Commit
•
4313d2c
1
Parent(s):
f2ab1f3
Model save
Browse files- README.md +57 -0
- all_results.json +9 -0
- generation_config.json +6 -0
- train_results.json +9 -0
- trainer_state.json +522 -0
README.md
ADDED
|
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: mistralai/Mistral-7B-v0.1
|
| 4 |
+
tags:
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
model-index:
|
| 7 |
+
- name: mistral-spin-hhrlhf
|
| 8 |
+
results: []
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 12 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 13 |
+
|
| 14 |
+
# mistral-spin-hhrlhf
|
| 15 |
+
|
| 16 |
+
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on an unknown dataset.
|
| 17 |
+
|
| 18 |
+
## Model description
|
| 19 |
+
|
| 20 |
+
More information needed
|
| 21 |
+
|
| 22 |
+
## Intended uses & limitations
|
| 23 |
+
|
| 24 |
+
More information needed
|
| 25 |
+
|
| 26 |
+
## Training and evaluation data
|
| 27 |
+
|
| 28 |
+
More information needed
|
| 29 |
+
|
| 30 |
+
## Training procedure
|
| 31 |
+
|
| 32 |
+
### Training hyperparameters
|
| 33 |
+
|
| 34 |
+
The following hyperparameters were used during training:
|
| 35 |
+
- learning_rate: 5e-07
|
| 36 |
+
- train_batch_size: 4
|
| 37 |
+
- eval_batch_size: 4
|
| 38 |
+
- seed: 42
|
| 39 |
+
- distributed_type: multi-GPU
|
| 40 |
+
- num_devices: 4
|
| 41 |
+
- total_train_batch_size: 16
|
| 42 |
+
- total_eval_batch_size: 16
|
| 43 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 44 |
+
- lr_scheduler_type: linear
|
| 45 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 46 |
+
- num_epochs: 1
|
| 47 |
+
|
| 48 |
+
### Training results
|
| 49 |
+
|
| 50 |
+
|
| 51 |
+
|
| 52 |
+
### Framework versions
|
| 53 |
+
|
| 54 |
+
- Transformers 4.43.3
|
| 55 |
+
- Pytorch 2.2.2+cu121
|
| 56 |
+
- Datasets 2.20.0
|
| 57 |
+
- Tokenizers 0.19.1
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.14877043307398835,
|
| 5 |
+
"train_runtime": 1689.625,
|
| 6 |
+
"train_samples": 5000,
|
| 7 |
+
"train_samples_per_second": 2.959,
|
| 8 |
+
"train_steps_per_second": 0.185
|
| 9 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 1,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"transformers_version": "4.43.3"
|
| 6 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.14877043307398835,
|
| 5 |
+
"train_runtime": 1689.625,
|
| 6 |
+
"train_samples": 5000,
|
| 7 |
+
"train_samples_per_second": 2.959,
|
| 8 |
+
"train_steps_per_second": 0.185
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,522 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 313,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.003194888178913738,
|
| 13 |
+
"grad_norm": 707.0901555036106,
|
| 14 |
+
"learning_rate": 1.5625e-08,
|
| 15 |
+
"logits/generated": -2.801048994064331,
|
| 16 |
+
"logits/real": -2.5173754692077637,
|
| 17 |
+
"logps/generated": -448.5753173828125,
|
| 18 |
+
"logps/real": -119.54361724853516,
|
| 19 |
+
"loss": 0.8855,
|
| 20 |
+
"rewards/accuracies": 0.0,
|
| 21 |
+
"rewards/generated": 0.0,
|
| 22 |
+
"rewards/margins": 0.0,
|
| 23 |
+
"rewards/real": 0.0,
|
| 24 |
+
"step": 1
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03194888178913738,
|
| 28 |
+
"grad_norm": 26.033036236831965,
|
| 29 |
+
"learning_rate": 1.5624999999999999e-07,
|
| 30 |
+
"logits/generated": -2.4279539585113525,
|
| 31 |
+
"logits/real": -2.416008234024048,
|
| 32 |
+
"logps/generated": -441.2065734863281,
|
| 33 |
+
"logps/real": -192.90866088867188,
|
| 34 |
+
"loss": 0.5492,
|
| 35 |
+
"rewards/accuracies": 0.8611111044883728,
|
| 36 |
+
"rewards/generated": -1.0092414617538452,
|
| 37 |
+
"rewards/margins": 1.2109122276306152,
|
| 38 |
+
"rewards/real": 0.20167087018489838,
|
| 39 |
+
"step": 10
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.06389776357827476,
|
| 43 |
+
"grad_norm": 2.607641570692266,
|
| 44 |
+
"learning_rate": 3.1249999999999997e-07,
|
| 45 |
+
"logits/generated": -2.0776472091674805,
|
| 46 |
+
"logits/real": -2.2696104049682617,
|
| 47 |
+
"logps/generated": -528.232666015625,
|
| 48 |
+
"logps/real": -141.57501220703125,
|
| 49 |
+
"loss": 0.1444,
|
| 50 |
+
"rewards/accuracies": 1.0,
|
| 51 |
+
"rewards/generated": -8.630824089050293,
|
| 52 |
+
"rewards/margins": 10.818326950073242,
|
| 53 |
+
"rewards/real": 2.187502384185791,
|
| 54 |
+
"step": 20
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.09584664536741214,
|
| 58 |
+
"grad_norm": 1.6797275561770264,
|
| 59 |
+
"learning_rate": 4.6874999999999996e-07,
|
| 60 |
+
"logits/generated": -2.0287694931030273,
|
| 61 |
+
"logits/real": -2.202587604522705,
|
| 62 |
+
"logps/generated": -561.3330078125,
|
| 63 |
+
"logps/real": -106.47761535644531,
|
| 64 |
+
"loss": 0.1182,
|
| 65 |
+
"rewards/accuracies": 1.0,
|
| 66 |
+
"rewards/generated": -12.7778902053833,
|
| 67 |
+
"rewards/margins": 15.439559936523438,
|
| 68 |
+
"rewards/real": 2.661670207977295,
|
| 69 |
+
"step": 30
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.12779552715654952,
|
| 73 |
+
"grad_norm": 1.6877722773345887,
|
| 74 |
+
"learning_rate": 4.857651245551601e-07,
|
| 75 |
+
"logits/generated": -2.2078909873962402,
|
| 76 |
+
"logits/real": -2.292013168334961,
|
| 77 |
+
"logps/generated": -599.3379516601562,
|
| 78 |
+
"logps/real": -124.68473052978516,
|
| 79 |
+
"loss": 0.1381,
|
| 80 |
+
"rewards/accuracies": 1.0,
|
| 81 |
+
"rewards/generated": -13.965204238891602,
|
| 82 |
+
"rewards/margins": 16.917484283447266,
|
| 83 |
+
"rewards/real": 2.9522790908813477,
|
| 84 |
+
"step": 40
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.1597444089456869,
|
| 88 |
+
"grad_norm": 1.97067074811704,
|
| 89 |
+
"learning_rate": 4.679715302491103e-07,
|
| 90 |
+
"logits/generated": -2.162761926651001,
|
| 91 |
+
"logits/real": -2.3204751014709473,
|
| 92 |
+
"logps/generated": -582.5556640625,
|
| 93 |
+
"logps/real": -136.84011840820312,
|
| 94 |
+
"loss": 0.143,
|
| 95 |
+
"rewards/accuracies": 1.0,
|
| 96 |
+
"rewards/generated": -14.243738174438477,
|
| 97 |
+
"rewards/margins": 17.179973602294922,
|
| 98 |
+
"rewards/real": 2.936236619949341,
|
| 99 |
+
"step": 50
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.19169329073482427,
|
| 103 |
+
"grad_norm": 1.8076057502854346,
|
| 104 |
+
"learning_rate": 4.501779359430605e-07,
|
| 105 |
+
"logits/generated": -2.0397229194641113,
|
| 106 |
+
"logits/real": -2.299813747406006,
|
| 107 |
+
"logps/generated": -601.5155029296875,
|
| 108 |
+
"logps/real": -158.8765869140625,
|
| 109 |
+
"loss": 0.1215,
|
| 110 |
+
"rewards/accuracies": 1.0,
|
| 111 |
+
"rewards/generated": -15.693455696105957,
|
| 112 |
+
"rewards/margins": 18.398269653320312,
|
| 113 |
+
"rewards/real": 2.7048134803771973,
|
| 114 |
+
"step": 60
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.22364217252396165,
|
| 118 |
+
"grad_norm": 1.5331566397357441,
|
| 119 |
+
"learning_rate": 4.3238434163701063e-07,
|
| 120 |
+
"logits/generated": -2.803489923477173,
|
| 121 |
+
"logits/real": -2.78045654296875,
|
| 122 |
+
"logps/generated": -619.4425048828125,
|
| 123 |
+
"logps/real": -137.57711791992188,
|
| 124 |
+
"loss": 0.1436,
|
| 125 |
+
"rewards/accuracies": 1.0,
|
| 126 |
+
"rewards/generated": -15.494806289672852,
|
| 127 |
+
"rewards/margins": 18.42218780517578,
|
| 128 |
+
"rewards/real": 2.927382707595825,
|
| 129 |
+
"step": 70
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.25559105431309903,
|
| 133 |
+
"grad_norm": 1.511875674922715,
|
| 134 |
+
"learning_rate": 4.1459074733096083e-07,
|
| 135 |
+
"logits/generated": -2.690621852874756,
|
| 136 |
+
"logits/real": -2.5948469638824463,
|
| 137 |
+
"logps/generated": -596.6559448242188,
|
| 138 |
+
"logps/real": -109.74491882324219,
|
| 139 |
+
"loss": 0.1233,
|
| 140 |
+
"rewards/accuracies": 1.0,
|
| 141 |
+
"rewards/generated": -16.492372512817383,
|
| 142 |
+
"rewards/margins": 19.500394821166992,
|
| 143 |
+
"rewards/real": 3.008024215698242,
|
| 144 |
+
"step": 80
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.28753993610223644,
|
| 148 |
+
"grad_norm": 1.3167801498083793,
|
| 149 |
+
"learning_rate": 3.9679715302491103e-07,
|
| 150 |
+
"logits/generated": -2.7257370948791504,
|
| 151 |
+
"logits/real": -2.683110237121582,
|
| 152 |
+
"logps/generated": -610.23583984375,
|
| 153 |
+
"logps/real": -146.4161834716797,
|
| 154 |
+
"loss": 0.1269,
|
| 155 |
+
"rewards/accuracies": 1.0,
|
| 156 |
+
"rewards/generated": -16.799118041992188,
|
| 157 |
+
"rewards/margins": 19.806835174560547,
|
| 158 |
+
"rewards/real": 3.007721185684204,
|
| 159 |
+
"step": 90
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.3194888178913738,
|
| 163 |
+
"grad_norm": 1.6900650289885557,
|
| 164 |
+
"learning_rate": 3.790035587188612e-07,
|
| 165 |
+
"logits/generated": -2.734708309173584,
|
| 166 |
+
"logits/real": -2.7301347255706787,
|
| 167 |
+
"logps/generated": -614.1470336914062,
|
| 168 |
+
"logps/real": -153.92105102539062,
|
| 169 |
+
"loss": 0.1487,
|
| 170 |
+
"rewards/accuracies": 1.0,
|
| 171 |
+
"rewards/generated": -17.10015106201172,
|
| 172 |
+
"rewards/margins": 20.158763885498047,
|
| 173 |
+
"rewards/real": 3.058610677719116,
|
| 174 |
+
"step": 100
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.3514376996805112,
|
| 178 |
+
"grad_norm": 1.3021727337855808,
|
| 179 |
+
"learning_rate": 3.6120996441281137e-07,
|
| 180 |
+
"logits/generated": -2.6350350379943848,
|
| 181 |
+
"logits/real": -2.7262301445007324,
|
| 182 |
+
"logps/generated": -596.9154663085938,
|
| 183 |
+
"logps/real": -122.75111389160156,
|
| 184 |
+
"loss": 0.1255,
|
| 185 |
+
"rewards/accuracies": 1.0,
|
| 186 |
+
"rewards/generated": -16.06097412109375,
|
| 187 |
+
"rewards/margins": 19.084787368774414,
|
| 188 |
+
"rewards/real": 3.023812770843506,
|
| 189 |
+
"step": 110
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.38338658146964855,
|
| 193 |
+
"grad_norm": 1.6018857166027896,
|
| 194 |
+
"learning_rate": 3.4341637010676157e-07,
|
| 195 |
+
"logits/generated": -2.7175872325897217,
|
| 196 |
+
"logits/real": -2.7377209663391113,
|
| 197 |
+
"logps/generated": -641.0619506835938,
|
| 198 |
+
"logps/real": -134.5575408935547,
|
| 199 |
+
"loss": 0.1344,
|
| 200 |
+
"rewards/accuracies": 1.0,
|
| 201 |
+
"rewards/generated": -18.354572296142578,
|
| 202 |
+
"rewards/margins": 21.459613800048828,
|
| 203 |
+
"rewards/real": 3.1050407886505127,
|
| 204 |
+
"step": 120
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.41533546325878595,
|
| 208 |
+
"grad_norm": 1.957334214600403,
|
| 209 |
+
"learning_rate": 3.256227758007117e-07,
|
| 210 |
+
"logits/generated": -2.6532580852508545,
|
| 211 |
+
"logits/real": -2.6732802391052246,
|
| 212 |
+
"logps/generated": -593.9332885742188,
|
| 213 |
+
"logps/real": -146.0222625732422,
|
| 214 |
+
"loss": 0.1467,
|
| 215 |
+
"rewards/accuracies": 1.0,
|
| 216 |
+
"rewards/generated": -16.877891540527344,
|
| 217 |
+
"rewards/margins": 19.932432174682617,
|
| 218 |
+
"rewards/real": 3.054542064666748,
|
| 219 |
+
"step": 130
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.4472843450479233,
|
| 223 |
+
"grad_norm": 1.3941013687837516,
|
| 224 |
+
"learning_rate": 3.078291814946619e-07,
|
| 225 |
+
"logits/generated": -2.655681848526001,
|
| 226 |
+
"logits/real": -2.6176505088806152,
|
| 227 |
+
"logps/generated": -563.309814453125,
|
| 228 |
+
"logps/real": -146.92739868164062,
|
| 229 |
+
"loss": 0.1287,
|
| 230 |
+
"rewards/accuracies": 1.0,
|
| 231 |
+
"rewards/generated": -14.66212272644043,
|
| 232 |
+
"rewards/margins": 17.799179077148438,
|
| 233 |
+
"rewards/real": 3.1370577812194824,
|
| 234 |
+
"step": 140
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.4792332268370607,
|
| 238 |
+
"grad_norm": 1.4882690191731298,
|
| 239 |
+
"learning_rate": 2.900355871886121e-07,
|
| 240 |
+
"logits/generated": -2.747220993041992,
|
| 241 |
+
"logits/real": -2.701289176940918,
|
| 242 |
+
"logps/generated": -579.5242919921875,
|
| 243 |
+
"logps/real": -130.2000274658203,
|
| 244 |
+
"loss": 0.1296,
|
| 245 |
+
"rewards/accuracies": 1.0,
|
| 246 |
+
"rewards/generated": -16.728702545166016,
|
| 247 |
+
"rewards/margins": 19.850534439086914,
|
| 248 |
+
"rewards/real": 3.1218338012695312,
|
| 249 |
+
"step": 150
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.5111821086261981,
|
| 253 |
+
"grad_norm": 1.5998263047919852,
|
| 254 |
+
"learning_rate": 2.7224199288256225e-07,
|
| 255 |
+
"logits/generated": -2.689751148223877,
|
| 256 |
+
"logits/real": -2.7137889862060547,
|
| 257 |
+
"logps/generated": -598.3092041015625,
|
| 258 |
+
"logps/real": -124.75559997558594,
|
| 259 |
+
"loss": 0.1466,
|
| 260 |
+
"rewards/accuracies": 1.0,
|
| 261 |
+
"rewards/generated": -16.721622467041016,
|
| 262 |
+
"rewards/margins": 19.949649810791016,
|
| 263 |
+
"rewards/real": 3.2280242443084717,
|
| 264 |
+
"step": 160
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.5431309904153354,
|
| 268 |
+
"grad_norm": 1.828096060541728,
|
| 269 |
+
"learning_rate": 2.5444839857651245e-07,
|
| 270 |
+
"logits/generated": -2.631895065307617,
|
| 271 |
+
"logits/real": -2.652967929840088,
|
| 272 |
+
"logps/generated": -586.2786254882812,
|
| 273 |
+
"logps/real": -147.9235382080078,
|
| 274 |
+
"loss": 0.1406,
|
| 275 |
+
"rewards/accuracies": 1.0,
|
| 276 |
+
"rewards/generated": -15.885740280151367,
|
| 277 |
+
"rewards/margins": 19.117094039916992,
|
| 278 |
+
"rewards/real": 3.231353282928467,
|
| 279 |
+
"step": 170
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.5750798722044729,
|
| 283 |
+
"grad_norm": 1.4197910889449137,
|
| 284 |
+
"learning_rate": 2.366548042704626e-07,
|
| 285 |
+
"logits/generated": -2.669239044189453,
|
| 286 |
+
"logits/real": -2.614443063735962,
|
| 287 |
+
"logps/generated": -651.5192260742188,
|
| 288 |
+
"logps/real": -123.7098617553711,
|
| 289 |
+
"loss": 0.1397,
|
| 290 |
+
"rewards/accuracies": 1.0,
|
| 291 |
+
"rewards/generated": -17.95551109313965,
|
| 292 |
+
"rewards/margins": 21.130229949951172,
|
| 293 |
+
"rewards/real": 3.1747171878814697,
|
| 294 |
+
"step": 180
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.6070287539936102,
|
| 298 |
+
"grad_norm": 1.623253543306337,
|
| 299 |
+
"learning_rate": 2.188612099644128e-07,
|
| 300 |
+
"logits/generated": -2.6021995544433594,
|
| 301 |
+
"logits/real": -2.6068227291107178,
|
| 302 |
+
"logps/generated": -603.6708374023438,
|
| 303 |
+
"logps/real": -133.12811279296875,
|
| 304 |
+
"loss": 0.1504,
|
| 305 |
+
"rewards/accuracies": 1.0,
|
| 306 |
+
"rewards/generated": -17.770116806030273,
|
| 307 |
+
"rewards/margins": 21.024080276489258,
|
| 308 |
+
"rewards/real": 3.2539620399475098,
|
| 309 |
+
"step": 190
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.6389776357827476,
|
| 313 |
+
"grad_norm": 1.6934472036753703,
|
| 314 |
+
"learning_rate": 2.0106761565836297e-07,
|
| 315 |
+
"logits/generated": -2.6339528560638428,
|
| 316 |
+
"logits/real": -2.5701403617858887,
|
| 317 |
+
"logps/generated": -586.7174072265625,
|
| 318 |
+
"logps/real": -126.43733978271484,
|
| 319 |
+
"loss": 0.1329,
|
| 320 |
+
"rewards/accuracies": 1.0,
|
| 321 |
+
"rewards/generated": -16.226985931396484,
|
| 322 |
+
"rewards/margins": 19.362442016601562,
|
| 323 |
+
"rewards/real": 3.1354544162750244,
|
| 324 |
+
"step": 200
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.670926517571885,
|
| 328 |
+
"grad_norm": 1.498510937597327,
|
| 329 |
+
"learning_rate": 1.8327402135231316e-07,
|
| 330 |
+
"logits/generated": -2.8424220085144043,
|
| 331 |
+
"logits/real": -2.6752820014953613,
|
| 332 |
+
"logps/generated": -605.6686401367188,
|
| 333 |
+
"logps/real": -122.51239013671875,
|
| 334 |
+
"loss": 0.1198,
|
| 335 |
+
"rewards/accuracies": 1.0,
|
| 336 |
+
"rewards/generated": -16.924907684326172,
|
| 337 |
+
"rewards/margins": 20.049123764038086,
|
| 338 |
+
"rewards/real": 3.1242165565490723,
|
| 339 |
+
"step": 210
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.7028753993610224,
|
| 343 |
+
"grad_norm": 1.4317115723777145,
|
| 344 |
+
"learning_rate": 1.6548042704626334e-07,
|
| 345 |
+
"logits/generated": -2.7989039421081543,
|
| 346 |
+
"logits/real": -2.6403791904449463,
|
| 347 |
+
"logps/generated": -624.21826171875,
|
| 348 |
+
"logps/real": -137.1665496826172,
|
| 349 |
+
"loss": 0.1386,
|
| 350 |
+
"rewards/accuracies": 1.0,
|
| 351 |
+
"rewards/generated": -18.16226577758789,
|
| 352 |
+
"rewards/margins": 21.232851028442383,
|
| 353 |
+
"rewards/real": 3.070587635040283,
|
| 354 |
+
"step": 220
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.7348242811501597,
|
| 358 |
+
"grad_norm": 1.303859194823977,
|
| 359 |
+
"learning_rate": 1.476868327402135e-07,
|
| 360 |
+
"logits/generated": -2.635648250579834,
|
| 361 |
+
"logits/real": -2.5836873054504395,
|
| 362 |
+
"logps/generated": -611.7884521484375,
|
| 363 |
+
"logps/real": -139.7376708984375,
|
| 364 |
+
"loss": 0.1379,
|
| 365 |
+
"rewards/accuracies": 1.0,
|
| 366 |
+
"rewards/generated": -16.920841217041016,
|
| 367 |
+
"rewards/margins": 20.23355484008789,
|
| 368 |
+
"rewards/real": 3.312713623046875,
|
| 369 |
+
"step": 230
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.7667731629392971,
|
| 373 |
+
"grad_norm": 1.671858324479984,
|
| 374 |
+
"learning_rate": 1.298932384341637e-07,
|
| 375 |
+
"logits/generated": -2.835536003112793,
|
| 376 |
+
"logits/real": -2.5438485145568848,
|
| 377 |
+
"logps/generated": -604.9158325195312,
|
| 378 |
+
"logps/real": -90.45211029052734,
|
| 379 |
+
"loss": 0.1236,
|
| 380 |
+
"rewards/accuracies": 1.0,
|
| 381 |
+
"rewards/generated": -18.167110443115234,
|
| 382 |
+
"rewards/margins": 21.106388092041016,
|
| 383 |
+
"rewards/real": 2.939276933670044,
|
| 384 |
+
"step": 240
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.7987220447284346,
|
| 388 |
+
"grad_norm": 1.4302765337757193,
|
| 389 |
+
"learning_rate": 1.1209964412811388e-07,
|
| 390 |
+
"logits/generated": -2.773268938064575,
|
| 391 |
+
"logits/real": -2.7092514038085938,
|
| 392 |
+
"logps/generated": -657.0368041992188,
|
| 393 |
+
"logps/real": -131.35690307617188,
|
| 394 |
+
"loss": 0.1284,
|
| 395 |
+
"rewards/accuracies": 1.0,
|
| 396 |
+
"rewards/generated": -19.432615280151367,
|
| 397 |
+
"rewards/margins": 22.64904022216797,
|
| 398 |
+
"rewards/real": 3.2164230346679688,
|
| 399 |
+
"step": 250
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.8306709265175719,
|
| 403 |
+
"grad_norm": 1.443073941719056,
|
| 404 |
+
"learning_rate": 9.430604982206406e-08,
|
| 405 |
+
"logits/generated": -2.756702184677124,
|
| 406 |
+
"logits/real": -2.677367687225342,
|
| 407 |
+
"logps/generated": -634.17041015625,
|
| 408 |
+
"logps/real": -103.53001403808594,
|
| 409 |
+
"loss": 0.1157,
|
| 410 |
+
"rewards/accuracies": 1.0,
|
| 411 |
+
"rewards/generated": -18.820682525634766,
|
| 412 |
+
"rewards/margins": 21.887561798095703,
|
| 413 |
+
"rewards/real": 3.0668790340423584,
|
| 414 |
+
"step": 260
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 0.8626198083067093,
|
| 418 |
+
"grad_norm": 1.8796921765501535,
|
| 419 |
+
"learning_rate": 7.651245551601423e-08,
|
| 420 |
+
"logits/generated": -2.7278664112091064,
|
| 421 |
+
"logits/real": -2.7242510318756104,
|
| 422 |
+
"logps/generated": -606.2032470703125,
|
| 423 |
+
"logps/real": -182.83935546875,
|
| 424 |
+
"loss": 0.1603,
|
| 425 |
+
"rewards/accuracies": 1.0,
|
| 426 |
+
"rewards/generated": -18.959697723388672,
|
| 427 |
+
"rewards/margins": 22.330434799194336,
|
| 428 |
+
"rewards/real": 3.3707375526428223,
|
| 429 |
+
"step": 270
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.8945686900958466,
|
| 433 |
+
"grad_norm": 1.553843219847652,
|
| 434 |
+
"learning_rate": 5.871886120996441e-08,
|
| 435 |
+
"logits/generated": -2.738983154296875,
|
| 436 |
+
"logits/real": -2.6845531463623047,
|
| 437 |
+
"logps/generated": -631.7647705078125,
|
| 438 |
+
"logps/real": -175.20382690429688,
|
| 439 |
+
"loss": 0.1264,
|
| 440 |
+
"rewards/accuracies": 1.0,
|
| 441 |
+
"rewards/generated": -19.304214477539062,
|
| 442 |
+
"rewards/margins": 22.55462646484375,
|
| 443 |
+
"rewards/real": 3.2504119873046875,
|
| 444 |
+
"step": 280
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.9265175718849841,
|
| 448 |
+
"grad_norm": 1.8666524089612875,
|
| 449 |
+
"learning_rate": 4.092526690391459e-08,
|
| 450 |
+
"logits/generated": -2.6846232414245605,
|
| 451 |
+
"logits/real": -2.6559813022613525,
|
| 452 |
+
"logps/generated": -589.41796875,
|
| 453 |
+
"logps/real": -121.3634262084961,
|
| 454 |
+
"loss": 0.1282,
|
| 455 |
+
"rewards/accuracies": 1.0,
|
| 456 |
+
"rewards/generated": -18.461414337158203,
|
| 457 |
+
"rewards/margins": 21.56503677368164,
|
| 458 |
+
"rewards/real": 3.103621006011963,
|
| 459 |
+
"step": 290
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.9584664536741214,
|
| 463 |
+
"grad_norm": 1.9808591697359095,
|
| 464 |
+
"learning_rate": 2.3131672597864765e-08,
|
| 465 |
+
"logits/generated": -2.789712905883789,
|
| 466 |
+
"logits/real": -2.6096606254577637,
|
| 467 |
+
"logps/generated": -629.02294921875,
|
| 468 |
+
"logps/real": -189.27085876464844,
|
| 469 |
+
"loss": 0.1435,
|
| 470 |
+
"rewards/accuracies": 1.0,
|
| 471 |
+
"rewards/generated": -18.036930084228516,
|
| 472 |
+
"rewards/margins": 21.457435607910156,
|
| 473 |
+
"rewards/real": 3.4205050468444824,
|
| 474 |
+
"step": 300
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 0.9904153354632588,
|
| 478 |
+
"grad_norm": 1.5167523084769354,
|
| 479 |
+
"learning_rate": 5.338078291814947e-09,
|
| 480 |
+
"logits/generated": -2.786799192428589,
|
| 481 |
+
"logits/real": -2.7553584575653076,
|
| 482 |
+
"logps/generated": -612.6634521484375,
|
| 483 |
+
"logps/real": -153.3998260498047,
|
| 484 |
+
"loss": 0.1292,
|
| 485 |
+
"rewards/accuracies": 1.0,
|
| 486 |
+
"rewards/generated": -17.887155532836914,
|
| 487 |
+
"rewards/margins": 21.20827293395996,
|
| 488 |
+
"rewards/real": 3.321115016937256,
|
| 489 |
+
"step": 310
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 1.0,
|
| 493 |
+
"step": 313,
|
| 494 |
+
"total_flos": 0.0,
|
| 495 |
+
"train_loss": 0.14877043307398835,
|
| 496 |
+
"train_runtime": 1689.625,
|
| 497 |
+
"train_samples_per_second": 2.959,
|
| 498 |
+
"train_steps_per_second": 0.185
|
| 499 |
+
}
|
| 500 |
+
],
|
| 501 |
+
"logging_steps": 10,
|
| 502 |
+
"max_steps": 313,
|
| 503 |
+
"num_input_tokens_seen": 0,
|
| 504 |
+
"num_train_epochs": 1,
|
| 505 |
+
"save_steps": 1000,
|
| 506 |
+
"stateful_callbacks": {
|
| 507 |
+
"TrainerControl": {
|
| 508 |
+
"args": {
|
| 509 |
+
"should_epoch_stop": false,
|
| 510 |
+
"should_evaluate": false,
|
| 511 |
+
"should_log": false,
|
| 512 |
+
"should_save": true,
|
| 513 |
+
"should_training_stop": true
|
| 514 |
+
},
|
| 515 |
+
"attributes": {}
|
| 516 |
+
}
|
| 517 |
+
},
|
| 518 |
+
"total_flos": 0.0,
|
| 519 |
+
"train_batch_size": 4,
|
| 520 |
+
"trial_name": null,
|
| 521 |
+
"trial_params": null
|
| 522 |
+
}
|