]
 66%|██████▌   | 7044/10712 [1:18:36<29:55,  2.04it/s]
 66%|██████▌   | 7045/10712 [1:18:36<29:56,  2.04it/s]
 66%|██████▌   | 7046/10712 [1:18:37<29:57,  2.04it/s]
 66%|██████▌   | 7047/10712 [1:18:37<29:55,  2.04it/s]
 66%|██████▌   | 7048/10712 [1:18:38<29:53,  2.04it/s]
 66%|██████▌   | 7049/10712 [1:18:38<29:54,  2.04it/s]
 66%|██████▌   | 7050/10712 [1:18:39<29:55,  2.04it/s]
                                                      {'loss': 3.6815, 'grad_norm': 0.20856554806232452, 'learning_rate': 0.00031575593406631694, 'epoch': 0.66}

 66%|██████▌   | 7050/10712 [1:18:39<29:55,  2.04it/s]
 66%|██████▌   | 7051/10712 [1:18:39<29:55,  2.04it/s]
 66%|██████▌   | 7052/10712 [1:18:40<29:53,  2.04it/s]
 66%|██████▌   | 7053/10712 [1:18:40<29:53,  2.04it/s]
 66%|██████▌   | 7054/10712 [1:18:41<29:54,  2.04it/s]
 66%|██████▌   | 7055/10712 [1:18:41<29:52,  2.04it/s]
 66%|██████▌   | 7056/10712 [1:18:42<29:51,  2.04it/s]
 66%|██████▌   | 7057/10712 [1:18:42<29:54,  2.04it/s]
 66%|██████▌   | 7058/10712 [1:18:43<29:50,  2.04it/s]
 66%|██████▌   | 7059/10712 [1:18:43<29:49,  2.04it/s]
 66%|██████▌   | 7060/10712 [1:18:44<29:48,  2.04it/s]
 66%|██████▌   | 7061/10712 [1:18:44<29:50,  2.04it/s]
 66%|██████▌   | 7062/10712 [1:18:45<29:47,  2.04it/s]
 66%|██████▌   | 7063/10712 [1:18:45<29:45,  2.04it/s]
 66%|██████▌   | 7064/10712 [1:18:46<29:47,  2.04it/s]
 66%|██████▌   | 7065/10712 [1:18:46<29:46,  2.04it/s]
 66%|██████▌   | 7066/10712 [1:18:47<29:45,  2.04it/s]
 66%|██████▌   | 7067/10712 [1:18:47<29:43,  2.04it/s]
 66%|██████▌   | 7068/10712 [1:18:47<29:43,  2.04it/s]
 66%|██████▌   | 7069/10712 [1:18:48<29:45,  2.04it/s]
 66%|██████▌   | 7070/10712 [1:18:48<29:43,  2.04it/s]
 66%|██████▌   | 7071/10712 [1:18:49<29:43,  2.04it/s]
 66%|██████▌   | 7072/10712 [1:18:49<29:43,  2.04it/s]
 66%|██████▌   | 7073/10712 [1:18:50<29:42,  2.04it/s]
 66%|██████▌   | 7074/10712 [1:18:50<29:40,  2.04it/s]
 66%|██████▌   | 7075/10712 [1:18:51<29:38,  2.04it/s]{'loss': 3.6849, 'grad_norm': 0.2075154334306717, 'learning_rate': 0.0003119751010914092, 'epoch': 0.66}
                                                      

 66%|██████▌   | 7075/10712 [1:18:51<29:38,  2.04it/s]
 66%|██████▌   | 7076/10712 [1:18:51<29:42,  2.04it/s]
 66%|██████▌   | 7077/10712 [1:18:52<29:42,  2.04it/s]
 66%|██████▌   | 7078/10712 [1:18:52<29:40,  2.04it/s]
 66%|██████▌   | 7079/10712 [1:18:53<29:40,  2.04it/s]
 66%|██████▌   | 7080/10712 [1:18:53<29:41,  2.04it/s]
 66%|██████▌   | 7081/10712 [1:18:54<29:38,  2.04it/s]
 66%|██████▌   | 7082/10712 [1:18:54<29:38,  2.04it/s]
 66%|██████▌   | 7083/10712 [1:18:55<29:37,  2.04it/s]
 66%|██████▌   | 7084/10712 [1:18:55<29:38,  2.04it/s]
 66%|██████▌   | 7085/10712 [1:18:56<29:36,  2.04it/s]
 66%|██████▌   | 7086/10712 [1:18:56<29:34,  2.04it/s]
 66%|██████▌   | 7087/10712 [1:18:57<29:33,  2.04it/s]
 66%|██████▌   | 7088/10712 [1:18:57<29:34,  2.04it/s]
 66%|██████▌   | 7089/10712 [1:18:58<29:35,  2.04it/s]
 66%|██████▌   | 7090/10712 [1:18:58<29:35,  2.04it/s]
 66%|██████▌   | 7091/10712 [1:18:59<29:34,  2.04it/s]
 66%|██████▌   | 7092/10712 [1:18:59<29:34,  2.04it/s]
 66%|██████▌   | 7093/10712 [1:19:00<29:35,  2.04it/s]
 66%|██████▌   | 7094/10712 [1:19:00<29:34,  2.04it/s]
 66%|██████▌   | 7095/10712 [1:19:01<29:31,  2.04it/s]
 66%|██████▌   | 7096/10712 [1:19:01<29:29,  2.04it/s]
 66%|██████▋   | 7097/10712 [1:19:02<29:29,  2.04it/s]
 66%|██████▋   | 7098/10712 [1:19:02<29:30,  2.04it/s]
 66%|██████▋   | 7099/10712 [1:19:03<29:30,  2.04it/s]
 66%|██████▋   | 7100/10712 [1:19:03<29:29,  2.04it/s]{'loss': 3.6701, 'grad_norm': 0.22216494381427765, 'learning_rate': 0.00030820674880823476, 'epoch': 0.66}

                                                      

 66%|██████▋   | 7100/10712 [1:19:03<29:29,  2.04it/s]
 66%|██████▋   | 7101/10712 [1:19:04<29:32,  2.04it/s]
 66%|██████▋   | 7102/10712 [1:19:04<29:31,  2.04it/s]
 66%|██████▋   | 7103/10712 [1:19:05<29:29,  2.04it/s]
 66%|██████▋   | 7104/10712 [1:19:05<29:29,  2.04it/s]
 66%|██████▋   | 7105/10712 [1:19:06<29:30,  2.04it/s]
 66%|██████▋   | 7106/10712 [1:19:06<29:29,  2.04it/s]
 66%|██████▋   | 7107/10712 [1:19:07<29:27,  2.04it/s]
 66%|██████▋   | 7108/10712 [1:19:07<29:25,  2.04it/s]
 66%|██████▋   | 7109/10712 [1:19:08<29:26,  2.04it/s]
 66%|██████▋   | 7110/10712 [1:19:08<29:26,  2.04it/s]
 66%|██████▋   | 7111/10712 [1:19:09<29:23,  2.04it/s]
 66%|██████▋   | 7112/10712 [1:19:09<29:22,  2.04it/s]
 66%|██████▋   | 7113/10712 [1:19:10<29:23,  2.04it/s]
 66%|██████▋   | 7114/10712 [1:19:10<29:22,  2.04it/s]
 66%|██████▋   | 7115/10712 [1:19:11<29:20,  2.04it/s]
 66%|██████▋   | 7116/10712 [1:19:11<29:20,  2.04it/s]
 66%|██████▋   | 7117/10712 [1:19:12<29:22,  2.04it/s]
 66%|██████▋   | 7118/10712 [1:19:12<29:20,  2.04it/s]
 66%|██████▋   | 7119/10712 [1:19:12<29:18,  2.04it/s]
 66%|██████▋   | 7120/10712 [1:19:13<29:19,  2.04it/s]
 66%|██████▋   | 7121/10712 [1:19:13<29:19,  2.04it/s]
 66%|██████▋   | 7122/10712 [1:19:14<29:18,  2.04it/s]
 66%|██████▋   | 7123/10712 [1:19:14<29:15,  2.04it/s]
 67%|██████▋   | 7124/10712 [1:19:15<29:18,  2.04it/s]
 67%|██████▋   | 7125/10712 [1:19:15<29:18,  2.04it/s]{'loss': 3.6708, 'grad_norm': 0.2090538591146469, 'learning_rate': 0.0003044511273519802, 'epoch': 0.67}
                                                      

 67%|██████▋   | 7125/10712 [1:19:15<29:18,  2.04it/s]
 67%|██████▋   | 7126/10712 [1:19:16<29:18,  2.04it/s]
 67%|██████▋   | 7127/10712 [1:19:16<29:15,  2.04it/s]
 67%|██████▋   | 7128/10712 [1:19:17<29:15,  2.04it/s]
 67%|██████▋   | 7129/10712 [1:19:17<29:17,  2.04it/s]
 67%|██████▋   | 7130/10712 [1:19:18<29:14,  2.04it/s]
 67%|██████▋   | 7131/10712 [1:19:18<29:13,  2.04it/s]
 67%|██████▋   | 7132/10712 [1:19:19<29:15,  2.04it/s]
 67%|██████▋   | 7133/10712 [1:19:19<29:13,  2.04it/s]
 67%|██████▋   | 7134/10712 [1:19:20<29:13,  2.04it/s]
 67%|██████▋   | 7135/10712 [1:19:20<29:14,  2.04it/s]
 67%|██████▋   | 7136/10712 [1:19:21<29:13,  2.04it/s]
 67%|██████▋   | 7137/10712 [1:19:21<29:11,  2.04it/s]
 67%|██████▋   | 7138/10712 [1:19:22<29:09,  2.04it/s]
 67%|██████▋   | 7139/10712 [1:19:22<29:10,  2.04it/s]
 67%|██████▋   | 7140/10712 [1:19:23<29:11,  2.04it/s]
 67%|██████▋   | 7141/10712 [1:19:23<29:09,  2.04it/s]
 67%|██████▋   | 7142/10712 [1:19:24<29:10,  2.04it/s]
 67%|██████▋   | 7143/10712 [1:19:24<29:08,  2.04it/s]
 67%|██████▋   | 7144/10712 [1:19:25<29:07,  2.04it/s]
 67%|██████▋   | 7145/10712 [1:19:25<29:05,  2.04it/s]
 67%|██████▋   | 7146/10712 [1:19:26<29:05,  2.04it/s]
 67%|██████▋   | 7147/10712 [1:19:26<29:04,  2.04it/s]
 67%|██████▋   | 7148/10712 [1:19:27<29:03,  2.04it/s]
 67%|██████▋   | 7149/10712 [1:19:27<29:02,  2.04it/s]
 67%|██████▋   | 7150/10712 [1:19:28<29:04,  2.04it/s]{'loss': 3.6668, 'grad_norm': 0.20870347321033478, 'learning_rate': 0.00030070848601278715, 'epoch': 0.67}
                                                      

 67%|██████▋   | 7150/10712 [1:19:28<29:04,  2.04it/s]
 67%|██████▋   | 7151/10712 [1:19:28<29:06,  2.04it/s]
 67%|██████▋   | 7152/10712 [1:19:29<29:05,  2.04it/s]
 67%|██████▋   | 7153/10712 [1:19:29<29:04,  2.04it/s]
 67%|██████▋   | 7154/10712 [1:19:30<29:04,  2.04it/s]
 67%|██████▋   | 7155/10712 [1:19:30<29:03,  2.04it/s]
 67%|██████▋   | 7156/10712 [1:19:31<29:02,  2.04it/s]
 67%|██████▋   | 7157/10712 [1:19:31<29:04,  2.04it/s]
 67%|██████▋   | 7158/10712 [1:19:32<29:02,  2.04it/s]
 67%|██████▋   | 7159/10712 [1:19:32<29:02,  2.04it/s]
 67%|██████▋   | 7160/10712 [1:19:33<29:01,  2.04it/s]
 67%|██████▋   | 7161/10712 [1:19:33<29:02,  2.04it/s]
 67%|██████▋   | 7162/10712 [1:19:34<28:59,  2.04it/s]
 67%|██████▋   | 7163/10712 [1:19:34<28:59,  2.04it/s]
 67%|██████▋   | 7164/10712 [1:19:35<28:59,  2.04it/s]
 67%|██████▋   | 7165/10712 [1:19:35<28:59,  2.04it/s]
 67%|██████▋   | 7166/10712 [1:19:36<28:56,  2.04it/s]
 67%|██████▋   | 7167/10712 [1:19:36<28:56,  2.04it/s]
 67%|██████▋   | 7168/10712 [1:19:36<28:58,  2.04it/s]
 67%|██████▋   | 7169/10712 [1:19:37<28:57,  2.04it/s]
 67%|██████▋   | 7170/10712 [1:19:37<28:56,  2.04it/s]
 67%|██████▋   | 7171/10712 [1:19:38<28:54,  2.04it/s]
 67%|██████▋   | 7172/10712 [1:19:38<28:54,  2.04it/s]
 67%|██████▋   | 7173/10712 [1:19:39<28:54,  2.04it/s]
 67%|██████▋   | 7174/10712 [1:19:39<28:53,  2.04it/s]
 67%|██████▋   | 7175/10712 [1:19:40<28:53,  2.04it/s]{'loss': 3.6697, 'grad_norm': 0.20599548518657684, 'learning_rate': 0.0002969790732192046, 'epoch': 0.67}
                                                      

 67%|██████▋   | 7175/10712 [1:19:40<28:53,  2.04it/s]
 67%|██████▋   | 7176/10712 [1:19:40<28:54,  2.04it/s]
 67%|██████▋   | 7177/10712 [1:19:41<28:53,  2.04it/s]
 67%|██████▋   | 7178/10712 [1:19:41<28:51,  2.04it/s]
 67%|██████▋   | 7179/10712 [1:19:42<28:51,  2.04it/s]
 67%|██████▋   | 7180/10712 [1:19:42<28:52,  2.04it/s]
 67%|██████▋   | 7181/10712 [1:19:43<28:50,  2.04it/s]
 67%|██████▋   | 7182/10712 [1:19:43<28:49,  2.04it/s]
 67%|██████▋   | 7183/10712 [1:19:44<28:51,  2.04it/s]
 67%|██████▋   | 7184/10712 [1:19:44<28:50,  2.04it/s]
 67%|██████▋   | 7185/10712 [1:19:45<28:49,  2.04it/s]
 67%|██████▋   | 7186/10712 [1:19:45<28:47,  2.04it/s]
 67%|██████▋   | 7187/10712 [1:19:46<28:49,  2.04it/s]
 67%|██████▋   | 7188/10712 [1:19:46<28:48,  2.04it/s]
 67%|██████▋   | 7189/10712 [1:19:47<28:45,  2.04it/s]
 67%|██████▋   | 7190/10712 [1:19:47<28:47,  2.04it/s]
 67%|██████▋   | 7191/10712 [1:19:48<28:48,  2.04it/s]
 67%|██████▋   | 7192/10712 [1:19:48<29:13,  2.01it/s]
 67%|██████▋   | 7193/10712 [1:19:49<29:05,  2.02it/s]
 67%|██████▋   | 7194/10712 [1:19:49<28:59,  2.02it/s]
 67%|██████▋   | 7195/10712 [1:19:50<28:55,  2.03it/s]
 67%|██████▋   | 7196/10712 [1:19:50<28:53,  2.03it/s]
 67%|██████▋   | 7197/10712 [1:19:51<28:51,  2.03it/s]
 67%|██████▋   | 7198/10712 [1:19:51<28:47,  2.03it/s]
 67%|██████▋   | 7199/10712 [1:19:52<28:43,  2.04it/s]
 67%|██████▋   | 7200/10712 [1:19:52<28:45,  2.04it/s]{'loss': 3.6632, 'grad_norm': 0.20758244395256042, 'learning_rate': 0.000293263136521699, 'epoch': 0.67}
                                                      

 67%|██████▋   | 7200/10712 [1:19:52<28:45,  2.04it/s]
 67%|██████▋   | 7201/10712 [1:19:53<28:46,  2.03it/s]
 67%|██████▋   | 7202/10712 [1:19:53<28:43,  2.04it/s]
 67%|██████▋   | 7203/10712 [1:19:54<28:44,  2.03it/s]
 67%|██████▋   | 7204/10712 [1:19:54<28:41,  2.04it/s]
 67%|██████▋   | 7205/10712 [1:19:55<28:39,  2.04it/s]
 67%|██████▋   | 7206/10712 [1:19:55<28:38,  2.04it/s]
 67%|██████▋   | 7207/10712 [1:19:56<28:38,  2.04it/s]
 67%|██████▋   | 7208/10712 [1:19:56<28:38,  2.04it/s]
 67%|██████▋   | 7209/10712 [1:19:57<28:34,  2.04it/s]
 67%|██████▋   | 7210/10712 [1:19:57<28:35,  2.04it/s]
 67%|██████▋   | 7211/10712 [1:19:58<28:37,  2.04it/s]
 67%|██████▋   | 7212/10712 [1:19:58<28:33,  2.04it/s]
 67%|██████▋   | 7213/10712 [1:19:59<28:31,  2.04it/s]
 67%|██████▋   | 7214/10712 [1:19:59<28:32,  2.04it/s]
 67%|██████▋   | 7215/10712 [1:20:00<28:33,  2.04it/s]
 67%|██████▋   | 7216/10712 [1:20:00<28:33,  2.04it/s]
 67%|██████▋   | 7217/10712 [1:20:01<28:30,  2.04it/s]
 67%|██████▋   | 7218/10712 [1:20:01<28:31,  2.04it/s]
 67%|██████▋   | 7219/10712 [1:20:02<28:32,  2.04it/s]
 67%|██████▋   | 7220/10712 [1:20:02<28:33,  2.04it/s]
 67%|██████▋   | 7221/10712 [1:20:03<28:32,  2.04it/s]
 67%|██████▋   | 7222/10712 [1:20:03<28:32,  2.04it/s]
 67%|██████▋   | 7223/10712 [1:20:03<28:32,  2.04it/s]
 67%|██████▋   | 7224/10712 [1:20:04<28:30,  2.04it/s]
 67%|██████▋   | 7225/10712 [1:20:04<28:31,  2.04it/s]{'loss': 3.6716, 'grad_norm': 0.20743732154369354, 'learning_rate': 0.0002895609225762222, 'epoch': 0.67}
                                                      

 67%|██████▋   | 7225/10712 [1:20:04<28:31,  2.04it/s]
 67%|██████▋   | 7226/10712 [1:20:05<28:32,  2.04it/s]
 67%|██████▋   | 7227/10712 [1:20:05<28:29,  2.04it/s]
 67%|██████▋   | 7228/10712 [1:20:06<28:28,  2.04it/s]
 67%|██████▋   | 7229/10712 [1:20:06<28:26,  2.04it/s]
 67%|██████▋   | 7230/10712 [1:20:07<28:28,  2.04it/s]
 68%|██████▊   | 7231/10712 [1:20:07<28:28,  2.04it/s]
 68%|██████▊   | 7232/10712 [1:20:08<28:26,  2.04it/s]
 68%|██████▊   | 7233/10712 [1:20:08<28:25,  2.04it/s]
 68%|██████▊   | 7234/10712 [1:20:09<28:24,  2.04it/s]
 68%|██████▊   | 7235/10712 [1:20:09<28:25,  2.04it/s]
 68%|██████▊   | 7236/10712 [1:20:10<28:24,  2.04it/s]
 68%|██████▊   | 7237/10712 [1:20:10<28:22,  2.04it/s]
 68%|██████▊   | 7238/10712 [1:20:11<28:22,  2.04it/s]
 68%|██████▊   | 7239/10712 [1:20:11<28:24,  2.04it/s]
 68%|██████▊   | 7240/10712 [1:20:12<28:22,  2.04it/s]
 68%|██████▊   | 7241/10712 [1:20:12<28:21,  2.04it/s]
 68%|██████▊   | 7242/10712 [1:20:13<28:21,  2.04it/s]
 68%|██████▊   | 7243/10712 [1:20:13<28:21,  2.04it/s]
 68%|██████▊   | 7244/10712 [1:20:14<28:20,  2.04it/s]
 68%|██████▊   | 7245/10712 [1:20:14<28:18,  2.04it/s]
 68%|██████▊   | 7246/10712 [1:20:15<28:19,  2.04it/s]
 68%|██████▊   | 7247/10712 [1:20:15<28:19,  2.04it/s]
 68%|██████▊   | 7248/10712 [1:20:16<28:18,  2.04it/s]
 68%|██████▊   | 7249/10712 [1:20:16<28:19,  2.04it/s]
 68%|██████▊   | 7250/10712 [1:20:17<28:20,  2.04it/s]
                                                      {'loss': 3.662, 'grad_norm': 0.2156381458044052, 'learning_rate': 0.0002858726771278391, 'epoch': 0.68}

 68%|██████▊   | 7250/10712 [1:20:17<28:20,  2.04it/s]
 68%|██████▊   | 7251/10712 [1:20:17<28:21,  2.03it/s]
 68%|██████▊   | 7252/10712 [1:20:18<28:19,  2.04it/s]
 68%|██████▊   | 7253/10712 [1:20:18<28:19,  2.04it/s]
 68%|██████▊   | 7254/10712 [1:20:19<28:16,  2.04it/s]
 68%|██████▊   | 7255/10712 [1:20:19<28:14,  2.04it/s]
 68%|██████▊   | 7256/10712 [1:20:20<28:13,  2.04it/s]
 68%|██████▊   | 7257/10712 [1:20:20<28:13,  2.04it/s]
 68%|██████▊   | 7258/10712 [1:20:21<28:12,  2.04it/s]
 68%|██████▊   | 7259/10712 [1:20:21<28:11,  2.04it/s]
 68%|██████▊   | 7260/10712 [1:20:22<28:11,  2.04it/s]
 68%|██████▊   | 7261/10712 [1:20:22<28:10,  2.04it/s]
 68%|██████▊   | 7262/10712 [1:20:23<28:08,  2.04it/s]
 68%|██████▊   | 7263/10712 [1:20:23<28:08,  2.04it/s]
 68%|██████▊   | 7264/10712 [1:20:24<28:08,  2.04it/s]
 68%|██████▊   | 7265/10712 [1:20:24<28:08,  2.04it/s]
 68%|██████▊   | 7266/10712 [1:20:25<28:06,  2.04it/s]
 68%|██████▊   | 7267/10712 [1:20:25<28:05,  2.04it/s]
 68%|██████▊   | 7268/10712 [1:20:26<28:08,  2.04it/s]
 68%|██████▊   | 7269/10712 [1:20:26<28:08,  2.04it/s]
 68%|██████▊   | 7270/10712 [1:20:27<28:05,  2.04it/s]
 68%|██████▊   | 7271/10712 [1:20:27<28:05,  2.04it/s]
 68%|██████▊   | 7272/10712 [1:20:28<28:06,  2.04it/s]
 68%|██████▊   | 7273/10712 [1:20:28<28:05,  2.04it/s]
 68%|██████▊   | 7274/10712 [1:20:28<28:02,  2.04it/s]
 68%|██████▊   | 7275/10712 [1:20:29<28:00,  2.04it/s]
                                                      
{'loss': 3.6651, 'grad_norm': 0.20926706492900848, 'learning_rate': 0.00028219864499441584, 'epoch': 0.68}

 68%|██████▊   | 7275/10712 [1:20:29<28:00,  2.04it/s]
 68%|██████▊   | 7276/10712 [1:20:29<28:04,  2.04it/s]
 68%|██████▊   | 7277/10712 [1:20:30<28:02,  2.04it/s]
 68%|██████▊   | 7278/10712 [1:20:30<28:03,  2.04it/s]
 68%|██████▊   | 7279/10712 [1:20:31<28:04,  2.04it/s]
 68%|██████▊   | 7280/10712 [1:20:31<28:02,  2.04it/s]
 68%|██████▊   | 7281/10712 [1:20:32<27:59,  2.04it/s]
 68%|██████▊   | 7282/10712 [1:20:32<28:00,  2.04it/s]
 68%|██████▊   | 7283/10712 [1:20:33<28:01,  2.04it/s]
 68%|██████▊   | 7284/10712 [1:20:33<28:00,  2.04it/s]
 68%|██████▊   | 7285/10712 [1:20:34<27:59,  2.04it/s]
 68%|██████▊   | 7286/10712 [1:20:34<28:00,  2.04it/s]
 68%|██████▊   | 7287/10712 [1:20:35<27:59,  2.04it/s]
 68%|██████▊   | 7288/10712 [1:20:35<27:57,  2.04it/s]
 68%|██████▊   | 7289/10712 [1:20:36<28:00,  2.04it/s]
 68%|██████▊   | 7290/10712 [1:20:36<28:00,  2.04it/s]
 68%|██████▊   | 7291/10712 [1:20:37<27:58,  2.04it/s]
 68%|██████▊   | 7292/10712 [1:20:37<27:58,  2.04it/s]
 68%|██████▊   | 7293/10712 [1:20:38<27:59,  2.04it/s]
 68%|██████▊   | 7294/10712 [1:20:38<27:57,  2.04it/s]
 68%|██████▊   | 7295/10712 [1:20:39<27:56,  2.04it/s]
 68%|██████▊   | 7296/10712 [1:20:39<27:56,  2.04it/s]
 68%|██████▊   | 7297/10712 [1:20:40<27:54,  2.04it/s]
 68%|██████▊   | 7298/10712 [1:20:40<27:53,  2.04it/s]
 68%|██████▊   | 7299/10712 [1:20:41<27:56,  2.04it/s]
 68%|██████▊   | 7300/10712 [1:20:41<27:54,  2.04it/s]
                                                      
{'loss': 3.66, 'grad_norm': 0.21162216365337372, 'learning_rate': 0.00027853907005036824, 'epoch': 0.68}

 68%|██████▊   | 7300/10712 [1:20:41<27:54,  2.04it/s]
 68%|██████▊   | 7301/10712 [1:20:42<28:01,  2.03it/s]
 68%|██████▊   | 7302/10712 [1:20:42<27:59,  2.03it/s]
 68%|██████▊   | 7303/10712 [1:20:43<27:55,  2.03it/s]
 68%|██████▊   | 7304/10712 [1:20:43<27:54,  2.04it/s]
 68%|██████▊   | 7305/10712 [1:20:44<27:54,  2.03it/s]
 68%|██████▊   | 7306/10712 [1:20:44<27:52,  2.04it/s]
 68%|██████▊   | 7307/10712 [1:20:45<27:50,  2.04it/s]
 68%|██████▊   | 7308/10712 [1:20:45<27:50,  2.04it/s]
 68%|██████▊   | 7309/10712 [1:20:46<27:48,  2.04it/s]
 68%|██████▊   | 7310/10712 [1:20:46<27:47,  2.04it/s]
 68%|██████▊   | 7311/10712 [1:20:47<27:46,  2.04it/s]
 68%|██████▊   | 7312/10712 [1:20:47<27:46,  2.04it/s]
 68%|██████▊   | 7313/10712 [1:20:48<27:48,  2.04it/s]
 68%|██████▊   | 7314/10712 [1:20:48<27:46,  2.04it/s]
 68%|██████▊   | 7315/10712 [1:20:49<27:47,  2.04it/s]
 68%|██████▊   | 7316/10712 [1:20:49<27:46,  2.04it/s]
 68%|██████▊   | 7317/10712 [1:20:50<27:44,  2.04it/s]
 68%|██████▊   | 7318/10712 [1:20:50<28:08,  2.01it/s]
 68%|██████▊   | 7319/10712 [1:20:51<28:00,  2.02it/s]
 68%|██████▊   | 7320/10712 [1:20:51<27:54,  2.03it/s]
 68%|██████▊   | 7321/10712 [1:20:52<27:50,  2.03it/s]
 68%|██████▊   | 7322/10712 [1:20:52<27:48,  2.03it/s]
 68%|██████▊   | 7323/10712 [1:20:53<27:47,  2.03it/s]
 68%|██████▊   | 7324/10712 [1:20:53<27:44,  2.04it/s]
 68%|██████▊   | 7325/10712 [1:20:54<27:40,  2.04it/s]
                                                      {'loss': 3.6577, 'grad_norm': 0.20539535582065582, 'learning_rate': 0.00027489419521047563, 'epoch': 0.68}

 68%|██████▊   | 7325/10712 [1:20:54<27:40,  2.04it/s]
 68%|██████▊   | 7326/10712 [1:20:54<27:43,  2.04it/s]
 68%|██████▊   | 7327/10712 [1:20:55<27:41,  2.04it/s]
 68%|██████▊   | 7328/10712 [1:20:55<27:41,  2.04it/s]
 68%|██████▊   | 7329/10712 [1:20:55<27:41,  2.04it/s]
 68%|██████▊   | 7330/10712 [1:20:56<27:39,  2.04it/s]
 68%|██████▊   | 7331/10712 [1:20:56<27:37,  2.04it/s]
 68%|██████▊   | 7332/10712 [1:20:57<27:37,  2.04it/s]
 68%|██████▊   | 7333/10712 [1:20:57<27:36,  2.04it/s]
 68%|██████▊   | 7334/10712 [1:20:58<27:36,  2.04it/s]
 68%|██████▊   | 7335/10712 [1:20:58<27:34,  2.04it/s]
 68%|██████▊   | 7336/10712 [1:20:59<27:33,  2.04it/s]
 68%|██████▊   | 7337/10712 [1:20:59<27:34,  2.04it/s]
 69%|██████▊   | 7338/10712 [1:21:00<27:34,  2.04it/s]
 69%|██████▊   | 7339/10712 [1:21:00<27:33,  2.04it/s]
 69%|██████▊   | 7340/10712 [1:21:01<27:32,  2.04it/s]
 69%|██████▊   | 7341/10712 [1:21:01<27:32,  2.04it/s]
 69%|██████▊   | 7342/10712 [1:21:02<27:30,  2.04it/s]
 69%|██████▊   | 7343/10712 [1:21:02<27:30,  2.04it/s]
 69%|██████▊   | 7344/10712 [1:21:03<27:31,  2.04it/s]
 69%|██████▊   | 7345/10712 [1:21:03<27:30,  2.04it/s]
 69%|██████▊   | 7346/10712 [1:21:04<27:28,  2.04it/s]
 69%|██████▊   | 7347/10712 [1:21:04<27:28,  2.04it/s]
 69%|██████▊   | 7348/10712 [1:21:05<27:32,  2.04it/s]
 69%|██████▊   | 7349/10712 [1:21:05<27:29,  2.04it/s]
 69%|██████▊   | 7350/10712 [1:21:06<27:27,  2.04it/s]
                                                      {'loss': 3.6604, 'grad_norm': 0.2114870697259903, 'learning_rate': 0.0002712642624137552, 'epoch': 0.69}

 69%|██████▊   | 7350/10712 [1:21:06<27:27,  2.04it/s]
 69%|██████▊   | 7351/10712 [1:21:06<27:30,  2.04it/s]
 69%|██████▊   | 7352/10712 [1:21:07<27:28,  2.04it/s]
 69%|██████▊   | 7353/10712 [1:21:07<27:25,  2.04it/s]
 69%|██████▊   | 7354/10712 [1:21:08<27:23,  2.04it/s]
 69%|██████▊   | 7355/10712 [1:21:08<27:25,  2.04it/s]
 69%|██████▊   | 7356/10712 [1:21:09<27:24,  2.04it/s]
 69%|██████▊   | 7357/10712 [1:21:09<27:21,  2.04it/s]
 69%|██████▊   | 7358/10712 [1:21:10<27:22,  2.04it/s]
 69%|██████▊   | 7359/10712 [1:21:10<27:22,  2.04it/s]
 69%|██████▊   | 7360/10712 [1:21:11<27:21,  2.04it/s]
 69%|██████▊   | 7361/10712 [1:21:11<27:20,  2.04it/s]
 69%|██████▊   | 7362/10712 [1:21:12<27:19,  2.04it/s]
 69%|██████▊   | 7363/10712 [1:21:12<27:20,  2.04it/s]
 69%|██████▊   | 7364/10712 [1:21:13<27:21,  2.04it/s]
 69%|██████▉   | 7365/10712 [1:21:13<27:21,  2.04it/s]
 69%|██████▉   | 7366/10712 [1:21:14<27:20,  2.04it/s]
 69%|██████▉   | 7367/10712 [1:21:14<27:20,  2.04it/s]
 69%|██████▉   | 7368/10712 [1:21:15<27:20,  2.04it/s]
 69%|██████▉   | 7369/10712 [1:21:15<27:19,  2.04it/s]
 69%|██████▉   | 7370/10712 [1:21:16<27:16,  2.04it/s]
 69%|██████▉   | 7371/10712 [1:21:16<27:14,  2.04it/s]
 69%|██████▉   | 7372/10712 [1:21:17<27:15,  2.04it/s]
 69%|██████▉   | 7373/10712 [1:21:17<27:17,  2.04it/s]
 69%|██████▉   | 7374/10712 [1:21:18<27:15,  2.04it/s]
 69%|██████▉   | 7375/10712 [1:21:18<27:14,  2.04it/s]
                                                      
{'loss': 3.6675, 'grad_norm': 0.22274236381053925, 'learning_rate': 0.0002676495126074034, 'epoch': 0.69}

 69%|██████▉   | 7375/10712 [1:21:18<27:14,  2.04it/s]
 69%|██████▉   | 7376/10712 [1:21:19<27:16,  2.04it/s]
 69%|██████▉   | 7377/10712 [1:21:19<27:16,  2.04it/s]
 69%|██████▉   | 7378/10712 [1:21:20<27:16,  2.04it/s]
 69%|██████▉   | 7379/10712 [1:21:20<27:13,  2.04it/s]
 69%|██████▉   | 7380/10712 [1:21:20<27:13,  2.04it/s]
 69%|██████▉   | 7381/10712 [1:21:21<27:15,  2.04it/s]
 69%|██████▉   | 7382/10712 [1:21:21<27:12,  2.04it/s]
 69%|██████▉   | 7383/10712 [1:21:22<27:12,  2.04it/s]
 69%|██████▉   | 7384/10712 [1:21:22<27:12,  2.04it/s]
 69%|██████▉   | 7385/10712 [1:21:23<27:12,  2.04it/s]
 69%|██████▉   | 7386/10712 [1:21:23<27:11,  2.04it/s]
 69%|██████▉   | 7387/10712 [1:21:24<27:12,  2.04it/s]
 69%|██████▉   | 7388/10712 [1:21:24<27:11,  2.04it/s]
 69%|██████▉   | 7389/10712 [1:21:25<27:10,  2.04it/s]
 69%|██████▉   | 7390/10712 [1:21:25<27:10,  2.04it/s]
 69%|██████▉   | 7391/10712 [1:21:26<27:10,  2.04it/s]
 69%|██████▉   | 7392/10712 [1:21:26<27:06,  2.04it/s]
 69%|██████▉   | 7393/10712 [1:21:27<27:05,  2.04it/s]
 69%|██████▉   | 7394/10712 [1:21:27<27:06,  2.04it/s]
 69%|██████▉   | 7395/10712 [1:21:28<27:06,  2.04it/s]
 69%|██████▉   | 7396/10712 [1:21:28<27:05,  2.04it/s]
 69%|██████▉   | 7397/10712 [1:21:29<27:04,  2.04it/s]
 69%|██████▉   | 7398/10712 [1:21:29<27:08,  2.03it/s]
 69%|██████▉   | 7399/10712 [1:21:30<27:06,  2.04it/s]
 69%|██████▉   | 7400/10712 [1:21:31<31:33,  1.75it/s]{'loss': 3.6577, 'grad_norm': 0.21110838651657104, 'learning_rate': 0.0002640501857308021, 'epoch': 0.69}

                                                      

 69%|██████▉   | 7400/10712 [1:21:31<31:33,  1.75it/s]
 69%|██████▉   | 7401/10712 [1:21:31<30:11,  1.83it/s]
 69%|██████▉   | 7402/10712 [1:21:32<29:14,  1.89it/s]
 69%|██████▉   | 7403/10712 [1:21:32<28:34,  1.93it/s]
 69%|██████▉   | 7404/10712 [1:21:33<28:06,  1.96it/s]
 69%|██████▉   | 7405/10712 [1:21:33<27:44,  1.99it/s]
 69%|██████▉   | 7406/10712 [1:21:34<27:31,  2.00it/s]
 69%|██████▉   | 7407/10712 [1:21:34<27:22,  2.01it/s]
 69%|██████▉   | 7408/10712 [1:21:34<27:14,  2.02it/s]
 69%|██████▉   | 7409/10712 [1:21:35<27:08,  2.03it/s]
 69%|██████▉   | 7410/10712 [1:21:35<27:05,  2.03it/s]
 69%|██████▉   | 7411/10712 [1:21:36<27:01,  2.04it/s]
 69%|██████▉   | 7412/10712 [1:21:36<27:00,  2.04it/s]
 69%|██████▉   | 7413/10712 [1:21:37<26:57,  2.04it/s]
 69%|██████▉   | 7414/10712 [1:21:37<26:58,  2.04it/s]
 69%|██████▉   | 7415/10712 [1:21:38<26:56,  2.04it/s]
 69%|██████▉   | 7416/10712 [1:21:38<26:53,  2.04it/s]
 69%|██████▉   | 7417/10712 [1:21:39<26:53,  2.04it/s]
 69%|██████▉   | 7418/10712 [1:21:39<26:52,  2.04it/s]
 69%|██████▉   | 7419/10712 [1:21:40<26:53,  2.04it/s]
 69%|██████▉   | 7420/10712 [1:21:40<26:53,  2.04it/s]
 69%|██████▉   | 7421/10712 [1:21:41<27:35,  1.99it/s]
 69%|██████▉   | 7422/10712 [1:21:41<27:22,  2.00it/s]
 69%|██████▉   | 7423/10712 [1:21:42<27:13,  2.01it/s]
 69%|██████▉   | 7424/10712 [1:21:42<27:06,  2.02it/s]
 69%|██████▉   | 7425/10712 [1:21:43<27:03,  2.02it/s]{'loss': 3.658, 'grad_norm': 0.21561577916145325, 'learning_rate': 0.0002604665206995922, 'epoch': 0.69}

                                                      

 69%|██████▉   | 7425/10712 [1:21:43<27:03,  2.02it/s]
 69%|██████▉   | 7426/10712 [1:21:43<27:01,  2.03it/s]
 69%|██████▉   | 7427/10712 [1:21:44<26:57,  2.03it/s]
 69%|██████▉   | 7428/10712 [1:21:44<26:56,  2.03it/s]
 69%|██████▉   | 7429/10712 [1:21:45<26:55,  2.03it/s]
 69%|██████▉   | 7430/10712 [1:21:45<26:53,  2.03it/s]
 69%|██████▉   | 7431/10712 [1:21:46<26:51,  2.04it/s]
 69%|██████▉   | 7432/10712 [1:21:46<26:49,  2.04it/s]
 69%|██████▉   | 7433/10712 [1:21:47<26:47,  2.04it/s]
 69%|██████▉   | 7434/10712 [1:21:47<26:48,  2.04it/s]
 69%|██████▉   | 7435/10712 [1:21:48<26:48,  2.04it/s]
 69%|██████▉   | 7436/10712 [1:21:48<26:46,  2.04it/s]
 69%|██████▉   | 7437/10712 [1:21:49<31:08,  1.75it/s]
 69%|██████▉   | 7438/10712 [1:21:50<29:49,  1.83it/s]
 69%|██████▉   | 7439/10712 [1:21:50<28:53,  1.89it/s]
 69%|██████▉   | 7440/10712 [1:21:50<28:15,  1.93it/s]
 69%|██████▉   | 7441/10712 [1:21:51<27:47,  1.96it/s]
 69%|██████▉   | 7442/10712 [1:21:51<27:27,  1.98it/s]
 69%|██████▉   | 7443/10712 [1:21:52<27:14,  2.00it/s]
 69%|██████▉   | 7444/10712 [1:21:52<27:06,  2.01it/s]
 70%|██████▉   | 7445/10712 [1:21:53<26:59,  2.02it/s]
 70%|██████▉   | 7446/10712 [1:21:53<26:53,  2.02it/s]
 70%|██████▉   | 7447/10712 [1:21:54<26:46,  2.03it/s]
 70%|██████▉   | 7448/10712 [1:21:54<26:46,  2.03it/s]
 70%|██████▉   | 7449/10712 [1:21:55<26:44,  2.03it/s]
 70%|██████▉   | 7450/10712 [1:21:55<26:41,  2.04it/s]
                                                      
{'loss': 3.6583, 'grad_norm': 0.21814173460006714, 'learning_rate': 0.0002568987553898142, 'epoch': 0.7}

 70%|██████▉   | 7450/10712 [1:21:55<26:41,  2.04it/s]
 70%|██████▉   | 7451/10712 [1:21:56<26:41,  2.04it/s]
 70%|██████▉   | 7452/10712 [1:21:56<26:40,  2.04it/s]
 70%|██████▉   | 7453/10712 [1:21:57<26:39,  2.04it/s]
 70%|██████▉   | 7454/10712 [1:21:57<26:37,  2.04it/s]
 70%|██████▉   | 7455/10712 [1:21:58<26:34,  2.04it/s]
 70%|██████▉   | 7456/10712 [1:21:58<26:35,  2.04it/s]
 70%|██████▉   | 7457/10712 [1:21:59<26:36,  2.04it/s]
 70%|██████▉   | 7458/10712 [1:21:59<26:35,  2.04it/s]
 70%|██████▉   | 7459/10712 [1:22:00<26:32,  2.04it/s]
 70%|██████▉   | 7460/10712 [1:22:00<26:33,  2.04it/s]
 70%|██████▉   | 7461/10712 [1:22:01<26:33,  2.04it/s]
 70%|██████▉   | 7462/10712 [1:22:01<26:32,  2.04it/s]
 70%|██████▉   | 7463/10712 [1:22:02<26:32,  2.04it/s]
 70%|██████▉   | 7464/10712 [1:22:02<26:34,  2.04it/s]
 70%|██████▉   | 7465/10712 [1:22:03<26:33,  2.04it/s]
 70%|██████▉   | 7466/10712 [1:22:03<26:30,  2.04it/s]
 70%|██████▉   | 7467/10712 [1:22:04<26:53,  2.01it/s]
 70%|██████▉   | 7468/10712 [1:22:04<26:48,  2.02it/s]
 70%|██████▉   | 7469/10712 [1:22:05<26:40,  2.03it/s]
 70%|██████▉   | 7470/10712 [1:22:05<26:35,  2.03it/s]
 70%|██████▉   | 7471/10712 [1:22:06<26:35,  2.03it/s]
 70%|██████▉   | 7472/10712 [1:22:06<26:32,  2.03it/s]
 70%|██████▉   | 7473/10712 [1:22:07<26:30,  2.04it/s]
 70%|██████▉   | 7474/10712 [1:22:07<26:28,  2.04it/s]
 70%|██████▉   | 7475/10712 [1:22:08<26:27,  2.04it/s]
                                                      
{'loss': 3.6564, 'grad_norm': 0.20911069214344025, 'learning_rate': 0.00025334712662211965, 'epoch': 0.7}

 70%|██████▉   | 7475/10712 [1:22:08<26:27,  2.04it/s]
 70%|██████▉   | 7476/10712 [1:22:08<26:27,  2.04it/s]
 70%|██████▉   | 7477/10712 [1:22:09<26:25,  2.04it/s]
 70%|██████▉   | 7478/10712 [1:22:09<26:25,  2.04it/s]
 70%|██████▉   | 7479/10712 [1:22:10<26:25,  2.04it/s]
 70%|██████▉   | 7480/10712 [1:22:10<26:24,  2.04it/s]
 70%|██████▉   | 7481/10712 [1:22:11<26:25,  2.04it/s]
 70%|██████▉   | 7482/10712 [1:22:11<26:26,  2.04it/s]
 70%|██████▉   | 7483/10712 [1:22:12<26:23,  2.04it/s]
 70%|██████▉   | 7484/10712 [1:22:12<26:23,  2.04it/s]
 70%|██████▉   | 7485/10712 [1:22:13<26:21,  2.04it/s]
 70%|██████▉   | 7486/10712 [1:22:13<26:22,  2.04it/s]
 70%|██████▉   | 7487/10712 [1:22:14<26:20,  2.04it/s]
 70%|██████▉   | 7488/10712 [1:22:14<26:19,  2.04it/s]
 70%|██████▉   | 7489/10712 [1:22:15<26:18,  2.04it/s]
 70%|██████▉   | 7490/10712 [1:22:15<26:19,  2.04it/s]
 70%|██████▉   | 7491/10712 [1:22:16<26:18,  2.04it/s]
 70%|██████▉   | 7492/10712 [1:22:16<26:17,  2.04it/s]
 70%|██████▉   | 7493/10712 [1:22:17<26:18,  2.04it/s]
 70%|██████▉   | 7494/10712 [1:22:17<26:18,  2.04it/s]
 70%|██████▉   | 7495/10712 [1:22:17<26:14,  2.04it/s]
 70%|██████▉   | 7496/10712 [1:22:18<26:14,  2.04it/s]
 70%|██████▉   | 7497/10712 [1:22:18<26:16,  2.04it/s]
 70%|██████▉   | 7498/10712 [1:22:19<26:17,  2.04it/s]
 70%|███████   | 7499/10712 [1:22:19<26:16,  2.04it/s]
 70%|███████   | 7500/10712 [1:22:20<26:15,  2.04it/s]
                                                      
{'loss': 3.6661, 'grad_norm': 0.21243244409561157, 'learning_rate': 0.0002498118701460503, 'epoch': 0.7}

 70%|███████   | 7500/10712 [1:22:20<26:15,  2.04it/s]
 70%|███████   | 7501/10712 [1:22:20<26:18,  2.03it/s]
 70%|███████   | 7502/10712 [1:22:21<26:15,  2.04it/s]
 70%|███████   | 7503/10712 [1:22:21<26:12,  2.04it/s]
 70%|███████   | 7504/10712 [1:22:22<26:10,  2.04it/s]
 70%|███████   | 7505/10712 [1:22:22<26:10,  2.04it/s]
 70%|███████   | 7506/10712 [1:22:23<26:12,  2.04it/s]
 70%|███████   | 7507/10712 [1:22:23<26:12,  2.04it/s]
 70%|███████   | 7508/10712 [1:22:24<26:12,  2.04it/s]
 70%|███████   | 7509/10712 [1:22:24<26:13,  2.04it/s]
 70%|███████   | 7510/10712 [1:22:25<26:13,  2.04it/s]
 70%|███████   | 7511/10712 [1:22:25<26:11,  2.04it/s]
 70%|███████   | 7512/10712 [1:22:26<26:09,  2.04it/s]
 70%|███████   | 7513/10712 [1:22:26<26:09,  2.04it/s]
 70%|███████   | 7514/10712 [1:22:27<26:09,  2.04it/s]
 70%|███████   | 7515/10712 [1:22:27<26:07,  2.04it/s]
 70%|███████   | 7516/10712 [1:22:28<26:05,  2.04it/s]
 70%|███████   | 7517/10712 [1:22:28<26:06,  2.04it/s]
 70%|███████   | 7518/10712 [1:22:29<26:06,  2.04it/s]
 70%|███████   | 7519/10712 [1:22:29<26:06,  2.04it/s]
 70%|███████   | 7520/10712 [1:22:30<26:04,  2.04it/s]
 70%|███████   | 7521/10712 [1:22:30<26:05,  2.04it/s]
 70%|███████   | 7522/10712 [1:22:31<26:05,  2.04it/s]
 70%|███████   | 7523/10712 [1:22:31<26:04,  2.04it/s]
 70%|███████   | 7524/10712 [1:22:32<26:03,  2.04it/s]
 70%|███████   | 7525/10712 [1:22:32<26:03,  2.04it/s]{'loss': 3.6553, 'grad_norm': 0.21632854640483856, 'learning_rate': 0.0002462932206243906, 'epoch': 0.7}

                                                      

 70%|███████   | 7525/10712 [1:22:32<26:03,  2.04it/s]
 70%|███████   | 7526/10712 [1:22:33<26:03,  2.04it/s]
 70%|███████   | 7527/10712 [1:22:33<26:00,  2.04it/s]
 70%|███████   | 7528/10712 [1:22:34<25:59,  2.04it/s]
 70%|███████   | 7529/10712 [1:22:34<26:00,  2.04it/s]
 70%|███████   | 7530/10712 [1:22:35<26:00,  2.04it/s]
 70%|███████   | 7531/10712 [1:22:35<25:59,  2.04it/s]
 70%|███████   | 7532/10712 [1:22:36<26:00,  2.04it/s]
 70%|███████   | 7533/10712 [1:22:36<25:57,  2.04it/s]
 70%|███████   | 7534/10712 [1:22:37<25:55,  2.04it/s]
 70%|███████   | 7535/10712 [1:22:37<25:56,  2.04it/s]
 70%|███████   | 7536/10712 [1:22:38<25:55,  2.04it/s]
 70%|███████   | 7537/10712 [1:22:38<25:55,  2.04it/s]
 70%|███████   | 7538/10712 [1:22:39<25:53,  2.04it/s]
 70%|███████   | 7539/10712 [1:22:39<25:53,  2.04it/s]
 70%|███████   | 7540/10712 [1:22:40<25:54,  2.04it/s]
 70%|███████   | 7541/10712 [1:22:40<25:53,  2.04it/s]
 70%|███████   | 7542/10712 [1:22:41<25:54,  2.04it/s]
 70%|███████   | 7543/10712 [1:22:41<25:54,  2.04it/s]
 70%|███████   | 7544/10712 [1:22:42<25:52,  2.04it/s]
 70%|███████   | 7545/10712 [1:22:42<25:49,  2.04it/s]
 70%|███████   | 7546/10712 [1:22:42<25:49,  2.04it/s]
 70%|███████   | 7547/10712 [1:22:43<25:52,  2.04it/s]
 70%|███████   | 7548/10712 [1:22:43<25:50,  2.04it/s]
 70%|███████   | 7549/10712 [1:22:44<25:47,  2.04it/s]
 70%|███████   | 7550/10712 [1:22:44<25:48,  2.04it/s]
                                                      
{'loss': 3.6518, 'grad_norm': 0.2012103796005249, 'learning_rate': 0.00024279141161759015, 'epoch': 0.7}

 70%|███████   | 7550/10712 [1:22:44<25:48,  2.04it/s]
 70%|███████   | 7551/10712 [1:22:45<25:50,  2.04it/s]
 71%|███████   | 7552/10712 [1:22:45<25:47,  2.04it/s]
 71%|███████   | 7553/10712 [1:22:46<25:47,  2.04it/s]
 71%|███████   | 7554/10712 [1:22:46<25:48,  2.04it/s]
 71%|███████   | 7555/10712 [1:22:47<25:47,  2.04it/s]
 71%|███████   | 7556/10712 [1:22:47<25:44,  2.04it/s]
 71%|███████   | 7557/10712 [1:22:48<25:46,  2.04it/s]
 71%|███████   | 7558/10712 [1:22:48<25:46,  2.04it/s]
 71%|███████   | 7559/10712 [1:22:49<25:43,  2.04it/s]
 71%|███████   | 7560/10712 [1:22:49<25:41,  2.05it/s]
 71%|███████   | 7561/10712 [1:22:50<25:41,  2.04it/s]
 71%|███████   | 7562/10712 [1:22:50<25:44,  2.04it/s]
 71%|███████   | 7563/10712 [1:22:51<25:43,  2.04it/s]
 71%|███████   | 7564/10712 [1:22:51<25:43,  2.04it/s]
 71%|███████   | 7565/10712 [1:22:52<25:43,  2.04it/s]
 71%|███████   | 7566/10712 [1:22:52<25:41,  2.04it/s]
 71%|███████   | 7567/10712 [1:22:53<25:40,  2.04it/s]
 71%|███████   | 7568/10712 [1:22:53<25:39,  2.04it/s]
 71%|███████   | 7569/10712 [1:22:54<25:40,  2.04it/s]
 71%|███████   | 7570/10712 [1:22:54<25:39,  2.04it/s]
 71%|███████   | 7571/10712 [1:22:55<25:37,  2.04it/s]
 71%|███████   | 7572/10712 [1:22:55<25:39,  2.04it/s]
 71%|███████   | 7573/10712 [1:22:56<25:40,  2.04it/s]
 71%|███████   | 7574/10712 [1:22:56<25:38,  2.04it/s]
 71%|███████   | 7575/10712 [1:22:57<25:39,  2.04it/s]{'loss': 3.6636, 'grad_norm': 0.2023627758026123, 'learning_rate': 0.00023930667556826173, 'epoch': 0.71}

                                                      

 71%|███████   | 7575/10712 [1:22:57<25:39,  2.04it/s]
 71%|███████   | 7576/10712 [1:22:57<25:40,  2.04it/s]
 71%|███████   | 7577/10712 [1:22:58<25:38,  2.04it/s]
 71%|███████   | 7578/10712 [1:22:58<25:36,  2.04it/s]
 71%|███████   | 7579/10712 [1:22:59<25:35,  2.04it/s]
 71%|███████   | 7580/10712 [1:22:59<25:36,  2.04it/s]
 71%|███████   | 7581/10712 [1:23:00<25:34,  2.04it/s]
 71%|███████   | 7582/10712 [1:23:00<25:35,  2.04it/s]
 71%|███████   | 7583/10712 [1:23:01<25:34,  2.04it/s]
 71%|███████   | 7584/10712 [1:23:01<25:33,  2.04it/s]
 71%|███████   | 7585/10712 [1:23:02<25:31,  2.04it/s]
 71%|███████   | 7586/10712 [1:23:02<25:29,  2.04it/s]
 71%|███████   | 7587/10712 [1:23:03<25:30,  2.04it/s]
 71%|███████   | 7588/10712 [1:23:03<25:31,  2.04it/s]
 71%|███████   | 7589/10712 [1:23:04<25:29,  2.04it/s]
 71%|███████   | 7590/10712 [1:23:04<25:28,  2.04it/s]
 71%|███████   | 7591/10712 [1:23:05<25:29,  2.04it/s]
 71%|███████   | 7592/10712 [1:23:05<25:30,  2.04it/s]
 71%|███████   | 7593/10712 [1:23:06<25:28,  2.04it/s]
 71%|███████   | 7594/10712 [1:23:06<25:27,  2.04it/s]
 71%|███████   | 7595/10712 [1:23:06<25:28,  2.04it/s]
 71%|███████   | 7596/10712 [1:23:07<25:28,  2.04it/s]
 71%|███████   | 7597/10712 [1:23:07<25:26,  2.04it/s]
 71%|███████   | 7598/10712 [1:23:08<25:25,  2.04it/s]
 71%|███████   | 7599/10712 [1:23:08<25:26,  2.04it/s]
 71%|███████   | 7600/10712 [1:23:09<25:24,  2.04it/s]{'loss': 3.6584, 'grad_norm': 0.20237842202186584, 'learning_rate': 0.00023583924378575154, 'epoch': 0.71}
                                                      

 71%|███████   | 7600/10712 [1:23:09<25:24,  2.04it/s]
 71%|███████   | 7601/10712 [1:23:09<25:24,  2.04it/s]
 71%|███████   | 7602/10712 [1:23:10<25:25,  2.04it/s]
 71%|███████   | 7603/10712 [1:23:10<25:24,  2.04it/s]
 71%|███████   | 7604/10712 [1:23:11<25:22,  2.04it/s]
 71%|███████   | 7605/10712 [1:23:11<25:21,  2.04it/s]
 71%|███████   | 7606/10712 [1:23:12<25:20,  2.04it/s]
 71%|███████   | 7607/10712 [1:23:12<25:21,  2.04it/s]
 71%|███████   | 7608/10712 [1:23:13<25:21,  2.04it/s]
 71%|███████   | 7609/10712 [1:23:13<25:19,  2.04it/s]
 71%|███████   | 7610/10712 [1:23:14<25:19,  2.04it/s]
 71%|███████   | 7611/10712 [1:23:14<25:20,  2.04it/s]
 71%|███████   | 7612/10712 [1:23:15<25:18,  2.04it/s]
 71%|███████   | 7613/10712 [1:23:15<25:18,  2.04it/s]
 71%|███████   | 7614/10712 [1:23:16<25:18,  2.04it/s]
 71%|███████   | 7615/10712 [1:23:16<25:18,  2.04it/s]
 71%|███████   | 7616/10712 [1:23:17<25:16,  2.04it/s]
 71%|███████   | 7617/10712 [1:23:17<25:16,  2.04it/s]
 71%|███████   | 7618/10712 [1:23:18<25:17,  2.04it/s]
 71%|███████   | 7619/10712 [1:23:18<25:15,  2.04it/s]
 71%|███████   | 7620/10712 [1:23:19<25:13,  2.04it/s]
 71%|███████   | 7621/10712 [1:23:19<25:13,  2.04it/s]
 71%|███████   | 7622/10712 [1:23:20<25:14,  2.04it/s]
 71%|███████   | 7623/10712 [1:23:20<25:13,  2.04it/s]
 71%|███████   | 7624/10712 [1:23:21<25:11,  2.04it/s]
 71%|███████   | 7625/10712 [1:23:21<25:12,  2.04it/s]
                                                      {'loss': 3.6604, 'grad_norm': 0.21091046929359436, 'learning_rate': 0.00023238934643078497, 'epoch': 0.71}

 71%|███████   | 7625/10712 [1:23:21<25:12,  2.04it/s]
 71%|███████   | 7626/10712 [1:23:22<25:13,  2.04it/s]
 71%|███████   | 7627/10712 [1:23:22<25:11,  2.04it/s]
 71%|███████   | 7628/10712 [1:23:23<25:10,  2.04it/s]
 71%|███████   | 7629/10712 [1:23:23<25:11,  2.04it/s]
 71%|███████   | 7630/10712 [1:23:24<25:10,  2.04it/s]
 71%|███████   | 7631/10712 [1:23:24<25:09,  2.04it/s]
 71%|███████   | 7632/10712 [1:23:25<25:08,  2.04it/s]
 71%|███████▏  | 7633/10712 [1:23:25<25:10,  2.04it/s]
 71%|███████▏  | 7634/10712 [1:23:26<25:08,  2.04it/s]
 71%|███████▏  | 7635/10712 [1:23:26<25:07,  2.04it/s]
 71%|███████▏  | 7636/10712 [1:23:27<25:08,  2.04it/s]
 71%|███████▏  | 7637/10712 [1:23:27<25:08,  2.04it/s]
 71%|███████▏  | 7638/10712 [1:23:28<25:07,  2.04it/s]
 71%|███████▏  | 7639/10712 [1:23:28<25:05,  2.04it/s]
 71%|███████▏  | 7640/10712 [1:23:29<25:03,  2.04it/s]
 71%|███████▏  | 7641/10712 [1:23:29<25:03,  2.04it/s]
 71%|███████▏  | 7642/10712 [1:23:30<25:04,  2.04it/s]
 71%|███████▏  | 7643/10712 [1:23:30<25:03,  2.04it/s]
 71%|███████▏  | 7644/10712 [1:23:31<25:02,  2.04it/s]
 71%|███████▏  | 7645/10712 [1:23:31<25:00,  2.04it/s]
 71%|███████▏  | 7646/10712 [1:23:31<24:59,  2.04it/s]
 71%|███████▏  | 7647/10712 [1:23:32<25:01,  2.04it/s]
 71%|███████▏  | 7648/10712 [1:23:32<25:01,  2.04it/s]
 71%|███████▏  | 7649/10712 [1:23:33<24:59,  2.04it/s]
 71%|███████▏  | 7650/10712 [1:23:33<24:58,  2.04it/s]{'loss': 3.6492, 'grad_norm': 0.20849977433681488, 'learning_rate': 0.0002289572125001901, 'epoch': 0.71}

                                                      

 71%|███████▏  | 7650/10712 [1:23:33<24:58,  2.04it/s]
 71%|███████▏  | 7651/10712 [1:23:34<25:01,  2.04it/s]
 71%|███████▏  | 7652/10712 [1:23:34<25:01,  2.04it/s]
 71%|███████▏  | 7653/10712 [1:23:35<25:00,  2.04it/s]
 71%|███████▏  | 7654/10712 [1:23:35<24:58,  2.04it/s]
 71%|███████▏  | 7655/10712 [1:23:36<24:59,  2.04it/s]
 71%|███████▏  | 7656/10712 [1:23:36<24:58,  2.04it/s]
 71%|███████▏  | 7657/10712 [1:23:37<24:56,  2.04it/s]
 71%|███████▏  | 7658/10712 [1:23:37<24:55,  2.04it/s]
 71%|███████▏  | 7659/10712 [1:23:38<24:54,  2.04it/s]
 72%|███████▏  | 7660/10712 [1:23:38<24:54,  2.04it/s]
 72%|███████▏  | 7661/10712 [1:23:39<24:51,  2.05it/s]
 72%|███████▏  | 7662/10712 [1:23:39<24:50,  2.05it/s]
 72%|███████▏  | 7663/10712 [1:23:40<24:51,  2.04it/s]
 72%|███████▏  | 7664/10712 [1:23:40<24:52,  2.04it/s]
 72%|███████▏  | 7665/10712 [1:23:41<24:52,  2.04it/s]
 72%|███████▏  | 7666/10712 [1:23:41<24:50,  2.04it/s]
 72%|███████▏  | 7667/10712 [1:23:42<24:52,  2.04it/s]
 72%|███████▏  | 7668/10712 [1:23:42<24:52,  2.04it/s]
 72%|███████▏  | 7669/10712 [1:23:43<24:51,  2.04it/s]
 72%|███████▏  | 7670/10712 [1:23:43<24:51,  2.04it/s]
 72%|███████▏  | 7671/10712 [1:23:44<24:51,  2.04it/s]
 72%|███████▏  | 7672/10712 [1:23:44<24:49,  2.04it/s]
 72%|███████▏  | 7673/10712 [1:23:45<24:48,  2.04it/s]
 72%|███████▏  | 7674/10712 [1:23:45<24:48,  2.04it/s]
 72%|███████▏  | 7675/10712 [1:23:46<24:48,  2.04it/s]
                                                      {'loss': 3.648, 'grad_norm': 0.20618891716003418, 'learning_rate': 0.00022554306981169614, 'epoch': 0.72}


 72%|███████▏  | 7675/10712 [1:23:46<24:48,  2.04it/s]
 72%|███████▏  | 7676/10712 [1:23:46<24:48,  2.04it/s]
 72%|███████▏  | 7677/10712 [1:23:47<24:46,  2.04it/s]
 72%|███████▏  | 7678/10712 [1:23:47<24:48,  2.04it/s]
 72%|███████▏  | 7679/10712 [1:23:48<24:47,  2.04it/s]
 72%|███████▏  | 7680/10712 [1:23:48<24:46,  2.04it/s]
 72%|███████▏  | 7681/10712 [1:23:49<24:46,  2.04it/s]
 72%|███████▏  | 7682/10712 [1:23:49<24:45,  2.04it/s]
 72%|███████▏  | 7683/10712 [1:23:50<24:44,  2.04it/s]
 72%|███████▏  | 7684/10712 [1:23:50<24:42,  2.04it/s]
 72%|███████▏  | 7685/10712 [1:23:51<24:42,  2.04it/s]
 72%|███████▏  | 7686/10712 [1:23:51<24:42,  2.04it/s]
 72%|███████▏  | 7687/10712 [1:23:52<24:41,  2.04it/s]
 72%|███████▏  | 7688/10712 [1:23:52<24:43,  2.04it/s]
 72%|███████▏  | 7689/10712 [1:23:53<24:41,  2.04it/s]
 72%|███████▏  | 7690/10712 [1:23:53<24:40,  2.04it/s]
 72%|███████▏  | 7691/10712 [1:23:54<24:38,  2.04it/s]
 72%|███████▏  | 7692/10712 [1:23:54<24:40,  2.04it/s]
 72%|███████▏  | 7693/10712 [1:23:55<24:39,  2.04it/s]
 72%|███████▏  | 7694/10712 [1:23:55<24:37,  2.04it/s]
 72%|███████▏  | 7695/10712 [1:23:55<24:39,  2.04it/s]
 72%|███████▏  | 7696/10712 [1:23:56<24:39,  2.04it/s]
 72%|███████▏  | 7697/10712 [1:23:56<24:38,  2.04it/s]
 72%|███████▏  | 7698/10712 [1:23:57<24:37,  2.04it/s]
 72%|███████▏  | 7699/10712 [1:23:57<24:37,  2.04it/s]
 72%|███████▏  | 7700/10712 [1:23:58<24:36,  2.04it/s]{'loss': 3.6489, 'grad_norm': 0.21610340476036072, 'learning_rate': 0.00022214714498881283, 'epoch': 0.72}
                                                      

 72%|███████▏  | 7700/10712 [1:23:58<24:36,  2.04it/s]
 72%|███████▏  | 7701/10712 [1:23:58<24:37,  2.04it/s]
 72%|███████▏  | 7702/10712 [1:23:59<24:36,  2.04it/s]
 72%|███████▏  | 7703/10712 [1:23:59<24:36,  2.04it/s]
 72%|███████▏  | 7704/10712 [1:24:00<24:34,  2.04it/s]
 72%|███████▏  | 7705/10712 [1:24:00<24:34,  2.04it/s]
 72%|███████▏  | 7706/10712 [1:24:01<24:34,  2.04it/s]
 72%|███████▏  | 7707/10712 [1:24:01<24:31,  2.04it/s]
 72%|███████▏  | 7708/10712 [1:24:02<24:31,  2.04it/s]
 72%|███████▏  | 7709/10712 [1:24:02<24:30,  2.04it/s]
 72%|███████▏  | 7710/10712 [1:24:03<24:30,  2.04it/s]
 72%|███████▏  | 7711/10712 [1:24:03<24:31,  2.04it/s]
 72%|███████▏  | 7712/10712 [1:24:04<24:29,  2.04it/s]
 72%|███████▏  | 7713/10712 [1:24:04<24:29,  2.04it/s]
 72%|███████▏  | 7714/10712 [1:24:05<24:29,  2.04it/s]
 72%|███████▏  | 7715/10712 [1:24:05<24:28,  2.04it/s]
 72%|███████▏  | 7716/10712 [1:24:06<24:26,  2.04it/s]
 72%|███████▏  | 7717/10712 [1:24:06<24:25,  2.04it/s]
 72%|███████▏  | 7718/10712 [1:24:07<24:27,  2.04it/s]
 72%|███████▏  | 7719/10712 [1:24:07<24:26,  2.04it/s]
 72%|███████▏  | 7720/10712 [1:24:08<24:23,  2.04it/s]
 72%|███████▏  | 7721/10712 [1:24:08<24:23,  2.04it/s]
 72%|███████▏  | 7722/10712 [1:24:09<24:25,  2.04it/s]
 72%|███████▏  | 7723/10712 [1:24:09<24:25,  2.04it/s]
 72%|███████▏  | 7724/10712 [1:24:10<24:24,  2.04it/s]
 72%|███████▏  | 7725/10712 [1:24:10<24:24,  2.04it/s]
                                                      
{'loss': 3.6531, 'grad_norm': 0.2155597060918808, 'learning_rate': 0.00021876966344578608, 'epoch': 0.72}

 72%|███████▏  | 7725/10712 [1:24:10<24:24,  2.04it/s]
 72%|███████▏  | 7726/10712 [1:24:11<24:25,  2.04it/s]
 72%|███████▏  | 7727/10712 [1:24:11<24:23,  2.04it/s]
 72%|███████▏  | 7728/10712 [1:24:12<24:22,  2.04it/s]
 72%|███████▏  | 7729/10712 [1:24:12<24:21,  2.04it/s]
 72%|███████▏  | 7730/10712 [1:24:13<24:21,  2.04it/s]
 72%|███████▏  | 7731/10712 [1:24:13<24:21,  2.04it/s]
 72%|███████▏  | 7732/10712 [1:24:14<24:19,  2.04it/s]
 72%|███████▏  | 7733/10712 [1:24:14<24:19,  2.04it/s]
 72%|███████▏  | 7734/10712 [1:24:15<24:20,  2.04it/s]
 72%|███████▏  | 7735/10712 [1:24:15<24:18,  2.04it/s]
 72%|███████▏  | 7736/10712 [1:24:16<24:17,  2.04it/s]
 72%|███████▏  | 7737/10712 [1:24:16<24:18,  2.04it/s]
 72%|███████▏  | 7738/10712 [1:24:17<24:17,  2.04it/s]
 72%|███████▏  | 7739/10712 [1:24:17<24:16,  2.04it/s]
 72%|███████▏  | 7740/10712 [1:24:18<24:14,  2.04it/s]
 72%|███████▏  | 7741/10712 [1:24:18<24:15,  2.04it/s]
 72%|███████▏  | 7742/10712 [1:24:19<24:15,  2.04it/s]
 72%|███████▏  | 7743/10712 [1:24:19<24:14,  2.04it/s]
 72%|███████▏  | 7744/10712 [1:24:20<24:13,  2.04it/s]
 72%|███████▏  | 7745/10712 [1:24:20<24:12,  2.04it/s]
 72%|███████▏  | 7746/10712 [1:24:20<24:13,  2.04it/s]
 72%|███████▏  | 7747/10712 [1:24:21<24:12,  2.04it/s]
 72%|███████▏  | 7748/10712 [1:24:21<24:10,  2.04it/s]
 72%|███████▏  | 7749/10712 [1:24:22<24:11,  2.04it/s]
 72%|███████▏  | 7750/10712 [1:24:22<24:10,  2.04it/s]
                                                      
{'loss': 3.6447, 'grad_norm': 0.2080143839120865, 'learning_rate': 0.00021541084937263684, 'epoch': 0.72}

 72%|███████▏  | 7750/10712 [1:24:22<24:10,  2.04it/s]
 72%|███████▏  | 7751/10712 [1:24:23<24:10,  2.04it/s]
 72%|███████▏  | 7752/10712 [1:24:23<24:09,  2.04it/s]
 72%|███████▏  | 7753/10712 [1:24:24<24:11,  2.04it/s]
 72%|███████▏  | 7754/10712 [1:24:24<24:11,  2.04it/s]
 72%|███████▏  | 7755/10712 [1:24:25<24:08,  2.04it/s]
 72%|███████▏  | 7756/10712 [1:24:25<24:08,  2.04it/s]
 72%|███████▏  | 7757/10712 [1:24:26<24:08,  2.04it/s]
 72%|███████▏  | 7758/10712 [1:24:26<24:06,  2.04it/s]
 72%|███████▏  | 7759/10712 [1:24:27<24:05,  2.04it/s]
 72%|███████▏  | 7760/10712 [1:24:27<24:05,  2.04it/s]
 72%|███████▏  | 7761/10712 [1:24:28<24:05,  2.04it/s]
 72%|███████▏  | 7762/10712 [1:24:28<24:05,  2.04it/s]
 72%|███████▏  | 7763/10712 [1:24:29<24:03,  2.04it/s]
 72%|███████▏  | 7764/10712 [1:24:29<24:05,  2.04it/s]
 72%|███████▏  | 7765/10712 [1:24:30<24:04,  2.04it/s]
 72%|███████▏  | 7766/10712 [1:24:30<24:02,  2.04it/s]
 73%|███████▎  | 7767/10712 [1:24:31<24:00,  2.04it/s]
 73%|███████▎  | 7768/10712 [1:24:31<24:01,  2.04it/s]
 73%|███████▎  | 7769/10712 [1:24:32<24:01,  2.04it/s]
 73%|███████▎  | 7770/10712 [1:24:32<23:59,  2.04it/s]
 73%|███████▎  | 7771/10712 [1:24:33<23:59,  2.04it/s]
 73%|███████▎  | 7772/10712 [1:24:33<24:01,  2.04it/s]
 73%|███████▎  | 7773/10712 [1:24:34<24:00,  2.04it/s]
 73%|███████▎  | 7774/10712 [1:24:34<23:59,  2.04it/s]
 73%|███████▎  | 7775/10712 [1:24:35<23:59,  2.04it/s]{'loss': 3.6465, 'grad_norm': 0.19942878186702728, 'learning_rate': 0.00021207092572027864, 'epoch': 0.73}

                                                      

 73%|███████▎  | 7775/10712 [1:24:35<23:59,  2.04it/s]
 73%|███████▎  | 7776/10712 [1:24:35<23:59,  2.04it/s]
 73%|███████▎  | 7777/10712 [1:24:36<23:59,  2.04it/s]
 73%|███████▎  | 7778/10712 [1:24:36<23:58,  2.04it/s]
 73%|███████▎  | 7779/10712 [1:24:37<23:58,  2.04it/s]
 73%|███████▎  | 7780/10712 [1:24:37<23:57,  2.04it/s]
 73%|███████▎  | 7781/10712 [1:24:38<23:56,  2.04it/s]
 73%|███████▎  | 7782/10712 [1:24:38<23:57,  2.04it/s]
 73%|███████▎  | 7783/10712 [1:24:39<23:55,  2.04it/s]
 73%|███████▎  | 7784/10712 [1:24:39<23:53,  2.04it/s]
 73%|███████▎  | 7785/10712 [1:24:40<23:52,  2.04it/s]
 73%|███████▎  | 7786/10712 [1:24:40<23:53,  2.04it/s]
 73%|███████▎  | 7787/10712 [1:24:41<23:54,  2.04it/s]
 73%|███████▎  | 7788/10712 [1:24:41<23:52,  2.04it/s]
 73%|███████▎  | 7789/10712 [1:24:42<23:51,  2.04it/s]
 73%|███████▎  | 7790/10712 [1:24:42<23:52,  2.04it/s]
 73%|███████▎  | 7791/10712 [1:24:43<23:51,  2.04it/s]
 73%|███████▎  | 7792/10712 [1:24:43<23:49,  2.04it/s]
 73%|███████▎  | 7793/10712 [1:24:44<23:48,  2.04it/s]
 73%|███████▎  | 7794/10712 [1:24:44<23:48,  2.04it/s]
 73%|███████▎  | 7795/10712 [1:24:44<23:50,  2.04it/s]
 73%|███████▎  | 7796/10712 [1:24:45<23:48,  2.04it/s]
 73%|███████▎  | 7797/10712 [1:24:45<23:48,  2.04it/s]
 73%|███████▎  | 7798/10712 [1:24:46<23:48,  2.04it/s]
 73%|███████▎  | 7799/10712 [1:24:46<23:48,  2.04it/s]
 73%|███████▎  | 7800/10712 [1:24:47<23:46,  2.04it/s]{'loss': 3.6409, 'grad_norm': 0.20163676142692566, 'learning_rate': 0.00020875011418571983, 'epoch': 0.73}

                                                      

 73%|███████▎  | 7800/10712 [1:24:47<23:46,  2.04it/s]
 73%|███████▎  | 7801/10712 [1:24:47<23:47,  2.04it/s]
 73%|███████▎  | 7802/10712 [1:24:48<23:47,  2.04it/s]
 73%|███████▎  | 7803/10712 [1:24:48<23:45,  2.04it/s]
 73%|███████▎  | 7804/10712 [1:24:49<23:47,  2.04it/s]
 73%|███████▎  | 7805/10712 [1:24:49<23:47,  2.04it/s]
 73%|███████▎  | 7806/10712 [1:24:50<23:45,  2.04it/s]
 73%|███████▎  | 7807/10712 [1:24:50<23:44,  2.04it/s]
 73%|███████▎  | 7808/10712 [1:24:51<23:44,  2.04it/s]
 73%|███████▎  | 7809/10712 [1:24:51<23:43,  2.04it/s]
 73%|███████▎  | 7810/10712 [1:24:52<23:41,  2.04it/s]
 73%|███████▎  | 7811/10712 [1:24:52<23:40,  2.04it/s]
 73%|███████▎  | 7812/10712 [1:24:53<23:41,  2.04it/s]
 73%|███████▎  | 7813/10712 [1:24:53<23:40,  2.04it/s]
 73%|███████▎  | 7814/10712 [1:24:54<23:38,  2.04it/s]
 73%|███████▎  | 7815/10712 [1:24:54<23:38,  2.04it/s]
 73%|███████▎  | 7816/10712 [1:24:55<23:38,  2.04it/s]
 73%|███████▎  | 7817/10712 [1:24:55<23:39,  2.04it/s]
 73%|███████▎  | 7818/10712 [1:24:56<23:37,  2.04it/s]
 73%|███████▎  | 7819/10712 [1:24:56<23:37,  2.04it/s]
 73%|███████▎  | 7820/10712 [1:24:57<23:38,  2.04it/s]
 73%|███████▎  | 7821/10712 [1:24:57<23:37,  2.04it/s]
 73%|███████▎  | 7822/10712 [1:24:58<23:37,  2.04it/s]
 73%|███████▎  | 7823/10712 [1:24:58<23:38,  2.04it/s]
 73%|███████▎  | 7824/10712 [1:24:59<23:37,  2.04it/s]
 73%|███████▎  | 7825/10712 [1:24:59<23:35,  2.04it/s]
                                                      
{'loss': 3.6372, 'grad_norm': 0.21982909739017487, 'learning_rate': 0.00020544863519734647, 'epoch': 0.73}

 73%|███████▎  | 7825/10712 [1:24:59<23:35,  2.04it/s]
 73%|███████▎  | 7826/10712 [1:25:00<23:37,  2.04it/s]
 73%|███████▎  | 7827/10712 [1:25:00<23:36,  2.04it/s]
 73%|███████▎  | 7828/10712 [1:25:01<23:34,  2.04it/s]
 73%|███████▎  | 7829/10712 [1:25:01<23:33,  2.04it/s]
 73%|███████▎  | 7830/10712 [1:25:02<23:33,  2.04it/s]
 73%|███████▎  | 7831/10712 [1:25:02<23:32,  2.04it/s]
 73%|███████▎  | 7832/10712 [1:25:03<23:31,  2.04it/s]
 73%|███████▎  | 7833/10712 [1:25:03<23:31,  2.04it/s]
 73%|███████▎  | 7834/10712 [1:25:04<23:30,  2.04it/s]
 73%|███████▎  | 7835/10712 [1:25:04<23:29,  2.04it/s]
 73%|███████▎  | 7836/10712 [1:25:05<23:28,  2.04it/s]
 73%|███████▎  | 7837/10712 [1:25:05<23:30,  2.04it/s]
 73%|███████▎  | 7838/10712 [1:25:06<23:28,  2.04it/s]
 73%|███████▎  | 7839/10712 [1:25:06<23:27,  2.04it/s]
 73%|███████▎  | 7840/10712 [1:25:07<23:27,  2.04it/s]
 73%|███████▎  | 7841/10712 [1:25:07<23:27,  2.04it/s]
 73%|███████▎  | 7842/10712 [1:25:08<23:26,  2.04it/s]
 73%|███████▎  | 7843/10712 [1:25:08<23:25,  2.04it/s]
 73%|███████▎  | 7844/10712 [1:25:09<23:26,  2.04it/s]
 73%|███████▎  | 7845/10712 [1:25:09<23:25,  2.04it/s]
 73%|███████▎  | 7846/10712 [1:25:09<23:23,  2.04it/s]
 73%|███████▎  | 7847/10712 [1:25:10<23:23,  2.04it/s]
 73%|███████▎  | 7848/10712 [1:25:10<23:23,  2.04it/s]
 73%|███████▎  | 7849/10712 [1:25:11<23:23,  2.04it/s]
 73%|███████▎  | 7850/10712 [1:25:11<23:21,  2.04it/s]
                                                      
{'loss': 3.6382, 'grad_norm': 0.21491530537605286, 'learning_rate': 0.000202166707900292, 'epoch': 0.73}

 73%|███████▎  | 7850/10712 [1:25:11<23:21,  2.04it/s]
 73%|███████▎  | 7851/10712 [1:25:12<23:22,  2.04it/s]
 73%|███████▎  | 7852/10712 [1:25:12<23:22,  2.04it/s]
 73%|███████▎  | 7853/10712 [1:25:13<23:21,  2.04it/s]
 73%|███████▎  | 7854/10712 [1:25:13<23:19,  2.04it/s]
 73%|███████▎  | 7855/10712 [1:25:14<23:19,  2.04it/s]
 73%|███████▎  | 7856/10712 [1:25:14<23:19,  2.04it/s]
 73%|███████▎  | 7857/10712 [1:25:15<23:20,  2.04it/s]
 73%|███████▎  | 7858/10712 [1:25:15<23:17,  2.04it/s]
 73%|███████▎  | 7859/10712 [1:25:16<23:16,  2.04it/s]
 73%|███████▎  | 7860/10712 [1:25:16<23:15,  2.04it/s]
 73%|███████▎  | 7861/10712 [1:25:17<23:16,  2.04it/s]
 73%|███████▎  | 7862/10712 [1:25:17<23:15,  2.04it/s]
 73%|███████▎  | 7863/10712 [1:25:18<23:13,  2.04it/s]
 73%|███████▎  | 7864/10712 [1:25:18<23:13,  2.04it/s]
 73%|███████▎  | 7865/10712 [1:25:19<23:15,  2.04it/s]
 73%|███████▎  | 7866/10712 [1:25:19<23:14,  2.04it/s]
 73%|███████▎  | 7867/10712 [1:25:20<23:12,  2.04it/s]
 73%|███████▎  | 7868/10712 [1:25:20<23:10,  2.05it/s]
 73%|███████▎  | 7869/10712 [1:25:21<23:12,  2.04it/s]
 73%|███████▎  | 7870/10712 [1:25:21<23:13,  2.04it/s]
 73%|███████▎  | 7871/10712 [1:25:22<23:11,  2.04it/s]
 73%|███████▎  | 7872/10712 [1:25:22<23:10,  2.04it/s]
 73%|███████▎  | 7873/10712 [1:25:23<23:11,  2.04it/s]
 74%|███████▎  | 7874/10712 [1:25:23<23:10,  2.04it/s]
 74%|███████▎  | 7875/10712 [1:25:24<23:09,  2.04it/s]{'loss': 3.6357, 'grad_norm': 0.2024458944797516, 'learning_rate': 0.0001989045501418899, 'epoch': 0.74}

                                                      

 74%|███████▎  | 7875/10712 [1:25:24<23:09,  2.04it/s]
 74%|███████▎  | 7876/10712 [1:25:24<23:10,  2.04it/s]
 74%|███████▎  | 7877/10712 [1:25:25<23:10,  2.04it/s]
 74%|███████▎  | 7878/10712 [1:25:25<23:08,  2.04it/s]
 74%|███████▎  | 7879/10712 [1:25:26<23:08,  2.04it/s]
 74%|███████▎  | 7880/10712 [1:25:26<23:08,  2.04it/s]
 74%|███████▎  | 7881/10712 [1:25:27<23:08,  2.04it/s]
 74%|███████▎  | 7882/10712 [1:25:27<23:07,  2.04it/s]
 74%|███████▎  | 7883/10712 [1:25:28<23:07,  2.04it/s]
 74%|███████▎  | 7884/10712 [1:25:28<23:06,  2.04it/s]
 74%|███████▎  | 7885/10712 [1:25:29<23:04,  2.04it/s]
 74%|███████▎  | 7886/10712 [1:25:29<23:05,  2.04it/s]
 74%|███████▎  | 7887/10712 [1:25:30<23:04,  2.04it/s]
 74%|███████▎  | 7888/10712 [1:25:30<23:04,  2.04it/s]
 74%|███████▎  | 7889/10712 [1:25:31<23:02,  2.04it/s]
 74%|███████▎  | 7890/10712 [1:25:31<23:01,  2.04it/s]
 74%|███████▎  | 7891/10712 [1:25:32<23:01,  2.04it/s]
 74%|███████▎  | 7892/10712 [1:25:32<23:01,  2.04it/s]
 74%|███████▎  | 7893/10712 [1:25:33<23:00,  2.04it/s]
 74%|███████▎  | 7894/10712 [1:25:33<22:59,  2.04it/s]
 74%|███████▎  | 7895/10712 [1:25:34<22:58,  2.04it/s]
 74%|███████▎  | 7896/10712 [1:25:34<22:58,  2.04it/s]
 74%|███████▎  | 7897/10712 [1:25:34<23:00,  2.04it/s]
 74%|███████▎  | 7898/10712 [1:25:35<22:59,  2.04it/s]
 74%|███████▎  | 7899/10712 [1:25:35<22:58,  2.04it/s]
 74%|███████▎  | 7900/10712 [1:25:36<22:58,  2.04it/s]{'loss': 3.6458, 'grad_norm': 0.20437031984329224, 'learning_rate': 0.00019566237845721407, 'epoch': 0.74}
                                                      

 74%|███████▎  | 7900/10712 [1:25:36<22:58,  2.04it/s]
 74%|███████▍  | 7901/10712 [1:25:36<23:01,  2.03it/s]
 74%|███████▍  | 7902/10712 [1:25:37<22:59,  2.04it/s]
 74%|███████▍  | 7903/10712 [1:25:37<22:58,  2.04it/s]
 74%|███████▍  | 7904/10712 [1:25:38<22:57,  2.04it/s]
 74%|███████▍  | 7905/10712 [1:25:38<22:56,  2.04it/s]
 74%|███████▍  | 7906/10712 [1:25:39<22:53,  2.04it/s]
 74%|███████▍  | 7907/10712 [1:25:39<22:54,  2.04it/s]
 74%|███████▍  | 7908/10712 [1:25:40<22:53,  2.04it/s]
 74%|███████▍  | 7909/10712 [1:25:40<22:51,  2.04it/s]
 74%|███████▍  | 7910/10712 [1:25:41<22:49,  2.05it/s]
 74%|███████▍  | 7911/10712 [1:25:41<22:50,  2.04it/s]
 74%|███████▍  | 7912/10712 [1:25:42<22:51,  2.04it/s]
 74%|███████▍  | 7913/10712 [1:25:42<22:50,  2.04it/s]
 74%|███████▍  | 7914/10712 [1:25:43<22:50,  2.04it/s]
 74%|███████▍  | 7915/10712 [1:25:43<22:48,  2.04it/s]
 74%|███████▍  | 7916/10712 [1:25:44<22:48,  2.04it/s]
 74%|███████▍  | 7917/10712 [1:25:44<22:49,  2.04it/s]
 74%|███████▍  | 7918/10712 [1:25:45<22:49,  2.04it/s]
 74%|███████▍  | 7919/10712 [1:25:45<22:48,  2.04it/s]
 74%|███████▍  | 7920/10712 [1:25:46<22:47,  2.04it/s]
 74%|███████▍  | 7921/10712 [1:25:46<22:47,  2.04it/s]
 74%|███████▍  | 7922/10712 [1:25:47<22:47,  2.04it/s]
 74%|███████▍  | 7923/10712 [1:25:47<22:46,  2.04it/s]
 74%|███████▍  | 7924/10712 [1:25:48<22:46,  2.04it/s]
 74%|███████▍  | 7925/10712 [1:25:48<22:44,  2.04it/s]
                                                      
{'loss': 3.6444, 'grad_norm': 0.2059612274169922, 'learning_rate': 0.00019244040805470552, 'epoch': 0.74}

 74%|███████▍  | 7925/10712 [1:25:48<22:44,  2.04it/s]
 74%|███████▍  | 7926/10712 [1:25:49<22:45,  2.04it/s]
 74%|███████▍  | 7927/10712 [1:25:49<22:46,  2.04it/s]
 74%|███████▍  | 7928/10712 [1:25:50<22:45,  2.04it/s]
 74%|███████▍  | 7929/10712 [1:25:50<22:43,  2.04it/s]
 74%|███████▍  | 7930/10712 [1:25:51<22:42,  2.04it/s]
 74%|███████▍  | 7931/10712 [1:25:51<22:44,  2.04it/s]
 74%|███████▍  | 7932/10712 [1:25:52<22:43,  2.04it/s]
 74%|███████▍  | 7933/10712 [1:25:52<22:43,  2.04it/s]
 74%|███████▍  | 7934/10712 [1:25:53<22:42,  2.04it/s]
 74%|███████▍  | 7935/10712 [1:25:53<22:41,  2.04it/s]
 74%|███████▍  | 7936/10712 [1:25:54<22:40,  2.04it/s]
 74%|███████▍  | 7937/10712 [1:25:54<22:38,  2.04it/s]
 74%|███████▍  | 7938/10712 [1:25:55<22:39,  2.04it/s]
 74%|███████▍  | 7939/10712 [1:25:55<22:39,  2.04it/s]
 74%|███████▍  | 7940/10712 [1:25:56<22:38,  2.04it/s]
 74%|███████▍  | 7941/10712 [1:25:56<22:37,  2.04it/s]
 74%|███████▍  | 7942/10712 [1:25:57<22:36,  2.04it/s]
 74%|███████▍  | 7943/10712 [1:25:57<22:36,  2.04it/s]
 74%|███████▍  | 7944/10712 [1:25:58<22:35,  2.04it/s]
 74%|███████▍  | 7945/10712 [1:25:58<22:34,  2.04it/s]
 74%|███████▍  | 7946/10712 [1:25:58<22:37,  2.04it/s]
 74%|███████▍  | 7947/10712 [1:25:59<22:36,  2.04it/s]
 74%|███████▍  | 7948/10712 [1:25:59<22:34,  2.04it/s]
 74%|███████▍  | 7949/10712 [1:26:00<22:33,  2.04it/s]
 74%|███████▍  | 7950/10712 [1:26:00<22:34,  2.04it/s]{'loss': 3.6461, 'grad_norm': 0.20775793492794037, 'learning_rate': 0.0001892388528018869, 'epoch': 0.74}
                                                      

 74%|███████▍  | 7950/10712 [1:26:00<22:34,  2.04it/s]
 74%|███████▍  | 7951/10712 [1:26:01<22:35,  2.04it/s]
 74%|███████▍  | 7952/10712 [1:26:01<22:32,  2.04it/s]
 74%|███████▍  | 7953/10712 [1:26:02<22:33,  2.04it/s]
 74%|███████▍  | 7954/10712 [1:26:02<22:33,  2.04it/s]
 74%|███████▍  | 7955/10712 [1:26:03<22:31,  2.04it/s]
 74%|███████▍  | 7956/10712 [1:26:03<22:31,  2.04it/s]
 74%|███████▍  | 7957/10712 [1:26:04<22:30,  2.04it/s]
 74%|███████▍  | 7958/10712 [1:26:04<22:29,  2.04it/s]
 74%|███████▍  | 7959/10712 [1:26:05<22:28,  2.04it/s]
 74%|███████▍  | 7960/10712 [1:26:05<22:28,  2.04it/s]
 74%|███████▍  | 7961/10712 [1:26:06<22:28,  2.04it/s]
 74%|███████▍  | 7962/10712 [1:26:06<22:27,  2.04it/s]
 74%|███████▍  | 7963/10712 [1:26:07<22:27,  2.04it/s]
 74%|███████▍  | 7964/10712 [1:26:07<22:27,  2.04it/s]
 74%|███████▍  | 7965/10712 [1:26:08<22:26,  2.04it/s]
 74%|███████▍  | 7966/10712 [1:26:08<22:24,  2.04it/s]
 74%|███████▍  | 7967/10712 [1:26:09<22:25,  2.04it/s]
 74%|███████▍  | 7968/10712 [1:26:09<22:24,  2.04it/s]
 74%|███████▍  | 7969/10712 [1:26:10<22:24,  2.04it/s]
 74%|███████▍  | 7970/10712 [1:26:10<22:22,  2.04it/s]
 74%|███████▍  | 7971/10712 [1:26:11<22:23,  2.04it/s]
 74%|███████▍  | 7972/10712 [1:26:11<22:22,  2.04it/s]
 74%|███████▍  | 7973/10712 [1:26:12<22:21,  2.04it/s]
 74%|███████▍  | 7974/10712 [1:26:12<22:21,  2.04it/s]
 74%|███████▍  | 7975/10712 [1:26:13<22:21,  2.04it/s]
                                                      
{'loss': 3.6325, 'grad_norm': 0.20349593460559845, 'learning_rate': 0.00018605792521116728, 'epoch': 0.74}

 74%|███████▍  | 7975/10712 [1:26:13<22:21,  2.04it/s]
 74%|███████▍  | 7976/10712 [1:26:13<22:21,  2.04it/s]
 74%|███████▍  | 7977/10712 [1:26:14<22:20,  2.04it/s]
 74%|███████▍  | 7978/10712 [1:26:14<22:19,  2.04it/s]
 74%|███████▍  | 7979/10712 [1:26:15<22:19,  2.04it/s]
 74%|███████▍  | 7980/10712 [1:26:15<22:17,  2.04it/s]
 75%|███████▍  | 7981/10712 [1:26:16<22:17,  2.04it/s]
 75%|███████▍  | 7982/10712 [1:26:16<22:16,  2.04it/s]
 75%|███████▍  | 7983/10712 [1:26:17<22:17,  2.04it/s]
 75%|███████▍  | 7984/10712 [1:26:17<22:15,  2.04it/s]
 75%|███████▍  | 7985/10712 [1:26:18<22:13,  2.04it/s]
 75%|███████▍  | 7986/10712 [1:26:18<22:14,  2.04it/s]
 75%|███████▍  | 7987/10712 [1:26:19<22:15,  2.04it/s]
 75%|███████▍  | 7988/10712 [1:26:19<22:14,  2.04it/s]
 75%|███████▍  | 7989/10712 [1:26:20<22:12,  2.04it/s]
 75%|███████▍  | 7990/10712 [1:26:20<22:11,  2.04it/s]
 75%|███████▍  | 7991/10712 [1:26:21<22:13,  2.04it/s]
 75%|███████▍  | 7992/10712 [1:26:21<22:12,  2.04it/s]
 75%|███████▍  | 7993/10712 [1:26:22<22:11,  2.04it/s]
 75%|███████▍  | 7994/10712 [1:26:22<22:11,  2.04it/s]
 75%|███████▍  | 7995/10712 [1:26:23<22:11,  2.04it/s]
 75%|███████▍  | 7996/10712 [1:26:23<22:11,  2.04it/s]
 75%|███████▍  | 7997/10712 [1:26:23<22:09,  2.04it/s]
 75%|███████▍  | 7998/10712 [1:26:24<22:10,  2.04it/s]
 75%|███████▍  | 7999/10712 [1:26:24<22:10,  2.04it/s]
 75%|███████▍  | 8000/10712 [1:26:25<22:08,  2.04it/s]{'loss': 3.6357, 'grad_norm': 0.20619972050189972, 'learning_rate': 0.0001828978364257349, 'epoch': 0.75}

                                                      

 75%|███████▍  | 8000/10712 [1:26:25<22:08,  2.04it/s]
 75%|███████▍  | 8001/10712 [1:26:25<22:09,  2.04it/s]
 75%|███████▍  | 8002/10712 [1:26:26<22:09,  2.04it/s]
 75%|███████▍  | 8003/10712 [1:26:26<22:08,  2.04it/s]
 75%|███████▍  | 8004/10712 [1:26:27<22:07,  2.04it/s]
 75%|███████▍  | 8005/10712 [1:26:27<22:06,  2.04it/s]
 75%|███████▍  | 8006/10712 [1:26:28<22:05,  2.04it/s]
 75%|███████▍  | 8007/10712 [1:26:28<22:06,  2.04it/s]
 75%|███████▍  | 8008/10712 [1:26:29<22:04,  2.04it/s]
 75%|███████▍  | 8009/10712 [1:26:29<22:03,  2.04it/s]
 75%|███████▍  | 8010/10712 [1:26:30<22:02,  2.04it/s]
 75%|███████▍  | 8011/10712 [1:26:30<22:04,  2.04it/s]
 75%|███████▍  | 8012/10712 [1:26:31<22:02,  2.04it/s]
 75%|███████▍  | 8013/10712 [1:26:31<22:01,  2.04it/s]
 75%|███████▍  | 8014/10712 [1:26:32<22:02,  2.04it/s]
 75%|███████▍  | 8015/10712 [1:26:32<22:02,  2.04it/s]
 75%|███████▍  | 8016/10712 [1:26:33<22:01,  2.04it/s]
 75%|███████▍  | 8017/10712 [1:26:33<22:00,  2.04it/s]
 75%|███████▍  | 8018/10712 [1:26:34<21:58,  2.04it/s]
 75%|███████▍  | 8019/10712 [1:26:34<22:00,  2.04it/s]
 75%|███████▍  | 8020/10712 [1:26:35<21:59,  2.04it/s]
 75%|███████▍  | 8021/10712 [1:26:35<21:57,  2.04it/s]
 75%|███████▍  | 8022/10712 [1:26:36<21:56,  2.04it/s]
 75%|███████▍  | 8023/10712 [1:26:36<21:59,  2.04it/s]
 75%|███████▍  | 8024/10712 [1:26:37<21:58,  2.04it/s]
 75%|███████▍  | 8025/10712 [1:26:37<21:56,  2.04it/s]
                                                      
{'loss': 3.6368, 'grad_norm': 0.2019481211900711, 'learning_rate': 0.0001797587962055431, 'epoch': 0.75}

 75%|███████▍  | 8025/10712 [1:26:37<21:56,  2.04it/s]
 75%|███████▍  | 8026/10712 [1:26:38<21:58,  2.04it/s]
 75%|███████▍  | 8027/10712 [1:26:38<21:57,  2.04it/s]
 75%|███████▍  | 8028/10712 [1:26:39<21:54,  2.04it/s]
 75%|███████▍  | 8029/10712 [1:26:39<21:55,  2.04it/s]
 75%|███████▍  | 8030/10712 [1:26:40<21:55,  2.04it/s]
 75%|███████▍  | 8031/10712 [1:26:40<21:55,  2.04it/s]
 75%|███████▍  | 8032/10712 [1:26:41<21:52,  2.04it/s]
 75%|███████▍  | 8033/10712 [1:26:41<21:52,  2.04it/s]
 75%|███████▌  | 8034/10712 [1:26:42<21:52,  2.04it/s]
 75%|███████▌  | 8035/10712 [1:26:42<21:51,  2.04it/s]
 75%|███████▌  | 8036/10712 [1:26:43<21:49,  2.04it/s]
 75%|███████▌  | 8037/10712 [1:26:43<21:50,  2.04it/s]
 75%|███████▌  | 8038/10712 [1:26:44<21:59,  2.03it/s]
 75%|███████▌  | 8039/10712 [1:26:44<21:58,  2.03it/s]
 75%|███████▌  | 8040/10712 [1:26:45<21:54,  2.03it/s]
 75%|███████▌  | 8041/10712 [1:26:45<21:52,  2.04it/s]
 75%|███████▌  | 8042/10712 [1:26:46<21:50,  2.04it/s]
 75%|███████▌  | 8043/10712 [1:26:46<21:50,  2.04it/s]
 75%|███████▌  | 8044/10712 [1:26:47<21:48,  2.04it/s]
 75%|███████▌  | 8045/10712 [1:26:47<21:46,  2.04it/s]
 75%|███████▌  | 8046/10712 [1:26:48<21:46,  2.04it/s]
 75%|███████▌  | 8047/10712 [1:26:48<21:49,  2.03it/s]
 75%|███████▌  | 8048/10712 [1:26:48<21:48,  2.04it/s]
 75%|███████▌  | 8049/10712 [1:26:49<21:48,  2.04it/s]
 75%|███████▌  | 8050/10712 [1:26:49<21:47,  2.04it/s]
                                                      
{'loss': 3.6401, 'grad_norm': 0.2016284316778183, 'learning_rate': 0.00017664101291338597, 'epoch': 0.75}

 75%|███████▌  | 8050/10712 [1:26:49<21:47,  2.04it/s]
 75%|███████▌  | 8051/10712 [1:26:50<21:46,  2.04it/s]
 75%|███████▌  | 8052/10712 [1:26:50<21:43,  2.04it/s]
 75%|███████▌  | 8053/10712 [1:26:51<21:43,  2.04it/s]
 75%|███████▌  | 8054/10712 [1:26:51<21:44,  2.04it/s]
 75%|███████▌  | 8055/10712 [1:26:52<21:43,  2.04it/s]
 75%|███████▌  | 8056/10712 [1:26:52<21:43,  2.04it/s]
 75%|███████▌  | 8057/10712 [1:26:53<21:41,  2.04it/s]
 75%|███████▌  | 8058/10712 [1:26:53<21:41,  2.04it/s]
 75%|███████▌  | 8059/10712 [1:26:54<21:41,  2.04it/s]
 75%|███████▌  | 8060/10712 [1:26:54<21:40,  2.04it/s]
 75%|███████▌  | 8061/10712 [1:26:55<21:38,  2.04it/s]
 75%|███████▌  | 8062/10712 [1:26:55<21:45,  2.03it/s]
 75%|███████▌  | 8063/10712 [1:26:56<21:42,  2.03it/s]
 75%|███████▌  | 8064/10712 [1:26:56<21:41,  2.03it/s]
 75%|███████▌  | 8065/10712 [1:26:57<21:41,  2.03it/s]
 75%|███████▌  | 8066/10712 [1:26:57<21:40,  2.04it/s]
 75%|███████▌  | 8067/10712 [1:26:58<21:38,  2.04it/s]
 75%|███████▌  | 8068/10712 [1:26:58<21:38,  2.04it/s]
 75%|███████▌  | 8069/10712 [1:26:59<21:38,  2.04it/s]
 75%|███████▌  | 8070/10712 [1:26:59<21:35,  2.04it/s]
 75%|███████▌  | 8071/10712 [1:27:00<21:34,  2.04it/s]
 75%|███████▌  | 8072/10712 [1:27:00<21:35,  2.04it/s]
 75%|███████▌  | 8073/10712 [1:27:01<21:35,  2.04it/s]
 75%|███████▌  | 8074/10712 [1:27:01<21:33,  2.04it/s]
 75%|███████▌  | 8075/10712 [1:27:02<21:33,  2.04it/s]
                                                      
{'loss': 3.6397, 'grad_norm': 0.20647446811199188, 'learning_rate': 0.00017354469350106845, 'epoch': 0.75}

 75%|███████▌  | 8075/10712 [1:27:02<21:33,  2.04it/s]
 75%|███████▌  | 8076/10712 [1:27:02<21:34,  2.04it/s]
 75%|███████▌  | 8077/10712 [1:27:03<21:32,  2.04it/s]
 75%|███████▌  | 8078/10712 [1:27:03<21:31,  2.04it/s]
 75%|███████▌  | 8079/10712 [1:27:04<21:31,  2.04it/s]
 75%|███████▌  | 8080/10712 [1:27:04<21:30,  2.04it/s]
 75%|███████▌  | 8081/10712 [1:27:05<21:30,  2.04it/s]
 75%|███████▌  | 8082/10712 [1:27:05<21:30,  2.04it/s]
 75%|███████▌  | 8083/10712 [1:27:06<21:28,  2.04it/s]
 75%|███████▌  | 8084/10712 [1:27:06<21:27,  2.04it/s]
 75%|███████▌  | 8085/10712 [1:27:07<21:26,  2.04it/s]
 75%|███████▌  | 8086/10712 [1:27:07<21:26,  2.04it/s]
 75%|███████▌  | 8087/10712 [1:27:08<21:25,  2.04it/s]
 76%|███████▌  | 8088/10712 [1:27:08<21:25,  2.04it/s]
 76%|███████▌  | 8089/10712 [1:27:09<21:25,  2.04it/s]
 76%|███████▌  | 8090/10712 [1:27:09<21:24,  2.04it/s]
 76%|███████▌  | 8091/10712 [1:27:10<21:24,  2.04it/s]
 76%|███████▌  | 8092/10712 [1:27:10<21:24,  2.04it/s]
 76%|███████▌  | 8093/10712 [1:27:11<21:24,  2.04it/s]
 76%|███████▌  | 8094/10712 [1:27:11<21:23,  2.04it/s]
 76%|███████▌  | 8095/10712 [1:27:12<21:21,  2.04it/s]
 76%|███████▌  | 8096/10712 [1:27:12<21:22,  2.04it/s]
 76%|███████▌  | 8097/10712 [1:27:13<21:21,  2.04it/s]
 76%|███████▌  | 8098/10712 [1:27:13<21:21,  2.04it/s]
 76%|███████▌  | 8099/10712 [1:27:14<21:20,  2.04it/s]
 76%|███████▌  | 8100/10712 [1:27:14<21:21,  2.04it/s]{'loss': 3.6381, 'grad_norm': 0.20886802673339844, 'learning_rate': 0.00017047004349566842, 'epoch': 0.76}

                                                      

 76%|███████▌  | 8100/10712 [1:27:14<21:21,  2.04it/s]
 76%|███████▌  | 8101/10712 [1:27:14<21:20,  2.04it/s]
 76%|███████▌  | 8102/10712 [1:27:15<21:18,  2.04it/s]
 76%|███████▌  | 8103/10712 [1:27:15<21:20,  2.04it/s]
 76%|███████▌  | 8104/10712 [1:27:16<21:19,  2.04it/s]
 76%|███████▌  | 8105/10712 [1:27:16<21:17,  2.04it/s]
 76%|███████▌  | 8106/10712 [1:27:17<21:18,  2.04it/s]
 76%|███████▌  | 8107/10712 [1:27:17<21:17,  2.04it/s]
 76%|███████▌  | 8108/10712 [1:27:18<21:15,  2.04it/s]
 76%|███████▌  | 8109/10712 [1:27:19<24:37,  1.76it/s]
 76%|███████▌  | 8110/10712 [1:27:19<23:34,  1.84it/s]
 76%|███████▌  | 8111/10712 [1:27:20<22:53,  1.89it/s]
 76%|███████▌  | 8112/10712 [1:27:20<22:23,  1.93it/s]
 76%|███████▌  | 8113/10712 [1:27:21<22:01,  1.97it/s]
 76%|███████▌  | 8114/10712 [1:27:21<21:45,  1.99it/s]
 76%|███████▌  | 8115/10712 [1:27:22<21:36,  2.00it/s]
 76%|███████▌  | 8116/10712 [1:27:22<21:28,  2.01it/s]
 76%|███████▌  | 8117/10712 [1:27:23<21:21,  2.02it/s]
 76%|███████▌  | 8118/10712 [1:27:23<21:17,  2.03it/s]
 76%|███████▌  | 8119/10712 [1:27:24<21:16,  2.03it/s]
 76%|███████▌  | 8120/10712 [1:27:24<21:14,  2.03it/s]
 76%|███████▌  | 8121/10712 [1:27:25<21:13,  2.04it/s]
 76%|███████▌  | 8122/10712 [1:27:25<21:12,  2.04it/s]
 76%|███████▌  | 8123/10712 [1:27:26<21:10,  2.04it/s]
 76%|███████▌  | 8124/10712 [1:27:26<21:09,  2.04it/s]
 76%|███████▌  | 8125/10712 [1:27:27<21:09,  2.04it/s]{'loss': 3.6441, 'grad_norm': 0.20522728562355042, 'learning_rate': 0.00016741726698589533, 'epoch': 0.76}
                                                      

 76%|███████▌  | 8125/10712 [1:27:27<21:09,  2.04it/s]
 76%|███████▌  | 8126/10712 [1:27:27<21:10,  2.04it/s]
 76%|███████▌  | 8127/10712 [1:27:27<21:09,  2.04it/s]
 76%|███████▌  | 8128/10712 [1:27:28<21:08,  2.04it/s]
 76%|███████▌  | 8129/10712 [1:27:28<21:07,  2.04it/s]
 76%|███████▌  | 8130/10712 [1:27:29<21:06,  2.04it/s]
 76%|███████▌  | 8131/10712 [1:27:29<21:05,  2.04it/s]
 76%|███████▌  | 8132/10712 [1:27:30<21:05,  2.04it/s]
 76%|███████▌  | 8133/10712 [1:27:30<21:04,  2.04it/s]
 76%|███████▌  | 8134/10712 [1:27:31<21:03,  2.04it/s]
 76%|███████▌  | 8135/10712 [1:27:31<21:02,  2.04it/s]
 76%|███████▌  | 8136/10712 [1:27:32<21:03,  2.04it/s]
 76%|███████▌  | 8137/10712 [1:27:32<21:03,  2.04it/s]
 76%|███████▌  | 8138/10712 [1:27:33<21:00,  2.04it/s]
 76%|███████▌  | 8139/10712 [1:27:33<20:59,  2.04it/s]
 76%|███████▌  | 8140/10712 [1:27:34<21:00,  2.04it/s]
 76%|███████▌  | 8141/10712 [1:27:34<20:58,  2.04it/s]
 76%|███████▌  | 8142/10712 [1:27:35<20:57,  2.04it/s]
 76%|███████▌  | 8143/10712 [1:27:35<20:55,  2.05it/s]
 76%|███████▌  | 8144/10712 [1:27:36<20:56,  2.04it/s]
 76%|███████▌  | 8145/10712 [1:27:36<20:57,  2.04it/s]
 76%|███████▌  | 8146/10712 [1:27:37<20:57,  2.04it/s]
 76%|███████▌  | 8147/10712 [1:27:37<20:55,  2.04it/s]
 76%|███████▌  | 8148/10712 [1:27:38<20:56,  2.04it/s]
 76%|███████▌  | 8149/10712 [1:27:38<20:56,  2.04it/s]
 76%|███████▌  | 8150/10712 [1:27:39<20:54,  2.04it/s]{'loss': 3.6389, 'grad_norm': 0.21200458705425262, 'learning_rate': 0.00016438656660854202, 'epoch': 0.76}

                                                      

 76%|███████▌  | 8150/10712 [1:27:39<20:54,  2.04it/s]
 76%|███████▌  | 8151/10712 [1:27:40<24:29,  1.74it/s]
 76%|███████▌  | 8152/10712 [1:27:40<23:25,  1.82it/s]
 76%|███████▌  | 8153/10712 [1:27:41<22:40,  1.88it/s]
 76%|███████▌  | 8154/10712 [1:27:41<22:08,  1.93it/s]
 76%|███████▌  | 8155/10712 [1:27:41<21:44,  1.96it/s]
 76%|███████▌  | 8156/10712 [1:27:42<21:28,  1.98it/s]
 76%|███████▌  | 8157/10712 [1:27:42<21:18,  2.00it/s]
 76%|███████▌  | 8158/10712 [1:27:43<21:09,  2.01it/s]
 76%|███████▌  | 8159/10712 [1:27:43<21:02,  2.02it/s]
 76%|███████▌  | 8160/10712 [1:27:44<20:58,  2.03it/s]
 76%|███████▌  | 8161/10712 [1:27:44<20:56,  2.03it/s]
 76%|███████▌  | 8162/10712 [1:27:45<20:54,  2.03it/s]
 76%|███████▌  | 8163/10712 [1:27:45<20:50,  2.04it/s]
 76%|███████▌  | 8164/10712 [1:27:46<20:50,  2.04it/s]
 76%|███████▌  | 8165/10712 [1:27:46<20:49,  2.04it/s]
 76%|███████▌  | 8166/10712 [1:27:47<20:47,  2.04it/s]
 76%|███████▌  | 8167/10712 [1:27:47<20:45,  2.04it/s]
 76%|███████▋  | 8168/10712 [1:27:48<20:45,  2.04it/s]
 76%|███████▋  | 8169/10712 [1:27:48<20:46,  2.04it/s]
 76%|███████▋  | 8170/10712 [1:27:49<20:45,  2.04it/s]
 76%|███████▋  | 8171/10712 [1:27:49<20:44,  2.04it/s]
 76%|███████▋  | 8172/10712 [1:27:50<20:45,  2.04it/s]
 76%|███████▋  | 8173/10712 [1:27:50<20:44,  2.04it/s]
 76%|███████▋  | 8174/10712 [1:27:51<20:43,  2.04it/s]
 76%|███████▋  | 8175/10712 [1:27:51<20:41,  2.04it/s]{'loss': 3.6327, 'grad_norm': 0.2091967612504959, 'learning_rate': 0.00016137814353503523, 'epoch': 0.76}
                                                      

 76%|███████▋  | 8175/10712 [1:27:51<20:41,  2.04it/s]
 76%|███████▋  | 8176/10712 [1:27:52<20:43,  2.04it/s]
 76%|███████▋  | 8177/10712 [1:27:52<20:41,  2.04it/s]
 76%|███████▋  | 8178/10712 [1:27:53<20:40,  2.04it/s]
 76%|███████▋  | 8179/10712 [1:27:53<20:39,  2.04it/s]
 76%|███████▋  | 8180/10712 [1:27:54<20:39,  2.04it/s]
 76%|███████▋  | 8181/10712 [1:27:54<20:40,  2.04it/s]
 76%|███████▋  | 8182/10712 [1:27:55<20:38,  2.04it/s]
 76%|███████▋  | 8183/10712 [1:27:55<20:38,  2.04it/s]
 76%|███████▋  | 8184/10712 [1:27:56<20:37,  2.04it/s]
 76%|███████▋  | 8185/10712 [1:27:56<20:38,  2.04it/s]
 76%|███████▋  | 8186/10712 [1:27:57<20:37,  2.04it/s]
 76%|███████▋  | 8187/10712 [1:27:57<20:37,  2.04it/s]
 76%|███████▋  | 8188/10712 [1:27:58<20:35,  2.04it/s]
 76%|███████▋  | 8189/10712 [1:27:58<20:35,  2.04it/s]
 76%|███████▋  | 8190/10712 [1:27:59<20:37,  2.04it/s]
 76%|███████▋  | 8191/10712 [1:27:59<20:36,  2.04it/s]
 76%|███████▋  | 8192/10712 [1:28:00<20:33,  2.04it/s]
 76%|███████▋  | 8193/10712 [1:28:00<20:33,  2.04it/s]
 76%|███████▋  | 8194/10712 [1:28:01<20:32,  2.04it/s]
 77%|███████▋  | 8195/10712 [1:28:01<20:34,  2.04it/s]
 77%|███████▋  | 8196/10712 [1:28:02<20:32,  2.04it/s]
 77%|███████▋  | 8197/10712 [1:28:02<20:32,  2.04it/s]
 77%|███████▋  | 8198/10712 [1:28:03<20:32,  2.04it/s]
 77%|███████▋  | 8199/10712 [1:28:03<20:32,  2.04it/s]
 77%|███████▋  | 8200/10712 [1:28:04<20:30,  2.04it/s]{'loss': 3.641, 'grad_norm': 0.21674151718616486, 'learning_rate': 0.00015839219745808174, 'epoch': 0.77}

                                                      

 77%|███████▋  | 8200/10712 [1:28:04<20:30,  2.04it/s]
 77%|███████▋  | 8201/10712 [1:28:04<20:31,  2.04it/s]
 77%|███████▋  | 8202/10712 [1:28:05<20:30,  2.04it/s]
 77%|███████▋  | 8203/10712 [1:28:05<20:32,  2.04it/s]
 77%|███████▋  | 8204/10712 [1:28:05<20:29,  2.04it/s]
 77%|███████▋  | 8205/10712 [1:28:06<20:29,  2.04it/s]
 77%|███████▋  | 8206/10712 [1:28:06<20:29,  2.04it/s]
 77%|███████▋  | 8207/10712 [1:28:07<20:29,  2.04it/s]
 77%|███████▋  | 8208/10712 [1:28:07<20:27,  2.04it/s]
 77%|███████▋  | 8209/10712 [1:28:08<20:26,  2.04it/s]
 77%|███████▋  | 8210/10712 [1:28:08<20:27,  2.04it/s]
 77%|███████▋  | 8211/10712 [1:28:09<20:25,  2.04it/s]
 77%|███████▋  | 8212/10712 [1:28:09<20:25,  2.04it/s]
 77%|███████▋  | 8213/10712 [1:28:10<20:25,  2.04it/s]
 77%|███████▋  | 8214/10712 [1:28:10<20:23,  2.04it/s]
 77%|███████▋  | 8215/10712 [1:28:11<20:22,  2.04it/s]
 77%|███████▋  | 8216/10712 [1:28:11<20:21,  2.04it/s]
 77%|███████▋  | 8217/10712 [1:28:12<20:22,  2.04it/s]
 77%|███████▋  | 8218/10712 [1:28:12<20:22,  2.04it/s]
 77%|███████▋  | 8219/10712 [1:28:13<20:20,  2.04it/s]
 77%|███████▋  | 8220/10712 [1:28:13<20:19,  2.04it/s]
 77%|███████▋  | 8221/10712 [1:28:14<20:21,  2.04it/s]
 77%|███████▋  | 8222/10712 [1:28:14<20:20,  2.04it/s]
 77%|███████▋  | 8223/10712 [1:28:15<20:18,  2.04it/s]
 77%|███████▋  | 8224/10712 [1:28:15<20:20,  2.04it/s]
 77%|███████▋  | 8225/10712 [1:28:16<20:20,  2.04it/s]{'loss': 3.6264, 'grad_norm': 0.2081044614315033, 'learning_rate': 0.0001554289265784129, 'epoch': 0.77}

                                                      

 77%|███████▋  | 8225/10712 [1:28:16<20:20,  2.04it/s]
 77%|███████▋  | 8226/10712 [1:28:16<20:18,  2.04it/s]
 77%|███████▋  | 8227/10712 [1:28:17<20:18,  2.04it/s]
 77%|███████▋  | 8228/10712 [1:28:17<20:19,  2.04it/s]
 77%|███████▋  | 8229/10712 [1:28:18<20:17,  2.04it/s]
 77%|███████▋  | 8230/10712 [1:28:18<20:16,  2.04it/s]
 77%|███████▋  | 8231/10712 [1:28:19<20:17,  2.04it/s]
 77%|███████▋  | 8232/10712 [1:28:19<20:17,  2.04it/s]
 77%|███████▋  | 8233/10712 [1:28:20<20:14,  2.04it/s]
 77%|███████▋  | 8234/10712 [1:28:20<20:14,  2.04it/s]
 77%|███████▋  | 8235/10712 [1:28:21<20:14,  2.04it/s]
 77%|███████▋  | 8236/10712 [1:28:21<20:13,  2.04it/s]
 77%|███████▋  | 8237/10712 [1:28:22<20:12,  2.04it/s]
 77%|███████▋  | 8238/10712 [1:28:22<20:13,  2.04it/s]
 77%|███████▋  | 8239/10712 [1:28:23<20:12,  2.04it/s]
 77%|███████▋  | 8240/10712 [1:28:23<20:10,  2.04it/s]
 77%|███████▋  | 8241/10712 [1:28:24<20:10,  2.04it/s]
 77%|███████▋  | 8242/10712 [1:28:24<20:11,  2.04it/s]
 77%|███████▋  | 8243/10712 [1:28:25<20:10,  2.04it/s]
 77%|███████▋  | 8244/10712 [1:28:25<20:09,  2.04it/s]
 77%|███████▋  | 8245/10712 [1:28:26<20:09,  2.04it/s]
 77%|███████▋  | 8246/10712 [1:28:26<20:09,  2.04it/s]
 77%|███████▋  | 8247/10712 [1:28:27<20:08,  2.04it/s]
 77%|███████▋  | 8248/10712 [1:28:27<20:07,  2.04it/s]
 77%|███████▋  | 8249/10712 [1:28:28<20:07,  2.04it/s]
 77%|███████▋  | 8250/10712 [1:28:28<20:07,  2.04it/s]{'loss': 3.6337, 'grad_norm': 0.21018819510936737, 'learning_rate': 0.00015248852759162934, 'epoch': 0.77}
                                                      

 77%|███████▋  | 8250/10712 [1:28:28<20:07,  2.04it/s]
 77%|███████▋  | 8251/10712 [1:28:29<20:06,  2.04it/s]
 77%|███████▋  | 8252/10712 [1:28:29<20:06,  2.04it/s]
 77%|███████▋  | 8253/10712 [1:28:30<20:05,  2.04it/s]
 77%|███████▋  | 8254/10712 [1:28:30<20:05,  2.04it/s]
 77%|███████▋  | 8255/10712 [1:28:30<20:04,  2.04it/s]
 77%|███████▋  | 8256/10712 [1:28:31<20:03,  2.04it/s]
 77%|███████▋  | 8257/10712 [1:28:31<20:03,  2.04it/s]
 77%|███████▋  | 8258/10712 [1:28:32<20:02,  2.04it/s]
 77%|███████▋  | 8259/10712 [1:28:32<20:03,  2.04it/s]
 77%|███████▋  | 8260/10712 [1:28:33<20:02,  2.04it/s]
 77%|███████▋  | 8261/10712 [1:28:33<20:00,  2.04it/s]
 77%|███████▋  | 8262/10712 [1:28:34<20:00,  2.04it/s]
 77%|███████▋  | 8263/10712 [1:28:34<20:01,  2.04it/s]
 77%|███████▋  | 8264/10712 [1:28:35<19:59,  2.04it/s]
 77%|███████▋  | 8265/10712 [1:28:35<19:58,  2.04it/s]
 77%|███████▋  | 8266/10712 [1:28:36<19:57,  2.04it/s]
 77%|███████▋  | 8267/10712 [1:28:36<19:58,  2.04it/s]
 77%|███████▋  | 8268/10712 [1:28:37<19:57,  2.04it/s]
 77%|███████▋  | 8269/10712 [1:28:37<19:57,  2.04it/s]
 77%|███████▋  | 8270/10712 [1:28:38<19:55,  2.04it/s]
 77%|███████▋  | 8271/10712 [1:28:38<19:57,  2.04it/s]
 77%|███████▋  | 8272/10712 [1:28:39<19:56,  2.04it/s]
 77%|███████▋  | 8273/10712 [1:28:39<19:55,  2.04it/s]
 77%|███████▋  | 8274/10712 [1:28:40<19:54,  2.04it/s]
 77%|███████▋  | 8275/10712 [1:28:40<19:54,  2.04it/s]
                                                      
{'loss': 3.6273, 'grad_norm': 0.21714285016059875, 'learning_rate': 0.00014957119567514355, 'epoch': 0.77}

 77%|███████▋  | 8275/10712 [1:28:40<19:54,  2.04it/s]
 77%|███████▋  | 8276/10712 [1:28:41<19:55,  2.04it/s]
 77%|███████▋  | 8277/10712 [1:28:41<19:53,  2.04it/s]
 77%|███████▋  | 8278/10712 [1:28:42<19:52,  2.04it/s]
 77%|███████▋  | 8279/10712 [1:28:42<19:53,  2.04it/s]
 77%|███████▋  | 8280/10712 [1:28:43<19:52,  2.04it/s]
 77%|███████▋  | 8281/10712 [1:28:43<19:51,  2.04it/s]
 77%|███████▋  | 8282/10712 [1:28:44<19:50,  2.04it/s]
 77%|███████▋  | 8283/10712 [1:28:44<19:50,  2.04it/s]
 77%|███████▋  | 8284/10712 [1:28:45<19:49,  2.04it/s]
 77%|███████▋  | 8285/10712 [1:28:45<19:48,  2.04it/s]
 77%|███████▋  | 8286/10712 [1:28:46<19:48,  2.04it/s]
 77%|███████▋  | 8287/10712 [1:28:46<19:49,  2.04it/s]
 77%|███████▋  | 8288/10712 [1:28:47<19:48,  2.04it/s]
 77%|███████▋  | 8289/10712 [1:28:47<19:47,  2.04it/s]
 77%|███████▋  | 8290/10712 [1:28:48<19:48,  2.04it/s]
 77%|███████▋  | 8291/10712 [1:28:48<19:47,  2.04it/s]
 77%|███████▋  | 8292/10712 [1:28:49<19:47,  2.04it/s]
 77%|███████▋  | 8293/10712 [1:28:49<19:45,  2.04it/s]
 77%|███████▋  | 8294/10712 [1:28:50<19:47,  2.04it/s]
 77%|███████▋  | 8295/10712 [1:28:50<19:46,  2.04it/s]
 77%|███████▋  | 8296/10712 [1:28:51<19:46,  2.04it/s]
 77%|███████▋  | 8297/10712 [1:28:51<19:45,  2.04it/s]
 77%|███████▋  | 8298/10712 [1:28:52<19:44,  2.04it/s]
 77%|███████▋  | 8299/10712 [1:28:52<19:43,  2.04it/s]
 77%|███████▋  | 8300/10712 [1:28:53<19:43,  2.04it/s]
                                                      {'loss': 3.6272, 'grad_norm': 0.21150019764900208, 'learning_rate': 0.00014667712447522592, 'epoch': 0.77}

 77%|███████▋  | 8300/10712 [1:28:53<19:43,  2.04it/s]
 77%|███████▋  | 8301/10712 [1:28:53<19:42,  2.04it/s]
 78%|███████▊  | 8302/10712 [1:28:54<19:42,  2.04it/s]
 78%|███████▊  | 8303/10712 [1:28:54<19:41,  2.04it/s]
 78%|███████▊  | 8304/10712 [1:28:55<19:42,  2.04it/s]
 78%|███████▊  | 8305/10712 [1:28:55<19:42,  2.04it/s]
 78%|███████▊  | 8306/10712 [1:28:56<19:40,  2.04it/s]
 78%|███████▊  | 8307/10712 [1:28:56<19:40,  2.04it/s]
 78%|███████▊  | 8308/10712 [1:28:56<19:38,  2.04it/s]
 78%|███████▊  | 8309/10712 [1:28:57<19:37,  2.04it/s]
 78%|███████▊  | 8310/10712 [1:28:57<19:37,  2.04it/s]
 78%|███████▊  | 8311/10712 [1:28:58<19:37,  2.04it/s]
 78%|███████▊  | 8312/10712 [1:28:58<19:36,  2.04it/s]
 78%|███████▊  | 8313/10712 [1:28:59<19:36,  2.04it/s]
 78%|███████▊  | 8314/10712 [1:28:59<19:36,  2.04it/s]
 78%|███████▊  | 8315/10712 [1:29:00<19:36,  2.04it/s]
 78%|███████▊  | 8316/10712 [1:29:00<19:35,  2.04it/s]
 78%|███████▊  | 8317/10712 [1:29:01<19:34,  2.04it/s]
 78%|███████▊  | 8318/10712 [1:29:01<19:35,  2.04it/s]
 78%|███████▊  | 8319/10712 [1:29:02<19:33,  2.04it/s]
 78%|███████▊  | 8320/10712 [1:29:02<19:32,  2.04it/s]
 78%|███████▊  | 8321/10712 [1:29:03<19:33,  2.04it/s]
 78%|███████▊  | 8322/10712 [1:29:03<19:32,  2.04it/s]
 78%|███████▊  | 8323/10712 [1:29:04<19:31,  2.04it/s]
 78%|███████▊  | 8324/10712 [1:29:04<19:30,  2.04it/s]
 78%|███████▊  | 8325/10712 [1:29:05<19:30,  2.04it/s]{'loss': 3.6291, 'grad_norm': 0.20886409282684326, 'learning_rate': 0.0001438065060941493, 'epoch': 0.78}
                                                      

 78%|███████▊  | 8325/10712 [1:29:05<19:30,  2.04it/s]
 78%|███████▊  | 8326/10712 [1:29:05<19:31,  2.04it/s]
 78%|███████▊  | 8327/10712 [1:29:06<19:30,  2.04it/s]
 78%|███████▊  | 8328/10712 [1:29:06<19:29,  2.04it/s]
 78%|███████▊  | 8329/10712 [1:29:07<19:27,  2.04it/s]
 78%|███████▊  | 8330/10712 [1:29:07<19:28,  2.04it/s]
 78%|███████▊  | 8331/10712 [1:29:08<19:27,  2.04it/s]
 78%|███████▊  | 8332/10712 [1:29:08<19:25,  2.04it/s]
 78%|███████▊  | 8333/10712 [1:29:09<19:25,  2.04it/s]
 78%|███████▊  | 8334/10712 [1:29:09<19:24,  2.04it/s]
 78%|███████▊  | 8335/10712 [1:29:10<19:25,  2.04it/s]
 78%|███████▊  | 8336/10712 [1:29:10<19:24,  2.04it/s]
 78%|███████▊  | 8337/10712 [1:29:11<19:23,  2.04it/s]
 78%|███████▊  | 8338/10712 [1:29:11<19:22,  2.04it/s]
 78%|███████▊  | 8339/10712 [1:29:12<19:23,  2.04it/s]
 78%|███████▊  | 8340/10712 [1:29:12<19:24,  2.04it/s]
 78%|███████▊  | 8341/10712 [1:29:13<19:23,  2.04it/s]
 78%|███████▊  | 8342/10712 [1:29:13<19:22,  2.04it/s]
 78%|███████▊  | 8343/10712 [1:29:14<19:21,  2.04it/s]
 78%|███████▊  | 8344/10712 [1:29:14<19:19,  2.04it/s]
 78%|███████▊  | 8345/10712 [1:29:15<19:19,  2.04it/s]
 78%|███████▊  | 8346/10712 [1:29:15<19:20,  2.04it/s]
 78%|███████▊  | 8347/10712 [1:29:16<19:18,  2.04it/s]
 78%|███████▊  | 8348/10712 [1:29:16<19:18,  2.04it/s]
 78%|███████▊  | 8349/10712 [1:29:17<19:18,  2.04it/s]
 78%|███████▊  | 8350/10712 [1:29:17<19:17,  2.04it/s]{'loss': 3.6209, 'grad_norm': 0.20838214457035065, 'learning_rate': 0.0001409595310774391, 'epoch': 0.78}

                                                      

 78%|███████▊  | 8350/10712 [1:29:17<19:17,  2.04it/s]
 78%|███████▊  | 8351/10712 [1:29:18<19:17,  2.04it/s]
 78%|███████▊  | 8352/10712 [1:29:18<19:16,  2.04it/s]
 78%|███████▊  | 8353/10712 [1:29:19<19:16,  2.04it/s]
 78%|███████▊  | 8354/10712 [1:29:19<19:15,  2.04it/s]
 78%|███████▊  | 8355/10712 [1:29:20<19:13,  2.04it/s]
 78%|███████▊  | 8356/10712 [1:29:20<19:13,  2.04it/s]
 78%|███████▊  | 8357/10712 [1:29:21<19:13,  2.04it/s]
 78%|███████▊  | 8358/10712 [1:29:21<19:12,  2.04it/s]
 78%|███████▊  | 8359/10712 [1:29:21<19:11,  2.04it/s]
 78%|███████▊  | 8360/10712 [1:29:22<19:11,  2.04it/s]
 78%|███████▊  | 8361/10712 [1:29:22<19:13,  2.04it/s]
 78%|███████▊  | 8362/10712 [1:29:23<19:11,  2.04it/s]
 78%|███████▊  | 8363/10712 [1:29:23<19:10,  2.04it/s]
 78%|███████▊  | 8364/10712 [1:29:24<19:09,  2.04it/s]
 78%|███████▊  | 8365/10712 [1:29:24<19:09,  2.04it/s]
 78%|███████▊  | 8366/10712 [1:29:25<19:09,  2.04it/s]
 78%|███████▊  | 8367/10712 [1:29:25<19:07,  2.04it/s]
 78%|███████▊  | 8368/10712 [1:29:26<19:08,  2.04it/s]
 78%|███████▊  | 8369/10712 [1:29:26<19:07,  2.04it/s]
 78%|███████▊  | 8370/10712 [1:29:27<19:06,  2.04it/s]
 78%|███████▊  | 8371/10712 [1:29:27<19:05,  2.04it/s]
 78%|███████▊  | 8372/10712 [1:29:28<19:05,  2.04it/s]
 78%|███████▊  | 8373/10712 [1:29:28<19:05,  2.04it/s]
 78%|███████▊  | 8374/10712 [1:29:29<19:04,  2.04it/s]
 78%|███████▊  | 8375/10712 [1:29:29<19:02,  2.05it/s]{'loss': 3.6311, 'grad_norm': 0.2053886353969574, 'learning_rate': 0.000138136388401224, 'epoch': 0.78}

                                                      

 78%|███████▊  | 8375/10712 [1:29:29<19:02,  2.05it/s]
 78%|███████▊  | 8376/10712 [1:29:30<19:05,  2.04it/s]
 78%|███████▊  | 8377/10712 [1:29:30<19:03,  2.04it/s]
 78%|███████▊  | 8378/10712 [1:29:31<19:02,  2.04it/s]
 78%|███████▊  | 8379/10712 [1:29:31<19:02,  2.04it/s]
 78%|███████▊  | 8380/10712 [1:29:32<19:02,  2.04it/s]
 78%|███████▊  | 8381/10712 [1:29:32<19:02,  2.04it/s]
 78%|███████▊  | 8382/10712 [1:29:33<19:00,  2.04it/s]
 78%|███████▊  | 8383/10712 [1:29:33<19:01,  2.04it/s]
 78%|███████▊  | 8384/10712 [1:29:34<19:02,  2.04it/s]
 78%|███████▊  | 8385/10712 [1:29:34<19:00,  2.04it/s]
 78%|███████▊  | 8386/10712 [1:29:35<18:59,  2.04it/s]
 78%|███████▊  | 8387/10712 [1:29:35<18:58,  2.04it/s]
 78%|███████▊  | 8388/10712 [1:29:36<18:59,  2.04it/s]
 78%|███████▊  | 8389/10712 [1:29:36<18:57,  2.04it/s]
 78%|███████▊  | 8390/10712 [1:29:37<18:56,  2.04it/s]
 78%|███████▊  | 8391/10712 [1:29:37<18:56,  2.04it/s]
 78%|███████▊  | 8392/10712 [1:29:38<18:56,  2.04it/s]
 78%|███████▊  | 8393/10712 [1:29:38<18:55,  2.04it/s]
 78%|███████▊  | 8394/10712 [1:29:39<18:55,  2.04it/s]
 78%|███████▊  | 8395/10712 [1:29:39<18:55,  2.04it/s]
 78%|███████▊  | 8396/10712 [1:29:40<18:56,  2.04it/s]
 78%|███████▊  | 8397/10712 [1:29:40<18:55,  2.04it/s]
 78%|███████▊  | 8398/10712 [1:29:41<18:54,  2.04it/s]
 78%|███████▊  | 8399/10712 [1:29:41<18:53,  2.04it/s]
 78%|███████▊  | 8400/10712 [1:29:42<18:52,  2.04it/s]{'loss': 3.6353, 'grad_norm': 0.20758990943431854, 'learning_rate': 0.0001353372654596935, 'epoch': 0.78}

                                                      

 78%|███████▊  | 8400/10712 [1:29:42<18:52,  2.04it/s]
 78%|███████▊  | 8401/10712 [1:29:42<18:53,  2.04it/s]
 78%|███████▊  | 8402/10712 [1:29:43<18:52,  2.04it/s]
 78%|███████▊  | 8403/10712 [1:29:43<18:52,  2.04it/s]
 78%|███████▊  | 8404/10712 [1:29:44<18:50,  2.04it/s]
 78%|███████▊  | 8405/10712 [1:29:44<18:49,  2.04it/s]
 78%|███████▊  | 8406/10712 [1:29:45<18:49,  2.04it/s]
 78%|███████▊  | 8407/10712 [1:29:45<18:50,  2.04it/s]
 78%|███████▊  | 8408/10712 [1:29:45<18:48,  2.04it/s]
 79%|███████▊  | 8409/10712 [1:29:46<18:48,  2.04it/s]
 79%|███████▊  | 8410/10712 [1:29:46<18:48,  2.04it/s]
 79%|███████▊  | 8411/10712 [1:29:47<18:48,  2.04it/s]
 79%|███████▊  | 8412/10712 [1:29:47<18:47,  2.04it/s]
 79%|███████▊  | 8413/10712 [1:29:48<18:46,  2.04it/s]
 79%|███████▊  | 8414/10712 [1:29:48<18:45,  2.04it/s]
 79%|███████▊  | 8415/10712 [1:29:49<18:45,  2.04it/s]
 79%|███████▊  | 8416/10712 [1:29:49<18:45,  2.04it/s]
 79%|███████▊  | 8417/10712 [1:29:50<18:44,  2.04it/s]
 79%|███████▊  | 8418/10712 [1:29:50<18:44,  2.04it/s]
 79%|███████▊  | 8419/10712 [1:29:51<18:44,  2.04it/s]
 79%|███████▊  | 8420/10712 [1:29:51<18:44,  2.04it/s]
 79%|███████▊  | 8421/10712 [1:29:52<18:43,  2.04it/s]
 79%|███████▊  | 8422/10712 [1:29:52<18:42,  2.04it/s]
 79%|███████▊  | 8423/10712 [1:29:53<18:42,  2.04it/s]
 79%|███████▊  | 8424/10712 [1:29:53<18:40,  2.04it/s]
 79%|███████▊  | 8425/10712 [1:29:54<18:40,  2.04it/s]{'loss': 3.6255, 'grad_norm': 0.20996366441249847, 'learning_rate': 0.0001325623480526576, 'epoch': 0.79}

                                                      

 79%|███████▊  | 8425/10712 [1:29:54<18:40,  2.04it/s]
 79%|███████▊  | 8426/10712 [1:29:54<18:42,  2.04it/s]
 79%|███████▊  | 8427/10712 [1:29:55<18:40,  2.04it/s]
 79%|███████▊  | 8428/10712 [1:29:55<18:39,  2.04it/s]
 79%|███████▊  | 8429/10712 [1:29:56<18:39,  2.04it/s]
 79%|███████▊  | 8430/10712 [1:29:56<18:38,  2.04it/s]
 79%|███████▊  | 8431/10712 [1:29:57<18:38,  2.04it/s]
 79%|███████▊  | 8432/10712 [1:29:57<18:38,  2.04it/s]
 79%|███████▊  | 8433/10712 [1:29:58<18:37,  2.04it/s]
 79%|███████▊  | 8434/10712 [1:29:58<18:36,  2.04it/s]
 79%|███████▊  | 8435/10712 [1:29:59<18:35,  2.04it/s]
 79%|███████▉  | 8436/10712 [1:29:59<18:36,  2.04it/s]
 79%|███████▉  | 8437/10712 [1:30:00<18:35,  2.04it/s]
 79%|███████▉  | 8438/10712 [1:30:00<18:35,  2.04it/s]
 79%|███████▉  | 8439/10712 [1:30:01<18:33,  2.04it/s]
 79%|███████▉  | 8440/10712 [1:30:01<18:33,  2.04it/s]
 79%|███████▉  | 8441/10712 [1:30:02<18:32,  2.04it/s]
 79%|███████▉  | 8442/10712 [1:30:02<18:31,  2.04it/s]
 79%|███████▉  | 8443/10712 [1:30:03<18:31,  2.04it/s]
 79%|███████▉  | 8444/10712 [1:30:03<18:32,  2.04it/s]
 79%|███████▉  | 8445/10712 [1:30:04<18:32,  2.04it/s]
 79%|███████▉  | 8446/10712 [1:30:04<18:31,  2.04it/s]
 79%|███████▉  | 8447/10712 [1:30:05<18:30,  2.04it/s]
 79%|███████▉  | 8448/10712 [1:30:05<18:30,  2.04it/s]
 79%|███████▉  | 8449/10712 [1:30:06<18:29,  2.04it/s]
 79%|███████▉  | 8450/10712 [1:30:06<18:29,  2.04it/s]{'loss': 3.6349, 'grad_norm': 0.20106133818626404, 'learning_rate': 0.00012981182037321516, 'epoch': 0.79}
                                                      

 79%|███████▉  | 8450/10712 [1:30:06<18:29,  2.04it/s]
 79%|███████▉  | 8451/10712 [1:30:07<18:29,  2.04it/s]
 79%|███████▉  | 8452/10712 [1:30:07<18:28,  2.04it/s]
 79%|███████▉  | 8453/10712 [1:30:08<18:29,  2.04it/s]
 79%|███████▉  | 8454/10712 [1:30:08<18:28,  2.04it/s]
 79%|███████▉  | 8455/10712 [1:30:09<18:26,  2.04it/s]
 79%|███████▉  | 8456/10712 [1:30:09<18:25,  2.04it/s]
 79%|███████▉  | 8457/10712 [1:30:10<18:25,  2.04it/s]
 79%|███████▉  | 8458/10712 [1:30:10<18:24,  2.04it/s]
 79%|███████▉  | 8459/10712 [1:30:10<18:24,  2.04it/s]
 79%|███████▉  | 8460/10712 [1:30:11<18:22,  2.04it/s]
 79%|███████▉  | 8461/10712 [1:30:11<18:22,  2.04it/s]
 79%|███████▉  | 8462/10712 [1:30:12<18:23,  2.04it/s]
 79%|███████▉  | 8463/10712 [1:30:12<18:23,  2.04it/s]
 79%|███████▉  | 8464/10712 [1:30:13<18:22,  2.04it/s]
 79%|███████▉  | 8465/10712 [1:30:13<18:22,  2.04it/s]
 79%|███████▉  | 8466/10712 [1:30:14<18:20,  2.04it/s]
 79%|███████▉  | 8467/10712 [1:30:14<18:20,  2.04it/s]
 79%|███████▉  | 8468/10712 [1:30:15<18:19,  2.04it/s]
 79%|███████▉  | 8469/10712 [1:30:15<18:19,  2.04it/s]
 79%|███████▉  | 8470/10712 [1:30:16<18:19,  2.04it/s]
 79%|███████▉  | 8471/10712 [1:30:16<18:18,  2.04it/s]
 79%|███████▉  | 8472/10712 [1:30:17<18:18,  2.04it/s]
 79%|███████▉  | 8473/10712 [1:30:17<18:17,  2.04it/s]
 79%|███████▉  | 8474/10712 [1:30:18<18:17,  2.04it/s]
 79%|███████▉  | 8475/10712 [1:30:18<18:16,  2.04it/s]{'loss': 3.6268, 'grad_norm': 0.20455597341060638, 'learning_rate': 0.00012708586499552644, 'epoch': 0.79}

                                                      

 79%|███████▉  | 8475/10712 [1:30:18<18:16,  2.04it/s]
 79%|███████▉  | 8476/10712 [1:30:19<18:16,  2.04it/s]
 79%|███████▉  | 8477/10712 [1:30:19<18:15,  2.04it/s]
 79%|███████▉  | 8478/10712 [1:30:20<18:14,  2.04it/s]
 79%|███████▉  | 8479/10712 [1:30:20<18:16,  2.04it/s]
 79%|███████▉  | 8480/10712 [1:30:21<18:14,  2.04it/s]
 79%|███████▉  | 8481/10712 [1:30:21<18:13,  2.04it/s]
 79%|███████▉  | 8482/10712 [1:30:22<18:14,  2.04it/s]
 79%|███████▉  | 8483/10712 [1:30:22<18:13,  2.04it/s]
 79%|███████▉  | 8484/10712 [1:30:23<18:12,  2.04it/s]
 79%|███████▉  | 8485/10712 [1:30:23<18:11,  2.04it/s]
 79%|███████▉  | 8486/10712 [1:30:24<18:10,  2.04it/s]
 79%|███████▉  | 8487/10712 [1:30:24<18:10,  2.04it/s]
 79%|███████▉  | 8488/10712 [1:30:25<18:08,  2.04it/s]
 79%|███████▉  | 8489/10712 [1:30:25<18:08,  2.04it/s]
 79%|███████▉  | 8490/10712 [1:30:26<18:09,  2.04it/s]
 79%|███████▉  | 8491/10712 [1:30:26<18:08,  2.04it/s]
 79%|███████▉  | 8492/10712 [1:30:27<18:07,  2.04it/s]
 79%|███████▉  | 8493/10712 [1:30:27<18:07,  2.04it/s]
 79%|███████▉  | 8494/10712 [1:30:28<18:06,  2.04it/s]
 79%|███████▉  | 8495/10712 [1:30:28<18:05,  2.04it/s]
 79%|███████▉  | 8496/10712 [1:30:29<18:07,  2.04it/s]
 79%|███████▉  | 8497/10712 [1:30:29<18:06,  2.04it/s]
 79%|███████▉  | 8498/10712 [1:30:30<18:06,  2.04it/s]
 79%|███████▉  | 8499/10712 [1:30:30<18:05,  2.04it/s]
 79%|███████▉  | 8500/10712 [1:30:31<18:05,  2.04it/s]{'loss': 3.6197, 'grad_norm': 0.19899408519268036, 'learning_rate': 0.000124384662862695, 'epoch': 0.79}

                                                      

 79%|███████▉  | 8500/10712 [1:30:31<18:05,  2.04it/s]
 79%|███████▉  | 8501/10712 [1:30:31<18:04,  2.04it/s]
 79%|███████▉  | 8502/10712 [1:30:32<18:03,  2.04it/s]
 79%|███████▉  | 8503/10712 [1:30:32<18:04,  2.04it/s]
 79%|███████▉  | 8504/10712 [1:30:33<18:03,  2.04it/s]
 79%|███████▉  | 8505/10712 [1:30:33<18:01,  2.04it/s]
 79%|███████▉  | 8506/10712 [1:30:34<18:02,  2.04it/s]
 79%|███████▉  | 8507/10712 [1:30:34<18:02,  2.04it/s]
 79%|███████▉  | 8508/10712 [1:30:35<18:00,  2.04it/s]
 79%|███████▉  | 8509/10712 [1:30:35<18:00,  2.04it/s]
 79%|███████▉  | 8510/10712 [1:30:36<18:01,  2.04it/s]
 79%|███████▉  | 8511/10712 [1:30:36<17:59,  2.04it/s]
 79%|███████▉  | 8512/10712 [1:30:36<17:59,  2.04it/s]
 79%|███████▉  | 8513/10712 [1:30:37<17:59,  2.04it/s]
 79%|███████▉  | 8514/10712 [1:30:37<17:58,  2.04it/s]
 79%|███████▉  | 8515/10712 [1:30:38<17:56,  2.04it/s]
 79%|███████▉  | 8516/10712 [1:30:38<17:54,  2.04it/s]
 80%|███████▉  | 8517/10712 [1:30:39<17:56,  2.04it/s]
 80%|███████▉  | 8518/10712 [1:30:39<17:55,  2.04it/s]
 80%|███████▉  | 8519/10712 [1:30:40<17:54,  2.04it/s]
 80%|███████▉  | 8520/10712 [1:30:40<17:53,  2.04it/s]
 80%|███████▉  | 8521/10712 [1:30:41<17:53,  2.04it/s]
 80%|███████▉  | 8522/10712 [1:30:41<17:52,  2.04it/s]
 80%|███████▉  | 8523/10712 [1:30:42<17:51,  2.04it/s]
 80%|███████▉  | 8524/10712 [1:30:42<17:52,  2.04it/s]
 80%|███████▉  | 8525/10712 [1:30:43<17:51,  2.04it/s]
                                                      
{'loss': 3.6329, 'grad_norm': 0.20471172034740448, 'learning_rate': 0.00012170839327475707, 'epoch': 0.8}

 80%|███████▉  | 8525/10712 [1:30:43<17:51,  2.04it/s]
 80%|███████▉  | 8526/10712 [1:30:43<17:51,  2.04it/s]
 80%|███████▉  | 8527/10712 [1:30:44<17:51,  2.04it/s]
 80%|███████▉  | 8528/10712 [1:30:44<17:51,  2.04it/s]
 80%|███████▉  | 8529/10712 [1:30:45<17:49,  2.04it/s]
 80%|███████▉  | 8530/10712 [1:30:45<17:48,  2.04it/s]
 80%|███████▉  | 8531/10712 [1:30:46<17:49,  2.04it/s]
 80%|███████▉  | 8532/10712 [1:30:46<17:48,  2.04it/s]
 80%|███████▉  | 8533/10712 [1:30:47<17:47,  2.04it/s]
 80%|███████▉  | 8534/10712 [1:30:47<17:47,  2.04it/s]
 80%|███████▉  | 8535/10712 [1:30:48<17:46,  2.04it/s]
 80%|███████▉  | 8536/10712 [1:30:48<17:46,  2.04it/s]
 80%|███████▉  | 8537/10712 [1:30:49<17:45,  2.04it/s]
 80%|███████▉  | 8538/10712 [1:30:49<17:45,  2.04it/s]
 80%|███████▉  | 8539/10712 [1:30:50<17:45,  2.04it/s]
 80%|███████▉  | 8540/10712 [1:30:50<17:43,  2.04it/s]
 80%|███████▉  | 8541/10712 [1:30:51<17:42,  2.04it/s]
 80%|███████▉  | 8542/10712 [1:30:51<17:41,  2.05it/s]
 80%|███████▉  | 8543/10712 [1:30:52<17:41,  2.04it/s]
 80%|███████▉  | 8544/10712 [1:30:52<17:41,  2.04it/s]
 80%|███████▉  | 8545/10712 [1:30:53<17:41,  2.04it/s]
 80%|███████▉  | 8546/10712 [1:30:53<17:40,  2.04it/s]
 80%|███████▉  | 8547/10712 [1:30:54<17:39,  2.04it/s]
 80%|███████▉  | 8548/10712 [1:30:54<17:40,  2.04it/s]
 80%|███████▉  | 8549/10712 [1:30:55<17:39,  2.04it/s]
 80%|███████▉  | 8550/10712 [1:30:55<17:38,  2.04it/s]
                                                      
{'loss': 3.6243, 'grad_norm': 0.20628201961517334, 'learning_rate': 0.00011905723387677897, 'epoch': 0.8}

 80%|███████▉  | 8550/10712 [1:30:55<17:38,  2.04it/s]
 80%|███████▉  | 8551/10712 [1:30:56<17:40,  2.04it/s]
 80%|███████▉  | 8552/10712 [1:30:56<17:39,  2.04it/s]
 80%|███████▉  | 8553/10712 [1:30:57<17:37,  2.04it/s]
 80%|███████▉  | 8554/10712 [1:30:57<17:36,  2.04it/s]
 80%|███████▉  | 8555/10712 [1:30:58<17:36,  2.04it/s]
 80%|███████▉  | 8556/10712 [1:30:58<17:37,  2.04it/s]
 80%|███████▉  | 8557/10712 [1:30:59<17:35,  2.04it/s]
 80%|███████▉  | 8558/10712 [1:30:59<17:34,  2.04it/s]
 80%|███████▉  | 8559/10712 [1:31:00<17:34,  2.04it/s]
 80%|███████▉  | 8560/10712 [1:31:00<17:34,  2.04it/s]
 80%|███████▉  | 8561/10712 [1:31:00<17:33,  2.04it/s]
 80%|███████▉  | 8562/10712 [1:31:01<17:32,  2.04it/s]
 80%|███████▉  | 8563/10712 [1:31:01<17:32,  2.04it/s]
 80%|███████▉  | 8564/10712 [1:31:02<17:34,  2.04it/s]
 80%|███████▉  | 8565/10712 [1:31:02<17:31,  2.04it/s]
 80%|███████▉  | 8566/10712 [1:31:03<17:30,  2.04it/s]
 80%|███████▉  | 8567/10712 [1:31:03<17:31,  2.04it/s]
 80%|███████▉  | 8568/10712 [1:31:04<17:31,  2.04it/s]
 80%|███████▉  | 8569/10712 [1:31:04<17:29,  2.04it/s]
 80%|████████  | 8570/10712 [1:31:05<17:29,  2.04it/s]
 80%|████████  | 8571/10712 [1:31:05<17:29,  2.04it/s]
 80%|████████  | 8572/10712 [1:31:06<17:28,  2.04it/s]
 80%|████████  | 8573/10712 [1:31:06<17:28,  2.04it/s]
 80%|████████  | 8574/10712 [1:31:07<17:27,  2.04it/s]
 80%|████████  | 8575/10712 [1:31:07<17:27,  2.04it/s]{'loss': 3.6205, 'grad_norm': 0.20690666139125824, 'learning_rate': 0.00011643136064706706, 'epoch': 0.8}

                                                      

 80%|████████  | 8575/10712 [1:31:07<17:27,  2.04it/s]
 80%|████████  | 8576/10712 [1:31:08<17:27,  2.04it/s]
 80%|████████  | 8577/10712 [1:31:08<17:26,  2.04it/s]
 80%|████████  | 8578/10712 [1:31:09<17:26,  2.04it/s]
 80%|████████  | 8579/10712 [1:31:09<17:25,  2.04it/s]
 80%|████████  | 8580/10712 [1:31:10<17:24,  2.04it/s]
 80%|████████  | 8581/10712 [1:31:10<17:25,  2.04it/s]
 80%|████████  | 8582/10712 [1:31:11<17:24,  2.04it/s]
 80%|████████  | 8583/10712 [1:31:11<17:23,  2.04it/s]
 80%|████████  | 8584/10712 [1:31:12<17:23,  2.04it/s]
 80%|████████  | 8585/10712 [1:31:12<17:22,  2.04it/s]
 80%|████████  | 8586/10712 [1:31:13<17:22,  2.04it/s]
 80%|████████  | 8587/10712 [1:31:13<17:20,  2.04it/s]
 80%|████████  | 8588/10712 [1:31:14<17:21,  2.04it/s]
 80%|████████  | 8589/10712 [1:31:14<17:21,  2.04it/s]
 80%|████████  | 8590/10712 [1:31:15<17:20,  2.04it/s]
 80%|████████  | 8591/10712 [1:31:15<17:19,  2.04it/s]
 80%|████████  | 8592/10712 [1:31:16<17:19,  2.04it/s]
 80%|████████  | 8593/10712 [1:31:16<17:20,  2.04it/s]
 80%|████████  | 8594/10712 [1:31:17<17:19,  2.04it/s]
 80%|████████  | 8595/10712 [1:31:17<17:17,  2.04it/s]
 80%|████████  | 8596/10712 [1:31:18<17:17,  2.04it/s]
 80%|████████  | 8597/10712 [1:31:18<17:16,  2.04it/s]
 80%|████████  | 8598/10712 [1:31:19<17:17,  2.04it/s]
 80%|████████  | 8599/10712 [1:31:19<17:15,  2.04it/s]
 80%|████████  | 8600/10712 [1:31:20<17:15,  2.04it/s]
                                                      
{'loss': 3.6301, 'grad_norm': 0.20802810788154602, 'learning_rate': 0.00011383094788548492, 'epoch': 0.8}

 80%|████████  | 8600/10712 [1:31:20<17:15,  2.04it/s]
 80%|████████  | 8601/10712 [1:31:20<17:16,  2.04it/s]
 80%|████████  | 8602/10712 [1:31:21<17:15,  2.04it/s]
 80%|████████  | 8603/10712 [1:31:21<17:13,  2.04it/s]
 80%|████████  | 8604/10712 [1:31:22<17:12,  2.04it/s]
 80%|████████  | 8605/10712 [1:31:22<17:13,  2.04it/s]
 80%|████████  | 8606/10712 [1:31:23<17:12,  2.04it/s]
 80%|████████  | 8607/10712 [1:31:23<17:10,  2.04it/s]
 80%|████████  | 8608/10712 [1:31:24<17:09,  2.04it/s]
 80%|████████  | 8609/10712 [1:31:24<17:10,  2.04it/s]
 80%|████████  | 8610/10712 [1:31:25<17:11,  2.04it/s]
 80%|████████  | 8611/10712 [1:31:25<17:10,  2.04it/s]
 80%|████████  | 8612/10712 [1:31:25<17:09,  2.04it/s]
 80%|████████  | 8613/10712 [1:31:26<17:09,  2.04it/s]
 80%|████████  | 8614/10712 [1:31:26<17:09,  2.04it/s]
 80%|████████  | 8615/10712 [1:31:27<17:07,  2.04it/s]
 80%|████████  | 8616/10712 [1:31:27<17:06,  2.04it/s]
 80%|████████  | 8617/10712 [1:31:28<17:06,  2.04it/s]
 80%|████████  | 8618/10712 [1:31:28<17:06,  2.04it/s]
 80%|████████  | 8619/10712 [1:31:29<17:05,  2.04it/s]
 80%|████████  | 8620/10712 [1:31:29<17:05,  2.04it/s]
 80%|████████  | 8621/10712 [1:31:30<17:05,  2.04it/s]
 80%|████████  | 8622/10712 [1:31:30<17:05,  2.04it/s]
 80%|████████  | 8623/10712 [1:31:31<17:03,  2.04it/s]
 81%|████████  | 8624/10712 [1:31:31<17:03,  2.04it/s]
 81%|████████  | 8625/10712 [1:31:32<17:03,  2.04it/s]
                                                      
{'loss': 3.6216, 'grad_norm': 0.20221181213855743, 'learning_rate': 0.00011125616820188544, 'epoch': 0.81}

 81%|████████  | 8625/10712 [1:31:32<17:03,  2.04it/s]
 81%|████████  | 8626/10712 [1:31:32<17:04,  2.04it/s]
 81%|████████  | 8627/10712 [1:31:33<17:04,  2.04it/s]
 81%|████████  | 8628/10712 [1:31:33<17:03,  2.04it/s]
 81%|████████  | 8629/10712 [1:31:34<17:01,  2.04it/s]
 81%|████████  | 8630/10712 [1:31:34<17:02,  2.04it/s]
 81%|████████  | 8631/10712 [1:31:35<17:01,  2.04it/s]
 81%|████████  | 8632/10712 [1:31:35<17:00,  2.04it/s]
 81%|████████  | 8633/10712 [1:31:36<16:59,  2.04it/s]
 81%|████████  | 8634/10712 [1:31:36<16:58,  2.04it/s]
 81%|████████  | 8635/10712 [1:31:37<16:58,  2.04it/s]
 81%|████████  | 8636/10712 [1:31:37<16:56,  2.04it/s]
 81%|████████  | 8637/10712 [1:31:38<16:56,  2.04it/s]
 81%|████████  | 8638/10712 [1:31:38<16:56,  2.04it/s]
 81%|████████  | 8639/10712 [1:31:39<16:56,  2.04it/s]
 81%|████████  | 8640/10712 [1:31:39<16:54,  2.04it/s]
 81%|████████  | 8641/10712 [1:31:40<16:54,  2.04it/s]
 81%|████████  | 8642/10712 [1:31:40<16:54,  2.04it/s]
 81%|████████  | 8643/10712 [1:31:41<16:54,  2.04it/s]
 81%|████████  | 8644/10712 [1:31:41<16:53,  2.04it/s]
 81%|████████  | 8645/10712 [1:31:42<16:53,  2.04it/s]
 81%|████████  | 8646/10712 [1:31:42<16:53,  2.04it/s]
 81%|████████  | 8647/10712 [1:31:43<16:52,  2.04it/s]
 81%|████████  | 8648/10712 [1:31:43<16:52,  2.04it/s]
 81%|████████  | 8649/10712 [1:31:44<16:51,  2.04it/s]
 81%|████████  | 8650/10712 [1:31:44<16:51,  2.04it/s]
                                                      {'loss': 3.6307, 'grad_norm': 0.2071399688720703, 'learning_rate': 0.00010870719250465161, 'epoch': 0.81}

 81%|████████  | 8650/10712 [1:31:44<16:51,  2.04it/s]
 81%|████████  | 8651/10712 [1:31:45<16:53,  2.03it/s]
 81%|████████  | 8652/10712 [1:31:45<16:51,  2.04it/s]
 81%|████████  | 8653/10712 [1:31:46<16:50,  2.04it/s]
 81%|████████  | 8654/10712 [1:31:46<16:50,  2.04it/s]
 81%|████████  | 8655/10712 [1:31:47<16:50,  2.03it/s]
 81%|████████  | 8656/10712 [1:31:47<16:48,  2.04it/s]
 81%|████████  | 8657/10712 [1:31:48<16:48,  2.04it/s]
 81%|████████  | 8658/10712 [1:31:48<16:48,  2.04it/s]
 81%|████████  | 8659/10712 [1:31:49<16:47,  2.04it/s]
 81%|████████  | 8660/10712 [1:31:49<16:47,  2.04it/s]
 81%|████████  | 8661/10712 [1:31:50<16:46,  2.04it/s]
 81%|████████  | 8662/10712 [1:31:50<16:45,  2.04it/s]
 81%|████████  | 8663/10712 [1:31:51<16:44,  2.04it/s]
 81%|████████  | 8664/10712 [1:31:51<16:44,  2.04it/s]
 81%|████████  | 8665/10712 [1:31:51<16:43,  2.04it/s]
 81%|████████  | 8666/10712 [1:31:52<16:42,  2.04it/s]
 81%|████████  | 8667/10712 [1:31:52<16:40,  2.04it/s]
 81%|████████  | 8668/10712 [1:31:53<16:42,  2.04it/s]
 81%|████████  | 8669/10712 [1:31:53<16:40,  2.04it/s]
 81%|████████  | 8670/10712 [1:31:54<16:39,  2.04it/s]
 81%|████████  | 8671/10712 [1:31:54<16:38,  2.04it/s]
 81%|████████  | 8672/10712 [1:31:55<16:39,  2.04it/s]
 81%|████████  | 8673/10712 [1:31:55<16:38,  2.04it/s]
 81%|████████  | 8674/10712 [1:31:56<16:37,  2.04it/s]
 81%|████████  | 8675/10712 [1:31:56<16:35,  2.05it/s]{'loss': 3.6224, 'grad_norm': 0.20434370636940002, 'learning_rate': 0.00010618418998935358, 'epoch': 0.81}
                                                      

 81%|████████  | 8675/10712 [1:31:56<16:35,  2.05it/s]
 81%|████████  | 8676/10712 [1:31:57<16:37,  2.04it/s]
 81%|████████  | 8677/10712 [1:31:57<16:37,  2.04it/s]
 81%|████████  | 8678/10712 [1:31:58<16:35,  2.04it/s]
 81%|████████  | 8679/10712 [1:31:58<16:35,  2.04it/s]
 81%|████████  | 8680/10712 [1:31:59<16:35,  2.04it/s]
 81%|████████  | 8681/10712 [1:31:59<16:35,  2.04it/s]
 81%|████████  | 8682/10712 [1:32:00<16:34,  2.04it/s]
 81%|████████  | 8683/10712 [1:32:00<16:33,  2.04it/s]
 81%|████████  | 8684/10712 [1:32:01<16:33,  2.04it/s]
 81%|████████  | 8685/10712 [1:32:01<16:32,  2.04it/s]
 81%|████████  | 8686/10712 [1:32:02<16:32,  2.04it/s]
 81%|████████  | 8687/10712 [1:32:02<16:31,  2.04it/s]
 81%|████████  | 8688/10712 [1:32:03<16:31,  2.04it/s]
 81%|████████  | 8689/10712 [1:32:03<16:32,  2.04it/s]
 81%|████████  | 8690/10712 [1:32:04<16:31,  2.04it/s]
 81%|████████  | 8691/10712 [1:32:04<16:29,  2.04it/s]
 81%|████████  | 8692/10712 [1:32:05<16:30,  2.04it/s]
 81%|████████  | 8693/10712 [1:32:05<16:29,  2.04it/s]
 81%|████████  | 8694/10712 [1:32:06<16:29,  2.04it/s]
 81%|████████  | 8695/10712 [1:32:06<16:27,  2.04it/s]
 81%|████████  | 8696/10712 [1:32:07<16:29,  2.04it/s]
 81%|████████  | 8697/10712 [1:32:07<16:28,  2.04it/s]
 81%|████████  | 8698/10712 [1:32:08<16:27,  2.04it/s]
 81%|████████  | 8699/10712 [1:32:08<16:26,  2.04it/s]
 81%|████████  | 8700/10712 [1:32:09<16:26,  2.04it/s]
                                                      
{'loss': 3.6168, 'grad_norm': 0.2029978334903717, 'learning_rate': 0.00010368732812751652, 'epoch': 0.81}

 81%|████████  | 8700/10712 [1:32:09<16:26,  2.04it/s]
 81%|████████  | 8701/10712 [1:32:09<16:26,  2.04it/s]
 81%|████████  | 8702/10712 [1:32:10<16:24,  2.04it/s]
 81%|████████  | 8703/10712 [1:32:10<16:24,  2.04it/s]
 81%|████████▏ | 8704/10712 [1:32:11<16:24,  2.04it/s]
 81%|████████▏ | 8705/10712 [1:32:11<16:24,  2.04it/s]
 81%|████████▏ | 8706/10712 [1:32:12<16:21,  2.04it/s]
 81%|████████▏ | 8707/10712 [1:32:12<16:22,  2.04it/s]
 81%|████████▏ | 8708/10712 [1:32:13<16:21,  2.04it/s]
 81%|████████▏ | 8709/10712 [1:32:13<16:21,  2.04it/s]
 81%|████████▏ | 8710/10712 [1:32:14<16:20,  2.04it/s]
 81%|████████▏ | 8711/10712 [1:32:14<16:20,  2.04it/s]
 81%|████████▏ | 8712/10712 [1:32:15<16:19,  2.04it/s]
 81%|████████▏ | 8713/10712 [1:32:15<16:18,  2.04it/s]
 81%|████████▏ | 8714/10712 [1:32:15<16:17,  2.04it/s]
 81%|████████▏ | 8715/10712 [1:32:16<16:17,  2.04it/s]
 81%|████████▏ | 8716/10712 [1:32:16<16:17,  2.04it/s]
 81%|████████▏ | 8717/10712 [1:32:17<16:16,  2.04it/s]
 81%|████████▏ | 8718/10712 [1:32:17<16:16,  2.04it/s]
 81%|████████▏ | 8719/10712 [1:32:18<16:17,  2.04it/s]
 81%|████████▏ | 8720/10712 [1:32:18<16:16,  2.04it/s]
 81%|████████▏ | 8721/10712 [1:32:19<16:15,  2.04it/s]
 81%|████████▏ | 8722/10712 [1:32:19<16:16,  2.04it/s]
 81%|████████▏ | 8723/10712 [1:32:20<16:15,  2.04it/s]
 81%|████████▏ | 8724/10712 [1:32:20<16:14,  2.04it/s]
 81%|████████▏ | 8725/10712 [1:32:21<16:14,  2.04it/s]
                                                      
{'loss': 3.6186, 'grad_norm': 0.21850678324699402, 'learning_rate': 0.00010121677265550522, 'epoch': 0.81}

 81%|████████▏ | 8725/10712 [1:32:21<16:14,  2.04it/s]
 81%|████████▏ | 8726/10712 [1:32:21<16:15,  2.04it/s]
 81%|████████▏ | 8727/10712 [1:32:22<16:13,  2.04it/s]
 81%|████████▏ | 8728/10712 [1:32:22<16:11,  2.04it/s]
 81%|████████▏ | 8729/10712 [1:32:23<16:11,  2.04it/s]
 81%|████████▏ | 8730/10712 [1:32:23<16:12,  2.04it/s]
 82%|████████▏ | 8731/10712 [1:32:24<16:10,  2.04it/s]
 82%|████████▏ | 8732/10712 [1:32:24<16:10,  2.04it/s]
 82%|████████▏ | 8733/10712 [1:32:25<16:09,  2.04it/s]
 82%|████████▏ | 8734/10712 [1:32:25<16:10,  2.04it/s]
 82%|████████▏ | 8735/10712 [1:32:26<16:08,  2.04it/s]
 82%|████████▏ | 8736/10712 [1:32:26<16:07,  2.04it/s]
 82%|████████▏ | 8737/10712 [1:32:27<16:07,  2.04it/s]
 82%|████████▏ | 8738/10712 [1:32:27<16:07,  2.04it/s]
 82%|████████▏ | 8739/10712 [1:32:28<16:07,  2.04it/s]
 82%|████████▏ | 8740/10712 [1:32:28<16:06,  2.04it/s]
 82%|████████▏ | 8741/10712 [1:32:29<16:05,  2.04it/s]
 82%|████████▏ | 8742/10712 [1:32:29<16:05,  2.04it/s]
 82%|████████▏ | 8743/10712 [1:32:30<16:05,  2.04it/s]
 82%|████████▏ | 8744/10712 [1:32:30<16:05,  2.04it/s]
 82%|████████▏ | 8745/10712 [1:32:31<16:04,  2.04it/s]
 82%|████████▏ | 8746/10712 [1:32:31<16:03,  2.04it/s]
 82%|████████▏ | 8747/10712 [1:32:32<16:02,  2.04it/s]
 82%|████████▏ | 8748/10712 [1:32:32<16:02,  2.04it/s]
 82%|████████▏ | 8749/10712 [1:32:33<16:01,  2.04it/s]
 82%|████████▏ | 8750/10712 [1:32:33<16:00,  2.04it/s]
                                                      {'loss': 3.6236, 'grad_norm': 0.20813213288784027, 'learning_rate': 9.877268756352204e-05, 'epoch': 0.82}

 82%|████████▏ | 8750/10712 [1:32:33<16:00,  2.04it/s]
 82%|████████▏ | 8751/10712 [1:32:34<16:01,  2.04it/s]
 82%|████████▏ | 8752/10712 [1:32:34<16:01,  2.04it/s]
 82%|████████▏ | 8753/10712 [1:32:35<15:59,  2.04it/s]
 82%|████████▏ | 8754/10712 [1:32:35<15:59,  2.04it/s]
 82%|████████▏ | 8755/10712 [1:32:36<15:59,  2.04it/s]
 82%|████████▏ | 8756/10712 [1:32:36<15:58,  2.04it/s]
 82%|████████▏ | 8757/10712 [1:32:37<15:57,  2.04it/s]
 82%|████████▏ | 8758/10712 [1:32:37<15:56,  2.04it/s]
 82%|████████▏ | 8759/10712 [1:32:38<15:57,  2.04it/s]
 82%|████████▏ | 8760/10712 [1:32:38<15:57,  2.04it/s]
 82%|████████▏ | 8761/10712 [1:32:39<15:55,  2.04it/s]
 82%|████████▏ | 8762/10712 [1:32:39<15:54,  2.04it/s]
 82%|████████▏ | 8763/10712 [1:32:40<15:54,  2.04it/s]
 82%|████████▏ | 8764/10712 [1:32:40<15:53,  2.04it/s]
 82%|████████▏ | 8765/10712 [1:32:40<15:53,  2.04it/s]
 82%|████████▏ | 8766/10712 [1:32:41<15:52,  2.04it/s]
 82%|████████▏ | 8767/10712 [1:32:41<15:54,  2.04it/s]
 82%|████████▏ | 8768/10712 [1:32:42<15:53,  2.04it/s]
 82%|████████▏ | 8769/10712 [1:32:42<15:52,  2.04it/s]
 82%|████████▏ | 8770/10712 [1:32:43<15:53,  2.04it/s]
 82%|████████▏ | 8771/10712 [1:32:43<15:52,  2.04it/s]
 82%|████████▏ | 8772/10712 [1:32:44<15:50,  2.04it/s]
 82%|████████▏ | 8773/10712 [1:32:44<15:50,  2.04it/s]
 82%|████████▏ | 8774/10712 [1:32:45<15:50,  2.04it/s]
 82%|████████▏ | 8775/10712 [1:32:45<15:49,  2.04it/s]
                                                      
{'loss': 3.6145, 'grad_norm': 0.20480607450008392, 'learning_rate': 9.635523508472227e-05, 'epoch': 0.82}

 82%|████████▏ | 8775/10712 [1:32:45<15:49,  2.04it/s]
 82%|████████▏ | 8776/10712 [1:32:46<15:49,  2.04it/s]
 82%|████████▏ | 8777/10712 [1:32:46<15:49,  2.04it/s]
 82%|████████▏ | 8778/10712 [1:32:47<15:48,  2.04it/s]
 82%|████████▏ | 8779/10712 [1:32:47<15:47,  2.04it/s]
 82%|████████▏ | 8780/10712 [1:32:48<15:48,  2.04it/s]
 82%|████████▏ | 8781/10712 [1:32:48<15:47,  2.04it/s]
 82%|████████▏ | 8782/10712 [1:32:49<15:46,  2.04it/s]
 82%|████████▏ | 8783/10712 [1:32:49<15:45,  2.04it/s]
 82%|████████▏ | 8784/10712 [1:32:50<15:45,  2.04it/s]
 82%|████████▏ | 8785/10712 [1:32:50<15:45,  2.04it/s]
 82%|████████▏ | 8786/10712 [1:32:51<15:43,  2.04it/s]
 82%|████████▏ | 8787/10712 [1:32:51<15:44,  2.04it/s]
 82%|████████▏ | 8788/10712 [1:32:52<15:43,  2.04it/s]
 82%|████████▏ | 8789/10712 [1:32:52<15:42,  2.04it/s]
 82%|████████▏ | 8790/10712 [1:32:53<15:43,  2.04it/s]
 82%|████████▏ | 8791/10712 [1:32:53<15:42,  2.04it/s]
 82%|████████▏ | 8792/10712 [1:32:54<15:41,  2.04it/s]
 82%|████████▏ | 8793/10712 [1:32:54<15:40,  2.04it/s]
 82%|████████▏ | 8794/10712 [1:32:55<15:40,  2.04it/s]
 82%|████████▏ | 8795/10712 [1:32:55<15:39,  2.04it/s]
 82%|████████▏ | 8796/10712 [1:32:56<15:39,  2.04it/s]
 82%|████████▏ | 8797/10712 [1:32:56<15:38,  2.04it/s]
 82%|████████▏ | 8798/10712 [1:32:57<15:38,  2.04it/s]
 82%|████████▏ | 8799/10712 [1:32:57<15:37,  2.04it/s]
 82%|████████▏ | 8800/10712 [1:32:58<15:36,  2.04it/s]{'loss': 3.6174, 'grad_norm': 0.20440484583377838, 'learning_rate': 9.396457568444517e-05, 'epoch': 0.82}

                                                      

 82%|████████▏ | 8800/10712 [1:32:58<15:36,  2.04it/s]
 82%|████████▏ | 8801/10712 [1:32:58<15:37,  2.04it/s]
 82%|████████▏ | 8802/10712 [1:32:59<15:36,  2.04it/s]
 82%|████████▏ | 8803/10712 [1:32:59<15:36,  2.04it/s]
 82%|████████▏ | 8804/10712 [1:33:00<15:35,  2.04it/s]
 82%|████████▏ | 8805/10712 [1:33:00<15:34,  2.04it/s]
 82%|████████▏ | 8806/10712 [1:33:01<15:34,  2.04it/s]
 82%|████████▏ | 8807/10712 [1:33:01<15:33,  2.04it/s]
 82%|████████▏ | 8808/10712 [1:33:02<15:33,  2.04it/s]
 82%|████████▏ | 8809/10712 [1:33:02<15:33,  2.04it/s]
 82%|████████▏ | 8810/10712 [1:33:03<15:32,  2.04it/s]
 82%|████████▏ | 8811/10712 [1:33:03<15:31,  2.04it/s]
 82%|████████▏ | 8812/10712 [1:33:04<15:31,  2.04it/s]
 82%|████████▏ | 8813/10712 [1:33:04<15:30,  2.04it/s]
 82%|████████▏ | 8814/10712 [1:33:05<15:29,  2.04it/s]
 82%|████████▏ | 8815/10712 [1:33:05<15:28,  2.04it/s]
 82%|████████▏ | 8816/10712 [1:33:05<15:29,  2.04it/s]
 82%|████████▏ | 8817/10712 [1:33:06<18:00,  1.75it/s]
 82%|████████▏ | 8818/10712 [1:33:07<17:13,  1.83it/s]
 82%|████████▏ | 8819/10712 [1:33:07<16:41,  1.89it/s]
 82%|████████▏ | 8820/10712 [1:33:08<16:18,  1.93it/s]
 82%|████████▏ | 8821/10712 [1:33:08<16:01,  1.97it/s]
 82%|████████▏ | 8822/10712 [1:33:09<15:51,  1.99it/s]
 82%|████████▏ | 8823/10712 [1:33:09<15:44,  2.00it/s]
 82%|████████▏ | 8824/10712 [1:33:10<15:38,  2.01it/s]
 82%|████████▏ | 8825/10712 [1:33:10<15:33,  2.02it/s]{'loss': 3.6236, 'grad_norm': 0.20961783826351166, 'learning_rate': 9.160086804956214e-05, 'epoch': 0.82}
                                                      

 82%|████████▏ | 8825/10712 [1:33:10<15:33,  2.02it/s]
 82%|████████▏ | 8826/10712 [1:33:11<15:31,  2.03it/s]
 82%|████████▏ | 8827/10712 [1:33:11<15:29,  2.03it/s]
 82%|████████▏ | 8828/10712 [1:33:12<15:27,  2.03it/s]
 82%|████████▏ | 8829/10712 [1:33:12<15:25,  2.03it/s]
 82%|████████▏ | 8830/10712 [1:33:13<15:24,  2.04it/s]
 82%|████████▏ | 8831/10712 [1:33:13<15:23,  2.04it/s]
 82%|████████▏ | 8832/10712 [1:33:14<15:22,  2.04it/s]
 82%|████████▏ | 8833/10712 [1:33:14<15:20,  2.04it/s]
 82%|████████▏ | 8834/10712 [1:33:15<15:20,  2.04it/s]
 82%|████████▏ | 8835/10712 [1:33:15<15:20,  2.04it/s]
 82%|████████▏ | 8836/10712 [1:33:16<15:19,  2.04it/s]
 82%|████████▏ | 8837/10712 [1:33:16<15:17,  2.04it/s]
 83%|████████▎ | 8838/10712 [1:33:17<15:17,  2.04it/s]
 83%|████████▎ | 8839/10712 [1:33:17<15:17,  2.04it/s]
 83%|████████▎ | 8840/10712 [1:33:18<15:16,  2.04it/s]
 83%|████████▎ | 8841/10712 [1:33:18<15:15,  2.04it/s]
 83%|████████▎ | 8842/10712 [1:33:18<15:15,  2.04it/s]
 83%|████████▎ | 8843/10712 [1:33:19<15:16,  2.04it/s]
 83%|████████▎ | 8844/10712 [1:33:19<15:15,  2.04it/s]
 83%|████████▎ | 8845/10712 [1:33:20<15:14,  2.04it/s]
 83%|████████▎ | 8846/10712 [1:33:20<15:14,  2.04it/s]
 83%|████████▎ | 8847/10712 [1:33:21<15:14,  2.04it/s]
 83%|████████▎ | 8848/10712 [1:33:21<15:13,  2.04it/s]
 83%|████████▎ | 8849/10712 [1:33:22<15:12,  2.04it/s]
 83%|████████▎ | 8850/10712 [1:33:22<15:12,  2.04it/s]{'loss': 3.6185, 'grad_norm': 0.21637071669101715, 'learning_rate': 8.926426907794455e-05, 'epoch': 0.83}

                                                      

 83%|████████▎ | 8850/10712 [1:33:22<15:12,  2.04it/s]
 83%|████████▎ | 8851/10712 [1:33:23<15:13,  2.04it/s]
 83%|████████▎ | 8852/10712 [1:33:23<15:11,  2.04it/s]
 83%|████████▎ | 8853/10712 [1:33:24<15:11,  2.04it/s]
 83%|████████▎ | 8854/10712 [1:33:24<15:11,  2.04it/s]
 83%|████████▎ | 8855/10712 [1:33:25<15:09,  2.04it/s]
 83%|████████▎ | 8856/10712 [1:33:25<15:08,  2.04it/s]
 83%|████████▎ | 8857/10712 [1:33:26<15:09,  2.04it/s]
 83%|████████▎ | 8858/10712 [1:33:26<15:08,  2.04it/s]
 83%|████████▎ | 8859/10712 [1:33:27<15:07,  2.04it/s]
 83%|████████▎ | 8860/10712 [1:33:28<17:33,  1.76it/s]
 83%|████████▎ | 8861/10712 [1:33:28<16:49,  1.83it/s]
 83%|████████▎ | 8862/10712 [1:33:29<16:18,  1.89it/s]
 83%|████████▎ | 8863/10712 [1:33:29<15:55,  1.93it/s]
 83%|████████▎ | 8864/10712 [1:33:30<15:40,  1.96it/s]
 83%|████████▎ | 8865/10712 [1:33:30<15:30,  1.98it/s]
 83%|████████▎ | 8866/10712 [1:33:31<15:22,  2.00it/s]
 83%|████████▎ | 8867/10712 [1:33:31<15:16,  2.01it/s]
 83%|████████▎ | 8868/10712 [1:33:32<15:11,  2.02it/s]
 83%|████████▎ | 8869/10712 [1:33:32<15:10,  2.02it/s]
 83%|████████▎ | 8870/10712 [1:33:32<15:07,  2.03it/s]
 83%|████████▎ | 8871/10712 [1:33:33<15:05,  2.03it/s]
 83%|████████▎ | 8872/10712 [1:33:33<15:02,  2.04it/s]
 83%|████████▎ | 8873/10712 [1:33:34<15:02,  2.04it/s]
 83%|████████▎ | 8874/10712 [1:33:34<15:01,  2.04it/s]
 83%|████████▎ | 8875/10712 [1:33:35<15:01,  2.04it/s]{'loss': 3.6167, 'grad_norm': 0.20315676927566528, 'learning_rate': 8.695493386804792e-05, 'epoch': 0.83}

                                                      

 83%|████████▎ | 8875/10712 [1:33:35<15:01,  2.04it/s]
 83%|████████▎ | 8876/10712 [1:33:35<15:00,  2.04it/s]
 83%|████████▎ | 8877/10712 [1:33:36<15:00,  2.04it/s]
 83%|████████▎ | 8878/10712 [1:33:36<15:00,  2.04it/s]
 83%|████████▎ | 8879/10712 [1:33:37<14:59,  2.04it/s]
 83%|████████▎ | 8880/10712 [1:33:37<14:57,  2.04it/s]
 83%|████████▎ | 8881/10712 [1:33:38<14:56,  2.04it/s]
 83%|████████▎ | 8882/10712 [1:33:38<14:57,  2.04it/s]
 83%|████████▎ | 8883/10712 [1:33:39<14:56,  2.04it/s]
 83%|████████▎ | 8884/10712 [1:33:39<14:56,  2.04it/s]
 83%|████████▎ | 8885/10712 [1:33:40<14:55,  2.04it/s]
 83%|████████▎ | 8886/10712 [1:33:40<14:55,  2.04it/s]
 83%|████████▎ | 8887/10712 [1:33:41<14:55,  2.04it/s]
 83%|████████▎ | 8888/10712 [1:33:41<14:54,  2.04it/s]
 83%|████████▎ | 8889/10712 [1:33:42<14:53,  2.04it/s]
 83%|████████▎ | 8890/10712 [1:33:42<14:53,  2.04it/s]
 83%|████████▎ | 8891/10712 [1:33:43<14:53,  2.04it/s]
 83%|████████▎ | 8892/10712 [1:33:43<14:52,  2.04it/s]
 83%|████████▎ | 8893/10712 [1:33:44<14:52,  2.04it/s]
 83%|████████▎ | 8894/10712 [1:33:44<14:52,  2.04it/s]
 83%|████████▎ | 8895/10712 [1:33:45<14:51,  2.04it/s]
 83%|████████▎ | 8896/10712 [1:33:45<14:50,  2.04it/s]
 83%|████████▎ | 8897/10712 [1:33:46<14:49,  2.04it/s]
 83%|████████▎ | 8898/10712 [1:33:46<14:50,  2.04it/s]
 83%|████████▎ | 8899/10712 [1:33:47<14:48,  2.04it/s]
 83%|████████▎ | 8900/10712 [1:33:47<14:48,  2.04it/s]
                                                      {'loss': 3.6121, 'grad_norm': 0.20434728264808655, 'learning_rate': 8.467301570861784e-05, 'epoch': 0.83}

 83%|████████▎ | 8900/10712 [1:33:47<14:48,  2.04it/s]
 83%|████████▎ | 8901/10712 [1:33:48<14:48,  2.04it/s]
 83%|████████▎ | 8902/10712 [1:33:48<14:47,  2.04it/s]
 83%|████████▎ | 8903/10712 [1:33:49<14:46,  2.04it/s]
 83%|████████▎ | 8904/10712 [1:33:49<14:46,  2.04it/s]
 83%|████████▎ | 8905/10712 [1:33:50<14:45,  2.04it/s]
 83%|████████▎ | 8906/10712 [1:33:50<14:44,  2.04it/s]
 83%|████████▎ | 8907/10712 [1:33:51<14:44,  2.04it/s]
 83%|████████▎ | 8908/10712 [1:33:51<14:45,  2.04it/s]
 83%|████████▎ | 8909/10712 [1:33:52<14:43,  2.04it/s]
 83%|████████▎ | 8910/10712 [1:33:52<14:41,  2.04it/s]
 83%|████████▎ | 8911/10712 [1:33:53<14:42,  2.04it/s]
 83%|████████▎ | 8912/10712 [1:33:53<14:42,  2.04it/s]
 83%|████████▎ | 8913/10712 [1:33:54<14:41,  2.04it/s]
 83%|████████▎ | 8914/10712 [1:33:54<14:40,  2.04it/s]
 83%|████████▎ | 8915/10712 [1:33:55<14:40,  2.04it/s]
 83%|████████▎ | 8916/10712 [1:33:55<14:40,  2.04it/s]
 83%|████████▎ | 8917/10712 [1:33:56<14:39,  2.04it/s]
 83%|████████▎ | 8918/10712 [1:33:56<14:38,  2.04it/s]
 83%|████████▎ | 8919/10712 [1:33:57<14:39,  2.04it/s]
 83%|████████▎ | 8920/10712 [1:33:57<14:37,  2.04it/s]
 83%|████████▎ | 8921/10712 [1:33:57<14:36,  2.04it/s]
 83%|████████▎ | 8922/10712 [1:33:58<14:37,  2.04it/s]
 83%|████████▎ | 8923/10712 [1:33:58<14:36,  2.04it/s]
 83%|████████▎ | 8924/10712 [1:33:59<14:35,  2.04it/s]
 83%|████████▎ | 8925/10712 [1:33:59<14:35,  2.04it/s]
                                                      
{'loss': 3.6117, 'grad_norm': 0.207365944981575, 'learning_rate': 8.241866606851428e-05, 'epoch': 0.83}

 83%|████████▎ | 8925/10712 [1:33:59<14:35,  2.04it/s]
 83%|████████▎ | 8926/10712 [1:34:00<14:36,  2.04it/s]
 83%|████████▎ | 8927/10712 [1:34:00<14:34,  2.04it/s]
 83%|████████▎ | 8928/10712 [1:34:01<14:33,  2.04it/s]
 83%|████████▎ | 8929/10712 [1:34:01<14:33,  2.04it/s]
 83%|████████▎ | 8930/10712 [1:34:02<14:33,  2.04it/s]
 83%|████████▎ | 8931/10712 [1:34:02<14:32,  2.04it/s]
 83%|████████▎ | 8932/10712 [1:34:03<14:31,  2.04it/s]
 83%|████████▎ | 8933/10712 [1:34:03<14:31,  2.04it/s]
 83%|████████▎ | 8934/10712 [1:34:04<14:31,  2.04it/s]
 83%|████████▎ | 8935/10712 [1:34:04<14:30,  2.04it/s]
 83%|████████▎ | 8936/10712 [1:34:05<14:29,  2.04it/s]
 83%|████████▎ | 8937/10712 [1:34:05<14:30,  2.04it/s]
 83%|████████▎ | 8938/10712 [1:34:06<14:30,  2.04it/s]
 83%|████████▎ | 8939/10712 [1:34:06<14:29,  2.04it/s]
 83%|████████▎ | 8940/10712 [1:34:07<14:29,  2.04it/s]
 83%|████████▎ | 8941/10712 [1:34:07<14:28,  2.04it/s]
 83%|████████▎ | 8942/10712 [1:34:08<14:27,  2.04it/s]
 83%|████████▎ | 8943/10712 [1:34:08<14:26,  2.04it/s]
 83%|████████▎ | 8944/10712 [1:34:09<14:26,  2.04it/s]
 84%|████████▎ | 8945/10712 [1:34:09<14:26,  2.04it/s]
 84%|████████▎ | 8946/10712 [1:34:10<14:26,  2.04it/s]
 84%|████████▎ | 8947/10712 [1:34:10<14:26,  2.04it/s]
 84%|████████▎ | 8948/10712 [1:34:11<14:25,  2.04it/s]
 84%|████████▎ | 8949/10712 [1:34:11<14:23,  2.04it/s]
 84%|████████▎ | 8950/10712 [1:34:12<14:23,  2.04it/s]{'loss': 3.6152, 'grad_norm': 0.20761935412883759, 'learning_rate': 8.019203458665802e-05, 'epoch': 0.84}
                                                      

 84%|████████▎ | 8950/10712 [1:34:12<14:23,  2.04it/s]
 84%|████████▎ | 8951/10712 [1:34:12<14:24,  2.04it/s]
 84%|████████▎ | 8952/10712 [1:34:13<14:23,  2.04it/s]
 84%|████████▎ | 8953/10712 [1:34:13<14:22,  2.04it/s]
 84%|████████▎ | 8954/10712 [1:34:14<14:22,  2.04it/s]
 84%|████████▎ | 8955/10712 [1:34:14<14:22,  2.04it/s]
 84%|████████▎ | 8956/10712 [1:34:15<14:20,  2.04it/s]
 84%|████████▎ | 8957/10712 [1:34:15<14:19,  2.04it/s]
 84%|████████▎ | 8958/10712 [1:34:16<14:19,  2.04it/s]
 84%|████████▎ | 8959/10712 [1:34:16<14:19,  2.04it/s]
 84%|████████▎ | 8960/10712 [1:34:17<14:18,  2.04it/s]
 84%|████████▎ | 8961/10712 [1:34:17<14:17,  2.04it/s]
 84%|████████▎ | 8962/10712 [1:34:18<14:17,  2.04it/s]
 84%|████████▎ | 8963/10712 [1:34:18<14:17,  2.04it/s]
 84%|████████▎ | 8964/10712 [1:34:19<14:16,  2.04it/s]
 84%|████████▎ | 8965/10712 [1:34:19<14:15,  2.04it/s]
 84%|████████▎ | 8966/10712 [1:34:20<14:16,  2.04it/s]
 84%|████████▎ | 8967/10712 [1:34:20<14:15,  2.04it/s]
 84%|████████▎ | 8968/10712 [1:34:21<14:14,  2.04it/s]
 84%|████████▎ | 8969/10712 [1:34:21<14:14,  2.04it/s]
 84%|████████▎ | 8970/10712 [1:34:22<14:14,  2.04it/s]
 84%|████████▎ | 8971/10712 [1:34:22<14:13,  2.04it/s]
 84%|████████▍ | 8972/10712 [1:34:22<14:12,  2.04it/s]
 84%|████████▍ | 8973/10712 [1:34:23<14:11,  2.04it/s]
 84%|████████▍ | 8974/10712 [1:34:23<14:12,  2.04it/s]
 84%|████████▍ | 8975/10712 [1:34:24<14:11,  2.04it/s]{'loss': 3.6135, 'grad_norm': 0.20291678607463837, 'learning_rate': 7.79932690620972e-05, 'epoch': 0.84}
                                                      

 84%|████████▍ | 8975/10712 [1:34:24<14:11,  2.04it/s]
 84%|████████▍ | 8976/10712 [1:34:24<14:12,  2.04it/s]
 84%|████████▍ | 8977/10712 [1:34:25<14:11,  2.04it/s]
 84%|████████▍ | 8978/10712 [1:34:25<14:10,  2.04it/s]
 84%|████████▍ | 8979/10712 [1:34:26<14:09,  2.04it/s]
 84%|████████▍ | 8980/10712 [1:34:26<14:08,  2.04it/s]
 84%|████████▍ | 8981/10712 [1:34:27<14:08,  2.04it/s]
 84%|████████▍ | 8982/10712 [1:34:27<14:07,  2.04it/s]
 84%|████████▍ | 8983/10712 [1:34:28<14:07,  2.04it/s]
 84%|████████▍ | 8984/10712 [1:34:28<14:07,  2.04it/s]
 84%|████████▍ | 8985/10712 [1:34:29<14:06,  2.04it/s]
 84%|████████▍ | 8986/10712 [1:34:29<14:05,  2.04it/s]
 84%|████████▍ | 8987/10712 [1:34:30<14:05,  2.04it/s]
 84%|████████▍ | 8988/10712 [1:34:30<14:05,  2.04it/s]
 84%|████████▍ | 8989/10712 [1:34:31<14:04,  2.04it/s]
 84%|████████▍ | 8990/10712 [1:34:31<14:04,  2.04it/s]
 84%|████████▍ | 8991/10712 [1:34:32<14:03,  2.04it/s]
 84%|████████▍ | 8992/10712 [1:34:32<14:03,  2.04it/s]
 84%|████████▍ | 8993/10712 [1:34:33<14:03,  2.04it/s]
 84%|████████▍ | 8994/10712 [1:34:33<14:01,  2.04it/s]
 84%|████████▍ | 8995/10712 [1:34:34<14:01,  2.04it/s]
 84%|████████▍ | 8996/10712 [1:34:34<14:00,  2.04it/s]
 84%|████████▍ | 8997/10712 [1:34:35<13:59,  2.04it/s]
 84%|████████▍ | 8998/10712 [1:34:35<13:59,  2.04it/s]
 84%|████████▍ | 8999/10712 [1:34:36<13:59,  2.04it/s]
 84%|████████▍ | 9000/10712 [1:34:36<13:59,  2.04it/s]
                                                      {'loss': 3.6071, 'grad_norm': 0.199598491191864, 'learning_rate': 7.582251544419754e-05, 'epoch': 0.84}

 84%|████████▍ | 9000/10712 [1:34:36<13:59,  2.04it/s]
 84%|████████▍ | 9001/10712 [1:34:37<13:59,  2.04it/s]
 84%|████████▍ | 9002/10712 [1:34:37<13:58,  2.04it/s]
 84%|████████▍ | 9003/10712 [1:34:38<13:58,  2.04it/s]
 84%|████████▍ | 9004/10712 [1:34:38<13:57,  2.04it/s]
 84%|████████▍ | 9005/10712 [1:34:39<13:56,  2.04it/s]
 84%|████████▍ | 9006/10712 [1:34:39<13:57,  2.04it/s]
 84%|████████▍ | 9007/10712 [1:34:40<13:56,  2.04it/s]
 84%|████████▍ | 9008/10712 [1:34:40<13:55,  2.04it/s]
 84%|████████▍ | 9009/10712 [1:34:41<13:54,  2.04it/s]
 84%|████████▍ | 9010/10712 [1:34:41<13:54,  2.04it/s]
 84%|████████▍ | 9011/10712 [1:34:42<13:54,  2.04it/s]
 84%|████████▍ | 9012/10712 [1:34:42<13:53,  2.04it/s]
 84%|████████▍ | 9013/10712 [1:34:43<13:52,  2.04it/s]
 84%|████████▍ | 9014/10712 [1:34:43<13:52,  2.04it/s]
 84%|████████▍ | 9015/10712 [1:34:44<13:53,  2.04it/s]
 84%|████████▍ | 9016/10712 [1:34:44<13:52,  2.04it/s]
 84%|████████▍ | 9017/10712 [1:34:45<13:51,  2.04it/s]
 84%|████████▍ | 9018/10712 [1:34:45<13:50,  2.04it/s]
 84%|████████▍ | 9019/10712 [1:34:46<13:50,  2.04it/s]
 84%|████████▍ | 9020/10712 [1:34:46<13:49,  2.04it/s]
 84%|████████▍ | 9021/10712 [1:34:47<13:48,  2.04it/s]
 84%|████████▍ | 9022/10712 [1:34:47<13:47,  2.04it/s]
 84%|████████▍ | 9023/10712 [1:34:47<13:47,  2.04it/s]
 84%|████████▍ | 9024/10712 [1:34:48<13:48,  2.04it/s]
 84%|████████▍ | 9025/10712 [1:34:48<13:47,  2.04it/s]{'loss': 3.6073, 'grad_norm': 0.2048661857843399, 'learning_rate': 7.367991782295391e-05, 'epoch': 0.84}

                                                      

 84%|████████▍ | 9025/10712 [1:34:48<13:47,  2.04it/s]
 84%|████████▍ | 9026/10712 [1:34:49<13:47,  2.04it/s]
 84%|████████▍ | 9027/10712 [1:34:49<13:46,  2.04it/s]
 84%|████████▍ | 9028/10712 [1:34:50<13:46,  2.04it/s]
 84%|████████▍ | 9029/10712 [1:34:50<13:44,  2.04it/s]
 84%|████████▍ | 9030/10712 [1:34:51<13:44,  2.04it/s]
 84%|████████▍ | 9031/10712 [1:34:51<13:43,  2.04it/s]
 84%|████████▍ | 9032/10712 [1:34:52<13:43,  2.04it/s]
 84%|████████▍ | 9033/10712 [1:34:52<13:42,  2.04it/s]
 84%|████████▍ | 9034/10712 [1:34:53<13:42,  2.04it/s]
 84%|████████▍ | 9035/10712 [1:34:53<13:42,  2.04it/s]
 84%|████████▍ | 9036/10712 [1:34:54<13:41,  2.04it/s]
 84%|████████▍ | 9037/10712 [1:34:54<13:42,  2.04it/s]
 84%|████████▍ | 9038/10712 [1:34:55<13:41,  2.04it/s]
 84%|████████▍ | 9039/10712 [1:34:55<13:40,  2.04it/s]
 84%|████████▍ | 9040/10712 [1:34:56<13:40,  2.04it/s]
 84%|████████▍ | 9041/10712 [1:34:56<13:40,  2.04it/s]
 84%|████████▍ | 9042/10712 [1:34:57<13:39,  2.04it/s]
 84%|████████▍ | 9043/10712 [1:34:57<13:40,  2.03it/s]
 84%|████████▍ | 9044/10712 [1:34:58<13:38,  2.04it/s]
 84%|████████▍ | 9045/10712 [1:34:58<13:37,  2.04it/s]
 84%|████████▍ | 9046/10712 [1:34:59<13:36,  2.04it/s]
 84%|████████▍ | 9047/10712 [1:34:59<13:36,  2.04it/s]
 84%|████████▍ | 9048/10712 [1:35:00<13:35,  2.04it/s]
 84%|████████▍ | 9049/10712 [1:35:00<13:35,  2.04it/s]
 84%|████████▍ | 9050/10712 [1:35:01<13:34,  2.04it/s]
                                                      
{'loss': 3.6073, 'grad_norm': 0.20698492228984833, 'learning_rate': 7.156561841942638e-05, 'epoch': 0.84}

 84%|████████▍ | 9050/10712 [1:35:01<13:34,  2.04it/s]
 84%|████████▍ | 9051/10712 [1:35:01<13:34,  2.04it/s]
 85%|████████▍ | 9052/10712 [1:35:02<13:33,  2.04it/s]
 85%|████████▍ | 9053/10712 [1:35:02<13:33,  2.04it/s]
 85%|████████▍ | 9054/10712 [1:35:03<13:33,  2.04it/s]
 85%|████████▍ | 9055/10712 [1:35:03<13:32,  2.04it/s]
 85%|████████▍ | 9056/10712 [1:35:04<13:31,  2.04it/s]
 85%|████████▍ | 9057/10712 [1:35:04<13:31,  2.04it/s]
 85%|████████▍ | 9058/10712 [1:35:05<13:31,  2.04it/s]
 85%|████████▍ | 9059/10712 [1:35:05<13:29,  2.04it/s]
 85%|████████▍ | 9060/10712 [1:35:06<13:30,  2.04it/s]
 85%|████████▍ | 9061/10712 [1:35:06<13:29,  2.04it/s]
 85%|████████▍ | 9062/10712 [1:35:07<13:28,  2.04it/s]
 85%|████████▍ | 9063/10712 [1:35:07<13:28,  2.04it/s]
 85%|████████▍ | 9064/10712 [1:35:08<13:28,  2.04it/s]
 85%|████████▍ | 9065/10712 [1:35:08<13:27,  2.04it/s]
 85%|████████▍ | 9066/10712 [1:35:09<13:26,  2.04it/s]
 85%|████████▍ | 9067/10712 [1:35:09<13:26,  2.04it/s]
 85%|████████▍ | 9068/10712 [1:35:10<13:25,  2.04it/s]
 85%|████████▍ | 9069/10712 [1:35:10<13:24,  2.04it/s]
 85%|████████▍ | 9070/10712 [1:35:11<13:23,  2.04it/s]
 85%|████████▍ | 9071/10712 [1:35:11<13:24,  2.04it/s]
 85%|████████▍ | 9072/10712 [1:35:12<13:24,  2.04it/s]
 85%|████████▍ | 9073/10712 [1:35:12<13:23,  2.04it/s]
 85%|████████▍ | 9074/10712 [1:35:12<13:23,  2.04it/s]
 85%|████████▍ | 9075/10712 [1:35:13<13:22,  2.04it/s]
                                                      {'loss': 3.6141, 'grad_norm': 0.1998717188835144, 'learning_rate': 6.947975757629937e-05, 'epoch': 0.85}

 85%|████████▍ | 9075/10712 [1:35:13<13:22,  2.04it/s]
 85%|████████▍ | 9076/10712 [1:35:13<13:22,  2.04it/s]
 85%|████████▍ | 9077/10712 [1:35:14<13:22,  2.04it/s]
 85%|████████▍ | 9078/10712 [1:35:14<13:22,  2.04it/s]
 85%|████████▍ | 9079/10712 [1:35:15<13:20,  2.04it/s]
 85%|████████▍ | 9080/10712 [1:35:15<13:20,  2.04it/s]
 85%|████████▍ | 9081/10712 [1:35:16<13:20,  2.04it/s]
 85%|████████▍ | 9082/10712 [1:35:16<13:20,  2.04it/s]
 85%|████████▍ | 9083/10712 [1:35:17<13:18,  2.04it/s]
 85%|████████▍ | 9084/10712 [1:35:17<13:18,  2.04it/s]
 85%|████████▍ | 9085/10712 [1:35:18<13:17,  2.04it/s]
 85%|████████▍ | 9086/10712 [1:35:18<13:17,  2.04it/s]
 85%|████████▍ | 9087/10712 [1:35:19<13:15,  2.04it/s]
 85%|████████▍ | 9088/10712 [1:35:19<13:15,  2.04it/s]
 85%|████████▍ | 9089/10712 [1:35:20<13:16,  2.04it/s]
 85%|████████▍ | 9090/10712 [1:35:20<13:15,  2.04it/s]
 85%|████████▍ | 9091/10712 [1:35:21<13:14,  2.04it/s]
 85%|████████▍ | 9092/10712 [1:35:21<13:14,  2.04it/s]
 85%|████████▍ | 9093/10712 [1:35:22<13:14,  2.04it/s]
 85%|████████▍ | 9094/10712 [1:35:22<13:13,  2.04it/s]
 85%|████████▍ | 9095/10712 [1:35:23<13:12,  2.04it/s]
 85%|████████▍ | 9096/10712 [1:35:23<13:12,  2.04it/s]
 85%|████████▍ | 9097/10712 [1:35:24<13:11,  2.04it/s]
 85%|████████▍ | 9098/10712 [1:35:24<13:11,  2.04it/s]
 85%|████████▍ | 9099/10712 [1:35:25<13:11,  2.04it/s]
 85%|████████▍ | 9100/10712 [1:35:25<13:10,  2.04it/s]{'loss': 3.6147, 'grad_norm': 0.20371860265731812, 'learning_rate': 6.742247374856664e-05, 'epoch': 0.85}
                                                      

 85%|████████▍ | 9100/10712 [1:35:25<13:10,  2.04it/s]
 85%|████████▍ | 9101/10712 [1:35:26<13:11,  2.04it/s]
 85%|████████▍ | 9102/10712 [1:35:26<13:09,  2.04it/s]
 85%|████████▍ | 9103/10712 [1:35:27<13:10,  2.04it/s]
 85%|████████▍ | 9104/10712 [1:35:27<13:09,  2.04it/s]
 85%|████████▍ | 9105/10712 [1:35:28<13:09,  2.04it/s]
 85%|████████▌ | 9106/10712 [1:35:28<13:09,  2.03it/s]
 85%|████████▌ | 9107/10712 [1:35:29<13:08,  2.04it/s]
 85%|████████▌ | 9108/10712 [1:35:29<13:07,  2.04it/s]
 85%|████████▌ | 9109/10712 [1:35:30<13:06,  2.04it/s]
 85%|████████▌ | 9110/10712 [1:35:30<13:05,  2.04it/s]
 85%|████████▌ | 9111/10712 [1:35:31<13:04,  2.04it/s]
 85%|████████▌ | 9112/10712 [1:35:31<13:03,  2.04it/s]
 85%|████████▌ | 9113/10712 [1:35:32<13:04,  2.04it/s]
 85%|████████▌ | 9114/10712 [1:35:32<13:03,  2.04it/s]
 85%|████████▌ | 9115/10712 [1:35:33<13:03,  2.04it/s]
 85%|████████▌ | 9116/10712 [1:35:33<13:02,  2.04it/s]
 85%|████████▌ | 9117/10712 [1:35:34<13:02,  2.04it/s]
 85%|████████▌ | 9118/10712 [1:35:34<13:01,  2.04it/s]
 85%|████████▌ | 9119/10712 [1:35:35<13:00,  2.04it/s]
 85%|████████▌ | 9120/10712 [1:35:35<13:00,  2.04it/s]
 85%|████████▌ | 9121/10712 [1:35:36<12:59,  2.04it/s]
 85%|████████▌ | 9122/10712 [1:35:36<12:58,  2.04it/s]
 85%|████████▌ | 9123/10712 [1:35:37<12:57,  2.04it/s]
 85%|████████▌ | 9124/10712 [1:35:37<12:58,  2.04it/s]
 85%|████████▌ | 9125/10712 [1:35:38<12:58,  2.04it/s]{'loss': 3.6092, 'grad_norm': 0.20381568372249603, 'learning_rate': 6.539390349434054e-05, 'epoch': 0.85}
                                                      

 85%|████████▌ | 9125/10712 [1:35:38<12:58,  2.04it/s]
 85%|████████▌ | 9126/10712 [1:35:38<12:58,  2.04it/s]
 85%|████████▌ | 9127/10712 [1:35:38<12:58,  2.04it/s]
 85%|████████▌ | 9128/10712 [1:35:39<12:57,  2.04it/s]
 85%|████████▌ | 9129/10712 [1:35:39<12:56,  2.04it/s]
 85%|████████▌ | 9130/10712 [1:35:40<12:55,  2.04it/s]
 85%|████████▌ | 9131/10712 [1:35:40<12:54,  2.04it/s]
 85%|████████▌ | 9132/10712 [1:35:41<12:54,  2.04it/s]
 85%|████████▌ | 9133/10712 [1:35:41<12:53,  2.04it/s]
 85%|████████▌ | 9134/10712 [1:35:42<12:53,  2.04it/s]
 85%|████████▌ | 9135/10712 [1:35:42<12:53,  2.04it/s]
 85%|████████▌ | 9136/10712 [1:35:43<12:52,  2.04it/s]
 85%|████████▌ | 9137/10712 [1:35:43<12:51,  2.04it/s]
 85%|████████▌ | 9138/10712 [1:35:44<12:51,  2.04it/s]
 85%|████████▌ | 9139/10712 [1:35:44<12:51,  2.04it/s]
 85%|████████▌ | 9140/10712 [1:35:45<12:50,  2.04it/s]
 85%|████████▌ | 9141/10712 [1:35:45<12:50,  2.04it/s]
 85%|████████▌ | 9142/10712 [1:35:46<12:50,  2.04it/s]
 85%|████████▌ | 9143/10712 [1:35:46<12:48,  2.04it/s]
 85%|████████▌ | 9144/10712 [1:35:47<12:48,  2.04it/s]
 85%|████████▌ | 9145/10712 [1:35:47<12:48,  2.04it/s]
 85%|████████▌ | 9146/10712 [1:35:48<12:48,  2.04it/s]
 85%|████████▌ | 9147/10712 [1:35:48<12:48,  2.04it/s]
 85%|████████▌ | 9148/10712 [1:35:49<12:47,  2.04it/s]
 85%|████████▌ | 9149/10712 [1:35:49<12:46,  2.04it/s]
 85%|████████▌ | 9150/10712 [1:35:50<12:46,  2.04it/s]{'loss': 3.6153, 'grad_norm': 0.19894002377986908, 'learning_rate': 6.339418146578746e-05, 'epoch': 0.85}

                                                      

 85%|████████▌ | 9150/10712 [1:35:50<12:46,  2.04it/s]
 85%|████████▌ | 9151/10712 [1:35:50<12:46,  2.04it/s]
 85%|████████▌ | 9152/10712 [1:35:51<12:45,  2.04it/s]
 85%|████████▌ | 9153/10712 [1:35:51<12:44,  2.04it/s]
 85%|████████▌ | 9154/10712 [1:35:52<12:44,  2.04it/s]
 85%|████████▌ | 9155/10712 [1:35:52<12:44,  2.04it/s]
 85%|████████▌ | 9156/10712 [1:35:53<12:42,  2.04it/s]
 85%|████████▌ | 9157/10712 [1:35:53<12:42,  2.04it/s]
 85%|████████▌ | 9158/10712 [1:35:54<12:41,  2.04it/s]
 86%|████████▌ | 9159/10712 [1:35:54<12:41,  2.04it/s]
 86%|████████▌ | 9160/10712 [1:35:55<12:40,  2.04it/s]
 86%|████████▌ | 9161/10712 [1:35:55<12:39,  2.04it/s]
 86%|████████▌ | 9162/10712 [1:35:56<12:39,  2.04it/s]
 86%|████████▌ | 9163/10712 [1:35:56<12:38,  2.04it/s]
 86%|████████▌ | 9164/10712 [1:35:57<12:38,  2.04it/s]
 86%|████████▌ | 9165/10712 [1:35:57<12:38,  2.04it/s]
 86%|████████▌ | 9166/10712 [1:35:58<12:38,  2.04it/s]
 86%|████████▌ | 9167/10712 [1:35:58<12:38,  2.04it/s]
 86%|████████▌ | 9168/10712 [1:35:59<12:36,  2.04it/s]
 86%|████████▌ | 9169/10712 [1:35:59<12:36,  2.04it/s]
 86%|████████▌ | 9170/10712 [1:36:00<12:36,  2.04it/s]
 86%|████████▌ | 9171/10712 [1:36:00<12:35,  2.04it/s]
 86%|████████▌ | 9172/10712 [1:36:01<12:34,  2.04it/s]
 86%|████████▌ | 9173/10712 [1:36:01<12:35,  2.04it/s]
 86%|████████▌ | 9174/10712 [1:36:02<12:35,  2.04it/s]
 86%|████████▌ | 9175/10712 [1:36:02<12:34,  2.04it/s]
                                                      {'loss': 3.6043, 'grad_norm': 0.20000189542770386, 'learning_rate': 6.142344040019049e-05, 'epoch': 0.86}


 86%|████████▌ | 9175/10712 [1:36:02<12:34,  2.04it/s]
 86%|████████▌ | 9176/10712 [1:36:03<12:34,  2.03it/s]
 86%|████████▌ | 9177/10712 [1:36:03<12:34,  2.04it/s]
 86%|████████▌ | 9178/10712 [1:36:04<12:32,  2.04it/s]
 86%|████████▌ | 9179/10712 [1:36:04<12:31,  2.04it/s]
 86%|████████▌ | 9180/10712 [1:36:04<12:31,  2.04it/s]
 86%|████████▌ | 9181/10712 [1:36:05<12:30,  2.04it/s]
 86%|████████▌ | 9182/10712 [1:36:05<12:30,  2.04it/s]
 86%|████████▌ | 9183/10712 [1:36:06<12:30,  2.04it/s]
 86%|████████▌ | 9184/10712 [1:36:06<12:29,  2.04it/s]
 86%|████████▌ | 9185/10712 [1:36:07<12:28,  2.04it/s]
 86%|████████▌ | 9186/10712 [1:36:07<12:28,  2.04it/s]
 86%|████████▌ | 9187/10712 [1:36:08<12:28,  2.04it/s]
 86%|████████▌ | 9188/10712 [1:36:08<12:27,  2.04it/s]
 86%|████████▌ | 9189/10712 [1:36:09<12:27,  2.04it/s]
 86%|████████▌ | 9190/10712 [1:36:09<12:26,  2.04it/s]
 86%|████████▌ | 9191/10712 [1:36:10<12:25,  2.04it/s]
 86%|████████▌ | 9192/10712 [1:36:10<12:24,  2.04it/s]
 86%|████████▌ | 9193/10712 [1:36:11<12:24,  2.04it/s]
 86%|████████▌ | 9194/10712 [1:36:11<12:23,  2.04it/s]
 86%|████████▌ | 9195/10712 [1:36:12<12:23,  2.04it/s]
 86%|████████▌ | 9196/10712 [1:36:12<12:22,  2.04it/s]
 86%|████████▌ | 9197/10712 [1:36:13<12:22,  2.04it/s]
 86%|████████▌ | 9198/10712 [1:36:13<12:22,  2.04it/s]
 86%|████████▌ | 9199/10712 [1:36:14<12:21,  2.04it/s]
 86%|████████▌ | 9200/10712 [1:36:14<12:22,  2.04it/s]{'loss': 3.6083, 'grad_norm': 0.20158882439136505, 'learning_rate': 5.948181111113765e-05, 'epoch': 0.86}

                                                      

 86%|████████▌ | 9200/10712 [1:36:14<12:22,  2.04it/s]
 86%|████████▌ | 9201/10712 [1:36:15<12:22,  2.04it/s]
 86%|████████▌ | 9202/10712 [1:36:15<12:21,  2.04it/s]
 86%|████████▌ | 9203/10712 [1:36:16<12:20,  2.04it/s]
 86%|████████▌ | 9204/10712 [1:36:16<12:19,  2.04it/s]
 86%|████████▌ | 9205/10712 [1:36:17<12:19,  2.04it/s]
 86%|████████▌ | 9206/10712 [1:36:17<12:18,  2.04it/s]
 86%|████████▌ | 9207/10712 [1:36:18<12:18,  2.04it/s]
 86%|████████▌ | 9208/10712 [1:36:18<12:18,  2.04it/s]
 86%|████████▌ | 9209/10712 [1:36:19<12:16,  2.04it/s]
 86%|████████▌ | 9210/10712 [1:36:19<12:16,  2.04it/s]
 86%|████████▌ | 9211/10712 [1:36:20<12:16,  2.04it/s]
 86%|████████▌ | 9212/10712 [1:36:20<12:16,  2.04it/s]
 86%|████████▌ | 9213/10712 [1:36:21<12:15,  2.04it/s]
 86%|████████▌ | 9214/10712 [1:36:21<12:14,  2.04it/s]
 86%|████████▌ | 9215/10712 [1:36:22<12:14,  2.04it/s]
 86%|████████▌ | 9216/10712 [1:36:22<12:13,  2.04it/s]
 86%|████████▌ | 9217/10712 [1:36:23<12:13,  2.04it/s]
 86%|████████▌ | 9218/10712 [1:36:23<12:12,  2.04it/s]
 86%|████████▌ | 9219/10712 [1:36:24<12:11,  2.04it/s]
 86%|████████▌ | 9220/10712 [1:36:24<12:10,  2.04it/s]
 86%|████████▌ | 9221/10712 [1:36:25<12:11,  2.04it/s]
 86%|████████▌ | 9222/10712 [1:36:25<12:10,  2.04it/s]
 86%|████████▌ | 9223/10712 [1:36:26<12:09,  2.04it/s]
 86%|████████▌ | 9224/10712 [1:36:26<12:09,  2.04it/s]
 86%|████████▌ | 9225/10712 [1:36:27<12:09,  2.04it/s]{'loss': 3.6033, 'grad_norm': 0.19748754799365997, 'learning_rate': 5.7569422479840036e-05, 'epoch': 0.86}
                                                      

 86%|████████▌ | 9225/10712 [1:36:27<12:09,  2.04it/s]
 86%|████████▌ | 9226/10712 [1:36:27<12:09,  2.04it/s]
 86%|████████▌ | 9227/10712 [1:36:28<12:09,  2.04it/s]
 86%|████████▌ | 9228/10712 [1:36:28<12:09,  2.04it/s]
 86%|████████▌ | 9229/10712 [1:36:29<12:08,  2.04it/s]
 86%|████████▌ | 9230/10712 [1:36:29<12:07,  2.04it/s]
 86%|████████▌ | 9231/10712 [1:36:30<12:08,  2.03it/s]
 86%|████████▌ | 9232/10712 [1:36:30<12:07,  2.04it/s]
 86%|████████▌ | 9233/10712 [1:36:30<12:05,  2.04it/s]
 86%|████████▌ | 9234/10712 [1:36:31<12:04,  2.04it/s]
 86%|████████▌ | 9235/10712 [1:36:31<12:04,  2.04it/s]
 86%|████████▌ | 9236/10712 [1:36:32<12:03,  2.04it/s]
 86%|████████▌ | 9237/10712 [1:36:32<12:02,  2.04it/s]
 86%|████████▌ | 9238/10712 [1:36:33<12:03,  2.04it/s]
 86%|████████▌ | 9239/10712 [1:36:33<12:02,  2.04it/s]
 86%|████████▋ | 9240/10712 [1:36:34<12:01,  2.04it/s]
 86%|████████▋ | 9241/10712 [1:36:34<12:00,  2.04it/s]
 86%|████████▋ | 9242/10712 [1:36:35<12:00,  2.04it/s]
 86%|████████▋ | 9243/10712 [1:36:35<12:00,  2.04it/s]
 86%|████████▋ | 9244/10712 [1:36:36<11:59,  2.04it/s]
 86%|████████▋ | 9245/10712 [1:36:36<11:59,  2.04it/s]
 86%|████████▋ | 9246/10712 [1:36:37<11:59,  2.04it/s]
 86%|████████▋ | 9247/10712 [1:36:37<11:58,  2.04it/s]
 86%|████████▋ | 9248/10712 [1:36:38<11:58,  2.04it/s]
 86%|████████▋ | 9249/10712 [1:36:38<11:57,  2.04it/s]
 86%|████████▋ | 9250/10712 [1:36:39<11:57,  2.04it/s]
                                                      
{'loss': 3.6156, 'grad_norm': 0.19858692586421967, 'learning_rate': 5.5686401446575555e-05, 'epoch': 0.86}

 86%|████████▋ | 9250/10712 [1:36:39<11:57,  2.04it/s]
 86%|████████▋ | 9251/10712 [1:36:39<11:56,  2.04it/s]
 86%|████████▋ | 9252/10712 [1:36:40<11:55,  2.04it/s]
 86%|████████▋ | 9253/10712 [1:36:40<11:55,  2.04it/s]
 86%|████████▋ | 9254/10712 [1:36:41<11:54,  2.04it/s]
 86%|████████▋ | 9255/10712 [1:36:41<11:53,  2.04it/s]
 86%|████████▋ | 9256/10712 [1:36:42<11:54,  2.04it/s]
 86%|████████▋ | 9257/10712 [1:36:42<11:53,  2.04it/s]
 86%|████████▋ | 9258/10712 [1:36:43<11:52,  2.04it/s]
 86%|████████▋ | 9259/10712 [1:36:43<11:51,  2.04it/s]
 86%|████████▋ | 9260/10712 [1:36:44<11:51,  2.04it/s]
 86%|████████▋ | 9261/10712 [1:36:44<11:51,  2.04it/s]
 86%|████████▋ | 9262/10712 [1:36:45<11:49,  2.04it/s]
 86%|████████▋ | 9263/10712 [1:36:45<11:49,  2.04it/s]
 86%|████████▋ | 9264/10712 [1:36:46<11:48,  2.04it/s]
 86%|████████▋ | 9265/10712 [1:36:46<11:49,  2.04it/s]
 87%|████████▋ | 9266/10712 [1:36:47<11:48,  2.04it/s]
 87%|████████▋ | 9267/10712 [1:36:47<11:47,  2.04it/s]
 87%|████████▋ | 9268/10712 [1:36:48<11:48,  2.04it/s]
 87%|████████▋ | 9269/10712 [1:36:48<11:47,  2.04it/s]
 87%|████████▋ | 9270/10712 [1:36:49<11:46,  2.04it/s]
 87%|████████▋ | 9271/10712 [1:36:49<11:46,  2.04it/s]
 87%|████████▋ | 9272/10712 [1:36:50<11:46,  2.04it/s]
 87%|████████▋ | 9273/10712 [1:36:50<11:46,  2.04it/s]
 87%|████████▋ | 9274/10712 [1:36:51<11:46,  2.04it/s]
 87%|████████▋ | 9275/10712 [1:36:51<11:45,  2.04it/s]
                                                      
{'loss': 3.6042, 'grad_norm': 0.20154999196529388, 'learning_rate': 5.383287300226419e-05, 'epoch': 0.87}

 87%|████████▋ | 9275/10712 [1:36:51<11:45,  2.04it/s]
 87%|████████▋ | 9276/10712 [1:36:52<11:44,  2.04it/s]
 87%|████████▋ | 9277/10712 [1:36:52<11:43,  2.04it/s]
 87%|████████▋ | 9278/10712 [1:36:53<11:43,  2.04it/s]
 87%|████████▋ | 9279/10712 [1:36:53<11:42,  2.04it/s]
 87%|████████▋ | 9280/10712 [1:36:54<11:41,  2.04it/s]
 87%|████████▋ | 9281/10712 [1:36:54<11:40,  2.04it/s]
 87%|████████▋ | 9282/10712 [1:36:55<11:40,  2.04it/s]
 87%|████████▋ | 9283/10712 [1:36:55<11:40,  2.04it/s]
 87%|████████▋ | 9284/10712 [1:36:55<11:40,  2.04it/s]
 87%|████████▋ | 9285/10712 [1:36:56<11:40,  2.04it/s]
 87%|████████▋ | 9286/10712 [1:36:56<11:40,  2.03it/s]
 87%|████████▋ | 9287/10712 [1:36:57<11:39,  2.04it/s]
 87%|████████▋ | 9288/10712 [1:36:57<11:39,  2.04it/s]
 87%|████████▋ | 9289/10712 [1:36:58<11:39,  2.03it/s]
 87%|████████▋ | 9290/10712 [1:36:58<11:39,  2.03it/s]
 87%|████████▋ | 9291/10712 [1:36:59<11:38,  2.04it/s]
 87%|████████▋ | 9292/10712 [1:36:59<11:36,  2.04it/s]
 87%|████████▋ | 9293/10712 [1:37:00<11:36,  2.04it/s]
 87%|████████▋ | 9294/10712 [1:37:00<11:37,  2.03it/s]
 87%|████████▋ | 9295/10712 [1:37:01<11:36,  2.03it/s]
 87%|████████▋ | 9296/10712 [1:37:01<11:35,  2.04it/s]
 87%|████████▋ | 9297/10712 [1:37:02<11:33,  2.04it/s]
 87%|████████▋ | 9298/10712 [1:37:02<11:34,  2.04it/s]
 87%|████████▋ | 9299/10712 [1:37:03<11:33,  2.04it/s]
 87%|████████▋ | 9300/10712 [1:37:03<11:32,  2.04it/s]
                                                      
{'loss': 3.6147, 'grad_norm': 0.20083454251289368, 'learning_rate': 5.2008960180170264e-05, 'epoch': 0.87}

 87%|████████▋ | 9300/10712 [1:37:03<11:32,  2.04it/s]
 87%|████████▋ | 9301/10712 [1:37:04<11:32,  2.04it/s]
 87%|████████▋ | 9302/10712 [1:37:04<11:32,  2.04it/s]
 87%|████████▋ | 9303/10712 [1:37:05<11:31,  2.04it/s]
 87%|████████▋ | 9304/10712 [1:37:05<11:30,  2.04it/s]
 87%|████████▋ | 9305/10712 [1:37:06<11:30,  2.04it/s]
 87%|████████▋ | 9306/10712 [1:37:06<11:30,  2.04it/s]
 87%|████████▋ | 9307/10712 [1:37:07<11:29,  2.04it/s]
 87%|████████▋ | 9308/10712 [1:37:07<11:29,  2.04it/s]
 87%|████████▋ | 9309/10712 [1:37:08<11:28,  2.04it/s]
 87%|████████▋ | 9310/10712 [1:37:08<11:27,  2.04it/s]
 87%|████████▋ | 9311/10712 [1:37:09<11:27,  2.04it/s]
 87%|████████▋ | 9312/10712 [1:37:09<11:27,  2.04it/s]
 87%|████████▋ | 9313/10712 [1:37:10<11:26,  2.04it/s]
 87%|████████▋ | 9314/10712 [1:37:10<11:25,  2.04it/s]
 87%|████████▋ | 9315/10712 [1:37:11<11:25,  2.04it/s]
 87%|████████▋ | 9316/10712 [1:37:11<11:24,  2.04it/s]
 87%|████████▋ | 9317/10712 [1:37:12<11:23,  2.04it/s]
 87%|████████▋ | 9318/10712 [1:37:12<11:23,  2.04it/s]
 87%|████████▋ | 9319/10712 [1:37:13<11:23,  2.04it/s]
 87%|████████▋ | 9320/10712 [1:37:13<11:22,  2.04it/s]
 87%|████████▋ | 9321/10712 [1:37:14<11:21,  2.04it/s]
 87%|████████▋ | 9322/10712 [1:37:14<11:22,  2.04it/s]
 87%|████████▋ | 9323/10712 [1:37:15<11:21,  2.04it/s]
 87%|████████▋ | 9324/10712 [1:37:15<11:20,  2.04it/s]
 87%|████████▋ | 9325/10712 [1:37:16<11:20,  2.04it/s]{'loss': 3.6032, 'grad_norm': 0.20275002717971802, 'learning_rate': 5.021478404773666e-05, 'epoch': 0.87}
                                                      

 87%|████████▋ | 9325/10712 [1:37:16<11:20,  2.04it/s]
 87%|████████▋ | 9326/10712 [1:37:16<11:20,  2.04it/s]
 87%|████████▋ | 9327/10712 [1:37:17<11:19,  2.04it/s]
 87%|████████▋ | 9328/10712 [1:37:17<11:19,  2.04it/s]
 87%|████████▋ | 9329/10712 [1:37:18<11:18,  2.04it/s]
 87%|████████▋ | 9330/10712 [1:37:18<11:17,  2.04it/s]
 87%|████████▋ | 9331/10712 [1:37:19<11:17,  2.04it/s]
 87%|████████▋ | 9332/10712 [1:37:19<11:17,  2.04it/s]
 87%|████████▋ | 9333/10712 [1:37:20<11:16,  2.04it/s]
 87%|████████▋ | 9334/10712 [1:37:20<11:15,  2.04it/s]
 87%|████████▋ | 9335/10712 [1:37:21<11:15,  2.04it/s]
 87%|████████▋ | 9336/10712 [1:37:21<11:15,  2.04it/s]
 87%|████████▋ | 9337/10712 [1:37:22<11:14,  2.04it/s]
 87%|████████▋ | 9338/10712 [1:37:22<11:13,  2.04it/s]
 87%|████████▋ | 9339/10712 [1:37:22<11:14,  2.04it/s]
 87%|████████▋ | 9340/10712 [1:37:23<11:13,  2.04it/s]
 87%|████████▋ | 9341/10712 [1:37:23<11:12,  2.04it/s]
 87%|████████▋ | 9342/10712 [1:37:24<11:11,  2.04it/s]
 87%|████████▋ | 9343/10712 [1:37:24<11:11,  2.04it/s]
 87%|████████▋ | 9344/10712 [1:37:25<11:10,  2.04it/s]
 87%|████████▋ | 9345/10712 [1:37:25<11:09,  2.04it/s]
 87%|████████▋ | 9346/10712 [1:37:26<11:10,  2.04it/s]
 87%|████████▋ | 9347/10712 [1:37:26<11:09,  2.04it/s]
 87%|████████▋ | 9348/10712 [1:37:27<11:08,  2.04it/s]
 87%|████████▋ | 9349/10712 [1:37:27<11:08,  2.04it/s]
 87%|████████▋ | 9350/10712 [1:37:28<11:08,  2.04it/s]{'loss': 3.6065, 'grad_norm': 0.20585286617279053, 'learning_rate': 4.8450463698547996e-05, 'epoch': 0.87}
                                                      

 87%|████████▋ | 9350/10712 [1:37:28<11:08,  2.04it/s]
 87%|████████▋ | 9351/10712 [1:37:28<11:07,  2.04it/s]
 87%|████████▋ | 9352/10712 [1:37:29<11:08,  2.04it/s]
 87%|████████▋ | 9353/10712 [1:37:29<11:07,  2.04it/s]
 87%|████████▋ | 9354/10712 [1:37:30<11:06,  2.04it/s]
 87%|████████▋ | 9355/10712 [1:37:30<11:05,  2.04it/s]
 87%|████████▋ | 9356/10712 [1:37:31<11:05,  2.04it/s]
 87%|████████▋ | 9357/10712 [1:37:31<11:05,  2.04it/s]
 87%|████████▋ | 9358/10712 [1:37:32<11:03,  2.04it/s]
 87%|████████▋ | 9359/10712 [1:37:32<11:04,  2.04it/s]
 87%|████████▋ | 9360/10712 [1:37:33<11:03,  2.04it/s]
 87%|████████▋ | 9361/10712 [1:37:33<11:03,  2.04it/s]
 87%|████████▋ | 9362/10712 [1:37:34<11:02,  2.04it/s]
 87%|████████▋ | 9363/10712 [1:37:34<11:01,  2.04it/s]
 87%|████████▋ | 9364/10712 [1:37:35<11:01,  2.04it/s]
 87%|████████▋ | 9365/10712 [1:37:35<11:00,  2.04it/s]
 87%|████████▋ | 9366/10712 [1:37:36<10:59,  2.04it/s]
 87%|████████▋ | 9367/10712 [1:37:36<10:59,  2.04it/s]
 87%|████████▋ | 9368/10712 [1:37:37<10:59,  2.04it/s]
 87%|████████▋ | 9369/10712 [1:37:37<10:58,  2.04it/s]
 87%|████████▋ | 9370/10712 [1:37:38<10:58,  2.04it/s]
 87%|████████▋ | 9371/10712 [1:37:38<10:57,  2.04it/s]
 87%|████████▋ | 9372/10712 [1:37:39<10:57,  2.04it/s]
 88%|████████▊ | 9373/10712 [1:37:39<10:56,  2.04it/s]
 88%|████████▊ | 9374/10712 [1:37:40<10:56,  2.04it/s]
 88%|████████▊ | 9375/10712 [1:37:40<10:55,  2.04it/s]{'loss': 3.6104, 'grad_norm': 0.19918066263198853, 'learning_rate': 4.6716116244425786e-05, 'epoch': 0.88}
                                                      

 88%|████████▊ | 9375/10712 [1:37:40<10:55,  2.04it/s]
 88%|████████▊ | 9376/10712 [1:37:41<10:54,  2.04it/s]
 88%|████████▊ | 9377/10712 [1:37:41<10:54,  2.04it/s]
 88%|████████▊ | 9378/10712 [1:37:42<10:54,  2.04it/s]
 88%|████████▊ | 9379/10712 [1:37:42<10:53,  2.04it/s]
 88%|████████▊ | 9380/10712 [1:37:43<10:52,  2.04it/s]
 88%|████████▊ | 9381/10712 [1:37:43<10:52,  2.04it/s]
 88%|████████▊ | 9382/10712 [1:37:44<10:52,  2.04it/s]
 88%|████████▊ | 9383/10712 [1:37:44<10:51,  2.04it/s]
 88%|████████▊ | 9384/10712 [1:37:45<10:50,  2.04it/s]
 88%|████████▊ | 9385/10712 [1:37:45<10:50,  2.04it/s]
 88%|████████▊ | 9386/10712 [1:37:46<10:50,  2.04it/s]
 88%|████████▊ | 9387/10712 [1:37:46<10:49,  2.04it/s]
 88%|████████▊ | 9388/10712 [1:37:47<10:48,  2.04it/s]
 88%|████████▊ | 9389/10712 [1:37:47<10:48,  2.04it/s]
 88%|████████▊ | 9390/10712 [1:37:47<10:48,  2.04it/s]
 88%|████████▊ | 9391/10712 [1:37:48<10:47,  2.04it/s]
 88%|████████▊ | 9392/10712 [1:37:48<10:46,  2.04it/s]
 88%|████████▊ | 9393/10712 [1:37:49<10:46,  2.04it/s]
 88%|████████▊ | 9394/10712 [1:37:49<10:45,  2.04it/s]
 88%|████████▊ | 9395/10712 [1:37:50<10:45,  2.04it/s]
 88%|████████▊ | 9396/10712 [1:37:50<10:44,  2.04it/s]
 88%|████████▊ | 9397/10712 [1:37:51<10:44,  2.04it/s]
 88%|████████▊ | 9398/10712 [1:37:51<10:44,  2.04it/s]
 88%|████████▊ | 9399/10712 [1:37:52<10:43,  2.04it/s]
 88%|████████▊ | 9400/10712 [1:37:52<10:43,  2.04it/s]
                                                      {'loss': 3.6098, 'grad_norm': 0.20281539857387543, 'learning_rate': 4.5011856807654794e-05, 'epoch': 0.88}


 88%|████████▊ | 9400/10712 [1:37:52<10:43,  2.04it/s]
 88%|████████▊ | 9401/10712 [1:37:53<10:43,  2.04it/s]
 88%|████████▊ | 9402/10712 [1:37:53<10:42,  2.04it/s]
 88%|████████▊ | 9403/10712 [1:37:54<10:42,  2.04it/s]
 88%|████████▊ | 9404/10712 [1:37:54<10:41,  2.04it/s]
 88%|████████▊ | 9405/10712 [1:37:55<10:41,  2.04it/s]
 88%|████████▊ | 9406/10712 [1:37:55<10:41,  2.04it/s]
 88%|████████▊ | 9407/10712 [1:37:56<10:40,  2.04it/s]
 88%|████████▊ | 9408/10712 [1:37:56<10:40,  2.04it/s]
 88%|████████▊ | 9409/10712 [1:37:57<10:38,  2.04it/s]
 88%|████████▊ | 9410/10712 [1:37:57<10:38,  2.04it/s]
 88%|████████▊ | 9411/10712 [1:37:58<10:37,  2.04it/s]
 88%|████████▊ | 9412/10712 [1:37:58<10:37,  2.04it/s]
 88%|████████▊ | 9413/10712 [1:37:59<10:37,  2.04it/s]
 88%|████████▊ | 9414/10712 [1:37:59<10:36,  2.04it/s]
 88%|████████▊ | 9415/10712 [1:38:00<10:35,  2.04it/s]
 88%|████████▊ | 9416/10712 [1:38:00<10:34,  2.04it/s]
 88%|████████▊ | 9417/10712 [1:38:01<10:34,  2.04it/s]
 88%|████████▊ | 9418/10712 [1:38:01<10:34,  2.04it/s]
 88%|████████▊ | 9419/10712 [1:38:02<10:33,  2.04it/s]
 88%|████████▊ | 9420/10712 [1:38:02<10:32,  2.04it/s]
 88%|████████▊ | 9421/10712 [1:38:03<10:33,  2.04it/s]
 88%|████████▊ | 9422/10712 [1:38:03<10:33,  2.04it/s]
 88%|████████▊ | 9423/10712 [1:38:04<10:32,  2.04it/s]
 88%|████████▊ | 9424/10712 [1:38:04<10:31,  2.04it/s]
 88%|████████▊ | 9425/10712 [1:38:05<10:31,  2.04it/s]{'loss': 3.602, 'grad_norm': 0.20339369773864746, 'learning_rate': 4.333779851334108e-05, 'epoch': 0.88}

                                                      

 88%|████████▊ | 9425/10712 [1:38:05<10:31,  2.04it/s]
 88%|████████▊ | 9426/10712 [1:38:05<10:31,  2.04it/s]
 88%|████████▊ | 9427/10712 [1:38:06<10:30,  2.04it/s]
 88%|████████▊ | 9428/10712 [1:38:06<10:29,  2.04it/s]
 88%|████████▊ | 9429/10712 [1:38:07<10:29,  2.04it/s]
 88%|████████▊ | 9430/10712 [1:38:07<10:29,  2.04it/s]
 88%|████████▊ | 9431/10712 [1:38:08<10:28,  2.04it/s]
 88%|████████▊ | 9432/10712 [1:38:08<10:27,  2.04it/s]
 88%|████████▊ | 9433/10712 [1:38:09<10:26,  2.04it/s]
 88%|████████▊ | 9434/10712 [1:38:09<10:27,  2.04it/s]
 88%|████████▊ | 9435/10712 [1:38:10<10:26,  2.04it/s]
 88%|████████▊ | 9436/10712 [1:38:10<10:26,  2.04it/s]
 88%|████████▊ | 9437/10712 [1:38:11<10:24,  2.04it/s]
 88%|████████▊ | 9438/10712 [1:38:11<10:25,  2.04it/s]
 88%|████████▊ | 9439/10712 [1:38:12<10:24,  2.04it/s]
 88%|████████▊ | 9440/10712 [1:38:12<10:23,  2.04it/s]
 88%|████████▊ | 9441/10712 [1:38:13<10:22,  2.04it/s]
 88%|████████▊ | 9442/10712 [1:38:13<10:23,  2.04it/s]
 88%|████████▊ | 9443/10712 [1:38:13<10:23,  2.04it/s]
 88%|████████▊ | 9444/10712 [1:38:14<10:21,  2.04it/s]
 88%|████████▊ | 9445/10712 [1:38:14<10:21,  2.04it/s]
 88%|████████▊ | 9446/10712 [1:38:15<10:21,  2.04it/s]
 88%|████████▊ | 9447/10712 [1:38:15<10:20,  2.04it/s]
 88%|████████▊ | 9448/10712 [1:38:16<10:19,  2.04it/s]
 88%|████████▊ | 9449/10712 [1:38:16<10:19,  2.04it/s]
 88%|████████▊ | 9450/10712 [1:38:17<10:18,  2.04it/s]
                                                      
{'loss': 3.6072, 'grad_norm': 0.20608074963092804, 'learning_rate': 4.169405248190361e-05, 'epoch': 0.88}

 88%|████████▊ | 9450/10712 [1:38:17<10:18,  2.04it/s]
 88%|████████▊ | 9451/10712 [1:38:17<10:19,  2.04it/s]
 88%|████████▊ | 9452/10712 [1:38:18<10:18,  2.04it/s]
 88%|████████▊ | 9453/10712 [1:38:18<10:17,  2.04it/s]
 88%|████████▊ | 9454/10712 [1:38:19<10:17,  2.04it/s]
 88%|████████▊ | 9455/10712 [1:38:19<10:17,  2.04it/s]
 88%|████████▊ | 9456/10712 [1:38:20<10:16,  2.04it/s]
 88%|████████▊ | 9457/10712 [1:38:20<10:16,  2.04it/s]
 88%|████████▊ | 9458/10712 [1:38:21<10:16,  2.03it/s]
 88%|████████▊ | 9459/10712 [1:38:21<10:15,  2.04it/s]
 88%|████████▊ | 9460/10712 [1:38:22<10:14,  2.04it/s]
 88%|████████▊ | 9461/10712 [1:38:22<10:14,  2.04it/s]
 88%|████████▊ | 9462/10712 [1:38:23<10:13,  2.04it/s]
 88%|████████▊ | 9463/10712 [1:38:23<10:12,  2.04it/s]
 88%|████████▊ | 9464/10712 [1:38:24<10:12,  2.04it/s]
 88%|████████▊ | 9465/10712 [1:38:24<10:12,  2.04it/s]
 88%|████████▊ | 9466/10712 [1:38:25<10:11,  2.04it/s]
 88%|████████▊ | 9467/10712 [1:38:25<10:11,  2.04it/s]
 88%|████████▊ | 9468/10712 [1:38:26<10:10,  2.04it/s]
 88%|████████▊ | 9469/10712 [1:38:26<10:09,  2.04it/s]
 88%|████████▊ | 9470/10712 [1:38:27<10:09,  2.04it/s]
 88%|████████▊ | 9471/10712 [1:38:27<10:08,  2.04it/s]
 88%|████████▊ | 9472/10712 [1:38:28<10:07,  2.04it/s]
 88%|████████▊ | 9473/10712 [1:38:28<10:07,  2.04it/s]
 88%|████████▊ | 9474/10712 [1:38:29<10:07,  2.04it/s]
 88%|████████▊ | 9475/10712 [1:38:29<10:06,  2.04it/s]{'loss': 3.6, 'grad_norm': 0.19659234583377838, 'learning_rate': 4.0080727821697695e-05, 'epoch': 0.88}
                                                      

 88%|████████▊ | 9475/10712 [1:38:29<10:06,  2.04it/s]
 88%|████████▊ | 9476/10712 [1:38:30<10:07,  2.03it/s]
 88%|████████▊ | 9477/10712 [1:38:30<10:07,  2.03it/s]
 88%|████████▊ | 9478/10712 [1:38:31<10:06,  2.03it/s]
 88%|████████▊ | 9479/10712 [1:38:31<10:05,  2.04it/s]
 88%|████████▊ | 9480/10712 [1:38:32<10:05,  2.04it/s]
 89%|████████▊ | 9481/10712 [1:38:32<10:04,  2.04it/s]
 89%|████████▊ | 9482/10712 [1:38:33<10:02,  2.04it/s]
 89%|████████▊ | 9483/10712 [1:38:33<10:02,  2.04it/s]
 89%|████████▊ | 9484/10712 [1:38:34<10:03,  2.04it/s]
 89%|████████▊ | 9485/10712 [1:38:34<10:01,  2.04it/s]
 89%|████████▊ | 9486/10712 [1:38:35<10:01,  2.04it/s]
 89%|████████▊ | 9487/10712 [1:38:35<10:00,  2.04it/s]
 89%|████████▊ | 9488/10712 [1:38:36<10:00,  2.04it/s]
 89%|████████▊ | 9489/10712 [1:38:36<10:00,  2.04it/s]
 89%|████████▊ | 9490/10712 [1:38:37<10:00,  2.04it/s]
 89%|████████▊ | 9491/10712 [1:38:37<09:59,  2.04it/s]
 89%|████████▊ | 9492/10712 [1:38:38<09:58,  2.04it/s]
 89%|████████▊ | 9493/10712 [1:38:38<09:57,  2.04it/s]
 89%|████████▊ | 9494/10712 [1:38:39<09:57,  2.04it/s]
 89%|████████▊ | 9495/10712 [1:38:39<09:56,  2.04it/s]
 89%|████████▊ | 9496/10712 [1:38:39<09:55,  2.04it/s]
 89%|████████▊ | 9497/10712 [1:38:40<09:55,  2.04it/s]
 89%|████████▊ | 9498/10712 [1:38:40<09:55,  2.04it/s]
 89%|████████▊ | 9499/10712 [1:38:41<09:54,  2.04it/s]
 89%|████████▊ | 9500/10712 [1:38:41<09:53,  2.04it/s]
                                                      {'loss': 3.6012, 'grad_norm': 0.20072096586227417, 'learning_rate': 3.849793162177323e-05, 'epoch': 0.89}

 89%|████████▊ | 9500/10712 [1:38:41<09:53,  2.04it/s]
 89%|████████▊ | 9501/10712 [1:38:42<09:54,  2.04it/s]
 89%|████████▊ | 9502/10712 [1:38:42<09:53,  2.04it/s]
 89%|████████▊ | 9503/10712 [1:38:43<09:52,  2.04it/s]
 89%|████████▊ | 9504/10712 [1:38:43<09:53,  2.04it/s]
 89%|████████▊ | 9505/10712 [1:38:44<09:52,  2.04it/s]
 89%|████████▊ | 9506/10712 [1:38:44<09:51,  2.04it/s]
 89%|████████▉ | 9507/10712 [1:38:45<09:52,  2.04it/s]
 89%|████████▉ | 9508/10712 [1:38:45<09:51,  2.03it/s]
 89%|████████▉ | 9509/10712 [1:38:46<09:51,  2.03it/s]
 89%|████████▉ | 9510/10712 [1:38:46<09:49,  2.04it/s]
 89%|████████▉ | 9511/10712 [1:38:47<09:49,  2.04it/s]
 89%|████████▉ | 9512/10712 [1:38:47<09:49,  2.04it/s]
 89%|████████▉ | 9513/10712 [1:38:48<09:48,  2.04it/s]
 89%|████████▉ | 9514/10712 [1:38:48<09:47,  2.04it/s]
 89%|████████▉ | 9515/10712 [1:38:49<09:47,  2.04it/s]
 89%|████████▉ | 9516/10712 [1:38:49<09:47,  2.04it/s]
 89%|████████▉ | 9517/10712 [1:38:50<09:46,  2.04it/s]
 89%|████████▉ | 9518/10712 [1:38:50<09:46,  2.04it/s]
 89%|████████▉ | 9519/10712 [1:38:51<09:45,  2.04it/s]
 89%|████████▉ | 9520/10712 [1:38:51<09:44,  2.04it/s]
 89%|████████▉ | 9521/10712 [1:38:52<09:43,  2.04it/s]
 89%|████████▉ | 9522/10712 [1:38:52<09:43,  2.04it/s]
 89%|████████▉ | 9523/10712 [1:38:53<09:42,  2.04it/s]
 89%|████████▉ | 9524/10712 [1:38:53<09:42,  2.04it/s]
 89%|████████▉ | 9525/10712 [1:38:54<09:42,  2.04it/s]{'loss': 3.6083, 'grad_norm': 0.19693852961063385, 'learning_rate': 3.69457689447657e-05, 'epoch': 0.89}
                                                      

 89%|████████▉ | 9525/10712 [1:38:54<09:42,  2.04it/s]
 89%|████████▉ | 9526/10712 [1:38:54<11:16,  1.75it/s]
 89%|████████▉ | 9527/10712 [1:38:55<10:47,  1.83it/s]
 89%|████████▉ | 9528/10712 [1:38:55<10:26,  1.89it/s]
 89%|████████▉ | 9529/10712 [1:38:56<10:13,  1.93it/s]
 89%|████████▉ | 9530/10712 [1:38:56<10:02,  1.96it/s]
 89%|████████▉ | 9531/10712 [1:38:57<09:55,  1.98it/s]
 89%|████████▉ | 9532/10712 [1:38:57<09:50,  2.00it/s]
 89%|████████▉ | 9533/10712 [1:38:58<09:45,  2.01it/s]
 89%|████████▉ | 9534/10712 [1:38:58<09:42,  2.02it/s]
 89%|████████▉ | 9535/10712 [1:38:59<09:40,  2.03it/s]
 89%|████████▉ | 9536/10712 [1:38:59<09:39,  2.03it/s]
 89%|████████▉ | 9537/10712 [1:39:00<09:37,  2.03it/s]
 89%|████████▉ | 9538/10712 [1:39:00<09:36,  2.04it/s]
 89%|████████▉ | 9539/10712 [1:39:01<09:35,  2.04it/s]
 89%|████████▉ | 9540/10712 [1:39:01<09:35,  2.04it/s]
 89%|████████▉ | 9541/10712 [1:39:02<09:33,  2.04it/s]
 89%|████████▉ | 9542/10712 [1:39:02<09:33,  2.04it/s]
 89%|████████▉ | 9543/10712 [1:39:03<09:33,  2.04it/s]
 89%|████████▉ | 9544/10712 [1:39:03<09:32,  2.04it/s]
 89%|████████▉ | 9545/10712 [1:39:04<09:32,  2.04it/s]
 89%|████████▉ | 9546/10712 [1:39:04<09:32,  2.04it/s]
 89%|████████▉ | 9547/10712 [1:39:05<09:31,  2.04it/s]
 89%|████████▉ | 9548/10712 [1:39:05<09:31,  2.04it/s]
 89%|████████▉ | 9549/10712 [1:39:06<09:30,  2.04it/s]
 89%|████████▉ | 9550/10712 [1:39:06<09:30,  2.04it/s]
                                                      {'loss': 3.6112, 'grad_norm': 0.1944766491651535, 'learning_rate': 3.5424342819923015e-05, 'epoch': 0.89}


 89%|████████▉ | 9550/10712 [1:39:06<09:30,  2.04it/s]
 89%|████████▉ | 9551/10712 [1:39:07<09:29,  2.04it/s]
 89%|████████▉ | 9552/10712 [1:39:07<09:29,  2.04it/s]
 89%|████████▉ | 9553/10712 [1:39:08<09:29,  2.04it/s]
 89%|████████▉ | 9554/10712 [1:39:08<09:28,  2.04it/s]
 89%|████████▉ | 9555/10712 [1:39:09<09:27,  2.04it/s]
 89%|████████▉ | 9556/10712 [1:39:09<09:27,  2.04it/s]
 89%|████████▉ | 9557/10712 [1:39:10<09:26,  2.04it/s]
 89%|████████▉ | 9558/10712 [1:39:10<09:25,  2.04it/s]
 89%|████████▉ | 9559/10712 [1:39:11<09:24,  2.04it/s]
 89%|████████▉ | 9560/10712 [1:39:11<09:23,  2.04it/s]
 89%|████████▉ | 9561/10712 [1:39:12<09:24,  2.04it/s]
 89%|████████▉ | 9562/10712 [1:39:12<09:24,  2.04it/s]
 89%|████████▉ | 9563/10712 [1:39:13<09:23,  2.04it/s]
 89%|████████▉ | 9564/10712 [1:39:13<09:22,  2.04it/s]
 89%|████████▉ | 9565/10712 [1:39:14<09:22,  2.04it/s]
 89%|████████▉ | 9566/10712 [1:39:14<09:22,  2.04it/s]
 89%|████████▉ | 9567/10712 [1:39:15<09:21,  2.04it/s]
 89%|████████▉ | 9568/10712 [1:39:15<09:20,  2.04it/s]
 89%|████████▉ | 9569/10712 [1:39:16<09:20,  2.04it/s]
 89%|████████▉ | 9570/10712 [1:39:16<09:20,  2.04it/s]
 89%|████████▉ | 9571/10712 [1:39:17<09:19,  2.04it/s]
 89%|████████▉ | 9572/10712 [1:39:17<09:18,  2.04it/s]
 89%|████████▉ | 9573/10712 [1:39:18<09:18,  2.04it/s]
 89%|████████▉ | 9574/10712 [1:39:18<10:47,  1.76it/s]
 89%|████████▉ | 9575/10712 [1:39:19<10:20,  1.83it/s]{'loss': 3.6017, 'grad_norm': 0.20685018599033356, 'learning_rate': 3.393375423626599e-05, 'epoch': 0.89}

                                                      

 89%|████████▉ | 9575/10712 [1:39:19<10:20,  1.83it/s]
 89%|████████▉ | 9576/10712 [1:39:19<10:01,  1.89it/s]
 89%|████████▉ | 9577/10712 [1:39:20<09:48,  1.93it/s]
 89%|████████▉ | 9578/10712 [1:39:20<09:37,  1.96it/s]
 89%|████████▉ | 9579/10712 [1:39:21<09:30,  1.99it/s]
 89%|████████▉ | 9580/10712 [1:39:21<09:24,  2.00it/s]
 89%|████████▉ | 9581/10712 [1:39:22<09:22,  2.01it/s]
 89%|████████▉ | 9582/10712 [1:39:22<09:19,  2.02it/s]
 89%|████████▉ | 9583/10712 [1:39:23<09:16,  2.03it/s]
 89%|████████▉ | 9584/10712 [1:39:23<09:14,  2.03it/s]
 89%|████████▉ | 9585/10712 [1:39:24<09:14,  2.03it/s]
 89%|████████▉ | 9586/10712 [1:39:24<09:12,  2.04it/s]
 89%|████████▉ | 9587/10712 [1:39:25<09:11,  2.04it/s]
 90%|████████▉ | 9588/10712 [1:39:25<09:11,  2.04it/s]
 90%|████████▉ | 9589/10712 [1:39:26<09:10,  2.04it/s]
 90%|████████▉ | 9590/10712 [1:39:26<09:09,  2.04it/s]
 90%|████████▉ | 9591/10712 [1:39:27<09:08,  2.04it/s]
 90%|████████▉ | 9592/10712 [1:39:27<09:08,  2.04it/s]
 90%|████████▉ | 9593/10712 [1:39:28<09:07,  2.04it/s]
 90%|████████▉ | 9594/10712 [1:39:28<09:07,  2.04it/s]
 90%|████████▉ | 9595/10712 [1:39:29<09:06,  2.04it/s]
 90%|████████▉ | 9596/10712 [1:39:29<09:06,  2.04it/s]
 90%|████████▉ | 9597/10712 [1:39:30<09:06,  2.04it/s]
 90%|████████▉ | 9598/10712 [1:39:30<09:05,  2.04it/s]
 90%|████████▉ | 9599/10712 [1:39:31<09:04,  2.04it/s]
 90%|████████▉ | 9600/10712 [1:39:31<09:04,  2.04it/s]{'loss': 3.602, 'grad_norm': 0.2023308128118515, 'learning_rate': 3.247410213588553e-05, 'epoch': 0.9}

                                                      

 90%|████████▉ | 9600/10712 [1:39:31<09:04,  2.04it/s]
 90%|████████▉ | 9601/10712 [1:39:32<09:04,  2.04it/s]
 90%|████████▉ | 9602/10712 [1:39:32<09:04,  2.04it/s]
 90%|████████▉ | 9603/10712 [1:39:32<09:03,  2.04it/s]
 90%|████████▉ | 9604/10712 [1:39:33<09:03,  2.04it/s]
 90%|████████▉ | 9605/10712 [1:39:33<09:02,  2.04it/s]
 90%|████████▉ | 9606/10712 [1:39:34<09:01,  2.04it/s]
 90%|████████▉ | 9607/10712 [1:39:34<09:00,  2.04it/s]
 90%|████████▉ | 9608/10712 [1:39:35<09:00,  2.04it/s]
 90%|████████▉ | 9609/10712 [1:39:35<09:00,  2.04it/s]
 90%|████████▉ | 9610/10712 [1:39:36<09:00,  2.04it/s]
 90%|████████▉ | 9611/10712 [1:39:36<08:59,  2.04it/s]
 90%|████████▉ | 9612/10712 [1:39:37<08:59,  2.04it/s]
 90%|████████▉ | 9613/10712 [1:39:37<08:58,  2.04it/s]
 90%|████████▉ | 9614/10712 [1:39:38<08:57,  2.04it/s]
 90%|████████▉ | 9615/10712 [1:39:38<08:57,  2.04it/s]
 90%|████████▉ | 9616/10712 [1:39:39<08:56,  2.04it/s]
 90%|████████▉ | 9617/10712 [1:39:39<08:55,  2.04it/s]
 90%|████████▉ | 9618/10712 [1:39:40<08:55,  2.04it/s]
 90%|████████▉ | 9619/10712 [1:39:40<08:55,  2.04it/s]
 90%|████████▉ | 9620/10712 [1:39:41<08:55,  2.04it/s]
 90%|████████▉ | 9621/10712 [1:39:41<08:54,  2.04it/s]
 90%|████████▉ | 9622/10712 [1:39:42<08:54,  2.04it/s]
 90%|████████▉ | 9623/10712 [1:39:42<08:53,  2.04it/s]
 90%|████████▉ | 9624/10712 [1:39:43<08:53,  2.04it/s]
 90%|████████▉ | 9625/10712 [1:39:43<08:52,  2.04it/s]{'loss': 3.6014, 'grad_norm': 0.20043201744556427, 'learning_rate': 3.1045483407374505e-05, 'epoch': 0.9}

                                                      

 90%|████████▉ | 9625/10712 [1:39:43<08:52,  2.04it/s]
 90%|████████▉ | 9626/10712 [1:39:44<08:52,  2.04it/s]
 90%|████████▉ | 9627/10712 [1:39:44<08:52,  2.04it/s]
 90%|████████▉ | 9628/10712 [1:39:45<08:51,  2.04it/s]
 90%|████████▉ | 9629/10712 [1:39:45<08:50,  2.04it/s]
 90%|████████▉ | 9630/10712 [1:39:46<08:50,  2.04it/s]
 90%|████████▉ | 9631/10712 [1:39:46<08:49,  2.04it/s]
 90%|████████▉ | 9632/10712 [1:39:47<08:49,  2.04it/s]
 90%|████████▉ | 9633/10712 [1:39:47<08:48,  2.04it/s]
 90%|████████▉ | 9634/10712 [1:39:48<08:48,  2.04it/s]
 90%|████████▉ | 9635/10712 [1:39:48<08:47,  2.04it/s]
 90%|████████▉ | 9636/10712 [1:39:49<08:46,  2.04it/s]
 90%|████████▉ | 9637/10712 [1:39:49<08:45,  2.04it/s]
 90%|████████▉ | 9638/10712 [1:39:50<08:46,  2.04it/s]
 90%|████████▉ | 9639/10712 [1:39:50<08:45,  2.04it/s]
 90%|████████▉ | 9640/10712 [1:39:51<08:44,  2.04it/s]
 90%|█████████ | 9641/10712 [1:39:51<08:43,  2.04it/s]
 90%|█████████ | 9642/10712 [1:39:52<08:43,  2.04it/s]
 90%|█████████ | 9643/10712 [1:39:52<08:44,  2.04it/s]
 90%|█████████ | 9644/10712 [1:39:53<08:42,  2.04it/s]
 90%|█████████ | 9645/10712 [1:39:53<08:42,  2.04it/s]
 90%|█████████ | 9646/10712 [1:39:54<08:42,  2.04it/s]
 90%|█████████ | 9647/10712 [1:39:54<08:41,  2.04it/s]
 90%|█████████ | 9648/10712 [1:39:55<08:41,  2.04it/s]
 90%|█████████ | 9649/10712 [1:39:55<08:40,  2.04it/s]
 90%|█████████ | 9650/10712 [1:39:56<08:40,  2.04it/s]{'loss': 3.6064, 'grad_norm': 0.20316214859485626, 'learning_rate': 2.9647992879397012e-05, 'epoch': 0.9}

                                                      

 90%|█████████ | 9650/10712 [1:39:56<08:40,  2.04it/s]
 90%|█████████ | 9651/10712 [1:39:56<08:40,  2.04it/s]
 90%|█████████ | 9652/10712 [1:39:56<08:39,  2.04it/s]
 90%|█████████ | 9653/10712 [1:39:57<08:39,  2.04it/s]
 90%|█████████ | 9654/10712 [1:39:57<08:39,  2.04it/s]
 90%|█████████ | 9655/10712 [1:39:58<08:38,  2.04it/s]
 90%|█████████ | 9656/10712 [1:39:58<08:37,  2.04it/s]
 90%|█████████ | 9657/10712 [1:39:59<08:36,  2.04it/s]
 90%|█████████ | 9658/10712 [1:39:59<08:36,  2.04it/s]
 90%|█████████ | 9659/10712 [1:40:00<08:35,  2.04it/s]
 90%|█████████ | 9660/10712 [1:40:00<08:35,  2.04it/s]
 90%|█████████ | 9661/10712 [1:40:01<08:34,  2.04it/s]
 90%|█████████ | 9662/10712 [1:40:01<08:34,  2.04it/s]
 90%|█████████ | 9663/10712 [1:40:02<08:34,  2.04it/s]
 90%|█████████ | 9664/10712 [1:40:02<08:33,  2.04it/s]
 90%|█████████ | 9665/10712 [1:40:03<08:33,  2.04it/s]
 90%|█████████ | 9666/10712 [1:40:03<08:32,  2.04it/s]
 90%|█████████ | 9667/10712 [1:40:04<08:32,  2.04it/s]
 90%|█████████ | 9668/10712 [1:40:04<08:32,  2.04it/s]
 90%|█████████ | 9669/10712 [1:40:05<08:31,  2.04it/s]
 90%|█████████ | 9670/10712 [1:40:05<08:31,  2.04it/s]
 90%|█████████ | 9671/10712 [1:40:06<08:30,  2.04it/s]
 90%|█████████ | 9672/10712 [1:40:06<08:29,  2.04it/s]
 90%|█████████ | 9673/10712 [1:40:07<08:29,  2.04it/s]
 90%|█████████ | 9674/10712 [1:40:07<08:28,  2.04it/s]
 90%|█████████ | 9675/10712 [1:40:08<08:27,  2.04it/s]
                                                      
{'loss': 3.6046, 'grad_norm': 0.199016734957695, 'learning_rate': 2.828172331439327e-05, 'epoch': 0.9}

 90%|█████████ | 9675/10712 [1:40:08<08:27,  2.04it/s]
 90%|█████████ | 9676/10712 [1:40:08<08:28,  2.04it/s]
 90%|█████████ | 9677/10712 [1:40:09<08:28,  2.04it/s]
 90%|█████████ | 9678/10712 [1:40:09<08:27,  2.04it/s]
 90%|█████████ | 9679/10712 [1:40:10<08:26,  2.04it/s]
 90%|█████████ | 9680/10712 [1:40:10<08:26,  2.04it/s]
 90%|█████████ | 9681/10712 [1:40:11<08:25,  2.04it/s]
 90%|█████████ | 9682/10712 [1:40:11<08:24,  2.04it/s]
 90%|█████████ | 9683/10712 [1:40:12<08:24,  2.04it/s]
 90%|█████████ | 9684/10712 [1:40:12<08:24,  2.04it/s]
 90%|█████████ | 9685/10712 [1:40:13<08:23,  2.04it/s]
 90%|█████████ | 9686/10712 [1:40:13<08:22,  2.04it/s]
 90%|█████████ | 9687/10712 [1:40:14<08:22,  2.04it/s]
 90%|█████████ | 9688/10712 [1:40:14<08:22,  2.04it/s]
 90%|█████████ | 9689/10712 [1:40:15<08:21,  2.04it/s]
 90%|█████████ | 9690/10712 [1:40:15<08:20,  2.04it/s]
 90%|█████████ | 9691/10712 [1:40:16<08:20,  2.04it/s]
 90%|█████████ | 9692/10712 [1:40:16<08:20,  2.04it/s]
 90%|█████████ | 9693/10712 [1:40:17<08:20,  2.04it/s]
 90%|█████████ | 9694/10712 [1:40:17<08:19,  2.04it/s]
 91%|█████████ | 9695/10712 [1:40:18<08:19,  2.04it/s]
 91%|█████████ | 9696/10712 [1:40:18<08:18,  2.04it/s]
 91%|█████████ | 9697/10712 [1:40:19<08:17,  2.04it/s]
 91%|█████████ | 9698/10712 [1:40:19<08:16,  2.04it/s]
 91%|█████████ | 9699/10712 [1:40:20<08:15,  2.04it/s]
 91%|█████████ | 9700/10712 [1:40:20<08:15,  2.04it/s]{'loss': 3.605, 'grad_norm': 0.19517090916633606, 'learning_rate': 2.6946765402422858e-05, 'epoch': 0.91}
                                                      

 91%|█████████ | 9700/10712 [1:40:20<08:15,  2.04it/s]
 91%|█████████ | 9701/10712 [1:40:21<08:15,  2.04it/s]
 91%|█████████ | 9702/10712 [1:40:21<08:14,  2.04it/s]
 91%|█████████ | 9703/10712 [1:40:21<08:14,  2.04it/s]
 91%|█████████ | 9704/10712 [1:40:22<08:13,  2.04it/s]
 91%|█████████ | 9705/10712 [1:40:22<08:13,  2.04it/s]
 91%|█████████ | 9706/10712 [1:40:23<08:13,  2.04it/s]
 91%|█████████ | 9707/10712 [1:40:23<08:12,  2.04it/s]
 91%|█████████ | 9708/10712 [1:40:24<08:12,  2.04it/s]
 91%|█████████ | 9709/10712 [1:40:24<08:11,  2.04it/s]
 91%|█████████ | 9710/10712 [1:40:25<08:11,  2.04it/s]
 91%|█████████ | 9711/10712 [1:40:25<08:10,  2.04it/s]
 91%|█████████ | 9712/10712 [1:40:26<08:09,  2.04it/s]
 91%|█████████ | 9713/10712 [1:40:26<08:09,  2.04it/s]
 91%|█████████ | 9714/10712 [1:40:27<08:09,  2.04it/s]
 91%|█████████ | 9715/10712 [1:40:27<08:08,  2.04it/s]
 91%|█████████ | 9716/10712 [1:40:28<08:07,  2.04it/s]
 91%|█████████ | 9717/10712 [1:40:28<08:08,  2.04it/s]
 91%|█████████ | 9718/10712 [1:40:29<08:07,  2.04it/s]
 91%|█████████ | 9719/10712 [1:40:29<08:06,  2.04it/s]
 91%|█████████ | 9720/10712 [1:40:30<08:05,  2.04it/s]
 91%|█████████ | 9721/10712 [1:40:30<08:05,  2.04it/s]
 91%|█████████ | 9722/10712 [1:40:31<08:04,  2.04it/s]
 91%|█████████ | 9723/10712 [1:40:31<08:04,  2.04it/s]
 91%|█████████ | 9724/10712 [1:40:32<08:04,  2.04it/s]
 91%|█████████ | 9725/10712 [1:40:32<08:03,  2.04it/s]{'loss': 3.611, 'grad_norm': 0.20043127238750458, 'learning_rate': 2.5643207755144615e-05, 'epoch': 0.91}

                                                      

 91%|█████████ | 9725/10712 [1:40:32<08:03,  2.04it/s]
 91%|█████████ | 9726/10712 [1:40:33<08:04,  2.04it/s]
 91%|█████████ | 9727/10712 [1:40:33<08:03,  2.04it/s]
 91%|█████████ | 9728/10712 [1:40:34<08:02,  2.04it/s]
 91%|█████████ | 9729/10712 [1:40:34<08:01,  2.04it/s]
 91%|█████████ | 9730/10712 [1:40:35<08:01,  2.04it/s]
 91%|█████████ | 9731/10712 [1:40:35<08:01,  2.04it/s]
 91%|█████████ | 9732/10712 [1:40:36<08:00,  2.04it/s]
 91%|█████████ | 9733/10712 [1:40:36<07:59,  2.04it/s]
 91%|█████████ | 9734/10712 [1:40:37<07:59,  2.04it/s]
 91%|█████████ | 9735/10712 [1:40:37<07:58,  2.04it/s]
 91%|█████████ | 9736/10712 [1:40:38<07:58,  2.04it/s]
 91%|█████████ | 9737/10712 [1:40:38<07:56,  2.04it/s]
 91%|█████████ | 9738/10712 [1:40:39<07:56,  2.04it/s]
 91%|█████████ | 9739/10712 [1:40:39<07:56,  2.04it/s]
 91%|█████████ | 9740/10712 [1:40:40<07:55,  2.04it/s]
 91%|█████████ | 9741/10712 [1:40:40<07:55,  2.04it/s]
 91%|█████████ | 9742/10712 [1:40:41<07:54,  2.04it/s]
 91%|█████████ | 9743/10712 [1:40:41<07:54,  2.04it/s]
 91%|█████████ | 9744/10712 [1:40:42<07:53,  2.04it/s]
 91%|█████████ | 9745/10712 [1:40:42<07:52,  2.05it/s]
 91%|█████████ | 9746/10712 [1:40:43<07:52,  2.04it/s]
 91%|█████████ | 9747/10712 [1:40:43<07:52,  2.04it/s]
 91%|█████████ | 9748/10712 [1:40:44<07:52,  2.04it/s]
 91%|█████████ | 9749/10712 [1:40:44<07:51,  2.04it/s]
 91%|█████████ | 9750/10712 [1:40:45<07:50,  2.04it/s]
                                                      
{'loss': 3.6013, 'grad_norm': 0.19403138756752014, 'learning_rate': 2.4371136899934532e-05, 'epoch': 0.91}

 91%|█████████ | 9750/10712 [1:40:45<07:50,  2.04it/s]
 91%|█████████ | 9751/10712 [1:40:45<07:51,  2.04it/s]
 91%|█████████ | 9752/10712 [1:40:46<07:50,  2.04it/s]
 91%|█████████ | 9753/10712 [1:40:46<07:49,  2.04it/s]
 91%|█████████ | 9754/10712 [1:40:46<07:49,  2.04it/s]
 91%|█████████ | 9755/10712 [1:40:47<07:48,  2.04it/s]
 91%|█████████ | 9756/10712 [1:40:47<07:48,  2.04it/s]
 91%|█████████ | 9757/10712 [1:40:48<07:47,  2.04it/s]
 91%|█████████ | 9758/10712 [1:40:48<07:47,  2.04it/s]
 91%|█████████ | 9759/10712 [1:40:49<07:46,  2.04it/s]
 91%|█████████ | 9760/10712 [1:40:49<07:45,  2.04it/s]
 91%|█████████ | 9761/10712 [1:40:50<07:45,  2.04it/s]
 91%|█████████ | 9762/10712 [1:40:50<07:45,  2.04it/s]
 91%|█████████ | 9763/10712 [1:40:51<07:44,  2.04it/s]
 91%|█████████ | 9764/10712 [1:40:51<07:44,  2.04it/s]
 91%|█████████ | 9765/10712 [1:40:52<07:43,  2.04it/s]
 91%|█████████ | 9766/10712 [1:40:52<07:43,  2.04it/s]
 91%|█████████ | 9767/10712 [1:40:53<07:43,  2.04it/s]
 91%|█████████ | 9768/10712 [1:40:53<07:42,  2.04it/s]
 91%|█████████ | 9769/10712 [1:40:54<07:42,  2.04it/s]
 91%|█████████ | 9770/10712 [1:40:54<07:41,  2.04it/s]
 91%|█████████ | 9771/10712 [1:40:55<07:41,  2.04it/s]
 91%|█████████ | 9772/10712 [1:40:55<07:40,  2.04it/s]
 91%|█████████ | 9773/10712 [1:40:56<07:40,  2.04it/s]
 91%|█████████ | 9774/10712 [1:40:56<07:39,  2.04it/s]
 91%|█████████▏| 9775/10712 [1:40:57<07:39,  2.04it/s]{'loss': 3.6019, 'grad_norm': 0.19612985849380493, 'learning_rate': 2.3130637274142775e-05, 'epoch': 0.91}
                                                      

 91%|█████████▏| 9775/10712 [1:40:57<07:39,  2.04it/s]
 91%|█████████▏| 9776/10712 [1:40:57<07:39,  2.04it/s]
 91%|█████████▏| 9777/10712 [1:40:58<07:38,  2.04it/s]
 91%|█████████▏| 9778/10712 [1:40:58<07:37,  2.04it/s]
 91%|█████████▏| 9779/10712 [1:40:59<07:37,  2.04it/s]
 91%|█████████▏| 9780/10712 [1:40:59<07:37,  2.04it/s]
 91%|█████████▏| 9781/10712 [1:41:00<07:36,  2.04it/s]
 91%|█████████▏| 9782/10712 [1:41:00<07:35,  2.04it/s]
 91%|█████████▏| 9783/10712 [1:41:01<07:34,  2.04it/s]
 91%|█████████▏| 9784/10712 [1:41:01<07:34,  2.04it/s]
 91%|█████████▏| 9785/10712 [1:41:02<07:34,  2.04it/s]
 91%|█████████▏| 9786/10712 [1:41:02<07:33,  2.04it/s]
 91%|█████████▏| 9787/10712 [1:41:03<07:32,  2.04it/s]
 91%|█████████▏| 9788/10712 [1:41:03<07:32,  2.04it/s]
 91%|█████████▏| 9789/10712 [1:41:04<07:32,  2.04it/s]
 91%|█████████▏| 9790/10712 [1:41:04<07:31,  2.04it/s]
 91%|█████████▏| 9791/10712 [1:41:05<07:31,  2.04it/s]
 91%|█████████▏| 9792/10712 [1:41:05<07:31,  2.04it/s]
 91%|█████████▏| 9793/10712 [1:41:06<07:30,  2.04it/s]
 91%|█████████▏| 9794/10712 [1:41:06<07:29,  2.04it/s]
 91%|█████████▏| 9795/10712 [1:41:07<07:29,  2.04it/s]
 91%|█████████▏| 9796/10712 [1:41:07<07:29,  2.04it/s]
 91%|█████████▏| 9797/10712 [1:41:08<07:28,  2.04it/s]
 91%|█████████▏| 9798/10712 [1:41:08<07:27,  2.04it/s]
 91%|█████████▏| 9799/10712 [1:41:09<07:27,  2.04it/s]
 91%|█████████▏| 9800/10712 [1:41:09<07:26,  2.04it/s]{'loss': 3.6059, 'grad_norm': 0.20165720582008362, 'learning_rate': 2.1921791219488398e-05, 'epoch': 0.91}
                                                      

 91%|█████████▏| 9800/10712 [1:41:09<07:26,  2.04it/s]
 91%|█████████▏| 9801/10712 [1:41:10<07:27,  2.04it/s]
 92%|█████████▏| 9802/10712 [1:41:10<07:26,  2.04it/s]
 92%|█████████▏| 9803/10712 [1:41:10<07:26,  2.04it/s]
 92%|█████████▏| 9804/10712 [1:41:11<07:25,  2.04it/s]
 92%|█████████▏| 9805/10712 [1:41:11<07:24,  2.04it/s]
 92%|█████████▏| 9806/10712 [1:41:12<07:24,  2.04it/s]
 92%|█████████▏| 9807/10712 [1:41:12<07:23,  2.04it/s]
 92%|█████████▏| 9808/10712 [1:41:13<07:22,  2.04it/s]
 92%|█████████▏| 9809/10712 [1:41:13<07:22,  2.04it/s]
 92%|█████████▏| 9810/10712 [1:41:14<07:22,  2.04it/s]
 92%|█████████▏| 9811/10712 [1:41:14<07:21,  2.04it/s]
 92%|█████████▏| 9812/10712 [1:41:15<07:21,  2.04it/s]
 92%|█████████▏| 9813/10712 [1:41:15<07:20,  2.04it/s]
 92%|█████████▏| 9814/10712 [1:41:16<07:20,  2.04it/s]
 92%|█████████▏| 9815/10712 [1:41:16<07:19,  2.04it/s]
 92%|█████████▏| 9816/10712 [1:41:17<07:19,  2.04it/s]
 92%|█████████▏| 9817/10712 [1:41:17<07:18,  2.04it/s]
 92%|█████████▏| 9818/10712 [1:41:18<07:18,  2.04it/s]
 92%|█████████▏| 9819/10712 [1:41:18<07:17,  2.04it/s]
 92%|█████████▏| 9820/10712 [1:41:19<07:17,  2.04it/s]
 92%|█████████▏| 9821/10712 [1:41:19<07:17,  2.04it/s]
 92%|█████████▏| 9822/10712 [1:41:20<07:16,  2.04it/s]
 92%|█████████▏| 9823/10712 [1:41:20<07:15,  2.04it/s]
 92%|█████████▏| 9824/10712 [1:41:21<07:15,  2.04it/s]
 92%|█████████▏| 9825/10712 [1:41:21<07:14,  2.04it/s]
                                                      
{'loss': 3.6055, 'grad_norm': 0.19655576348304749, 'learning_rate': 2.0744678976594156e-05, 'epoch': 0.92}

 92%|█████████▏| 9825/10712 [1:41:21<07:14,  2.04it/s]
 92%|█████████▏| 9826/10712 [1:41:22<07:15,  2.03it/s]
 92%|█████████▏| 9827/10712 [1:41:22<07:14,  2.04it/s]
 92%|█████████▏| 9828/10712 [1:41:23<07:14,  2.04it/s]
 92%|█████████▏| 9829/10712 [1:41:23<07:13,  2.04it/s]
 92%|█████████▏| 9830/10712 [1:41:24<07:12,  2.04it/s]
 92%|█████████▏| 9831/10712 [1:41:24<07:11,  2.04it/s]
 92%|█████████▏| 9832/10712 [1:41:25<07:11,  2.04it/s]
 92%|█████████▏| 9833/10712 [1:41:25<07:11,  2.04it/s]
 92%|█████████▏| 9834/10712 [1:41:26<07:10,  2.04it/s]
 92%|█████████▏| 9835/10712 [1:41:26<07:09,  2.04it/s]
 92%|█████████▏| 9836/10712 [1:41:27<07:09,  2.04it/s]
 92%|█████████▏| 9837/10712 [1:41:27<07:08,  2.04it/s]
 92%|█████████▏| 9838/10712 [1:41:28<07:08,  2.04it/s]
 92%|█████████▏| 9839/10712 [1:41:28<07:07,  2.04it/s]
 92%|█████████▏| 9840/10712 [1:41:29<07:07,  2.04it/s]
 92%|█████████▏| 9841/10712 [1:41:29<07:06,  2.04it/s]
 92%|█████████▏| 9842/10712 [1:41:30<07:06,  2.04it/s]
 92%|█████████▏| 9843/10712 [1:41:30<07:06,  2.04it/s]
 92%|█████████▏| 9844/10712 [1:41:31<07:05,  2.04it/s]
 92%|█████████▏| 9845/10712 [1:41:31<07:05,  2.04it/s]
 92%|█████████▏| 9846/10712 [1:41:32<07:04,  2.04it/s]
 92%|█████████▏| 9847/10712 [1:41:32<07:04,  2.04it/s]
 92%|█████████▏| 9848/10712 [1:41:33<07:03,  2.04it/s]
 92%|█████████▏| 9849/10712 [1:41:33<07:02,  2.04it/s]
 92%|█████████▏| 9850/10712 [1:41:34<07:02,  2.04it/s]{'loss': 3.6009, 'grad_norm': 0.2005862295627594, 'learning_rate': 1.959937867965983e-05, 'epoch': 0.92}

                                                      

 92%|█████████▏| 9850/10712 [1:41:34<07:02,  2.04it/s]
 92%|█████████▏| 9851/10712 [1:41:34<07:02,  2.04it/s]
 92%|█████████▏| 9852/10712 [1:41:35<07:02,  2.04it/s]
 92%|█████████▏| 9853/10712 [1:41:35<07:01,  2.04it/s]
 92%|█████████▏| 9854/10712 [1:41:36<07:00,  2.04it/s]
 92%|█████████▏| 9855/10712 [1:41:36<07:00,  2.04it/s]
 92%|█████████▏| 9856/10712 [1:41:36<06:59,  2.04it/s]
 92%|█████████▏| 9857/10712 [1:41:37<06:59,  2.04it/s]
 92%|█████████▏| 9858/10712 [1:41:37<06:58,  2.04it/s]
 92%|█████████▏| 9859/10712 [1:41:38<06:58,  2.04it/s]
 92%|█████████▏| 9860/10712 [1:41:38<06:57,  2.04it/s]
 92%|█████████▏| 9861/10712 [1:41:39<06:57,  2.04it/s]
 92%|█████████▏| 9862/10712 [1:41:39<06:56,  2.04it/s]
 92%|█████████▏| 9863/10712 [1:41:40<06:56,  2.04it/s]
 92%|█████████▏| 9864/10712 [1:41:40<06:55,  2.04it/s]
 92%|█████████▏| 9865/10712 [1:41:41<06:54,  2.04it/s]
 92%|█████████▏| 9866/10712 [1:41:41<06:54,  2.04it/s]
 92%|█████████▏| 9867/10712 [1:41:42<06:54,  2.04it/s]
 92%|█████████▏| 9868/10712 [1:41:42<06:53,  2.04it/s]
 92%|█████████▏| 9869/10712 [1:41:43<06:53,  2.04it/s]
 92%|█████████▏| 9870/10712 [1:41:43<06:53,  2.04it/s]
 92%|█████████▏| 9871/10712 [1:41:44<06:52,  2.04it/s]
 92%|█████████▏| 9872/10712 [1:41:44<06:51,  2.04it/s]
 92%|█████████▏| 9873/10712 [1:41:45<06:50,  2.04it/s]
 92%|█████████▏| 9874/10712 [1:41:45<06:50,  2.04it/s]
 92%|█████████▏| 9875/10712 [1:41:46<06:50,  2.04it/s]{'loss': 3.6047, 'grad_norm': 0.2050933688879013, 'learning_rate': 1.8485966351276417e-05, 'epoch': 0.92}
                                                      

 92%|█████████▏| 9875/10712 [1:41:46<06:50,  2.04it/s]
 92%|█████████▏| 9876/10712 [1:41:46<06:50,  2.04it/s]
 92%|█████████▏| 9877/10712 [1:41:47<06:49,  2.04it/s]
 92%|█████████▏| 9878/10712 [1:41:47<06:49,  2.04it/s]
 92%|█████████▏| 9879/10712 [1:41:48<06:48,  2.04it/s]
 92%|█████████▏| 9880/10712 [1:41:48<06:48,  2.04it/s]
 92%|█████████▏| 9881/10712 [1:41:49<06:47,  2.04it/s]
 92%|█████████▏| 9882/10712 [1:41:49<06:46,  2.04it/s]
 92%|█████████▏| 9883/10712 [1:41:50<06:46,  2.04it/s]
 92%|█████████▏| 9884/10712 [1:41:50<06:45,  2.04it/s]
 92%|█████████▏| 9885/10712 [1:41:51<06:45,  2.04it/s]
 92%|█████████▏| 9886/10712 [1:41:51<06:44,  2.04it/s]
 92%|█████████▏| 9887/10712 [1:41:52<06:44,  2.04it/s]
 92%|█████████▏| 9888/10712 [1:41:52<06:44,  2.04it/s]
 92%|█████████▏| 9889/10712 [1:41:53<06:43,  2.04it/s]
 92%|█████████▏| 9890/10712 [1:41:53<06:41,  2.04it/s]
 92%|█████████▏| 9891/10712 [1:41:54<06:42,  2.04it/s]
 92%|█████████▏| 9892/10712 [1:41:54<06:41,  2.04it/s]
 92%|█████████▏| 9893/10712 [1:41:55<06:41,  2.04it/s]
 92%|█████████▏| 9894/10712 [1:41:55<06:41,  2.04it/s]
 92%|█████████▏| 9895/10712 [1:41:56<06:40,  2.04it/s]
 92%|█████████▏| 9896/10712 [1:41:56<06:40,  2.04it/s]
 92%|█████████▏| 9897/10712 [1:41:57<06:39,  2.04it/s]
 92%|█████████▏| 9898/10712 [1:41:57<06:38,  2.04it/s]
 92%|█████████▏| 9899/10712 [1:41:58<06:38,  2.04it/s]
 92%|█████████▏| 9900/10712 [1:41:58<06:37,  2.04it/s]{'loss': 3.5998, 'grad_norm': 0.19912250339984894, 'learning_rate': 1.7404515897379126e-05, 'epoch': 0.92}
                                                      

 92%|█████████▏| 9900/10712 [1:41:58<06:37,  2.04it/s]
 92%|█████████▏| 9901/10712 [1:41:59<06:38,  2.03it/s]
 92%|█████████▏| 9902/10712 [1:41:59<06:43,  2.01it/s]
 92%|█████████▏| 9903/10712 [1:42:00<06:40,  2.02it/s]
 92%|█████████▏| 9904/10712 [1:42:00<06:39,  2.02it/s]
 92%|█████████▏| 9905/10712 [1:42:01<06:37,  2.03it/s]
 92%|█████████▏| 9906/10712 [1:42:01<06:36,  2.04it/s]
 92%|█████████▏| 9907/10712 [1:42:02<06:34,  2.04it/s]
 92%|█████████▏| 9908/10712 [1:42:02<06:34,  2.04it/s]
 93%|█████████▎| 9909/10712 [1:42:02<06:34,  2.04it/s]
 93%|█████████▎| 9910/10712 [1:42:03<06:33,  2.04it/s]
 93%|█████████▎| 9911/10712 [1:42:03<06:32,  2.04it/s]
 93%|█████████▎| 9912/10712 [1:42:04<06:32,  2.04it/s]
 93%|█████████▎| 9913/10712 [1:42:04<06:31,  2.04it/s]
 93%|█████████▎| 9914/10712 [1:42:05<06:31,  2.04it/s]
 93%|█████████▎| 9915/10712 [1:42:05<06:30,  2.04it/s]
 93%|█████████▎| 9916/10712 [1:42:06<06:30,  2.04it/s]
 93%|█████████▎| 9917/10712 [1:42:06<06:30,  2.04it/s]
 93%|█████████▎| 9918/10712 [1:42:07<06:29,  2.04it/s]
 93%|█████████▎| 9919/10712 [1:42:07<06:28,  2.04it/s]
 93%|█████████▎| 9920/10712 [1:42:08<06:28,  2.04it/s]
 93%|█████████▎| 9921/10712 [1:42:08<06:27,  2.04it/s]
 93%|█████████▎| 9922/10712 [1:42:09<06:27,  2.04it/s]
 93%|█████████▎| 9923/10712 [1:42:09<06:26,  2.04it/s]
 93%|█████████▎| 9924/10712 [1:42:10<06:26,  2.04it/s]
 93%|█████████▎| 9925/10712 [1:42:10<06:25,  2.04it/s]{'loss': 3.5991, 'grad_norm': 0.19864162802696228, 'learning_rate': 1.6355099102342562e-05, 'epoch': 0.93}

                                                      

 93%|█████████▎| 9925/10712 [1:42:10<06:25,  2.04it/s]
 93%|█████████▎| 9926/10712 [1:42:11<06:25,  2.04it/s]
 93%|█████████▎| 9927/10712 [1:42:11<06:25,  2.04it/s]
 93%|█████████▎| 9928/10712 [1:42:12<06:24,  2.04it/s]
 93%|█████████▎| 9929/10712 [1:42:12<06:23,  2.04it/s]
 93%|█████████▎| 9930/10712 [1:42:13<06:23,  2.04it/s]
 93%|█████████▎| 9931/10712 [1:42:13<06:23,  2.04it/s]
 93%|█████████▎| 9932/10712 [1:42:14<06:22,  2.04it/s]
 93%|█████████▎| 9933/10712 [1:42:14<06:21,  2.04it/s]
 93%|█████████▎| 9934/10712 [1:42:15<06:21,  2.04it/s]
 93%|█████████▎| 9935/10712 [1:42:15<06:21,  2.04it/s]
 93%|█████████▎| 9936/10712 [1:42:16<06:20,  2.04it/s]
 93%|█████████▎| 9937/10712 [1:42:16<06:19,  2.04it/s]
 93%|█████████▎| 9938/10712 [1:42:17<06:19,  2.04it/s]
 93%|█████████▎| 9939/10712 [1:42:17<06:19,  2.04it/s]
 93%|█████████▎| 9940/10712 [1:42:18<06:18,  2.04it/s]
 93%|█████████▎| 9941/10712 [1:42:18<06:17,  2.04it/s]
 93%|█████████▎| 9942/10712 [1:42:19<06:17,  2.04it/s]
 93%|█████████▎| 9943/10712 [1:42:19<06:17,  2.04it/s]
 93%|█████████▎| 9944/10712 [1:42:20<06:16,  2.04it/s]
 93%|█████████▎| 9945/10712 [1:42:20<06:15,  2.04it/s]
 93%|█████████▎| 9946/10712 [1:42:21<06:15,  2.04it/s]
 93%|█████████▎| 9947/10712 [1:42:21<06:14,  2.04it/s]
 93%|█████████▎| 9948/10712 [1:42:22<06:13,  2.04it/s]
 93%|█████████▎| 9949/10712 [1:42:22<06:13,  2.04it/s]
 93%|█████████▎| 9950/10712 [1:42:23<06:13,  2.04it/s]{'loss': 3.6063, 'grad_norm': 0.20128130912780762, 'learning_rate': 1.5337785624215004e-05, 'epoch': 0.93}
                                                      

 93%|█████████▎| 9950/10712 [1:42:23<06:13,  2.04it/s]
 93%|█████████▎| 9951/10712 [1:42:23<06:13,  2.04it/s]
 93%|█████████▎| 9952/10712 [1:42:24<06:12,  2.04it/s]
 93%|█████████▎| 9953/10712 [1:42:24<06:12,  2.04it/s]
 93%|█████████▎| 9954/10712 [1:42:25<06:11,  2.04it/s]
 93%|█████████▎| 9955/10712 [1:42:25<06:10,  2.04it/s]
 93%|█████████▎| 9956/10712 [1:42:26<06:09,  2.04it/s]
 93%|█████████▎| 9957/10712 [1:42:26<06:09,  2.04it/s]
 93%|█████████▎| 9958/10712 [1:42:26<06:09,  2.04it/s]
 93%|█████████▎| 9959/10712 [1:42:27<06:09,  2.04it/s]
 93%|█████████▎| 9960/10712 [1:42:27<06:08,  2.04it/s]
 93%|█████████▎| 9961/10712 [1:42:28<06:08,  2.04it/s]
 93%|█████████▎| 9962/10712 [1:42:28<06:07,  2.04it/s]
 93%|█████████▎| 9963/10712 [1:42:29<06:06,  2.04it/s]
 93%|█████████▎| 9964/10712 [1:42:29<06:06,  2.04it/s]
 93%|█████████▎| 9965/10712 [1:42:30<06:06,  2.04it/s]
 93%|█████████▎| 9966/10712 [1:42:30<06:05,  2.04it/s]
 93%|█████████▎| 9967/10712 [1:42:31<06:05,  2.04it/s]
 93%|█████████▎| 9968/10712 [1:42:31<06:04,  2.04it/s]
 93%|█████████▎| 9969/10712 [1:42:32<06:04,  2.04it/s]
 93%|█████████▎| 9970/10712 [1:42:32<06:03,  2.04it/s]
 93%|█████████▎| 9971/10712 [1:42:33<06:03,  2.04it/s]
 93%|█████████▎| 9972/10712 [1:42:33<06:02,  2.04it/s]
 93%|█████████▎| 9973/10712 [1:42:34<06:02,  2.04it/s]
 93%|█████████▎| 9974/10712 [1:42:34<06:01,  2.04it/s]
 93%|█████████▎| 9975/10712 [1:42:35<06:01,  2.04it/s]{'loss': 3.6072, 'grad_norm': 0.1993636190891266, 'learning_rate': 1.435264299009531e-05, 'epoch': 0.93}

                                                      

 93%|█████████▎| 9975/10712 [1:42:35<06:01,  2.04it/s]
 93%|█████████▎| 9976/10712 [1:42:35<06:01,  2.04it/s]
 93%|█████████▎| 9977/10712 [1:42:36<06:01,  2.04it/s]
 93%|█████████▎| 9978/10712 [1:42:36<06:00,  2.04it/s]
 93%|█████████▎| 9979/10712 [1:42:37<05:59,  2.04it/s]
 93%|█████████▎| 9980/10712 [1:42:37<05:58,  2.04it/s]
 93%|█████████▎| 9981/10712 [1:42:38<05:58,  2.04it/s]
 93%|█████████▎| 9982/10712 [1:42:38<05:58,  2.04it/s]
 93%|█████████▎| 9983/10712 [1:42:39<05:57,  2.04it/s]
 93%|█████████▎| 9984/10712 [1:42:39<05:57,  2.04it/s]
 93%|█████████▎| 9985/10712 [1:42:40<05:56,  2.04it/s]
 93%|█████████▎| 9986/10712 [1:42:40<05:56,  2.04it/s]
 93%|█████████▎| 9987/10712 [1:42:41<05:55,  2.04it/s]
 93%|█████████▎| 9988/10712 [1:42:41<05:55,  2.04it/s]
 93%|█████████▎| 9989/10712 [1:42:42<05:54,  2.04it/s]
 93%|█████████▎| 9990/10712 [1:42:42<05:54,  2.04it/s]
 93%|█████████▎| 9991/10712 [1:42:43<05:53,  2.04it/s]
 93%|█████████▎| 9992/10712 [1:42:43<05:53,  2.03it/s]
 93%|█████████▎| 9993/10712 [1:42:44<05:53,  2.03it/s]
 93%|█████████▎| 9994/10712 [1:42:44<05:52,  2.04it/s]
 93%|█████████▎| 9995/10712 [1:42:45<05:52,  2.04it/s]
 93%|█████████▎| 9996/10712 [1:42:45<05:51,  2.04it/s]
 93%|█████████▎| 9997/10712 [1:42:46<05:50,  2.04it/s]
 93%|█████████▎| 9998/10712 [1:42:46<05:50,  2.04it/s]
 93%|█████████▎| 9999/10712 [1:42:47<05:49,  2.04it/s]
 93%|█████████▎| 10000/10712 [1:42:47<05:49,  2.04it/s]{'loss': 3.6081, 'grad_norm': 0.19500356912612915, 'learning_rate': 1.339973659165028e-05, 'epoch': 0.93}
                                                       

 93%|█████████▎| 10000/10712 [1:42:47<05:49,  2.04it/s]
 93%|█████████▎| 10001/10712 [1:42:48<05:48,  2.04it/s]
 93%|█████████▎| 10002/10712 [1:42:48<05:48,  2.04it/s]
 93%|█████████▎| 10003/10712 [1:42:49<05:47,  2.04it/s]
 93%|█████████▎| 10004/10712 [1:42:49<05:46,  2.04it/s]
 93%|█████████▎| 10005/10712 [1:42:50<05:46,  2.04it/s]
 93%|█████████▎| 10006/10712 [1:42:50<05:46,  2.04it/s]
 93%|█████████▎| 10007/10712 [1:42:51<05:45,  2.04it/s]
 93%|█████████▎| 10008/10712 [1:42:51<05:44,  2.04it/s]
 93%|█████████▎| 10009/10712 [1:42:52<05:44,  2.04it/s]
 93%|█████████▎| 10010/10712 [1:42:52<05:44,  2.04it/s]
 93%|█████████▎| 10011/10712 [1:42:52<05:43,  2.04it/s]
 93%|█████████▎| 10012/10712 [1:42:53<05:42,  2.04it/s]
 93%|█████████▎| 10013/10712 [1:42:53<05:42,  2.04it/s]
 93%|█████████▎| 10014/10712 [1:42:54<05:41,  2.04it/s]
 93%|█████████▎| 10015/10712 [1:42:54<05:41,  2.04it/s]
 94%|█████████▎| 10016/10712 [1:42:55<05:40,  2.04it/s]
 94%|█████████▎| 10017/10712 [1:42:55<05:40,  2.04it/s]
 94%|█████████▎| 10018/10712 [1:42:56<05:40,  2.04it/s]
 94%|█████████▎| 10019/10712 [1:42:56<05:39,  2.04it/s]
 94%|█████████▎| 10020/10712 [1:42:57<05:39,  2.04it/s]
 94%|█████████▎| 10021/10712 [1:42:57<05:38,  2.04it/s]
 94%|█████████▎| 10022/10712 [1:42:58<05:38,  2.04it/s]
 94%|█████████▎| 10023/10712 [1:42:58<05:37,  2.04it/s]
 94%|█████████▎| 10024/10712 [1:42:59<05:37,  2.04it/s]
 94%|█████████▎| 10025/10712 [1:42:59<05:36,  2.04it/s]
                                                       
{'loss': 3.5938, 'grad_norm': 0.19525422155857086, 'learning_rate': 1.2479129680773971e-05, 'epoch': 0.94}

 94%|█████████▎| 10025/10712 [1:42:59<05:36,  2.04it/s]
 94%|█████████▎| 10026/10712 [1:43:00<05:36,  2.04it/s]
 94%|█████████▎| 10027/10712 [1:43:00<05:36,  2.04it/s]
 94%|█████████▎| 10028/10712 [1:43:01<05:35,  2.04it/s]
 94%|█████████▎| 10029/10712 [1:43:01<05:34,  2.04it/s]
 94%|█████████▎| 10030/10712 [1:43:02<05:33,  2.04it/s]
 94%|█████████▎| 10031/10712 [1:43:02<05:33,  2.04it/s]
 94%|█████████▎| 10032/10712 [1:43:03<05:33,  2.04it/s]
 94%|█████████▎| 10033/10712 [1:43:03<05:32,  2.04it/s]
 94%|█████████▎| 10034/10712 [1:43:04<05:31,  2.04it/s]
 94%|█████████▎| 10035/10712 [1:43:04<05:31,  2.04it/s]
 94%|█████████▎| 10036/10712 [1:43:05<05:31,  2.04it/s]
 94%|█████████▎| 10037/10712 [1:43:05<05:30,  2.04it/s]
 94%|█████████▎| 10038/10712 [1:43:06<05:29,  2.04it/s]
 94%|█████████▎| 10039/10712 [1:43:06<05:29,  2.04it/s]
 94%|█████████▎| 10040/10712 [1:43:07<05:29,  2.04it/s]
 94%|█████████▎| 10041/10712 [1:43:07<05:28,  2.04it/s]
 94%|█████████▎| 10042/10712 [1:43:08<05:28,  2.04it/s]
 94%|█████████▍| 10043/10712 [1:43:08<05:28,  2.04it/s]
 94%|█████████▍| 10044/10712 [1:43:09<05:27,  2.04it/s]
 94%|█████████▍| 10045/10712 [1:43:09<05:26,  2.04it/s]
 94%|█████████▍| 10046/10712 [1:43:10<05:26,  2.04it/s]
 94%|█████████▍| 10047/10712 [1:43:10<05:25,  2.04it/s]
 94%|█████████▍| 10048/10712 [1:43:11<05:25,  2.04it/s]
 94%|█████████▍| 10049/10712 [1:43:11<05:24,  2.04it/s]
 94%|█████████▍| 10050/10712 [1:43:12<05:24,  2.04it/s]{'loss': 3.591, 'grad_norm': 0.19812296330928802, 'learning_rate': 1.1590883365389549e-05, 'epoch': 0.94}
                                                       

 94%|█████████▍| 10050/10712 [1:43:12<05:24,  2.04it/s]
 94%|█████████▍| 10051/10712 [1:43:12<05:24,  2.04it/s]
 94%|█████████▍| 10052/10712 [1:43:13<05:23,  2.04it/s]
 94%|█████████▍| 10053/10712 [1:43:13<05:22,  2.04it/s]
 94%|█████████▍| 10054/10712 [1:43:14<05:22,  2.04it/s]
 94%|█████████▍| 10055/10712 [1:43:14<05:22,  2.04it/s]
 94%|█████████▍| 10056/10712 [1:43:15<05:21,  2.04it/s]
 94%|█████████▍| 10057/10712 [1:43:15<05:20,  2.04it/s]
 94%|█████████▍| 10058/10712 [1:43:16<05:20,  2.04it/s]
 94%|█████████▍| 10059/10712 [1:43:16<05:20,  2.04it/s]
 94%|█████████▍| 10060/10712 [1:43:17<05:19,  2.04it/s]
 94%|█████████▍| 10061/10712 [1:43:17<05:19,  2.04it/s]
 94%|█████████▍| 10062/10712 [1:43:17<05:18,  2.04it/s]
 94%|█████████▍| 10063/10712 [1:43:18<05:17,  2.04it/s]
 94%|█████████▍| 10064/10712 [1:43:18<05:17,  2.04it/s]
 94%|█████████▍| 10065/10712 [1:43:19<05:16,  2.05it/s]
 94%|█████████▍| 10066/10712 [1:43:19<05:15,  2.05it/s]
 94%|█████████▍| 10067/10712 [1:43:20<05:15,  2.04it/s]
 94%|█████████▍| 10068/10712 [1:43:20<05:15,  2.04it/s]
 94%|█████████▍| 10069/10712 [1:43:21<05:14,  2.04it/s]
 94%|█████████▍| 10070/10712 [1:43:21<05:14,  2.04it/s]
 94%|█████████▍| 10071/10712 [1:43:22<05:14,  2.04it/s]
 94%|█████████▍| 10072/10712 [1:43:22<05:13,  2.04it/s]
 94%|█████████▍| 10073/10712 [1:43:23<05:12,  2.04it/s]
 94%|█████████▍| 10074/10712 [1:43:23<05:12,  2.04it/s]
 94%|█████████▍| 10075/10712 [1:43:24<05:12,  2.04it/s]
                                                       
{'loss': 3.6009, 'grad_norm': 0.19712331891059875, 'learning_rate': 1.0735056605392756e-05, 'epoch': 0.94}

 94%|█████████▍| 10075/10712 [1:43:24<05:12,  2.04it/s]
 94%|█████████▍| 10076/10712 [1:43:24<05:12,  2.04it/s]
 94%|█████████▍| 10077/10712 [1:43:25<05:11,  2.04it/s]
 94%|█████████▍| 10078/10712 [1:43:25<05:11,  2.04it/s]
 94%|█████████▍| 10079/10712 [1:43:26<05:10,  2.04it/s]
 94%|█████████▍| 10080/10712 [1:43:26<05:09,  2.04it/s]
 94%|█████████▍| 10081/10712 [1:43:27<05:09,  2.04it/s]
 94%|█████████▍| 10082/10712 [1:43:27<05:08,  2.04it/s]
 94%|█████████▍| 10083/10712 [1:43:28<05:08,  2.04it/s]
 94%|█████████▍| 10084/10712 [1:43:28<05:08,  2.04it/s]
 94%|█████████▍| 10085/10712 [1:43:29<05:07,  2.04it/s]
 94%|█████████▍| 10086/10712 [1:43:29<05:06,  2.04it/s]
 94%|█████████▍| 10087/10712 [1:43:30<05:05,  2.04it/s]
 94%|█████████▍| 10088/10712 [1:43:30<05:05,  2.04it/s]
 94%|█████████▍| 10089/10712 [1:43:31<05:05,  2.04it/s]
 94%|█████████▍| 10090/10712 [1:43:31<05:04,  2.04it/s]
 94%|█████████▍| 10091/10712 [1:43:32<05:04,  2.04it/s]
 94%|█████████▍| 10092/10712 [1:43:32<05:04,  2.04it/s]
 94%|█████████▍| 10093/10712 [1:43:33<05:03,  2.04it/s]
 94%|█████████▍| 10094/10712 [1:43:33<05:03,  2.04it/s]
 94%|█████████▍| 10095/10712 [1:43:34<05:03,  2.04it/s]
 94%|█████████▍| 10096/10712 [1:43:34<05:02,  2.04it/s]
 94%|█████████▍| 10097/10712 [1:43:35<05:01,  2.04it/s]
 94%|█████████▍| 10098/10712 [1:43:35<05:00,  2.04it/s]
 94%|█████████▍| 10099/10712 [1:43:36<05:00,  2.04it/s]
 94%|█████████▍| 10100/10712 [1:43:36<05:00,  2.04it/s]
                                                       
{'loss': 3.5919, 'grad_norm': 0.19641870260238647, 'learning_rate': 9.911706208738547e-06, 'epoch': 0.94}

 94%|█████████▍| 10100/10712 [1:43:36<05:00,  2.04it/s]
 94%|█████████▍| 10101/10712 [1:43:37<05:00,  2.03it/s]
 94%|█████████▍| 10102/10712 [1:43:37<04:59,  2.03it/s]
 94%|█████████▍| 10103/10712 [1:43:38<04:59,  2.04it/s]
 94%|█████████▍| 10104/10712 [1:43:38<04:58,  2.04it/s]
 94%|█████████▍| 10105/10712 [1:43:39<04:57,  2.04it/s]
 94%|█████████▍| 10106/10712 [1:43:39<04:57,  2.04it/s]
 94%|█████████▍| 10107/10712 [1:43:40<04:57,  2.04it/s]
 94%|█████████▍| 10108/10712 [1:43:40<04:56,  2.04it/s]
 94%|█████████▍| 10109/10712 [1:43:41<04:56,  2.04it/s]
 94%|█████████▍| 10110/10712 [1:43:41<04:55,  2.04it/s]
 94%|█████████▍| 10111/10712 [1:43:42<04:55,  2.04it/s]
 94%|█████████▍| 10112/10712 [1:43:42<04:54,  2.04it/s]
 94%|█████████▍| 10113/10712 [1:43:42<04:53,  2.04it/s]
 94%|█████████▍| 10114/10712 [1:43:43<04:53,  2.04it/s]
 94%|█████████▍| 10115/10712 [1:43:43<04:52,  2.04it/s]
 94%|█████████▍| 10116/10712 [1:43:44<04:52,  2.04it/s]
 94%|█████████▍| 10117/10712 [1:43:44<04:51,  2.04it/s]
 94%|█████████▍| 10118/10712 [1:43:45<04:51,  2.04it/s]
 94%|█████████▍| 10119/10712 [1:43:45<04:50,  2.04it/s]
 94%|█████████▍| 10120/10712 [1:43:46<04:50,  2.04it/s]
 94%|█████████▍| 10121/10712 [1:43:46<04:49,  2.04it/s]
 94%|█████████▍| 10122/10712 [1:43:47<04:49,  2.04it/s]
 95%|█████████▍| 10123/10712 [1:43:47<04:49,  2.04it/s]
 95%|█████████▍| 10124/10712 [1:43:48<04:48,  2.04it/s]
 95%|█████████▍| 10125/10712 [1:43:48<04:47,  2.04it/s]
                                                       
{'loss': 3.5974, 'grad_norm': 0.19581060111522675, 'learning_rate': 9.120886827669928e-06, 'epoch': 0.95}

 95%|█████████▍| 10125/10712 [1:43:48<04:47,  2.04it/s]
 95%|█████████▍| 10126/10712 [1:43:49<04:47,  2.04it/s]
 95%|█████████▍| 10127/10712 [1:43:49<04:47,  2.04it/s]
 95%|█████████▍| 10128/10712 [1:43:50<04:46,  2.04it/s]
 95%|█████████▍| 10129/10712 [1:43:50<04:45,  2.04it/s]
 95%|█████████▍| 10130/10712 [1:43:51<04:45,  2.04it/s]
 95%|█████████▍| 10131/10712 [1:43:51<04:44,  2.04it/s]
 95%|█████████▍| 10132/10712 [1:43:52<04:44,  2.04it/s]
 95%|█████████▍| 10133/10712 [1:43:52<04:44,  2.04it/s]
 95%|█████████▍| 10134/10712 [1:43:53<04:43,  2.04it/s]
 95%|█████████▍| 10135/10712 [1:43:53<04:42,  2.04it/s]
 95%|█████████▍| 10136/10712 [1:43:54<04:42,  2.04it/s]
 95%|█████████▍| 10137/10712 [1:43:54<04:41,  2.04it/s]
 95%|█████████▍| 10138/10712 [1:43:55<04:41,  2.04it/s]
 95%|█████████▍| 10139/10712 [1:43:55<04:40,  2.04it/s]
 95%|█████████▍| 10140/10712 [1:43:56<04:40,  2.04it/s]
 95%|█████████▍| 10141/10712 [1:43:56<04:39,  2.04it/s]
 95%|█████████▍| 10142/10712 [1:43:57<04:39,  2.04it/s]
 95%|█████████▍| 10143/10712 [1:43:57<04:39,  2.04it/s]
 95%|█████████▍| 10144/10712 [1:43:58<04:38,  2.04it/s]
 95%|█████████▍| 10145/10712 [1:43:58<04:37,  2.04it/s]
 95%|█████████▍| 10146/10712 [1:43:59<04:37,  2.04it/s]
 95%|█████████▍| 10147/10712 [1:43:59<04:37,  2.04it/s]
 95%|█████████▍| 10148/10712 [1:44:00<04:36,  2.04it/s]
 95%|█████████▍| 10149/10712 [1:44:00<04:35,  2.04it/s]
 95%|█████████▍| 10150/10712 [1:44:01<04:35,  2.04it/s]{'loss': 3.6039, 'grad_norm': 0.19446131587028503, 'learning_rate': 8.3626509550907e-06, 'epoch': 0.95}
                                                       

 95%|█████████▍| 10150/10712 [1:44:01<04:35,  2.04it/s]
 95%|█████████▍| 10151/10712 [1:44:01<04:35,  2.04it/s]
 95%|█████████▍| 10152/10712 [1:44:02<04:34,  2.04it/s]
 95%|█████████▍| 10153/10712 [1:44:02<04:34,  2.04it/s]
 95%|█████████▍| 10154/10712 [1:44:03<04:33,  2.04it/s]
 95%|█████████▍| 10155/10712 [1:44:03<04:32,  2.04it/s]
 95%|█████████▍| 10156/10712 [1:44:04<04:32,  2.04it/s]
 95%|█████████▍| 10157/10712 [1:44:04<04:32,  2.04it/s]
 95%|█████████▍| 10158/10712 [1:44:05<04:31,  2.04it/s]
 95%|█████████▍| 10159/10712 [1:44:05<04:31,  2.04it/s]
 95%|█████████▍| 10160/10712 [1:44:06<04:30,  2.04it/s]
 95%|█████████▍| 10161/10712 [1:44:06<04:30,  2.04it/s]
 95%|█████████▍| 10162/10712 [1:44:07<04:29,  2.04it/s]
 95%|█████████▍| 10163/10712 [1:44:07<04:28,  2.04it/s]
 95%|█████████▍| 10164/10712 [1:44:08<04:28,  2.04it/s]
 95%|█████████▍| 10165/10712 [1:44:08<04:28,  2.04it/s]
 95%|█████████▍| 10166/10712 [1:44:08<04:27,  2.04it/s]
 95%|█████████▍| 10167/10712 [1:44:09<04:27,  2.04it/s]
 95%|█████████▍| 10168/10712 [1:44:09<04:26,  2.04it/s]
 95%|█████████▍| 10169/10712 [1:44:10<04:26,  2.04it/s]
 95%|█████████▍| 10170/10712 [1:44:10<04:25,  2.04it/s]
 95%|█████████▍| 10171/10712 [1:44:11<04:25,  2.04it/s]
 95%|█████████▍| 10172/10712 [1:44:11<04:24,  2.04it/s]
 95%|█████████▍| 10173/10712 [1:44:12<04:24,  2.04it/s]
 95%|█████████▍| 10174/10712 [1:44:12<04:23,  2.04it/s]
 95%|█████████▍| 10175/10712 [1:44:13<04:23,  2.04it/s]{'loss': 3.5921, 'grad_norm': 0.19672468304634094, 'learning_rate': 7.637048921080636e-06, 'epoch': 0.95}

                                                       

 95%|█████████▍| 10175/10712 [1:44:13<04:23,  2.04it/s]
 95%|█████████▍| 10176/10712 [1:44:13<04:23,  2.04it/s]
 95%|█████████▌| 10177/10712 [1:44:14<04:22,  2.04it/s]
 95%|█████████▌| 10178/10712 [1:44:14<04:22,  2.04it/s]
 95%|█████████▌| 10179/10712 [1:44:15<04:21,  2.04it/s]
 95%|█████████▌| 10180/10712 [1:44:15<04:20,  2.04it/s]
 95%|█████████▌| 10181/10712 [1:44:16<04:20,  2.04it/s]
 95%|█████████▌| 10182/10712 [1:44:16<04:19,  2.04it/s]
 95%|█████████▌| 10183/10712 [1:44:17<04:19,  2.04it/s]
 95%|█████████▌| 10184/10712 [1:44:17<04:18,  2.04it/s]
 95%|█████████▌| 10185/10712 [1:44:18<04:18,  2.04it/s]
 95%|█████████▌| 10186/10712 [1:44:18<04:17,  2.04it/s]
 95%|█████████▌| 10187/10712 [1:44:19<04:16,  2.04it/s]
 95%|█████████▌| 10188/10712 [1:44:19<04:16,  2.04it/s]
 95%|█████████▌| 10189/10712 [1:44:20<04:16,  2.04it/s]
 95%|█████████▌| 10190/10712 [1:44:20<04:15,  2.04it/s]
 95%|█████████▌| 10191/10712 [1:44:21<04:15,  2.04it/s]
 95%|█████████▌| 10192/10712 [1:44:21<04:14,  2.04it/s]
 95%|█████████▌| 10193/10712 [1:44:22<04:14,  2.04it/s]
 95%|█████████▌| 10194/10712 [1:44:22<04:14,  2.04it/s]
 95%|█████████▌| 10195/10712 [1:44:23<04:13,  2.04it/s]
 95%|█████████▌| 10196/10712 [1:44:23<04:13,  2.04it/s]
 95%|█████████▌| 10197/10712 [1:44:24<04:12,  2.04it/s]
 95%|█████████▌| 10198/10712 [1:44:24<04:12,  2.04it/s]
 95%|█████████▌| 10199/10712 [1:44:25<04:11,  2.04it/s]
 95%|█████████▌| 10200/10712 [1:44:25<04:11,  2.04it/s]{'loss': 3.6044, 'grad_norm': 0.19667847454547882, 'learning_rate': 6.9441288895552015e-06, 'epoch': 0.95}

                                                       

 95%|█████████▌| 10200/10712 [1:44:25<04:11,  2.04it/s]
 95%|█████████▌| 10201/10712 [1:44:26<04:10,  2.04it/s]
 95%|█████████▌| 10202/10712 [1:44:26<04:09,  2.04it/s]
 95%|█████████▌| 10203/10712 [1:44:27<04:09,  2.04it/s]
 95%|█████████▌| 10204/10712 [1:44:27<04:09,  2.04it/s]
 95%|█████████▌| 10205/10712 [1:44:28<04:08,  2.04it/s]
 95%|█████████▌| 10206/10712 [1:44:28<04:07,  2.04it/s]
 95%|█████████▌| 10207/10712 [1:44:29<04:07,  2.04it/s]
 95%|█████████▌| 10208/10712 [1:44:29<04:07,  2.04it/s]
 95%|█████████▌| 10209/10712 [1:44:30<04:06,  2.04it/s]
 95%|█████████▌| 10210/10712 [1:44:30<04:05,  2.04it/s]
 95%|█████████▌| 10211/10712 [1:44:31<04:05,  2.04it/s]
 95%|█████████▌| 10212/10712 [1:44:31<04:05,  2.04it/s]
 95%|█████████▌| 10213/10712 [1:44:32<04:04,  2.04it/s]
 95%|█████████▌| 10214/10712 [1:44:32<04:03,  2.04it/s]
 95%|█████████▌| 10215/10712 [1:44:33<04:03,  2.04it/s]
 95%|█████████▌| 10216/10712 [1:44:33<04:03,  2.04it/s]
 95%|█████████▌| 10217/10712 [1:44:33<04:02,  2.04it/s]
 95%|█████████▌| 10218/10712 [1:44:34<04:02,  2.04it/s]
 95%|█████████▌| 10219/10712 [1:44:34<04:01,  2.04it/s]
 95%|█████████▌| 10220/10712 [1:44:35<04:01,  2.04it/s]
 95%|█████████▌| 10221/10712 [1:44:35<04:00,  2.04it/s]
 95%|█████████▌| 10222/10712 [1:44:36<03:59,  2.04it/s]
 95%|█████████▌| 10223/10712 [1:44:36<03:59,  2.04it/s]
 95%|█████████▌| 10224/10712 [1:44:37<03:58,  2.04it/s]
 95%|█████████▌| 10225/10712 [1:44:37<03:58,  2.04it/s]{'loss': 3.6003, 'grad_norm': 0.2012169063091278, 'learning_rate': 6.283936855068006e-06, 'epoch': 0.95}
                                                       

 95%|█████████▌| 10225/10712 [1:44:37<03:58,  2.04it/s]
 95%|█████████▌| 10226/10712 [1:44:38<03:58,  2.04it/s]
 95%|█████████▌| 10227/10712 [1:44:38<03:57,  2.04it/s]
 95%|█████████▌| 10228/10712 [1:44:39<03:57,  2.04it/s]
 95%|█████████▌| 10229/10712 [1:44:39<03:56,  2.04it/s]
 96%|█████████▌| 10230/10712 [1:44:40<03:56,  2.04it/s]
 96%|█████████▌| 10231/10712 [1:44:40<03:55,  2.04it/s]
 96%|█████████▌| 10232/10712 [1:44:41<03:55,  2.04it/s]
 96%|█████████▌| 10233/10712 [1:44:41<03:54,  2.04it/s]
 96%|█████████▌| 10234/10712 [1:44:42<04:31,  1.76it/s]
 96%|█████████▌| 10235/10712 [1:44:43<04:19,  1.84it/s]
 96%|█████████▌| 10236/10712 [1:44:43<04:11,  1.89it/s]
 96%|█████████▌| 10237/10712 [1:44:44<04:05,  1.93it/s]
 96%|█████████▌| 10238/10712 [1:44:44<04:01,  1.96it/s]
 96%|█████████▌| 10239/10712 [1:44:45<03:58,  1.99it/s]
 96%|█████████▌| 10240/10712 [1:44:45<03:55,  2.00it/s]
 96%|█████████▌| 10241/10712 [1:44:46<03:54,  2.01it/s]
 96%|█████████▌| 10242/10712 [1:44:46<03:52,  2.02it/s]
 96%|█████████▌| 10243/10712 [1:44:46<03:51,  2.03it/s]
 96%|█████████▌| 10244/10712 [1:44:47<03:50,  2.03it/s]
 96%|█████████▌| 10245/10712 [1:44:47<03:49,  2.04it/s]
 96%|█████████▌| 10246/10712 [1:44:48<03:48,  2.04it/s]
 96%|█████████▌| 10247/10712 [1:44:48<03:48,  2.04it/s]
 96%|█████████▌| 10248/10712 [1:44:49<03:47,  2.04it/s]
 96%|█████████▌| 10249/10712 [1:44:49<03:47,  2.04it/s]
 96%|█████████▌| 10250/10712 [1:44:50<03:46,  2.04it/s]{'loss': 3.6011, 'grad_norm': 0.19849394261837006, 'learning_rate': 5.656516639758136e-06, 'epoch': 0.96}
                                                       

 96%|█████████▌| 10250/10712 [1:44:50<03:46,  2.04it/s]
 96%|█████████▌| 10251/10712 [1:44:50<03:46,  2.03it/s]
 96%|█████████▌| 10252/10712 [1:44:51<03:45,  2.04it/s]
 96%|█████████▌| 10253/10712 [1:44:51<03:45,  2.04it/s]
 96%|█████████▌| 10254/10712 [1:44:52<03:44,  2.04it/s]
 96%|█████████▌| 10255/10712 [1:44:52<03:44,  2.04it/s]
 96%|█████████▌| 10256/10712 [1:44:53<03:43,  2.04it/s]
 96%|█████████▌| 10257/10712 [1:44:53<03:43,  2.04it/s]
 96%|█████████▌| 10258/10712 [1:44:54<03:42,  2.04it/s]
 96%|█████████▌| 10259/10712 [1:44:54<03:41,  2.04it/s]
 96%|█████████▌| 10260/10712 [1:44:55<03:41,  2.04it/s]
 96%|█████████▌| 10261/10712 [1:44:55<03:41,  2.04it/s]
 96%|█████████▌| 10262/10712 [1:44:56<03:40,  2.04it/s]
 96%|█████████▌| 10263/10712 [1:44:56<03:40,  2.04it/s]
 96%|█████████▌| 10264/10712 [1:44:57<03:39,  2.04it/s]
 96%|█████████▌| 10265/10712 [1:44:57<03:39,  2.04it/s]
 96%|█████████▌| 10266/10712 [1:44:58<03:38,  2.04it/s]
 96%|█████████▌| 10267/10712 [1:44:58<03:38,  2.04it/s]
 96%|█████████▌| 10268/10712 [1:44:59<03:37,  2.04it/s]
 96%|█████████▌| 10269/10712 [1:44:59<03:37,  2.04it/s]
 96%|█████████▌| 10270/10712 [1:45:00<03:36,  2.04it/s]
 96%|█████████▌| 10271/10712 [1:45:00<03:35,  2.04it/s]
 96%|█████████▌| 10272/10712 [1:45:01<03:35,  2.04it/s]
 96%|█████████▌| 10273/10712 [1:45:01<03:35,  2.04it/s]
 96%|█████████▌| 10274/10712 [1:45:02<03:34,  2.04it/s]
 96%|█████████▌| 10275/10712 [1:45:02<03:34,  2.04it/s]{'loss': 3.597, 'grad_norm': 0.19683994352817535, 'learning_rate': 5.0619098904411964e-06, 'epoch': 0.96}
                                                       

 96%|█████████▌| 10275/10712 [1:45:02<03:34,  2.04it/s]
 96%|█████████▌| 10276/10712 [1:45:03<03:33,  2.04it/s]
 96%|█████████▌| 10277/10712 [1:45:03<03:33,  2.04it/s]
 96%|█████████▌| 10278/10712 [1:45:04<03:32,  2.04it/s]
 96%|█████████▌| 10279/10712 [1:45:04<03:32,  2.04it/s]
 96%|█████████▌| 10280/10712 [1:45:05<03:31,  2.04it/s]
 96%|█████████▌| 10281/10712 [1:45:05<03:31,  2.04it/s]
 96%|█████████▌| 10282/10712 [1:45:06<03:30,  2.04it/s]
 96%|█████████▌| 10283/10712 [1:45:06<04:04,  1.76it/s]
 96%|█████████▌| 10284/10712 [1:45:07<03:53,  1.83it/s]
 96%|█████████▌| 10285/10712 [1:45:07<03:45,  1.89it/s]
 96%|█████████▌| 10286/10712 [1:45:08<03:40,  1.93it/s]
 96%|█████████▌| 10287/10712 [1:45:08<03:36,  1.96it/s]
 96%|█████████▌| 10288/10712 [1:45:09<03:33,  1.99it/s]
 96%|█████████▌| 10289/10712 [1:45:09<03:31,  2.00it/s]
 96%|█████████▌| 10290/10712 [1:45:10<03:29,  2.01it/s]
 96%|█████████▌| 10291/10712 [1:45:10<03:28,  2.02it/s]
 96%|█████████▌| 10292/10712 [1:45:11<03:27,  2.03it/s]
 96%|█████████▌| 10293/10712 [1:45:11<03:26,  2.03it/s]
 96%|█████████▌| 10294/10712 [1:45:12<03:25,  2.03it/s]
 96%|█████████▌| 10295/10712 [1:45:12<03:24,  2.04it/s]
 96%|█████████▌| 10296/10712 [1:45:13<03:23,  2.04it/s]
 96%|█████████▌| 10297/10712 [1:45:13<03:23,  2.04it/s]
 96%|█████████▌| 10298/10712 [1:45:14<03:22,  2.04it/s]
 96%|█████████▌| 10299/10712 [1:45:14<03:22,  2.04it/s]
 96%|█████████▌| 10300/10712 [1:45:15<03:21,  2.04it/s]{'loss': 3.5992, 'grad_norm': 0.1958075314760208, 'learning_rate': 4.500156075844919e-06, 'epoch': 0.96}
                                                       

 96%|█████████▌| 10300/10712 [1:45:15<03:21,  2.04it/s]
 96%|█████████▌| 10301/10712 [1:45:15<03:21,  2.04it/s]
 96%|█████████▌| 10302/10712 [1:45:16<03:21,  2.04it/s]
 96%|█████████▌| 10303/10712 [1:45:16<03:20,  2.04it/s]
 96%|█████████▌| 10304/10712 [1:45:17<03:19,  2.04it/s]
 96%|█████████▌| 10305/10712 [1:45:17<03:19,  2.04it/s]
 96%|█████████▌| 10306/10712 [1:45:18<03:19,  2.04it/s]
 96%|█████████▌| 10307/10712 [1:45:18<03:18,  2.04it/s]
 96%|█████████▌| 10308/10712 [1:45:19<03:17,  2.04it/s]
 96%|█████████▌| 10309/10712 [1:45:19<03:17,  2.04it/s]
 96%|█████████▌| 10310/10712 [1:45:20<03:17,  2.04it/s]
 96%|█████████▋| 10311/10712 [1:45:20<03:16,  2.04it/s]
 96%|█████████▋| 10312/10712 [1:45:21<03:15,  2.04it/s]
 96%|█████████▋| 10313/10712 [1:45:21<03:15,  2.04it/s]
 96%|█████████▋| 10314/10712 [1:45:22<03:15,  2.04it/s]
 96%|█████████▋| 10315/10712 [1:45:22<03:14,  2.04it/s]
 96%|█████████▋| 10316/10712 [1:45:23<03:13,  2.04it/s]
 96%|█████████▋| 10317/10712 [1:45:23<03:13,  2.04it/s]
 96%|█████████▋| 10318/10712 [1:45:24<03:13,  2.04it/s]
 96%|█████████▋| 10319/10712 [1:45:24<03:12,  2.04it/s]
 96%|█████████▋| 10320/10712 [1:45:24<03:12,  2.04it/s]
 96%|█████████▋| 10321/10712 [1:45:25<03:11,  2.04it/s]
 96%|█████████▋| 10322/10712 [1:45:25<03:11,  2.04it/s]
 96%|█████████▋| 10323/10712 [1:45:26<03:10,  2.04it/s]
 96%|█████████▋| 10324/10712 [1:45:26<03:10,  2.04it/s]
 96%|█████████▋| 10325/10712 [1:45:27<03:09,  2.04it/s]
                                                       
{'loss': 3.598, 'grad_norm': 0.19910714030265808, 'learning_rate': 3.971292483989309e-06, 'epoch': 0.96}

 96%|█████████▋| 10325/10712 [1:45:27<03:09,  2.04it/s]
 96%|█████████▋| 10326/10712 [1:45:27<03:09,  2.04it/s]
 96%|█████████▋| 10327/10712 [1:45:28<03:08,  2.04it/s]
 96%|█████████▋| 10328/10712 [1:45:28<03:08,  2.04it/s]
 96%|█████████▋| 10329/10712 [1:45:29<03:07,  2.04it/s]
 96%|█████████▋| 10330/10712 [1:45:29<03:07,  2.04it/s]
 96%|█████████▋| 10331/10712 [1:45:30<03:06,  2.04it/s]
 96%|█████████▋| 10332/10712 [1:45:30<03:06,  2.04it/s]
 96%|█████████▋| 10333/10712 [1:45:31<03:05,  2.04it/s]
 96%|█████████▋| 10334/10712 [1:45:31<03:05,  2.04it/s]
 96%|█████████▋| 10335/10712 [1:45:32<03:04,  2.04it/s]
 96%|█████████▋| 10336/10712 [1:45:32<03:04,  2.04it/s]
 96%|█████████▋| 10337/10712 [1:45:33<03:03,  2.04it/s]
 97%|█████████▋| 10338/10712 [1:45:33<03:03,  2.04it/s]
 97%|█████████▋| 10339/10712 [1:45:34<03:02,  2.04it/s]
 97%|█████████▋| 10340/10712 [1:45:34<03:02,  2.04it/s]
 97%|█████████▋| 10341/10712 [1:45:35<03:01,  2.04it/s]
 97%|█████████▋| 10342/10712 [1:45:35<03:01,  2.04it/s]
 97%|█████████▋| 10343/10712 [1:45:36<03:00,  2.04it/s]
 97%|█████████▋| 10344/10712 [1:45:36<03:00,  2.04it/s]
 97%|█████████▋| 10345/10712 [1:45:37<02:59,  2.04it/s]
 97%|█████████▋| 10346/10712 [1:45:37<02:59,  2.04it/s]
 97%|█████████▋| 10347/10712 [1:45:38<02:58,  2.04it/s]
 97%|█████████▋| 10348/10712 [1:45:38<02:58,  2.04it/s]
 97%|█████████▋| 10349/10712 [1:45:39<02:57,  2.04it/s]
 97%|█████████▋| 10350/10712 [1:45:39<02:57,  2.04it/s]{'loss': 3.6001, 'grad_norm': 0.195979043841362, 'learning_rate': 3.4753542197114575e-06, 'epoch': 0.97}
                                                       

 97%|█████████▋| 10350/10712 [1:45:39<02:57,  2.04it/s]
 97%|█████████▋| 10351/10712 [1:45:40<02:57,  2.04it/s]
 97%|█████████▋| 10352/10712 [1:45:40<02:56,  2.04it/s]
 97%|█████████▋| 10353/10712 [1:45:41<02:56,  2.04it/s]
 97%|█████████▋| 10354/10712 [1:45:41<02:55,  2.04it/s]
 97%|█████████▋| 10355/10712 [1:45:42<02:54,  2.04it/s]
 97%|█████████▋| 10356/10712 [1:45:42<02:54,  2.04it/s]
 97%|█████████▋| 10357/10712 [1:45:43<02:53,  2.04it/s]
 97%|█████████▋| 10358/10712 [1:45:43<02:53,  2.04it/s]
 97%|█████████▋| 10359/10712 [1:45:44<02:53,  2.04it/s]
 97%|█████████▋| 10360/10712 [1:45:44<02:52,  2.04it/s]
 97%|█████████▋| 10361/10712 [1:45:45<02:52,  2.04it/s]
 97%|█████████▋| 10362/10712 [1:45:45<02:51,  2.04it/s]
 97%|█████████▋| 10363/10712 [1:45:46<02:51,  2.04it/s]
 97%|█████████▋| 10364/10712 [1:45:46<02:50,  2.04it/s]
 97%|█████████▋| 10365/10712 [1:45:47<02:50,  2.04it/s]
 97%|█████████▋| 10366/10712 [1:45:47<02:49,  2.04it/s]
 97%|█████████▋| 10367/10712 [1:45:48<02:49,  2.04it/s]
 97%|█████████▋| 10368/10712 [1:45:48<02:48,  2.04it/s]
 97%|█████████▋| 10369/10712 [1:45:49<02:48,  2.04it/s]
 97%|█████████▋| 10370/10712 [1:45:49<02:47,  2.04it/s]
 97%|█████████▋| 10371/10712 [1:45:49<02:47,  2.04it/s]
 97%|█████████▋| 10372/10712 [1:45:50<02:46,  2.04it/s]
 97%|█████████▋| 10373/10712 [1:45:50<02:46,  2.04it/s]
 97%|█████████▋| 10374/10712 [1:45:51<02:45,  2.04it/s]
 97%|█████████▋| 10375/10712 [1:45:51<02:44,  2.04it/s]{'loss': 3.6016, 'grad_norm': 0.19631534814834595, 'learning_rate': 3.0123742023356325e-06, 'epoch': 0.97}

                                                       

 97%|█████████▋| 10375/10712 [1:45:51<02:44,  2.04it/s]
 97%|█████████▋| 10376/10712 [1:45:52<02:44,  2.04it/s]
 97%|█████████▋| 10377/10712 [1:45:52<02:44,  2.04it/s]
 97%|█████████▋| 10378/10712 [1:45:53<02:43,  2.04it/s]
 97%|█████████▋| 10379/10712 [1:45:53<02:43,  2.04it/s]
 97%|█████████▋| 10380/10712 [1:45:54<02:42,  2.04it/s]
 97%|█████████▋| 10381/10712 [1:45:54<02:42,  2.04it/s]
 97%|█████████▋| 10382/10712 [1:45:55<02:41,  2.04it/s]
 97%|█████████▋| 10383/10712 [1:45:55<02:41,  2.04it/s]
 97%|█████████▋| 10384/10712 [1:45:56<02:40,  2.04it/s]
 97%|█████████▋| 10385/10712 [1:45:56<02:40,  2.04it/s]
 97%|█████████▋| 10386/10712 [1:45:57<02:39,  2.04it/s]
 97%|█████████▋| 10387/10712 [1:45:57<02:39,  2.04it/s]
 97%|█████████▋| 10388/10712 [1:45:58<02:38,  2.04it/s]
 97%|█████████▋| 10389/10712 [1:45:58<02:38,  2.04it/s]
 97%|█████████▋| 10390/10712 [1:45:59<02:37,  2.04it/s]
 97%|█████████▋| 10391/10712 [1:45:59<02:37,  2.04it/s]
 97%|█████████▋| 10392/10712 [1:46:00<02:36,  2.04it/s]
 97%|█████████▋| 10393/10712 [1:46:00<02:36,  2.04it/s]
 97%|█████████▋| 10394/10712 [1:46:01<02:35,  2.04it/s]
 97%|█████████▋| 10395/10712 [1:46:01<02:35,  2.04it/s]
 97%|█████████▋| 10396/10712 [1:46:02<02:34,  2.04it/s]
 97%|█████████▋| 10397/10712 [1:46:02<02:34,  2.04it/s]
 97%|█████████▋| 10398/10712 [1:46:03<02:34,  2.04it/s]
 97%|█████████▋| 10399/10712 [1:46:03<02:33,  2.04it/s]
 97%|█████████▋| 10400/10712 [1:46:04<02:32,  2.04it/s]{'loss': 3.5983, 'grad_norm': 0.19586975872516632, 'learning_rate': 2.582383163487745e-06, 'epoch': 0.97}

                                                       

 97%|█████████▋| 10400/10712 [1:46:04<02:32,  2.04it/s]
 97%|█████████▋| 10401/10712 [1:46:04<02:32,  2.04it/s]
 97%|█████████▋| 10402/10712 [1:46:05<02:32,  2.04it/s]
 97%|█████████▋| 10403/10712 [1:46:05<02:31,  2.04it/s]
 97%|█████████▋| 10404/10712 [1:46:06<02:30,  2.04it/s]
 97%|█████████▋| 10405/10712 [1:46:06<02:30,  2.04it/s]
 97%|█████████▋| 10406/10712 [1:46:07<02:30,  2.04it/s]
 97%|█████████▋| 10407/10712 [1:46:07<02:29,  2.04it/s]
 97%|█████████▋| 10408/10712 [1:46:08<02:29,  2.04it/s]
 97%|█████████▋| 10409/10712 [1:46:08<02:28,  2.04it/s]
 97%|█████████▋| 10410/10712 [1:46:09<02:28,  2.04it/s]
 97%|█████████▋| 10411/10712 [1:46:09<02:27,  2.04it/s]
 97%|█████████▋| 10412/10712 [1:46:10<02:27,  2.04it/s]
 97%|█████████▋| 10413/10712 [1:46:10<02:26,  2.04it/s]
 97%|█████████▋| 10414/10712 [1:46:11<02:26,  2.04it/s]
 97%|█████████▋| 10415/10712 [1:46:11<02:25,  2.04it/s]
 97%|█████████▋| 10416/10712 [1:46:12<02:25,  2.04it/s]
 97%|█████████▋| 10417/10712 [1:46:12<02:24,  2.04it/s]
 97%|█████████▋| 10418/10712 [1:46:13<02:24,  2.04it/s]
 97%|█████████▋| 10419/10712 [1:46:13<02:23,  2.04it/s]
 97%|█████████▋| 10420/10712 [1:46:14<02:23,  2.04it/s]
 97%|█████████▋| 10421/10712 [1:46:14<02:22,  2.04it/s]
 97%|█████████▋| 10422/10712 [1:46:14<02:22,  2.04it/s]
 97%|█████████▋| 10423/10712 [1:46:15<02:21,  2.04it/s]
 97%|█████████▋| 10424/10712 [1:46:15<02:21,  2.04it/s]
 97%|█████████▋| 10425/10712 [1:46:16<02:20,  2.04it/s]{'loss': 3.5902, 'grad_norm': 0.19625885784626007, 'learning_rate': 2.1854096450558136e-06, 'epoch': 0.97}

                                                       

 97%|█████████▋| 10425/10712 [1:46:16<02:20,  2.04it/s]
 97%|█████████▋| 10426/10712 [1:46:16<02:20,  2.04it/s]
 97%|█████████▋| 10427/10712 [1:46:17<02:19,  2.04it/s]
 97%|█████████▋| 10428/10712 [1:46:17<02:19,  2.04it/s]
 97%|█████████▋| 10429/10712 [1:46:18<02:18,  2.04it/s]
 97%|█████████▋| 10430/10712 [1:46:18<02:18,  2.04it/s]
 97%|█████████▋| 10431/10712 [1:46:19<02:17,  2.04it/s]
 97%|█████████▋| 10432/10712 [1:46:19<02:17,  2.04it/s]
 97%|█████████▋| 10433/10712 [1:46:20<02:16,  2.04it/s]
 97%|█████████▋| 10434/10712 [1:46:20<02:16,  2.04it/s]
 97%|█████████▋| 10435/10712 [1:46:21<02:15,  2.04it/s]
 97%|█████████▋| 10436/10712 [1:46:21<02:15,  2.04it/s]
 97%|█████████▋| 10437/10712 [1:46:22<02:14,  2.04it/s]
 97%|█████████▋| 10438/10712 [1:46:22<02:14,  2.04it/s]
 97%|█████████▋| 10439/10712 [1:46:23<02:13,  2.04it/s]
 97%|█████████▋| 10440/10712 [1:46:23<02:13,  2.04it/s]
 97%|█████████▋| 10441/10712 [1:46:24<02:12,  2.04it/s]
 97%|█████████▋| 10442/10712 [1:46:24<02:12,  2.04it/s]
 97%|█████████▋| 10443/10712 [1:46:25<02:12,  2.04it/s]
 97%|█████████▋| 10444/10712 [1:46:25<02:11,  2.04it/s]
 98%|█████████▊| 10445/10712 [1:46:26<02:10,  2.04it/s]
 98%|█████████▊| 10446/10712 [1:46:26<02:10,  2.04it/s]
 98%|█████████▊| 10447/10712 [1:46:27<02:10,  2.04it/s]
 98%|█████████▊| 10448/10712 [1:46:27<02:09,  2.04it/s]
 98%|█████████▊| 10449/10712 [1:46:28<02:08,  2.04it/s]
 98%|█████████▊| 10450/10712 [1:46:28<02:08,  2.04it/s]{'loss': 3.5976, 'grad_norm': 0.19550135731697083, 'learning_rate': 1.8214799972953166e-06, 'epoch': 0.98}

                                                       

 98%|█████████▊| 10450/10712 [1:46:28<02:08,  2.04it/s]
 98%|█████████▊| 10451/10712 [1:46:29<02:08,  2.04it/s]
 98%|█████████▊| 10452/10712 [1:46:29<02:07,  2.04it/s]
 98%|█████████▊| 10453/10712 [1:46:30<02:07,  2.04it/s]
 98%|█████████▊| 10454/10712 [1:46:30<02:06,  2.04it/s]
 98%|█████████▊| 10455/10712 [1:46:31<02:06,  2.04it/s]
 98%|█████████▊| 10456/10712 [1:46:31<02:05,  2.04it/s]
 98%|█████████▊| 10457/10712 [1:46:32<02:05,  2.04it/s]
 98%|█████████▊| 10458/10712 [1:46:32<02:04,  2.04it/s]
 98%|█████████▊| 10459/10712 [1:46:33<02:04,  2.04it/s]
 98%|█████████▊| 10460/10712 [1:46:33<02:03,  2.04it/s]
 98%|█████████▊| 10461/10712 [1:46:34<02:03,  2.04it/s]
 98%|█████████▊| 10462/10712 [1:46:34<02:02,  2.04it/s]
 98%|█████████▊| 10463/10712 [1:46:35<02:02,  2.04it/s]
 98%|█████████▊| 10464/10712 [1:46:35<02:01,  2.04it/s]
 98%|█████████▊| 10465/10712 [1:46:36<02:01,  2.04it/s]
 98%|█████████▊| 10466/10712 [1:46:36<02:00,  2.04it/s]
 98%|█████████▊| 10467/10712 [1:46:37<01:59,  2.04it/s]
 98%|█████████▊| 10468/10712 [1:46:37<01:59,  2.04it/s]
 98%|█████████▊| 10469/10712 [1:46:38<01:59,  2.04it/s]
 98%|█████████▊| 10470/10712 [1:46:38<01:58,  2.04it/s]
 98%|█████████▊| 10471/10712 [1:46:39<01:57,  2.04it/s]
 98%|█████████▊| 10472/10712 [1:46:39<01:57,  2.04it/s]
 98%|█████████▊| 10473/10712 [1:46:39<01:57,  2.04it/s]
 98%|█████████▊| 10474/10712 [1:46:40<01:56,  2.04it/s]
 98%|█████████▊| 10475/10712 [1:46:40<01:56,  2.04it/s]{'loss': 3.6031, 'grad_norm': 0.19436602294445038, 'learning_rate': 1.490618377079922e-06, 'epoch': 0.98}

                                                       

 98%|█████████▊| 10475/10712 [1:46:40<01:56,  2.04it/s]
 98%|█████████▊| 10476/10712 [1:46:41<01:55,  2.04it/s]
 98%|█████████▊| 10477/10712 [1:46:41<01:55,  2.04it/s]
 98%|█████████▊| 10478/10712 [1:46:42<01:54,  2.04it/s]
 98%|█████████▊| 10479/10712 [1:46:42<01:54,  2.04it/s]
 98%|█████████▊| 10480/10712 [1:46:43<01:53,  2.04it/s]
 98%|█████████▊| 10481/10712 [1:46:43<01:53,  2.04it/s]
 98%|█████████▊| 10482/10712 [1:46:44<01:52,  2.04it/s]
 98%|█████████▊| 10483/10712 [1:46:44<01:52,  2.04it/s]
 98%|█████████▊| 10484/10712 [1:46:45<01:51,  2.04it/s]
 98%|█████████▊| 10485/10712 [1:46:45<01:51,  2.04it/s]
 98%|█████████▊| 10486/10712 [1:46:46<01:50,  2.04it/s]
 98%|█████████▊| 10487/10712 [1:46:46<01:50,  2.04it/s]
 98%|█████████▊| 10488/10712 [1:46:47<01:49,  2.04it/s]
 98%|█████████▊| 10489/10712 [1:46:47<01:49,  2.04it/s]
 98%|█████████▊| 10490/10712 [1:46:48<01:48,  2.04it/s]
 98%|█████████▊| 10491/10712 [1:46:48<01:48,  2.04it/s]
 98%|█████████▊| 10492/10712 [1:46:49<01:47,  2.04it/s]
 98%|█████████▊| 10493/10712 [1:46:49<01:47,  2.04it/s]
 98%|█████████▊| 10494/10712 [1:46:50<01:46,  2.04it/s]
 98%|█████████▊| 10495/10712 [1:46:50<01:46,  2.04it/s]
 98%|█████████▊| 10496/10712 [1:46:51<01:45,  2.04it/s]
 98%|█████████▊| 10497/10712 [1:46:51<01:45,  2.04it/s]
 98%|█████████▊| 10498/10712 [1:46:52<01:44,  2.04it/s]
 98%|█████████▊| 10499/10712 [1:46:52<01:44,  2.04it/s]
 98%|█████████▊| 10500/10712 [1:46:53<01:43,  2.04it/s]
                                                       
{'loss': 3.5983, 'grad_norm': 0.1943190097808838, 'learning_rate': 1.1928467462982706e-06, 'epoch': 0.98}

 98%|█████████▊| 10500/10712 [1:46:53<01:43,  2.04it/s]
 98%|█████████▊| 10501/10712 [1:46:53<01:43,  2.04it/s]
 98%|█████████▊| 10502/10712 [1:46:54<01:42,  2.04it/s]
 98%|█████████▊| 10503/10712 [1:46:54<01:42,  2.04it/s]
 98%|█████████▊| 10504/10712 [1:46:55<01:41,  2.04it/s]
 98%|█████████▊| 10505/10712 [1:46:55<01:41,  2.04it/s]
 98%|█████████▊| 10506/10712 [1:46:56<01:41,  2.04it/s]
 98%|█████████▊| 10507/10712 [1:46:56<01:40,  2.04it/s]
 98%|█████████▊| 10508/10712 [1:46:57<01:40,  2.04it/s]
 98%|█████████▊| 10509/10712 [1:46:57<01:39,  2.04it/s]
 98%|█████████▊| 10510/10712 [1:46:58<01:38,  2.04it/s]
 98%|█████████▊| 10511/10712 [1:46:58<01:38,  2.04it/s]
 98%|█████████▊| 10512/10712 [1:46:59<01:38,  2.04it/s]
 98%|█████████▊| 10513/10712 [1:46:59<01:37,  2.04it/s]
 98%|█████████▊| 10514/10712 [1:47:00<01:36,  2.04it/s]
 98%|█████████▊| 10515/10712 [1:47:00<01:36,  2.04it/s]
 98%|█████████▊| 10516/10712 [1:47:01<01:36,  2.04it/s]
 98%|█████████▊| 10517/10712 [1:47:01<01:35,  2.04it/s]
 98%|█████████▊| 10518/10712 [1:47:02<01:35,  2.04it/s]
 98%|█████████▊| 10519/10712 [1:47:02<01:34,  2.04it/s]
 98%|█████████▊| 10520/10712 [1:47:03<01:34,  2.04it/s]
 98%|█████████▊| 10521/10712 [1:47:03<01:33,  2.04it/s]
 98%|█████████▊| 10522/10712 [1:47:04<01:33,  2.04it/s]
 98%|█████████▊| 10523/10712 [1:47:04<01:32,  2.04it/s]
 98%|█████████▊| 10524/10712 [1:47:05<01:32,  2.04it/s]
 98%|█████████▊| 10525/10712 [1:47:05<01:31,  2.04it/s]{'loss': 3.6074, 'grad_norm': 0.19465430080890656, 'learning_rate': 9.281848703961426e-07, 'epoch': 0.98}
                                                       

 98%|█████████▊| 10525/10712 [1:47:05<01:31,  2.04it/s]
 98%|█████████▊| 10526/10712 [1:47:05<01:31,  2.04it/s]
 98%|█████████▊| 10527/10712 [1:47:06<01:30,  2.04it/s]
 98%|█████████▊| 10528/10712 [1:47:06<01:30,  2.04it/s]
 98%|█████████▊| 10529/10712 [1:47:07<01:29,  2.04it/s]
 98%|█████████▊| 10530/10712 [1:47:07<01:29,  2.04it/s]
 98%|█████████▊| 10531/10712 [1:47:08<01:28,  2.04it/s]
 98%|█████████▊| 10532/10712 [1:47:08<01:28,  2.04it/s]
 98%|█████████▊| 10533/10712 [1:47:09<01:27,  2.04it/s]
 98%|█████████▊| 10534/10712 [1:47:09<01:27,  2.04it/s]
 98%|█████████▊| 10535/10712 [1:47:10<01:26,  2.04it/s]
 98%|█████████▊| 10536/10712 [1:47:10<01:26,  2.04it/s]
 98%|█████████▊| 10537/10712 [1:47:11<01:25,  2.04it/s]
 98%|█████████▊| 10538/10712 [1:47:11<01:25,  2.04it/s]
 98%|█████████▊| 10539/10712 [1:47:12<01:24,  2.04it/s]
 98%|█████████▊| 10540/10712 [1:47:12<01:24,  2.04it/s]
 98%|█████████▊| 10541/10712 [1:47:13<01:23,  2.04it/s]
 98%|█████████▊| 10542/10712 [1:47:13<01:23,  2.04it/s]
 98%|█████████▊| 10543/10712 [1:47:14<01:22,  2.04it/s]
 98%|█████████▊| 10544/10712 [1:47:14<01:22,  2.04it/s]
 98%|█████████▊| 10545/10712 [1:47:15<01:21,  2.04it/s]
 98%|█████████▊| 10546/10712 [1:47:15<01:21,  2.04it/s]
 98%|█████████▊| 10547/10712 [1:47:16<01:20,  2.04it/s]
 98%|█████████▊| 10548/10712 [1:47:16<01:20,  2.04it/s]
 98%|█████████▊| 10549/10712 [1:47:17<01:19,  2.04it/s]
 98%|█████████▊| 10550/10712 [1:47:17<01:19,  2.04it/s]{'loss': 3.5994, 'grad_norm': 0.19594857096672058, 'learning_rate': 6.966503170642292e-07, 'epoch': 0.98}

                                                       

 98%|█████████▊| 10550/10712 [1:47:17<01:19,  2.04it/s]
 98%|█████████▊| 10551/10712 [1:47:18<01:18,  2.04it/s]
 99%|█████████▊| 10552/10712 [1:47:18<01:18,  2.04it/s]
 99%|█████████▊| 10553/10712 [1:47:19<01:17,  2.04it/s]
 99%|█████████▊| 10554/10712 [1:47:19<01:17,  2.04it/s]
 99%|█████████▊| 10555/10712 [1:47:20<01:16,  2.04it/s]
 99%|█████████▊| 10556/10712 [1:47:20<01:16,  2.04it/s]
 99%|█████████▊| 10557/10712 [1:47:21<01:15,  2.04it/s]
 99%|█████████▊| 10558/10712 [1:47:21<01:15,  2.04it/s]
 99%|█████████▊| 10559/10712 [1:47:22<01:14,  2.04it/s]
 99%|█████████▊| 10560/10712 [1:47:22<01:14,  2.04it/s]
 99%|█████████▊| 10561/10712 [1:47:23<01:13,  2.04it/s]
 99%|█████████▊| 10562/10712 [1:47:23<01:13,  2.04it/s]
 99%|█████████▊| 10563/10712 [1:47:24<01:12,  2.04it/s]
 99%|█████████▊| 10564/10712 [1:47:24<01:12,  2.04it/s]
 99%|█████████▊| 10565/10712 [1:47:25<01:12,  2.04it/s]
 99%|█████████▊| 10566/10712 [1:47:25<01:11,  2.04it/s]
 99%|█████████▊| 10567/10712 [1:47:26<01:11,  2.04it/s]
 99%|█████████▊| 10568/10712 [1:47:26<01:10,  2.04it/s]
 99%|█████████▊| 10569/10712 [1:47:27<01:10,  2.04it/s]
 99%|█████████▊| 10570/10712 [1:47:27<01:09,  2.04it/s]
 99%|█████████▊| 10571/10712 [1:47:28<01:09,  2.04it/s]
 99%|█████████▊| 10572/10712 [1:47:28<01:08,  2.04it/s]
 99%|█████████▊| 10573/10712 [1:47:29<01:08,  2.04it/s]
 99%|█████████▊| 10574/10712 [1:47:29<01:07,  2.04it/s]
 99%|█████████▊| 10575/10712 [1:47:29<01:07,  2.04it/s]{'loss': 3.5973, 'grad_norm': 0.19539114832878113, 'learning_rate': 4.982584550723978e-07, 'epoch': 0.99}
                                                       

 99%|█████████▊| 10575/10712 [1:47:30<01:07,  2.04it/s]
 99%|█████████▊| 10576/10712 [1:47:30<01:06,  2.04it/s]
 99%|█████████▊| 10577/10712 [1:47:30<01:06,  2.04it/s]
 99%|█████████▊| 10578/10712 [1:47:31<01:05,  2.04it/s]
 99%|█████████▉| 10579/10712 [1:47:31<01:05,  2.04it/s]
 99%|█████████▉| 10580/10712 [1:47:32<01:04,  2.04it/s]
 99%|█████████▉| 10581/10712 [1:47:32<01:04,  2.04it/s]
 99%|█████████▉| 10582/10712 [1:47:33<01:03,  2.04it/s]
 99%|█████████▉| 10583/10712 [1:47:33<01:03,  2.04it/s]
 99%|█████████▉| 10584/10712 [1:47:34<01:02,  2.04it/s]
 99%|█████████▉| 10585/10712 [1:47:34<01:02,  2.04it/s]
 99%|█████████▉| 10586/10712 [1:47:35<01:01,  2.04it/s]
 99%|█████████▉| 10587/10712 [1:47:35<01:01,  2.04it/s]
 99%|█████████▉| 10588/10712 [1:47:36<01:00,  2.04it/s]
 99%|█████████▉| 10589/10712 [1:47:36<01:00,  2.04it/s]
 99%|█████████▉| 10590/10712 [1:47:37<00:59,  2.04it/s]
 99%|█████████▉| 10591/10712 [1:47:37<00:59,  2.04it/s]
 99%|█████████▉| 10592/10712 [1:47:38<00:58,  2.04it/s]
 99%|█████████▉| 10593/10712 [1:47:38<00:58,  2.04it/s]
 99%|█████████▉| 10594/10712 [1:47:39<00:57,  2.04it/s]
 99%|█████████▉| 10595/10712 [1:47:39<00:57,  2.04it/s]
 99%|█████████▉| 10596/10712 [1:47:40<00:56,  2.04it/s]
 99%|█████████▉| 10597/10712 [1:47:40<00:56,  2.04it/s]
 99%|█████████▉| 10598/10712 [1:47:41<00:55,  2.04it/s]
 99%|█████████▉| 10599/10712 [1:47:41<00:55,  2.04it/s]
 99%|█████████▉| 10600/10712 [1:47:42<00:54,  2.04it/s]{'loss': 3.6063, 'grad_norm': 0.19407439231872559, 'learning_rate': 3.3302245324928757e-07, 'epoch': 0.99}
                                                       

 99%|█████████▉| 10600/10712 [1:47:42<00:54,  2.04it/s]
 99%|█████████▉| 10601/10712 [1:47:42<00:54,  2.04it/s]
 99%|█████████▉| 10602/10712 [1:47:43<00:53,  2.04it/s]
 99%|█████████▉| 10603/10712 [1:47:43<00:53,  2.04it/s]
 99%|█████████▉| 10604/10712 [1:47:44<00:52,  2.04it/s]
 99%|█████████▉| 10605/10712 [1:47:44<00:52,  2.04it/s]
 99%|█████████▉| 10606/10712 [1:47:45<00:52,  2.04it/s]
 99%|█████████▉| 10607/10712 [1:47:45<00:51,  2.04it/s]
 99%|█████████▉| 10608/10712 [1:47:46<00:51,  2.04it/s]
 99%|█████████▉| 10609/10712 [1:47:46<00:50,  2.04it/s]
 99%|█████████▉| 10610/10712 [1:47:47<00:50,  2.04it/s]
 99%|█████████▉| 10611/10712 [1:47:47<00:49,  2.04it/s]
 99%|█████████▉| 10612/10712 [1:47:48<00:49,  2.04it/s]
 99%|█████████▉| 10613/10712 [1:47:48<00:48,  2.04it/s]
 99%|█████████▉| 10614/10712 [1:47:49<00:48,  2.04it/s]
 99%|█████████▉| 10615/10712 [1:47:49<00:47,  2.04it/s]
 99%|█████████▉| 10616/10712 [1:47:50<00:47,  2.04it/s]
 99%|█████████▉| 10617/10712 [1:47:50<00:46,  2.04it/s]
 99%|█████████▉| 10618/10712 [1:47:51<00:46,  2.04it/s]
 99%|█████████▉| 10619/10712 [1:47:51<00:45,  2.04it/s]
 99%|█████████▉| 10620/10712 [1:47:52<00:45,  2.04it/s]
 99%|█████████▉| 10621/10712 [1:47:52<00:44,  2.04it/s]
 99%|█████████▉| 10622/10712 [1:47:53<00:44,  2.04it/s]
 99%|█████████▉| 10623/10712 [1:47:53<00:43,  2.04it/s]
 99%|█████████▉| 10624/10712 [1:47:54<00:43,  2.04it/s]
 99%|█████████▉| 10625/10712 [1:47:54<00:42,  2.04it/s]
                                                       
{'loss': 3.5918, 'grad_norm': 0.19808201491832733, 'learning_rate': 2.0095327960817367e-07, 'epoch': 0.99}

 99%|█████████▉| 10625/10712 [1:47:54<00:42,  2.04it/s]
 99%|█████████▉| 10626/10712 [1:47:55<00:42,  2.03it/s]
 99%|█████████▉| 10627/10712 [1:47:55<00:41,  2.04it/s]
 99%|█████████▉| 10628/10712 [1:47:55<00:41,  2.04it/s]
 99%|█████████▉| 10629/10712 [1:47:56<00:40,  2.04it/s]
 99%|█████████▉| 10630/10712 [1:47:56<00:40,  2.04it/s]
 99%|█████████▉| 10631/10712 [1:47:57<00:39,  2.04it/s]
 99%|█████████▉| 10632/10712 [1:47:57<00:39,  2.03it/s]
 99%|█████████▉| 10633/10712 [1:47:58<00:38,  2.04it/s]
 99%|█████████▉| 10634/10712 [1:47:58<00:38,  2.04it/s]
 99%|█████████▉| 10635/10712 [1:47:59<00:37,  2.04it/s]
 99%|█████████▉| 10636/10712 [1:47:59<00:37,  2.04it/s]
 99%|█████████▉| 10637/10712 [1:48:00<00:36,  2.04it/s]
 99%|█████████▉| 10638/10712 [1:48:00<00:36,  2.04it/s]
 99%|█████████▉| 10639/10712 [1:48:01<00:35,  2.04it/s]
 99%|█████████▉| 10640/10712 [1:48:01<00:35,  2.04it/s]
 99%|█████████▉| 10641/10712 [1:48:02<00:34,  2.04it/s]
 99%|█████████▉| 10642/10712 [1:48:02<00:34,  2.04it/s]
 99%|█████████▉| 10643/10712 [1:48:03<00:33,  2.04it/s]
 99%|█████████▉| 10644/10712 [1:48:03<00:33,  2.04it/s]
 99%|█████████▉| 10645/10712 [1:48:04<00:32,  2.04it/s]
 99%|█████████▉| 10646/10712 [1:48:04<00:32,  2.04it/s]
 99%|█████████▉| 10647/10712 [1:48:05<00:31,  2.04it/s]
 99%|█████████▉| 10648/10712 [1:48:05<00:31,  2.04it/s]
 99%|█████████▉| 10649/10712 [1:48:06<00:30,  2.04it/s]
 99%|█████████▉| 10650/10712 [1:48:06<00:30,  2.04it/s]{'loss': 3.5995, 'grad_norm': 0.19517377018928528, 'learning_rate': 1.0205970061916148e-07, 'epoch': 0.99}
                                                       

 99%|█████████▉| 10650/10712 [1:48:06<00:30,  2.04it/s]
 99%|█████████▉| 10651/10712 [1:48:07<00:29,  2.04it/s]
 99%|█████████▉| 10652/10712 [1:48:07<00:29,  2.04it/s]
 99%|█████████▉| 10653/10712 [1:48:08<00:28,  2.04it/s]
 99%|█████████▉| 10654/10712 [1:48:08<00:28,  2.04it/s]
 99%|█████████▉| 10655/10712 [1:48:09<00:27,  2.04it/s]
 99%|█████████▉| 10656/10712 [1:48:09<00:27,  2.04it/s]
 99%|█████████▉| 10657/10712 [1:48:10<00:26,  2.04it/s]
 99%|█████████▉| 10658/10712 [1:48:10<00:26,  2.04it/s]
100%|█████████▉| 10659/10712 [1:48:11<00:25,  2.04it/s]
100%|█████████▉| 10660/10712 [1:48:11<00:25,  2.04it/s]
100%|█████████▉| 10661/10712 [1:48:12<00:24,  2.04it/s]
100%|█████████▉| 10662/10712 [1:48:12<00:24,  2.04it/s]
100%|█████████▉| 10663/10712 [1:48:13<00:24,  2.04it/s]
100%|█████████▉| 10664/10712 [1:48:13<00:23,  2.04it/s]
100%|█████████▉| 10665/10712 [1:48:14<00:23,  2.04it/s]
100%|█████████▉| 10666/10712 [1:48:14<00:22,  2.04it/s]
100%|█████████▉| 10667/10712 [1:48:15<00:22,  2.04it/s]
100%|█████████▉| 10668/10712 [1:48:15<00:21,  2.04it/s]
100%|█████████▉| 10669/10712 [1:48:16<00:21,  2.04it/s]
100%|█████████▉| 10670/10712 [1:48:16<00:20,  2.04it/s]
100%|█████████▉| 10671/10712 [1:48:17<00:20,  2.04it/s]
100%|█████████▉| 10672/10712 [1:48:17<00:19,  2.04it/s]
100%|█████████▉| 10673/10712 [1:48:18<00:19,  2.04it/s]
100%|█████████▉| 10674/10712 [1:48:18<00:18,  2.04it/s]
100%|█████████▉| 10675/10712 [1:48:19<00:18,  2.04it/s]
                                                       
{'loss': 3.6005, 'grad_norm': 0.1984858363866806, 'learning_rate': 3.6348280627040987e-08, 'epoch': 1.0}

100%|█████████▉| 10675/10712 [1:48:19<00:18,  2.04it/s]
100%|█████████▉| 10676/10712 [1:48:19<00:17,  2.04it/s]
100%|█████████▉| 10677/10712 [1:48:20<00:17,  2.04it/s]
100%|█████████▉| 10678/10712 [1:48:20<00:16,  2.04it/s]
100%|█████████▉| 10679/10712 [1:48:21<00:16,  2.04it/s]
100%|█████████▉| 10680/10712 [1:48:21<00:15,  2.04it/s]
100%|█████████▉| 10681/10712 [1:48:21<00:15,  2.04it/s]
100%|█████████▉| 10682/10712 [1:48:22<00:14,  2.04it/s]
100%|█████████▉| 10683/10712 [1:48:22<00:14,  2.04it/s]
100%|█████████▉| 10684/10712 [1:48:23<00:13,  2.04it/s]
100%|█████████▉| 10685/10712 [1:48:23<00:13,  2.04it/s]
100%|█████████▉| 10686/10712 [1:48:24<00:12,  2.04it/s]
100%|█████████▉| 10687/10712 [1:48:24<00:12,  2.04it/s]
100%|█████████▉| 10688/10712 [1:48:25<00:11,  2.04it/s]
100%|█████████▉| 10689/10712 [1:48:25<00:11,  2.04it/s]
100%|█████████▉| 10690/10712 [1:48:26<00:10,  2.04it/s]
100%|█████████▉| 10691/10712 [1:48:26<00:10,  2.04it/s]
100%|█████████▉| 10692/10712 [1:48:27<00:09,  2.04it/s]
100%|█████████▉| 10693/10712 [1:48:27<00:09,  2.04it/s]
100%|█████████▉| 10694/10712 [1:48:28<00:08,  2.04it/s]
100%|█████████▉| 10695/10712 [1:48:28<00:08,  2.04it/s]
100%|█████████▉| 10696/10712 [1:48:29<00:07,  2.04it/s]
100%|█████████▉| 10697/10712 [1:48:29<00:07,  2.04it/s]
100%|█████████▉| 10698/10712 [1:48:30<00:06,  2.04it/s]
100%|█████████▉| 10699/10712 [1:48:30<00:06,  2.04it/s]
100%|█████████▉| 10700/10712 [1:48:31<00:05,  2.04it/s]
                                                       
{'loss': 3.5977, 'grad_norm': 0.19480589032173157, 'learning_rate': 3.823381415801741e-09, 'epoch': 1.0}

100%|█████████▉| 10700/10712 [1:48:31<00:05,  2.04it/s]
100%|█████████▉| 10701/10712 [1:48:31<00:05,  2.04it/s]
100%|█████████▉| 10702/10712 [1:48:32<00:04,  2.04it/s]
100%|█████████▉| 10703/10712 [1:48:32<00:04,  2.04it/s]
100%|█████████▉| 10704/10712 [1:48:33<00:03,  2.04it/s]
100%|█████████▉| 10705/10712 [1:48:33<00:03,  2.04it/s]
100%|█████████▉| 10706/10712 [1:48:34<00:02,  2.04it/s]
100%|█████████▉| 10707/10712 [1:48:34<00:02,  2.04it/s]
100%|█████████▉| 10708/10712 [1:48:35<00:01,  2.04it/s]
100%|█████████▉| 10709/10712 [1:48:35<00:01,  2.04it/s]
100%|█████████▉| 10710/10712 [1:48:36<00:00,  2.04it/s]
100%|█████████▉| 10711/10712 [1:48:36<00:00,  2.04it/s]
100%|██████████| 10712/10712 [1:48:38<00:00,  1.35it/s]
                                                       
{'train_runtime': 6535.7777, 'train_samples_per_second': 1678.278, 'train_steps_per_second': 1.639, 'train_loss': 3.98263484350864, 'epoch': 1.0}

100%|██████████| 10712/10712 [1:48:55<00:00,  1.35it/s]
100%|██████████| 10712/10712 [1:48:55<00:00,  1.64it/s]
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.