Upload folder using huggingface_hub
Browse files- checkpoint-1000/config.json +25 -0
- checkpoint-1000/generation_config.json +6 -0
- checkpoint-1000/optimizer.pt +3 -0
- checkpoint-1000/pytorch_model.bin +3 -0
- checkpoint-1000/rng_state.pth +3 -0
- checkpoint-1000/scheduler.pt +3 -0
- checkpoint-1000/special_tokens_map.json +14 -0
- checkpoint-1000/tokenizer.json +0 -0
- checkpoint-1000/tokenizer_config.json +10 -0
- checkpoint-1000/trainer_state.json +661 -0
- checkpoint-1000/training_args.bin +3 -0
- checkpoint-2000/config.json +25 -0
- checkpoint-2000/generation_config.json +6 -0
- checkpoint-2000/optimizer.pt +3 -0
- checkpoint-2000/pytorch_model.bin +3 -0
- checkpoint-2000/rng_state.pth +3 -0
- checkpoint-2000/scheduler.pt +3 -0
- checkpoint-2000/special_tokens_map.json +14 -0
- checkpoint-2000/tokenizer.json +0 -0
- checkpoint-2000/tokenizer_config.json +10 -0
- checkpoint-2000/trainer_state.json +1306 -0
- checkpoint-2000/training_args.bin +3 -0
- config.json +25 -0
- generation_config.json +6 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +14 -0
- tokenizer.json +0 -0
- tokenizer_config.json +10 -0
- training_args.bin +3 -0
checkpoint-1000/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "EleutherAI/pythia-70m-deduped",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"GPTNeoXForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"eos_token_id": 0,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_size": 512,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 2048,
|
| 12 |
+
"layer_norm_eps": 1e-05,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "gpt_neox",
|
| 15 |
+
"num_attention_heads": 8,
|
| 16 |
+
"num_hidden_layers": 6,
|
| 17 |
+
"rotary_emb_base": 10000,
|
| 18 |
+
"rotary_pct": 0.25,
|
| 19 |
+
"tie_word_embeddings": false,
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0.dev0",
|
| 22 |
+
"use_cache": true,
|
| 23 |
+
"use_parallel_residual": true,
|
| 24 |
+
"vocab_size": 50288
|
| 25 |
+
}
|
checkpoint-1000/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"eos_token_id": 0,
|
| 5 |
+
"transformers_version": "4.28.0.dev0"
|
| 6 |
+
}
|
checkpoint-1000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b07ac90ca497c2d68f2de635fd5e7a1f5f15b0da437b8d5d96ecc6f4821a789e
|
| 3 |
+
size 563326405
|
checkpoint-1000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4be53587133a680397004c5f91da72b5ecd41f75e228e67f917ce9dd3c07e995
|
| 3 |
+
size 306838155
|
checkpoint-1000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bef1f5a9d2e04387e25ae037dc21f0d007f2d49e037447909baf4ee6b4256660
|
| 3 |
+
size 14575
|
checkpoint-1000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:462c009a3e31fdee826b495fda5ad5b00113e60e337dc5d530aa689850317695
|
| 3 |
+
size 627
|
checkpoint-1000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<|prefix_begin|>",
|
| 4 |
+
"<|system|>",
|
| 5 |
+
"<|prefix_end|>",
|
| 6 |
+
"<|assistant|>",
|
| 7 |
+
"<|prompter|>"
|
| 8 |
+
],
|
| 9 |
+
"bos_token": "<|endoftext|>",
|
| 10 |
+
"eos_token": "<|endoftext|>",
|
| 11 |
+
"pad_token": "<|padding|>",
|
| 12 |
+
"sep_token": "<|endoftext|>",
|
| 13 |
+
"unk_token": "<|endoftext|>"
|
| 14 |
+
}
|
checkpoint-1000/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"bos_token": "<|endoftext|>",
|
| 4 |
+
"clean_up_tokenization_spaces": true,
|
| 5 |
+
"eos_token": "<|endoftext|>",
|
| 6 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 7 |
+
"special_tokens_map_file": "/admin/home-hailey/.cache/huggingface/hub/models--EleutherAI--gpt-neox-20b/snapshots/4e49eadb5d14bd22f314ec3f45b69a87b88c7691/special_tokens_map.json",
|
| 8 |
+
"tokenizer_class": "GPTNeoXTokenizer",
|
| 9 |
+
"unk_token": "<|endoftext|>"
|
| 10 |
+
}
|
checkpoint-1000/trainer_state.json
ADDED
|
@@ -0,0 +1,661 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.2629806342969407,
|
| 5 |
+
"global_step": 1000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 8e-08,
|
| 13 |
+
"loss": 11.8368,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.03,
|
| 18 |
+
"learning_rate": 1.6e-07,
|
| 19 |
+
"loss": 11.103,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.04,
|
| 24 |
+
"learning_rate": 2.4e-07,
|
| 25 |
+
"loss": 11.6842,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.05,
|
| 30 |
+
"learning_rate": 3.2e-07,
|
| 31 |
+
"loss": 11.7279,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.06,
|
| 36 |
+
"learning_rate": 4e-07,
|
| 37 |
+
"loss": 11.6439,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.08,
|
| 42 |
+
"learning_rate": 4.8e-07,
|
| 43 |
+
"loss": 11.1668,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.09,
|
| 48 |
+
"learning_rate": 5.6e-07,
|
| 49 |
+
"loss": 11.5542,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.1,
|
| 54 |
+
"learning_rate": 6.4e-07,
|
| 55 |
+
"loss": 11.2201,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.11,
|
| 60 |
+
"learning_rate": 7.2e-07,
|
| 61 |
+
"loss": 10.0118,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.13,
|
| 66 |
+
"learning_rate": 8e-07,
|
| 67 |
+
"loss": 9.2907,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.14,
|
| 72 |
+
"learning_rate": 8.799999999999999e-07,
|
| 73 |
+
"loss": 8.9216,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.15,
|
| 78 |
+
"learning_rate": 9.6e-07,
|
| 79 |
+
"loss": 8.9268,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.16,
|
| 84 |
+
"learning_rate": 1.04e-06,
|
| 85 |
+
"loss": 8.6545,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.18,
|
| 90 |
+
"learning_rate": 1.12e-06,
|
| 91 |
+
"loss": 8.7888,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.19,
|
| 96 |
+
"learning_rate": 1.2e-06,
|
| 97 |
+
"loss": 9.0201,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.2,
|
| 102 |
+
"learning_rate": 1.28e-06,
|
| 103 |
+
"loss": 8.0512,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.21,
|
| 108 |
+
"learning_rate": 1.3600000000000001e-06,
|
| 109 |
+
"loss": 5.9266,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.23,
|
| 114 |
+
"learning_rate": 1.44e-06,
|
| 115 |
+
"loss": 5.3527,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.24,
|
| 120 |
+
"learning_rate": 1.5199999999999998e-06,
|
| 121 |
+
"loss": 4.9847,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.25,
|
| 126 |
+
"learning_rate": 1.6e-06,
|
| 127 |
+
"loss": 4.8117,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.25,
|
| 132 |
+
"eval_webgpt_accuracy": 0.3513391153050626,
|
| 133 |
+
"eval_webgpt_loss": 4.794132709503174,
|
| 134 |
+
"eval_webgpt_runtime": 39.5471,
|
| 135 |
+
"eval_webgpt_samples_per_second": 90.095,
|
| 136 |
+
"eval_webgpt_steps_per_second": 22.53,
|
| 137 |
+
"step": 200
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.27,
|
| 141 |
+
"learning_rate": 1.6799999999999998e-06,
|
| 142 |
+
"loss": 4.7854,
|
| 143 |
+
"step": 210
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.28,
|
| 147 |
+
"learning_rate": 1.7599999999999999e-06,
|
| 148 |
+
"loss": 4.5555,
|
| 149 |
+
"step": 220
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.29,
|
| 153 |
+
"learning_rate": 1.84e-06,
|
| 154 |
+
"loss": 4.4819,
|
| 155 |
+
"step": 230
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.3,
|
| 159 |
+
"learning_rate": 1.92e-06,
|
| 160 |
+
"loss": 4.3504,
|
| 161 |
+
"step": 240
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.32,
|
| 165 |
+
"learning_rate": 2e-06,
|
| 166 |
+
"loss": 4.1471,
|
| 167 |
+
"step": 250
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 0.33,
|
| 171 |
+
"learning_rate": 2.08e-06,
|
| 172 |
+
"loss": 4.4132,
|
| 173 |
+
"step": 260
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 0.34,
|
| 177 |
+
"learning_rate": 2.16e-06,
|
| 178 |
+
"loss": 4.1309,
|
| 179 |
+
"step": 270
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.35,
|
| 183 |
+
"learning_rate": 2.24e-06,
|
| 184 |
+
"loss": 4.4153,
|
| 185 |
+
"step": 280
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.37,
|
| 189 |
+
"learning_rate": 2.32e-06,
|
| 190 |
+
"loss": 4.3256,
|
| 191 |
+
"step": 290
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.38,
|
| 195 |
+
"learning_rate": 2.4e-06,
|
| 196 |
+
"loss": 4.3798,
|
| 197 |
+
"step": 300
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.39,
|
| 201 |
+
"learning_rate": 2.48e-06,
|
| 202 |
+
"loss": 4.3665,
|
| 203 |
+
"step": 310
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.4,
|
| 207 |
+
"learning_rate": 2.56e-06,
|
| 208 |
+
"loss": 4.2545,
|
| 209 |
+
"step": 320
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.42,
|
| 213 |
+
"learning_rate": 2.64e-06,
|
| 214 |
+
"loss": 4.1912,
|
| 215 |
+
"step": 330
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.43,
|
| 219 |
+
"learning_rate": 2.7200000000000002e-06,
|
| 220 |
+
"loss": 4.2359,
|
| 221 |
+
"step": 340
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.44,
|
| 225 |
+
"learning_rate": 2.8e-06,
|
| 226 |
+
"loss": 4.2145,
|
| 227 |
+
"step": 350
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.45,
|
| 231 |
+
"learning_rate": 2.88e-06,
|
| 232 |
+
"loss": 4.0716,
|
| 233 |
+
"step": 360
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.47,
|
| 237 |
+
"learning_rate": 2.96e-06,
|
| 238 |
+
"loss": 4.0429,
|
| 239 |
+
"step": 370
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.48,
|
| 243 |
+
"learning_rate": 3.0399999999999997e-06,
|
| 244 |
+
"loss": 4.1404,
|
| 245 |
+
"step": 380
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 0.49,
|
| 249 |
+
"learning_rate": 3.1199999999999998e-06,
|
| 250 |
+
"loss": 3.9878,
|
| 251 |
+
"step": 390
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.51,
|
| 255 |
+
"learning_rate": 3.2e-06,
|
| 256 |
+
"loss": 4.1169,
|
| 257 |
+
"step": 400
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.51,
|
| 261 |
+
"eval_webgpt_accuracy": 0.3585799846229516,
|
| 262 |
+
"eval_webgpt_loss": 3.982567310333252,
|
| 263 |
+
"eval_webgpt_runtime": 39.5324,
|
| 264 |
+
"eval_webgpt_samples_per_second": 90.129,
|
| 265 |
+
"eval_webgpt_steps_per_second": 22.538,
|
| 266 |
+
"step": 400
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.52,
|
| 270 |
+
"learning_rate": 3.2799999999999995e-06,
|
| 271 |
+
"loss": 3.9303,
|
| 272 |
+
"step": 410
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.53,
|
| 276 |
+
"learning_rate": 3.3599999999999996e-06,
|
| 277 |
+
"loss": 4.0221,
|
| 278 |
+
"step": 420
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.54,
|
| 282 |
+
"learning_rate": 3.4399999999999997e-06,
|
| 283 |
+
"loss": 3.8958,
|
| 284 |
+
"step": 430
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.56,
|
| 288 |
+
"learning_rate": 3.5199999999999998e-06,
|
| 289 |
+
"loss": 3.8434,
|
| 290 |
+
"step": 440
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.57,
|
| 294 |
+
"learning_rate": 3.6e-06,
|
| 295 |
+
"loss": 4.0083,
|
| 296 |
+
"step": 450
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.58,
|
| 300 |
+
"learning_rate": 3.68e-06,
|
| 301 |
+
"loss": 3.9172,
|
| 302 |
+
"step": 460
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.59,
|
| 306 |
+
"learning_rate": 3.7599999999999996e-06,
|
| 307 |
+
"loss": 4.0955,
|
| 308 |
+
"step": 470
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.61,
|
| 312 |
+
"learning_rate": 3.84e-06,
|
| 313 |
+
"loss": 4.1085,
|
| 314 |
+
"step": 480
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.62,
|
| 318 |
+
"learning_rate": 3.92e-06,
|
| 319 |
+
"loss": 4.149,
|
| 320 |
+
"step": 490
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.63,
|
| 324 |
+
"learning_rate": 4e-06,
|
| 325 |
+
"loss": 4.0345,
|
| 326 |
+
"step": 500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.64,
|
| 330 |
+
"learning_rate": 4.08e-06,
|
| 331 |
+
"loss": 3.9929,
|
| 332 |
+
"step": 510
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.66,
|
| 336 |
+
"learning_rate": 4.16e-06,
|
| 337 |
+
"loss": 3.7969,
|
| 338 |
+
"step": 520
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.67,
|
| 342 |
+
"learning_rate": 4.24e-06,
|
| 343 |
+
"loss": 3.6102,
|
| 344 |
+
"step": 530
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.68,
|
| 348 |
+
"learning_rate": 4.32e-06,
|
| 349 |
+
"loss": 3.744,
|
| 350 |
+
"step": 540
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.69,
|
| 354 |
+
"learning_rate": 4.4e-06,
|
| 355 |
+
"loss": 3.8242,
|
| 356 |
+
"step": 550
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.71,
|
| 360 |
+
"learning_rate": 4.48e-06,
|
| 361 |
+
"loss": 3.8092,
|
| 362 |
+
"step": 560
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.72,
|
| 366 |
+
"learning_rate": 4.5599999999999995e-06,
|
| 367 |
+
"loss": 3.8835,
|
| 368 |
+
"step": 570
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.73,
|
| 372 |
+
"learning_rate": 4.64e-06,
|
| 373 |
+
"loss": 3.7335,
|
| 374 |
+
"step": 580
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.75,
|
| 378 |
+
"learning_rate": 4.72e-06,
|
| 379 |
+
"loss": 3.8462,
|
| 380 |
+
"step": 590
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.76,
|
| 384 |
+
"learning_rate": 4.8e-06,
|
| 385 |
+
"loss": 3.7422,
|
| 386 |
+
"step": 600
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.76,
|
| 390 |
+
"eval_webgpt_accuracy": 0.36334147099294245,
|
| 391 |
+
"eval_webgpt_loss": 3.8047080039978027,
|
| 392 |
+
"eval_webgpt_runtime": 39.6534,
|
| 393 |
+
"eval_webgpt_samples_per_second": 89.854,
|
| 394 |
+
"eval_webgpt_steps_per_second": 22.47,
|
| 395 |
+
"step": 600
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.77,
|
| 399 |
+
"learning_rate": 4.88e-06,
|
| 400 |
+
"loss": 3.7068,
|
| 401 |
+
"step": 610
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.78,
|
| 405 |
+
"learning_rate": 4.96e-06,
|
| 406 |
+
"loss": 3.69,
|
| 407 |
+
"step": 620
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.8,
|
| 411 |
+
"learning_rate": 5.04e-06,
|
| 412 |
+
"loss": 3.65,
|
| 413 |
+
"step": 630
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.81,
|
| 417 |
+
"learning_rate": 5.12e-06,
|
| 418 |
+
"loss": 3.6972,
|
| 419 |
+
"step": 640
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.82,
|
| 423 |
+
"learning_rate": 5.2e-06,
|
| 424 |
+
"loss": 3.7907,
|
| 425 |
+
"step": 650
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.83,
|
| 429 |
+
"learning_rate": 5.28e-06,
|
| 430 |
+
"loss": 3.7505,
|
| 431 |
+
"step": 660
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 0.85,
|
| 435 |
+
"learning_rate": 5.36e-06,
|
| 436 |
+
"loss": 3.8109,
|
| 437 |
+
"step": 670
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.86,
|
| 441 |
+
"learning_rate": 5.4400000000000004e-06,
|
| 442 |
+
"loss": 3.8612,
|
| 443 |
+
"step": 680
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.87,
|
| 447 |
+
"learning_rate": 5.52e-06,
|
| 448 |
+
"loss": 3.7077,
|
| 449 |
+
"step": 690
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.88,
|
| 453 |
+
"learning_rate": 5.6e-06,
|
| 454 |
+
"loss": 3.8546,
|
| 455 |
+
"step": 700
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.9,
|
| 459 |
+
"learning_rate": 5.68e-06,
|
| 460 |
+
"loss": 3.8801,
|
| 461 |
+
"step": 710
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.91,
|
| 465 |
+
"learning_rate": 5.76e-06,
|
| 466 |
+
"loss": 3.7939,
|
| 467 |
+
"step": 720
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 0.92,
|
| 471 |
+
"learning_rate": 5.84e-06,
|
| 472 |
+
"loss": 3.6499,
|
| 473 |
+
"step": 730
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.93,
|
| 477 |
+
"learning_rate": 5.92e-06,
|
| 478 |
+
"loss": 4.0742,
|
| 479 |
+
"step": 740
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.95,
|
| 483 |
+
"learning_rate": 6e-06,
|
| 484 |
+
"loss": 3.8072,
|
| 485 |
+
"step": 750
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.96,
|
| 489 |
+
"learning_rate": 6.079999999999999e-06,
|
| 490 |
+
"loss": 3.7959,
|
| 491 |
+
"step": 760
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.97,
|
| 495 |
+
"learning_rate": 6.1599999999999995e-06,
|
| 496 |
+
"loss": 3.7556,
|
| 497 |
+
"step": 770
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.99,
|
| 501 |
+
"learning_rate": 6.2399999999999995e-06,
|
| 502 |
+
"loss": 3.8263,
|
| 503 |
+
"step": 780
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 1.0,
|
| 507 |
+
"learning_rate": 6.32e-06,
|
| 508 |
+
"loss": 3.9162,
|
| 509 |
+
"step": 790
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 1.01,
|
| 513 |
+
"learning_rate": 6.4e-06,
|
| 514 |
+
"loss": 3.7188,
|
| 515 |
+
"step": 800
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.01,
|
| 519 |
+
"eval_webgpt_accuracy": 0.3654184548784667,
|
| 520 |
+
"eval_webgpt_loss": 3.7724239826202393,
|
| 521 |
+
"eval_webgpt_runtime": 39.717,
|
| 522 |
+
"eval_webgpt_samples_per_second": 89.71,
|
| 523 |
+
"eval_webgpt_steps_per_second": 22.434,
|
| 524 |
+
"step": 800
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 1.02,
|
| 528 |
+
"learning_rate": 6.48e-06,
|
| 529 |
+
"loss": 3.7938,
|
| 530 |
+
"step": 810
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 1.04,
|
| 534 |
+
"learning_rate": 6.559999999999999e-06,
|
| 535 |
+
"loss": 3.6401,
|
| 536 |
+
"step": 820
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.05,
|
| 540 |
+
"learning_rate": 6.639999999999999e-06,
|
| 541 |
+
"loss": 3.7234,
|
| 542 |
+
"step": 830
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.06,
|
| 546 |
+
"learning_rate": 6.719999999999999e-06,
|
| 547 |
+
"loss": 3.6234,
|
| 548 |
+
"step": 840
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.07,
|
| 552 |
+
"learning_rate": 6.799999999999999e-06,
|
| 553 |
+
"loss": 3.7614,
|
| 554 |
+
"step": 850
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 1.09,
|
| 558 |
+
"learning_rate": 6.879999999999999e-06,
|
| 559 |
+
"loss": 3.7001,
|
| 560 |
+
"step": 860
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.1,
|
| 564 |
+
"learning_rate": 6.9599999999999994e-06,
|
| 565 |
+
"loss": 3.8073,
|
| 566 |
+
"step": 870
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 1.11,
|
| 570 |
+
"learning_rate": 7.0399999999999995e-06,
|
| 571 |
+
"loss": 3.6328,
|
| 572 |
+
"step": 880
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 1.12,
|
| 576 |
+
"learning_rate": 7.12e-06,
|
| 577 |
+
"loss": 3.9974,
|
| 578 |
+
"step": 890
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.14,
|
| 582 |
+
"learning_rate": 7.2e-06,
|
| 583 |
+
"loss": 3.6881,
|
| 584 |
+
"step": 900
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.15,
|
| 588 |
+
"learning_rate": 7.28e-06,
|
| 589 |
+
"loss": 3.7669,
|
| 590 |
+
"step": 910
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.16,
|
| 594 |
+
"learning_rate": 7.36e-06,
|
| 595 |
+
"loss": 3.5966,
|
| 596 |
+
"step": 920
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 1.17,
|
| 600 |
+
"learning_rate": 7.44e-06,
|
| 601 |
+
"loss": 3.748,
|
| 602 |
+
"step": 930
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 1.19,
|
| 606 |
+
"learning_rate": 7.519999999999999e-06,
|
| 607 |
+
"loss": 3.7903,
|
| 608 |
+
"step": 940
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.2,
|
| 612 |
+
"learning_rate": 7.599999999999999e-06,
|
| 613 |
+
"loss": 3.7352,
|
| 614 |
+
"step": 950
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.21,
|
| 618 |
+
"learning_rate": 7.68e-06,
|
| 619 |
+
"loss": 3.5784,
|
| 620 |
+
"step": 960
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.23,
|
| 624 |
+
"learning_rate": 7.76e-06,
|
| 625 |
+
"loss": 3.5768,
|
| 626 |
+
"step": 970
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.24,
|
| 630 |
+
"learning_rate": 7.84e-06,
|
| 631 |
+
"loss": 3.7704,
|
| 632 |
+
"step": 980
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.25,
|
| 636 |
+
"learning_rate": 7.92e-06,
|
| 637 |
+
"loss": 3.6493,
|
| 638 |
+
"step": 990
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 1.26,
|
| 642 |
+
"learning_rate": 8e-06,
|
| 643 |
+
"loss": 3.6712,
|
| 644 |
+
"step": 1000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 1.26,
|
| 648 |
+
"eval_webgpt_accuracy": 0.36571671085272284,
|
| 649 |
+
"eval_webgpt_loss": 3.766130208969116,
|
| 650 |
+
"eval_webgpt_runtime": 39.6543,
|
| 651 |
+
"eval_webgpt_samples_per_second": 89.852,
|
| 652 |
+
"eval_webgpt_steps_per_second": 22.469,
|
| 653 |
+
"step": 1000
|
| 654 |
+
}
|
| 655 |
+
],
|
| 656 |
+
"max_steps": 2373,
|
| 657 |
+
"num_train_epochs": 3,
|
| 658 |
+
"total_flos": 1227277565165568.0,
|
| 659 |
+
"trial_name": null,
|
| 660 |
+
"trial_params": null
|
| 661 |
+
}
|
checkpoint-1000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d1417a2d058e79629fcaab14e3d9fabd6e4e21479447cbffb50c01997da34240
|
| 3 |
+
size 3515
|
checkpoint-2000/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "EleutherAI/pythia-70m-deduped",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"GPTNeoXForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"eos_token_id": 0,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_size": 512,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 2048,
|
| 12 |
+
"layer_norm_eps": 1e-05,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "gpt_neox",
|
| 15 |
+
"num_attention_heads": 8,
|
| 16 |
+
"num_hidden_layers": 6,
|
| 17 |
+
"rotary_emb_base": 10000,
|
| 18 |
+
"rotary_pct": 0.25,
|
| 19 |
+
"tie_word_embeddings": false,
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0.dev0",
|
| 22 |
+
"use_cache": true,
|
| 23 |
+
"use_parallel_residual": true,
|
| 24 |
+
"vocab_size": 50288
|
| 25 |
+
}
|
checkpoint-2000/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"eos_token_id": 0,
|
| 5 |
+
"transformers_version": "4.28.0.dev0"
|
| 6 |
+
}
|
checkpoint-2000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d1a3a7a00af0df7fe1324c03786b2edc775658290ec7d05141e29ea4e50c5cc4
|
| 3 |
+
size 563326405
|
checkpoint-2000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4fb4001ba611cf90180b52080e486191af146cfcf59b80bd4f279250a42f8cb
|
| 3 |
+
size 306838155
|
checkpoint-2000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d281150009ff5a1b39c6921325fc94924b7db4d40b63f3a979faec74a64a5915
|
| 3 |
+
size 14575
|
checkpoint-2000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:faea6a21c95612a8a10023dab14944867c642fe7ec5c77878ade9e37a58b4bca
|
| 3 |
+
size 627
|
checkpoint-2000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<|prefix_begin|>",
|
| 4 |
+
"<|system|>",
|
| 5 |
+
"<|prefix_end|>",
|
| 6 |
+
"<|assistant|>",
|
| 7 |
+
"<|prompter|>"
|
| 8 |
+
],
|
| 9 |
+
"bos_token": "<|endoftext|>",
|
| 10 |
+
"eos_token": "<|endoftext|>",
|
| 11 |
+
"pad_token": "<|padding|>",
|
| 12 |
+
"sep_token": "<|endoftext|>",
|
| 13 |
+
"unk_token": "<|endoftext|>"
|
| 14 |
+
}
|
checkpoint-2000/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-2000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"bos_token": "<|endoftext|>",
|
| 4 |
+
"clean_up_tokenization_spaces": true,
|
| 5 |
+
"eos_token": "<|endoftext|>",
|
| 6 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 7 |
+
"special_tokens_map_file": "/admin/home-hailey/.cache/huggingface/hub/models--EleutherAI--gpt-neox-20b/snapshots/4e49eadb5d14bd22f314ec3f45b69a87b88c7691/special_tokens_map.json",
|
| 8 |
+
"tokenizer_class": "GPTNeoXTokenizer",
|
| 9 |
+
"unk_token": "<|endoftext|>"
|
| 10 |
+
}
|
checkpoint-2000/trainer_state.json
ADDED
|
@@ -0,0 +1,1306 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.5259612685938815,
|
| 5 |
+
"global_step": 2000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 8e-08,
|
| 13 |
+
"loss": 11.8368,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.03,
|
| 18 |
+
"learning_rate": 1.6e-07,
|
| 19 |
+
"loss": 11.103,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.04,
|
| 24 |
+
"learning_rate": 2.4e-07,
|
| 25 |
+
"loss": 11.6842,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.05,
|
| 30 |
+
"learning_rate": 3.2e-07,
|
| 31 |
+
"loss": 11.7279,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.06,
|
| 36 |
+
"learning_rate": 4e-07,
|
| 37 |
+
"loss": 11.6439,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.08,
|
| 42 |
+
"learning_rate": 4.8e-07,
|
| 43 |
+
"loss": 11.1668,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.09,
|
| 48 |
+
"learning_rate": 5.6e-07,
|
| 49 |
+
"loss": 11.5542,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.1,
|
| 54 |
+
"learning_rate": 6.4e-07,
|
| 55 |
+
"loss": 11.2201,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.11,
|
| 60 |
+
"learning_rate": 7.2e-07,
|
| 61 |
+
"loss": 10.0118,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.13,
|
| 66 |
+
"learning_rate": 8e-07,
|
| 67 |
+
"loss": 9.2907,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.14,
|
| 72 |
+
"learning_rate": 8.799999999999999e-07,
|
| 73 |
+
"loss": 8.9216,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.15,
|
| 78 |
+
"learning_rate": 9.6e-07,
|
| 79 |
+
"loss": 8.9268,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.16,
|
| 84 |
+
"learning_rate": 1.04e-06,
|
| 85 |
+
"loss": 8.6545,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.18,
|
| 90 |
+
"learning_rate": 1.12e-06,
|
| 91 |
+
"loss": 8.7888,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.19,
|
| 96 |
+
"learning_rate": 1.2e-06,
|
| 97 |
+
"loss": 9.0201,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.2,
|
| 102 |
+
"learning_rate": 1.28e-06,
|
| 103 |
+
"loss": 8.0512,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.21,
|
| 108 |
+
"learning_rate": 1.3600000000000001e-06,
|
| 109 |
+
"loss": 5.9266,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.23,
|
| 114 |
+
"learning_rate": 1.44e-06,
|
| 115 |
+
"loss": 5.3527,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.24,
|
| 120 |
+
"learning_rate": 1.5199999999999998e-06,
|
| 121 |
+
"loss": 4.9847,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.25,
|
| 126 |
+
"learning_rate": 1.6e-06,
|
| 127 |
+
"loss": 4.8117,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.25,
|
| 132 |
+
"eval_webgpt_accuracy": 0.3513391153050626,
|
| 133 |
+
"eval_webgpt_loss": 4.794132709503174,
|
| 134 |
+
"eval_webgpt_runtime": 39.5471,
|
| 135 |
+
"eval_webgpt_samples_per_second": 90.095,
|
| 136 |
+
"eval_webgpt_steps_per_second": 22.53,
|
| 137 |
+
"step": 200
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.27,
|
| 141 |
+
"learning_rate": 1.6799999999999998e-06,
|
| 142 |
+
"loss": 4.7854,
|
| 143 |
+
"step": 210
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.28,
|
| 147 |
+
"learning_rate": 1.7599999999999999e-06,
|
| 148 |
+
"loss": 4.5555,
|
| 149 |
+
"step": 220
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.29,
|
| 153 |
+
"learning_rate": 1.84e-06,
|
| 154 |
+
"loss": 4.4819,
|
| 155 |
+
"step": 230
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.3,
|
| 159 |
+
"learning_rate": 1.92e-06,
|
| 160 |
+
"loss": 4.3504,
|
| 161 |
+
"step": 240
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.32,
|
| 165 |
+
"learning_rate": 2e-06,
|
| 166 |
+
"loss": 4.1471,
|
| 167 |
+
"step": 250
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 0.33,
|
| 171 |
+
"learning_rate": 2.08e-06,
|
| 172 |
+
"loss": 4.4132,
|
| 173 |
+
"step": 260
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 0.34,
|
| 177 |
+
"learning_rate": 2.16e-06,
|
| 178 |
+
"loss": 4.1309,
|
| 179 |
+
"step": 270
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.35,
|
| 183 |
+
"learning_rate": 2.24e-06,
|
| 184 |
+
"loss": 4.4153,
|
| 185 |
+
"step": 280
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.37,
|
| 189 |
+
"learning_rate": 2.32e-06,
|
| 190 |
+
"loss": 4.3256,
|
| 191 |
+
"step": 290
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.38,
|
| 195 |
+
"learning_rate": 2.4e-06,
|
| 196 |
+
"loss": 4.3798,
|
| 197 |
+
"step": 300
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.39,
|
| 201 |
+
"learning_rate": 2.48e-06,
|
| 202 |
+
"loss": 4.3665,
|
| 203 |
+
"step": 310
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.4,
|
| 207 |
+
"learning_rate": 2.56e-06,
|
| 208 |
+
"loss": 4.2545,
|
| 209 |
+
"step": 320
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.42,
|
| 213 |
+
"learning_rate": 2.64e-06,
|
| 214 |
+
"loss": 4.1912,
|
| 215 |
+
"step": 330
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.43,
|
| 219 |
+
"learning_rate": 2.7200000000000002e-06,
|
| 220 |
+
"loss": 4.2359,
|
| 221 |
+
"step": 340
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.44,
|
| 225 |
+
"learning_rate": 2.8e-06,
|
| 226 |
+
"loss": 4.2145,
|
| 227 |
+
"step": 350
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.45,
|
| 231 |
+
"learning_rate": 2.88e-06,
|
| 232 |
+
"loss": 4.0716,
|
| 233 |
+
"step": 360
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.47,
|
| 237 |
+
"learning_rate": 2.96e-06,
|
| 238 |
+
"loss": 4.0429,
|
| 239 |
+
"step": 370
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.48,
|
| 243 |
+
"learning_rate": 3.0399999999999997e-06,
|
| 244 |
+
"loss": 4.1404,
|
| 245 |
+
"step": 380
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 0.49,
|
| 249 |
+
"learning_rate": 3.1199999999999998e-06,
|
| 250 |
+
"loss": 3.9878,
|
| 251 |
+
"step": 390
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.51,
|
| 255 |
+
"learning_rate": 3.2e-06,
|
| 256 |
+
"loss": 4.1169,
|
| 257 |
+
"step": 400
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.51,
|
| 261 |
+
"eval_webgpt_accuracy": 0.3585799846229516,
|
| 262 |
+
"eval_webgpt_loss": 3.982567310333252,
|
| 263 |
+
"eval_webgpt_runtime": 39.5324,
|
| 264 |
+
"eval_webgpt_samples_per_second": 90.129,
|
| 265 |
+
"eval_webgpt_steps_per_second": 22.538,
|
| 266 |
+
"step": 400
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.52,
|
| 270 |
+
"learning_rate": 3.2799999999999995e-06,
|
| 271 |
+
"loss": 3.9303,
|
| 272 |
+
"step": 410
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.53,
|
| 276 |
+
"learning_rate": 3.3599999999999996e-06,
|
| 277 |
+
"loss": 4.0221,
|
| 278 |
+
"step": 420
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.54,
|
| 282 |
+
"learning_rate": 3.4399999999999997e-06,
|
| 283 |
+
"loss": 3.8958,
|
| 284 |
+
"step": 430
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.56,
|
| 288 |
+
"learning_rate": 3.5199999999999998e-06,
|
| 289 |
+
"loss": 3.8434,
|
| 290 |
+
"step": 440
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.57,
|
| 294 |
+
"learning_rate": 3.6e-06,
|
| 295 |
+
"loss": 4.0083,
|
| 296 |
+
"step": 450
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.58,
|
| 300 |
+
"learning_rate": 3.68e-06,
|
| 301 |
+
"loss": 3.9172,
|
| 302 |
+
"step": 460
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.59,
|
| 306 |
+
"learning_rate": 3.7599999999999996e-06,
|
| 307 |
+
"loss": 4.0955,
|
| 308 |
+
"step": 470
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.61,
|
| 312 |
+
"learning_rate": 3.84e-06,
|
| 313 |
+
"loss": 4.1085,
|
| 314 |
+
"step": 480
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.62,
|
| 318 |
+
"learning_rate": 3.92e-06,
|
| 319 |
+
"loss": 4.149,
|
| 320 |
+
"step": 490
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.63,
|
| 324 |
+
"learning_rate": 4e-06,
|
| 325 |
+
"loss": 4.0345,
|
| 326 |
+
"step": 500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.64,
|
| 330 |
+
"learning_rate": 4.08e-06,
|
| 331 |
+
"loss": 3.9929,
|
| 332 |
+
"step": 510
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.66,
|
| 336 |
+
"learning_rate": 4.16e-06,
|
| 337 |
+
"loss": 3.7969,
|
| 338 |
+
"step": 520
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.67,
|
| 342 |
+
"learning_rate": 4.24e-06,
|
| 343 |
+
"loss": 3.6102,
|
| 344 |
+
"step": 530
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.68,
|
| 348 |
+
"learning_rate": 4.32e-06,
|
| 349 |
+
"loss": 3.744,
|
| 350 |
+
"step": 540
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.69,
|
| 354 |
+
"learning_rate": 4.4e-06,
|
| 355 |
+
"loss": 3.8242,
|
| 356 |
+
"step": 550
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.71,
|
| 360 |
+
"learning_rate": 4.48e-06,
|
| 361 |
+
"loss": 3.8092,
|
| 362 |
+
"step": 560
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.72,
|
| 366 |
+
"learning_rate": 4.5599999999999995e-06,
|
| 367 |
+
"loss": 3.8835,
|
| 368 |
+
"step": 570
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.73,
|
| 372 |
+
"learning_rate": 4.64e-06,
|
| 373 |
+
"loss": 3.7335,
|
| 374 |
+
"step": 580
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.75,
|
| 378 |
+
"learning_rate": 4.72e-06,
|
| 379 |
+
"loss": 3.8462,
|
| 380 |
+
"step": 590
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.76,
|
| 384 |
+
"learning_rate": 4.8e-06,
|
| 385 |
+
"loss": 3.7422,
|
| 386 |
+
"step": 600
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.76,
|
| 390 |
+
"eval_webgpt_accuracy": 0.36334147099294245,
|
| 391 |
+
"eval_webgpt_loss": 3.8047080039978027,
|
| 392 |
+
"eval_webgpt_runtime": 39.6534,
|
| 393 |
+
"eval_webgpt_samples_per_second": 89.854,
|
| 394 |
+
"eval_webgpt_steps_per_second": 22.47,
|
| 395 |
+
"step": 600
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.77,
|
| 399 |
+
"learning_rate": 4.88e-06,
|
| 400 |
+
"loss": 3.7068,
|
| 401 |
+
"step": 610
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.78,
|
| 405 |
+
"learning_rate": 4.96e-06,
|
| 406 |
+
"loss": 3.69,
|
| 407 |
+
"step": 620
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.8,
|
| 411 |
+
"learning_rate": 5.04e-06,
|
| 412 |
+
"loss": 3.65,
|
| 413 |
+
"step": 630
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.81,
|
| 417 |
+
"learning_rate": 5.12e-06,
|
| 418 |
+
"loss": 3.6972,
|
| 419 |
+
"step": 640
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.82,
|
| 423 |
+
"learning_rate": 5.2e-06,
|
| 424 |
+
"loss": 3.7907,
|
| 425 |
+
"step": 650
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.83,
|
| 429 |
+
"learning_rate": 5.28e-06,
|
| 430 |
+
"loss": 3.7505,
|
| 431 |
+
"step": 660
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 0.85,
|
| 435 |
+
"learning_rate": 5.36e-06,
|
| 436 |
+
"loss": 3.8109,
|
| 437 |
+
"step": 670
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.86,
|
| 441 |
+
"learning_rate": 5.4400000000000004e-06,
|
| 442 |
+
"loss": 3.8612,
|
| 443 |
+
"step": 680
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.87,
|
| 447 |
+
"learning_rate": 5.52e-06,
|
| 448 |
+
"loss": 3.7077,
|
| 449 |
+
"step": 690
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.88,
|
| 453 |
+
"learning_rate": 5.6e-06,
|
| 454 |
+
"loss": 3.8546,
|
| 455 |
+
"step": 700
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.9,
|
| 459 |
+
"learning_rate": 5.68e-06,
|
| 460 |
+
"loss": 3.8801,
|
| 461 |
+
"step": 710
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.91,
|
| 465 |
+
"learning_rate": 5.76e-06,
|
| 466 |
+
"loss": 3.7939,
|
| 467 |
+
"step": 720
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 0.92,
|
| 471 |
+
"learning_rate": 5.84e-06,
|
| 472 |
+
"loss": 3.6499,
|
| 473 |
+
"step": 730
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.93,
|
| 477 |
+
"learning_rate": 5.92e-06,
|
| 478 |
+
"loss": 4.0742,
|
| 479 |
+
"step": 740
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.95,
|
| 483 |
+
"learning_rate": 6e-06,
|
| 484 |
+
"loss": 3.8072,
|
| 485 |
+
"step": 750
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.96,
|
| 489 |
+
"learning_rate": 6.079999999999999e-06,
|
| 490 |
+
"loss": 3.7959,
|
| 491 |
+
"step": 760
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.97,
|
| 495 |
+
"learning_rate": 6.1599999999999995e-06,
|
| 496 |
+
"loss": 3.7556,
|
| 497 |
+
"step": 770
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.99,
|
| 501 |
+
"learning_rate": 6.2399999999999995e-06,
|
| 502 |
+
"loss": 3.8263,
|
| 503 |
+
"step": 780
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 1.0,
|
| 507 |
+
"learning_rate": 6.32e-06,
|
| 508 |
+
"loss": 3.9162,
|
| 509 |
+
"step": 790
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 1.01,
|
| 513 |
+
"learning_rate": 6.4e-06,
|
| 514 |
+
"loss": 3.7188,
|
| 515 |
+
"step": 800
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.01,
|
| 519 |
+
"eval_webgpt_accuracy": 0.3654184548784667,
|
| 520 |
+
"eval_webgpt_loss": 3.7724239826202393,
|
| 521 |
+
"eval_webgpt_runtime": 39.717,
|
| 522 |
+
"eval_webgpt_samples_per_second": 89.71,
|
| 523 |
+
"eval_webgpt_steps_per_second": 22.434,
|
| 524 |
+
"step": 800
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 1.02,
|
| 528 |
+
"learning_rate": 6.48e-06,
|
| 529 |
+
"loss": 3.7938,
|
| 530 |
+
"step": 810
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 1.04,
|
| 534 |
+
"learning_rate": 6.559999999999999e-06,
|
| 535 |
+
"loss": 3.6401,
|
| 536 |
+
"step": 820
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.05,
|
| 540 |
+
"learning_rate": 6.639999999999999e-06,
|
| 541 |
+
"loss": 3.7234,
|
| 542 |
+
"step": 830
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.06,
|
| 546 |
+
"learning_rate": 6.719999999999999e-06,
|
| 547 |
+
"loss": 3.6234,
|
| 548 |
+
"step": 840
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.07,
|
| 552 |
+
"learning_rate": 6.799999999999999e-06,
|
| 553 |
+
"loss": 3.7614,
|
| 554 |
+
"step": 850
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 1.09,
|
| 558 |
+
"learning_rate": 6.879999999999999e-06,
|
| 559 |
+
"loss": 3.7001,
|
| 560 |
+
"step": 860
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.1,
|
| 564 |
+
"learning_rate": 6.9599999999999994e-06,
|
| 565 |
+
"loss": 3.8073,
|
| 566 |
+
"step": 870
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 1.11,
|
| 570 |
+
"learning_rate": 7.0399999999999995e-06,
|
| 571 |
+
"loss": 3.6328,
|
| 572 |
+
"step": 880
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 1.12,
|
| 576 |
+
"learning_rate": 7.12e-06,
|
| 577 |
+
"loss": 3.9974,
|
| 578 |
+
"step": 890
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.14,
|
| 582 |
+
"learning_rate": 7.2e-06,
|
| 583 |
+
"loss": 3.6881,
|
| 584 |
+
"step": 900
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.15,
|
| 588 |
+
"learning_rate": 7.28e-06,
|
| 589 |
+
"loss": 3.7669,
|
| 590 |
+
"step": 910
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.16,
|
| 594 |
+
"learning_rate": 7.36e-06,
|
| 595 |
+
"loss": 3.5966,
|
| 596 |
+
"step": 920
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 1.17,
|
| 600 |
+
"learning_rate": 7.44e-06,
|
| 601 |
+
"loss": 3.748,
|
| 602 |
+
"step": 930
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 1.19,
|
| 606 |
+
"learning_rate": 7.519999999999999e-06,
|
| 607 |
+
"loss": 3.7903,
|
| 608 |
+
"step": 940
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.2,
|
| 612 |
+
"learning_rate": 7.599999999999999e-06,
|
| 613 |
+
"loss": 3.7352,
|
| 614 |
+
"step": 950
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.21,
|
| 618 |
+
"learning_rate": 7.68e-06,
|
| 619 |
+
"loss": 3.5784,
|
| 620 |
+
"step": 960
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.23,
|
| 624 |
+
"learning_rate": 7.76e-06,
|
| 625 |
+
"loss": 3.5768,
|
| 626 |
+
"step": 970
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.24,
|
| 630 |
+
"learning_rate": 7.84e-06,
|
| 631 |
+
"loss": 3.7704,
|
| 632 |
+
"step": 980
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.25,
|
| 636 |
+
"learning_rate": 7.92e-06,
|
| 637 |
+
"loss": 3.6493,
|
| 638 |
+
"step": 990
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 1.26,
|
| 642 |
+
"learning_rate": 8e-06,
|
| 643 |
+
"loss": 3.6712,
|
| 644 |
+
"step": 1000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 1.26,
|
| 648 |
+
"eval_webgpt_accuracy": 0.36571671085272284,
|
| 649 |
+
"eval_webgpt_loss": 3.766130208969116,
|
| 650 |
+
"eval_webgpt_runtime": 39.6543,
|
| 651 |
+
"eval_webgpt_samples_per_second": 89.852,
|
| 652 |
+
"eval_webgpt_steps_per_second": 22.469,
|
| 653 |
+
"step": 1000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.28,
|
| 657 |
+
"learning_rate": 7.941733430444282e-06,
|
| 658 |
+
"loss": 3.8326,
|
| 659 |
+
"step": 1010
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 1.29,
|
| 663 |
+
"learning_rate": 7.883466860888565e-06,
|
| 664 |
+
"loss": 3.6731,
|
| 665 |
+
"step": 1020
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 1.3,
|
| 669 |
+
"learning_rate": 7.825200291332846e-06,
|
| 670 |
+
"loss": 3.6476,
|
| 671 |
+
"step": 1030
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 1.31,
|
| 675 |
+
"learning_rate": 7.76693372177713e-06,
|
| 676 |
+
"loss": 3.7165,
|
| 677 |
+
"step": 1040
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 1.33,
|
| 681 |
+
"learning_rate": 7.708667152221412e-06,
|
| 682 |
+
"loss": 3.6997,
|
| 683 |
+
"step": 1050
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.34,
|
| 687 |
+
"learning_rate": 7.650400582665694e-06,
|
| 688 |
+
"loss": 3.7811,
|
| 689 |
+
"step": 1060
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.35,
|
| 693 |
+
"learning_rate": 7.592134013109977e-06,
|
| 694 |
+
"loss": 3.6025,
|
| 695 |
+
"step": 1070
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.36,
|
| 699 |
+
"learning_rate": 7.533867443554261e-06,
|
| 700 |
+
"loss": 3.8738,
|
| 701 |
+
"step": 1080
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 1.38,
|
| 705 |
+
"learning_rate": 7.475600873998543e-06,
|
| 706 |
+
"loss": 3.7493,
|
| 707 |
+
"step": 1090
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 1.39,
|
| 711 |
+
"learning_rate": 7.4173343044428255e-06,
|
| 712 |
+
"loss": 3.8542,
|
| 713 |
+
"step": 1100
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 1.4,
|
| 717 |
+
"learning_rate": 7.359067734887107e-06,
|
| 718 |
+
"loss": 3.4613,
|
| 719 |
+
"step": 1110
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 1.41,
|
| 723 |
+
"learning_rate": 7.300801165331391e-06,
|
| 724 |
+
"loss": 3.7582,
|
| 725 |
+
"step": 1120
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.43,
|
| 729 |
+
"learning_rate": 7.242534595775673e-06,
|
| 730 |
+
"loss": 3.6542,
|
| 731 |
+
"step": 1130
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.44,
|
| 735 |
+
"learning_rate": 7.184268026219956e-06,
|
| 736 |
+
"loss": 3.6867,
|
| 737 |
+
"step": 1140
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.45,
|
| 741 |
+
"learning_rate": 7.126001456664238e-06,
|
| 742 |
+
"loss": 3.6046,
|
| 743 |
+
"step": 1150
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 1.47,
|
| 747 |
+
"learning_rate": 7.067734887108521e-06,
|
| 748 |
+
"loss": 3.7645,
|
| 749 |
+
"step": 1160
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 1.48,
|
| 753 |
+
"learning_rate": 7.009468317552803e-06,
|
| 754 |
+
"loss": 3.8135,
|
| 755 |
+
"step": 1170
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 1.49,
|
| 759 |
+
"learning_rate": 6.951201747997087e-06,
|
| 760 |
+
"loss": 3.8123,
|
| 761 |
+
"step": 1180
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 1.5,
|
| 765 |
+
"learning_rate": 6.8929351784413686e-06,
|
| 766 |
+
"loss": 3.743,
|
| 767 |
+
"step": 1190
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.52,
|
| 771 |
+
"learning_rate": 6.834668608885651e-06,
|
| 772 |
+
"loss": 3.8658,
|
| 773 |
+
"step": 1200
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.52,
|
| 777 |
+
"eval_webgpt_accuracy": 0.36873122296639843,
|
| 778 |
+
"eval_webgpt_loss": 3.7487637996673584,
|
| 779 |
+
"eval_webgpt_runtime": 39.5934,
|
| 780 |
+
"eval_webgpt_samples_per_second": 89.99,
|
| 781 |
+
"eval_webgpt_steps_per_second": 22.504,
|
| 782 |
+
"step": 1200
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 1.53,
|
| 786 |
+
"learning_rate": 6.776402039329934e-06,
|
| 787 |
+
"loss": 3.7027,
|
| 788 |
+
"step": 1210
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.54,
|
| 792 |
+
"learning_rate": 6.718135469774217e-06,
|
| 793 |
+
"loss": 3.6325,
|
| 794 |
+
"step": 1220
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.55,
|
| 798 |
+
"learning_rate": 6.659868900218499e-06,
|
| 799 |
+
"loss": 3.7518,
|
| 800 |
+
"step": 1230
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.57,
|
| 804 |
+
"learning_rate": 6.601602330662782e-06,
|
| 805 |
+
"loss": 3.5668,
|
| 806 |
+
"step": 1240
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 1.58,
|
| 810 |
+
"learning_rate": 6.543335761107064e-06,
|
| 811 |
+
"loss": 3.7264,
|
| 812 |
+
"step": 1250
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 1.59,
|
| 816 |
+
"learning_rate": 6.485069191551347e-06,
|
| 817 |
+
"loss": 3.5763,
|
| 818 |
+
"step": 1260
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 1.6,
|
| 822 |
+
"learning_rate": 6.426802621995629e-06,
|
| 823 |
+
"loss": 3.7075,
|
| 824 |
+
"step": 1270
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 1.62,
|
| 828 |
+
"learning_rate": 6.3685360524399125e-06,
|
| 829 |
+
"loss": 3.6597,
|
| 830 |
+
"step": 1280
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.63,
|
| 834 |
+
"learning_rate": 6.310269482884194e-06,
|
| 835 |
+
"loss": 3.8605,
|
| 836 |
+
"step": 1290
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.64,
|
| 840 |
+
"learning_rate": 6.252002913328477e-06,
|
| 841 |
+
"loss": 3.7568,
|
| 842 |
+
"step": 1300
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.65,
|
| 846 |
+
"learning_rate": 6.19373634377276e-06,
|
| 847 |
+
"loss": 3.6982,
|
| 848 |
+
"step": 1310
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 1.67,
|
| 852 |
+
"learning_rate": 6.135469774217043e-06,
|
| 853 |
+
"loss": 3.8453,
|
| 854 |
+
"step": 1320
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 1.68,
|
| 858 |
+
"learning_rate": 6.0772032046613246e-06,
|
| 859 |
+
"loss": 3.8372,
|
| 860 |
+
"step": 1330
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 1.69,
|
| 864 |
+
"learning_rate": 6.018936635105608e-06,
|
| 865 |
+
"loss": 3.5192,
|
| 866 |
+
"step": 1340
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.71,
|
| 870 |
+
"learning_rate": 5.96067006554989e-06,
|
| 871 |
+
"loss": 3.7037,
|
| 872 |
+
"step": 1350
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.72,
|
| 876 |
+
"learning_rate": 5.902403495994173e-06,
|
| 877 |
+
"loss": 3.4107,
|
| 878 |
+
"step": 1360
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.73,
|
| 882 |
+
"learning_rate": 5.844136926438456e-06,
|
| 883 |
+
"loss": 3.5911,
|
| 884 |
+
"step": 1370
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.74,
|
| 888 |
+
"learning_rate": 5.785870356882738e-06,
|
| 889 |
+
"loss": 3.7885,
|
| 890 |
+
"step": 1380
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 1.76,
|
| 894 |
+
"learning_rate": 5.72760378732702e-06,
|
| 895 |
+
"loss": 3.5698,
|
| 896 |
+
"step": 1390
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 1.77,
|
| 900 |
+
"learning_rate": 5.669337217771304e-06,
|
| 901 |
+
"loss": 3.6686,
|
| 902 |
+
"step": 1400
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 1.77,
|
| 906 |
+
"eval_webgpt_accuracy": 0.3689563590269814,
|
| 907 |
+
"eval_webgpt_loss": 3.7346436977386475,
|
| 908 |
+
"eval_webgpt_runtime": 39.5649,
|
| 909 |
+
"eval_webgpt_samples_per_second": 90.055,
|
| 910 |
+
"eval_webgpt_steps_per_second": 22.52,
|
| 911 |
+
"step": 1400
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 1.78,
|
| 915 |
+
"learning_rate": 5.611070648215586e-06,
|
| 916 |
+
"loss": 3.7345,
|
| 917 |
+
"step": 1410
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 1.79,
|
| 921 |
+
"learning_rate": 5.5528040786598685e-06,
|
| 922 |
+
"loss": 3.8317,
|
| 923 |
+
"step": 1420
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 1.81,
|
| 927 |
+
"learning_rate": 5.49453750910415e-06,
|
| 928 |
+
"loss": 3.7166,
|
| 929 |
+
"step": 1430
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 1.82,
|
| 933 |
+
"learning_rate": 5.436270939548434e-06,
|
| 934 |
+
"loss": 3.7229,
|
| 935 |
+
"step": 1440
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.83,
|
| 939 |
+
"learning_rate": 5.378004369992716e-06,
|
| 940 |
+
"loss": 3.4168,
|
| 941 |
+
"step": 1450
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.84,
|
| 945 |
+
"learning_rate": 5.319737800436999e-06,
|
| 946 |
+
"loss": 3.9726,
|
| 947 |
+
"step": 1460
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.86,
|
| 951 |
+
"learning_rate": 5.2614712308812814e-06,
|
| 952 |
+
"loss": 3.8361,
|
| 953 |
+
"step": 1470
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 1.87,
|
| 957 |
+
"learning_rate": 5.203204661325564e-06,
|
| 958 |
+
"loss": 3.5802,
|
| 959 |
+
"step": 1480
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 1.88,
|
| 963 |
+
"learning_rate": 5.144938091769846e-06,
|
| 964 |
+
"loss": 3.9139,
|
| 965 |
+
"step": 1490
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 1.89,
|
| 969 |
+
"learning_rate": 5.08667152221413e-06,
|
| 970 |
+
"loss": 3.5712,
|
| 971 |
+
"step": 1500
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 1.91,
|
| 975 |
+
"learning_rate": 5.028404952658412e-06,
|
| 976 |
+
"loss": 3.6163,
|
| 977 |
+
"step": 1510
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 1.92,
|
| 981 |
+
"learning_rate": 4.970138383102694e-06,
|
| 982 |
+
"loss": 3.5838,
|
| 983 |
+
"step": 1520
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.93,
|
| 987 |
+
"learning_rate": 4.911871813546977e-06,
|
| 988 |
+
"loss": 3.8421,
|
| 989 |
+
"step": 1530
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.94,
|
| 993 |
+
"learning_rate": 4.85360524399126e-06,
|
| 994 |
+
"loss": 3.6685,
|
| 995 |
+
"step": 1540
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 1.96,
|
| 999 |
+
"learning_rate": 4.795338674435542e-06,
|
| 1000 |
+
"loss": 3.6735,
|
| 1001 |
+
"step": 1550
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 1.97,
|
| 1005 |
+
"learning_rate": 4.737072104879825e-06,
|
| 1006 |
+
"loss": 3.7081,
|
| 1007 |
+
"step": 1560
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 1.98,
|
| 1011 |
+
"learning_rate": 4.678805535324107e-06,
|
| 1012 |
+
"loss": 3.7911,
|
| 1013 |
+
"step": 1570
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 2.0,
|
| 1017 |
+
"learning_rate": 4.62053896576839e-06,
|
| 1018 |
+
"loss": 3.7296,
|
| 1019 |
+
"step": 1580
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 2.01,
|
| 1023 |
+
"learning_rate": 4.562272396212672e-06,
|
| 1024 |
+
"loss": 3.6247,
|
| 1025 |
+
"step": 1590
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 2.02,
|
| 1029 |
+
"learning_rate": 4.5040058266569555e-06,
|
| 1030 |
+
"loss": 3.5536,
|
| 1031 |
+
"step": 1600
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 2.02,
|
| 1035 |
+
"eval_webgpt_accuracy": 0.3708860692093479,
|
| 1036 |
+
"eval_webgpt_loss": 3.709535837173462,
|
| 1037 |
+
"eval_webgpt_runtime": 39.7168,
|
| 1038 |
+
"eval_webgpt_samples_per_second": 89.71,
|
| 1039 |
+
"eval_webgpt_steps_per_second": 22.434,
|
| 1040 |
+
"step": 1600
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 2.03,
|
| 1044 |
+
"learning_rate": 4.4457392571012374e-06,
|
| 1045 |
+
"loss": 3.3749,
|
| 1046 |
+
"step": 1610
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.05,
|
| 1050 |
+
"learning_rate": 4.38747268754552e-06,
|
| 1051 |
+
"loss": 3.6204,
|
| 1052 |
+
"step": 1620
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.06,
|
| 1056 |
+
"learning_rate": 4.329206117989803e-06,
|
| 1057 |
+
"loss": 3.6534,
|
| 1058 |
+
"step": 1630
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 2.07,
|
| 1062 |
+
"learning_rate": 4.270939548434086e-06,
|
| 1063 |
+
"loss": 3.769,
|
| 1064 |
+
"step": 1640
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 2.08,
|
| 1068 |
+
"learning_rate": 4.212672978878368e-06,
|
| 1069 |
+
"loss": 3.4946,
|
| 1070 |
+
"step": 1650
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.1,
|
| 1074 |
+
"learning_rate": 4.154406409322651e-06,
|
| 1075 |
+
"loss": 3.4767,
|
| 1076 |
+
"step": 1660
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 2.11,
|
| 1080 |
+
"learning_rate": 4.096139839766933e-06,
|
| 1081 |
+
"loss": 4.0031,
|
| 1082 |
+
"step": 1670
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 2.12,
|
| 1086 |
+
"learning_rate": 4.037873270211216e-06,
|
| 1087 |
+
"loss": 3.5801,
|
| 1088 |
+
"step": 1680
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.13,
|
| 1092 |
+
"learning_rate": 3.979606700655499e-06,
|
| 1093 |
+
"loss": 3.7839,
|
| 1094 |
+
"step": 1690
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 2.15,
|
| 1098 |
+
"learning_rate": 3.921340131099781e-06,
|
| 1099 |
+
"loss": 3.389,
|
| 1100 |
+
"step": 1700
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 2.16,
|
| 1104 |
+
"learning_rate": 3.863073561544064e-06,
|
| 1105 |
+
"loss": 3.5539,
|
| 1106 |
+
"step": 1710
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 2.17,
|
| 1110 |
+
"learning_rate": 3.8048069919883465e-06,
|
| 1111 |
+
"loss": 3.7058,
|
| 1112 |
+
"step": 1720
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 2.18,
|
| 1116 |
+
"learning_rate": 3.746540422432629e-06,
|
| 1117 |
+
"loss": 3.6334,
|
| 1118 |
+
"step": 1730
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 2.2,
|
| 1122 |
+
"learning_rate": 3.6882738528769115e-06,
|
| 1123 |
+
"loss": 3.5749,
|
| 1124 |
+
"step": 1740
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 2.21,
|
| 1128 |
+
"learning_rate": 3.6300072833211943e-06,
|
| 1129 |
+
"loss": 3.5804,
|
| 1130 |
+
"step": 1750
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.22,
|
| 1134 |
+
"learning_rate": 3.571740713765477e-06,
|
| 1135 |
+
"loss": 3.5608,
|
| 1136 |
+
"step": 1760
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.24,
|
| 1140 |
+
"learning_rate": 3.5134741442097594e-06,
|
| 1141 |
+
"loss": 3.8894,
|
| 1142 |
+
"step": 1770
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 2.25,
|
| 1146 |
+
"learning_rate": 3.455207574654042e-06,
|
| 1147 |
+
"loss": 3.4811,
|
| 1148 |
+
"step": 1780
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 2.26,
|
| 1152 |
+
"learning_rate": 3.396941005098325e-06,
|
| 1153 |
+
"loss": 3.5773,
|
| 1154 |
+
"step": 1790
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 2.27,
|
| 1158 |
+
"learning_rate": 3.338674435542607e-06,
|
| 1159 |
+
"loss": 3.4083,
|
| 1160 |
+
"step": 1800
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 2.27,
|
| 1164 |
+
"eval_webgpt_accuracy": 0.37110112351569674,
|
| 1165 |
+
"eval_webgpt_loss": 3.7004425525665283,
|
| 1166 |
+
"eval_webgpt_runtime": 39.7103,
|
| 1167 |
+
"eval_webgpt_samples_per_second": 89.725,
|
| 1168 |
+
"eval_webgpt_steps_per_second": 22.438,
|
| 1169 |
+
"step": 1800
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 2.29,
|
| 1173 |
+
"learning_rate": 3.28040786598689e-06,
|
| 1174 |
+
"loss": 3.5236,
|
| 1175 |
+
"step": 1810
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 2.3,
|
| 1179 |
+
"learning_rate": 3.2221412964311723e-06,
|
| 1180 |
+
"loss": 3.6029,
|
| 1181 |
+
"step": 1820
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 2.31,
|
| 1185 |
+
"learning_rate": 3.163874726875455e-06,
|
| 1186 |
+
"loss": 3.6437,
|
| 1187 |
+
"step": 1830
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 2.32,
|
| 1191 |
+
"learning_rate": 3.105608157319738e-06,
|
| 1192 |
+
"loss": 3.5334,
|
| 1193 |
+
"step": 1840
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 2.34,
|
| 1197 |
+
"learning_rate": 3.04734158776402e-06,
|
| 1198 |
+
"loss": 3.5593,
|
| 1199 |
+
"step": 1850
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 2.35,
|
| 1203 |
+
"learning_rate": 2.989075018208303e-06,
|
| 1204 |
+
"loss": 3.8394,
|
| 1205 |
+
"step": 1860
|
| 1206 |
+
},
|
| 1207 |
+
{
|
| 1208 |
+
"epoch": 2.36,
|
| 1209 |
+
"learning_rate": 2.9308084486525856e-06,
|
| 1210 |
+
"loss": 3.5156,
|
| 1211 |
+
"step": 1870
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 2.37,
|
| 1215 |
+
"learning_rate": 2.872541879096868e-06,
|
| 1216 |
+
"loss": 3.5369,
|
| 1217 |
+
"step": 1880
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 2.39,
|
| 1221 |
+
"learning_rate": 2.8142753095411507e-06,
|
| 1222 |
+
"loss": 3.5477,
|
| 1223 |
+
"step": 1890
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 2.4,
|
| 1227 |
+
"learning_rate": 2.7560087399854335e-06,
|
| 1228 |
+
"loss": 3.6827,
|
| 1229 |
+
"step": 1900
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 2.41,
|
| 1233 |
+
"learning_rate": 2.697742170429716e-06,
|
| 1234 |
+
"loss": 3.5929,
|
| 1235 |
+
"step": 1910
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 2.42,
|
| 1239 |
+
"learning_rate": 2.6394756008739986e-06,
|
| 1240 |
+
"loss": 3.7087,
|
| 1241 |
+
"step": 1920
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 2.44,
|
| 1245 |
+
"learning_rate": 2.581209031318281e-06,
|
| 1246 |
+
"loss": 3.4699,
|
| 1247 |
+
"step": 1930
|
| 1248 |
+
},
|
| 1249 |
+
{
|
| 1250 |
+
"epoch": 2.45,
|
| 1251 |
+
"learning_rate": 2.5229424617625636e-06,
|
| 1252 |
+
"loss": 3.5238,
|
| 1253 |
+
"step": 1940
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 2.46,
|
| 1257 |
+
"learning_rate": 2.4646758922068464e-06,
|
| 1258 |
+
"loss": 3.6169,
|
| 1259 |
+
"step": 1950
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 2.48,
|
| 1263 |
+
"learning_rate": 2.4064093226511287e-06,
|
| 1264 |
+
"loss": 3.536,
|
| 1265 |
+
"step": 1960
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 2.49,
|
| 1269 |
+
"learning_rate": 2.3481427530954115e-06,
|
| 1270 |
+
"loss": 3.4735,
|
| 1271 |
+
"step": 1970
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 2.5,
|
| 1275 |
+
"learning_rate": 2.2898761835396942e-06,
|
| 1276 |
+
"loss": 3.616,
|
| 1277 |
+
"step": 1980
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 2.51,
|
| 1281 |
+
"learning_rate": 2.2316096139839766e-06,
|
| 1282 |
+
"loss": 3.4397,
|
| 1283 |
+
"step": 1990
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 2.53,
|
| 1287 |
+
"learning_rate": 2.1733430444282593e-06,
|
| 1288 |
+
"loss": 3.6732,
|
| 1289 |
+
"step": 2000
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 2.53,
|
| 1293 |
+
"eval_webgpt_accuracy": 0.37167770768266506,
|
| 1294 |
+
"eval_webgpt_loss": 3.6911776065826416,
|
| 1295 |
+
"eval_webgpt_runtime": 39.7365,
|
| 1296 |
+
"eval_webgpt_samples_per_second": 89.666,
|
| 1297 |
+
"eval_webgpt_steps_per_second": 22.423,
|
| 1298 |
+
"step": 2000
|
| 1299 |
+
}
|
| 1300 |
+
],
|
| 1301 |
+
"max_steps": 2373,
|
| 1302 |
+
"num_train_epochs": 3,
|
| 1303 |
+
"total_flos": 2456896174817280.0,
|
| 1304 |
+
"trial_name": null,
|
| 1305 |
+
"trial_params": null
|
| 1306 |
+
}
|
checkpoint-2000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d1417a2d058e79629fcaab14e3d9fabd6e4e21479447cbffb50c01997da34240
|
| 3 |
+
size 3515
|
config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "EleutherAI/pythia-70m-deduped",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"GPTNeoXForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"eos_token_id": 0,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_size": 512,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 2048,
|
| 12 |
+
"layer_norm_eps": 1e-05,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "gpt_neox",
|
| 15 |
+
"num_attention_heads": 8,
|
| 16 |
+
"num_hidden_layers": 6,
|
| 17 |
+
"rotary_emb_base": 10000,
|
| 18 |
+
"rotary_pct": 0.25,
|
| 19 |
+
"tie_word_embeddings": false,
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0.dev0",
|
| 22 |
+
"use_cache": true,
|
| 23 |
+
"use_parallel_residual": true,
|
| 24 |
+
"vocab_size": 50288
|
| 25 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"eos_token_id": 0,
|
| 5 |
+
"transformers_version": "4.28.0.dev0"
|
| 6 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c65e9104e29069dd415f699ef552a4e8c7080c2e23666b6bf9c335fa18925099
|
| 3 |
+
size 306838155
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<|prefix_begin|>",
|
| 4 |
+
"<|system|>",
|
| 5 |
+
"<|prefix_end|>",
|
| 6 |
+
"<|assistant|>",
|
| 7 |
+
"<|prompter|>"
|
| 8 |
+
],
|
| 9 |
+
"bos_token": "<|endoftext|>",
|
| 10 |
+
"eos_token": "<|endoftext|>",
|
| 11 |
+
"pad_token": "<|padding|>",
|
| 12 |
+
"sep_token": "<|endoftext|>",
|
| 13 |
+
"unk_token": "<|endoftext|>"
|
| 14 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"bos_token": "<|endoftext|>",
|
| 4 |
+
"clean_up_tokenization_spaces": true,
|
| 5 |
+
"eos_token": "<|endoftext|>",
|
| 6 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 7 |
+
"special_tokens_map_file": "/admin/home-hailey/.cache/huggingface/hub/models--EleutherAI--gpt-neox-20b/snapshots/4e49eadb5d14bd22f314ec3f45b69a87b88c7691/special_tokens_map.json",
|
| 8 |
+
"tokenizer_class": "GPTNeoXTokenizer",
|
| 9 |
+
"unk_token": "<|endoftext|>"
|
| 10 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d1417a2d058e79629fcaab14e3d9fabd6e4e21479447cbffb50c01997da34240
|
| 3 |
+
size 3515
|