Initial commit
Browse files- config.json +43 -0
- merges.txt +0 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +1 -0
- tokenizer_config.json +1 -0
- trainer_state.json +1456 -0
- training_args.bin +3 -0
- vocab.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "distilgpt2",
|
| 3 |
+
"_num_labels": 1,
|
| 4 |
+
"activation_function": "gelu_new",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"GPT2LMHeadModel"
|
| 7 |
+
],
|
| 8 |
+
"attn_pdrop": 0.1,
|
| 9 |
+
"bos_token_id": 50256,
|
| 10 |
+
"embd_pdrop": 0.1,
|
| 11 |
+
"eos_token_id": 50256,
|
| 12 |
+
"gradient_checkpointing": false,
|
| 13 |
+
"id2label": {
|
| 14 |
+
"0": "LABEL_0"
|
| 15 |
+
},
|
| 16 |
+
"initializer_range": 0.02,
|
| 17 |
+
"label2id": {
|
| 18 |
+
"LABEL_0": 0
|
| 19 |
+
},
|
| 20 |
+
"layer_norm_epsilon": 1e-05,
|
| 21 |
+
"model_type": "gpt2",
|
| 22 |
+
"n_ctx": 1024,
|
| 23 |
+
"n_embd": 768,
|
| 24 |
+
"n_head": 12,
|
| 25 |
+
"n_inner": null,
|
| 26 |
+
"n_layer": 6,
|
| 27 |
+
"n_positions": 1024,
|
| 28 |
+
"resid_pdrop": 0.1,
|
| 29 |
+
"summary_activation": null,
|
| 30 |
+
"summary_first_dropout": 0.1,
|
| 31 |
+
"summary_proj_to_labels": true,
|
| 32 |
+
"summary_type": "cls_index",
|
| 33 |
+
"summary_use_proj": true,
|
| 34 |
+
"task_specific_params": {
|
| 35 |
+
"text-generation": {
|
| 36 |
+
"do_sample": true,
|
| 37 |
+
"max_length": 50
|
| 38 |
+
}
|
| 39 |
+
},
|
| 40 |
+
"transformers_version": "4.2.2",
|
| 41 |
+
"use_cache": true,
|
| 42 |
+
"vocab_size": 50257
|
| 43 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f2ad73549bef0ba13b6251b5c3716921a1d12dc0beb3545926d9cc4d489fd796
|
| 3 |
+
size 655348487
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ec8a4731fcf721d0ca983f965e07d4dd9d3ab13652b524d6986d742adf34d531
|
| 3 |
+
size 333975623
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2db1ac53d50a2406b4d55d32b4dcef6a00e4475802e2e9d9ecc5cb112d069555
|
| 3 |
+
size 623
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "unk_token": "<|endoftext|>"}
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"unk_token": "<|endoftext|>", "bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "add_prefix_space": false, "model_max_length": 1024, "special_tokens_map_file": null, "name_or_path": "distilgpt2"}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1456 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 16.402405686167306,
|
| 5 |
+
"global_step": 120000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.07,
|
| 12 |
+
"learning_rate": 5e-05,
|
| 13 |
+
"loss": 3.1294,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.14,
|
| 18 |
+
"learning_rate": 4.999658259859204e-05,
|
| 19 |
+
"loss": 3.1422,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.21,
|
| 24 |
+
"learning_rate": 4.999316519718406e-05,
|
| 25 |
+
"loss": 3.1479,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.27,
|
| 30 |
+
"learning_rate": 4.998974779577609e-05,
|
| 31 |
+
"loss": 3.1412,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.34,
|
| 36 |
+
"learning_rate": 4.998633039436813e-05,
|
| 37 |
+
"loss": 3.1387,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.41,
|
| 42 |
+
"learning_rate": 4.9982912992960154e-05,
|
| 43 |
+
"loss": 3.1344,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.48,
|
| 48 |
+
"learning_rate": 4.997949559155218e-05,
|
| 49 |
+
"loss": 3.1353,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.55,
|
| 54 |
+
"learning_rate": 4.9976078190144217e-05,
|
| 55 |
+
"loss": 3.1283,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.62,
|
| 60 |
+
"learning_rate": 4.9972660788736244e-05,
|
| 61 |
+
"loss": 3.1291,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.68,
|
| 66 |
+
"learning_rate": 4.996924338732828e-05,
|
| 67 |
+
"loss": 3.1226,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.75,
|
| 72 |
+
"learning_rate": 4.9965825985920306e-05,
|
| 73 |
+
"loss": 3.1173,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.82,
|
| 78 |
+
"learning_rate": 4.996240858451234e-05,
|
| 79 |
+
"loss": 3.1126,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.89,
|
| 84 |
+
"learning_rate": 4.995899118310437e-05,
|
| 85 |
+
"loss": 3.1117,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.96,
|
| 90 |
+
"learning_rate": 4.99555737816964e-05,
|
| 91 |
+
"loss": 3.1037,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 1.03,
|
| 96 |
+
"learning_rate": 4.995215638028843e-05,
|
| 97 |
+
"loss": 3.0981,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.09,
|
| 102 |
+
"learning_rate": 4.994873897888046e-05,
|
| 103 |
+
"loss": 3.0682,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 1.16,
|
| 108 |
+
"learning_rate": 4.994532157747249e-05,
|
| 109 |
+
"loss": 3.0697,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 1.23,
|
| 114 |
+
"learning_rate": 4.994190417606453e-05,
|
| 115 |
+
"loss": 3.0673,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.3,
|
| 120 |
+
"learning_rate": 4.993848677465655e-05,
|
| 121 |
+
"loss": 3.0703,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.37,
|
| 126 |
+
"learning_rate": 4.993506937324858e-05,
|
| 127 |
+
"loss": 3.0624,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.44,
|
| 132 |
+
"learning_rate": 4.993165197184062e-05,
|
| 133 |
+
"loss": 3.0618,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 1.5,
|
| 138 |
+
"learning_rate": 4.9928234570432645e-05,
|
| 139 |
+
"loss": 3.0612,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.57,
|
| 144 |
+
"learning_rate": 4.992481716902467e-05,
|
| 145 |
+
"loss": 3.0651,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 1.64,
|
| 150 |
+
"learning_rate": 4.992139976761671e-05,
|
| 151 |
+
"loss": 3.0555,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 1.71,
|
| 156 |
+
"learning_rate": 4.9917982366208735e-05,
|
| 157 |
+
"loss": 3.0609,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.78,
|
| 162 |
+
"learning_rate": 4.991456496480077e-05,
|
| 163 |
+
"loss": 3.0564,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.85,
|
| 168 |
+
"learning_rate": 4.99111475633928e-05,
|
| 169 |
+
"loss": 3.0581,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.91,
|
| 174 |
+
"learning_rate": 4.990773016198483e-05,
|
| 175 |
+
"loss": 3.054,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 1.98,
|
| 180 |
+
"learning_rate": 4.990431276057686e-05,
|
| 181 |
+
"loss": 3.0559,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 2.05,
|
| 186 |
+
"learning_rate": 4.990089535916889e-05,
|
| 187 |
+
"loss": 3.0273,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 2.12,
|
| 192 |
+
"learning_rate": 4.989747795776092e-05,
|
| 193 |
+
"loss": 3.0221,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 2.19,
|
| 198 |
+
"learning_rate": 4.989406055635295e-05,
|
| 199 |
+
"loss": 3.0218,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 2.26,
|
| 204 |
+
"learning_rate": 4.989064315494498e-05,
|
| 205 |
+
"loss": 3.0245,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 2.32,
|
| 210 |
+
"learning_rate": 4.988722575353702e-05,
|
| 211 |
+
"loss": 3.0237,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 2.39,
|
| 216 |
+
"learning_rate": 4.988380835212904e-05,
|
| 217 |
+
"loss": 3.0224,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 2.46,
|
| 222 |
+
"learning_rate": 4.988039095072107e-05,
|
| 223 |
+
"loss": 3.0183,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 2.53,
|
| 228 |
+
"learning_rate": 4.987697354931311e-05,
|
| 229 |
+
"loss": 3.0113,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 2.6,
|
| 234 |
+
"learning_rate": 4.9873556147905135e-05,
|
| 235 |
+
"loss": 3.0211,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 2.67,
|
| 240 |
+
"learning_rate": 4.987013874649716e-05,
|
| 241 |
+
"loss": 3.0253,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.73,
|
| 246 |
+
"learning_rate": 4.98667213450892e-05,
|
| 247 |
+
"loss": 3.0143,
|
| 248 |
+
"step": 20000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 2.8,
|
| 252 |
+
"learning_rate": 4.9863303943681225e-05,
|
| 253 |
+
"loss": 3.0192,
|
| 254 |
+
"step": 20500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 2.87,
|
| 258 |
+
"learning_rate": 4.985988654227326e-05,
|
| 259 |
+
"loss": 3.012,
|
| 260 |
+
"step": 21000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 2.94,
|
| 264 |
+
"learning_rate": 4.985646914086529e-05,
|
| 265 |
+
"loss": 3.016,
|
| 266 |
+
"step": 21500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 3.01,
|
| 270 |
+
"learning_rate": 4.985305173945732e-05,
|
| 271 |
+
"loss": 3.0087,
|
| 272 |
+
"step": 22000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 3.08,
|
| 276 |
+
"learning_rate": 4.984963433804935e-05,
|
| 277 |
+
"loss": 2.9676,
|
| 278 |
+
"step": 22500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 3.14,
|
| 282 |
+
"learning_rate": 4.9846216936641384e-05,
|
| 283 |
+
"loss": 2.9644,
|
| 284 |
+
"step": 23000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 3.21,
|
| 288 |
+
"learning_rate": 4.984279953523341e-05,
|
| 289 |
+
"loss": 2.9714,
|
| 290 |
+
"step": 23500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 3.28,
|
| 294 |
+
"learning_rate": 4.983938213382544e-05,
|
| 295 |
+
"loss": 2.9816,
|
| 296 |
+
"step": 24000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 3.35,
|
| 300 |
+
"learning_rate": 4.9835964732417474e-05,
|
| 301 |
+
"loss": 2.9759,
|
| 302 |
+
"step": 24500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 3.42,
|
| 306 |
+
"learning_rate": 4.983254733100951e-05,
|
| 307 |
+
"loss": 2.977,
|
| 308 |
+
"step": 25000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 3.49,
|
| 312 |
+
"learning_rate": 4.982912992960153e-05,
|
| 313 |
+
"loss": 2.9841,
|
| 314 |
+
"step": 25500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 3.55,
|
| 318 |
+
"learning_rate": 4.982571252819356e-05,
|
| 319 |
+
"loss": 2.9731,
|
| 320 |
+
"step": 26000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 3.62,
|
| 324 |
+
"learning_rate": 4.98222951267856e-05,
|
| 325 |
+
"loss": 2.9716,
|
| 326 |
+
"step": 26500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 3.69,
|
| 330 |
+
"learning_rate": 4.9818877725377625e-05,
|
| 331 |
+
"loss": 2.9762,
|
| 332 |
+
"step": 27000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 3.76,
|
| 336 |
+
"learning_rate": 4.981546032396965e-05,
|
| 337 |
+
"loss": 2.9774,
|
| 338 |
+
"step": 27500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 3.83,
|
| 342 |
+
"learning_rate": 4.981204292256169e-05,
|
| 343 |
+
"loss": 2.9743,
|
| 344 |
+
"step": 28000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.9,
|
| 348 |
+
"learning_rate": 4.9808625521153715e-05,
|
| 349 |
+
"loss": 2.9737,
|
| 350 |
+
"step": 28500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 3.96,
|
| 354 |
+
"learning_rate": 4.980520811974575e-05,
|
| 355 |
+
"loss": 2.9701,
|
| 356 |
+
"step": 29000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 4.03,
|
| 360 |
+
"learning_rate": 4.980179071833778e-05,
|
| 361 |
+
"loss": 2.9589,
|
| 362 |
+
"step": 29500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 4.1,
|
| 366 |
+
"learning_rate": 4.9798373316929805e-05,
|
| 367 |
+
"loss": 2.926,
|
| 368 |
+
"step": 30000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 4.17,
|
| 372 |
+
"learning_rate": 4.979495591552184e-05,
|
| 373 |
+
"loss": 2.9327,
|
| 374 |
+
"step": 30500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 4.24,
|
| 378 |
+
"learning_rate": 4.9791538514113874e-05,
|
| 379 |
+
"loss": 2.9401,
|
| 380 |
+
"step": 31000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 4.31,
|
| 384 |
+
"learning_rate": 4.97881211127059e-05,
|
| 385 |
+
"loss": 2.9345,
|
| 386 |
+
"step": 31500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 4.37,
|
| 390 |
+
"learning_rate": 4.978470371129793e-05,
|
| 391 |
+
"loss": 2.9348,
|
| 392 |
+
"step": 32000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 4.44,
|
| 396 |
+
"learning_rate": 4.9781286309889964e-05,
|
| 397 |
+
"loss": 2.9386,
|
| 398 |
+
"step": 32500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 4.51,
|
| 402 |
+
"learning_rate": 4.977786890848199e-05,
|
| 403 |
+
"loss": 2.9312,
|
| 404 |
+
"step": 33000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 4.58,
|
| 408 |
+
"learning_rate": 4.977445150707402e-05,
|
| 409 |
+
"loss": 2.9361,
|
| 410 |
+
"step": 33500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 4.65,
|
| 414 |
+
"learning_rate": 4.9771034105666054e-05,
|
| 415 |
+
"loss": 2.9316,
|
| 416 |
+
"step": 34000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 4.72,
|
| 420 |
+
"learning_rate": 4.976761670425809e-05,
|
| 421 |
+
"loss": 2.9397,
|
| 422 |
+
"step": 34500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 4.78,
|
| 426 |
+
"learning_rate": 4.9764199302850116e-05,
|
| 427 |
+
"loss": 2.9377,
|
| 428 |
+
"step": 35000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 4.85,
|
| 432 |
+
"learning_rate": 4.9760781901442144e-05,
|
| 433 |
+
"loss": 2.9378,
|
| 434 |
+
"step": 35500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 4.92,
|
| 438 |
+
"learning_rate": 4.975736450003418e-05,
|
| 439 |
+
"loss": 2.9306,
|
| 440 |
+
"step": 36000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 4.99,
|
| 444 |
+
"learning_rate": 4.9753947098626206e-05,
|
| 445 |
+
"loss": 2.9339,
|
| 446 |
+
"step": 36500
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 5.06,
|
| 450 |
+
"learning_rate": 4.975052969721824e-05,
|
| 451 |
+
"loss": 2.8964,
|
| 452 |
+
"step": 37000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 5.13,
|
| 456 |
+
"learning_rate": 4.974711229581027e-05,
|
| 457 |
+
"loss": 2.8929,
|
| 458 |
+
"step": 37500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 5.19,
|
| 462 |
+
"learning_rate": 4.9743694894402296e-05,
|
| 463 |
+
"loss": 2.901,
|
| 464 |
+
"step": 38000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 5.26,
|
| 468 |
+
"learning_rate": 4.974027749299433e-05,
|
| 469 |
+
"loss": 2.8934,
|
| 470 |
+
"step": 38500
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 5.33,
|
| 474 |
+
"learning_rate": 4.9736860091586364e-05,
|
| 475 |
+
"loss": 2.9019,
|
| 476 |
+
"step": 39000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 5.4,
|
| 480 |
+
"learning_rate": 4.973344269017839e-05,
|
| 481 |
+
"loss": 2.9037,
|
| 482 |
+
"step": 39500
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 5.47,
|
| 486 |
+
"learning_rate": 4.973002528877042e-05,
|
| 487 |
+
"loss": 2.8915,
|
| 488 |
+
"step": 40000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 5.54,
|
| 492 |
+
"learning_rate": 4.9726607887362454e-05,
|
| 493 |
+
"loss": 2.899,
|
| 494 |
+
"step": 40500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 5.6,
|
| 498 |
+
"learning_rate": 4.972319048595448e-05,
|
| 499 |
+
"loss": 2.9026,
|
| 500 |
+
"step": 41000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 5.67,
|
| 504 |
+
"learning_rate": 4.971977308454651e-05,
|
| 505 |
+
"loss": 2.9062,
|
| 506 |
+
"step": 41500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 5.74,
|
| 510 |
+
"learning_rate": 4.9716355683138544e-05,
|
| 511 |
+
"loss": 2.8974,
|
| 512 |
+
"step": 42000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 5.81,
|
| 516 |
+
"learning_rate": 4.971293828173058e-05,
|
| 517 |
+
"loss": 2.8969,
|
| 518 |
+
"step": 42500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 5.88,
|
| 522 |
+
"learning_rate": 4.9709520880322606e-05,
|
| 523 |
+
"loss": 2.8972,
|
| 524 |
+
"step": 43000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 5.95,
|
| 528 |
+
"learning_rate": 4.9706103478914634e-05,
|
| 529 |
+
"loss": 2.8994,
|
| 530 |
+
"step": 43500
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 6.01,
|
| 534 |
+
"learning_rate": 4.970268607750667e-05,
|
| 535 |
+
"loss": 2.8925,
|
| 536 |
+
"step": 44000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 6.08,
|
| 540 |
+
"learning_rate": 4.9699268676098696e-05,
|
| 541 |
+
"loss": 2.8637,
|
| 542 |
+
"step": 44500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 6.15,
|
| 546 |
+
"learning_rate": 4.969585127469073e-05,
|
| 547 |
+
"loss": 2.859,
|
| 548 |
+
"step": 45000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 6.22,
|
| 552 |
+
"learning_rate": 4.969243387328276e-05,
|
| 553 |
+
"loss": 2.8624,
|
| 554 |
+
"step": 45500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 6.29,
|
| 558 |
+
"learning_rate": 4.9689016471874786e-05,
|
| 559 |
+
"loss": 2.8595,
|
| 560 |
+
"step": 46000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 6.36,
|
| 564 |
+
"learning_rate": 4.968559907046682e-05,
|
| 565 |
+
"loss": 2.8632,
|
| 566 |
+
"step": 46500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 6.42,
|
| 570 |
+
"learning_rate": 4.9682181669058855e-05,
|
| 571 |
+
"loss": 2.8609,
|
| 572 |
+
"step": 47000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 6.49,
|
| 576 |
+
"learning_rate": 4.9678764267650876e-05,
|
| 577 |
+
"loss": 2.8702,
|
| 578 |
+
"step": 47500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 6.56,
|
| 582 |
+
"learning_rate": 4.967534686624291e-05,
|
| 583 |
+
"loss": 2.8714,
|
| 584 |
+
"step": 48000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 6.63,
|
| 588 |
+
"learning_rate": 4.9671929464834945e-05,
|
| 589 |
+
"loss": 2.8636,
|
| 590 |
+
"step": 48500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 6.7,
|
| 594 |
+
"learning_rate": 4.966851206342697e-05,
|
| 595 |
+
"loss": 2.8666,
|
| 596 |
+
"step": 49000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 6.77,
|
| 600 |
+
"learning_rate": 4.9665094662019e-05,
|
| 601 |
+
"loss": 2.8701,
|
| 602 |
+
"step": 49500
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 6.83,
|
| 606 |
+
"learning_rate": 4.9661677260611034e-05,
|
| 607 |
+
"loss": 2.87,
|
| 608 |
+
"step": 50000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 6.9,
|
| 612 |
+
"learning_rate": 4.965825985920307e-05,
|
| 613 |
+
"loss": 2.8608,
|
| 614 |
+
"step": 50500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 6.97,
|
| 618 |
+
"learning_rate": 4.96548424577951e-05,
|
| 619 |
+
"loss": 2.867,
|
| 620 |
+
"step": 51000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 7.04,
|
| 624 |
+
"learning_rate": 4.9651425056387124e-05,
|
| 625 |
+
"loss": 2.8444,
|
| 626 |
+
"step": 51500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 7.11,
|
| 630 |
+
"learning_rate": 4.964800765497916e-05,
|
| 631 |
+
"loss": 2.8337,
|
| 632 |
+
"step": 52000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 7.18,
|
| 636 |
+
"learning_rate": 4.9644590253571186e-05,
|
| 637 |
+
"loss": 2.8337,
|
| 638 |
+
"step": 52500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 7.24,
|
| 642 |
+
"learning_rate": 4.964117285216322e-05,
|
| 643 |
+
"loss": 2.8319,
|
| 644 |
+
"step": 53000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 7.31,
|
| 648 |
+
"learning_rate": 4.963775545075525e-05,
|
| 649 |
+
"loss": 2.8317,
|
| 650 |
+
"step": 53500
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 7.38,
|
| 654 |
+
"learning_rate": 4.9634338049347276e-05,
|
| 655 |
+
"loss": 2.8342,
|
| 656 |
+
"step": 54000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 7.45,
|
| 660 |
+
"learning_rate": 4.963092064793931e-05,
|
| 661 |
+
"loss": 2.8273,
|
| 662 |
+
"step": 54500
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 7.52,
|
| 666 |
+
"learning_rate": 4.962750324653134e-05,
|
| 667 |
+
"loss": 2.8347,
|
| 668 |
+
"step": 55000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 7.59,
|
| 672 |
+
"learning_rate": 4.9624085845123366e-05,
|
| 673 |
+
"loss": 2.8383,
|
| 674 |
+
"step": 55500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 7.65,
|
| 678 |
+
"learning_rate": 4.96206684437154e-05,
|
| 679 |
+
"loss": 2.8313,
|
| 680 |
+
"step": 56000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 7.72,
|
| 684 |
+
"learning_rate": 4.9617251042307435e-05,
|
| 685 |
+
"loss": 2.8387,
|
| 686 |
+
"step": 56500
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 7.79,
|
| 690 |
+
"learning_rate": 4.961383364089946e-05,
|
| 691 |
+
"loss": 2.834,
|
| 692 |
+
"step": 57000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 7.86,
|
| 696 |
+
"learning_rate": 4.961041623949149e-05,
|
| 697 |
+
"loss": 2.8377,
|
| 698 |
+
"step": 57500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 7.93,
|
| 702 |
+
"learning_rate": 4.9606998838083525e-05,
|
| 703 |
+
"loss": 2.8365,
|
| 704 |
+
"step": 58000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 8.0,
|
| 708 |
+
"learning_rate": 4.960358143667555e-05,
|
| 709 |
+
"loss": 2.8338,
|
| 710 |
+
"step": 58500
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 8.06,
|
| 714 |
+
"learning_rate": 4.960016403526758e-05,
|
| 715 |
+
"loss": 2.7898,
|
| 716 |
+
"step": 59000
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 8.13,
|
| 720 |
+
"learning_rate": 4.9596746633859615e-05,
|
| 721 |
+
"loss": 2.7955,
|
| 722 |
+
"step": 59500
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 8.2,
|
| 726 |
+
"learning_rate": 4.959332923245165e-05,
|
| 727 |
+
"loss": 2.7901,
|
| 728 |
+
"step": 60000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 8.27,
|
| 732 |
+
"learning_rate": 4.958991183104368e-05,
|
| 733 |
+
"loss": 2.7974,
|
| 734 |
+
"step": 60500
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 8.34,
|
| 738 |
+
"learning_rate": 4.9586494429635705e-05,
|
| 739 |
+
"loss": 2.8055,
|
| 740 |
+
"step": 61000
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 8.41,
|
| 744 |
+
"learning_rate": 4.958307702822774e-05,
|
| 745 |
+
"loss": 2.8088,
|
| 746 |
+
"step": 61500
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 8.47,
|
| 750 |
+
"learning_rate": 4.957965962681977e-05,
|
| 751 |
+
"loss": 2.8061,
|
| 752 |
+
"step": 62000
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 8.54,
|
| 756 |
+
"learning_rate": 4.95762422254118e-05,
|
| 757 |
+
"loss": 2.8005,
|
| 758 |
+
"step": 62500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 8.61,
|
| 762 |
+
"learning_rate": 4.957282482400383e-05,
|
| 763 |
+
"loss": 2.8056,
|
| 764 |
+
"step": 63000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 8.68,
|
| 768 |
+
"learning_rate": 4.9569407422595856e-05,
|
| 769 |
+
"loss": 2.8096,
|
| 770 |
+
"step": 63500
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 8.75,
|
| 774 |
+
"learning_rate": 4.956599002118789e-05,
|
| 775 |
+
"loss": 2.8077,
|
| 776 |
+
"step": 64000
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 8.82,
|
| 780 |
+
"learning_rate": 4.9562572619779925e-05,
|
| 781 |
+
"loss": 2.8107,
|
| 782 |
+
"step": 64500
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 8.88,
|
| 786 |
+
"learning_rate": 4.955915521837195e-05,
|
| 787 |
+
"loss": 2.8123,
|
| 788 |
+
"step": 65000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 8.95,
|
| 792 |
+
"learning_rate": 4.955573781696398e-05,
|
| 793 |
+
"loss": 2.8121,
|
| 794 |
+
"step": 65500
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 9.02,
|
| 798 |
+
"learning_rate": 4.9552320415556015e-05,
|
| 799 |
+
"loss": 2.7953,
|
| 800 |
+
"step": 66000
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 9.09,
|
| 804 |
+
"learning_rate": 4.954890301414804e-05,
|
| 805 |
+
"loss": 2.7705,
|
| 806 |
+
"step": 66500
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 9.16,
|
| 810 |
+
"learning_rate": 4.954548561274007e-05,
|
| 811 |
+
"loss": 2.7682,
|
| 812 |
+
"step": 67000
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 9.23,
|
| 816 |
+
"learning_rate": 4.9542068211332105e-05,
|
| 817 |
+
"loss": 2.7656,
|
| 818 |
+
"step": 67500
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 9.29,
|
| 822 |
+
"learning_rate": 4.953865080992414e-05,
|
| 823 |
+
"loss": 2.7742,
|
| 824 |
+
"step": 68000
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 9.36,
|
| 828 |
+
"learning_rate": 4.953523340851617e-05,
|
| 829 |
+
"loss": 2.7677,
|
| 830 |
+
"step": 68500
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 9.43,
|
| 834 |
+
"learning_rate": 4.9531816007108195e-05,
|
| 835 |
+
"loss": 2.7772,
|
| 836 |
+
"step": 69000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 9.5,
|
| 840 |
+
"learning_rate": 4.952839860570023e-05,
|
| 841 |
+
"loss": 2.7756,
|
| 842 |
+
"step": 69500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 9.57,
|
| 846 |
+
"learning_rate": 4.952498120429226e-05,
|
| 847 |
+
"loss": 2.7735,
|
| 848 |
+
"step": 70000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 9.64,
|
| 852 |
+
"learning_rate": 4.952156380288429e-05,
|
| 853 |
+
"loss": 2.7794,
|
| 854 |
+
"step": 70500
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 9.7,
|
| 858 |
+
"learning_rate": 4.951814640147632e-05,
|
| 859 |
+
"loss": 2.7803,
|
| 860 |
+
"step": 71000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 9.77,
|
| 864 |
+
"learning_rate": 4.951472900006835e-05,
|
| 865 |
+
"loss": 2.7819,
|
| 866 |
+
"step": 71500
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 9.84,
|
| 870 |
+
"learning_rate": 4.951131159866038e-05,
|
| 871 |
+
"loss": 2.7798,
|
| 872 |
+
"step": 72000
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 9.91,
|
| 876 |
+
"learning_rate": 4.9507894197252416e-05,
|
| 877 |
+
"loss": 2.7858,
|
| 878 |
+
"step": 72500
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 9.98,
|
| 882 |
+
"learning_rate": 4.950447679584444e-05,
|
| 883 |
+
"loss": 2.7831,
|
| 884 |
+
"step": 73000
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 10.05,
|
| 888 |
+
"learning_rate": 4.950105939443647e-05,
|
| 889 |
+
"loss": 2.7485,
|
| 890 |
+
"step": 73500
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 10.11,
|
| 894 |
+
"learning_rate": 4.9497641993028506e-05,
|
| 895 |
+
"loss": 2.7361,
|
| 896 |
+
"step": 74000
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 10.18,
|
| 900 |
+
"learning_rate": 4.949422459162053e-05,
|
| 901 |
+
"loss": 2.739,
|
| 902 |
+
"step": 74500
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 10.25,
|
| 906 |
+
"learning_rate": 4.949080719021256e-05,
|
| 907 |
+
"loss": 2.7434,
|
| 908 |
+
"step": 75000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 10.32,
|
| 912 |
+
"learning_rate": 4.9487389788804595e-05,
|
| 913 |
+
"loss": 2.7427,
|
| 914 |
+
"step": 75500
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 10.39,
|
| 918 |
+
"learning_rate": 4.948397238739663e-05,
|
| 919 |
+
"loss": 2.7424,
|
| 920 |
+
"step": 76000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 10.46,
|
| 924 |
+
"learning_rate": 4.948055498598866e-05,
|
| 925 |
+
"loss": 2.756,
|
| 926 |
+
"step": 76500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 10.52,
|
| 930 |
+
"learning_rate": 4.9477137584580685e-05,
|
| 931 |
+
"loss": 2.7434,
|
| 932 |
+
"step": 77000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 10.59,
|
| 936 |
+
"learning_rate": 4.947372018317272e-05,
|
| 937 |
+
"loss": 2.7512,
|
| 938 |
+
"step": 77500
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 10.66,
|
| 942 |
+
"learning_rate": 4.947030278176475e-05,
|
| 943 |
+
"loss": 2.7527,
|
| 944 |
+
"step": 78000
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 10.73,
|
| 948 |
+
"learning_rate": 4.946688538035678e-05,
|
| 949 |
+
"loss": 2.7605,
|
| 950 |
+
"step": 78500
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 10.8,
|
| 954 |
+
"learning_rate": 4.946346797894881e-05,
|
| 955 |
+
"loss": 2.7624,
|
| 956 |
+
"step": 79000
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 10.87,
|
| 960 |
+
"learning_rate": 4.946005057754084e-05,
|
| 961 |
+
"loss": 2.7563,
|
| 962 |
+
"step": 79500
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 10.93,
|
| 966 |
+
"learning_rate": 4.945663317613287e-05,
|
| 967 |
+
"loss": 2.7531,
|
| 968 |
+
"step": 80000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 11.0,
|
| 972 |
+
"learning_rate": 4.9453215774724906e-05,
|
| 973 |
+
"loss": 2.7539,
|
| 974 |
+
"step": 80500
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 11.07,
|
| 978 |
+
"learning_rate": 4.944979837331693e-05,
|
| 979 |
+
"loss": 2.7114,
|
| 980 |
+
"step": 81000
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 11.14,
|
| 984 |
+
"learning_rate": 4.944638097190896e-05,
|
| 985 |
+
"loss": 2.7128,
|
| 986 |
+
"step": 81500
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 11.21,
|
| 990 |
+
"learning_rate": 4.9442963570500996e-05,
|
| 991 |
+
"loss": 2.7177,
|
| 992 |
+
"step": 82000
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 11.28,
|
| 996 |
+
"learning_rate": 4.9439546169093024e-05,
|
| 997 |
+
"loss": 2.7212,
|
| 998 |
+
"step": 82500
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 11.34,
|
| 1002 |
+
"learning_rate": 4.943612876768505e-05,
|
| 1003 |
+
"loss": 2.7224,
|
| 1004 |
+
"step": 83000
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 11.41,
|
| 1008 |
+
"learning_rate": 4.9432711366277086e-05,
|
| 1009 |
+
"loss": 2.7234,
|
| 1010 |
+
"step": 83500
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 11.48,
|
| 1014 |
+
"learning_rate": 4.9429293964869113e-05,
|
| 1015 |
+
"loss": 2.7233,
|
| 1016 |
+
"step": 84000
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 11.55,
|
| 1020 |
+
"learning_rate": 4.942587656346115e-05,
|
| 1021 |
+
"loss": 2.7297,
|
| 1022 |
+
"step": 84500
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 11.62,
|
| 1026 |
+
"learning_rate": 4.9422459162053176e-05,
|
| 1027 |
+
"loss": 2.7211,
|
| 1028 |
+
"step": 85000
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 11.69,
|
| 1032 |
+
"learning_rate": 4.941904176064521e-05,
|
| 1033 |
+
"loss": 2.7282,
|
| 1034 |
+
"step": 85500
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 11.76,
|
| 1038 |
+
"learning_rate": 4.941562435923724e-05,
|
| 1039 |
+
"loss": 2.7272,
|
| 1040 |
+
"step": 86000
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 11.82,
|
| 1044 |
+
"learning_rate": 4.941220695782927e-05,
|
| 1045 |
+
"loss": 2.7325,
|
| 1046 |
+
"step": 86500
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 11.89,
|
| 1050 |
+
"learning_rate": 4.94087895564213e-05,
|
| 1051 |
+
"loss": 2.7275,
|
| 1052 |
+
"step": 87000
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 11.96,
|
| 1056 |
+
"learning_rate": 4.940537215501333e-05,
|
| 1057 |
+
"loss": 2.7303,
|
| 1058 |
+
"step": 87500
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 12.03,
|
| 1062 |
+
"learning_rate": 4.940195475360536e-05,
|
| 1063 |
+
"loss": 2.7091,
|
| 1064 |
+
"step": 88000
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 12.1,
|
| 1068 |
+
"learning_rate": 4.9398537352197397e-05,
|
| 1069 |
+
"loss": 2.6863,
|
| 1070 |
+
"step": 88500
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 12.17,
|
| 1074 |
+
"learning_rate": 4.939511995078942e-05,
|
| 1075 |
+
"loss": 2.6867,
|
| 1076 |
+
"step": 89000
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 12.23,
|
| 1080 |
+
"learning_rate": 4.939170254938145e-05,
|
| 1081 |
+
"loss": 2.6934,
|
| 1082 |
+
"step": 89500
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 12.3,
|
| 1086 |
+
"learning_rate": 4.9388285147973486e-05,
|
| 1087 |
+
"loss": 2.6936,
|
| 1088 |
+
"step": 90000
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 12.37,
|
| 1092 |
+
"learning_rate": 4.9384867746565514e-05,
|
| 1093 |
+
"loss": 2.6953,
|
| 1094 |
+
"step": 90500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 12.44,
|
| 1098 |
+
"learning_rate": 4.938145034515754e-05,
|
| 1099 |
+
"loss": 2.6973,
|
| 1100 |
+
"step": 91000
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 12.51,
|
| 1104 |
+
"learning_rate": 4.9378032943749576e-05,
|
| 1105 |
+
"loss": 2.7022,
|
| 1106 |
+
"step": 91500
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 12.58,
|
| 1110 |
+
"learning_rate": 4.9374615542341604e-05,
|
| 1111 |
+
"loss": 2.6977,
|
| 1112 |
+
"step": 92000
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 12.64,
|
| 1116 |
+
"learning_rate": 4.937119814093364e-05,
|
| 1117 |
+
"loss": 2.6996,
|
| 1118 |
+
"step": 92500
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 12.71,
|
| 1122 |
+
"learning_rate": 4.9367780739525666e-05,
|
| 1123 |
+
"loss": 2.7008,
|
| 1124 |
+
"step": 93000
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 12.78,
|
| 1128 |
+
"learning_rate": 4.93643633381177e-05,
|
| 1129 |
+
"loss": 2.7091,
|
| 1130 |
+
"step": 93500
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 12.85,
|
| 1134 |
+
"learning_rate": 4.936094593670973e-05,
|
| 1135 |
+
"loss": 2.7057,
|
| 1136 |
+
"step": 94000
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 12.92,
|
| 1140 |
+
"learning_rate": 4.935752853530176e-05,
|
| 1141 |
+
"loss": 2.7084,
|
| 1142 |
+
"step": 94500
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 12.99,
|
| 1146 |
+
"learning_rate": 4.935411113389379e-05,
|
| 1147 |
+
"loss": 2.7114,
|
| 1148 |
+
"step": 95000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 13.05,
|
| 1152 |
+
"learning_rate": 4.935069373248582e-05,
|
| 1153 |
+
"loss": 2.6727,
|
| 1154 |
+
"step": 95500
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 13.12,
|
| 1158 |
+
"learning_rate": 4.934727633107785e-05,
|
| 1159 |
+
"loss": 2.6586,
|
| 1160 |
+
"step": 96000
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 13.19,
|
| 1164 |
+
"learning_rate": 4.934385892966989e-05,
|
| 1165 |
+
"loss": 2.6631,
|
| 1166 |
+
"step": 96500
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 13.26,
|
| 1170 |
+
"learning_rate": 4.934044152826191e-05,
|
| 1171 |
+
"loss": 2.6632,
|
| 1172 |
+
"step": 97000
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 13.33,
|
| 1176 |
+
"learning_rate": 4.933702412685394e-05,
|
| 1177 |
+
"loss": 2.6727,
|
| 1178 |
+
"step": 97500
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 13.4,
|
| 1182 |
+
"learning_rate": 4.933360672544598e-05,
|
| 1183 |
+
"loss": 2.6671,
|
| 1184 |
+
"step": 98000
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 13.46,
|
| 1188 |
+
"learning_rate": 4.9330189324038004e-05,
|
| 1189 |
+
"loss": 2.686,
|
| 1190 |
+
"step": 98500
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 13.53,
|
| 1194 |
+
"learning_rate": 4.932677192263003e-05,
|
| 1195 |
+
"loss": 2.6796,
|
| 1196 |
+
"step": 99000
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 13.6,
|
| 1200 |
+
"learning_rate": 4.9323354521222067e-05,
|
| 1201 |
+
"loss": 2.6745,
|
| 1202 |
+
"step": 99500
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 13.67,
|
| 1206 |
+
"learning_rate": 4.9319937119814094e-05,
|
| 1207 |
+
"loss": 2.6825,
|
| 1208 |
+
"step": 100000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 13.74,
|
| 1212 |
+
"learning_rate": 4.931651971840613e-05,
|
| 1213 |
+
"loss": 2.6857,
|
| 1214 |
+
"step": 100500
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 13.81,
|
| 1218 |
+
"learning_rate": 4.9313102316998156e-05,
|
| 1219 |
+
"loss": 2.6812,
|
| 1220 |
+
"step": 101000
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 13.87,
|
| 1224 |
+
"learning_rate": 4.9309684915590184e-05,
|
| 1225 |
+
"loss": 2.6805,
|
| 1226 |
+
"step": 101500
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 13.94,
|
| 1230 |
+
"learning_rate": 4.930626751418222e-05,
|
| 1231 |
+
"loss": 2.6857,
|
| 1232 |
+
"step": 102000
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 14.01,
|
| 1236 |
+
"learning_rate": 4.930285011277425e-05,
|
| 1237 |
+
"loss": 2.6768,
|
| 1238 |
+
"step": 102500
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 14.08,
|
| 1242 |
+
"learning_rate": 4.929943271136628e-05,
|
| 1243 |
+
"loss": 2.6435,
|
| 1244 |
+
"step": 103000
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 14.15,
|
| 1248 |
+
"learning_rate": 4.929601530995831e-05,
|
| 1249 |
+
"loss": 2.6508,
|
| 1250 |
+
"step": 103500
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 14.22,
|
| 1254 |
+
"learning_rate": 4.929259790855034e-05,
|
| 1255 |
+
"loss": 2.6444,
|
| 1256 |
+
"step": 104000
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 14.28,
|
| 1260 |
+
"learning_rate": 4.928918050714238e-05,
|
| 1261 |
+
"loss": 2.6396,
|
| 1262 |
+
"step": 104500
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 14.35,
|
| 1266 |
+
"learning_rate": 4.92857631057344e-05,
|
| 1267 |
+
"loss": 2.6452,
|
| 1268 |
+
"step": 105000
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 14.42,
|
| 1272 |
+
"learning_rate": 4.928234570432643e-05,
|
| 1273 |
+
"loss": 2.6535,
|
| 1274 |
+
"step": 105500
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 14.49,
|
| 1278 |
+
"learning_rate": 4.927892830291847e-05,
|
| 1279 |
+
"loss": 2.6482,
|
| 1280 |
+
"step": 106000
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 14.56,
|
| 1284 |
+
"learning_rate": 4.9275510901510495e-05,
|
| 1285 |
+
"loss": 2.6567,
|
| 1286 |
+
"step": 106500
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 14.63,
|
| 1290 |
+
"learning_rate": 4.927209350010252e-05,
|
| 1291 |
+
"loss": 2.6589,
|
| 1292 |
+
"step": 107000
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 14.69,
|
| 1296 |
+
"learning_rate": 4.926867609869456e-05,
|
| 1297 |
+
"loss": 2.6564,
|
| 1298 |
+
"step": 107500
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 14.76,
|
| 1302 |
+
"learning_rate": 4.9265258697286585e-05,
|
| 1303 |
+
"loss": 2.6579,
|
| 1304 |
+
"step": 108000
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 14.83,
|
| 1308 |
+
"learning_rate": 4.926184129587861e-05,
|
| 1309 |
+
"loss": 2.658,
|
| 1310 |
+
"step": 108500
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 14.9,
|
| 1314 |
+
"learning_rate": 4.925842389447065e-05,
|
| 1315 |
+
"loss": 2.6658,
|
| 1316 |
+
"step": 109000
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 14.97,
|
| 1320 |
+
"learning_rate": 4.9255006493062674e-05,
|
| 1321 |
+
"loss": 2.6535,
|
| 1322 |
+
"step": 109500
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 15.04,
|
| 1326 |
+
"learning_rate": 4.925158909165471e-05,
|
| 1327 |
+
"loss": 2.6404,
|
| 1328 |
+
"step": 110000
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 15.1,
|
| 1332 |
+
"learning_rate": 4.9248171690246737e-05,
|
| 1333 |
+
"loss": 2.6185,
|
| 1334 |
+
"step": 110500
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 15.17,
|
| 1338 |
+
"learning_rate": 4.924475428883877e-05,
|
| 1339 |
+
"loss": 2.6158,
|
| 1340 |
+
"step": 111000
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 15.24,
|
| 1344 |
+
"learning_rate": 4.92413368874308e-05,
|
| 1345 |
+
"loss": 2.621,
|
| 1346 |
+
"step": 111500
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 15.31,
|
| 1350 |
+
"learning_rate": 4.923791948602283e-05,
|
| 1351 |
+
"loss": 2.6296,
|
| 1352 |
+
"step": 112000
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 15.38,
|
| 1356 |
+
"learning_rate": 4.923450208461486e-05,
|
| 1357 |
+
"loss": 2.6284,
|
| 1358 |
+
"step": 112500
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 15.45,
|
| 1362 |
+
"learning_rate": 4.923108468320689e-05,
|
| 1363 |
+
"loss": 2.6252,
|
| 1364 |
+
"step": 113000
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 15.51,
|
| 1368 |
+
"learning_rate": 4.922766728179892e-05,
|
| 1369 |
+
"loss": 2.6355,
|
| 1370 |
+
"step": 113500
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 15.58,
|
| 1374 |
+
"learning_rate": 4.922424988039096e-05,
|
| 1375 |
+
"loss": 2.6337,
|
| 1376 |
+
"step": 114000
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 15.65,
|
| 1380 |
+
"learning_rate": 4.922083247898298e-05,
|
| 1381 |
+
"loss": 2.6337,
|
| 1382 |
+
"step": 114500
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 15.72,
|
| 1386 |
+
"learning_rate": 4.921741507757501e-05,
|
| 1387 |
+
"loss": 2.6354,
|
| 1388 |
+
"step": 115000
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 15.79,
|
| 1392 |
+
"learning_rate": 4.921399767616705e-05,
|
| 1393 |
+
"loss": 2.6371,
|
| 1394 |
+
"step": 115500
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 15.86,
|
| 1398 |
+
"learning_rate": 4.9210580274759075e-05,
|
| 1399 |
+
"loss": 2.6427,
|
| 1400 |
+
"step": 116000
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 15.92,
|
| 1404 |
+
"learning_rate": 4.92071628733511e-05,
|
| 1405 |
+
"loss": 2.639,
|
| 1406 |
+
"step": 116500
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 15.99,
|
| 1410 |
+
"learning_rate": 4.920374547194314e-05,
|
| 1411 |
+
"loss": 2.6426,
|
| 1412 |
+
"step": 117000
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 16.06,
|
| 1416 |
+
"learning_rate": 4.9200328070535165e-05,
|
| 1417 |
+
"loss": 2.5977,
|
| 1418 |
+
"step": 117500
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 16.13,
|
| 1422 |
+
"learning_rate": 4.91969106691272e-05,
|
| 1423 |
+
"loss": 2.5961,
|
| 1424 |
+
"step": 118000
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 16.2,
|
| 1428 |
+
"learning_rate": 4.919349326771923e-05,
|
| 1429 |
+
"loss": 2.6007,
|
| 1430 |
+
"step": 118500
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 16.27,
|
| 1434 |
+
"learning_rate": 4.919007586631126e-05,
|
| 1435 |
+
"loss": 2.6015,
|
| 1436 |
+
"step": 119000
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 16.33,
|
| 1440 |
+
"learning_rate": 4.918665846490329e-05,
|
| 1441 |
+
"loss": 2.6064,
|
| 1442 |
+
"step": 119500
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 16.4,
|
| 1446 |
+
"learning_rate": 4.9183241063495324e-05,
|
| 1447 |
+
"loss": 2.6125,
|
| 1448 |
+
"step": 120000
|
| 1449 |
+
}
|
| 1450 |
+
],
|
| 1451 |
+
"max_steps": 7316000,
|
| 1452 |
+
"num_train_epochs": 1000,
|
| 1453 |
+
"total_flos": 241553911465377792,
|
| 1454 |
+
"trial_name": null,
|
| 1455 |
+
"trial_params": null
|
| 1456 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ff665c6e301aa4e24e46b78e013048f01082a4fab4837d2da90ed3e2aa86e29c
|
| 3 |
+
size 2031
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|