Training in progress, epoch 18, checkpoint
Browse files- checkpoint-4464/added_tokens.json +3 -0
- checkpoint-4464/config.json +42 -0
- checkpoint-4464/generation_config.json +8 -0
- checkpoint-4464/merges.txt +0 -0
- checkpoint-4464/optimizer.pt +3 -0
- checkpoint-4464/pytorch_model.bin +3 -0
- checkpoint-4464/rng_state.pth +3 -0
- checkpoint-4464/scheduler.pt +3 -0
- checkpoint-4464/special_tokens_map.json +6 -0
- checkpoint-4464/tokenizer.json +0 -0
- checkpoint-4464/tokenizer_config.json +20 -0
- checkpoint-4464/trainer_state.json +895 -0
- checkpoint-4464/training_args.bin +3 -0
- checkpoint-4464/vocab.json +0 -0
checkpoint-4464/added_tokens.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<|endoftext|>": 50256
|
| 3 |
+
}
|
checkpoint-4464/config.json
ADDED
|
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "gpt2-xl",
|
| 3 |
+
"activation_function": "gelu_new",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"GPT2LMHeadModel"
|
| 6 |
+
],
|
| 7 |
+
"attn_pdrop": 0.1,
|
| 8 |
+
"bos_token_id": 50256,
|
| 9 |
+
"do_sample": true,
|
| 10 |
+
"embd_pdrop": 0.1,
|
| 11 |
+
"eos_token_id": 50256,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"layer_norm_epsilon": 1e-05,
|
| 14 |
+
"max_length": 50,
|
| 15 |
+
"model_type": "gpt2",
|
| 16 |
+
"n_ctx": 1024,
|
| 17 |
+
"n_embd": 1600,
|
| 18 |
+
"n_head": 25,
|
| 19 |
+
"n_inner": null,
|
| 20 |
+
"n_layer": 48,
|
| 21 |
+
"n_positions": 1024,
|
| 22 |
+
"output_past": true,
|
| 23 |
+
"reorder_and_upcast_attn": false,
|
| 24 |
+
"resid_pdrop": 0.1,
|
| 25 |
+
"scale_attn_by_inverse_layer_idx": false,
|
| 26 |
+
"scale_attn_weights": true,
|
| 27 |
+
"summary_activation": null,
|
| 28 |
+
"summary_first_dropout": 0.1,
|
| 29 |
+
"summary_proj_to_labels": true,
|
| 30 |
+
"summary_type": "cls_index",
|
| 31 |
+
"summary_use_proj": true,
|
| 32 |
+
"task_specific_params": {
|
| 33 |
+
"text-generation": {
|
| 34 |
+
"do_sample": true,
|
| 35 |
+
"max_length": 50
|
| 36 |
+
}
|
| 37 |
+
},
|
| 38 |
+
"torch_dtype": "float32",
|
| 39 |
+
"transformers_version": "4.34.0",
|
| 40 |
+
"use_cache": true,
|
| 41 |
+
"vocab_size": 50257
|
| 42 |
+
}
|
checkpoint-4464/generation_config.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 50256,
|
| 4 |
+
"do_sample": true,
|
| 5 |
+
"eos_token_id": 50256,
|
| 6 |
+
"max_length": 50,
|
| 7 |
+
"transformers_version": "4.34.0"
|
| 8 |
+
}
|
checkpoint-4464/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-4464/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:095e8668e214f07f70ebdc780a2727d3392a64d85a57739b763adc24e2739fad
|
| 3 |
+
size 12461385454
|
checkpoint-4464/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c555bf5fe3e458cde242705a7b21d8bb75b46e300887a0abb53341c32662cc84
|
| 3 |
+
size 6230637102
|
checkpoint-4464/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4d8d1973d4cef01b39c6c9ba103b2a92389aea783666f8d03a28b8769ea0ae04
|
| 3 |
+
size 14244
|
checkpoint-4464/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2a6ad7defe23aec3c5110c788b398a4402830bbd860fdcd099a3f0f6f6f54766
|
| 3 |
+
size 1064
|
checkpoint-4464/special_tokens_map.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<|endoftext|>",
|
| 3 |
+
"eos_token": "<|endoftext|>",
|
| 4 |
+
"pad_token": "<|endoftext|>",
|
| 5 |
+
"unk_token": "<|endoftext|>"
|
| 6 |
+
}
|
checkpoint-4464/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-4464/tokenizer_config.json
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"50256": {
|
| 5 |
+
"content": "<|endoftext|>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": false,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
}
|
| 12 |
+
},
|
| 13 |
+
"additional_special_tokens": [],
|
| 14 |
+
"bos_token": "<|endoftext|>",
|
| 15 |
+
"clean_up_tokenization_spaces": true,
|
| 16 |
+
"eos_token": "<|endoftext|>",
|
| 17 |
+
"model_max_length": 1024,
|
| 18 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 19 |
+
"unk_token": "<|endoftext|>"
|
| 20 |
+
}
|
checkpoint-4464/trainer_state.json
ADDED
|
@@ -0,0 +1,895 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 18.0,
|
| 5 |
+
"eval_steps": 248,
|
| 6 |
+
"global_step": 4464,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.2,
|
| 13 |
+
"learning_rate": 3e-05,
|
| 14 |
+
"loss": 2.8428,
|
| 15 |
+
"step": 50
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.4,
|
| 19 |
+
"learning_rate": 3e-05,
|
| 20 |
+
"loss": 2.1589,
|
| 21 |
+
"step": 100
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.6,
|
| 25 |
+
"learning_rate": 3e-05,
|
| 26 |
+
"loss": 2.0527,
|
| 27 |
+
"step": 150
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.81,
|
| 31 |
+
"learning_rate": 3e-05,
|
| 32 |
+
"loss": 1.9823,
|
| 33 |
+
"step": 200
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 1.0,
|
| 37 |
+
"eval_accuracy": 0.6398457059679767,
|
| 38 |
+
"eval_loss": 1.6149228811264038,
|
| 39 |
+
"eval_runtime": 17.4504,
|
| 40 |
+
"eval_samples_per_second": 28.653,
|
| 41 |
+
"eval_steps_per_second": 1.834,
|
| 42 |
+
"step": 248
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 1.0,
|
| 46 |
+
"eval_exact_match": 9.0,
|
| 47 |
+
"eval_f1": 13.45666666666666,
|
| 48 |
+
"eval_qa_bleu": 7.063432441196736,
|
| 49 |
+
"eval_qa_exact_match": 0.09,
|
| 50 |
+
"eval_recite_bleu": 15.344658855358656,
|
| 51 |
+
"eval_recite_exact_match": 0.0,
|
| 52 |
+
"step": 248
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.01,
|
| 56 |
+
"learning_rate": 3e-05,
|
| 57 |
+
"loss": 1.9153,
|
| 58 |
+
"step": 250
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 1.21,
|
| 62 |
+
"learning_rate": 3e-05,
|
| 63 |
+
"loss": 1.49,
|
| 64 |
+
"step": 300
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 1.41,
|
| 68 |
+
"learning_rate": 3e-05,
|
| 69 |
+
"loss": 1.4406,
|
| 70 |
+
"step": 350
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.61,
|
| 74 |
+
"learning_rate": 3e-05,
|
| 75 |
+
"loss": 1.4161,
|
| 76 |
+
"step": 400
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 1.81,
|
| 80 |
+
"learning_rate": 3e-05,
|
| 81 |
+
"loss": 1.3868,
|
| 82 |
+
"step": 450
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 2.0,
|
| 86 |
+
"eval_accuracy": 0.6700320232896652,
|
| 87 |
+
"eval_loss": 1.1928881406784058,
|
| 88 |
+
"eval_runtime": 15.4486,
|
| 89 |
+
"eval_samples_per_second": 32.365,
|
| 90 |
+
"eval_steps_per_second": 2.071,
|
| 91 |
+
"step": 496
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 2.0,
|
| 95 |
+
"eval_exact_match": 9.2,
|
| 96 |
+
"eval_f1": 14.735098500533276,
|
| 97 |
+
"eval_qa_bleu": 7.255412763224127,
|
| 98 |
+
"eval_qa_exact_match": 0.088,
|
| 99 |
+
"eval_recite_bleu": 17.979965007898493,
|
| 100 |
+
"eval_recite_exact_match": 0.0,
|
| 101 |
+
"step": 496
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 2.02,
|
| 105 |
+
"learning_rate": 3e-05,
|
| 106 |
+
"loss": 1.309,
|
| 107 |
+
"step": 500
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 2.22,
|
| 111 |
+
"learning_rate": 3e-05,
|
| 112 |
+
"loss": 0.9401,
|
| 113 |
+
"step": 550
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 2.42,
|
| 117 |
+
"learning_rate": 3e-05,
|
| 118 |
+
"loss": 0.9429,
|
| 119 |
+
"step": 600
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.62,
|
| 123 |
+
"learning_rate": 3e-05,
|
| 124 |
+
"loss": 0.9115,
|
| 125 |
+
"step": 650
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 2.82,
|
| 129 |
+
"learning_rate": 3e-05,
|
| 130 |
+
"loss": 0.9102,
|
| 131 |
+
"step": 700
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 3.0,
|
| 135 |
+
"eval_accuracy": 0.6992489082969432,
|
| 136 |
+
"eval_loss": 0.8513368368148804,
|
| 137 |
+
"eval_runtime": 15.7421,
|
| 138 |
+
"eval_samples_per_second": 31.762,
|
| 139 |
+
"eval_steps_per_second": 2.033,
|
| 140 |
+
"step": 744
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 3.0,
|
| 144 |
+
"eval_exact_match": 10.2,
|
| 145 |
+
"eval_f1": 15.888450046685344,
|
| 146 |
+
"eval_qa_bleu": 7.2379640900667646,
|
| 147 |
+
"eval_qa_exact_match": 0.098,
|
| 148 |
+
"eval_recite_bleu": 20.876580135938546,
|
| 149 |
+
"eval_recite_exact_match": 0.0,
|
| 150 |
+
"step": 744
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 3.02,
|
| 154 |
+
"learning_rate": 3e-05,
|
| 155 |
+
"loss": 0.8302,
|
| 156 |
+
"step": 750
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 3.23,
|
| 160 |
+
"learning_rate": 3e-05,
|
| 161 |
+
"loss": 0.5813,
|
| 162 |
+
"step": 800
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 3.43,
|
| 166 |
+
"learning_rate": 3e-05,
|
| 167 |
+
"loss": 0.5844,
|
| 168 |
+
"step": 850
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 3.63,
|
| 172 |
+
"learning_rate": 3e-05,
|
| 173 |
+
"loss": 0.5776,
|
| 174 |
+
"step": 900
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 3.83,
|
| 178 |
+
"learning_rate": 3e-05,
|
| 179 |
+
"loss": 0.5869,
|
| 180 |
+
"step": 950
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 4.0,
|
| 184 |
+
"eval_accuracy": 0.721787481804949,
|
| 185 |
+
"eval_loss": 0.6180558204650879,
|
| 186 |
+
"eval_runtime": 15.5183,
|
| 187 |
+
"eval_samples_per_second": 32.22,
|
| 188 |
+
"eval_steps_per_second": 2.062,
|
| 189 |
+
"step": 992
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 4.0,
|
| 193 |
+
"eval_exact_match": 13.6,
|
| 194 |
+
"eval_f1": 19.457685255920552,
|
| 195 |
+
"eval_qa_bleu": 7.372812651620524,
|
| 196 |
+
"eval_qa_exact_match": 0.126,
|
| 197 |
+
"eval_recite_bleu": 24.119469757278573,
|
| 198 |
+
"eval_recite_exact_match": 0.0,
|
| 199 |
+
"step": 992
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 4.03,
|
| 203 |
+
"learning_rate": 3e-05,
|
| 204 |
+
"loss": 0.5394,
|
| 205 |
+
"step": 1000
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 4.23,
|
| 209 |
+
"learning_rate": 3e-05,
|
| 210 |
+
"loss": 0.3787,
|
| 211 |
+
"step": 1050
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 4.44,
|
| 215 |
+
"learning_rate": 3e-05,
|
| 216 |
+
"loss": 0.3865,
|
| 217 |
+
"step": 1100
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 4.64,
|
| 221 |
+
"learning_rate": 3e-05,
|
| 222 |
+
"loss": 0.3734,
|
| 223 |
+
"step": 1150
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 4.84,
|
| 227 |
+
"learning_rate": 3e-05,
|
| 228 |
+
"loss": 0.364,
|
| 229 |
+
"step": 1200
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 5.0,
|
| 233 |
+
"eval_accuracy": 0.7352052401746725,
|
| 234 |
+
"eval_loss": 0.49626532196998596,
|
| 235 |
+
"eval_runtime": 15.762,
|
| 236 |
+
"eval_samples_per_second": 31.722,
|
| 237 |
+
"eval_steps_per_second": 2.03,
|
| 238 |
+
"step": 1240
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 5.0,
|
| 242 |
+
"eval_exact_match": 11.6,
|
| 243 |
+
"eval_f1": 17.050505050505052,
|
| 244 |
+
"eval_qa_bleu": 9.3668737780184,
|
| 245 |
+
"eval_qa_exact_match": 0.114,
|
| 246 |
+
"eval_recite_bleu": 28.6471128712922,
|
| 247 |
+
"eval_recite_exact_match": 0.004,
|
| 248 |
+
"step": 1240
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 5.04,
|
| 252 |
+
"learning_rate": 3e-05,
|
| 253 |
+
"loss": 0.345,
|
| 254 |
+
"step": 1250
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 5.24,
|
| 258 |
+
"learning_rate": 3e-05,
|
| 259 |
+
"loss": 0.2508,
|
| 260 |
+
"step": 1300
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 5.44,
|
| 264 |
+
"learning_rate": 3e-05,
|
| 265 |
+
"loss": 0.26,
|
| 266 |
+
"step": 1350
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 5.65,
|
| 270 |
+
"learning_rate": 3e-05,
|
| 271 |
+
"loss": 0.2618,
|
| 272 |
+
"step": 1400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 5.85,
|
| 276 |
+
"learning_rate": 3e-05,
|
| 277 |
+
"loss": 0.2706,
|
| 278 |
+
"step": 1450
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 6.0,
|
| 282 |
+
"eval_accuracy": 0.7419825327510917,
|
| 283 |
+
"eval_loss": 0.441346138715744,
|
| 284 |
+
"eval_runtime": 15.4217,
|
| 285 |
+
"eval_samples_per_second": 32.422,
|
| 286 |
+
"eval_steps_per_second": 2.075,
|
| 287 |
+
"step": 1488
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 6.0,
|
| 291 |
+
"eval_exact_match": 17.6,
|
| 292 |
+
"eval_f1": 23.964126984126988,
|
| 293 |
+
"eval_qa_bleu": 13.1735935431327,
|
| 294 |
+
"eval_qa_exact_match": 0.164,
|
| 295 |
+
"eval_recite_bleu": 35.20769002447571,
|
| 296 |
+
"eval_recite_exact_match": 0.008,
|
| 297 |
+
"step": 1488
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 6.05,
|
| 301 |
+
"learning_rate": 3e-05,
|
| 302 |
+
"loss": 0.2459,
|
| 303 |
+
"step": 1500
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 6.25,
|
| 307 |
+
"learning_rate": 3e-05,
|
| 308 |
+
"loss": 0.1802,
|
| 309 |
+
"step": 1550
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 6.45,
|
| 313 |
+
"learning_rate": 3e-05,
|
| 314 |
+
"loss": 0.1853,
|
| 315 |
+
"step": 1600
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 6.65,
|
| 319 |
+
"learning_rate": 3e-05,
|
| 320 |
+
"loss": 0.1905,
|
| 321 |
+
"step": 1650
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 6.85,
|
| 325 |
+
"learning_rate": 3e-05,
|
| 326 |
+
"loss": 0.1906,
|
| 327 |
+
"step": 1700
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 7.0,
|
| 331 |
+
"eval_accuracy": 0.7457903930131005,
|
| 332 |
+
"eval_loss": 0.4084840714931488,
|
| 333 |
+
"eval_runtime": 15.7338,
|
| 334 |
+
"eval_samples_per_second": 31.779,
|
| 335 |
+
"eval_steps_per_second": 2.034,
|
| 336 |
+
"step": 1736
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 7.0,
|
| 340 |
+
"eval_exact_match": 20.2,
|
| 341 |
+
"eval_f1": 25.946536796536805,
|
| 342 |
+
"eval_qa_bleu": 13.109935567750034,
|
| 343 |
+
"eval_qa_exact_match": 0.198,
|
| 344 |
+
"eval_recite_bleu": 41.370445416956926,
|
| 345 |
+
"eval_recite_exact_match": 0.018,
|
| 346 |
+
"step": 1736
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 7.06,
|
| 350 |
+
"learning_rate": 3e-05,
|
| 351 |
+
"loss": 0.1735,
|
| 352 |
+
"step": 1750
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 7.26,
|
| 356 |
+
"learning_rate": 3e-05,
|
| 357 |
+
"loss": 0.1402,
|
| 358 |
+
"step": 1800
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 7.46,
|
| 362 |
+
"learning_rate": 3e-05,
|
| 363 |
+
"loss": 0.1423,
|
| 364 |
+
"step": 1850
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 7.66,
|
| 368 |
+
"learning_rate": 3e-05,
|
| 369 |
+
"loss": 0.1468,
|
| 370 |
+
"step": 1900
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 7.86,
|
| 374 |
+
"learning_rate": 3e-05,
|
| 375 |
+
"loss": 0.1474,
|
| 376 |
+
"step": 1950
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 8.0,
|
| 380 |
+
"eval_accuracy": 0.7476826783114993,
|
| 381 |
+
"eval_loss": 0.4009522497653961,
|
| 382 |
+
"eval_runtime": 15.0512,
|
| 383 |
+
"eval_samples_per_second": 33.22,
|
| 384 |
+
"eval_steps_per_second": 2.126,
|
| 385 |
+
"step": 1984
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 8.0,
|
| 389 |
+
"eval_exact_match": 17.4,
|
| 390 |
+
"eval_f1": 24.613044733044738,
|
| 391 |
+
"eval_qa_bleu": 11.328428780013391,
|
| 392 |
+
"eval_qa_exact_match": 0.162,
|
| 393 |
+
"eval_recite_bleu": 45.62279446830165,
|
| 394 |
+
"eval_recite_exact_match": 0.036,
|
| 395 |
+
"step": 1984
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 8.06,
|
| 399 |
+
"learning_rate": 3e-05,
|
| 400 |
+
"loss": 0.1383,
|
| 401 |
+
"step": 2000
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 8.27,
|
| 405 |
+
"learning_rate": 3e-05,
|
| 406 |
+
"loss": 0.1181,
|
| 407 |
+
"step": 2050
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 8.47,
|
| 411 |
+
"learning_rate": 3e-05,
|
| 412 |
+
"loss": 0.1212,
|
| 413 |
+
"step": 2100
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 8.67,
|
| 417 |
+
"learning_rate": 3e-05,
|
| 418 |
+
"loss": 0.118,
|
| 419 |
+
"step": 2150
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 8.87,
|
| 423 |
+
"learning_rate": 3e-05,
|
| 424 |
+
"loss": 0.1264,
|
| 425 |
+
"step": 2200
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 9.0,
|
| 429 |
+
"eval_accuracy": 0.7484366812227075,
|
| 430 |
+
"eval_loss": 0.39788317680358887,
|
| 431 |
+
"eval_runtime": 15.8038,
|
| 432 |
+
"eval_samples_per_second": 31.638,
|
| 433 |
+
"eval_steps_per_second": 2.025,
|
| 434 |
+
"step": 2232
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 9.0,
|
| 438 |
+
"eval_exact_match": 16.4,
|
| 439 |
+
"eval_f1": 23.748124098124105,
|
| 440 |
+
"eval_qa_bleu": 10.821376307326117,
|
| 441 |
+
"eval_qa_exact_match": 0.148,
|
| 442 |
+
"eval_recite_bleu": 46.63269966666572,
|
| 443 |
+
"eval_recite_exact_match": 0.04,
|
| 444 |
+
"step": 2232
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 9.07,
|
| 448 |
+
"learning_rate": 3e-05,
|
| 449 |
+
"loss": 0.121,
|
| 450 |
+
"step": 2250
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 9.27,
|
| 454 |
+
"learning_rate": 3e-05,
|
| 455 |
+
"loss": 0.1036,
|
| 456 |
+
"step": 2300
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 9.48,
|
| 460 |
+
"learning_rate": 3e-05,
|
| 461 |
+
"loss": 0.1053,
|
| 462 |
+
"step": 2350
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 9.68,
|
| 466 |
+
"learning_rate": 3e-05,
|
| 467 |
+
"loss": 0.1129,
|
| 468 |
+
"step": 2400
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 9.88,
|
| 472 |
+
"learning_rate": 3e-05,
|
| 473 |
+
"loss": 0.1091,
|
| 474 |
+
"step": 2450
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 10.0,
|
| 478 |
+
"eval_accuracy": 0.7487336244541485,
|
| 479 |
+
"eval_loss": 0.4059823751449585,
|
| 480 |
+
"eval_runtime": 15.4716,
|
| 481 |
+
"eval_samples_per_second": 32.317,
|
| 482 |
+
"eval_steps_per_second": 2.068,
|
| 483 |
+
"step": 2480
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 10.0,
|
| 487 |
+
"eval_exact_match": 16.6,
|
| 488 |
+
"eval_f1": 24.02516483516484,
|
| 489 |
+
"eval_qa_bleu": 9.045796919947849,
|
| 490 |
+
"eval_qa_exact_match": 0.156,
|
| 491 |
+
"eval_recite_bleu": 50.48708434203323,
|
| 492 |
+
"eval_recite_exact_match": 0.042,
|
| 493 |
+
"step": 2480
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 10.08,
|
| 497 |
+
"learning_rate": 3e-05,
|
| 498 |
+
"loss": 0.1062,
|
| 499 |
+
"step": 2500
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 10.28,
|
| 503 |
+
"learning_rate": 3e-05,
|
| 504 |
+
"loss": 0.098,
|
| 505 |
+
"step": 2550
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 10.48,
|
| 509 |
+
"learning_rate": 3e-05,
|
| 510 |
+
"loss": 0.103,
|
| 511 |
+
"step": 2600
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 10.69,
|
| 515 |
+
"learning_rate": 3e-05,
|
| 516 |
+
"loss": 0.1015,
|
| 517 |
+
"step": 2650
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 10.89,
|
| 521 |
+
"learning_rate": 3e-05,
|
| 522 |
+
"loss": 0.1072,
|
| 523 |
+
"step": 2700
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 11.0,
|
| 527 |
+
"eval_accuracy": 0.749018922852984,
|
| 528 |
+
"eval_loss": 0.4050058424472809,
|
| 529 |
+
"eval_runtime": 15.8782,
|
| 530 |
+
"eval_samples_per_second": 31.49,
|
| 531 |
+
"eval_steps_per_second": 2.015,
|
| 532 |
+
"step": 2728
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 11.0,
|
| 536 |
+
"eval_exact_match": 17.0,
|
| 537 |
+
"eval_f1": 23.9401875901876,
|
| 538 |
+
"eval_qa_bleu": 11.704401309760495,
|
| 539 |
+
"eval_qa_exact_match": 0.154,
|
| 540 |
+
"eval_recite_bleu": 49.91003032025176,
|
| 541 |
+
"eval_recite_exact_match": 0.054,
|
| 542 |
+
"step": 2728
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 11.09,
|
| 546 |
+
"learning_rate": 3e-05,
|
| 547 |
+
"loss": 0.1013,
|
| 548 |
+
"step": 2750
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 11.29,
|
| 552 |
+
"learning_rate": 3e-05,
|
| 553 |
+
"loss": 0.0969,
|
| 554 |
+
"step": 2800
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 11.49,
|
| 558 |
+
"learning_rate": 3e-05,
|
| 559 |
+
"loss": 0.0971,
|
| 560 |
+
"step": 2850
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 11.69,
|
| 564 |
+
"learning_rate": 3e-05,
|
| 565 |
+
"loss": 0.0984,
|
| 566 |
+
"step": 2900
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 11.9,
|
| 570 |
+
"learning_rate": 3e-05,
|
| 571 |
+
"loss": 0.0969,
|
| 572 |
+
"step": 2950
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 12.0,
|
| 576 |
+
"eval_accuracy": 0.7491644832605532,
|
| 577 |
+
"eval_loss": 0.4080793559551239,
|
| 578 |
+
"eval_runtime": 15.5323,
|
| 579 |
+
"eval_samples_per_second": 32.191,
|
| 580 |
+
"eval_steps_per_second": 2.06,
|
| 581 |
+
"step": 2976
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 12.0,
|
| 585 |
+
"eval_exact_match": 18.2,
|
| 586 |
+
"eval_f1": 25.5360361860362,
|
| 587 |
+
"eval_qa_bleu": 14.03126374073827,
|
| 588 |
+
"eval_qa_exact_match": 0.176,
|
| 589 |
+
"eval_recite_bleu": 49.89862429102314,
|
| 590 |
+
"eval_recite_exact_match": 0.064,
|
| 591 |
+
"step": 2976
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 12.1,
|
| 595 |
+
"learning_rate": 3e-05,
|
| 596 |
+
"loss": 0.0977,
|
| 597 |
+
"step": 3000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 12.3,
|
| 601 |
+
"learning_rate": 3e-05,
|
| 602 |
+
"loss": 0.0917,
|
| 603 |
+
"step": 3050
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 12.5,
|
| 607 |
+
"learning_rate": 3e-05,
|
| 608 |
+
"loss": 0.0927,
|
| 609 |
+
"step": 3100
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 12.7,
|
| 613 |
+
"learning_rate": 3e-05,
|
| 614 |
+
"loss": 0.0963,
|
| 615 |
+
"step": 3150
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 12.9,
|
| 619 |
+
"learning_rate": 3e-05,
|
| 620 |
+
"loss": 0.0935,
|
| 621 |
+
"step": 3200
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 13.0,
|
| 625 |
+
"eval_accuracy": 0.7494788937409025,
|
| 626 |
+
"eval_loss": 0.4145064055919647,
|
| 627 |
+
"eval_runtime": 15.6596,
|
| 628 |
+
"eval_samples_per_second": 31.929,
|
| 629 |
+
"eval_steps_per_second": 2.043,
|
| 630 |
+
"step": 3224
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 13.0,
|
| 634 |
+
"eval_exact_match": 19.4,
|
| 635 |
+
"eval_f1": 26.85320346320347,
|
| 636 |
+
"eval_qa_bleu": 12.354122306398153,
|
| 637 |
+
"eval_qa_exact_match": 0.18,
|
| 638 |
+
"eval_recite_bleu": 51.76883509669797,
|
| 639 |
+
"eval_recite_exact_match": 0.068,
|
| 640 |
+
"step": 3224
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 13.1,
|
| 644 |
+
"learning_rate": 3e-05,
|
| 645 |
+
"loss": 0.0897,
|
| 646 |
+
"step": 3250
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 13.31,
|
| 650 |
+
"learning_rate": 3e-05,
|
| 651 |
+
"loss": 0.0905,
|
| 652 |
+
"step": 3300
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 13.51,
|
| 656 |
+
"learning_rate": 3e-05,
|
| 657 |
+
"loss": 0.0906,
|
| 658 |
+
"step": 3350
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 13.71,
|
| 662 |
+
"learning_rate": 3e-05,
|
| 663 |
+
"loss": 0.0929,
|
| 664 |
+
"step": 3400
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 13.91,
|
| 668 |
+
"learning_rate": 3e-05,
|
| 669 |
+
"loss": 0.0932,
|
| 670 |
+
"step": 3450
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 14.0,
|
| 674 |
+
"eval_accuracy": 0.7494177583697235,
|
| 675 |
+
"eval_loss": 0.4077925384044647,
|
| 676 |
+
"eval_runtime": 15.5414,
|
| 677 |
+
"eval_samples_per_second": 32.172,
|
| 678 |
+
"eval_steps_per_second": 2.059,
|
| 679 |
+
"step": 3472
|
| 680 |
+
},
|
| 681 |
+
{
|
| 682 |
+
"epoch": 14.0,
|
| 683 |
+
"eval_exact_match": 20.0,
|
| 684 |
+
"eval_f1": 27.0681240981241,
|
| 685 |
+
"eval_qa_bleu": 11.488299016816185,
|
| 686 |
+
"eval_qa_exact_match": 0.182,
|
| 687 |
+
"eval_recite_bleu": 50.751064727490615,
|
| 688 |
+
"eval_recite_exact_match": 0.072,
|
| 689 |
+
"step": 3472
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 14.11,
|
| 693 |
+
"learning_rate": 3e-05,
|
| 694 |
+
"loss": 0.0918,
|
| 695 |
+
"step": 3500
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 14.31,
|
| 699 |
+
"learning_rate": 3e-05,
|
| 700 |
+
"loss": 0.0855,
|
| 701 |
+
"step": 3550
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 14.52,
|
| 705 |
+
"learning_rate": 3e-05,
|
| 706 |
+
"loss": 0.0918,
|
| 707 |
+
"step": 3600
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 14.72,
|
| 711 |
+
"learning_rate": 3e-05,
|
| 712 |
+
"loss": 0.0951,
|
| 713 |
+
"step": 3650
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 14.92,
|
| 717 |
+
"learning_rate": 3e-05,
|
| 718 |
+
"loss": 0.0929,
|
| 719 |
+
"step": 3700
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 15.0,
|
| 723 |
+
"eval_accuracy": 0.7494439592430859,
|
| 724 |
+
"eval_loss": 0.41397154331207275,
|
| 725 |
+
"eval_runtime": 15.6399,
|
| 726 |
+
"eval_samples_per_second": 31.969,
|
| 727 |
+
"eval_steps_per_second": 2.046,
|
| 728 |
+
"step": 3720
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 15.0,
|
| 732 |
+
"eval_exact_match": 21.4,
|
| 733 |
+
"eval_f1": 28.090981240981243,
|
| 734 |
+
"eval_qa_bleu": 12.708765778160004,
|
| 735 |
+
"eval_qa_exact_match": 0.204,
|
| 736 |
+
"eval_recite_bleu": 51.72639724879611,
|
| 737 |
+
"eval_recite_exact_match": 0.068,
|
| 738 |
+
"step": 3720
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 15.12,
|
| 742 |
+
"learning_rate": 3e-05,
|
| 743 |
+
"loss": 0.0875,
|
| 744 |
+
"step": 3750
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 15.32,
|
| 748 |
+
"learning_rate": 3e-05,
|
| 749 |
+
"loss": 0.0889,
|
| 750 |
+
"step": 3800
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 15.52,
|
| 754 |
+
"learning_rate": 3e-05,
|
| 755 |
+
"loss": 0.0871,
|
| 756 |
+
"step": 3850
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 15.73,
|
| 760 |
+
"learning_rate": 3e-05,
|
| 761 |
+
"loss": 0.0918,
|
| 762 |
+
"step": 3900
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 15.93,
|
| 766 |
+
"learning_rate": 3e-05,
|
| 767 |
+
"loss": 0.0951,
|
| 768 |
+
"step": 3950
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 16.0,
|
| 772 |
+
"eval_accuracy": 0.7494730713245997,
|
| 773 |
+
"eval_loss": 0.4144821763038635,
|
| 774 |
+
"eval_runtime": 15.6836,
|
| 775 |
+
"eval_samples_per_second": 31.88,
|
| 776 |
+
"eval_steps_per_second": 2.04,
|
| 777 |
+
"step": 3968
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 16.0,
|
| 781 |
+
"eval_exact_match": 19.8,
|
| 782 |
+
"eval_f1": 27.501433011433015,
|
| 783 |
+
"eval_qa_bleu": 16.46590504145556,
|
| 784 |
+
"eval_qa_exact_match": 0.186,
|
| 785 |
+
"eval_recite_bleu": 52.1066910776491,
|
| 786 |
+
"eval_recite_exact_match": 0.074,
|
| 787 |
+
"step": 3968
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 16.13,
|
| 791 |
+
"learning_rate": 3e-05,
|
| 792 |
+
"loss": 0.0818,
|
| 793 |
+
"step": 4000
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 16.33,
|
| 797 |
+
"learning_rate": 3e-05,
|
| 798 |
+
"loss": 0.0894,
|
| 799 |
+
"step": 4050
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 16.53,
|
| 803 |
+
"learning_rate": 3e-05,
|
| 804 |
+
"loss": 0.0897,
|
| 805 |
+
"step": 4100
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 16.73,
|
| 809 |
+
"learning_rate": 3e-05,
|
| 810 |
+
"loss": 0.0914,
|
| 811 |
+
"step": 4150
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 16.94,
|
| 815 |
+
"learning_rate": 3e-05,
|
| 816 |
+
"loss": 0.0926,
|
| 817 |
+
"step": 4200
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 17.0,
|
| 821 |
+
"eval_accuracy": 0.7494847161572052,
|
| 822 |
+
"eval_loss": 0.4133778512477875,
|
| 823 |
+
"eval_runtime": 15.6296,
|
| 824 |
+
"eval_samples_per_second": 31.991,
|
| 825 |
+
"eval_steps_per_second": 2.047,
|
| 826 |
+
"step": 4216
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 17.0,
|
| 830 |
+
"eval_exact_match": 19.2,
|
| 831 |
+
"eval_f1": 26.86869424692955,
|
| 832 |
+
"eval_qa_bleu": 15.6053335659073,
|
| 833 |
+
"eval_qa_exact_match": 0.178,
|
| 834 |
+
"eval_recite_bleu": 52.354652374608186,
|
| 835 |
+
"eval_recite_exact_match": 0.068,
|
| 836 |
+
"step": 4216
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 17.14,
|
| 840 |
+
"learning_rate": 3e-05,
|
| 841 |
+
"loss": 0.0884,
|
| 842 |
+
"step": 4250
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 17.34,
|
| 846 |
+
"learning_rate": 3e-05,
|
| 847 |
+
"loss": 0.085,
|
| 848 |
+
"step": 4300
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 17.54,
|
| 852 |
+
"learning_rate": 3e-05,
|
| 853 |
+
"loss": 0.0908,
|
| 854 |
+
"step": 4350
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 17.74,
|
| 858 |
+
"learning_rate": 3e-05,
|
| 859 |
+
"loss": 0.0898,
|
| 860 |
+
"step": 4400
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 17.94,
|
| 864 |
+
"learning_rate": 3e-05,
|
| 865 |
+
"loss": 0.0946,
|
| 866 |
+
"step": 4450
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 18.0,
|
| 870 |
+
"eval_accuracy": 0.7493420669577875,
|
| 871 |
+
"eval_loss": 0.42565715312957764,
|
| 872 |
+
"eval_runtime": 15.8896,
|
| 873 |
+
"eval_samples_per_second": 31.467,
|
| 874 |
+
"eval_steps_per_second": 2.014,
|
| 875 |
+
"step": 4464
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 18.0,
|
| 879 |
+
"eval_exact_match": 21.2,
|
| 880 |
+
"eval_f1": 29.34194877018406,
|
| 881 |
+
"eval_qa_bleu": 13.32335745765551,
|
| 882 |
+
"eval_qa_exact_match": 0.196,
|
| 883 |
+
"eval_recite_bleu": 53.51642016865907,
|
| 884 |
+
"eval_recite_exact_match": 0.074,
|
| 885 |
+
"step": 4464
|
| 886 |
+
}
|
| 887 |
+
],
|
| 888 |
+
"logging_steps": 50,
|
| 889 |
+
"max_steps": 4960,
|
| 890 |
+
"num_train_epochs": 20,
|
| 891 |
+
"save_steps": 500,
|
| 892 |
+
"total_flos": 2.071633241373696e+17,
|
| 893 |
+
"trial_name": null,
|
| 894 |
+
"trial_params": null
|
| 895 |
+
}
|
checkpoint-4464/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:acdd81e53b27f902dc4c1896d0c5253609dec7800ac5e5d60f5a323620c45c4a
|
| 3 |
+
size 4728
|
checkpoint-4464/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|