Upload folder using huggingface_hub
Browse files- config.json +45 -0
- merges.txt +0 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +15 -0
- tokenizer.json +0 -0
- tokenizer_config.json +16 -0
- trainer_state.json +1528 -0
- training_args.bin +3 -0
- vocab.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "roberta-base",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"RobertaForTokenClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": 0,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"hidden_act": "gelu",
|
| 11 |
+
"hidden_dropout_prob": 0.1,
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"id2label": {
|
| 14 |
+
"0": "O",
|
| 15 |
+
"1": "B-GENE",
|
| 16 |
+
"2": "I-GENE",
|
| 17 |
+
"3": "B-DRUG",
|
| 18 |
+
"4": "I-DRUG",
|
| 19 |
+
"5": "B-DISEASE",
|
| 20 |
+
"6": "I-DISEASE"
|
| 21 |
+
},
|
| 22 |
+
"initializer_range": 0.02,
|
| 23 |
+
"intermediate_size": 3072,
|
| 24 |
+
"label2id": {
|
| 25 |
+
"B-DISEASE": 5,
|
| 26 |
+
"B-DRUG": 3,
|
| 27 |
+
"B-GENE": 1,
|
| 28 |
+
"I-DISEASE": 6,
|
| 29 |
+
"I-DRUG": 4,
|
| 30 |
+
"I-GENE": 2,
|
| 31 |
+
"O": 0
|
| 32 |
+
},
|
| 33 |
+
"layer_norm_eps": 1e-05,
|
| 34 |
+
"max_position_embeddings": 514,
|
| 35 |
+
"model_type": "roberta",
|
| 36 |
+
"num_attention_heads": 12,
|
| 37 |
+
"num_hidden_layers": 12,
|
| 38 |
+
"pad_token_id": 1,
|
| 39 |
+
"position_embedding_type": "absolute",
|
| 40 |
+
"torch_dtype": "float32",
|
| 41 |
+
"transformers_version": "4.24.0",
|
| 42 |
+
"type_vocab_size": 1,
|
| 43 |
+
"use_cache": true,
|
| 44 |
+
"vocab_size": 50265
|
| 45 |
+
}
|
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 sha256:2af4a4c50e4cb55207b7607b3b1986413622e319aac10e2a8deb7a48c24a0a17
|
| 3 |
+
size 992649285
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:85882f23215fb5199c3265310ca347438dbef8b5dd7002e7ef3b82174ad4a236
|
| 3 |
+
size 496314413
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1fa88decf973bed3833ccf02efaec6a3bb6b2caedc6cd9c35381717d53ba19c2
|
| 3 |
+
size 16619
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:97a256e0525a47eb457c7f53901d5c9aa56d2b471248091a69fc730a0d8b9fea
|
| 3 |
+
size 557
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0264819fb4e648770b35735a3e892f6e30c67eda23735302ddf4f6065e43f951
|
| 3 |
+
size 627
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": {
|
| 6 |
+
"content": "<mask>",
|
| 7 |
+
"lstrip": true,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"bos_token": "<s>",
|
| 4 |
+
"cls_token": "<s>",
|
| 5 |
+
"eos_token": "</s>",
|
| 6 |
+
"errors": "replace",
|
| 7 |
+
"mask_token": "<mask>",
|
| 8 |
+
"model_max_length": 512,
|
| 9 |
+
"name_or_path": "roberta-base",
|
| 10 |
+
"pad_token": "<pad>",
|
| 11 |
+
"sep_token": "</s>",
|
| 12 |
+
"special_tokens_map_file": null,
|
| 13 |
+
"tokenizer_class": "RobertaTokenizer",
|
| 14 |
+
"trim_offsets": true,
|
| 15 |
+
"unk_token": "<unk>"
|
| 16 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1528 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9981779291768993,
|
| 5 |
+
"global_step": 126000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
+
"learning_rate": 4.999342470094273e-05,
|
| 13 |
+
"loss": 0.0915,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.01,
|
| 18 |
+
"learning_rate": 4.998682299506193e-05,
|
| 19 |
+
"loss": 0.0476,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 4.998022128918113e-05,
|
| 25 |
+
"loss": 0.0439,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.02,
|
| 30 |
+
"learning_rate": 4.997361958330033e-05,
|
| 31 |
+
"loss": 0.0416,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.02,
|
| 36 |
+
"learning_rate": 4.996701787741953e-05,
|
| 37 |
+
"loss": 0.04,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.02,
|
| 42 |
+
"learning_rate": 4.996041617153873e-05,
|
| 43 |
+
"loss": 0.0388,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.03,
|
| 48 |
+
"learning_rate": 4.995381446565793e-05,
|
| 49 |
+
"loss": 0.0384,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.03,
|
| 54 |
+
"learning_rate": 4.994721275977713e-05,
|
| 55 |
+
"loss": 0.037,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.04,
|
| 60 |
+
"learning_rate": 4.994061105389633e-05,
|
| 61 |
+
"loss": 0.0362,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.04,
|
| 66 |
+
"learning_rate": 4.993400934801553e-05,
|
| 67 |
+
"loss": 0.0359,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.04,
|
| 72 |
+
"learning_rate": 4.9927407642134734e-05,
|
| 73 |
+
"loss": 0.0353,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.05,
|
| 78 |
+
"learning_rate": 4.992080593625393e-05,
|
| 79 |
+
"loss": 0.0349,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.05,
|
| 84 |
+
"learning_rate": 4.991420423037313e-05,
|
| 85 |
+
"loss": 0.0348,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.06,
|
| 90 |
+
"learning_rate": 4.990760252449233e-05,
|
| 91 |
+
"loss": 0.0338,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.06,
|
| 96 |
+
"learning_rate": 4.990100081861153e-05,
|
| 97 |
+
"loss": 0.0332,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.06,
|
| 102 |
+
"learning_rate": 4.989439911273073e-05,
|
| 103 |
+
"loss": 0.0332,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.07,
|
| 108 |
+
"learning_rate": 4.9887797406849934e-05,
|
| 109 |
+
"loss": 0.0331,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.07,
|
| 114 |
+
"learning_rate": 4.9881195700969135e-05,
|
| 115 |
+
"loss": 0.0325,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.08,
|
| 120 |
+
"learning_rate": 4.987459399508833e-05,
|
| 121 |
+
"loss": 0.0322,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.08,
|
| 126 |
+
"learning_rate": 4.986799228920753e-05,
|
| 127 |
+
"loss": 0.0321,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.08,
|
| 132 |
+
"learning_rate": 4.986139058332673e-05,
|
| 133 |
+
"loss": 0.0318,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.09,
|
| 138 |
+
"learning_rate": 4.985478887744593e-05,
|
| 139 |
+
"loss": 0.0317,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.09,
|
| 144 |
+
"learning_rate": 4.9848187171565134e-05,
|
| 145 |
+
"loss": 0.0313,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.1,
|
| 150 |
+
"learning_rate": 4.9841585465684335e-05,
|
| 151 |
+
"loss": 0.0312,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.1,
|
| 156 |
+
"learning_rate": 4.9834983759803536e-05,
|
| 157 |
+
"loss": 0.031,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.1,
|
| 162 |
+
"learning_rate": 4.982838205392274e-05,
|
| 163 |
+
"loss": 0.0307,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.11,
|
| 168 |
+
"learning_rate": 4.982178034804194e-05,
|
| 169 |
+
"loss": 0.0306,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.11,
|
| 174 |
+
"learning_rate": 4.981517864216114e-05,
|
| 175 |
+
"loss": 0.0301,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.11,
|
| 180 |
+
"learning_rate": 4.980859013969209e-05,
|
| 181 |
+
"loss": 0.0301,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.12,
|
| 186 |
+
"learning_rate": 4.9801988433811294e-05,
|
| 187 |
+
"loss": 0.0301,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.12,
|
| 192 |
+
"learning_rate": 4.979539993134226e-05,
|
| 193 |
+
"loss": 0.0301,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.13,
|
| 198 |
+
"learning_rate": 4.978879822546146e-05,
|
| 199 |
+
"loss": 0.0298,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.13,
|
| 204 |
+
"learning_rate": 4.978219651958066e-05,
|
| 205 |
+
"loss": 0.0298,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.13,
|
| 210 |
+
"learning_rate": 4.977559481369986e-05,
|
| 211 |
+
"loss": 0.0294,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.14,
|
| 216 |
+
"learning_rate": 4.976899310781906e-05,
|
| 217 |
+
"loss": 0.0293,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.14,
|
| 222 |
+
"learning_rate": 4.976239140193826e-05,
|
| 223 |
+
"loss": 0.0294,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.15,
|
| 228 |
+
"learning_rate": 4.975578969605746e-05,
|
| 229 |
+
"loss": 0.0292,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.15,
|
| 234 |
+
"learning_rate": 4.974918799017666e-05,
|
| 235 |
+
"loss": 0.0293,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.15,
|
| 240 |
+
"learning_rate": 4.9742586284295864e-05,
|
| 241 |
+
"loss": 0.029,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.16,
|
| 246 |
+
"learning_rate": 4.9735984578415065e-05,
|
| 247 |
+
"loss": 0.0288,
|
| 248 |
+
"step": 20000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.16,
|
| 252 |
+
"learning_rate": 4.9729396075946025e-05,
|
| 253 |
+
"loss": 0.0286,
|
| 254 |
+
"step": 20500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.17,
|
| 258 |
+
"learning_rate": 4.9722794370065227e-05,
|
| 259 |
+
"loss": 0.0289,
|
| 260 |
+
"step": 21000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.17,
|
| 264 |
+
"learning_rate": 4.971619266418443e-05,
|
| 265 |
+
"loss": 0.0285,
|
| 266 |
+
"step": 21500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.17,
|
| 270 |
+
"learning_rate": 4.970959095830363e-05,
|
| 271 |
+
"loss": 0.0287,
|
| 272 |
+
"step": 22000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.18,
|
| 276 |
+
"learning_rate": 4.970298925242283e-05,
|
| 277 |
+
"loss": 0.0284,
|
| 278 |
+
"step": 22500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.18,
|
| 282 |
+
"learning_rate": 4.969641395336555e-05,
|
| 283 |
+
"loss": 0.0284,
|
| 284 |
+
"step": 23000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.19,
|
| 288 |
+
"learning_rate": 4.968981224748475e-05,
|
| 289 |
+
"loss": 0.0284,
|
| 290 |
+
"step": 23500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.19,
|
| 294 |
+
"learning_rate": 4.968321054160395e-05,
|
| 295 |
+
"loss": 0.0282,
|
| 296 |
+
"step": 24000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.19,
|
| 300 |
+
"learning_rate": 4.967662203913491e-05,
|
| 301 |
+
"loss": 0.0278,
|
| 302 |
+
"step": 24500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.2,
|
| 306 |
+
"learning_rate": 4.9670020333254114e-05,
|
| 307 |
+
"loss": 0.0281,
|
| 308 |
+
"step": 25000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.2,
|
| 312 |
+
"learning_rate": 4.9663418627373315e-05,
|
| 313 |
+
"loss": 0.028,
|
| 314 |
+
"step": 25500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.21,
|
| 318 |
+
"learning_rate": 4.9656816921492516e-05,
|
| 319 |
+
"loss": 0.0278,
|
| 320 |
+
"step": 26000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.21,
|
| 324 |
+
"learning_rate": 4.965021521561172e-05,
|
| 325 |
+
"loss": 0.0278,
|
| 326 |
+
"step": 26500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.21,
|
| 330 |
+
"learning_rate": 4.964361350973092e-05,
|
| 331 |
+
"loss": 0.0276,
|
| 332 |
+
"step": 27000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.22,
|
| 336 |
+
"learning_rate": 4.963701180385012e-05,
|
| 337 |
+
"loss": 0.0276,
|
| 338 |
+
"step": 27500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.22,
|
| 342 |
+
"learning_rate": 4.963042330138108e-05,
|
| 343 |
+
"loss": 0.0276,
|
| 344 |
+
"step": 28000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.23,
|
| 348 |
+
"learning_rate": 4.962382159550028e-05,
|
| 349 |
+
"loss": 0.0275,
|
| 350 |
+
"step": 28500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.23,
|
| 354 |
+
"learning_rate": 4.961721988961948e-05,
|
| 355 |
+
"loss": 0.0274,
|
| 356 |
+
"step": 29000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.23,
|
| 360 |
+
"learning_rate": 4.961061818373868e-05,
|
| 361 |
+
"loss": 0.0274,
|
| 362 |
+
"step": 29500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.24,
|
| 366 |
+
"learning_rate": 4.960401647785788e-05,
|
| 367 |
+
"loss": 0.0276,
|
| 368 |
+
"step": 30000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.24,
|
| 372 |
+
"learning_rate": 4.959741477197708e-05,
|
| 373 |
+
"loss": 0.0272,
|
| 374 |
+
"step": 30500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.25,
|
| 378 |
+
"learning_rate": 4.9590826269508046e-05,
|
| 379 |
+
"loss": 0.0272,
|
| 380 |
+
"step": 31000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.25,
|
| 384 |
+
"learning_rate": 4.958422456362725e-05,
|
| 385 |
+
"loss": 0.027,
|
| 386 |
+
"step": 31500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.25,
|
| 390 |
+
"learning_rate": 4.957762285774644e-05,
|
| 391 |
+
"loss": 0.027,
|
| 392 |
+
"step": 32000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.26,
|
| 396 |
+
"learning_rate": 4.957102115186564e-05,
|
| 397 |
+
"loss": 0.027,
|
| 398 |
+
"step": 32500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.26,
|
| 402 |
+
"learning_rate": 4.9564419445984844e-05,
|
| 403 |
+
"loss": 0.0271,
|
| 404 |
+
"step": 33000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.27,
|
| 408 |
+
"learning_rate": 4.9557817740104045e-05,
|
| 409 |
+
"loss": 0.0268,
|
| 410 |
+
"step": 33500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.27,
|
| 414 |
+
"learning_rate": 4.9551216034223246e-05,
|
| 415 |
+
"loss": 0.0272,
|
| 416 |
+
"step": 34000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.27,
|
| 420 |
+
"learning_rate": 4.954461432834245e-05,
|
| 421 |
+
"loss": 0.0269,
|
| 422 |
+
"step": 34500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.28,
|
| 426 |
+
"learning_rate": 4.953801262246165e-05,
|
| 427 |
+
"loss": 0.0268,
|
| 428 |
+
"step": 35000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.28,
|
| 432 |
+
"learning_rate": 4.953141091658084e-05,
|
| 433 |
+
"loss": 0.027,
|
| 434 |
+
"step": 35500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.29,
|
| 438 |
+
"learning_rate": 4.9524809210700044e-05,
|
| 439 |
+
"loss": 0.0269,
|
| 440 |
+
"step": 36000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.29,
|
| 444 |
+
"learning_rate": 4.9518207504819245e-05,
|
| 445 |
+
"loss": 0.0266,
|
| 446 |
+
"step": 36500
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.29,
|
| 450 |
+
"learning_rate": 4.951163220576197e-05,
|
| 451 |
+
"loss": 0.0265,
|
| 452 |
+
"step": 37000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.3,
|
| 456 |
+
"learning_rate": 4.950503049988117e-05,
|
| 457 |
+
"loss": 0.0265,
|
| 458 |
+
"step": 37500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.3,
|
| 462 |
+
"learning_rate": 4.9498428794000374e-05,
|
| 463 |
+
"loss": 0.0265,
|
| 464 |
+
"step": 38000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.3,
|
| 468 |
+
"learning_rate": 4.9491827088119575e-05,
|
| 469 |
+
"loss": 0.0267,
|
| 470 |
+
"step": 38500
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.31,
|
| 474 |
+
"learning_rate": 4.948522538223877e-05,
|
| 475 |
+
"loss": 0.0263,
|
| 476 |
+
"step": 39000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.31,
|
| 480 |
+
"learning_rate": 4.947862367635797e-05,
|
| 481 |
+
"loss": 0.0262,
|
| 482 |
+
"step": 39500
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.32,
|
| 486 |
+
"learning_rate": 4.947203517388894e-05,
|
| 487 |
+
"loss": 0.0261,
|
| 488 |
+
"step": 40000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.32,
|
| 492 |
+
"learning_rate": 4.946543346800814e-05,
|
| 493 |
+
"loss": 0.0264,
|
| 494 |
+
"step": 40500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.32,
|
| 498 |
+
"learning_rate": 4.945883176212733e-05,
|
| 499 |
+
"loss": 0.0266,
|
| 500 |
+
"step": 41000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.33,
|
| 504 |
+
"learning_rate": 4.9452230056246534e-05,
|
| 505 |
+
"loss": 0.0262,
|
| 506 |
+
"step": 41500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.33,
|
| 510 |
+
"learning_rate": 4.94456415537775e-05,
|
| 511 |
+
"loss": 0.0263,
|
| 512 |
+
"step": 42000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.34,
|
| 516 |
+
"learning_rate": 4.94390398478967e-05,
|
| 517 |
+
"loss": 0.0261,
|
| 518 |
+
"step": 42500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.34,
|
| 522 |
+
"learning_rate": 4.94324381420159e-05,
|
| 523 |
+
"loss": 0.026,
|
| 524 |
+
"step": 43000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.34,
|
| 528 |
+
"learning_rate": 4.94258364361351e-05,
|
| 529 |
+
"loss": 0.0258,
|
| 530 |
+
"step": 43500
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.35,
|
| 534 |
+
"learning_rate": 4.94192347302543e-05,
|
| 535 |
+
"loss": 0.0261,
|
| 536 |
+
"step": 44000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.35,
|
| 540 |
+
"learning_rate": 4.941264622778527e-05,
|
| 541 |
+
"loss": 0.0261,
|
| 542 |
+
"step": 44500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.36,
|
| 546 |
+
"learning_rate": 4.940604452190446e-05,
|
| 547 |
+
"loss": 0.0262,
|
| 548 |
+
"step": 45000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.36,
|
| 552 |
+
"learning_rate": 4.939944281602366e-05,
|
| 553 |
+
"loss": 0.026,
|
| 554 |
+
"step": 45500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.36,
|
| 558 |
+
"learning_rate": 4.939284111014286e-05,
|
| 559 |
+
"loss": 0.0259,
|
| 560 |
+
"step": 46000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.37,
|
| 564 |
+
"learning_rate": 4.9386239404262064e-05,
|
| 565 |
+
"loss": 0.0259,
|
| 566 |
+
"step": 46500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.37,
|
| 570 |
+
"learning_rate": 4.9379637698381266e-05,
|
| 571 |
+
"loss": 0.0258,
|
| 572 |
+
"step": 47000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.38,
|
| 576 |
+
"learning_rate": 4.937303599250047e-05,
|
| 577 |
+
"loss": 0.0257,
|
| 578 |
+
"step": 47500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.38,
|
| 582 |
+
"learning_rate": 4.936643428661967e-05,
|
| 583 |
+
"loss": 0.0257,
|
| 584 |
+
"step": 48000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.38,
|
| 588 |
+
"learning_rate": 4.935983258073886e-05,
|
| 589 |
+
"loss": 0.0254,
|
| 590 |
+
"step": 48500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.39,
|
| 594 |
+
"learning_rate": 4.935323087485806e-05,
|
| 595 |
+
"loss": 0.0255,
|
| 596 |
+
"step": 49000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.39,
|
| 600 |
+
"learning_rate": 4.9346629168977265e-05,
|
| 601 |
+
"loss": 0.0253,
|
| 602 |
+
"step": 49500
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.4,
|
| 606 |
+
"learning_rate": 4.9340027463096466e-05,
|
| 607 |
+
"loss": 0.0255,
|
| 608 |
+
"step": 50000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.4,
|
| 612 |
+
"learning_rate": 4.9333438960627426e-05,
|
| 613 |
+
"loss": 0.0258,
|
| 614 |
+
"step": 50500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.4,
|
| 618 |
+
"learning_rate": 4.932683725474663e-05,
|
| 619 |
+
"loss": 0.0254,
|
| 620 |
+
"step": 51000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.41,
|
| 624 |
+
"learning_rate": 4.932023554886583e-05,
|
| 625 |
+
"loss": 0.0255,
|
| 626 |
+
"step": 51500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.41,
|
| 630 |
+
"learning_rate": 4.931363384298503e-05,
|
| 631 |
+
"loss": 0.0255,
|
| 632 |
+
"step": 52000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.42,
|
| 636 |
+
"learning_rate": 4.930703213710423e-05,
|
| 637 |
+
"loss": 0.0254,
|
| 638 |
+
"step": 52500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 0.42,
|
| 642 |
+
"learning_rate": 4.930044363463519e-05,
|
| 643 |
+
"loss": 0.0253,
|
| 644 |
+
"step": 53000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.42,
|
| 648 |
+
"learning_rate": 4.929384192875439e-05,
|
| 649 |
+
"loss": 0.0252,
|
| 650 |
+
"step": 53500
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.43,
|
| 654 |
+
"learning_rate": 4.9287240222873593e-05,
|
| 655 |
+
"loss": 0.0251,
|
| 656 |
+
"step": 54000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.43,
|
| 660 |
+
"learning_rate": 4.9280638516992795e-05,
|
| 661 |
+
"loss": 0.0254,
|
| 662 |
+
"step": 54500
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.44,
|
| 666 |
+
"learning_rate": 4.9274036811111996e-05,
|
| 667 |
+
"loss": 0.0251,
|
| 668 |
+
"step": 55000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.44,
|
| 672 |
+
"learning_rate": 4.92674351052312e-05,
|
| 673 |
+
"loss": 0.0252,
|
| 674 |
+
"step": 55500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.44,
|
| 678 |
+
"learning_rate": 4.926083339935039e-05,
|
| 679 |
+
"loss": 0.0251,
|
| 680 |
+
"step": 56000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.45,
|
| 684 |
+
"learning_rate": 4.925424489688136e-05,
|
| 685 |
+
"loss": 0.0251,
|
| 686 |
+
"step": 56500
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 0.45,
|
| 690 |
+
"learning_rate": 4.924764319100056e-05,
|
| 691 |
+
"loss": 0.0251,
|
| 692 |
+
"step": 57000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.46,
|
| 696 |
+
"learning_rate": 4.924104148511976e-05,
|
| 697 |
+
"loss": 0.0252,
|
| 698 |
+
"step": 57500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.46,
|
| 702 |
+
"learning_rate": 4.9234439779238955e-05,
|
| 703 |
+
"loss": 0.0251,
|
| 704 |
+
"step": 58000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.46,
|
| 708 |
+
"learning_rate": 4.922785127676992e-05,
|
| 709 |
+
"loss": 0.0251,
|
| 710 |
+
"step": 58500
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.47,
|
| 714 |
+
"learning_rate": 4.9221249570889124e-05,
|
| 715 |
+
"loss": 0.025,
|
| 716 |
+
"step": 59000
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.47,
|
| 720 |
+
"learning_rate": 4.9214647865008325e-05,
|
| 721 |
+
"loss": 0.0249,
|
| 722 |
+
"step": 59500
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.48,
|
| 726 |
+
"learning_rate": 4.920804615912752e-05,
|
| 727 |
+
"loss": 0.0249,
|
| 728 |
+
"step": 60000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 0.48,
|
| 732 |
+
"learning_rate": 4.920144445324672e-05,
|
| 733 |
+
"loss": 0.025,
|
| 734 |
+
"step": 60500
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 0.48,
|
| 738 |
+
"learning_rate": 4.919484274736592e-05,
|
| 739 |
+
"loss": 0.0249,
|
| 740 |
+
"step": 61000
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.49,
|
| 744 |
+
"learning_rate": 4.918825424489689e-05,
|
| 745 |
+
"loss": 0.0251,
|
| 746 |
+
"step": 61500
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.49,
|
| 750 |
+
"learning_rate": 4.918165253901608e-05,
|
| 751 |
+
"loss": 0.0251,
|
| 752 |
+
"step": 62000
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.5,
|
| 756 |
+
"learning_rate": 4.9175050833135284e-05,
|
| 757 |
+
"loss": 0.0249,
|
| 758 |
+
"step": 62500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.5,
|
| 762 |
+
"learning_rate": 4.916846233066625e-05,
|
| 763 |
+
"loss": 0.0247,
|
| 764 |
+
"step": 63000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.5,
|
| 768 |
+
"learning_rate": 4.916186062478545e-05,
|
| 769 |
+
"loss": 0.025,
|
| 770 |
+
"step": 63500
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 0.51,
|
| 774 |
+
"learning_rate": 4.915525891890465e-05,
|
| 775 |
+
"loss": 0.0249,
|
| 776 |
+
"step": 64000
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.51,
|
| 780 |
+
"learning_rate": 4.914865721302385e-05,
|
| 781 |
+
"loss": 0.0249,
|
| 782 |
+
"step": 64500
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.51,
|
| 786 |
+
"learning_rate": 4.914205550714305e-05,
|
| 787 |
+
"loss": 0.0249,
|
| 788 |
+
"step": 65000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.52,
|
| 792 |
+
"learning_rate": 4.913545380126225e-05,
|
| 793 |
+
"loss": 0.0249,
|
| 794 |
+
"step": 65500
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.52,
|
| 798 |
+
"learning_rate": 4.912885209538145e-05,
|
| 799 |
+
"loss": 0.0248,
|
| 800 |
+
"step": 66000
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.53,
|
| 804 |
+
"learning_rate": 4.912226359291241e-05,
|
| 805 |
+
"loss": 0.0248,
|
| 806 |
+
"step": 66500
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.53,
|
| 810 |
+
"learning_rate": 4.911566188703161e-05,
|
| 811 |
+
"loss": 0.0247,
|
| 812 |
+
"step": 67000
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.53,
|
| 816 |
+
"learning_rate": 4.9109060181150814e-05,
|
| 817 |
+
"loss": 0.0247,
|
| 818 |
+
"step": 67500
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.54,
|
| 822 |
+
"learning_rate": 4.9102458475270015e-05,
|
| 823 |
+
"loss": 0.0245,
|
| 824 |
+
"step": 68000
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.54,
|
| 828 |
+
"learning_rate": 4.9095856769389216e-05,
|
| 829 |
+
"loss": 0.0248,
|
| 830 |
+
"step": 68500
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.55,
|
| 834 |
+
"learning_rate": 4.908926826692018e-05,
|
| 835 |
+
"loss": 0.0248,
|
| 836 |
+
"step": 69000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.55,
|
| 840 |
+
"learning_rate": 4.908267976445113e-05,
|
| 841 |
+
"loss": 0.0248,
|
| 842 |
+
"step": 69500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.55,
|
| 846 |
+
"learning_rate": 4.907607805857033e-05,
|
| 847 |
+
"loss": 0.0243,
|
| 848 |
+
"step": 70000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.56,
|
| 852 |
+
"learning_rate": 4.90694895561013e-05,
|
| 853 |
+
"loss": 0.0246,
|
| 854 |
+
"step": 70500
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.56,
|
| 858 |
+
"learning_rate": 4.906290105363226e-05,
|
| 859 |
+
"loss": 0.0245,
|
| 860 |
+
"step": 71000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.57,
|
| 864 |
+
"learning_rate": 4.905629934775146e-05,
|
| 865 |
+
"loss": 0.0245,
|
| 866 |
+
"step": 71500
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.57,
|
| 870 |
+
"learning_rate": 4.904969764187066e-05,
|
| 871 |
+
"loss": 0.0243,
|
| 872 |
+
"step": 72000
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.57,
|
| 876 |
+
"learning_rate": 4.904309593598986e-05,
|
| 877 |
+
"loss": 0.0244,
|
| 878 |
+
"step": 72500
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.58,
|
| 882 |
+
"learning_rate": 4.9036494230109064e-05,
|
| 883 |
+
"loss": 0.0246,
|
| 884 |
+
"step": 73000
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.58,
|
| 888 |
+
"learning_rate": 4.902989252422826e-05,
|
| 889 |
+
"loss": 0.0243,
|
| 890 |
+
"step": 73500
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.59,
|
| 894 |
+
"learning_rate": 4.902329081834746e-05,
|
| 895 |
+
"loss": 0.0244,
|
| 896 |
+
"step": 74000
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.59,
|
| 900 |
+
"learning_rate": 4.901668911246666e-05,
|
| 901 |
+
"loss": 0.0242,
|
| 902 |
+
"step": 74500
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.59,
|
| 906 |
+
"learning_rate": 4.901008740658586e-05,
|
| 907 |
+
"loss": 0.0242,
|
| 908 |
+
"step": 75000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.6,
|
| 912 |
+
"learning_rate": 4.900348570070506e-05,
|
| 913 |
+
"loss": 0.0242,
|
| 914 |
+
"step": 75500
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.6,
|
| 918 |
+
"learning_rate": 4.8996883994824264e-05,
|
| 919 |
+
"loss": 0.0244,
|
| 920 |
+
"step": 76000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.61,
|
| 924 |
+
"learning_rate": 4.8990282288943465e-05,
|
| 925 |
+
"loss": 0.0241,
|
| 926 |
+
"step": 76500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.61,
|
| 930 |
+
"learning_rate": 4.8983680583062666e-05,
|
| 931 |
+
"loss": 0.0243,
|
| 932 |
+
"step": 77000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.61,
|
| 936 |
+
"learning_rate": 4.897707887718187e-05,
|
| 937 |
+
"loss": 0.0242,
|
| 938 |
+
"step": 77500
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 0.62,
|
| 942 |
+
"learning_rate": 4.897047717130107e-05,
|
| 943 |
+
"loss": 0.0244,
|
| 944 |
+
"step": 78000
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 0.62,
|
| 948 |
+
"learning_rate": 4.896388866883203e-05,
|
| 949 |
+
"loss": 0.0243,
|
| 950 |
+
"step": 78500
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.63,
|
| 954 |
+
"learning_rate": 4.8957286962951224e-05,
|
| 955 |
+
"loss": 0.0242,
|
| 956 |
+
"step": 79000
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 0.63,
|
| 960 |
+
"learning_rate": 4.8950685257070425e-05,
|
| 961 |
+
"loss": 0.0243,
|
| 962 |
+
"step": 79500
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.63,
|
| 966 |
+
"learning_rate": 4.8944083551189626e-05,
|
| 967 |
+
"loss": 0.0241,
|
| 968 |
+
"step": 80000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.64,
|
| 972 |
+
"learning_rate": 4.893748184530883e-05,
|
| 973 |
+
"loss": 0.0244,
|
| 974 |
+
"step": 80500
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 0.64,
|
| 978 |
+
"learning_rate": 4.893088013942803e-05,
|
| 979 |
+
"loss": 0.024,
|
| 980 |
+
"step": 81000
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 0.65,
|
| 984 |
+
"learning_rate": 4.8924278433547236e-05,
|
| 985 |
+
"loss": 0.0242,
|
| 986 |
+
"step": 81500
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.65,
|
| 990 |
+
"learning_rate": 4.891767672766644e-05,
|
| 991 |
+
"loss": 0.0239,
|
| 992 |
+
"step": 82000
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.65,
|
| 996 |
+
"learning_rate": 4.891107502178563e-05,
|
| 997 |
+
"loss": 0.0243,
|
| 998 |
+
"step": 82500
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 0.66,
|
| 1002 |
+
"learning_rate": 4.890448651931659e-05,
|
| 1003 |
+
"loss": 0.024,
|
| 1004 |
+
"step": 83000
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.66,
|
| 1008 |
+
"learning_rate": 4.889788481343579e-05,
|
| 1009 |
+
"loss": 0.0239,
|
| 1010 |
+
"step": 83500
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.67,
|
| 1014 |
+
"learning_rate": 4.8891283107554994e-05,
|
| 1015 |
+
"loss": 0.024,
|
| 1016 |
+
"step": 84000
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 0.67,
|
| 1020 |
+
"learning_rate": 4.8884681401674195e-05,
|
| 1021 |
+
"loss": 0.024,
|
| 1022 |
+
"step": 84500
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 0.67,
|
| 1026 |
+
"learning_rate": 4.8878079695793396e-05,
|
| 1027 |
+
"loss": 0.0239,
|
| 1028 |
+
"step": 85000
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 0.68,
|
| 1032 |
+
"learning_rate": 4.88714779899126e-05,
|
| 1033 |
+
"loss": 0.0239,
|
| 1034 |
+
"step": 85500
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 0.68,
|
| 1038 |
+
"learning_rate": 4.88648762840318e-05,
|
| 1039 |
+
"loss": 0.0238,
|
| 1040 |
+
"step": 86000
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.69,
|
| 1044 |
+
"learning_rate": 4.8858274578151e-05,
|
| 1045 |
+
"loss": 0.0236,
|
| 1046 |
+
"step": 86500
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.69,
|
| 1050 |
+
"learning_rate": 4.885168607568196e-05,
|
| 1051 |
+
"loss": 0.024,
|
| 1052 |
+
"step": 87000
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.69,
|
| 1056 |
+
"learning_rate": 4.884508436980116e-05,
|
| 1057 |
+
"loss": 0.0239,
|
| 1058 |
+
"step": 87500
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 0.7,
|
| 1062 |
+
"learning_rate": 4.883848266392036e-05,
|
| 1063 |
+
"loss": 0.0239,
|
| 1064 |
+
"step": 88000
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 0.7,
|
| 1068 |
+
"learning_rate": 4.8831880958039564e-05,
|
| 1069 |
+
"loss": 0.0239,
|
| 1070 |
+
"step": 88500
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.71,
|
| 1074 |
+
"learning_rate": 4.8825279252158765e-05,
|
| 1075 |
+
"loss": 0.0237,
|
| 1076 |
+
"step": 89000
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.71,
|
| 1080 |
+
"learning_rate": 4.8818677546277966e-05,
|
| 1081 |
+
"loss": 0.0239,
|
| 1082 |
+
"step": 89500
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.71,
|
| 1086 |
+
"learning_rate": 4.881207584039716e-05,
|
| 1087 |
+
"loss": 0.0238,
|
| 1088 |
+
"step": 90000
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.72,
|
| 1092 |
+
"learning_rate": 4.880547413451636e-05,
|
| 1093 |
+
"loss": 0.0238,
|
| 1094 |
+
"step": 90500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.72,
|
| 1098 |
+
"learning_rate": 4.879887242863556e-05,
|
| 1099 |
+
"loss": 0.0238,
|
| 1100 |
+
"step": 91000
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 0.72,
|
| 1104 |
+
"learning_rate": 4.8792270722754764e-05,
|
| 1105 |
+
"loss": 0.0237,
|
| 1106 |
+
"step": 91500
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 0.73,
|
| 1110 |
+
"learning_rate": 4.8785682220285724e-05,
|
| 1111 |
+
"loss": 0.0236,
|
| 1112 |
+
"step": 92000
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.73,
|
| 1116 |
+
"learning_rate": 4.8779080514404925e-05,
|
| 1117 |
+
"loss": 0.0235,
|
| 1118 |
+
"step": 92500
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.74,
|
| 1122 |
+
"learning_rate": 4.8772478808524127e-05,
|
| 1123 |
+
"loss": 0.0235,
|
| 1124 |
+
"step": 93000
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.74,
|
| 1128 |
+
"learning_rate": 4.876587710264333e-05,
|
| 1129 |
+
"loss": 0.0237,
|
| 1130 |
+
"step": 93500
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.74,
|
| 1134 |
+
"learning_rate": 4.875928860017429e-05,
|
| 1135 |
+
"loss": 0.0237,
|
| 1136 |
+
"step": 94000
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.75,
|
| 1140 |
+
"learning_rate": 4.875268689429349e-05,
|
| 1141 |
+
"loss": 0.0238,
|
| 1142 |
+
"step": 94500
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 0.75,
|
| 1146 |
+
"learning_rate": 4.874608518841269e-05,
|
| 1147 |
+
"loss": 0.0237,
|
| 1148 |
+
"step": 95000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 0.76,
|
| 1152 |
+
"learning_rate": 4.873948348253189e-05,
|
| 1153 |
+
"loss": 0.0237,
|
| 1154 |
+
"step": 95500
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.76,
|
| 1158 |
+
"learning_rate": 4.873288177665109e-05,
|
| 1159 |
+
"loss": 0.0234,
|
| 1160 |
+
"step": 96000
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.76,
|
| 1164 |
+
"learning_rate": 4.872629327418205e-05,
|
| 1165 |
+
"loss": 0.0236,
|
| 1166 |
+
"step": 96500
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 0.77,
|
| 1170 |
+
"learning_rate": 4.8719691568301254e-05,
|
| 1171 |
+
"loss": 0.0235,
|
| 1172 |
+
"step": 97000
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.77,
|
| 1176 |
+
"learning_rate": 4.8713089862420456e-05,
|
| 1177 |
+
"loss": 0.0237,
|
| 1178 |
+
"step": 97500
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.78,
|
| 1182 |
+
"learning_rate": 4.870648815653966e-05,
|
| 1183 |
+
"loss": 0.0235,
|
| 1184 |
+
"step": 98000
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 0.78,
|
| 1188 |
+
"learning_rate": 4.869988645065886e-05,
|
| 1189 |
+
"loss": 0.0233,
|
| 1190 |
+
"step": 98500
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 0.78,
|
| 1194 |
+
"learning_rate": 4.869328474477805e-05,
|
| 1195 |
+
"loss": 0.0235,
|
| 1196 |
+
"step": 99000
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.79,
|
| 1200 |
+
"learning_rate": 4.868668303889725e-05,
|
| 1201 |
+
"loss": 0.0238,
|
| 1202 |
+
"step": 99500
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.79,
|
| 1206 |
+
"learning_rate": 4.8680081333016454e-05,
|
| 1207 |
+
"loss": 0.0232,
|
| 1208 |
+
"step": 100000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 0.8,
|
| 1212 |
+
"learning_rate": 4.8673479627135656e-05,
|
| 1213 |
+
"loss": 0.0234,
|
| 1214 |
+
"step": 100500
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.8,
|
| 1218 |
+
"learning_rate": 4.8666891124666616e-05,
|
| 1219 |
+
"loss": 0.0235,
|
| 1220 |
+
"step": 101000
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.8,
|
| 1224 |
+
"learning_rate": 4.866028941878582e-05,
|
| 1225 |
+
"loss": 0.0237,
|
| 1226 |
+
"step": 101500
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 0.81,
|
| 1230 |
+
"learning_rate": 4.865368771290502e-05,
|
| 1231 |
+
"loss": 0.0234,
|
| 1232 |
+
"step": 102000
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 0.81,
|
| 1236 |
+
"learning_rate": 4.864708600702422e-05,
|
| 1237 |
+
"loss": 0.0234,
|
| 1238 |
+
"step": 102500
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.82,
|
| 1242 |
+
"learning_rate": 4.864048430114342e-05,
|
| 1243 |
+
"loss": 0.0237,
|
| 1244 |
+
"step": 103000
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 0.82,
|
| 1248 |
+
"learning_rate": 4.863388259526262e-05,
|
| 1249 |
+
"loss": 0.0233,
|
| 1250 |
+
"step": 103500
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 0.82,
|
| 1254 |
+
"learning_rate": 4.862728088938182e-05,
|
| 1255 |
+
"loss": 0.0236,
|
| 1256 |
+
"step": 104000
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.83,
|
| 1260 |
+
"learning_rate": 4.862067918350102e-05,
|
| 1261 |
+
"loss": 0.0233,
|
| 1262 |
+
"step": 104500
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.83,
|
| 1266 |
+
"learning_rate": 4.861407747762022e-05,
|
| 1267 |
+
"loss": 0.0236,
|
| 1268 |
+
"step": 105000
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 0.84,
|
| 1272 |
+
"learning_rate": 4.8607488975151186e-05,
|
| 1273 |
+
"loss": 0.0234,
|
| 1274 |
+
"step": 105500
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 0.84,
|
| 1278 |
+
"learning_rate": 4.860088726927039e-05,
|
| 1279 |
+
"loss": 0.0232,
|
| 1280 |
+
"step": 106000
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.84,
|
| 1284 |
+
"learning_rate": 4.859429876680134e-05,
|
| 1285 |
+
"loss": 0.0232,
|
| 1286 |
+
"step": 106500
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 0.85,
|
| 1290 |
+
"learning_rate": 4.858769706092055e-05,
|
| 1291 |
+
"loss": 0.023,
|
| 1292 |
+
"step": 107000
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 0.85,
|
| 1296 |
+
"learning_rate": 4.858109535503975e-05,
|
| 1297 |
+
"loss": 0.023,
|
| 1298 |
+
"step": 107500
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.86,
|
| 1302 |
+
"learning_rate": 4.857449364915895e-05,
|
| 1303 |
+
"loss": 0.0234,
|
| 1304 |
+
"step": 108000
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.86,
|
| 1308 |
+
"learning_rate": 4.8567891943278145e-05,
|
| 1309 |
+
"loss": 0.0232,
|
| 1310 |
+
"step": 108500
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 0.86,
|
| 1314 |
+
"learning_rate": 4.8561303440809106e-05,
|
| 1315 |
+
"loss": 0.0231,
|
| 1316 |
+
"step": 109000
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 0.87,
|
| 1320 |
+
"learning_rate": 4.855470173492831e-05,
|
| 1321 |
+
"loss": 0.0233,
|
| 1322 |
+
"step": 109500
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.87,
|
| 1326 |
+
"learning_rate": 4.854810002904751e-05,
|
| 1327 |
+
"loss": 0.0233,
|
| 1328 |
+
"step": 110000
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 0.88,
|
| 1332 |
+
"learning_rate": 4.854149832316671e-05,
|
| 1333 |
+
"loss": 0.0231,
|
| 1334 |
+
"step": 110500
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 0.88,
|
| 1338 |
+
"learning_rate": 4.853489661728591e-05,
|
| 1339 |
+
"loss": 0.0232,
|
| 1340 |
+
"step": 111000
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 0.88,
|
| 1344 |
+
"learning_rate": 4.852829491140511e-05,
|
| 1345 |
+
"loss": 0.0232,
|
| 1346 |
+
"step": 111500
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.89,
|
| 1350 |
+
"learning_rate": 4.852169320552431e-05,
|
| 1351 |
+
"loss": 0.0231,
|
| 1352 |
+
"step": 112000
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 0.89,
|
| 1356 |
+
"learning_rate": 4.851510470305527e-05,
|
| 1357 |
+
"loss": 0.023,
|
| 1358 |
+
"step": 112500
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 0.9,
|
| 1362 |
+
"learning_rate": 4.8508502997174474e-05,
|
| 1363 |
+
"loss": 0.023,
|
| 1364 |
+
"step": 113000
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.9,
|
| 1368 |
+
"learning_rate": 4.8501901291293675e-05,
|
| 1369 |
+
"loss": 0.0231,
|
| 1370 |
+
"step": 113500
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 0.9,
|
| 1374 |
+
"learning_rate": 4.8495299585412876e-05,
|
| 1375 |
+
"loss": 0.023,
|
| 1376 |
+
"step": 114000
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 0.91,
|
| 1380 |
+
"learning_rate": 4.848869787953208e-05,
|
| 1381 |
+
"loss": 0.0231,
|
| 1382 |
+
"step": 114500
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 0.91,
|
| 1386 |
+
"learning_rate": 4.848209617365128e-05,
|
| 1387 |
+
"loss": 0.0227,
|
| 1388 |
+
"step": 115000
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.91,
|
| 1392 |
+
"learning_rate": 4.847549446777048e-05,
|
| 1393 |
+
"loss": 0.0228,
|
| 1394 |
+
"step": 115500
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 0.92,
|
| 1398 |
+
"learning_rate": 4.8468892761889674e-05,
|
| 1399 |
+
"loss": 0.023,
|
| 1400 |
+
"step": 116000
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 0.92,
|
| 1404 |
+
"learning_rate": 4.8462304259420634e-05,
|
| 1405 |
+
"loss": 0.0228,
|
| 1406 |
+
"step": 116500
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.93,
|
| 1410 |
+
"learning_rate": 4.8455702553539836e-05,
|
| 1411 |
+
"loss": 0.023,
|
| 1412 |
+
"step": 117000
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 0.93,
|
| 1416 |
+
"learning_rate": 4.844910084765904e-05,
|
| 1417 |
+
"loss": 0.023,
|
| 1418 |
+
"step": 117500
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 0.93,
|
| 1422 |
+
"learning_rate": 4.844249914177824e-05,
|
| 1423 |
+
"loss": 0.0229,
|
| 1424 |
+
"step": 118000
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 0.94,
|
| 1428 |
+
"learning_rate": 4.84359106393092e-05,
|
| 1429 |
+
"loss": 0.023,
|
| 1430 |
+
"step": 118500
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.94,
|
| 1434 |
+
"learning_rate": 4.84293089334284e-05,
|
| 1435 |
+
"loss": 0.023,
|
| 1436 |
+
"step": 119000
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 0.95,
|
| 1440 |
+
"learning_rate": 4.84227072275476e-05,
|
| 1441 |
+
"loss": 0.0229,
|
| 1442 |
+
"step": 119500
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 0.95,
|
| 1446 |
+
"learning_rate": 4.84161055216668e-05,
|
| 1447 |
+
"loss": 0.0229,
|
| 1448 |
+
"step": 120000
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 0.95,
|
| 1452 |
+
"learning_rate": 4.8409503815786e-05,
|
| 1453 |
+
"loss": 0.023,
|
| 1454 |
+
"step": 120500
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 0.96,
|
| 1458 |
+
"learning_rate": 4.8402902109905204e-05,
|
| 1459 |
+
"loss": 0.0229,
|
| 1460 |
+
"step": 121000
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 0.96,
|
| 1464 |
+
"learning_rate": 4.8396300404024405e-05,
|
| 1465 |
+
"loss": 0.0228,
|
| 1466 |
+
"step": 121500
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 0.97,
|
| 1470 |
+
"learning_rate": 4.8389698698143606e-05,
|
| 1471 |
+
"loss": 0.0229,
|
| 1472 |
+
"step": 122000
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.97,
|
| 1476 |
+
"learning_rate": 4.838311019567457e-05,
|
| 1477 |
+
"loss": 0.023,
|
| 1478 |
+
"step": 122500
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 0.97,
|
| 1482 |
+
"learning_rate": 4.837650848979377e-05,
|
| 1483 |
+
"loss": 0.023,
|
| 1484 |
+
"step": 123000
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 0.98,
|
| 1488 |
+
"learning_rate": 4.836990678391297e-05,
|
| 1489 |
+
"loss": 0.0229,
|
| 1490 |
+
"step": 123500
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 0.98,
|
| 1494 |
+
"learning_rate": 4.836330507803217e-05,
|
| 1495 |
+
"loss": 0.0227,
|
| 1496 |
+
"step": 124000
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 0.99,
|
| 1500 |
+
"learning_rate": 4.8356716575563124e-05,
|
| 1501 |
+
"loss": 0.0231,
|
| 1502 |
+
"step": 124500
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 0.99,
|
| 1506 |
+
"learning_rate": 4.8350114869682325e-05,
|
| 1507 |
+
"loss": 0.0228,
|
| 1508 |
+
"step": 125000
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 0.99,
|
| 1512 |
+
"learning_rate": 4.8343513163801526e-05,
|
| 1513 |
+
"loss": 0.0228,
|
| 1514 |
+
"step": 125500
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.0,
|
| 1518 |
+
"learning_rate": 4.833691145792073e-05,
|
| 1519 |
+
"loss": 0.0229,
|
| 1520 |
+
"step": 126000
|
| 1521 |
+
}
|
| 1522 |
+
],
|
| 1523 |
+
"max_steps": 3786900,
|
| 1524 |
+
"num_train_epochs": 30,
|
| 1525 |
+
"total_flos": 4.741138645915031e+18,
|
| 1526 |
+
"trial_name": null,
|
| 1527 |
+
"trial_params": null
|
| 1528 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:96e075df303840a8497aa0801a4a5fb4ff95e195a7d9443f10c4ec77fc049b47
|
| 3 |
+
size 3387
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|