Training in progress, epoch 13, checkpoint
Browse files- checkpoint-18876/added_tokens.json +3 -0
- checkpoint-18876/config.json +42 -0
- checkpoint-18876/generation_config.json +8 -0
- checkpoint-18876/merges.txt +0 -0
- checkpoint-18876/optimizer.pt +3 -0
- checkpoint-18876/pytorch_model.bin +3 -0
- checkpoint-18876/rng_state.pth +3 -0
- checkpoint-18876/scheduler.pt +3 -0
- checkpoint-18876/special_tokens_map.json +6 -0
- checkpoint-18876/tokenizer.json +0 -0
- checkpoint-18876/tokenizer_config.json +20 -0
- checkpoint-18876/trainer_state.json +650 -0
- checkpoint-18876/training_args.bin +3 -0
- checkpoint-18876/vocab.json +0 -0
checkpoint-18876/added_tokens.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<|endoftext|>": 50256
|
| 3 |
+
}
|
checkpoint-18876/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-18876/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-18876/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-18876/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f5301217cc42425d06b1d81750e11b546b61d605cd2c88f8affe59a0f3c2509c
|
| 3 |
+
size 12461385454
|
checkpoint-18876/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:42c65ab27f6b8190ecf07b681ebfba74eda8bfcd5f18c7d2dcccdc442bf8dd86
|
| 3 |
+
size 6230637102
|
checkpoint-18876/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d59474a54d536d90e6f2a07dbffa292fcf8e4629ddb33cc1f350261e7307a6e5
|
| 3 |
+
size 14244
|
checkpoint-18876/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:19b54fcff02f86eef6a3ed8ef4747d39acf5939de13a9bcbeffa6601581c337f
|
| 3 |
+
size 1064
|
checkpoint-18876/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-18876/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-18876/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-18876/trainer_state.json
ADDED
|
@@ -0,0 +1,650 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 13.0,
|
| 5 |
+
"eval_steps": 1452,
|
| 6 |
+
"global_step": 18876,
|
| 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": 2.2721,
|
| 15 |
+
"step": 291
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.4,
|
| 19 |
+
"learning_rate": 3e-05,
|
| 20 |
+
"loss": 2.0539,
|
| 21 |
+
"step": 582
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.6,
|
| 25 |
+
"learning_rate": 3e-05,
|
| 26 |
+
"loss": 2.0079,
|
| 27 |
+
"step": 873
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.8,
|
| 31 |
+
"learning_rate": 3e-05,
|
| 32 |
+
"loss": 1.9686,
|
| 33 |
+
"step": 1164
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 1.0,
|
| 37 |
+
"eval_accuracy": 0.668149394347241,
|
| 38 |
+
"eval_loss": 1.5662202835083008,
|
| 39 |
+
"eval_runtime": 137.7572,
|
| 40 |
+
"eval_samples_per_second": 36.296,
|
| 41 |
+
"eval_steps_per_second": 2.272,
|
| 42 |
+
"step": 1452
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 1.0,
|
| 46 |
+
"eval_exact_match": 9.04,
|
| 47 |
+
"eval_f1": 13.087046780017364,
|
| 48 |
+
"eval_qa_bleu": 5.297766450359973,
|
| 49 |
+
"eval_qa_exact_match": 0.085,
|
| 50 |
+
"eval_recite_bleu": 14.716037077738477,
|
| 51 |
+
"eval_recite_exact_match": 0.0,
|
| 52 |
+
"step": 1452
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.0,
|
| 56 |
+
"learning_rate": 3e-05,
|
| 57 |
+
"loss": 1.9076,
|
| 58 |
+
"step": 1455
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 1.2,
|
| 62 |
+
"learning_rate": 3e-05,
|
| 63 |
+
"loss": 1.5211,
|
| 64 |
+
"step": 1746
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 1.4,
|
| 68 |
+
"learning_rate": 3e-05,
|
| 69 |
+
"loss": 1.5039,
|
| 70 |
+
"step": 2037
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.6,
|
| 74 |
+
"learning_rate": 3e-05,
|
| 75 |
+
"loss": 1.4834,
|
| 76 |
+
"step": 2328
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 1.8,
|
| 80 |
+
"learning_rate": 3e-05,
|
| 81 |
+
"loss": 1.4604,
|
| 82 |
+
"step": 2619
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 2.0,
|
| 86 |
+
"eval_accuracy": 0.6914514131897712,
|
| 87 |
+
"eval_loss": 1.206884741783142,
|
| 88 |
+
"eval_runtime": 139.7305,
|
| 89 |
+
"eval_samples_per_second": 35.783,
|
| 90 |
+
"eval_steps_per_second": 2.24,
|
| 91 |
+
"step": 2904
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 2.0,
|
| 95 |
+
"eval_exact_match": 11.64,
|
| 96 |
+
"eval_f1": 16.917593370997537,
|
| 97 |
+
"eval_qa_bleu": 7.627957176996819,
|
| 98 |
+
"eval_qa_exact_match": 0.1106,
|
| 99 |
+
"eval_recite_bleu": 18.356743056908567,
|
| 100 |
+
"eval_recite_exact_match": 0.0,
|
| 101 |
+
"step": 2904
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 2.0,
|
| 105 |
+
"learning_rate": 3e-05,
|
| 106 |
+
"loss": 1.4368,
|
| 107 |
+
"step": 2910
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 2.2,
|
| 111 |
+
"learning_rate": 3e-05,
|
| 112 |
+
"loss": 1.0705,
|
| 113 |
+
"step": 3201
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 2.4,
|
| 117 |
+
"learning_rate": 3e-05,
|
| 118 |
+
"loss": 1.0685,
|
| 119 |
+
"step": 3492
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.61,
|
| 123 |
+
"learning_rate": 3e-05,
|
| 124 |
+
"loss": 1.0554,
|
| 125 |
+
"step": 3783
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 2.81,
|
| 129 |
+
"learning_rate": 3e-05,
|
| 130 |
+
"loss": 1.0522,
|
| 131 |
+
"step": 4074
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 3.0,
|
| 135 |
+
"eval_accuracy": 0.7148896366083446,
|
| 136 |
+
"eval_loss": 0.9045500159263611,
|
| 137 |
+
"eval_runtime": 141.6455,
|
| 138 |
+
"eval_samples_per_second": 35.299,
|
| 139 |
+
"eval_steps_per_second": 2.21,
|
| 140 |
+
"step": 4356
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 3.0,
|
| 144 |
+
"eval_exact_match": 12.84,
|
| 145 |
+
"eval_f1": 18.555356874635468,
|
| 146 |
+
"eval_qa_bleu": 8.395879921702353,
|
| 147 |
+
"eval_qa_exact_match": 0.1188,
|
| 148 |
+
"eval_recite_bleu": 20.824221134285214,
|
| 149 |
+
"eval_recite_exact_match": 0.0,
|
| 150 |
+
"step": 4356
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 3.01,
|
| 154 |
+
"learning_rate": 3e-05,
|
| 155 |
+
"loss": 1.0375,
|
| 156 |
+
"step": 4365
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 3.21,
|
| 160 |
+
"learning_rate": 3e-05,
|
| 161 |
+
"loss": 0.7373,
|
| 162 |
+
"step": 4656
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 3.41,
|
| 166 |
+
"learning_rate": 3e-05,
|
| 167 |
+
"loss": 0.7497,
|
| 168 |
+
"step": 4947
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 3.61,
|
| 172 |
+
"learning_rate": 3e-05,
|
| 173 |
+
"loss": 0.7479,
|
| 174 |
+
"step": 5238
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 3.81,
|
| 178 |
+
"learning_rate": 3e-05,
|
| 179 |
+
"loss": 0.7455,
|
| 180 |
+
"step": 5529
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 4.0,
|
| 184 |
+
"eval_accuracy": 0.7340508748317631,
|
| 185 |
+
"eval_loss": 0.6843858957290649,
|
| 186 |
+
"eval_runtime": 143.5169,
|
| 187 |
+
"eval_samples_per_second": 34.839,
|
| 188 |
+
"eval_steps_per_second": 2.181,
|
| 189 |
+
"step": 5808
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 4.0,
|
| 193 |
+
"eval_exact_match": 13.44,
|
| 194 |
+
"eval_f1": 19.418847592695993,
|
| 195 |
+
"eval_qa_bleu": 8.731752840499793,
|
| 196 |
+
"eval_qa_exact_match": 0.1256,
|
| 197 |
+
"eval_recite_bleu": 24.11683390270685,
|
| 198 |
+
"eval_recite_exact_match": 0.0002,
|
| 199 |
+
"step": 5808
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 4.01,
|
| 203 |
+
"learning_rate": 3e-05,
|
| 204 |
+
"loss": 0.7226,
|
| 205 |
+
"step": 5820
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 4.21,
|
| 209 |
+
"learning_rate": 3e-05,
|
| 210 |
+
"loss": 0.5117,
|
| 211 |
+
"step": 6111
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 4.41,
|
| 215 |
+
"learning_rate": 3e-05,
|
| 216 |
+
"loss": 0.5208,
|
| 217 |
+
"step": 6402
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 4.61,
|
| 221 |
+
"learning_rate": 3e-05,
|
| 222 |
+
"loss": 0.5271,
|
| 223 |
+
"step": 6693
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 4.81,
|
| 227 |
+
"learning_rate": 3e-05,
|
| 228 |
+
"loss": 0.5307,
|
| 229 |
+
"step": 6984
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 5.0,
|
| 233 |
+
"eval_accuracy": 0.7474745625841185,
|
| 234 |
+
"eval_loss": 0.5420497059822083,
|
| 235 |
+
"eval_runtime": 144.0475,
|
| 236 |
+
"eval_samples_per_second": 34.711,
|
| 237 |
+
"eval_steps_per_second": 2.173,
|
| 238 |
+
"step": 7260
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 5.0,
|
| 242 |
+
"eval_exact_match": 15.08,
|
| 243 |
+
"eval_f1": 21.4491845879152,
|
| 244 |
+
"eval_qa_bleu": 10.89471390420015,
|
| 245 |
+
"eval_qa_exact_match": 0.1434,
|
| 246 |
+
"eval_recite_bleu": 27.62075386163555,
|
| 247 |
+
"eval_recite_exact_match": 0.0014,
|
| 248 |
+
"step": 7260
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 5.01,
|
| 252 |
+
"learning_rate": 3e-05,
|
| 253 |
+
"loss": 0.5165,
|
| 254 |
+
"step": 7275
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 5.21,
|
| 258 |
+
"learning_rate": 3e-05,
|
| 259 |
+
"loss": 0.3624,
|
| 260 |
+
"step": 7566
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 5.41,
|
| 264 |
+
"learning_rate": 3e-05,
|
| 265 |
+
"loss": 0.3748,
|
| 266 |
+
"step": 7857
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 5.61,
|
| 270 |
+
"learning_rate": 3e-05,
|
| 271 |
+
"loss": 0.383,
|
| 272 |
+
"step": 8148
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 5.81,
|
| 276 |
+
"learning_rate": 3e-05,
|
| 277 |
+
"loss": 0.3796,
|
| 278 |
+
"step": 8439
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 6.0,
|
| 282 |
+
"eval_accuracy": 0.7560010767160161,
|
| 283 |
+
"eval_loss": 0.4608670473098755,
|
| 284 |
+
"eval_runtime": 144.2314,
|
| 285 |
+
"eval_samples_per_second": 34.667,
|
| 286 |
+
"eval_steps_per_second": 2.17,
|
| 287 |
+
"step": 8712
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 6.0,
|
| 291 |
+
"eval_exact_match": 16.22,
|
| 292 |
+
"eval_f1": 22.5212141898349,
|
| 293 |
+
"eval_qa_bleu": 11.517856911949949,
|
| 294 |
+
"eval_qa_exact_match": 0.1524,
|
| 295 |
+
"eval_recite_bleu": 32.1802083615024,
|
| 296 |
+
"eval_recite_exact_match": 0.0044,
|
| 297 |
+
"step": 8712
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 6.01,
|
| 301 |
+
"learning_rate": 3e-05,
|
| 302 |
+
"loss": 0.3803,
|
| 303 |
+
"step": 8730
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 6.21,
|
| 307 |
+
"learning_rate": 3e-05,
|
| 308 |
+
"loss": 0.27,
|
| 309 |
+
"step": 9021
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 6.41,
|
| 313 |
+
"learning_rate": 3e-05,
|
| 314 |
+
"loss": 0.28,
|
| 315 |
+
"step": 9312
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 6.61,
|
| 319 |
+
"learning_rate": 3e-05,
|
| 320 |
+
"loss": 0.2889,
|
| 321 |
+
"step": 9603
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 6.81,
|
| 325 |
+
"learning_rate": 3e-05,
|
| 326 |
+
"loss": 0.2912,
|
| 327 |
+
"step": 9894
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 7.0,
|
| 331 |
+
"eval_accuracy": 0.7603886944818304,
|
| 332 |
+
"eval_loss": 0.43113213777542114,
|
| 333 |
+
"eval_runtime": 144.1381,
|
| 334 |
+
"eval_samples_per_second": 34.689,
|
| 335 |
+
"eval_steps_per_second": 2.172,
|
| 336 |
+
"step": 10164
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 7.0,
|
| 340 |
+
"eval_exact_match": 17.64,
|
| 341 |
+
"eval_f1": 24.23915265275915,
|
| 342 |
+
"eval_qa_bleu": 12.404229060978194,
|
| 343 |
+
"eval_qa_exact_match": 0.1648,
|
| 344 |
+
"eval_recite_bleu": 37.23348279932964,
|
| 345 |
+
"eval_recite_exact_match": 0.0102,
|
| 346 |
+
"step": 10164
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 7.01,
|
| 350 |
+
"learning_rate": 3e-05,
|
| 351 |
+
"loss": 0.2835,
|
| 352 |
+
"step": 10185
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 7.21,
|
| 356 |
+
"learning_rate": 3e-05,
|
| 357 |
+
"loss": 0.2131,
|
| 358 |
+
"step": 10476
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 7.42,
|
| 362 |
+
"learning_rate": 3e-05,
|
| 363 |
+
"loss": 0.2213,
|
| 364 |
+
"step": 10767
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 7.62,
|
| 368 |
+
"learning_rate": 3e-05,
|
| 369 |
+
"loss": 0.226,
|
| 370 |
+
"step": 11058
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 7.82,
|
| 374 |
+
"learning_rate": 3e-05,
|
| 375 |
+
"loss": 0.2282,
|
| 376 |
+
"step": 11349
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 8.0,
|
| 380 |
+
"eval_accuracy": 0.7626543741588157,
|
| 381 |
+
"eval_loss": 0.41835975646972656,
|
| 382 |
+
"eval_runtime": 144.0991,
|
| 383 |
+
"eval_samples_per_second": 34.698,
|
| 384 |
+
"eval_steps_per_second": 2.172,
|
| 385 |
+
"step": 11616
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 8.0,
|
| 389 |
+
"eval_exact_match": 17.78,
|
| 390 |
+
"eval_f1": 25.32648324184451,
|
| 391 |
+
"eval_qa_bleu": 12.460459536013193,
|
| 392 |
+
"eval_qa_exact_match": 0.1676,
|
| 393 |
+
"eval_recite_bleu": 40.14518456576004,
|
| 394 |
+
"eval_recite_exact_match": 0.0168,
|
| 395 |
+
"step": 11616
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 8.02,
|
| 399 |
+
"learning_rate": 3e-05,
|
| 400 |
+
"loss": 0.1645,
|
| 401 |
+
"step": 11640
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 8.22,
|
| 405 |
+
"learning_rate": 3e-05,
|
| 406 |
+
"loss": 0.1768,
|
| 407 |
+
"step": 11931
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 8.42,
|
| 411 |
+
"learning_rate": 3e-05,
|
| 412 |
+
"loss": 0.1851,
|
| 413 |
+
"step": 12222
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 8.62,
|
| 417 |
+
"learning_rate": 3e-05,
|
| 418 |
+
"loss": 0.1895,
|
| 419 |
+
"step": 12513
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 8.82,
|
| 423 |
+
"learning_rate": 3e-05,
|
| 424 |
+
"loss": 0.1905,
|
| 425 |
+
"step": 12804
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 9.0,
|
| 429 |
+
"eval_accuracy": 0.7639776581426648,
|
| 430 |
+
"eval_loss": 0.4136393666267395,
|
| 431 |
+
"eval_runtime": 146.8079,
|
| 432 |
+
"eval_samples_per_second": 34.058,
|
| 433 |
+
"eval_steps_per_second": 2.132,
|
| 434 |
+
"step": 13068
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 9.0,
|
| 438 |
+
"eval_exact_match": 18.58,
|
| 439 |
+
"eval_f1": 26.144983963782675,
|
| 440 |
+
"eval_qa_bleu": 12.723130940050861,
|
| 441 |
+
"eval_qa_exact_match": 0.1766,
|
| 442 |
+
"eval_recite_bleu": 44.36820281112791,
|
| 443 |
+
"eval_recite_exact_match": 0.0212,
|
| 444 |
+
"step": 13068
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 9.02,
|
| 448 |
+
"learning_rate": 3e-05,
|
| 449 |
+
"loss": 0.1919,
|
| 450 |
+
"step": 13095
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 9.22,
|
| 454 |
+
"learning_rate": 3e-05,
|
| 455 |
+
"loss": 0.156,
|
| 456 |
+
"step": 13386
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 9.42,
|
| 460 |
+
"learning_rate": 3e-05,
|
| 461 |
+
"loss": 0.158,
|
| 462 |
+
"step": 13677
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 9.62,
|
| 466 |
+
"learning_rate": 3e-05,
|
| 467 |
+
"loss": 0.1656,
|
| 468 |
+
"step": 13968
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 9.82,
|
| 472 |
+
"learning_rate": 3e-05,
|
| 473 |
+
"loss": 0.1687,
|
| 474 |
+
"step": 14259
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 10.0,
|
| 478 |
+
"eval_accuracy": 0.7647776581426649,
|
| 479 |
+
"eval_loss": 0.41753506660461426,
|
| 480 |
+
"eval_runtime": 149.0935,
|
| 481 |
+
"eval_samples_per_second": 33.536,
|
| 482 |
+
"eval_steps_per_second": 2.099,
|
| 483 |
+
"step": 14520
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 10.0,
|
| 487 |
+
"eval_exact_match": 19.8,
|
| 488 |
+
"eval_f1": 27.235079497903637,
|
| 489 |
+
"eval_qa_bleu": 13.874335090146856,
|
| 490 |
+
"eval_qa_exact_match": 0.1886,
|
| 491 |
+
"eval_recite_bleu": 46.53984717815193,
|
| 492 |
+
"eval_recite_exact_match": 0.0276,
|
| 493 |
+
"step": 14520
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 10.02,
|
| 497 |
+
"learning_rate": 3e-05,
|
| 498 |
+
"loss": 0.1684,
|
| 499 |
+
"step": 14550
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 10.22,
|
| 503 |
+
"learning_rate": 3e-05,
|
| 504 |
+
"loss": 0.1396,
|
| 505 |
+
"step": 14841
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 10.42,
|
| 509 |
+
"learning_rate": 3e-05,
|
| 510 |
+
"loss": 0.1457,
|
| 511 |
+
"step": 15132
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 10.62,
|
| 515 |
+
"learning_rate": 3e-05,
|
| 516 |
+
"loss": 0.1542,
|
| 517 |
+
"step": 15423
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 10.82,
|
| 521 |
+
"learning_rate": 3e-05,
|
| 522 |
+
"loss": 0.1553,
|
| 523 |
+
"step": 15714
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 11.0,
|
| 527 |
+
"eval_accuracy": 0.7651154777927321,
|
| 528 |
+
"eval_loss": 0.4211733937263489,
|
| 529 |
+
"eval_runtime": 150.5664,
|
| 530 |
+
"eval_samples_per_second": 33.208,
|
| 531 |
+
"eval_steps_per_second": 2.079,
|
| 532 |
+
"step": 15972
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 11.0,
|
| 536 |
+
"eval_exact_match": 19.1,
|
| 537 |
+
"eval_f1": 26.696266939458862,
|
| 538 |
+
"eval_qa_bleu": 14.962573389382232,
|
| 539 |
+
"eval_qa_exact_match": 0.1814,
|
| 540 |
+
"eval_recite_bleu": 46.889182609591046,
|
| 541 |
+
"eval_recite_exact_match": 0.032,
|
| 542 |
+
"step": 15972
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 11.02,
|
| 546 |
+
"learning_rate": 3e-05,
|
| 547 |
+
"loss": 0.1549,
|
| 548 |
+
"step": 16005
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 11.22,
|
| 552 |
+
"learning_rate": 3e-05,
|
| 553 |
+
"loss": 0.1335,
|
| 554 |
+
"step": 16296
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 11.42,
|
| 558 |
+
"learning_rate": 3e-05,
|
| 559 |
+
"loss": 0.1381,
|
| 560 |
+
"step": 16587
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 11.62,
|
| 564 |
+
"learning_rate": 3e-05,
|
| 565 |
+
"loss": 0.143,
|
| 566 |
+
"step": 16878
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 11.82,
|
| 570 |
+
"learning_rate": 3e-05,
|
| 571 |
+
"loss": 0.1447,
|
| 572 |
+
"step": 17169
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 12.0,
|
| 576 |
+
"eval_accuracy": 0.7653222072678331,
|
| 577 |
+
"eval_loss": 0.4283375144004822,
|
| 578 |
+
"eval_runtime": 152.1173,
|
| 579 |
+
"eval_samples_per_second": 32.869,
|
| 580 |
+
"eval_steps_per_second": 2.058,
|
| 581 |
+
"step": 17424
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 12.0,
|
| 585 |
+
"eval_exact_match": 19.7,
|
| 586 |
+
"eval_f1": 27.086380949947664,
|
| 587 |
+
"eval_qa_bleu": 13.161442853987554,
|
| 588 |
+
"eval_qa_exact_match": 0.1884,
|
| 589 |
+
"eval_recite_bleu": 48.94647260503517,
|
| 590 |
+
"eval_recite_exact_match": 0.0318,
|
| 591 |
+
"step": 17424
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 12.02,
|
| 595 |
+
"learning_rate": 3e-05,
|
| 596 |
+
"loss": 0.1449,
|
| 597 |
+
"step": 17460
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 12.23,
|
| 601 |
+
"learning_rate": 3e-05,
|
| 602 |
+
"loss": 0.1267,
|
| 603 |
+
"step": 17751
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 12.43,
|
| 607 |
+
"learning_rate": 3e-05,
|
| 608 |
+
"loss": 0.131,
|
| 609 |
+
"step": 18042
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 12.63,
|
| 613 |
+
"learning_rate": 3e-05,
|
| 614 |
+
"loss": 0.1366,
|
| 615 |
+
"step": 18333
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 12.83,
|
| 619 |
+
"learning_rate": 3e-05,
|
| 620 |
+
"loss": 0.1388,
|
| 621 |
+
"step": 18624
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 13.0,
|
| 625 |
+
"eval_accuracy": 0.7655897711978465,
|
| 626 |
+
"eval_loss": 0.4287155270576477,
|
| 627 |
+
"eval_runtime": 152.6887,
|
| 628 |
+
"eval_samples_per_second": 32.746,
|
| 629 |
+
"eval_steps_per_second": 2.05,
|
| 630 |
+
"step": 18876
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 13.0,
|
| 634 |
+
"eval_exact_match": 20.12,
|
| 635 |
+
"eval_f1": 27.673234730738653,
|
| 636 |
+
"eval_qa_bleu": 14.175044837001105,
|
| 637 |
+
"eval_qa_exact_match": 0.191,
|
| 638 |
+
"eval_recite_bleu": 49.79523539787241,
|
| 639 |
+
"eval_recite_exact_match": 0.0446,
|
| 640 |
+
"step": 18876
|
| 641 |
+
}
|
| 642 |
+
],
|
| 643 |
+
"logging_steps": 291,
|
| 644 |
+
"max_steps": 29040,
|
| 645 |
+
"num_train_epochs": 20,
|
| 646 |
+
"save_steps": 500,
|
| 647 |
+
"total_flos": 8.616909555105792e+17,
|
| 648 |
+
"trial_name": null,
|
| 649 |
+
"trial_params": null
|
| 650 |
+
}
|
checkpoint-18876/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4daeee080fafab59faec01202ee535e417ff0b7cbe131d7c41eeff253356d347
|
| 3 |
+
size 4728
|
checkpoint-18876/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|