Update from ubuntu
Browse files- README.md +1 -0
- config.json +29 -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 +1 -0
- tokenizer.json +0 -0
- tokenizer_config.json +1 -0
- trainer_state.json +2596 -0
- training_args.bin +3 -0
README.md
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
| 1 |
+
hello
|
config.json
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "xlm-roberta-base",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"XLMRobertaForSequenceClassification"
|
| 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 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 3072,
|
| 15 |
+
"layer_norm_eps": 1e-05,
|
| 16 |
+
"max_position_embeddings": 514,
|
| 17 |
+
"model_type": "xlm-roberta",
|
| 18 |
+
"num_attention_heads": 12,
|
| 19 |
+
"num_hidden_layers": 12,
|
| 20 |
+
"output_past": true,
|
| 21 |
+
"pad_token_id": 1,
|
| 22 |
+
"position_embedding_type": "absolute",
|
| 23 |
+
"problem_type": "single_label_classification",
|
| 24 |
+
"torch_dtype": "float32",
|
| 25 |
+
"transformers_version": "4.12.3",
|
| 26 |
+
"type_vocab_size": 1,
|
| 27 |
+
"use_cache": true,
|
| 28 |
+
"vocab_size": 250002
|
| 29 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b670cff935cfa966bc8ea23f4f716d95b37db87e607e3b6229932502f2cc23a2
|
| 3 |
+
size 558084205
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:245aacb587794ce0dd261d5cb53488f2995e7b2827270023bc732b663d5ba20a
|
| 3 |
+
size 1112266157
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cd09167824b340ea25826de3f56cf6f1a9dfa924aab1df7342a2b276da2f1702
|
| 3 |
+
size 14503
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bc5eefe7da6f58b199da3df3d8086cf40bf53e353a3b4e222e86dc35290f3ba5
|
| 3 |
+
size 559
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8c0976bde8727a36cf57e2a8476acca29bc89861be8ea075e35dc6fa22ec32e4
|
| 3 |
+
size 623
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
| 1 |
+
{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": false}}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
| 1 |
+
{"bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "model_max_length": 256, "special_tokens_map_file": null, "name_or_path": "xlm-roberta-base", "tokenizer_class": "XLMRobertaTokenizer"}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2596 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"global_step": 10000,
|
| 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": 2.5000000000000004e-07,
|
| 13 |
+
"loss": 0.7169,
|
| 14 |
+
"step": 25
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.01,
|
| 18 |
+
"learning_rate": 5.000000000000001e-07,
|
| 19 |
+
"loss": 0.7164,
|
| 20 |
+
"step": 50
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 7.5e-07,
|
| 25 |
+
"loss": 0.7121,
|
| 26 |
+
"step": 75
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.01,
|
| 30 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 31 |
+
"loss": 0.7051,
|
| 32 |
+
"step": 100
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.01,
|
| 36 |
+
"learning_rate": 1.2400000000000002e-06,
|
| 37 |
+
"loss": 0.6987,
|
| 38 |
+
"step": 125
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.01,
|
| 42 |
+
"learning_rate": 1.48e-06,
|
| 43 |
+
"loss": 0.6951,
|
| 44 |
+
"step": 150
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.02,
|
| 48 |
+
"learning_rate": 1.73e-06,
|
| 49 |
+
"loss": 0.6815,
|
| 50 |
+
"step": 175
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.02,
|
| 54 |
+
"learning_rate": 1.98e-06,
|
| 55 |
+
"loss": 0.659,
|
| 56 |
+
"step": 200
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.02,
|
| 60 |
+
"learning_rate": 2.2300000000000002e-06,
|
| 61 |
+
"loss": 0.6204,
|
| 62 |
+
"step": 225
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.03,
|
| 66 |
+
"learning_rate": 2.4800000000000004e-06,
|
| 67 |
+
"loss": 0.586,
|
| 68 |
+
"step": 250
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.03,
|
| 72 |
+
"learning_rate": 2.7300000000000005e-06,
|
| 73 |
+
"loss": 0.5613,
|
| 74 |
+
"step": 275
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.03,
|
| 78 |
+
"learning_rate": 2.97e-06,
|
| 79 |
+
"loss": 0.5296,
|
| 80 |
+
"step": 300
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.03,
|
| 84 |
+
"learning_rate": 3.2200000000000005e-06,
|
| 85 |
+
"loss": 0.4957,
|
| 86 |
+
"step": 325
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.04,
|
| 90 |
+
"learning_rate": 3.4700000000000002e-06,
|
| 91 |
+
"loss": 0.4583,
|
| 92 |
+
"step": 350
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.04,
|
| 96 |
+
"learning_rate": 3.7200000000000004e-06,
|
| 97 |
+
"loss": 0.4508,
|
| 98 |
+
"step": 375
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.04,
|
| 102 |
+
"learning_rate": 3.97e-06,
|
| 103 |
+
"loss": 0.3925,
|
| 104 |
+
"step": 400
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.04,
|
| 108 |
+
"learning_rate": 4.22e-06,
|
| 109 |
+
"loss": 0.3554,
|
| 110 |
+
"step": 425
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.04,
|
| 114 |
+
"learning_rate": 4.47e-06,
|
| 115 |
+
"loss": 0.3238,
|
| 116 |
+
"step": 450
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.05,
|
| 120 |
+
"learning_rate": 4.7200000000000005e-06,
|
| 121 |
+
"loss": 0.2907,
|
| 122 |
+
"step": 475
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.05,
|
| 126 |
+
"learning_rate": 4.970000000000001e-06,
|
| 127 |
+
"loss": 0.2741,
|
| 128 |
+
"step": 500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.05,
|
| 132 |
+
"eval_accuracy": 0.9106,
|
| 133 |
+
"eval_loss": 0.2429841309785843,
|
| 134 |
+
"eval_runtime": 157.4737,
|
| 135 |
+
"eval_samples_per_second": 63.503,
|
| 136 |
+
"eval_steps_per_second": 0.349,
|
| 137 |
+
"step": 500
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.05,
|
| 141 |
+
"learning_rate": 5.220000000000001e-06,
|
| 142 |
+
"loss": 0.2663,
|
| 143 |
+
"step": 525
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.06,
|
| 147 |
+
"learning_rate": 5.470000000000001e-06,
|
| 148 |
+
"loss": 0.247,
|
| 149 |
+
"step": 550
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.06,
|
| 153 |
+
"learning_rate": 5.72e-06,
|
| 154 |
+
"loss": 0.21,
|
| 155 |
+
"step": 575
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.06,
|
| 159 |
+
"learning_rate": 5.9700000000000004e-06,
|
| 160 |
+
"loss": 0.2073,
|
| 161 |
+
"step": 600
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.06,
|
| 165 |
+
"learning_rate": 6.220000000000001e-06,
|
| 166 |
+
"loss": 0.2088,
|
| 167 |
+
"step": 625
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 0.07,
|
| 171 |
+
"learning_rate": 6.470000000000001e-06,
|
| 172 |
+
"loss": 0.2218,
|
| 173 |
+
"step": 650
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 0.07,
|
| 177 |
+
"learning_rate": 6.720000000000001e-06,
|
| 178 |
+
"loss": 0.2045,
|
| 179 |
+
"step": 675
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.07,
|
| 183 |
+
"learning_rate": 6.97e-06,
|
| 184 |
+
"loss": 0.1942,
|
| 185 |
+
"step": 700
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.07,
|
| 189 |
+
"learning_rate": 7.22e-06,
|
| 190 |
+
"loss": 0.1776,
|
| 191 |
+
"step": 725
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.07,
|
| 195 |
+
"learning_rate": 7.4700000000000005e-06,
|
| 196 |
+
"loss": 0.1816,
|
| 197 |
+
"step": 750
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.08,
|
| 201 |
+
"learning_rate": 7.72e-06,
|
| 202 |
+
"loss": 0.177,
|
| 203 |
+
"step": 775
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.08,
|
| 207 |
+
"learning_rate": 7.970000000000002e-06,
|
| 208 |
+
"loss": 0.1678,
|
| 209 |
+
"step": 800
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.08,
|
| 213 |
+
"learning_rate": 8.220000000000001e-06,
|
| 214 |
+
"loss": 0.1682,
|
| 215 |
+
"step": 825
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.09,
|
| 219 |
+
"learning_rate": 8.47e-06,
|
| 220 |
+
"loss": 0.1619,
|
| 221 |
+
"step": 850
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.09,
|
| 225 |
+
"learning_rate": 8.720000000000001e-06,
|
| 226 |
+
"loss": 0.1631,
|
| 227 |
+
"step": 875
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.09,
|
| 231 |
+
"learning_rate": 8.97e-06,
|
| 232 |
+
"loss": 0.1629,
|
| 233 |
+
"step": 900
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.09,
|
| 237 |
+
"learning_rate": 9.220000000000002e-06,
|
| 238 |
+
"loss": 0.1671,
|
| 239 |
+
"step": 925
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.1,
|
| 243 |
+
"learning_rate": 9.47e-06,
|
| 244 |
+
"loss": 0.1479,
|
| 245 |
+
"step": 950
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 0.1,
|
| 249 |
+
"learning_rate": 9.72e-06,
|
| 250 |
+
"loss": 0.1494,
|
| 251 |
+
"step": 975
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.1,
|
| 255 |
+
"learning_rate": 9.970000000000001e-06,
|
| 256 |
+
"loss": 0.1624,
|
| 257 |
+
"step": 1000
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.1,
|
| 261 |
+
"eval_accuracy": 0.9495,
|
| 262 |
+
"eval_loss": 0.14469695091247559,
|
| 263 |
+
"eval_runtime": 146.152,
|
| 264 |
+
"eval_samples_per_second": 68.422,
|
| 265 |
+
"eval_steps_per_second": 0.376,
|
| 266 |
+
"step": 1000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.1,
|
| 270 |
+
"learning_rate": 9.975555555555556e-06,
|
| 271 |
+
"loss": 0.149,
|
| 272 |
+
"step": 1025
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.1,
|
| 276 |
+
"learning_rate": 9.947777777777779e-06,
|
| 277 |
+
"loss": 0.1482,
|
| 278 |
+
"step": 1050
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.11,
|
| 282 |
+
"learning_rate": 9.920000000000002e-06,
|
| 283 |
+
"loss": 0.1335,
|
| 284 |
+
"step": 1075
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.11,
|
| 288 |
+
"learning_rate": 9.892222222222223e-06,
|
| 289 |
+
"loss": 0.1405,
|
| 290 |
+
"step": 1100
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.11,
|
| 294 |
+
"learning_rate": 9.864444444444446e-06,
|
| 295 |
+
"loss": 0.1226,
|
| 296 |
+
"step": 1125
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.12,
|
| 300 |
+
"learning_rate": 9.836666666666668e-06,
|
| 301 |
+
"loss": 0.1263,
|
| 302 |
+
"step": 1150
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.12,
|
| 306 |
+
"learning_rate": 9.80888888888889e-06,
|
| 307 |
+
"loss": 0.1383,
|
| 308 |
+
"step": 1175
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.12,
|
| 312 |
+
"learning_rate": 9.781111111111113e-06,
|
| 313 |
+
"loss": 0.1338,
|
| 314 |
+
"step": 1200
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.12,
|
| 318 |
+
"learning_rate": 9.753333333333335e-06,
|
| 319 |
+
"loss": 0.1226,
|
| 320 |
+
"step": 1225
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.12,
|
| 324 |
+
"learning_rate": 9.725555555555557e-06,
|
| 325 |
+
"loss": 0.1342,
|
| 326 |
+
"step": 1250
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.13,
|
| 330 |
+
"learning_rate": 9.697777777777778e-06,
|
| 331 |
+
"loss": 0.137,
|
| 332 |
+
"step": 1275
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.13,
|
| 336 |
+
"learning_rate": 9.67e-06,
|
| 337 |
+
"loss": 0.1345,
|
| 338 |
+
"step": 1300
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.13,
|
| 342 |
+
"learning_rate": 9.642222222222223e-06,
|
| 343 |
+
"loss": 0.1304,
|
| 344 |
+
"step": 1325
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.14,
|
| 348 |
+
"learning_rate": 9.614444444444445e-06,
|
| 349 |
+
"loss": 0.1351,
|
| 350 |
+
"step": 1350
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.14,
|
| 354 |
+
"learning_rate": 9.586666666666667e-06,
|
| 355 |
+
"loss": 0.1232,
|
| 356 |
+
"step": 1375
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.14,
|
| 360 |
+
"learning_rate": 9.55888888888889e-06,
|
| 361 |
+
"loss": 0.1275,
|
| 362 |
+
"step": 1400
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.14,
|
| 366 |
+
"learning_rate": 9.531111111111112e-06,
|
| 367 |
+
"loss": 0.1218,
|
| 368 |
+
"step": 1425
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.14,
|
| 372 |
+
"learning_rate": 9.503333333333334e-06,
|
| 373 |
+
"loss": 0.119,
|
| 374 |
+
"step": 1450
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.15,
|
| 378 |
+
"learning_rate": 9.475555555555557e-06,
|
| 379 |
+
"loss": 0.1188,
|
| 380 |
+
"step": 1475
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.15,
|
| 384 |
+
"learning_rate": 9.447777777777778e-06,
|
| 385 |
+
"loss": 0.1196,
|
| 386 |
+
"step": 1500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.15,
|
| 390 |
+
"eval_accuracy": 0.9481,
|
| 391 |
+
"eval_loss": 0.13628898561000824,
|
| 392 |
+
"eval_runtime": 151.5738,
|
| 393 |
+
"eval_samples_per_second": 65.974,
|
| 394 |
+
"eval_steps_per_second": 0.363,
|
| 395 |
+
"step": 1500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.15,
|
| 399 |
+
"learning_rate": 9.42e-06,
|
| 400 |
+
"loss": 0.1205,
|
| 401 |
+
"step": 1525
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.15,
|
| 405 |
+
"learning_rate": 9.392222222222222e-06,
|
| 406 |
+
"loss": 0.1105,
|
| 407 |
+
"step": 1550
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.16,
|
| 411 |
+
"learning_rate": 9.364444444444445e-06,
|
| 412 |
+
"loss": 0.122,
|
| 413 |
+
"step": 1575
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.16,
|
| 417 |
+
"learning_rate": 9.336666666666666e-06,
|
| 418 |
+
"loss": 0.1313,
|
| 419 |
+
"step": 1600
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.16,
|
| 423 |
+
"learning_rate": 9.30888888888889e-06,
|
| 424 |
+
"loss": 0.1181,
|
| 425 |
+
"step": 1625
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.17,
|
| 429 |
+
"learning_rate": 9.281111111111112e-06,
|
| 430 |
+
"loss": 0.1216,
|
| 431 |
+
"step": 1650
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 0.17,
|
| 435 |
+
"learning_rate": 9.253333333333333e-06,
|
| 436 |
+
"loss": 0.1193,
|
| 437 |
+
"step": 1675
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.17,
|
| 441 |
+
"learning_rate": 9.225555555555556e-06,
|
| 442 |
+
"loss": 0.1128,
|
| 443 |
+
"step": 1700
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.17,
|
| 447 |
+
"learning_rate": 9.197777777777779e-06,
|
| 448 |
+
"loss": 0.1098,
|
| 449 |
+
"step": 1725
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.17,
|
| 453 |
+
"learning_rate": 9.17e-06,
|
| 454 |
+
"loss": 0.1003,
|
| 455 |
+
"step": 1750
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.18,
|
| 459 |
+
"learning_rate": 9.142222222222223e-06,
|
| 460 |
+
"loss": 0.121,
|
| 461 |
+
"step": 1775
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.18,
|
| 465 |
+
"learning_rate": 9.114444444444444e-06,
|
| 466 |
+
"loss": 0.1176,
|
| 467 |
+
"step": 1800
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 0.18,
|
| 471 |
+
"learning_rate": 9.086666666666667e-06,
|
| 472 |
+
"loss": 0.1112,
|
| 473 |
+
"step": 1825
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.18,
|
| 477 |
+
"learning_rate": 9.05888888888889e-06,
|
| 478 |
+
"loss": 0.1078,
|
| 479 |
+
"step": 1850
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.19,
|
| 483 |
+
"learning_rate": 9.031111111111111e-06,
|
| 484 |
+
"loss": 0.1121,
|
| 485 |
+
"step": 1875
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.19,
|
| 489 |
+
"learning_rate": 9.003333333333334e-06,
|
| 490 |
+
"loss": 0.1034,
|
| 491 |
+
"step": 1900
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.19,
|
| 495 |
+
"learning_rate": 8.975555555555557e-06,
|
| 496 |
+
"loss": 0.1108,
|
| 497 |
+
"step": 1925
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.2,
|
| 501 |
+
"learning_rate": 8.947777777777778e-06,
|
| 502 |
+
"loss": 0.1052,
|
| 503 |
+
"step": 1950
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 0.2,
|
| 507 |
+
"learning_rate": 8.920000000000001e-06,
|
| 508 |
+
"loss": 0.0919,
|
| 509 |
+
"step": 1975
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.2,
|
| 513 |
+
"learning_rate": 8.892222222222224e-06,
|
| 514 |
+
"loss": 0.1195,
|
| 515 |
+
"step": 2000
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.2,
|
| 519 |
+
"eval_accuracy": 0.9547,
|
| 520 |
+
"eval_loss": 0.1344193071126938,
|
| 521 |
+
"eval_runtime": 151.3523,
|
| 522 |
+
"eval_samples_per_second": 66.071,
|
| 523 |
+
"eval_steps_per_second": 0.363,
|
| 524 |
+
"step": 2000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.2,
|
| 528 |
+
"learning_rate": 8.864444444444445e-06,
|
| 529 |
+
"loss": 0.0975,
|
| 530 |
+
"step": 2025
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.2,
|
| 534 |
+
"learning_rate": 8.836666666666668e-06,
|
| 535 |
+
"loss": 0.115,
|
| 536 |
+
"step": 2050
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.21,
|
| 540 |
+
"learning_rate": 8.80888888888889e-06,
|
| 541 |
+
"loss": 0.1068,
|
| 542 |
+
"step": 2075
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.21,
|
| 546 |
+
"learning_rate": 8.781111111111112e-06,
|
| 547 |
+
"loss": 0.0911,
|
| 548 |
+
"step": 2100
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.21,
|
| 552 |
+
"learning_rate": 8.753333333333333e-06,
|
| 553 |
+
"loss": 0.1056,
|
| 554 |
+
"step": 2125
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.21,
|
| 558 |
+
"learning_rate": 8.725555555555556e-06,
|
| 559 |
+
"loss": 0.1117,
|
| 560 |
+
"step": 2150
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.22,
|
| 564 |
+
"learning_rate": 8.697777777777779e-06,
|
| 565 |
+
"loss": 0.1086,
|
| 566 |
+
"step": 2175
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.22,
|
| 570 |
+
"learning_rate": 8.67e-06,
|
| 571 |
+
"loss": 0.1084,
|
| 572 |
+
"step": 2200
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.22,
|
| 576 |
+
"learning_rate": 8.642222222222223e-06,
|
| 577 |
+
"loss": 0.116,
|
| 578 |
+
"step": 2225
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.23,
|
| 582 |
+
"learning_rate": 8.614444444444446e-06,
|
| 583 |
+
"loss": 0.0924,
|
| 584 |
+
"step": 2250
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.23,
|
| 588 |
+
"learning_rate": 8.586666666666667e-06,
|
| 589 |
+
"loss": 0.1115,
|
| 590 |
+
"step": 2275
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.23,
|
| 594 |
+
"learning_rate": 8.55888888888889e-06,
|
| 595 |
+
"loss": 0.0984,
|
| 596 |
+
"step": 2300
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.23,
|
| 600 |
+
"learning_rate": 8.531111111111113e-06,
|
| 601 |
+
"loss": 0.101,
|
| 602 |
+
"step": 2325
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.23,
|
| 606 |
+
"learning_rate": 8.503333333333334e-06,
|
| 607 |
+
"loss": 0.0994,
|
| 608 |
+
"step": 2350
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.24,
|
| 612 |
+
"learning_rate": 8.475555555555555e-06,
|
| 613 |
+
"loss": 0.102,
|
| 614 |
+
"step": 2375
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.24,
|
| 618 |
+
"learning_rate": 8.447777777777778e-06,
|
| 619 |
+
"loss": 0.1065,
|
| 620 |
+
"step": 2400
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.24,
|
| 624 |
+
"learning_rate": 8.42e-06,
|
| 625 |
+
"loss": 0.105,
|
| 626 |
+
"step": 2425
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.24,
|
| 630 |
+
"learning_rate": 8.392222222222222e-06,
|
| 631 |
+
"loss": 0.104,
|
| 632 |
+
"step": 2450
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.25,
|
| 636 |
+
"learning_rate": 8.364444444444445e-06,
|
| 637 |
+
"loss": 0.1191,
|
| 638 |
+
"step": 2475
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 0.25,
|
| 642 |
+
"learning_rate": 8.337777777777777e-06,
|
| 643 |
+
"loss": 0.0973,
|
| 644 |
+
"step": 2500
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.25,
|
| 648 |
+
"eval_accuracy": 0.9553,
|
| 649 |
+
"eval_loss": 0.130845844745636,
|
| 650 |
+
"eval_runtime": 152.4264,
|
| 651 |
+
"eval_samples_per_second": 65.605,
|
| 652 |
+
"eval_steps_per_second": 0.361,
|
| 653 |
+
"step": 2500
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.25,
|
| 657 |
+
"learning_rate": 8.31e-06,
|
| 658 |
+
"loss": 0.1062,
|
| 659 |
+
"step": 2525
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 0.26,
|
| 663 |
+
"learning_rate": 8.282222222222223e-06,
|
| 664 |
+
"loss": 0.1026,
|
| 665 |
+
"step": 2550
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 0.26,
|
| 669 |
+
"learning_rate": 8.254444444444444e-06,
|
| 670 |
+
"loss": 0.1072,
|
| 671 |
+
"step": 2575
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 0.26,
|
| 675 |
+
"learning_rate": 8.226666666666667e-06,
|
| 676 |
+
"loss": 0.1015,
|
| 677 |
+
"step": 2600
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 0.26,
|
| 681 |
+
"learning_rate": 8.19888888888889e-06,
|
| 682 |
+
"loss": 0.0997,
|
| 683 |
+
"step": 2625
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.27,
|
| 687 |
+
"learning_rate": 8.171111111111113e-06,
|
| 688 |
+
"loss": 0.1102,
|
| 689 |
+
"step": 2650
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.27,
|
| 693 |
+
"learning_rate": 8.143333333333334e-06,
|
| 694 |
+
"loss": 0.1022,
|
| 695 |
+
"step": 2675
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.27,
|
| 699 |
+
"learning_rate": 8.115555555555557e-06,
|
| 700 |
+
"loss": 0.0958,
|
| 701 |
+
"step": 2700
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 0.27,
|
| 705 |
+
"learning_rate": 8.08777777777778e-06,
|
| 706 |
+
"loss": 0.102,
|
| 707 |
+
"step": 2725
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 0.28,
|
| 711 |
+
"learning_rate": 8.06e-06,
|
| 712 |
+
"loss": 0.1028,
|
| 713 |
+
"step": 2750
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.28,
|
| 717 |
+
"learning_rate": 8.032222222222222e-06,
|
| 718 |
+
"loss": 0.1041,
|
| 719 |
+
"step": 2775
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 0.28,
|
| 723 |
+
"learning_rate": 8.004444444444445e-06,
|
| 724 |
+
"loss": 0.1026,
|
| 725 |
+
"step": 2800
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.28,
|
| 729 |
+
"learning_rate": 7.976666666666668e-06,
|
| 730 |
+
"loss": 0.1158,
|
| 731 |
+
"step": 2825
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.28,
|
| 735 |
+
"learning_rate": 7.948888888888889e-06,
|
| 736 |
+
"loss": 0.1059,
|
| 737 |
+
"step": 2850
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.29,
|
| 741 |
+
"learning_rate": 7.921111111111112e-06,
|
| 742 |
+
"loss": 0.0931,
|
| 743 |
+
"step": 2875
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 0.29,
|
| 747 |
+
"learning_rate": 7.893333333333335e-06,
|
| 748 |
+
"loss": 0.0988,
|
| 749 |
+
"step": 2900
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 0.29,
|
| 753 |
+
"learning_rate": 7.865555555555556e-06,
|
| 754 |
+
"loss": 0.1071,
|
| 755 |
+
"step": 2925
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 0.29,
|
| 759 |
+
"learning_rate": 7.837777777777779e-06,
|
| 760 |
+
"loss": 0.1026,
|
| 761 |
+
"step": 2950
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.3,
|
| 765 |
+
"learning_rate": 7.810000000000001e-06,
|
| 766 |
+
"loss": 0.108,
|
| 767 |
+
"step": 2975
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.3,
|
| 771 |
+
"learning_rate": 7.782222222222223e-06,
|
| 772 |
+
"loss": 0.1039,
|
| 773 |
+
"step": 3000
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.3,
|
| 777 |
+
"eval_accuracy": 0.9479,
|
| 778 |
+
"eval_loss": 0.15836530923843384,
|
| 779 |
+
"eval_runtime": 144.1908,
|
| 780 |
+
"eval_samples_per_second": 69.353,
|
| 781 |
+
"eval_steps_per_second": 0.381,
|
| 782 |
+
"step": 3000
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.3,
|
| 786 |
+
"learning_rate": 7.754444444444445e-06,
|
| 787 |
+
"loss": 0.1099,
|
| 788 |
+
"step": 3025
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.3,
|
| 792 |
+
"learning_rate": 7.726666666666667e-06,
|
| 793 |
+
"loss": 0.1021,
|
| 794 |
+
"step": 3050
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.31,
|
| 798 |
+
"learning_rate": 7.69888888888889e-06,
|
| 799 |
+
"loss": 0.0981,
|
| 800 |
+
"step": 3075
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.31,
|
| 804 |
+
"learning_rate": 7.67111111111111e-06,
|
| 805 |
+
"loss": 0.1085,
|
| 806 |
+
"step": 3100
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.31,
|
| 810 |
+
"learning_rate": 7.643333333333334e-06,
|
| 811 |
+
"loss": 0.0973,
|
| 812 |
+
"step": 3125
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.32,
|
| 816 |
+
"learning_rate": 7.6155555555555564e-06,
|
| 817 |
+
"loss": 0.107,
|
| 818 |
+
"step": 3150
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.32,
|
| 822 |
+
"learning_rate": 7.5877777777777784e-06,
|
| 823 |
+
"loss": 0.1035,
|
| 824 |
+
"step": 3175
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.32,
|
| 828 |
+
"learning_rate": 7.5600000000000005e-06,
|
| 829 |
+
"loss": 0.098,
|
| 830 |
+
"step": 3200
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.32,
|
| 834 |
+
"learning_rate": 7.532222222222223e-06,
|
| 835 |
+
"loss": 0.1066,
|
| 836 |
+
"step": 3225
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.33,
|
| 840 |
+
"learning_rate": 7.504444444444445e-06,
|
| 841 |
+
"loss": 0.0945,
|
| 842 |
+
"step": 3250
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.33,
|
| 846 |
+
"learning_rate": 7.476666666666667e-06,
|
| 847 |
+
"loss": 0.1128,
|
| 848 |
+
"step": 3275
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.33,
|
| 852 |
+
"learning_rate": 7.44888888888889e-06,
|
| 853 |
+
"loss": 0.1143,
|
| 854 |
+
"step": 3300
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.33,
|
| 858 |
+
"learning_rate": 7.421111111111112e-06,
|
| 859 |
+
"loss": 0.1122,
|
| 860 |
+
"step": 3325
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.34,
|
| 864 |
+
"learning_rate": 7.393333333333333e-06,
|
| 865 |
+
"loss": 0.1083,
|
| 866 |
+
"step": 3350
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.34,
|
| 870 |
+
"learning_rate": 7.3655555555555554e-06,
|
| 871 |
+
"loss": 0.108,
|
| 872 |
+
"step": 3375
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.34,
|
| 876 |
+
"learning_rate": 7.337777777777778e-06,
|
| 877 |
+
"loss": 0.1222,
|
| 878 |
+
"step": 3400
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.34,
|
| 882 |
+
"learning_rate": 7.31e-06,
|
| 883 |
+
"loss": 0.1238,
|
| 884 |
+
"step": 3425
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.34,
|
| 888 |
+
"learning_rate": 7.282222222222222e-06,
|
| 889 |
+
"loss": 0.1102,
|
| 890 |
+
"step": 3450
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.35,
|
| 894 |
+
"learning_rate": 7.254444444444445e-06,
|
| 895 |
+
"loss": 0.1293,
|
| 896 |
+
"step": 3475
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.35,
|
| 900 |
+
"learning_rate": 7.226666666666667e-06,
|
| 901 |
+
"loss": 0.1183,
|
| 902 |
+
"step": 3500
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.35,
|
| 906 |
+
"eval_accuracy": 0.952,
|
| 907 |
+
"eval_loss": 0.141157865524292,
|
| 908 |
+
"eval_runtime": 146.5068,
|
| 909 |
+
"eval_samples_per_second": 68.256,
|
| 910 |
+
"eval_steps_per_second": 0.375,
|
| 911 |
+
"step": 3500
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.35,
|
| 915 |
+
"learning_rate": 7.19888888888889e-06,
|
| 916 |
+
"loss": 0.1272,
|
| 917 |
+
"step": 3525
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 0.35,
|
| 921 |
+
"learning_rate": 7.171111111111112e-06,
|
| 922 |
+
"loss": 0.1219,
|
| 923 |
+
"step": 3550
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 0.36,
|
| 927 |
+
"learning_rate": 7.143333333333334e-06,
|
| 928 |
+
"loss": 0.1157,
|
| 929 |
+
"step": 3575
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 0.36,
|
| 933 |
+
"learning_rate": 7.115555555555557e-06,
|
| 934 |
+
"loss": 0.1077,
|
| 935 |
+
"step": 3600
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 0.36,
|
| 939 |
+
"learning_rate": 7.087777777777778e-06,
|
| 940 |
+
"loss": 0.1179,
|
| 941 |
+
"step": 3625
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.36,
|
| 945 |
+
"learning_rate": 7.06e-06,
|
| 946 |
+
"loss": 0.1219,
|
| 947 |
+
"step": 3650
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.37,
|
| 951 |
+
"learning_rate": 7.032222222222222e-06,
|
| 952 |
+
"loss": 0.1155,
|
| 953 |
+
"step": 3675
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 0.37,
|
| 957 |
+
"learning_rate": 7.004444444444445e-06,
|
| 958 |
+
"loss": 0.1293,
|
| 959 |
+
"step": 3700
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 0.37,
|
| 963 |
+
"learning_rate": 6.976666666666667e-06,
|
| 964 |
+
"loss": 0.1277,
|
| 965 |
+
"step": 3725
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 0.38,
|
| 969 |
+
"learning_rate": 6.948888888888889e-06,
|
| 970 |
+
"loss": 0.1191,
|
| 971 |
+
"step": 3750
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.38,
|
| 975 |
+
"learning_rate": 6.921111111111112e-06,
|
| 976 |
+
"loss": 0.1328,
|
| 977 |
+
"step": 3775
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 0.38,
|
| 981 |
+
"learning_rate": 6.893333333333334e-06,
|
| 982 |
+
"loss": 0.1188,
|
| 983 |
+
"step": 3800
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.38,
|
| 987 |
+
"learning_rate": 6.865555555555557e-06,
|
| 988 |
+
"loss": 0.1124,
|
| 989 |
+
"step": 3825
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.39,
|
| 993 |
+
"learning_rate": 6.837777777777779e-06,
|
| 994 |
+
"loss": 0.1272,
|
| 995 |
+
"step": 3850
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 0.39,
|
| 999 |
+
"learning_rate": 6.810000000000001e-06,
|
| 1000 |
+
"loss": 0.1149,
|
| 1001 |
+
"step": 3875
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 0.39,
|
| 1005 |
+
"learning_rate": 6.782222222222222e-06,
|
| 1006 |
+
"loss": 0.1173,
|
| 1007 |
+
"step": 3900
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 0.39,
|
| 1011 |
+
"learning_rate": 6.754444444444445e-06,
|
| 1012 |
+
"loss": 0.1283,
|
| 1013 |
+
"step": 3925
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.4,
|
| 1017 |
+
"learning_rate": 6.726666666666667e-06,
|
| 1018 |
+
"loss": 0.1104,
|
| 1019 |
+
"step": 3950
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 0.4,
|
| 1023 |
+
"learning_rate": 6.698888888888889e-06,
|
| 1024 |
+
"loss": 0.1336,
|
| 1025 |
+
"step": 3975
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.4,
|
| 1029 |
+
"learning_rate": 6.671111111111112e-06,
|
| 1030 |
+
"loss": 0.1103,
|
| 1031 |
+
"step": 4000
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.4,
|
| 1035 |
+
"eval_accuracy": 0.9528,
|
| 1036 |
+
"eval_loss": 0.1352531760931015,
|
| 1037 |
+
"eval_runtime": 148.8877,
|
| 1038 |
+
"eval_samples_per_second": 67.165,
|
| 1039 |
+
"eval_steps_per_second": 0.369,
|
| 1040 |
+
"step": 4000
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.4,
|
| 1044 |
+
"learning_rate": 6.643333333333334e-06,
|
| 1045 |
+
"loss": 0.1158,
|
| 1046 |
+
"step": 4025
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.41,
|
| 1050 |
+
"learning_rate": 6.615555555555556e-06,
|
| 1051 |
+
"loss": 0.1001,
|
| 1052 |
+
"step": 4050
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.41,
|
| 1056 |
+
"learning_rate": 6.587777777777779e-06,
|
| 1057 |
+
"loss": 0.1147,
|
| 1058 |
+
"step": 4075
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 0.41,
|
| 1062 |
+
"learning_rate": 6.560000000000001e-06,
|
| 1063 |
+
"loss": 0.1122,
|
| 1064 |
+
"step": 4100
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 0.41,
|
| 1068 |
+
"learning_rate": 6.532222222222223e-06,
|
| 1069 |
+
"loss": 0.1053,
|
| 1070 |
+
"step": 4125
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.41,
|
| 1074 |
+
"learning_rate": 6.504444444444446e-06,
|
| 1075 |
+
"loss": 0.1126,
|
| 1076 |
+
"step": 4150
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.42,
|
| 1080 |
+
"learning_rate": 6.476666666666667e-06,
|
| 1081 |
+
"loss": 0.1198,
|
| 1082 |
+
"step": 4175
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.42,
|
| 1086 |
+
"learning_rate": 6.448888888888889e-06,
|
| 1087 |
+
"loss": 0.1138,
|
| 1088 |
+
"step": 4200
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.42,
|
| 1092 |
+
"learning_rate": 6.421111111111112e-06,
|
| 1093 |
+
"loss": 0.1145,
|
| 1094 |
+
"step": 4225
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.42,
|
| 1098 |
+
"learning_rate": 6.393333333333334e-06,
|
| 1099 |
+
"loss": 0.1186,
|
| 1100 |
+
"step": 4250
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 0.43,
|
| 1104 |
+
"learning_rate": 6.365555555555556e-06,
|
| 1105 |
+
"loss": 0.1152,
|
| 1106 |
+
"step": 4275
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 0.43,
|
| 1110 |
+
"learning_rate": 6.3377777777777786e-06,
|
| 1111 |
+
"loss": 0.108,
|
| 1112 |
+
"step": 4300
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.43,
|
| 1116 |
+
"learning_rate": 6.3100000000000006e-06,
|
| 1117 |
+
"loss": 0.122,
|
| 1118 |
+
"step": 4325
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.43,
|
| 1122 |
+
"learning_rate": 6.282222222222223e-06,
|
| 1123 |
+
"loss": 0.1036,
|
| 1124 |
+
"step": 4350
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.44,
|
| 1128 |
+
"learning_rate": 6.2544444444444455e-06,
|
| 1129 |
+
"loss": 0.1058,
|
| 1130 |
+
"step": 4375
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.44,
|
| 1134 |
+
"learning_rate": 6.2266666666666675e-06,
|
| 1135 |
+
"loss": 0.1083,
|
| 1136 |
+
"step": 4400
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.44,
|
| 1140 |
+
"learning_rate": 6.1988888888888895e-06,
|
| 1141 |
+
"loss": 0.1252,
|
| 1142 |
+
"step": 4425
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 0.45,
|
| 1146 |
+
"learning_rate": 6.171111111111112e-06,
|
| 1147 |
+
"loss": 0.1181,
|
| 1148 |
+
"step": 4450
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 0.45,
|
| 1152 |
+
"learning_rate": 6.1433333333333335e-06,
|
| 1153 |
+
"loss": 0.1092,
|
| 1154 |
+
"step": 4475
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.45,
|
| 1158 |
+
"learning_rate": 6.1155555555555555e-06,
|
| 1159 |
+
"loss": 0.1084,
|
| 1160 |
+
"step": 4500
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.45,
|
| 1164 |
+
"eval_accuracy": 0.9578,
|
| 1165 |
+
"eval_loss": 0.11692265421152115,
|
| 1166 |
+
"eval_runtime": 147.2838,
|
| 1167 |
+
"eval_samples_per_second": 67.896,
|
| 1168 |
+
"eval_steps_per_second": 0.373,
|
| 1169 |
+
"step": 4500
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 0.45,
|
| 1173 |
+
"learning_rate": 6.087777777777778e-06,
|
| 1174 |
+
"loss": 0.0946,
|
| 1175 |
+
"step": 4525
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.46,
|
| 1179 |
+
"learning_rate": 6.0600000000000004e-06,
|
| 1180 |
+
"loss": 0.1256,
|
| 1181 |
+
"step": 4550
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 0.46,
|
| 1185 |
+
"learning_rate": 6.0322222222222224e-06,
|
| 1186 |
+
"loss": 0.1077,
|
| 1187 |
+
"step": 4575
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 0.46,
|
| 1191 |
+
"learning_rate": 6.004444444444445e-06,
|
| 1192 |
+
"loss": 0.1159,
|
| 1193 |
+
"step": 4600
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.46,
|
| 1197 |
+
"learning_rate": 5.976666666666667e-06,
|
| 1198 |
+
"loss": 0.0976,
|
| 1199 |
+
"step": 4625
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.47,
|
| 1203 |
+
"learning_rate": 5.948888888888889e-06,
|
| 1204 |
+
"loss": 0.1128,
|
| 1205 |
+
"step": 4650
|
| 1206 |
+
},
|
| 1207 |
+
{
|
| 1208 |
+
"epoch": 0.47,
|
| 1209 |
+
"learning_rate": 5.921111111111112e-06,
|
| 1210 |
+
"loss": 0.1148,
|
| 1211 |
+
"step": 4675
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 0.47,
|
| 1215 |
+
"learning_rate": 5.893333333333334e-06,
|
| 1216 |
+
"loss": 0.0886,
|
| 1217 |
+
"step": 4700
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 0.47,
|
| 1221 |
+
"learning_rate": 5.8666666666666675e-06,
|
| 1222 |
+
"loss": 0.1125,
|
| 1223 |
+
"step": 4725
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 0.47,
|
| 1227 |
+
"learning_rate": 5.8388888888888895e-06,
|
| 1228 |
+
"loss": 0.116,
|
| 1229 |
+
"step": 4750
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 0.48,
|
| 1233 |
+
"learning_rate": 5.8111111111111116e-06,
|
| 1234 |
+
"loss": 0.1091,
|
| 1235 |
+
"step": 4775
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.48,
|
| 1239 |
+
"learning_rate": 5.7833333333333344e-06,
|
| 1240 |
+
"loss": 0.1155,
|
| 1241 |
+
"step": 4800
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.48,
|
| 1245 |
+
"learning_rate": 5.7555555555555564e-06,
|
| 1246 |
+
"loss": 0.1104,
|
| 1247 |
+
"step": 4825
|
| 1248 |
+
},
|
| 1249 |
+
{
|
| 1250 |
+
"epoch": 0.48,
|
| 1251 |
+
"learning_rate": 5.7277777777777785e-06,
|
| 1252 |
+
"loss": 0.1037,
|
| 1253 |
+
"step": 4850
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 0.49,
|
| 1257 |
+
"learning_rate": 5.7e-06,
|
| 1258 |
+
"loss": 0.1101,
|
| 1259 |
+
"step": 4875
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 0.49,
|
| 1263 |
+
"learning_rate": 5.6722222222222225e-06,
|
| 1264 |
+
"loss": 0.1039,
|
| 1265 |
+
"step": 4900
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 0.49,
|
| 1269 |
+
"learning_rate": 5.6444444444444445e-06,
|
| 1270 |
+
"loss": 0.109,
|
| 1271 |
+
"step": 4925
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 0.49,
|
| 1275 |
+
"learning_rate": 5.6166666666666665e-06,
|
| 1276 |
+
"loss": 0.1196,
|
| 1277 |
+
"step": 4950
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.5,
|
| 1281 |
+
"learning_rate": 5.588888888888889e-06,
|
| 1282 |
+
"loss": 0.1116,
|
| 1283 |
+
"step": 4975
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.5,
|
| 1287 |
+
"learning_rate": 5.561111111111111e-06,
|
| 1288 |
+
"loss": 0.1041,
|
| 1289 |
+
"step": 5000
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 0.5,
|
| 1293 |
+
"eval_accuracy": 0.952,
|
| 1294 |
+
"eval_loss": 0.1560685634613037,
|
| 1295 |
+
"eval_runtime": 155.1342,
|
| 1296 |
+
"eval_samples_per_second": 64.46,
|
| 1297 |
+
"eval_steps_per_second": 0.355,
|
| 1298 |
+
"step": 5000
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.5,
|
| 1302 |
+
"learning_rate": 5.533333333333334e-06,
|
| 1303 |
+
"loss": 0.1097,
|
| 1304 |
+
"step": 5025
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.51,
|
| 1308 |
+
"learning_rate": 5.505555555555556e-06,
|
| 1309 |
+
"loss": 0.1029,
|
| 1310 |
+
"step": 5050
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 0.51,
|
| 1314 |
+
"learning_rate": 5.477777777777778e-06,
|
| 1315 |
+
"loss": 0.1015,
|
| 1316 |
+
"step": 5075
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 0.51,
|
| 1320 |
+
"learning_rate": 5.450000000000001e-06,
|
| 1321 |
+
"loss": 0.1203,
|
| 1322 |
+
"step": 5100
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.51,
|
| 1326 |
+
"learning_rate": 5.422222222222223e-06,
|
| 1327 |
+
"loss": 0.0967,
|
| 1328 |
+
"step": 5125
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 0.52,
|
| 1332 |
+
"learning_rate": 5.394444444444444e-06,
|
| 1333 |
+
"loss": 0.1082,
|
| 1334 |
+
"step": 5150
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 0.52,
|
| 1338 |
+
"learning_rate": 5.366666666666666e-06,
|
| 1339 |
+
"loss": 0.1175,
|
| 1340 |
+
"step": 5175
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 0.52,
|
| 1344 |
+
"learning_rate": 5.338888888888889e-06,
|
| 1345 |
+
"loss": 0.1043,
|
| 1346 |
+
"step": 5200
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.52,
|
| 1350 |
+
"learning_rate": 5.311111111111111e-06,
|
| 1351 |
+
"loss": 0.1035,
|
| 1352 |
+
"step": 5225
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 0.53,
|
| 1356 |
+
"learning_rate": 5.283333333333333e-06,
|
| 1357 |
+
"loss": 0.1043,
|
| 1358 |
+
"step": 5250
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 0.53,
|
| 1362 |
+
"learning_rate": 5.255555555555556e-06,
|
| 1363 |
+
"loss": 0.1002,
|
| 1364 |
+
"step": 5275
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.53,
|
| 1368 |
+
"learning_rate": 5.227777777777778e-06,
|
| 1369 |
+
"loss": 0.0964,
|
| 1370 |
+
"step": 5300
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 0.53,
|
| 1374 |
+
"learning_rate": 5.2e-06,
|
| 1375 |
+
"loss": 0.1069,
|
| 1376 |
+
"step": 5325
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 0.54,
|
| 1380 |
+
"learning_rate": 5.172222222222223e-06,
|
| 1381 |
+
"loss": 0.1089,
|
| 1382 |
+
"step": 5350
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 0.54,
|
| 1386 |
+
"learning_rate": 5.144444444444445e-06,
|
| 1387 |
+
"loss": 0.1092,
|
| 1388 |
+
"step": 5375
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.54,
|
| 1392 |
+
"learning_rate": 5.116666666666668e-06,
|
| 1393 |
+
"loss": 0.1026,
|
| 1394 |
+
"step": 5400
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 0.54,
|
| 1398 |
+
"learning_rate": 5.088888888888889e-06,
|
| 1399 |
+
"loss": 0.0982,
|
| 1400 |
+
"step": 5425
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 0.55,
|
| 1404 |
+
"learning_rate": 5.061111111111111e-06,
|
| 1405 |
+
"loss": 0.0983,
|
| 1406 |
+
"step": 5450
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.55,
|
| 1410 |
+
"learning_rate": 5.033333333333333e-06,
|
| 1411 |
+
"loss": 0.0978,
|
| 1412 |
+
"step": 5475
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 0.55,
|
| 1416 |
+
"learning_rate": 5.005555555555556e-06,
|
| 1417 |
+
"loss": 0.0966,
|
| 1418 |
+
"step": 5500
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 0.55,
|
| 1422 |
+
"eval_accuracy": 0.9549,
|
| 1423 |
+
"eval_loss": 0.12704403698444366,
|
| 1424 |
+
"eval_runtime": 152.5968,
|
| 1425 |
+
"eval_samples_per_second": 65.532,
|
| 1426 |
+
"eval_steps_per_second": 0.36,
|
| 1427 |
+
"step": 5500
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.55,
|
| 1431 |
+
"learning_rate": 4.977777777777778e-06,
|
| 1432 |
+
"loss": 0.1094,
|
| 1433 |
+
"step": 5525
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 0.56,
|
| 1437 |
+
"learning_rate": 4.95e-06,
|
| 1438 |
+
"loss": 0.1042,
|
| 1439 |
+
"step": 5550
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 0.56,
|
| 1443 |
+
"learning_rate": 4.922222222222223e-06,
|
| 1444 |
+
"loss": 0.1,
|
| 1445 |
+
"step": 5575
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 0.56,
|
| 1449 |
+
"learning_rate": 4.894444444444445e-06,
|
| 1450 |
+
"loss": 0.1079,
|
| 1451 |
+
"step": 5600
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.56,
|
| 1455 |
+
"learning_rate": 4.866666666666667e-06,
|
| 1456 |
+
"loss": 0.1017,
|
| 1457 |
+
"step": 5625
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 0.56,
|
| 1461 |
+
"learning_rate": 4.838888888888889e-06,
|
| 1462 |
+
"loss": 0.0953,
|
| 1463 |
+
"step": 5650
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 0.57,
|
| 1467 |
+
"learning_rate": 4.811111111111111e-06,
|
| 1468 |
+
"loss": 0.094,
|
| 1469 |
+
"step": 5675
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 0.57,
|
| 1473 |
+
"learning_rate": 4.783333333333334e-06,
|
| 1474 |
+
"loss": 0.0966,
|
| 1475 |
+
"step": 5700
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 0.57,
|
| 1479 |
+
"learning_rate": 4.755555555555556e-06,
|
| 1480 |
+
"loss": 0.1056,
|
| 1481 |
+
"step": 5725
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 0.57,
|
| 1485 |
+
"learning_rate": 4.727777777777779e-06,
|
| 1486 |
+
"loss": 0.104,
|
| 1487 |
+
"step": 5750
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 0.58,
|
| 1491 |
+
"learning_rate": 4.7e-06,
|
| 1492 |
+
"loss": 0.0862,
|
| 1493 |
+
"step": 5775
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.58,
|
| 1497 |
+
"learning_rate": 4.672222222222223e-06,
|
| 1498 |
+
"loss": 0.1007,
|
| 1499 |
+
"step": 5800
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 0.58,
|
| 1503 |
+
"learning_rate": 4.644444444444445e-06,
|
| 1504 |
+
"loss": 0.1061,
|
| 1505 |
+
"step": 5825
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 0.58,
|
| 1509 |
+
"learning_rate": 4.616666666666667e-06,
|
| 1510 |
+
"loss": 0.0974,
|
| 1511 |
+
"step": 5850
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 0.59,
|
| 1515 |
+
"learning_rate": 4.58888888888889e-06,
|
| 1516 |
+
"loss": 0.097,
|
| 1517 |
+
"step": 5875
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 0.59,
|
| 1521 |
+
"learning_rate": 4.561111111111112e-06,
|
| 1522 |
+
"loss": 0.1002,
|
| 1523 |
+
"step": 5900
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 0.59,
|
| 1527 |
+
"learning_rate": 4.533333333333334e-06,
|
| 1528 |
+
"loss": 0.1069,
|
| 1529 |
+
"step": 5925
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 0.59,
|
| 1533 |
+
"learning_rate": 4.505555555555556e-06,
|
| 1534 |
+
"loss": 0.1128,
|
| 1535 |
+
"step": 5950
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.6,
|
| 1539 |
+
"learning_rate": 4.477777777777778e-06,
|
| 1540 |
+
"loss": 0.0957,
|
| 1541 |
+
"step": 5975
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 0.6,
|
| 1545 |
+
"learning_rate": 4.450000000000001e-06,
|
| 1546 |
+
"loss": 0.1,
|
| 1547 |
+
"step": 6000
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 0.6,
|
| 1551 |
+
"eval_accuracy": 0.9594,
|
| 1552 |
+
"eval_loss": 0.11799036711454391,
|
| 1553 |
+
"eval_runtime": 145.5771,
|
| 1554 |
+
"eval_samples_per_second": 68.692,
|
| 1555 |
+
"eval_steps_per_second": 0.378,
|
| 1556 |
+
"step": 6000
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.6,
|
| 1560 |
+
"learning_rate": 4.422222222222223e-06,
|
| 1561 |
+
"loss": 0.098,
|
| 1562 |
+
"step": 6025
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 0.6,
|
| 1566 |
+
"learning_rate": 4.3944444444444455e-06,
|
| 1567 |
+
"loss": 0.098,
|
| 1568 |
+
"step": 6050
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 0.61,
|
| 1572 |
+
"learning_rate": 4.366666666666667e-06,
|
| 1573 |
+
"loss": 0.1047,
|
| 1574 |
+
"step": 6075
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 0.61,
|
| 1578 |
+
"learning_rate": 4.3388888888888895e-06,
|
| 1579 |
+
"loss": 0.0998,
|
| 1580 |
+
"step": 6100
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 0.61,
|
| 1584 |
+
"learning_rate": 4.3111111111111115e-06,
|
| 1585 |
+
"loss": 0.115,
|
| 1586 |
+
"step": 6125
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 0.61,
|
| 1590 |
+
"learning_rate": 4.2833333333333335e-06,
|
| 1591 |
+
"loss": 0.0965,
|
| 1592 |
+
"step": 6150
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 0.62,
|
| 1596 |
+
"learning_rate": 4.255555555555556e-06,
|
| 1597 |
+
"loss": 0.1031,
|
| 1598 |
+
"step": 6175
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.62,
|
| 1602 |
+
"learning_rate": 4.227777777777778e-06,
|
| 1603 |
+
"loss": 0.1105,
|
| 1604 |
+
"step": 6200
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 0.62,
|
| 1608 |
+
"learning_rate": 4.2000000000000004e-06,
|
| 1609 |
+
"loss": 0.105,
|
| 1610 |
+
"step": 6225
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 0.62,
|
| 1614 |
+
"learning_rate": 4.1722222222222225e-06,
|
| 1615 |
+
"loss": 0.1022,
|
| 1616 |
+
"step": 6250
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 0.63,
|
| 1620 |
+
"learning_rate": 4.1444444444444445e-06,
|
| 1621 |
+
"loss": 0.0991,
|
| 1622 |
+
"step": 6275
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 0.63,
|
| 1626 |
+
"learning_rate": 4.116666666666667e-06,
|
| 1627 |
+
"loss": 0.1004,
|
| 1628 |
+
"step": 6300
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 0.63,
|
| 1632 |
+
"learning_rate": 4.088888888888889e-06,
|
| 1633 |
+
"loss": 0.1009,
|
| 1634 |
+
"step": 6325
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 0.64,
|
| 1638 |
+
"learning_rate": 4.061111111111111e-06,
|
| 1639 |
+
"loss": 0.1041,
|
| 1640 |
+
"step": 6350
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.64,
|
| 1644 |
+
"learning_rate": 4.033333333333333e-06,
|
| 1645 |
+
"loss": 0.091,
|
| 1646 |
+
"step": 6375
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 0.64,
|
| 1650 |
+
"learning_rate": 4.005555555555556e-06,
|
| 1651 |
+
"loss": 0.1042,
|
| 1652 |
+
"step": 6400
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 0.64,
|
| 1656 |
+
"learning_rate": 3.977777777777778e-06,
|
| 1657 |
+
"loss": 0.0923,
|
| 1658 |
+
"step": 6425
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 0.65,
|
| 1662 |
+
"learning_rate": 3.95e-06,
|
| 1663 |
+
"loss": 0.0957,
|
| 1664 |
+
"step": 6450
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 0.65,
|
| 1668 |
+
"learning_rate": 3.922222222222223e-06,
|
| 1669 |
+
"loss": 0.0927,
|
| 1670 |
+
"step": 6475
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 0.65,
|
| 1674 |
+
"learning_rate": 3.894444444444444e-06,
|
| 1675 |
+
"loss": 0.0965,
|
| 1676 |
+
"step": 6500
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.65,
|
| 1680 |
+
"eval_accuracy": 0.9551,
|
| 1681 |
+
"eval_loss": 0.12714883685112,
|
| 1682 |
+
"eval_runtime": 146.7045,
|
| 1683 |
+
"eval_samples_per_second": 68.164,
|
| 1684 |
+
"eval_steps_per_second": 0.375,
|
| 1685 |
+
"step": 6500
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 0.65,
|
| 1689 |
+
"learning_rate": 3.866666666666667e-06,
|
| 1690 |
+
"loss": 0.1084,
|
| 1691 |
+
"step": 6525
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 0.66,
|
| 1695 |
+
"learning_rate": 3.838888888888889e-06,
|
| 1696 |
+
"loss": 0.1,
|
| 1697 |
+
"step": 6550
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.66,
|
| 1701 |
+
"learning_rate": 3.8111111111111117e-06,
|
| 1702 |
+
"loss": 0.0976,
|
| 1703 |
+
"step": 6575
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.66,
|
| 1707 |
+
"learning_rate": 3.7833333333333337e-06,
|
| 1708 |
+
"loss": 0.0974,
|
| 1709 |
+
"step": 6600
|
| 1710 |
+
},
|
| 1711 |
+
{
|
| 1712 |
+
"epoch": 0.66,
|
| 1713 |
+
"learning_rate": 3.7555555555555557e-06,
|
| 1714 |
+
"loss": 0.0927,
|
| 1715 |
+
"step": 6625
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 0.67,
|
| 1719 |
+
"learning_rate": 3.727777777777778e-06,
|
| 1720 |
+
"loss": 0.1042,
|
| 1721 |
+
"step": 6650
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 0.67,
|
| 1725 |
+
"learning_rate": 3.7e-06,
|
| 1726 |
+
"loss": 0.0954,
|
| 1727 |
+
"step": 6675
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 0.67,
|
| 1731 |
+
"learning_rate": 3.6722222222222226e-06,
|
| 1732 |
+
"loss": 0.0904,
|
| 1733 |
+
"step": 6700
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 0.67,
|
| 1737 |
+
"learning_rate": 3.644444444444445e-06,
|
| 1738 |
+
"loss": 0.0976,
|
| 1739 |
+
"step": 6725
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.68,
|
| 1743 |
+
"learning_rate": 3.616666666666667e-06,
|
| 1744 |
+
"loss": 0.0978,
|
| 1745 |
+
"step": 6750
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.68,
|
| 1749 |
+
"learning_rate": 3.588888888888889e-06,
|
| 1750 |
+
"loss": 0.0975,
|
| 1751 |
+
"step": 6775
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 0.68,
|
| 1755 |
+
"learning_rate": 3.561111111111111e-06,
|
| 1756 |
+
"loss": 0.1042,
|
| 1757 |
+
"step": 6800
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 0.68,
|
| 1761 |
+
"learning_rate": 3.5333333333333335e-06,
|
| 1762 |
+
"loss": 0.0997,
|
| 1763 |
+
"step": 6825
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 0.69,
|
| 1767 |
+
"learning_rate": 3.505555555555556e-06,
|
| 1768 |
+
"loss": 0.0996,
|
| 1769 |
+
"step": 6850
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 0.69,
|
| 1773 |
+
"learning_rate": 3.4777777777777784e-06,
|
| 1774 |
+
"loss": 0.1028,
|
| 1775 |
+
"step": 6875
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 0.69,
|
| 1779 |
+
"learning_rate": 3.45e-06,
|
| 1780 |
+
"loss": 0.0917,
|
| 1781 |
+
"step": 6900
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 0.69,
|
| 1785 |
+
"learning_rate": 3.4222222222222224e-06,
|
| 1786 |
+
"loss": 0.0857,
|
| 1787 |
+
"step": 6925
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.69,
|
| 1791 |
+
"learning_rate": 3.3944444444444445e-06,
|
| 1792 |
+
"loss": 0.0913,
|
| 1793 |
+
"step": 6950
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 0.7,
|
| 1797 |
+
"learning_rate": 3.366666666666667e-06,
|
| 1798 |
+
"loss": 0.0885,
|
| 1799 |
+
"step": 6975
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 0.7,
|
| 1803 |
+
"learning_rate": 3.3388888888888893e-06,
|
| 1804 |
+
"loss": 0.1008,
|
| 1805 |
+
"step": 7000
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 0.7,
|
| 1809 |
+
"eval_accuracy": 0.9545,
|
| 1810 |
+
"eval_loss": 0.12300202995538712,
|
| 1811 |
+
"eval_runtime": 180.9152,
|
| 1812 |
+
"eval_samples_per_second": 55.275,
|
| 1813 |
+
"eval_steps_per_second": 0.304,
|
| 1814 |
+
"step": 7000
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 0.7,
|
| 1818 |
+
"learning_rate": 3.3111111111111118e-06,
|
| 1819 |
+
"loss": 0.1016,
|
| 1820 |
+
"step": 7025
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 0.7,
|
| 1824 |
+
"learning_rate": 3.2833333333333334e-06,
|
| 1825 |
+
"loss": 0.0993,
|
| 1826 |
+
"step": 7050
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 0.71,
|
| 1830 |
+
"learning_rate": 3.255555555555556e-06,
|
| 1831 |
+
"loss": 0.0938,
|
| 1832 |
+
"step": 7075
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 0.71,
|
| 1836 |
+
"learning_rate": 3.227777777777778e-06,
|
| 1837 |
+
"loss": 0.0941,
|
| 1838 |
+
"step": 7100
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 0.71,
|
| 1842 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 1843 |
+
"loss": 0.0906,
|
| 1844 |
+
"step": 7125
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 0.71,
|
| 1848 |
+
"learning_rate": 3.1722222222222227e-06,
|
| 1849 |
+
"loss": 0.0857,
|
| 1850 |
+
"step": 7150
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.72,
|
| 1854 |
+
"learning_rate": 3.144444444444445e-06,
|
| 1855 |
+
"loss": 0.0876,
|
| 1856 |
+
"step": 7175
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 0.72,
|
| 1860 |
+
"learning_rate": 3.1166666666666668e-06,
|
| 1861 |
+
"loss": 0.089,
|
| 1862 |
+
"step": 7200
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 0.72,
|
| 1866 |
+
"learning_rate": 3.088888888888889e-06,
|
| 1867 |
+
"loss": 0.0919,
|
| 1868 |
+
"step": 7225
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 0.72,
|
| 1872 |
+
"learning_rate": 3.0611111111111112e-06,
|
| 1873 |
+
"loss": 0.0887,
|
| 1874 |
+
"step": 7250
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 0.73,
|
| 1878 |
+
"learning_rate": 3.0333333333333337e-06,
|
| 1879 |
+
"loss": 0.0882,
|
| 1880 |
+
"step": 7275
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 0.73,
|
| 1884 |
+
"learning_rate": 3.005555555555556e-06,
|
| 1885 |
+
"loss": 0.0915,
|
| 1886 |
+
"step": 7300
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 0.73,
|
| 1890 |
+
"learning_rate": 2.9777777777777777e-06,
|
| 1891 |
+
"loss": 0.0835,
|
| 1892 |
+
"step": 7325
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.73,
|
| 1896 |
+
"learning_rate": 2.95e-06,
|
| 1897 |
+
"loss": 0.0909,
|
| 1898 |
+
"step": 7350
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 0.74,
|
| 1902 |
+
"learning_rate": 2.9222222222222226e-06,
|
| 1903 |
+
"loss": 0.0934,
|
| 1904 |
+
"step": 7375
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 0.74,
|
| 1908 |
+
"learning_rate": 2.8944444444444446e-06,
|
| 1909 |
+
"loss": 0.0941,
|
| 1910 |
+
"step": 7400
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 0.74,
|
| 1914 |
+
"learning_rate": 2.866666666666667e-06,
|
| 1915 |
+
"loss": 0.0885,
|
| 1916 |
+
"step": 7425
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 0.74,
|
| 1920 |
+
"learning_rate": 2.8388888888888895e-06,
|
| 1921 |
+
"loss": 0.0843,
|
| 1922 |
+
"step": 7450
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 0.75,
|
| 1926 |
+
"learning_rate": 2.811111111111111e-06,
|
| 1927 |
+
"loss": 0.0894,
|
| 1928 |
+
"step": 7475
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 0.75,
|
| 1932 |
+
"learning_rate": 2.7833333333333335e-06,
|
| 1933 |
+
"loss": 0.1005,
|
| 1934 |
+
"step": 7500
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.75,
|
| 1938 |
+
"eval_accuracy": 0.9566,
|
| 1939 |
+
"eval_loss": 0.1161317378282547,
|
| 1940 |
+
"eval_runtime": 156.7623,
|
| 1941 |
+
"eval_samples_per_second": 63.791,
|
| 1942 |
+
"eval_steps_per_second": 0.351,
|
| 1943 |
+
"step": 7500
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 0.75,
|
| 1947 |
+
"learning_rate": 2.755555555555556e-06,
|
| 1948 |
+
"loss": 0.0928,
|
| 1949 |
+
"step": 7525
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 0.76,
|
| 1953 |
+
"learning_rate": 2.727777777777778e-06,
|
| 1954 |
+
"loss": 0.0872,
|
| 1955 |
+
"step": 7550
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.76,
|
| 1959 |
+
"learning_rate": 2.7000000000000004e-06,
|
| 1960 |
+
"loss": 0.0908,
|
| 1961 |
+
"step": 7575
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 0.76,
|
| 1965 |
+
"learning_rate": 2.672222222222223e-06,
|
| 1966 |
+
"loss": 0.0889,
|
| 1967 |
+
"step": 7600
|
| 1968 |
+
},
|
| 1969 |
+
{
|
| 1970 |
+
"epoch": 0.76,
|
| 1971 |
+
"learning_rate": 2.6444444444444444e-06,
|
| 1972 |
+
"loss": 0.0902,
|
| 1973 |
+
"step": 7625
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 0.77,
|
| 1977 |
+
"learning_rate": 2.616666666666667e-06,
|
| 1978 |
+
"loss": 0.0907,
|
| 1979 |
+
"step": 7650
|
| 1980 |
+
},
|
| 1981 |
+
{
|
| 1982 |
+
"epoch": 0.77,
|
| 1983 |
+
"learning_rate": 2.5888888888888893e-06,
|
| 1984 |
+
"loss": 0.0853,
|
| 1985 |
+
"step": 7675
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 0.77,
|
| 1989 |
+
"learning_rate": 2.5611111111111113e-06,
|
| 1990 |
+
"loss": 0.1042,
|
| 1991 |
+
"step": 7700
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 0.77,
|
| 1995 |
+
"learning_rate": 2.5333333333333338e-06,
|
| 1996 |
+
"loss": 0.0931,
|
| 1997 |
+
"step": 7725
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.78,
|
| 2001 |
+
"learning_rate": 2.5055555555555554e-06,
|
| 2002 |
+
"loss": 0.0949,
|
| 2003 |
+
"step": 7750
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 0.78,
|
| 2007 |
+
"learning_rate": 2.4777777777777782e-06,
|
| 2008 |
+
"loss": 0.0746,
|
| 2009 |
+
"step": 7775
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 0.78,
|
| 2013 |
+
"learning_rate": 2.4500000000000003e-06,
|
| 2014 |
+
"loss": 0.0874,
|
| 2015 |
+
"step": 7800
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 0.78,
|
| 2019 |
+
"learning_rate": 2.4222222222222223e-06,
|
| 2020 |
+
"loss": 0.0856,
|
| 2021 |
+
"step": 7825
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 0.79,
|
| 2025 |
+
"learning_rate": 2.3944444444444447e-06,
|
| 2026 |
+
"loss": 0.0802,
|
| 2027 |
+
"step": 7850
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 0.79,
|
| 2031 |
+
"learning_rate": 2.3666666666666667e-06,
|
| 2032 |
+
"loss": 0.0711,
|
| 2033 |
+
"step": 7875
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 0.79,
|
| 2037 |
+
"learning_rate": 2.338888888888889e-06,
|
| 2038 |
+
"loss": 0.0828,
|
| 2039 |
+
"step": 7900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.79,
|
| 2043 |
+
"learning_rate": 2.311111111111111e-06,
|
| 2044 |
+
"loss": 0.0852,
|
| 2045 |
+
"step": 7925
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 0.8,
|
| 2049 |
+
"learning_rate": 2.2833333333333336e-06,
|
| 2050 |
+
"loss": 0.0823,
|
| 2051 |
+
"step": 7950
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 0.8,
|
| 2055 |
+
"learning_rate": 2.2555555555555557e-06,
|
| 2056 |
+
"loss": 0.0853,
|
| 2057 |
+
"step": 7975
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 0.8,
|
| 2061 |
+
"learning_rate": 2.2277777777777777e-06,
|
| 2062 |
+
"loss": 0.0868,
|
| 2063 |
+
"step": 8000
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 0.8,
|
| 2067 |
+
"eval_accuracy": 0.9543,
|
| 2068 |
+
"eval_loss": 0.12787339091300964,
|
| 2069 |
+
"eval_runtime": 146.1122,
|
| 2070 |
+
"eval_samples_per_second": 68.441,
|
| 2071 |
+
"eval_steps_per_second": 0.376,
|
| 2072 |
+
"step": 8000
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 0.8,
|
| 2076 |
+
"learning_rate": 2.2e-06,
|
| 2077 |
+
"loss": 0.0888,
|
| 2078 |
+
"step": 8025
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 0.81,
|
| 2082 |
+
"learning_rate": 2.1722222222222226e-06,
|
| 2083 |
+
"loss": 0.089,
|
| 2084 |
+
"step": 8050
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 0.81,
|
| 2088 |
+
"learning_rate": 2.1444444444444446e-06,
|
| 2089 |
+
"loss": 0.0798,
|
| 2090 |
+
"step": 8075
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 0.81,
|
| 2094 |
+
"learning_rate": 2.116666666666667e-06,
|
| 2095 |
+
"loss": 0.0798,
|
| 2096 |
+
"step": 8100
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 0.81,
|
| 2100 |
+
"learning_rate": 2.088888888888889e-06,
|
| 2101 |
+
"loss": 0.0783,
|
| 2102 |
+
"step": 8125
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.81,
|
| 2106 |
+
"learning_rate": 2.061111111111111e-06,
|
| 2107 |
+
"loss": 0.0872,
|
| 2108 |
+
"step": 8150
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 0.82,
|
| 2112 |
+
"learning_rate": 2.0333333333333335e-06,
|
| 2113 |
+
"loss": 0.0756,
|
| 2114 |
+
"step": 8175
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 0.82,
|
| 2118 |
+
"learning_rate": 2.0055555555555555e-06,
|
| 2119 |
+
"loss": 0.0891,
|
| 2120 |
+
"step": 8200
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 0.82,
|
| 2124 |
+
"learning_rate": 1.977777777777778e-06,
|
| 2125 |
+
"loss": 0.0878,
|
| 2126 |
+
"step": 8225
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 0.82,
|
| 2130 |
+
"learning_rate": 1.9500000000000004e-06,
|
| 2131 |
+
"loss": 0.0965,
|
| 2132 |
+
"step": 8250
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 0.83,
|
| 2136 |
+
"learning_rate": 1.9222222222222224e-06,
|
| 2137 |
+
"loss": 0.0844,
|
| 2138 |
+
"step": 8275
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 0.83,
|
| 2142 |
+
"learning_rate": 1.8944444444444446e-06,
|
| 2143 |
+
"loss": 0.1039,
|
| 2144 |
+
"step": 8300
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.83,
|
| 2148 |
+
"learning_rate": 1.8666666666666669e-06,
|
| 2149 |
+
"loss": 0.0919,
|
| 2150 |
+
"step": 8325
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 0.83,
|
| 2154 |
+
"learning_rate": 1.8388888888888889e-06,
|
| 2155 |
+
"loss": 0.0914,
|
| 2156 |
+
"step": 8350
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 0.84,
|
| 2160 |
+
"learning_rate": 1.8111111111111113e-06,
|
| 2161 |
+
"loss": 0.1006,
|
| 2162 |
+
"step": 8375
|
| 2163 |
+
},
|
| 2164 |
+
{
|
| 2165 |
+
"epoch": 0.84,
|
| 2166 |
+
"learning_rate": 1.7833333333333336e-06,
|
| 2167 |
+
"loss": 0.0891,
|
| 2168 |
+
"step": 8400
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 0.84,
|
| 2172 |
+
"learning_rate": 1.7555555555555556e-06,
|
| 2173 |
+
"loss": 0.0945,
|
| 2174 |
+
"step": 8425
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 0.84,
|
| 2178 |
+
"learning_rate": 1.727777777777778e-06,
|
| 2179 |
+
"loss": 0.0798,
|
| 2180 |
+
"step": 8450
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 0.85,
|
| 2184 |
+
"learning_rate": 1.7000000000000002e-06,
|
| 2185 |
+
"loss": 0.0888,
|
| 2186 |
+
"step": 8475
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.85,
|
| 2190 |
+
"learning_rate": 1.6722222222222223e-06,
|
| 2191 |
+
"loss": 0.0747,
|
| 2192 |
+
"step": 8500
|
| 2193 |
+
},
|
| 2194 |
+
{
|
| 2195 |
+
"epoch": 0.85,
|
| 2196 |
+
"eval_accuracy": 0.9578,
|
| 2197 |
+
"eval_loss": 0.12338610738515854,
|
| 2198 |
+
"eval_runtime": 146.8921,
|
| 2199 |
+
"eval_samples_per_second": 68.077,
|
| 2200 |
+
"eval_steps_per_second": 0.374,
|
| 2201 |
+
"step": 8500
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 0.85,
|
| 2205 |
+
"learning_rate": 1.6444444444444447e-06,
|
| 2206 |
+
"loss": 0.084,
|
| 2207 |
+
"step": 8525
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.85,
|
| 2211 |
+
"learning_rate": 1.6166666666666667e-06,
|
| 2212 |
+
"loss": 0.0764,
|
| 2213 |
+
"step": 8550
|
| 2214 |
+
},
|
| 2215 |
+
{
|
| 2216 |
+
"epoch": 0.86,
|
| 2217 |
+
"learning_rate": 1.588888888888889e-06,
|
| 2218 |
+
"loss": 0.082,
|
| 2219 |
+
"step": 8575
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"epoch": 0.86,
|
| 2223 |
+
"learning_rate": 1.5611111111111114e-06,
|
| 2224 |
+
"loss": 0.0892,
|
| 2225 |
+
"step": 8600
|
| 2226 |
+
},
|
| 2227 |
+
{
|
| 2228 |
+
"epoch": 0.86,
|
| 2229 |
+
"learning_rate": 1.5333333333333334e-06,
|
| 2230 |
+
"loss": 0.0853,
|
| 2231 |
+
"step": 8625
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 0.86,
|
| 2235 |
+
"learning_rate": 1.5055555555555556e-06,
|
| 2236 |
+
"loss": 0.0833,
|
| 2237 |
+
"step": 8650
|
| 2238 |
+
},
|
| 2239 |
+
{
|
| 2240 |
+
"epoch": 0.87,
|
| 2241 |
+
"learning_rate": 1.477777777777778e-06,
|
| 2242 |
+
"loss": 0.0917,
|
| 2243 |
+
"step": 8675
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 0.87,
|
| 2247 |
+
"learning_rate": 1.45e-06,
|
| 2248 |
+
"loss": 0.0783,
|
| 2249 |
+
"step": 8700
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.87,
|
| 2253 |
+
"learning_rate": 1.4222222222222223e-06,
|
| 2254 |
+
"loss": 0.0959,
|
| 2255 |
+
"step": 8725
|
| 2256 |
+
},
|
| 2257 |
+
{
|
| 2258 |
+
"epoch": 0.88,
|
| 2259 |
+
"learning_rate": 1.3944444444444446e-06,
|
| 2260 |
+
"loss": 0.0758,
|
| 2261 |
+
"step": 8750
|
| 2262 |
+
},
|
| 2263 |
+
{
|
| 2264 |
+
"epoch": 0.88,
|
| 2265 |
+
"learning_rate": 1.3666666666666668e-06,
|
| 2266 |
+
"loss": 0.0806,
|
| 2267 |
+
"step": 8775
|
| 2268 |
+
},
|
| 2269 |
+
{
|
| 2270 |
+
"epoch": 0.88,
|
| 2271 |
+
"learning_rate": 1.338888888888889e-06,
|
| 2272 |
+
"loss": 0.0808,
|
| 2273 |
+
"step": 8800
|
| 2274 |
+
},
|
| 2275 |
+
{
|
| 2276 |
+
"epoch": 0.88,
|
| 2277 |
+
"learning_rate": 1.3111111111111112e-06,
|
| 2278 |
+
"loss": 0.0808,
|
| 2279 |
+
"step": 8825
|
| 2280 |
+
},
|
| 2281 |
+
{
|
| 2282 |
+
"epoch": 0.89,
|
| 2283 |
+
"learning_rate": 1.2833333333333335e-06,
|
| 2284 |
+
"loss": 0.0805,
|
| 2285 |
+
"step": 8850
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 0.89,
|
| 2289 |
+
"learning_rate": 1.2555555555555557e-06,
|
| 2290 |
+
"loss": 0.0822,
|
| 2291 |
+
"step": 8875
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.89,
|
| 2295 |
+
"learning_rate": 1.227777777777778e-06,
|
| 2296 |
+
"loss": 0.0861,
|
| 2297 |
+
"step": 8900
|
| 2298 |
+
},
|
| 2299 |
+
{
|
| 2300 |
+
"epoch": 0.89,
|
| 2301 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 2302 |
+
"loss": 0.094,
|
| 2303 |
+
"step": 8925
|
| 2304 |
+
},
|
| 2305 |
+
{
|
| 2306 |
+
"epoch": 0.9,
|
| 2307 |
+
"learning_rate": 1.1722222222222224e-06,
|
| 2308 |
+
"loss": 0.0798,
|
| 2309 |
+
"step": 8950
|
| 2310 |
+
},
|
| 2311 |
+
{
|
| 2312 |
+
"epoch": 0.9,
|
| 2313 |
+
"learning_rate": 1.1444444444444446e-06,
|
| 2314 |
+
"loss": 0.0881,
|
| 2315 |
+
"step": 8975
|
| 2316 |
+
},
|
| 2317 |
+
{
|
| 2318 |
+
"epoch": 0.9,
|
| 2319 |
+
"learning_rate": 1.1166666666666666e-06,
|
| 2320 |
+
"loss": 0.1015,
|
| 2321 |
+
"step": 9000
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 0.9,
|
| 2325 |
+
"eval_accuracy": 0.9578,
|
| 2326 |
+
"eval_loss": 0.11542944610118866,
|
| 2327 |
+
"eval_runtime": 146.1261,
|
| 2328 |
+
"eval_samples_per_second": 68.434,
|
| 2329 |
+
"eval_steps_per_second": 0.376,
|
| 2330 |
+
"step": 9000
|
| 2331 |
+
},
|
| 2332 |
+
{
|
| 2333 |
+
"epoch": 0.9,
|
| 2334 |
+
"learning_rate": 1.0888888888888889e-06,
|
| 2335 |
+
"loss": 0.0892,
|
| 2336 |
+
"step": 9025
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 0.91,
|
| 2340 |
+
"learning_rate": 1.0611111111111113e-06,
|
| 2341 |
+
"loss": 0.0888,
|
| 2342 |
+
"step": 9050
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 0.91,
|
| 2346 |
+
"learning_rate": 1.0333333333333333e-06,
|
| 2347 |
+
"loss": 0.0877,
|
| 2348 |
+
"step": 9075
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 0.91,
|
| 2352 |
+
"learning_rate": 1.0055555555555556e-06,
|
| 2353 |
+
"loss": 0.1019,
|
| 2354 |
+
"step": 9100
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.91,
|
| 2358 |
+
"learning_rate": 9.77777777777778e-07,
|
| 2359 |
+
"loss": 0.0959,
|
| 2360 |
+
"step": 9125
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"epoch": 0.92,
|
| 2364 |
+
"learning_rate": 9.500000000000001e-07,
|
| 2365 |
+
"loss": 0.0919,
|
| 2366 |
+
"step": 9150
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 0.92,
|
| 2370 |
+
"learning_rate": 9.222222222222222e-07,
|
| 2371 |
+
"loss": 0.0957,
|
| 2372 |
+
"step": 9175
|
| 2373 |
+
},
|
| 2374 |
+
{
|
| 2375 |
+
"epoch": 0.92,
|
| 2376 |
+
"learning_rate": 8.944444444444445e-07,
|
| 2377 |
+
"loss": 0.0943,
|
| 2378 |
+
"step": 9200
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 0.92,
|
| 2382 |
+
"learning_rate": 8.666666666666668e-07,
|
| 2383 |
+
"loss": 0.1008,
|
| 2384 |
+
"step": 9225
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 0.93,
|
| 2388 |
+
"learning_rate": 8.388888888888889e-07,
|
| 2389 |
+
"loss": 0.0911,
|
| 2390 |
+
"step": 9250
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 0.93,
|
| 2394 |
+
"learning_rate": 8.111111111111112e-07,
|
| 2395 |
+
"loss": 0.0967,
|
| 2396 |
+
"step": 9275
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.93,
|
| 2400 |
+
"learning_rate": 7.833333333333335e-07,
|
| 2401 |
+
"loss": 0.0922,
|
| 2402 |
+
"step": 9300
|
| 2403 |
+
},
|
| 2404 |
+
{
|
| 2405 |
+
"epoch": 0.93,
|
| 2406 |
+
"learning_rate": 7.555555555555556e-07,
|
| 2407 |
+
"loss": 0.0969,
|
| 2408 |
+
"step": 9325
|
| 2409 |
+
},
|
| 2410 |
+
{
|
| 2411 |
+
"epoch": 0.94,
|
| 2412 |
+
"learning_rate": 7.277777777777778e-07,
|
| 2413 |
+
"loss": 0.1004,
|
| 2414 |
+
"step": 9350
|
| 2415 |
+
},
|
| 2416 |
+
{
|
| 2417 |
+
"epoch": 0.94,
|
| 2418 |
+
"learning_rate": 7.000000000000001e-07,
|
| 2419 |
+
"loss": 0.0875,
|
| 2420 |
+
"step": 9375
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 0.94,
|
| 2424 |
+
"learning_rate": 6.722222222222223e-07,
|
| 2425 |
+
"loss": 0.0918,
|
| 2426 |
+
"step": 9400
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 0.94,
|
| 2430 |
+
"learning_rate": 6.444444444444445e-07,
|
| 2431 |
+
"loss": 0.0946,
|
| 2432 |
+
"step": 9425
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 0.94,
|
| 2436 |
+
"learning_rate": 6.166666666666668e-07,
|
| 2437 |
+
"loss": 0.0878,
|
| 2438 |
+
"step": 9450
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 0.95,
|
| 2442 |
+
"learning_rate": 5.888888888888889e-07,
|
| 2443 |
+
"loss": 0.0967,
|
| 2444 |
+
"step": 9475
|
| 2445 |
+
},
|
| 2446 |
+
{
|
| 2447 |
+
"epoch": 0.95,
|
| 2448 |
+
"learning_rate": 5.611111111111111e-07,
|
| 2449 |
+
"loss": 0.0973,
|
| 2450 |
+
"step": 9500
|
| 2451 |
+
},
|
| 2452 |
+
{
|
| 2453 |
+
"epoch": 0.95,
|
| 2454 |
+
"eval_accuracy": 0.9573,
|
| 2455 |
+
"eval_loss": 0.12337620556354523,
|
| 2456 |
+
"eval_runtime": 145.4242,
|
| 2457 |
+
"eval_samples_per_second": 68.764,
|
| 2458 |
+
"eval_steps_per_second": 0.378,
|
| 2459 |
+
"step": 9500
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 0.95,
|
| 2463 |
+
"learning_rate": 5.333333333333335e-07,
|
| 2464 |
+
"loss": 0.0863,
|
| 2465 |
+
"step": 9525
|
| 2466 |
+
},
|
| 2467 |
+
{
|
| 2468 |
+
"epoch": 0.95,
|
| 2469 |
+
"learning_rate": 5.055555555555556e-07,
|
| 2470 |
+
"loss": 0.0944,
|
| 2471 |
+
"step": 9550
|
| 2472 |
+
},
|
| 2473 |
+
{
|
| 2474 |
+
"epoch": 0.96,
|
| 2475 |
+
"learning_rate": 4.777777777777778e-07,
|
| 2476 |
+
"loss": 0.0927,
|
| 2477 |
+
"step": 9575
|
| 2478 |
+
},
|
| 2479 |
+
{
|
| 2480 |
+
"epoch": 0.96,
|
| 2481 |
+
"learning_rate": 4.5000000000000003e-07,
|
| 2482 |
+
"loss": 0.09,
|
| 2483 |
+
"step": 9600
|
| 2484 |
+
},
|
| 2485 |
+
{
|
| 2486 |
+
"epoch": 0.96,
|
| 2487 |
+
"learning_rate": 4.2222222222222226e-07,
|
| 2488 |
+
"loss": 0.0919,
|
| 2489 |
+
"step": 9625
|
| 2490 |
+
},
|
| 2491 |
+
{
|
| 2492 |
+
"epoch": 0.96,
|
| 2493 |
+
"learning_rate": 3.9444444444444444e-07,
|
| 2494 |
+
"loss": 0.1121,
|
| 2495 |
+
"step": 9650
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 0.97,
|
| 2499 |
+
"learning_rate": 3.666666666666667e-07,
|
| 2500 |
+
"loss": 0.0811,
|
| 2501 |
+
"step": 9675
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 0.97,
|
| 2505 |
+
"learning_rate": 3.4000000000000003e-07,
|
| 2506 |
+
"loss": 0.1077,
|
| 2507 |
+
"step": 9700
|
| 2508 |
+
},
|
| 2509 |
+
{
|
| 2510 |
+
"epoch": 0.97,
|
| 2511 |
+
"learning_rate": 3.1222222222222226e-07,
|
| 2512 |
+
"loss": 0.093,
|
| 2513 |
+
"step": 9725
|
| 2514 |
+
},
|
| 2515 |
+
{
|
| 2516 |
+
"epoch": 0.97,
|
| 2517 |
+
"learning_rate": 2.844444444444445e-07,
|
| 2518 |
+
"loss": 0.0869,
|
| 2519 |
+
"step": 9750
|
| 2520 |
+
},
|
| 2521 |
+
{
|
| 2522 |
+
"epoch": 0.98,
|
| 2523 |
+
"learning_rate": 2.566666666666667e-07,
|
| 2524 |
+
"loss": 0.1002,
|
| 2525 |
+
"step": 9775
|
| 2526 |
+
},
|
| 2527 |
+
{
|
| 2528 |
+
"epoch": 0.98,
|
| 2529 |
+
"learning_rate": 2.2888888888888892e-07,
|
| 2530 |
+
"loss": 0.0925,
|
| 2531 |
+
"step": 9800
|
| 2532 |
+
},
|
| 2533 |
+
{
|
| 2534 |
+
"epoch": 0.98,
|
| 2535 |
+
"learning_rate": 2.0111111111111112e-07,
|
| 2536 |
+
"loss": 0.087,
|
| 2537 |
+
"step": 9825
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 0.98,
|
| 2541 |
+
"learning_rate": 1.7333333333333335e-07,
|
| 2542 |
+
"loss": 0.0909,
|
| 2543 |
+
"step": 9850
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 0.99,
|
| 2547 |
+
"learning_rate": 1.4555555555555558e-07,
|
| 2548 |
+
"loss": 0.1016,
|
| 2549 |
+
"step": 9875
|
| 2550 |
+
},
|
| 2551 |
+
{
|
| 2552 |
+
"epoch": 0.99,
|
| 2553 |
+
"learning_rate": 1.1777777777777778e-07,
|
| 2554 |
+
"loss": 0.0943,
|
| 2555 |
+
"step": 9900
|
| 2556 |
+
},
|
| 2557 |
+
{
|
| 2558 |
+
"epoch": 0.99,
|
| 2559 |
+
"learning_rate": 9e-08,
|
| 2560 |
+
"loss": 0.1004,
|
| 2561 |
+
"step": 9925
|
| 2562 |
+
},
|
| 2563 |
+
{
|
| 2564 |
+
"epoch": 0.99,
|
| 2565 |
+
"learning_rate": 6.222222222222223e-08,
|
| 2566 |
+
"loss": 0.099,
|
| 2567 |
+
"step": 9950
|
| 2568 |
+
},
|
| 2569 |
+
{
|
| 2570 |
+
"epoch": 1.0,
|
| 2571 |
+
"learning_rate": 3.4444444444444444e-08,
|
| 2572 |
+
"loss": 0.0976,
|
| 2573 |
+
"step": 9975
|
| 2574 |
+
},
|
| 2575 |
+
{
|
| 2576 |
+
"epoch": 1.0,
|
| 2577 |
+
"learning_rate": 6.666666666666667e-09,
|
| 2578 |
+
"loss": 0.0901,
|
| 2579 |
+
"step": 10000
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 1.0,
|
| 2583 |
+
"eval_accuracy": 0.9562,
|
| 2584 |
+
"eval_loss": 0.12256840616464615,
|
| 2585 |
+
"eval_runtime": 150.3828,
|
| 2586 |
+
"eval_samples_per_second": 66.497,
|
| 2587 |
+
"eval_steps_per_second": 0.366,
|
| 2588 |
+
"step": 10000
|
| 2589 |
+
}
|
| 2590 |
+
],
|
| 2591 |
+
"max_steps": 10000,
|
| 2592 |
+
"num_train_epochs": 9223372036854775807,
|
| 2593 |
+
"total_flos": 2.418844065244224e+17,
|
| 2594 |
+
"trial_name": null,
|
| 2595 |
+
"trial_params": null
|
| 2596 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:202c7fe46ae36dee00f881f3d2e8ecb66afb40a1f5d2bd0f09838e761908a7d5
|
| 3 |
+
size 2799
|