First commit of the disfluency classifier
Browse files- config.json +27 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +7 -0
- tokenizer.json +0 -0
- tokenizer_config.json +14 -0
- trainer_state.json +2092 -0
- training_args.bin +3 -0
- vocab.txt +0 -0
config.json
ADDED
|
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "bert-base-cased",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"gradient_checkpointing": false,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 3072,
|
| 14 |
+
"layer_norm_eps": 1e-12,
|
| 15 |
+
"max_position_embeddings": 512,
|
| 16 |
+
"model_type": "bert",
|
| 17 |
+
"num_attention_heads": 12,
|
| 18 |
+
"num_hidden_layers": 12,
|
| 19 |
+
"pad_token_id": 0,
|
| 20 |
+
"position_embedding_type": "absolute",
|
| 21 |
+
"problem_type": "single_label_classification",
|
| 22 |
+
"torch_dtype": "float32",
|
| 23 |
+
"transformers_version": "4.24.0.dev0",
|
| 24 |
+
"type_vocab_size": 2,
|
| 25 |
+
"use_cache": true,
|
| 26 |
+
"vocab_size": 28996
|
| 27 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:42e3d540365896fc45793fa96e40d4c0523582f9a074764d931153b050575bab
|
| 3 |
+
size 866617781
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e8a42457ff4c4ec18021fa483036e31b694bdf9e0293034becb02be6277907c7
|
| 3 |
+
size 433321674
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6141f35a2d0809dcaef833ea5ba6f3a33ecf1499103a608e2b0ed57ce0a35795
|
| 3 |
+
size 14649
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3ddb5d0db0067f7d5c36deee166966fa255ba5dbe3ac42959cd554fad412d075
|
| 3 |
+
size 623
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": "[CLS]",
|
| 3 |
+
"mask_token": "[MASK]",
|
| 4 |
+
"pad_token": "[PAD]",
|
| 5 |
+
"sep_token": "[SEP]",
|
| 6 |
+
"unk_token": "[UNK]"
|
| 7 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": "[CLS]",
|
| 3 |
+
"do_lower_case": false,
|
| 4 |
+
"mask_token": "[MASK]",
|
| 5 |
+
"model_max_length": 512,
|
| 6 |
+
"name_or_path": "bert-base-cased",
|
| 7 |
+
"pad_token": "[PAD]",
|
| 8 |
+
"sep_token": "[SEP]",
|
| 9 |
+
"special_tokens_map_file": null,
|
| 10 |
+
"strip_accents": null,
|
| 11 |
+
"tokenize_chinese_chars": true,
|
| 12 |
+
"tokenizer_class": "BertTokenizer",
|
| 13 |
+
"unk_token": "[UNK]"
|
| 14 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2092 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.901960784313726,
|
| 5 |
+
"global_step": 170000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 13 |
+
"loss": 0.6866,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.03,
|
| 18 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 19 |
+
"loss": 0.513,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.04,
|
| 24 |
+
"learning_rate": 3e-06,
|
| 25 |
+
"loss": 0.3957,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.06,
|
| 30 |
+
"learning_rate": 4.000000000000001e-06,
|
| 31 |
+
"loss": 0.4037,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.07,
|
| 36 |
+
"learning_rate": 5e-06,
|
| 37 |
+
"loss": 0.377,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.09,
|
| 42 |
+
"learning_rate": 6e-06,
|
| 43 |
+
"loss": 0.3838,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.1,
|
| 48 |
+
"learning_rate": 7e-06,
|
| 49 |
+
"loss": 0.3804,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.12,
|
| 54 |
+
"learning_rate": 8.000000000000001e-06,
|
| 55 |
+
"loss": 0.3842,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.13,
|
| 60 |
+
"learning_rate": 9e-06,
|
| 61 |
+
"loss": 0.368,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.14,
|
| 66 |
+
"learning_rate": 1e-05,
|
| 67 |
+
"loss": 0.3753,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.16,
|
| 72 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 73 |
+
"loss": 0.3637,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.17,
|
| 78 |
+
"learning_rate": 1.2e-05,
|
| 79 |
+
"loss": 0.3884,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.19,
|
| 84 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 85 |
+
"loss": 0.3814,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.2,
|
| 90 |
+
"learning_rate": 1.4e-05,
|
| 91 |
+
"loss": 0.3741,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.22,
|
| 96 |
+
"learning_rate": 1.5000000000000002e-05,
|
| 97 |
+
"loss": 0.3502,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.23,
|
| 102 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 103 |
+
"loss": 0.3667,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.25,
|
| 108 |
+
"learning_rate": 1.7e-05,
|
| 109 |
+
"loss": 0.37,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.26,
|
| 114 |
+
"learning_rate": 1.8e-05,
|
| 115 |
+
"loss": 0.3754,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.27,
|
| 120 |
+
"learning_rate": 1.9e-05,
|
| 121 |
+
"loss": 0.3505,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.29,
|
| 126 |
+
"learning_rate": 2e-05,
|
| 127 |
+
"loss": 0.3663,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.3,
|
| 132 |
+
"learning_rate": 1.9938800489596085e-05,
|
| 133 |
+
"loss": 0.3551,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.32,
|
| 138 |
+
"learning_rate": 1.987760097919217e-05,
|
| 139 |
+
"loss": 0.3738,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.33,
|
| 144 |
+
"learning_rate": 1.9816401468788252e-05,
|
| 145 |
+
"loss": 0.3595,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.35,
|
| 150 |
+
"learning_rate": 1.9755201958384335e-05,
|
| 151 |
+
"loss": 0.3704,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.36,
|
| 156 |
+
"learning_rate": 1.969400244798042e-05,
|
| 157 |
+
"loss": 0.3719,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.37,
|
| 162 |
+
"learning_rate": 1.9632802937576502e-05,
|
| 163 |
+
"loss": 0.3682,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.39,
|
| 168 |
+
"learning_rate": 1.9571603427172585e-05,
|
| 169 |
+
"loss": 0.3772,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.4,
|
| 174 |
+
"learning_rate": 1.951040391676867e-05,
|
| 175 |
+
"loss": 0.3707,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.42,
|
| 180 |
+
"learning_rate": 1.944920440636475e-05,
|
| 181 |
+
"loss": 0.3637,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.43,
|
| 186 |
+
"learning_rate": 1.9388004895960835e-05,
|
| 187 |
+
"loss": 0.3543,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.45,
|
| 192 |
+
"learning_rate": 1.932680538555692e-05,
|
| 193 |
+
"loss": 0.3656,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.46,
|
| 198 |
+
"learning_rate": 1.9265605875153e-05,
|
| 199 |
+
"loss": 0.3551,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.48,
|
| 204 |
+
"learning_rate": 1.9204406364749086e-05,
|
| 205 |
+
"loss": 0.3686,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.49,
|
| 210 |
+
"learning_rate": 1.9143206854345166e-05,
|
| 211 |
+
"loss": 0.3533,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5,
|
| 216 |
+
"learning_rate": 1.908200734394125e-05,
|
| 217 |
+
"loss": 0.3564,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.52,
|
| 222 |
+
"learning_rate": 1.9020807833537332e-05,
|
| 223 |
+
"loss": 0.3597,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.53,
|
| 228 |
+
"learning_rate": 1.8959608323133416e-05,
|
| 229 |
+
"loss": 0.3608,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.55,
|
| 234 |
+
"learning_rate": 1.88984088127295e-05,
|
| 235 |
+
"loss": 0.3603,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.56,
|
| 240 |
+
"learning_rate": 1.8837209302325582e-05,
|
| 241 |
+
"loss": 0.3622,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.58,
|
| 246 |
+
"learning_rate": 1.8776009791921666e-05,
|
| 247 |
+
"loss": 0.3642,
|
| 248 |
+
"step": 20000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.59,
|
| 252 |
+
"learning_rate": 1.871481028151775e-05,
|
| 253 |
+
"loss": 0.3719,
|
| 254 |
+
"step": 20500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.61,
|
| 258 |
+
"learning_rate": 1.8653610771113833e-05,
|
| 259 |
+
"loss": 0.3595,
|
| 260 |
+
"step": 21000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.62,
|
| 264 |
+
"learning_rate": 1.8592411260709916e-05,
|
| 265 |
+
"loss": 0.3693,
|
| 266 |
+
"step": 21500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.63,
|
| 270 |
+
"learning_rate": 1.8531211750306e-05,
|
| 271 |
+
"loss": 0.3625,
|
| 272 |
+
"step": 22000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.65,
|
| 276 |
+
"learning_rate": 1.8470012239902083e-05,
|
| 277 |
+
"loss": 0.3513,
|
| 278 |
+
"step": 22500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.66,
|
| 282 |
+
"learning_rate": 1.8408812729498166e-05,
|
| 283 |
+
"loss": 0.3613,
|
| 284 |
+
"step": 23000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.68,
|
| 288 |
+
"learning_rate": 1.834761321909425e-05,
|
| 289 |
+
"loss": 0.3583,
|
| 290 |
+
"step": 23500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.69,
|
| 294 |
+
"learning_rate": 1.8286413708690333e-05,
|
| 295 |
+
"loss": 0.3645,
|
| 296 |
+
"step": 24000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.71,
|
| 300 |
+
"learning_rate": 1.8225214198286416e-05,
|
| 301 |
+
"loss": 0.3575,
|
| 302 |
+
"step": 24500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.72,
|
| 306 |
+
"learning_rate": 1.81640146878825e-05,
|
| 307 |
+
"loss": 0.3646,
|
| 308 |
+
"step": 25000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.74,
|
| 312 |
+
"learning_rate": 1.810281517747858e-05,
|
| 313 |
+
"loss": 0.3497,
|
| 314 |
+
"step": 25500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.75,
|
| 318 |
+
"learning_rate": 1.8041615667074666e-05,
|
| 319 |
+
"loss": 0.357,
|
| 320 |
+
"step": 26000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.76,
|
| 324 |
+
"learning_rate": 1.798041615667075e-05,
|
| 325 |
+
"loss": 0.3658,
|
| 326 |
+
"step": 26500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.78,
|
| 330 |
+
"learning_rate": 1.791921664626683e-05,
|
| 331 |
+
"loss": 0.3442,
|
| 332 |
+
"step": 27000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.79,
|
| 336 |
+
"learning_rate": 1.7858017135862916e-05,
|
| 337 |
+
"loss": 0.3582,
|
| 338 |
+
"step": 27500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.81,
|
| 342 |
+
"learning_rate": 1.7796817625458996e-05,
|
| 343 |
+
"loss": 0.3679,
|
| 344 |
+
"step": 28000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.82,
|
| 348 |
+
"learning_rate": 1.773561811505508e-05,
|
| 349 |
+
"loss": 0.3575,
|
| 350 |
+
"step": 28500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.84,
|
| 354 |
+
"learning_rate": 1.7674418604651163e-05,
|
| 355 |
+
"loss": 0.3492,
|
| 356 |
+
"step": 29000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.85,
|
| 360 |
+
"learning_rate": 1.7613219094247247e-05,
|
| 361 |
+
"loss": 0.3463,
|
| 362 |
+
"step": 29500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.87,
|
| 366 |
+
"learning_rate": 1.755201958384333e-05,
|
| 367 |
+
"loss": 0.3457,
|
| 368 |
+
"step": 30000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.88,
|
| 372 |
+
"learning_rate": 1.7490820073439413e-05,
|
| 373 |
+
"loss": 0.3631,
|
| 374 |
+
"step": 30500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.89,
|
| 378 |
+
"learning_rate": 1.7429620563035497e-05,
|
| 379 |
+
"loss": 0.3587,
|
| 380 |
+
"step": 31000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.91,
|
| 384 |
+
"learning_rate": 1.736842105263158e-05,
|
| 385 |
+
"loss": 0.3577,
|
| 386 |
+
"step": 31500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.92,
|
| 390 |
+
"learning_rate": 1.7307221542227663e-05,
|
| 391 |
+
"loss": 0.3476,
|
| 392 |
+
"step": 32000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.94,
|
| 396 |
+
"learning_rate": 1.7246022031823747e-05,
|
| 397 |
+
"loss": 0.357,
|
| 398 |
+
"step": 32500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.95,
|
| 402 |
+
"learning_rate": 1.718482252141983e-05,
|
| 403 |
+
"loss": 0.3657,
|
| 404 |
+
"step": 33000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.97,
|
| 408 |
+
"learning_rate": 1.7123623011015914e-05,
|
| 409 |
+
"loss": 0.3622,
|
| 410 |
+
"step": 33500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.98,
|
| 414 |
+
"learning_rate": 1.7062423500611997e-05,
|
| 415 |
+
"loss": 0.3585,
|
| 416 |
+
"step": 34000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.99,
|
| 420 |
+
"learning_rate": 1.700122399020808e-05,
|
| 421 |
+
"loss": 0.3441,
|
| 422 |
+
"step": 34500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.0,
|
| 426 |
+
"eval_accuracy": 0.8873362445414847,
|
| 427 |
+
"eval_loss": 0.3056604862213135,
|
| 428 |
+
"eval_runtime": 105.089,
|
| 429 |
+
"eval_samples_per_second": 65.373,
|
| 430 |
+
"eval_steps_per_second": 8.174,
|
| 431 |
+
"step": 34680
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.01,
|
| 435 |
+
"learning_rate": 1.6940024479804164e-05,
|
| 436 |
+
"loss": 0.3411,
|
| 437 |
+
"step": 35000
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.02,
|
| 441 |
+
"learning_rate": 1.6878824969400247e-05,
|
| 442 |
+
"loss": 0.3405,
|
| 443 |
+
"step": 35500
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.04,
|
| 447 |
+
"learning_rate": 1.681762545899633e-05,
|
| 448 |
+
"loss": 0.3493,
|
| 449 |
+
"step": 36000
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 1.05,
|
| 453 |
+
"learning_rate": 1.675642594859241e-05,
|
| 454 |
+
"loss": 0.3418,
|
| 455 |
+
"step": 36500
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 1.07,
|
| 459 |
+
"learning_rate": 1.6695226438188497e-05,
|
| 460 |
+
"loss": 0.3631,
|
| 461 |
+
"step": 37000
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 1.08,
|
| 465 |
+
"learning_rate": 1.663402692778458e-05,
|
| 466 |
+
"loss": 0.3478,
|
| 467 |
+
"step": 37500
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 1.1,
|
| 471 |
+
"learning_rate": 1.657282741738066e-05,
|
| 472 |
+
"loss": 0.3571,
|
| 473 |
+
"step": 38000
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.11,
|
| 477 |
+
"learning_rate": 1.6511627906976747e-05,
|
| 478 |
+
"loss": 0.3454,
|
| 479 |
+
"step": 38500
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.12,
|
| 483 |
+
"learning_rate": 1.6450428396572827e-05,
|
| 484 |
+
"loss": 0.3325,
|
| 485 |
+
"step": 39000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.14,
|
| 489 |
+
"learning_rate": 1.6389228886168914e-05,
|
| 490 |
+
"loss": 0.3618,
|
| 491 |
+
"step": 39500
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 1.15,
|
| 495 |
+
"learning_rate": 1.6328029375764994e-05,
|
| 496 |
+
"loss": 0.3607,
|
| 497 |
+
"step": 40000
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 1.17,
|
| 501 |
+
"learning_rate": 1.6266829865361078e-05,
|
| 502 |
+
"loss": 0.3562,
|
| 503 |
+
"step": 40500
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 1.18,
|
| 507 |
+
"learning_rate": 1.620563035495716e-05,
|
| 508 |
+
"loss": 0.3476,
|
| 509 |
+
"step": 41000
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 1.2,
|
| 513 |
+
"learning_rate": 1.6144430844553244e-05,
|
| 514 |
+
"loss": 0.3554,
|
| 515 |
+
"step": 41500
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.21,
|
| 519 |
+
"learning_rate": 1.6083231334149328e-05,
|
| 520 |
+
"loss": 0.3462,
|
| 521 |
+
"step": 42000
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.23,
|
| 525 |
+
"learning_rate": 1.602203182374541e-05,
|
| 526 |
+
"loss": 0.3716,
|
| 527 |
+
"step": 42500
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.24,
|
| 531 |
+
"learning_rate": 1.5960832313341494e-05,
|
| 532 |
+
"loss": 0.3578,
|
| 533 |
+
"step": 43000
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 1.25,
|
| 537 |
+
"learning_rate": 1.5899632802937578e-05,
|
| 538 |
+
"loss": 0.3642,
|
| 539 |
+
"step": 43500
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 1.27,
|
| 543 |
+
"learning_rate": 1.583843329253366e-05,
|
| 544 |
+
"loss": 0.3545,
|
| 545 |
+
"step": 44000
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 1.28,
|
| 549 |
+
"learning_rate": 1.5777233782129745e-05,
|
| 550 |
+
"loss": 0.346,
|
| 551 |
+
"step": 44500
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 1.3,
|
| 555 |
+
"learning_rate": 1.5716034271725828e-05,
|
| 556 |
+
"loss": 0.3498,
|
| 557 |
+
"step": 45000
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.31,
|
| 561 |
+
"learning_rate": 1.565483476132191e-05,
|
| 562 |
+
"loss": 0.3523,
|
| 563 |
+
"step": 45500
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.33,
|
| 567 |
+
"learning_rate": 1.5593635250917995e-05,
|
| 568 |
+
"loss": 0.3527,
|
| 569 |
+
"step": 46000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.34,
|
| 573 |
+
"learning_rate": 1.5532435740514078e-05,
|
| 574 |
+
"loss": 0.3466,
|
| 575 |
+
"step": 46500
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 1.36,
|
| 579 |
+
"learning_rate": 1.547123623011016e-05,
|
| 580 |
+
"loss": 0.3584,
|
| 581 |
+
"step": 47000
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 1.37,
|
| 585 |
+
"learning_rate": 1.541003671970624e-05,
|
| 586 |
+
"loss": 0.3737,
|
| 587 |
+
"step": 47500
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 1.38,
|
| 591 |
+
"learning_rate": 1.5348837209302328e-05,
|
| 592 |
+
"loss": 0.3542,
|
| 593 |
+
"step": 48000
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 1.4,
|
| 597 |
+
"learning_rate": 1.528763769889841e-05,
|
| 598 |
+
"loss": 0.362,
|
| 599 |
+
"step": 48500
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.41,
|
| 603 |
+
"learning_rate": 1.5226438188494492e-05,
|
| 604 |
+
"loss": 0.3588,
|
| 605 |
+
"step": 49000
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.43,
|
| 609 |
+
"learning_rate": 1.5165238678090577e-05,
|
| 610 |
+
"loss": 0.3471,
|
| 611 |
+
"step": 49500
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.44,
|
| 615 |
+
"learning_rate": 1.510403916768666e-05,
|
| 616 |
+
"loss": 0.3514,
|
| 617 |
+
"step": 50000
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 1.46,
|
| 621 |
+
"learning_rate": 1.5042839657282743e-05,
|
| 622 |
+
"loss": 0.3488,
|
| 623 |
+
"step": 50500
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 1.47,
|
| 627 |
+
"learning_rate": 1.4981640146878827e-05,
|
| 628 |
+
"loss": 0.352,
|
| 629 |
+
"step": 51000
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 1.49,
|
| 633 |
+
"learning_rate": 1.4920440636474908e-05,
|
| 634 |
+
"loss": 0.3617,
|
| 635 |
+
"step": 51500
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 1.5,
|
| 639 |
+
"learning_rate": 1.4859241126070994e-05,
|
| 640 |
+
"loss": 0.3546,
|
| 641 |
+
"step": 52000
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.51,
|
| 645 |
+
"learning_rate": 1.4798041615667075e-05,
|
| 646 |
+
"loss": 0.3501,
|
| 647 |
+
"step": 52500
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.53,
|
| 651 |
+
"learning_rate": 1.4736842105263159e-05,
|
| 652 |
+
"loss": 0.3508,
|
| 653 |
+
"step": 53000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.54,
|
| 657 |
+
"learning_rate": 1.4675642594859244e-05,
|
| 658 |
+
"loss": 0.358,
|
| 659 |
+
"step": 53500
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 1.56,
|
| 663 |
+
"learning_rate": 1.4614443084455325e-05,
|
| 664 |
+
"loss": 0.3474,
|
| 665 |
+
"step": 54000
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 1.57,
|
| 669 |
+
"learning_rate": 1.4553243574051407e-05,
|
| 670 |
+
"loss": 0.3487,
|
| 671 |
+
"step": 54500
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 1.59,
|
| 675 |
+
"learning_rate": 1.4492044063647492e-05,
|
| 676 |
+
"loss": 0.3532,
|
| 677 |
+
"step": 55000
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 1.6,
|
| 681 |
+
"learning_rate": 1.4430844553243575e-05,
|
| 682 |
+
"loss": 0.3429,
|
| 683 |
+
"step": 55500
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.61,
|
| 687 |
+
"learning_rate": 1.4369645042839659e-05,
|
| 688 |
+
"loss": 0.3509,
|
| 689 |
+
"step": 56000
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.63,
|
| 693 |
+
"learning_rate": 1.4308445532435742e-05,
|
| 694 |
+
"loss": 0.3478,
|
| 695 |
+
"step": 56500
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.64,
|
| 699 |
+
"learning_rate": 1.4247246022031824e-05,
|
| 700 |
+
"loss": 0.3505,
|
| 701 |
+
"step": 57000
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 1.66,
|
| 705 |
+
"learning_rate": 1.4186046511627909e-05,
|
| 706 |
+
"loss": 0.35,
|
| 707 |
+
"step": 57500
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 1.67,
|
| 711 |
+
"learning_rate": 1.412484700122399e-05,
|
| 712 |
+
"loss": 0.3605,
|
| 713 |
+
"step": 58000
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 1.69,
|
| 717 |
+
"learning_rate": 1.4063647490820074e-05,
|
| 718 |
+
"loss": 0.3541,
|
| 719 |
+
"step": 58500
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 1.7,
|
| 723 |
+
"learning_rate": 1.4002447980416159e-05,
|
| 724 |
+
"loss": 0.3576,
|
| 725 |
+
"step": 59000
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.72,
|
| 729 |
+
"learning_rate": 1.394124847001224e-05,
|
| 730 |
+
"loss": 0.3477,
|
| 731 |
+
"step": 59500
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.73,
|
| 735 |
+
"learning_rate": 1.3880048959608322e-05,
|
| 736 |
+
"loss": 0.347,
|
| 737 |
+
"step": 60000
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.74,
|
| 741 |
+
"learning_rate": 1.3818849449204408e-05,
|
| 742 |
+
"loss": 0.344,
|
| 743 |
+
"step": 60500
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 1.76,
|
| 747 |
+
"learning_rate": 1.3757649938800491e-05,
|
| 748 |
+
"loss": 0.3617,
|
| 749 |
+
"step": 61000
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 1.77,
|
| 753 |
+
"learning_rate": 1.3696450428396574e-05,
|
| 754 |
+
"loss": 0.3447,
|
| 755 |
+
"step": 61500
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 1.79,
|
| 759 |
+
"learning_rate": 1.3635250917992658e-05,
|
| 760 |
+
"loss": 0.3481,
|
| 761 |
+
"step": 62000
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 1.8,
|
| 765 |
+
"learning_rate": 1.357405140758874e-05,
|
| 766 |
+
"loss": 0.3522,
|
| 767 |
+
"step": 62500
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.82,
|
| 771 |
+
"learning_rate": 1.3512851897184824e-05,
|
| 772 |
+
"loss": 0.3509,
|
| 773 |
+
"step": 63000
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.83,
|
| 777 |
+
"learning_rate": 1.3451652386780906e-05,
|
| 778 |
+
"loss": 0.3529,
|
| 779 |
+
"step": 63500
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.85,
|
| 783 |
+
"learning_rate": 1.339045287637699e-05,
|
| 784 |
+
"loss": 0.3502,
|
| 785 |
+
"step": 64000
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 1.86,
|
| 789 |
+
"learning_rate": 1.3329253365973075e-05,
|
| 790 |
+
"loss": 0.3455,
|
| 791 |
+
"step": 64500
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 1.87,
|
| 795 |
+
"learning_rate": 1.3268053855569156e-05,
|
| 796 |
+
"loss": 0.3567,
|
| 797 |
+
"step": 65000
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 1.89,
|
| 801 |
+
"learning_rate": 1.3206854345165238e-05,
|
| 802 |
+
"loss": 0.3316,
|
| 803 |
+
"step": 65500
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 1.9,
|
| 807 |
+
"learning_rate": 1.3145654834761323e-05,
|
| 808 |
+
"loss": 0.3535,
|
| 809 |
+
"step": 66000
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.92,
|
| 813 |
+
"learning_rate": 1.3084455324357406e-05,
|
| 814 |
+
"loss": 0.37,
|
| 815 |
+
"step": 66500
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.93,
|
| 819 |
+
"learning_rate": 1.302325581395349e-05,
|
| 820 |
+
"loss": 0.3531,
|
| 821 |
+
"step": 67000
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.95,
|
| 825 |
+
"learning_rate": 1.2962056303549573e-05,
|
| 826 |
+
"loss": 0.3728,
|
| 827 |
+
"step": 67500
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 1.96,
|
| 831 |
+
"learning_rate": 1.2900856793145655e-05,
|
| 832 |
+
"loss": 0.3517,
|
| 833 |
+
"step": 68000
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 1.98,
|
| 837 |
+
"learning_rate": 1.283965728274174e-05,
|
| 838 |
+
"loss": 0.3531,
|
| 839 |
+
"step": 68500
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 1.99,
|
| 843 |
+
"learning_rate": 1.2778457772337822e-05,
|
| 844 |
+
"loss": 0.3645,
|
| 845 |
+
"step": 69000
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 2.0,
|
| 849 |
+
"eval_accuracy": 0.8902474526928675,
|
| 850 |
+
"eval_loss": 0.28919732570648193,
|
| 851 |
+
"eval_runtime": 105.8357,
|
| 852 |
+
"eval_samples_per_second": 64.912,
|
| 853 |
+
"eval_steps_per_second": 8.116,
|
| 854 |
+
"step": 69360
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 2.0,
|
| 858 |
+
"learning_rate": 1.2717258261933905e-05,
|
| 859 |
+
"loss": 0.3729,
|
| 860 |
+
"step": 69500
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 2.02,
|
| 864 |
+
"learning_rate": 1.265605875152999e-05,
|
| 865 |
+
"loss": 0.3533,
|
| 866 |
+
"step": 70000
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 2.03,
|
| 870 |
+
"learning_rate": 1.2594859241126072e-05,
|
| 871 |
+
"loss": 0.3419,
|
| 872 |
+
"step": 70500
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 2.05,
|
| 876 |
+
"learning_rate": 1.2533659730722153e-05,
|
| 877 |
+
"loss": 0.3424,
|
| 878 |
+
"step": 71000
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 2.06,
|
| 882 |
+
"learning_rate": 1.2472460220318238e-05,
|
| 883 |
+
"loss": 0.3411,
|
| 884 |
+
"step": 71500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.08,
|
| 888 |
+
"learning_rate": 1.2411260709914322e-05,
|
| 889 |
+
"loss": 0.3358,
|
| 890 |
+
"step": 72000
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 2.09,
|
| 894 |
+
"learning_rate": 1.2350061199510405e-05,
|
| 895 |
+
"loss": 0.3383,
|
| 896 |
+
"step": 72500
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 2.1,
|
| 900 |
+
"learning_rate": 1.2288861689106489e-05,
|
| 901 |
+
"loss": 0.3483,
|
| 902 |
+
"step": 73000
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 2.12,
|
| 906 |
+
"learning_rate": 1.222766217870257e-05,
|
| 907 |
+
"loss": 0.3569,
|
| 908 |
+
"step": 73500
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 2.13,
|
| 912 |
+
"learning_rate": 1.2166462668298655e-05,
|
| 913 |
+
"loss": 0.358,
|
| 914 |
+
"step": 74000
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 2.15,
|
| 918 |
+
"learning_rate": 1.2105263157894737e-05,
|
| 919 |
+
"loss": 0.3548,
|
| 920 |
+
"step": 74500
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 2.16,
|
| 924 |
+
"learning_rate": 1.204406364749082e-05,
|
| 925 |
+
"loss": 0.3455,
|
| 926 |
+
"step": 75000
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 2.18,
|
| 930 |
+
"learning_rate": 1.1982864137086905e-05,
|
| 931 |
+
"loss": 0.3434,
|
| 932 |
+
"step": 75500
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 2.19,
|
| 936 |
+
"learning_rate": 1.1921664626682987e-05,
|
| 937 |
+
"loss": 0.3542,
|
| 938 |
+
"step": 76000
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 2.21,
|
| 942 |
+
"learning_rate": 1.1860465116279072e-05,
|
| 943 |
+
"loss": 0.3472,
|
| 944 |
+
"step": 76500
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 2.22,
|
| 948 |
+
"learning_rate": 1.1799265605875154e-05,
|
| 949 |
+
"loss": 0.3349,
|
| 950 |
+
"step": 77000
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 2.23,
|
| 954 |
+
"learning_rate": 1.1738066095471237e-05,
|
| 955 |
+
"loss": 0.343,
|
| 956 |
+
"step": 77500
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 2.25,
|
| 960 |
+
"learning_rate": 1.167686658506732e-05,
|
| 961 |
+
"loss": 0.3399,
|
| 962 |
+
"step": 78000
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.26,
|
| 966 |
+
"learning_rate": 1.1615667074663404e-05,
|
| 967 |
+
"loss": 0.3455,
|
| 968 |
+
"step": 78500
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.28,
|
| 972 |
+
"learning_rate": 1.1554467564259486e-05,
|
| 973 |
+
"loss": 0.3514,
|
| 974 |
+
"step": 79000
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 2.29,
|
| 978 |
+
"learning_rate": 1.149326805385557e-05,
|
| 979 |
+
"loss": 0.3436,
|
| 980 |
+
"step": 79500
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 2.31,
|
| 984 |
+
"learning_rate": 1.1432068543451652e-05,
|
| 985 |
+
"loss": 0.3484,
|
| 986 |
+
"step": 80000
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 2.32,
|
| 990 |
+
"learning_rate": 1.1370869033047736e-05,
|
| 991 |
+
"loss": 0.3398,
|
| 992 |
+
"step": 80500
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 2.34,
|
| 996 |
+
"learning_rate": 1.1309669522643821e-05,
|
| 997 |
+
"loss": 0.3455,
|
| 998 |
+
"step": 81000
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 2.35,
|
| 1002 |
+
"learning_rate": 1.1248470012239903e-05,
|
| 1003 |
+
"loss": 0.3629,
|
| 1004 |
+
"step": 81500
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.36,
|
| 1008 |
+
"learning_rate": 1.1187270501835988e-05,
|
| 1009 |
+
"loss": 0.3481,
|
| 1010 |
+
"step": 82000
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.38,
|
| 1014 |
+
"learning_rate": 1.112607099143207e-05,
|
| 1015 |
+
"loss": 0.3483,
|
| 1016 |
+
"step": 82500
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 2.39,
|
| 1020 |
+
"learning_rate": 1.1064871481028153e-05,
|
| 1021 |
+
"loss": 0.3478,
|
| 1022 |
+
"step": 83000
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 2.41,
|
| 1026 |
+
"learning_rate": 1.1003671970624236e-05,
|
| 1027 |
+
"loss": 0.3357,
|
| 1028 |
+
"step": 83500
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 2.42,
|
| 1032 |
+
"learning_rate": 1.094247246022032e-05,
|
| 1033 |
+
"loss": 0.3733,
|
| 1034 |
+
"step": 84000
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 2.44,
|
| 1038 |
+
"learning_rate": 1.0881272949816401e-05,
|
| 1039 |
+
"loss": 0.346,
|
| 1040 |
+
"step": 84500
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 2.45,
|
| 1044 |
+
"learning_rate": 1.0820073439412486e-05,
|
| 1045 |
+
"loss": 0.3526,
|
| 1046 |
+
"step": 85000
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.47,
|
| 1050 |
+
"learning_rate": 1.0758873929008568e-05,
|
| 1051 |
+
"loss": 0.3383,
|
| 1052 |
+
"step": 85500
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.48,
|
| 1056 |
+
"learning_rate": 1.0697674418604651e-05,
|
| 1057 |
+
"loss": 0.3447,
|
| 1058 |
+
"step": 86000
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 2.49,
|
| 1062 |
+
"learning_rate": 1.0636474908200736e-05,
|
| 1063 |
+
"loss": 0.3582,
|
| 1064 |
+
"step": 86500
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 2.51,
|
| 1068 |
+
"learning_rate": 1.0575275397796818e-05,
|
| 1069 |
+
"loss": 0.3549,
|
| 1070 |
+
"step": 87000
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.52,
|
| 1074 |
+
"learning_rate": 1.0514075887392903e-05,
|
| 1075 |
+
"loss": 0.376,
|
| 1076 |
+
"step": 87500
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 2.54,
|
| 1080 |
+
"learning_rate": 1.0452876376988985e-05,
|
| 1081 |
+
"loss": 0.3545,
|
| 1082 |
+
"step": 88000
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 2.55,
|
| 1086 |
+
"learning_rate": 1.0391676866585068e-05,
|
| 1087 |
+
"loss": 0.3642,
|
| 1088 |
+
"step": 88500
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.57,
|
| 1092 |
+
"learning_rate": 1.0330477356181152e-05,
|
| 1093 |
+
"loss": 0.3634,
|
| 1094 |
+
"step": 89000
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 2.58,
|
| 1098 |
+
"learning_rate": 1.0269277845777235e-05,
|
| 1099 |
+
"loss": 0.3633,
|
| 1100 |
+
"step": 89500
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 2.6,
|
| 1104 |
+
"learning_rate": 1.0208078335373317e-05,
|
| 1105 |
+
"loss": 0.3502,
|
| 1106 |
+
"step": 90000
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 2.61,
|
| 1110 |
+
"learning_rate": 1.0146878824969402e-05,
|
| 1111 |
+
"loss": 0.3485,
|
| 1112 |
+
"step": 90500
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 2.62,
|
| 1116 |
+
"learning_rate": 1.0085679314565483e-05,
|
| 1117 |
+
"loss": 0.3527,
|
| 1118 |
+
"step": 91000
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 2.64,
|
| 1122 |
+
"learning_rate": 1.0024479804161567e-05,
|
| 1123 |
+
"loss": 0.3552,
|
| 1124 |
+
"step": 91500
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 2.65,
|
| 1128 |
+
"learning_rate": 9.963280293757652e-06,
|
| 1129 |
+
"loss": 0.3445,
|
| 1130 |
+
"step": 92000
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.67,
|
| 1134 |
+
"learning_rate": 9.902080783353734e-06,
|
| 1135 |
+
"loss": 0.3601,
|
| 1136 |
+
"step": 92500
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.68,
|
| 1140 |
+
"learning_rate": 9.840881272949817e-06,
|
| 1141 |
+
"loss": 0.3644,
|
| 1142 |
+
"step": 93000
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 2.7,
|
| 1146 |
+
"learning_rate": 9.7796817625459e-06,
|
| 1147 |
+
"loss": 0.355,
|
| 1148 |
+
"step": 93500
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 2.71,
|
| 1152 |
+
"learning_rate": 9.718482252141984e-06,
|
| 1153 |
+
"loss": 0.3531,
|
| 1154 |
+
"step": 94000
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 2.72,
|
| 1158 |
+
"learning_rate": 9.657282741738067e-06,
|
| 1159 |
+
"loss": 0.3382,
|
| 1160 |
+
"step": 94500
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 2.74,
|
| 1164 |
+
"learning_rate": 9.59608323133415e-06,
|
| 1165 |
+
"loss": 0.3503,
|
| 1166 |
+
"step": 95000
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 2.75,
|
| 1170 |
+
"learning_rate": 9.534883720930234e-06,
|
| 1171 |
+
"loss": 0.3471,
|
| 1172 |
+
"step": 95500
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.77,
|
| 1176 |
+
"learning_rate": 9.473684210526315e-06,
|
| 1177 |
+
"loss": 0.3496,
|
| 1178 |
+
"step": 96000
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 2.78,
|
| 1182 |
+
"learning_rate": 9.412484700122399e-06,
|
| 1183 |
+
"loss": 0.3402,
|
| 1184 |
+
"step": 96500
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 2.8,
|
| 1188 |
+
"learning_rate": 9.351285189718482e-06,
|
| 1189 |
+
"loss": 0.355,
|
| 1190 |
+
"step": 97000
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 2.81,
|
| 1194 |
+
"learning_rate": 9.290085679314567e-06,
|
| 1195 |
+
"loss": 0.3342,
|
| 1196 |
+
"step": 97500
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.83,
|
| 1200 |
+
"learning_rate": 9.228886168910649e-06,
|
| 1201 |
+
"loss": 0.3456,
|
| 1202 |
+
"step": 98000
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 2.84,
|
| 1206 |
+
"learning_rate": 9.167686658506732e-06,
|
| 1207 |
+
"loss": 0.3553,
|
| 1208 |
+
"step": 98500
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 2.85,
|
| 1212 |
+
"learning_rate": 9.106487148102816e-06,
|
| 1213 |
+
"loss": 0.3475,
|
| 1214 |
+
"step": 99000
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.87,
|
| 1218 |
+
"learning_rate": 9.045287637698899e-06,
|
| 1219 |
+
"loss": 0.3507,
|
| 1220 |
+
"step": 99500
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 2.88,
|
| 1224 |
+
"learning_rate": 8.984088127294983e-06,
|
| 1225 |
+
"loss": 0.3457,
|
| 1226 |
+
"step": 100000
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 2.9,
|
| 1230 |
+
"learning_rate": 8.922888616891066e-06,
|
| 1231 |
+
"loss": 0.3445,
|
| 1232 |
+
"step": 100500
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 2.91,
|
| 1236 |
+
"learning_rate": 8.86168910648715e-06,
|
| 1237 |
+
"loss": 0.3623,
|
| 1238 |
+
"step": 101000
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 2.93,
|
| 1242 |
+
"learning_rate": 8.800489596083231e-06,
|
| 1243 |
+
"loss": 0.3477,
|
| 1244 |
+
"step": 101500
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 2.94,
|
| 1248 |
+
"learning_rate": 8.739290085679314e-06,
|
| 1249 |
+
"loss": 0.3363,
|
| 1250 |
+
"step": 102000
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 2.96,
|
| 1254 |
+
"learning_rate": 8.678090575275398e-06,
|
| 1255 |
+
"loss": 0.3445,
|
| 1256 |
+
"step": 102500
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.97,
|
| 1260 |
+
"learning_rate": 8.616891064871483e-06,
|
| 1261 |
+
"loss": 0.3453,
|
| 1262 |
+
"step": 103000
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.98,
|
| 1266 |
+
"learning_rate": 8.555691554467564e-06,
|
| 1267 |
+
"loss": 0.3509,
|
| 1268 |
+
"step": 103500
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 3.0,
|
| 1272 |
+
"learning_rate": 8.494492044063648e-06,
|
| 1273 |
+
"loss": 0.3516,
|
| 1274 |
+
"step": 104000
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 3.0,
|
| 1278 |
+
"eval_accuracy": 0.8893740902474527,
|
| 1279 |
+
"eval_loss": 0.28782936930656433,
|
| 1280 |
+
"eval_runtime": 104.9747,
|
| 1281 |
+
"eval_samples_per_second": 65.444,
|
| 1282 |
+
"eval_steps_per_second": 8.183,
|
| 1283 |
+
"step": 104040
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 3.01,
|
| 1287 |
+
"learning_rate": 8.433292533659731e-06,
|
| 1288 |
+
"loss": 0.3375,
|
| 1289 |
+
"step": 104500
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 3.03,
|
| 1293 |
+
"learning_rate": 8.372093023255815e-06,
|
| 1294 |
+
"loss": 0.3318,
|
| 1295 |
+
"step": 105000
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"epoch": 3.04,
|
| 1299 |
+
"learning_rate": 8.310893512851898e-06,
|
| 1300 |
+
"loss": 0.3481,
|
| 1301 |
+
"step": 105500
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 3.06,
|
| 1305 |
+
"learning_rate": 8.249694002447981e-06,
|
| 1306 |
+
"loss": 0.3387,
|
| 1307 |
+
"step": 106000
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 3.07,
|
| 1311 |
+
"learning_rate": 8.188494492044065e-06,
|
| 1312 |
+
"loss": 0.3336,
|
| 1313 |
+
"step": 106500
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 3.09,
|
| 1317 |
+
"learning_rate": 8.127294981640148e-06,
|
| 1318 |
+
"loss": 0.3451,
|
| 1319 |
+
"step": 107000
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 3.1,
|
| 1323 |
+
"learning_rate": 8.06609547123623e-06,
|
| 1324 |
+
"loss": 0.33,
|
| 1325 |
+
"step": 107500
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 3.11,
|
| 1329 |
+
"learning_rate": 8.004895960832313e-06,
|
| 1330 |
+
"loss": 0.3304,
|
| 1331 |
+
"step": 108000
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 3.13,
|
| 1335 |
+
"learning_rate": 7.943696450428398e-06,
|
| 1336 |
+
"loss": 0.3429,
|
| 1337 |
+
"step": 108500
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 3.14,
|
| 1341 |
+
"learning_rate": 7.88249694002448e-06,
|
| 1342 |
+
"loss": 0.3267,
|
| 1343 |
+
"step": 109000
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 3.16,
|
| 1347 |
+
"learning_rate": 7.821297429620563e-06,
|
| 1348 |
+
"loss": 0.3357,
|
| 1349 |
+
"step": 109500
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 3.17,
|
| 1353 |
+
"learning_rate": 7.760097919216647e-06,
|
| 1354 |
+
"loss": 0.346,
|
| 1355 |
+
"step": 110000
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 3.19,
|
| 1359 |
+
"learning_rate": 7.69889840881273e-06,
|
| 1360 |
+
"loss": 0.3401,
|
| 1361 |
+
"step": 110500
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 3.2,
|
| 1365 |
+
"learning_rate": 7.637698898408813e-06,
|
| 1366 |
+
"loss": 0.3478,
|
| 1367 |
+
"step": 111000
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 3.22,
|
| 1371 |
+
"learning_rate": 7.576499388004897e-06,
|
| 1372 |
+
"loss": 0.338,
|
| 1373 |
+
"step": 111500
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 3.23,
|
| 1377 |
+
"learning_rate": 7.51529987760098e-06,
|
| 1378 |
+
"loss": 0.3375,
|
| 1379 |
+
"step": 112000
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 3.24,
|
| 1383 |
+
"learning_rate": 7.4541003671970636e-06,
|
| 1384 |
+
"loss": 0.3342,
|
| 1385 |
+
"step": 112500
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 3.26,
|
| 1389 |
+
"learning_rate": 7.392900856793146e-06,
|
| 1390 |
+
"loss": 0.3433,
|
| 1391 |
+
"step": 113000
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 3.27,
|
| 1395 |
+
"learning_rate": 7.3317013463892295e-06,
|
| 1396 |
+
"loss": 0.3502,
|
| 1397 |
+
"step": 113500
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 3.29,
|
| 1401 |
+
"learning_rate": 7.270501835985313e-06,
|
| 1402 |
+
"loss": 0.3409,
|
| 1403 |
+
"step": 114000
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 3.3,
|
| 1407 |
+
"learning_rate": 7.209302325581395e-06,
|
| 1408 |
+
"loss": 0.3335,
|
| 1409 |
+
"step": 114500
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 3.32,
|
| 1413 |
+
"learning_rate": 7.148102815177479e-06,
|
| 1414 |
+
"loss": 0.3508,
|
| 1415 |
+
"step": 115000
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 3.33,
|
| 1419 |
+
"learning_rate": 7.086903304773562e-06,
|
| 1420 |
+
"loss": 0.3485,
|
| 1421 |
+
"step": 115500
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 3.34,
|
| 1425 |
+
"learning_rate": 7.025703794369646e-06,
|
| 1426 |
+
"loss": 0.3517,
|
| 1427 |
+
"step": 116000
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 3.36,
|
| 1431 |
+
"learning_rate": 6.964504283965728e-06,
|
| 1432 |
+
"loss": 0.3336,
|
| 1433 |
+
"step": 116500
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 3.37,
|
| 1437 |
+
"learning_rate": 6.903304773561812e-06,
|
| 1438 |
+
"loss": 0.3591,
|
| 1439 |
+
"step": 117000
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 3.39,
|
| 1443 |
+
"learning_rate": 6.842105263157896e-06,
|
| 1444 |
+
"loss": 0.3456,
|
| 1445 |
+
"step": 117500
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 3.4,
|
| 1449 |
+
"learning_rate": 6.780905752753979e-06,
|
| 1450 |
+
"loss": 0.338,
|
| 1451 |
+
"step": 118000
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 3.42,
|
| 1455 |
+
"learning_rate": 6.7197062423500616e-06,
|
| 1456 |
+
"loss": 0.3456,
|
| 1457 |
+
"step": 118500
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 3.43,
|
| 1461 |
+
"learning_rate": 6.658506731946145e-06,
|
| 1462 |
+
"loss": 0.3559,
|
| 1463 |
+
"step": 119000
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 3.45,
|
| 1467 |
+
"learning_rate": 6.597307221542228e-06,
|
| 1468 |
+
"loss": 0.3462,
|
| 1469 |
+
"step": 119500
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 3.46,
|
| 1473 |
+
"learning_rate": 6.536107711138311e-06,
|
| 1474 |
+
"loss": 0.3343,
|
| 1475 |
+
"step": 120000
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 3.47,
|
| 1479 |
+
"learning_rate": 6.474908200734394e-06,
|
| 1480 |
+
"loss": 0.3238,
|
| 1481 |
+
"step": 120500
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 3.49,
|
| 1485 |
+
"learning_rate": 6.413708690330478e-06,
|
| 1486 |
+
"loss": 0.3459,
|
| 1487 |
+
"step": 121000
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 3.5,
|
| 1491 |
+
"learning_rate": 6.352509179926562e-06,
|
| 1492 |
+
"loss": 0.362,
|
| 1493 |
+
"step": 121500
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 3.52,
|
| 1497 |
+
"learning_rate": 6.2913096695226435e-06,
|
| 1498 |
+
"loss": 0.3579,
|
| 1499 |
+
"step": 122000
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 3.53,
|
| 1503 |
+
"learning_rate": 6.230110159118728e-06,
|
| 1504 |
+
"loss": 0.3487,
|
| 1505 |
+
"step": 122500
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 3.55,
|
| 1509 |
+
"learning_rate": 6.168910648714811e-06,
|
| 1510 |
+
"loss": 0.3362,
|
| 1511 |
+
"step": 123000
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 3.56,
|
| 1515 |
+
"learning_rate": 6.1077111383108945e-06,
|
| 1516 |
+
"loss": 0.3466,
|
| 1517 |
+
"step": 123500
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 3.58,
|
| 1521 |
+
"learning_rate": 6.046511627906977e-06,
|
| 1522 |
+
"loss": 0.3487,
|
| 1523 |
+
"step": 124000
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 3.59,
|
| 1527 |
+
"learning_rate": 5.98531211750306e-06,
|
| 1528 |
+
"loss": 0.3408,
|
| 1529 |
+
"step": 124500
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 3.6,
|
| 1533 |
+
"learning_rate": 5.924112607099144e-06,
|
| 1534 |
+
"loss": 0.3456,
|
| 1535 |
+
"step": 125000
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 3.62,
|
| 1539 |
+
"learning_rate": 5.862913096695227e-06,
|
| 1540 |
+
"loss": 0.347,
|
| 1541 |
+
"step": 125500
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 3.63,
|
| 1545 |
+
"learning_rate": 5.80171358629131e-06,
|
| 1546 |
+
"loss": 0.3368,
|
| 1547 |
+
"step": 126000
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 3.65,
|
| 1551 |
+
"learning_rate": 5.740514075887393e-06,
|
| 1552 |
+
"loss": 0.3428,
|
| 1553 |
+
"step": 126500
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 3.66,
|
| 1557 |
+
"learning_rate": 5.679314565483477e-06,
|
| 1558 |
+
"loss": 0.3376,
|
| 1559 |
+
"step": 127000
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 3.68,
|
| 1563 |
+
"learning_rate": 5.618115055079559e-06,
|
| 1564 |
+
"loss": 0.3434,
|
| 1565 |
+
"step": 127500
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 3.69,
|
| 1569 |
+
"learning_rate": 5.556915544675643e-06,
|
| 1570 |
+
"loss": 0.3393,
|
| 1571 |
+
"step": 128000
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 3.71,
|
| 1575 |
+
"learning_rate": 5.4957160342717266e-06,
|
| 1576 |
+
"loss": 0.3397,
|
| 1577 |
+
"step": 128500
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 3.72,
|
| 1581 |
+
"learning_rate": 5.43451652386781e-06,
|
| 1582 |
+
"loss": 0.3281,
|
| 1583 |
+
"step": 129000
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 3.73,
|
| 1587 |
+
"learning_rate": 5.3733170134638925e-06,
|
| 1588 |
+
"loss": 0.3318,
|
| 1589 |
+
"step": 129500
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 3.75,
|
| 1593 |
+
"learning_rate": 5.312117503059976e-06,
|
| 1594 |
+
"loss": 0.3417,
|
| 1595 |
+
"step": 130000
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 3.76,
|
| 1599 |
+
"learning_rate": 5.250917992656059e-06,
|
| 1600 |
+
"loss": 0.3502,
|
| 1601 |
+
"step": 130500
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 3.78,
|
| 1605 |
+
"learning_rate": 5.189718482252143e-06,
|
| 1606 |
+
"loss": 0.3476,
|
| 1607 |
+
"step": 131000
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 3.79,
|
| 1611 |
+
"learning_rate": 5.128518971848225e-06,
|
| 1612 |
+
"loss": 0.3351,
|
| 1613 |
+
"step": 131500
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 3.81,
|
| 1617 |
+
"learning_rate": 5.0673194614443085e-06,
|
| 1618 |
+
"loss": 0.3328,
|
| 1619 |
+
"step": 132000
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 3.82,
|
| 1623 |
+
"learning_rate": 5.006119951040393e-06,
|
| 1624 |
+
"loss": 0.3388,
|
| 1625 |
+
"step": 132500
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 3.84,
|
| 1629 |
+
"learning_rate": 4.944920440636475e-06,
|
| 1630 |
+
"loss": 0.3465,
|
| 1631 |
+
"step": 133000
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 3.85,
|
| 1635 |
+
"learning_rate": 4.883720930232559e-06,
|
| 1636 |
+
"loss": 0.3549,
|
| 1637 |
+
"step": 133500
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 3.86,
|
| 1641 |
+
"learning_rate": 4.822521419828642e-06,
|
| 1642 |
+
"loss": 0.3524,
|
| 1643 |
+
"step": 134000
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 3.88,
|
| 1647 |
+
"learning_rate": 4.7613219094247246e-06,
|
| 1648 |
+
"loss": 0.3398,
|
| 1649 |
+
"step": 134500
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 3.89,
|
| 1653 |
+
"learning_rate": 4.700122399020808e-06,
|
| 1654 |
+
"loss": 0.3547,
|
| 1655 |
+
"step": 135000
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 3.91,
|
| 1659 |
+
"learning_rate": 4.638922888616891e-06,
|
| 1660 |
+
"loss": 0.3386,
|
| 1661 |
+
"step": 135500
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 3.92,
|
| 1665 |
+
"learning_rate": 4.577723378212975e-06,
|
| 1666 |
+
"loss": 0.3498,
|
| 1667 |
+
"step": 136000
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"epoch": 3.94,
|
| 1671 |
+
"learning_rate": 4.516523867809058e-06,
|
| 1672 |
+
"loss": 0.3341,
|
| 1673 |
+
"step": 136500
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 3.95,
|
| 1677 |
+
"learning_rate": 4.4553243574051415e-06,
|
| 1678 |
+
"loss": 0.3401,
|
| 1679 |
+
"step": 137000
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 3.96,
|
| 1683 |
+
"learning_rate": 4.394124847001224e-06,
|
| 1684 |
+
"loss": 0.3359,
|
| 1685 |
+
"step": 137500
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 3.98,
|
| 1689 |
+
"learning_rate": 4.332925336597308e-06,
|
| 1690 |
+
"loss": 0.3511,
|
| 1691 |
+
"step": 138000
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 3.99,
|
| 1695 |
+
"learning_rate": 4.271725826193391e-06,
|
| 1696 |
+
"loss": 0.3306,
|
| 1697 |
+
"step": 138500
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 4.0,
|
| 1701 |
+
"eval_accuracy": 0.8962154294032023,
|
| 1702 |
+
"eval_loss": 0.292516827583313,
|
| 1703 |
+
"eval_runtime": 105.0784,
|
| 1704 |
+
"eval_samples_per_second": 65.38,
|
| 1705 |
+
"eval_steps_per_second": 8.175,
|
| 1706 |
+
"step": 138720
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 4.01,
|
| 1710 |
+
"learning_rate": 4.210526315789474e-06,
|
| 1711 |
+
"loss": 0.3299,
|
| 1712 |
+
"step": 139000
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 4.02,
|
| 1716 |
+
"learning_rate": 4.1493268053855575e-06,
|
| 1717 |
+
"loss": 0.3361,
|
| 1718 |
+
"step": 139500
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 4.04,
|
| 1722 |
+
"learning_rate": 4.08812729498164e-06,
|
| 1723 |
+
"loss": 0.3293,
|
| 1724 |
+
"step": 140000
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 4.05,
|
| 1728 |
+
"learning_rate": 4.026927784577723e-06,
|
| 1729 |
+
"loss": 0.3197,
|
| 1730 |
+
"step": 140500
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 4.07,
|
| 1734 |
+
"learning_rate": 3.965728274173807e-06,
|
| 1735 |
+
"loss": 0.3438,
|
| 1736 |
+
"step": 141000
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 4.08,
|
| 1740 |
+
"learning_rate": 3.90452876376989e-06,
|
| 1741 |
+
"loss": 0.3336,
|
| 1742 |
+
"step": 141500
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 4.09,
|
| 1746 |
+
"learning_rate": 3.8433292533659735e-06,
|
| 1747 |
+
"loss": 0.3318,
|
| 1748 |
+
"step": 142000
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 4.11,
|
| 1752 |
+
"learning_rate": 3.782129742962057e-06,
|
| 1753 |
+
"loss": 0.3402,
|
| 1754 |
+
"step": 142500
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 4.12,
|
| 1758 |
+
"learning_rate": 3.72093023255814e-06,
|
| 1759 |
+
"loss": 0.3291,
|
| 1760 |
+
"step": 143000
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 4.14,
|
| 1764 |
+
"learning_rate": 3.6597307221542232e-06,
|
| 1765 |
+
"loss": 0.3319,
|
| 1766 |
+
"step": 143500
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 4.15,
|
| 1770 |
+
"learning_rate": 3.598531211750306e-06,
|
| 1771 |
+
"loss": 0.3278,
|
| 1772 |
+
"step": 144000
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 4.17,
|
| 1776 |
+
"learning_rate": 3.537331701346389e-06,
|
| 1777 |
+
"loss": 0.3332,
|
| 1778 |
+
"step": 144500
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 4.18,
|
| 1782 |
+
"learning_rate": 3.476132190942473e-06,
|
| 1783 |
+
"loss": 0.3366,
|
| 1784 |
+
"step": 145000
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 4.2,
|
| 1788 |
+
"learning_rate": 3.414932680538556e-06,
|
| 1789 |
+
"loss": 0.3364,
|
| 1790 |
+
"step": 145500
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 4.21,
|
| 1794 |
+
"learning_rate": 3.3537331701346393e-06,
|
| 1795 |
+
"loss": 0.3333,
|
| 1796 |
+
"step": 146000
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 4.22,
|
| 1800 |
+
"learning_rate": 3.2925336597307222e-06,
|
| 1801 |
+
"loss": 0.3264,
|
| 1802 |
+
"step": 146500
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 4.24,
|
| 1806 |
+
"learning_rate": 3.2313341493268056e-06,
|
| 1807 |
+
"loss": 0.3213,
|
| 1808 |
+
"step": 147000
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 4.25,
|
| 1812 |
+
"learning_rate": 3.1701346389228886e-06,
|
| 1813 |
+
"loss": 0.3323,
|
| 1814 |
+
"step": 147500
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 4.27,
|
| 1818 |
+
"learning_rate": 3.1089351285189724e-06,
|
| 1819 |
+
"loss": 0.3389,
|
| 1820 |
+
"step": 148000
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 4.28,
|
| 1824 |
+
"learning_rate": 3.0477356181150553e-06,
|
| 1825 |
+
"loss": 0.3235,
|
| 1826 |
+
"step": 148500
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 4.3,
|
| 1830 |
+
"learning_rate": 2.9865361077111387e-06,
|
| 1831 |
+
"loss": 0.3395,
|
| 1832 |
+
"step": 149000
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 4.31,
|
| 1836 |
+
"learning_rate": 2.9253365973072217e-06,
|
| 1837 |
+
"loss": 0.3305,
|
| 1838 |
+
"step": 149500
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 4.33,
|
| 1842 |
+
"learning_rate": 2.8641370869033055e-06,
|
| 1843 |
+
"loss": 0.3263,
|
| 1844 |
+
"step": 150000
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 4.34,
|
| 1848 |
+
"learning_rate": 2.8029375764993884e-06,
|
| 1849 |
+
"loss": 0.3425,
|
| 1850 |
+
"step": 150500
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 4.35,
|
| 1854 |
+
"learning_rate": 2.7417380660954714e-06,
|
| 1855 |
+
"loss": 0.3285,
|
| 1856 |
+
"step": 151000
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 4.37,
|
| 1860 |
+
"learning_rate": 2.6805385556915547e-06,
|
| 1861 |
+
"loss": 0.3256,
|
| 1862 |
+
"step": 151500
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 4.38,
|
| 1866 |
+
"learning_rate": 2.6193390452876377e-06,
|
| 1867 |
+
"loss": 0.3289,
|
| 1868 |
+
"step": 152000
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 4.4,
|
| 1872 |
+
"learning_rate": 2.558139534883721e-06,
|
| 1873 |
+
"loss": 0.3445,
|
| 1874 |
+
"step": 152500
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 4.41,
|
| 1878 |
+
"learning_rate": 2.4969400244798045e-06,
|
| 1879 |
+
"loss": 0.3381,
|
| 1880 |
+
"step": 153000
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 4.43,
|
| 1884 |
+
"learning_rate": 2.4357405140758874e-06,
|
| 1885 |
+
"loss": 0.3312,
|
| 1886 |
+
"step": 153500
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 4.44,
|
| 1890 |
+
"learning_rate": 2.374541003671971e-06,
|
| 1891 |
+
"loss": 0.3235,
|
| 1892 |
+
"step": 154000
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 4.46,
|
| 1896 |
+
"learning_rate": 2.313341493268054e-06,
|
| 1897 |
+
"loss": 0.3379,
|
| 1898 |
+
"step": 154500
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 4.47,
|
| 1902 |
+
"learning_rate": 2.252141982864137e-06,
|
| 1903 |
+
"loss": 0.3359,
|
| 1904 |
+
"step": 155000
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 4.48,
|
| 1908 |
+
"learning_rate": 2.1909424724602205e-06,
|
| 1909 |
+
"loss": 0.3199,
|
| 1910 |
+
"step": 155500
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 4.5,
|
| 1914 |
+
"learning_rate": 2.129742962056304e-06,
|
| 1915 |
+
"loss": 0.3291,
|
| 1916 |
+
"step": 156000
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 4.51,
|
| 1920 |
+
"learning_rate": 2.068543451652387e-06,
|
| 1921 |
+
"loss": 0.3218,
|
| 1922 |
+
"step": 156500
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 4.53,
|
| 1926 |
+
"learning_rate": 2.00734394124847e-06,
|
| 1927 |
+
"loss": 0.3362,
|
| 1928 |
+
"step": 157000
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 4.54,
|
| 1932 |
+
"learning_rate": 1.9461444308445536e-06,
|
| 1933 |
+
"loss": 0.3308,
|
| 1934 |
+
"step": 157500
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 4.56,
|
| 1938 |
+
"learning_rate": 1.8849449204406368e-06,
|
| 1939 |
+
"loss": 0.3229,
|
| 1940 |
+
"step": 158000
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 4.57,
|
| 1944 |
+
"learning_rate": 1.82374541003672e-06,
|
| 1945 |
+
"loss": 0.3406,
|
| 1946 |
+
"step": 158500
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 4.58,
|
| 1950 |
+
"learning_rate": 1.7625458996328033e-06,
|
| 1951 |
+
"loss": 0.3372,
|
| 1952 |
+
"step": 159000
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 4.6,
|
| 1956 |
+
"learning_rate": 1.7013463892288863e-06,
|
| 1957 |
+
"loss": 0.3181,
|
| 1958 |
+
"step": 159500
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 4.61,
|
| 1962 |
+
"learning_rate": 1.6401468788249694e-06,
|
| 1963 |
+
"loss": 0.3182,
|
| 1964 |
+
"step": 160000
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 4.63,
|
| 1968 |
+
"learning_rate": 1.5789473684210526e-06,
|
| 1969 |
+
"loss": 0.3362,
|
| 1970 |
+
"step": 160500
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 4.64,
|
| 1974 |
+
"learning_rate": 1.517747858017136e-06,
|
| 1975 |
+
"loss": 0.3269,
|
| 1976 |
+
"step": 161000
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 4.66,
|
| 1980 |
+
"learning_rate": 1.4565483476132191e-06,
|
| 1981 |
+
"loss": 0.3409,
|
| 1982 |
+
"step": 161500
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 4.67,
|
| 1986 |
+
"learning_rate": 1.3953488372093025e-06,
|
| 1987 |
+
"loss": 0.3411,
|
| 1988 |
+
"step": 162000
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 4.69,
|
| 1992 |
+
"learning_rate": 1.3341493268053857e-06,
|
| 1993 |
+
"loss": 0.3064,
|
| 1994 |
+
"step": 162500
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 4.7,
|
| 1998 |
+
"learning_rate": 1.2729498164014688e-06,
|
| 1999 |
+
"loss": 0.3364,
|
| 2000 |
+
"step": 163000
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 4.71,
|
| 2004 |
+
"learning_rate": 1.2117503059975522e-06,
|
| 2005 |
+
"loss": 0.3444,
|
| 2006 |
+
"step": 163500
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 4.73,
|
| 2010 |
+
"learning_rate": 1.1505507955936354e-06,
|
| 2011 |
+
"loss": 0.33,
|
| 2012 |
+
"step": 164000
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 4.74,
|
| 2016 |
+
"learning_rate": 1.0893512851897185e-06,
|
| 2017 |
+
"loss": 0.3212,
|
| 2018 |
+
"step": 164500
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 4.76,
|
| 2022 |
+
"learning_rate": 1.0281517747858017e-06,
|
| 2023 |
+
"loss": 0.3356,
|
| 2024 |
+
"step": 165000
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 4.77,
|
| 2028 |
+
"learning_rate": 9.66952264381885e-07,
|
| 2029 |
+
"loss": 0.319,
|
| 2030 |
+
"step": 165500
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 4.79,
|
| 2034 |
+
"learning_rate": 9.057527539779683e-07,
|
| 2035 |
+
"loss": 0.3363,
|
| 2036 |
+
"step": 166000
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 4.8,
|
| 2040 |
+
"learning_rate": 8.445532435740515e-07,
|
| 2041 |
+
"loss": 0.3466,
|
| 2042 |
+
"step": 166500
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 4.82,
|
| 2046 |
+
"learning_rate": 7.833537331701346e-07,
|
| 2047 |
+
"loss": 0.3301,
|
| 2048 |
+
"step": 167000
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 4.83,
|
| 2052 |
+
"learning_rate": 7.221542227662179e-07,
|
| 2053 |
+
"loss": 0.3286,
|
| 2054 |
+
"step": 167500
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 4.84,
|
| 2058 |
+
"learning_rate": 6.609547123623011e-07,
|
| 2059 |
+
"loss": 0.3233,
|
| 2060 |
+
"step": 168000
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 4.86,
|
| 2064 |
+
"learning_rate": 5.997552019583844e-07,
|
| 2065 |
+
"loss": 0.3333,
|
| 2066 |
+
"step": 168500
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 4.87,
|
| 2070 |
+
"learning_rate": 5.385556915544677e-07,
|
| 2071 |
+
"loss": 0.3129,
|
| 2072 |
+
"step": 169000
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 4.89,
|
| 2076 |
+
"learning_rate": 4.773561811505508e-07,
|
| 2077 |
+
"loss": 0.338,
|
| 2078 |
+
"step": 169500
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 4.9,
|
| 2082 |
+
"learning_rate": 4.1615667074663406e-07,
|
| 2083 |
+
"loss": 0.3297,
|
| 2084 |
+
"step": 170000
|
| 2085 |
+
}
|
| 2086 |
+
],
|
| 2087 |
+
"max_steps": 173400,
|
| 2088 |
+
"num_train_epochs": 5,
|
| 2089 |
+
"total_flos": 3.578310352896e+17,
|
| 2090 |
+
"trial_name": null,
|
| 2091 |
+
"trial_params": null
|
| 2092 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bc3f728bb954ad04a9a7104f6368ca5eb4dc802140c56db01820de86abb371aa
|
| 3 |
+
size 3439
|
vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|