Upload folder using huggingface_hub
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- .gitattributes +8 -0
- checkpoint-10536/config.json +49 -0
- checkpoint-10536/model.safetensors +3 -0
- checkpoint-10536/optimizer.pt +3 -0
- checkpoint-10536/rng_state.pth +3 -0
- checkpoint-10536/scaler.pt +3 -0
- checkpoint-10536/scheduler.pt +3 -0
- checkpoint-10536/special_tokens_map.json +15 -0
- checkpoint-10536/tokenizer.json +3 -0
- checkpoint-10536/tokenizer_config.json +55 -0
- checkpoint-10536/trainer_state.json +1570 -0
- checkpoint-10536/training_args.bin +3 -0
- checkpoint-12292/config.json +49 -0
- checkpoint-12292/model.safetensors +3 -0
- checkpoint-12292/optimizer.pt +3 -0
- checkpoint-12292/rng_state.pth +3 -0
- checkpoint-12292/scaler.pt +3 -0
- checkpoint-12292/scheduler.pt +3 -0
- checkpoint-12292/special_tokens_map.json +15 -0
- checkpoint-12292/tokenizer.json +3 -0
- checkpoint-12292/tokenizer_config.json +55 -0
- checkpoint-12292/trainer_state.json +1826 -0
- checkpoint-12292/training_args.bin +3 -0
- checkpoint-1756/config.json +49 -0
- checkpoint-1756/model.safetensors +3 -0
- checkpoint-1756/optimizer.pt +3 -0
- checkpoint-1756/rng_state.pth +3 -0
- checkpoint-1756/scaler.pt +3 -0
- checkpoint-1756/scheduler.pt +3 -0
- checkpoint-1756/special_tokens_map.json +15 -0
- checkpoint-1756/tokenizer.json +3 -0
- checkpoint-1756/tokenizer_config.json +55 -0
- checkpoint-1756/trainer_state.json +290 -0
- checkpoint-1756/training_args.bin +3 -0
- checkpoint-3512/config.json +49 -0
- checkpoint-3512/model.safetensors +3 -0
- checkpoint-3512/optimizer.pt +3 -0
- checkpoint-3512/rng_state.pth +3 -0
- checkpoint-3512/scaler.pt +3 -0
- checkpoint-3512/scheduler.pt +3 -0
- checkpoint-3512/special_tokens_map.json +15 -0
- checkpoint-3512/tokenizer.json +3 -0
- checkpoint-3512/tokenizer_config.json +55 -0
- checkpoint-3512/trainer_state.json +546 -0
- checkpoint-3512/training_args.bin +3 -0
- checkpoint-5268/config.json +49 -0
- checkpoint-5268/model.safetensors +3 -0
- checkpoint-5268/optimizer.pt +3 -0
- checkpoint-5268/rng_state.pth +3 -0
- checkpoint-5268/scaler.pt +3 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,11 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
checkpoint-10536/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
checkpoint-12292/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 38 |
+
checkpoint-1756/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 39 |
+
checkpoint-3512/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 40 |
+
checkpoint-5268/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 41 |
+
checkpoint-7024/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 42 |
+
checkpoint-8780/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 43 |
+
tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
checkpoint-10536/config.json
ADDED
|
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"XLMRobertaForTokenClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "O",
|
| 14 |
+
"1": "B-PER",
|
| 15 |
+
"2": "I-PER",
|
| 16 |
+
"3": "B-ORG",
|
| 17 |
+
"4": "I-ORG",
|
| 18 |
+
"5": "B-LOC",
|
| 19 |
+
"6": "I-LOC",
|
| 20 |
+
"7": "B-MISC",
|
| 21 |
+
"8": "I-MISC"
|
| 22 |
+
},
|
| 23 |
+
"initializer_range": 0.02,
|
| 24 |
+
"intermediate_size": 3072,
|
| 25 |
+
"label2id": {
|
| 26 |
+
"B-LOC": 5,
|
| 27 |
+
"B-MISC": 7,
|
| 28 |
+
"B-ORG": 3,
|
| 29 |
+
"B-PER": 1,
|
| 30 |
+
"I-LOC": 6,
|
| 31 |
+
"I-MISC": 8,
|
| 32 |
+
"I-ORG": 4,
|
| 33 |
+
"I-PER": 2,
|
| 34 |
+
"O": 0
|
| 35 |
+
},
|
| 36 |
+
"layer_norm_eps": 1e-05,
|
| 37 |
+
"max_position_embeddings": 514,
|
| 38 |
+
"model_type": "xlm-roberta",
|
| 39 |
+
"num_attention_heads": 12,
|
| 40 |
+
"num_hidden_layers": 12,
|
| 41 |
+
"output_past": true,
|
| 42 |
+
"pad_token_id": 1,
|
| 43 |
+
"position_embedding_type": "absolute",
|
| 44 |
+
"torch_dtype": "float32",
|
| 45 |
+
"transformers_version": "4.51.3",
|
| 46 |
+
"type_vocab_size": 1,
|
| 47 |
+
"use_cache": true,
|
| 48 |
+
"vocab_size": 250002
|
| 49 |
+
}
|
checkpoint-10536/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6440a058003a17dd1e8e9f37290457a8d15df03bda5f80db000d0ec55c3b65a3
|
| 3 |
+
size 1109863956
|
checkpoint-10536/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e5aebcdf200787a3d36ad26e97baf2f95a1edbcaa27dfb7de4afd552f457a17d
|
| 3 |
+
size 2219846202
|
checkpoint-10536/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:083825a219aea1d69cbec6e4b693fbcf2e11dad1312750efc78367126c76fbe2
|
| 3 |
+
size 14244
|
checkpoint-10536/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:10d19f6e3641d81395be9b8e86101ef67c4f91a8f20888a2ad5331c197d7e6e6
|
| 3 |
+
size 988
|
checkpoint-10536/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4cd02bf0a083f5bb6fbdae03e00fa68b9cfc022b252c7ffd71dcb9e74d3421a
|
| 3 |
+
size 1064
|
checkpoint-10536/special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": {
|
| 6 |
+
"content": "<mask>",
|
| 7 |
+
"lstrip": true,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
checkpoint-10536/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3a56def25aa40facc030ea8b0b87f3688e4b3c39eb8b45d5702b3a1300fe2a20
|
| 3 |
+
size 17082734
|
checkpoint-10536/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"250001": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": true,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 512,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizer",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
checkpoint-10536/trainer_state.json
ADDED
|
@@ -0,0 +1,1570 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 10536,
|
| 3 |
+
"best_metric": 0.9557566616390145,
|
| 4 |
+
"best_model_checkpoint": "C:\\Users\\Shara\\projects\\models\\run_16-lr_2e-05-acc_1-wd_0.01-bs_8-ep_7\\checkpoint-10536",
|
| 5 |
+
"epoch": 6.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 10536,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.02847380410022779,
|
| 14 |
+
"grad_norm": 4.8302321434021,
|
| 15 |
+
"learning_rate": 1.992840872111943e-05,
|
| 16 |
+
"loss": 0.9994,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.05694760820045558,
|
| 21 |
+
"grad_norm": 3.4978649616241455,
|
| 22 |
+
"learning_rate": 1.9847054995118778e-05,
|
| 23 |
+
"loss": 0.3939,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.08542141230068337,
|
| 28 |
+
"grad_norm": 8.353672981262207,
|
| 29 |
+
"learning_rate": 1.9765701269118128e-05,
|
| 30 |
+
"loss": 0.2882,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.11389521640091116,
|
| 35 |
+
"grad_norm": 9.822871208190918,
|
| 36 |
+
"learning_rate": 1.9684347543117477e-05,
|
| 37 |
+
"loss": 0.1728,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.14236902050113895,
|
| 42 |
+
"grad_norm": 3.1411221027374268,
|
| 43 |
+
"learning_rate": 1.9602993817116827e-05,
|
| 44 |
+
"loss": 0.1251,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.17084282460136674,
|
| 49 |
+
"grad_norm": 8.493424415588379,
|
| 50 |
+
"learning_rate": 1.9521640091116173e-05,
|
| 51 |
+
"loss": 0.1405,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.19931662870159453,
|
| 56 |
+
"grad_norm": 4.757007122039795,
|
| 57 |
+
"learning_rate": 1.9440286365115523e-05,
|
| 58 |
+
"loss": 0.1431,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.22779043280182232,
|
| 63 |
+
"grad_norm": 4.339049816131592,
|
| 64 |
+
"learning_rate": 1.9358932639114873e-05,
|
| 65 |
+
"loss": 0.0912,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.25626423690205014,
|
| 70 |
+
"grad_norm": 14.540772438049316,
|
| 71 |
+
"learning_rate": 1.9277578913114222e-05,
|
| 72 |
+
"loss": 0.1048,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.2847380410022779,
|
| 77 |
+
"grad_norm": 10.503073692321777,
|
| 78 |
+
"learning_rate": 1.9196225187113572e-05,
|
| 79 |
+
"loss": 0.1224,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.3132118451025057,
|
| 84 |
+
"grad_norm": 2.738131284713745,
|
| 85 |
+
"learning_rate": 1.911487146111292e-05,
|
| 86 |
+
"loss": 0.1025,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.3416856492027335,
|
| 91 |
+
"grad_norm": 6.681102752685547,
|
| 92 |
+
"learning_rate": 1.903351773511227e-05,
|
| 93 |
+
"loss": 0.0869,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.3701594533029613,
|
| 98 |
+
"grad_norm": 1.4330339431762695,
|
| 99 |
+
"learning_rate": 1.895216400911162e-05,
|
| 100 |
+
"loss": 0.052,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.39863325740318906,
|
| 105 |
+
"grad_norm": 2.581470012664795,
|
| 106 |
+
"learning_rate": 1.8870810283110967e-05,
|
| 107 |
+
"loss": 0.0808,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.4271070615034169,
|
| 112 |
+
"grad_norm": 5.825100898742676,
|
| 113 |
+
"learning_rate": 1.8789456557110317e-05,
|
| 114 |
+
"loss": 0.0716,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.45558086560364464,
|
| 119 |
+
"grad_norm": 1.74166738986969,
|
| 120 |
+
"learning_rate": 1.8708102831109666e-05,
|
| 121 |
+
"loss": 0.0806,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.48405466970387245,
|
| 126 |
+
"grad_norm": 7.3097405433654785,
|
| 127 |
+
"learning_rate": 1.8626749105109016e-05,
|
| 128 |
+
"loss": 0.0666,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.5125284738041003,
|
| 133 |
+
"grad_norm": 2.220766067504883,
|
| 134 |
+
"learning_rate": 1.8545395379108362e-05,
|
| 135 |
+
"loss": 0.0768,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.541002277904328,
|
| 140 |
+
"grad_norm": 4.0566511154174805,
|
| 141 |
+
"learning_rate": 1.8464041653107715e-05,
|
| 142 |
+
"loss": 0.0707,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.5694760820045558,
|
| 147 |
+
"grad_norm": 0.2833240330219269,
|
| 148 |
+
"learning_rate": 1.8382687927107065e-05,
|
| 149 |
+
"loss": 0.0858,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.5979498861047836,
|
| 154 |
+
"grad_norm": 21.131481170654297,
|
| 155 |
+
"learning_rate": 1.8301334201106415e-05,
|
| 156 |
+
"loss": 0.0595,
|
| 157 |
+
"step": 1050
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.6264236902050114,
|
| 161 |
+
"grad_norm": 27.655433654785156,
|
| 162 |
+
"learning_rate": 1.821998047510576e-05,
|
| 163 |
+
"loss": 0.0466,
|
| 164 |
+
"step": 1100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.6548974943052391,
|
| 168 |
+
"grad_norm": 3.4721717834472656,
|
| 169 |
+
"learning_rate": 1.813862674910511e-05,
|
| 170 |
+
"loss": 0.0776,
|
| 171 |
+
"step": 1150
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.683371298405467,
|
| 175 |
+
"grad_norm": 2.4811441898345947,
|
| 176 |
+
"learning_rate": 1.805727302310446e-05,
|
| 177 |
+
"loss": 0.0647,
|
| 178 |
+
"step": 1200
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.7118451025056948,
|
| 182 |
+
"grad_norm": 6.169066429138184,
|
| 183 |
+
"learning_rate": 1.797591929710381e-05,
|
| 184 |
+
"loss": 0.0724,
|
| 185 |
+
"step": 1250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.7403189066059226,
|
| 189 |
+
"grad_norm": 8.984748840332031,
|
| 190 |
+
"learning_rate": 1.7894565571103156e-05,
|
| 191 |
+
"loss": 0.0838,
|
| 192 |
+
"step": 1300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.7687927107061503,
|
| 196 |
+
"grad_norm": 5.944505214691162,
|
| 197 |
+
"learning_rate": 1.7813211845102506e-05,
|
| 198 |
+
"loss": 0.0683,
|
| 199 |
+
"step": 1350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.7972665148063781,
|
| 203 |
+
"grad_norm": 0.28812381625175476,
|
| 204 |
+
"learning_rate": 1.7731858119101856e-05,
|
| 205 |
+
"loss": 0.0614,
|
| 206 |
+
"step": 1400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.8257403189066059,
|
| 210 |
+
"grad_norm": 6.176011085510254,
|
| 211 |
+
"learning_rate": 1.7650504393101205e-05,
|
| 212 |
+
"loss": 0.0615,
|
| 213 |
+
"step": 1450
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.8542141230068337,
|
| 217 |
+
"grad_norm": 2.5244405269622803,
|
| 218 |
+
"learning_rate": 1.7569150667100555e-05,
|
| 219 |
+
"loss": 0.0714,
|
| 220 |
+
"step": 1500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.8826879271070615,
|
| 224 |
+
"grad_norm": 3.3508074283599854,
|
| 225 |
+
"learning_rate": 1.7487796941099904e-05,
|
| 226 |
+
"loss": 0.0563,
|
| 227 |
+
"step": 1550
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.9111617312072893,
|
| 231 |
+
"grad_norm": 9.980842590332031,
|
| 232 |
+
"learning_rate": 1.7406443215099254e-05,
|
| 233 |
+
"loss": 0.0647,
|
| 234 |
+
"step": 1600
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.9396355353075171,
|
| 238 |
+
"grad_norm": 3.6959774494171143,
|
| 239 |
+
"learning_rate": 1.7325089489098604e-05,
|
| 240 |
+
"loss": 0.0682,
|
| 241 |
+
"step": 1650
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.9681093394077449,
|
| 245 |
+
"grad_norm": 0.626620888710022,
|
| 246 |
+
"learning_rate": 1.724373576309795e-05,
|
| 247 |
+
"loss": 0.0499,
|
| 248 |
+
"step": 1700
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.9965831435079726,
|
| 252 |
+
"grad_norm": 7.145672798156738,
|
| 253 |
+
"learning_rate": 1.71623820370973e-05,
|
| 254 |
+
"loss": 0.0569,
|
| 255 |
+
"step": 1750
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.0,
|
| 259 |
+
"eval_f1": 0.9242246437552389,
|
| 260 |
+
"eval_loss": 0.05600914731621742,
|
| 261 |
+
"eval_precision": 0.9206746826987308,
|
| 262 |
+
"eval_recall": 0.927802086839448,
|
| 263 |
+
"eval_runtime": 3.7538,
|
| 264 |
+
"eval_samples_per_second": 865.798,
|
| 265 |
+
"eval_steps_per_second": 108.425,
|
| 266 |
+
"step": 1756
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.0250569476082005,
|
| 270 |
+
"grad_norm": 4.280463695526123,
|
| 271 |
+
"learning_rate": 1.708102831109665e-05,
|
| 272 |
+
"loss": 0.0673,
|
| 273 |
+
"step": 1800
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.0535307517084282,
|
| 277 |
+
"grad_norm": 0.06692535430192947,
|
| 278 |
+
"learning_rate": 1.6999674585096e-05,
|
| 279 |
+
"loss": 0.0403,
|
| 280 |
+
"step": 1850
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 1.082004555808656,
|
| 284 |
+
"grad_norm": 0.03789375349879265,
|
| 285 |
+
"learning_rate": 1.691832085909535e-05,
|
| 286 |
+
"loss": 0.0496,
|
| 287 |
+
"step": 1900
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.1104783599088839,
|
| 291 |
+
"grad_norm": 3.698348045349121,
|
| 292 |
+
"learning_rate": 1.6836967133094698e-05,
|
| 293 |
+
"loss": 0.0381,
|
| 294 |
+
"step": 1950
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.1389521640091116,
|
| 298 |
+
"grad_norm": 2.753032684326172,
|
| 299 |
+
"learning_rate": 1.6755613407094048e-05,
|
| 300 |
+
"loss": 0.0432,
|
| 301 |
+
"step": 2000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.1674259681093395,
|
| 305 |
+
"grad_norm": 0.06336130946874619,
|
| 306 |
+
"learning_rate": 1.6674259681093398e-05,
|
| 307 |
+
"loss": 0.0482,
|
| 308 |
+
"step": 2050
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.1958997722095672,
|
| 312 |
+
"grad_norm": 3.9503254890441895,
|
| 313 |
+
"learning_rate": 1.6592905955092744e-05,
|
| 314 |
+
"loss": 0.0436,
|
| 315 |
+
"step": 2100
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.224373576309795,
|
| 319 |
+
"grad_norm": 0.08679741621017456,
|
| 320 |
+
"learning_rate": 1.6511552229092093e-05,
|
| 321 |
+
"loss": 0.0346,
|
| 322 |
+
"step": 2150
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 1.2528473804100229,
|
| 326 |
+
"grad_norm": 2.1995127201080322,
|
| 327 |
+
"learning_rate": 1.6430198503091443e-05,
|
| 328 |
+
"loss": 0.0312,
|
| 329 |
+
"step": 2200
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.2813211845102506,
|
| 333 |
+
"grad_norm": 5.7709808349609375,
|
| 334 |
+
"learning_rate": 1.634884477709079e-05,
|
| 335 |
+
"loss": 0.032,
|
| 336 |
+
"step": 2250
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 1.3097949886104785,
|
| 340 |
+
"grad_norm": 0.2031829059123993,
|
| 341 |
+
"learning_rate": 1.626749105109014e-05,
|
| 342 |
+
"loss": 0.053,
|
| 343 |
+
"step": 2300
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 1.3382687927107062,
|
| 347 |
+
"grad_norm": 4.292157173156738,
|
| 348 |
+
"learning_rate": 1.6186137325089492e-05,
|
| 349 |
+
"loss": 0.0413,
|
| 350 |
+
"step": 2350
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.366742596810934,
|
| 354 |
+
"grad_norm": 0.07821401953697205,
|
| 355 |
+
"learning_rate": 1.6104783599088842e-05,
|
| 356 |
+
"loss": 0.0304,
|
| 357 |
+
"step": 2400
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.3952164009111616,
|
| 361 |
+
"grad_norm": 0.5930206179618835,
|
| 362 |
+
"learning_rate": 1.6023429873088188e-05,
|
| 363 |
+
"loss": 0.0525,
|
| 364 |
+
"step": 2450
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 1.4236902050113895,
|
| 368 |
+
"grad_norm": 1.6114712953567505,
|
| 369 |
+
"learning_rate": 1.5942076147087538e-05,
|
| 370 |
+
"loss": 0.0391,
|
| 371 |
+
"step": 2500
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.4521640091116172,
|
| 375 |
+
"grad_norm": 0.28632065653800964,
|
| 376 |
+
"learning_rate": 1.5860722421086887e-05,
|
| 377 |
+
"loss": 0.0491,
|
| 378 |
+
"step": 2550
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 1.4806378132118452,
|
| 382 |
+
"grad_norm": 1.0017669200897217,
|
| 383 |
+
"learning_rate": 1.5779368695086237e-05,
|
| 384 |
+
"loss": 0.0395,
|
| 385 |
+
"step": 2600
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 1.5091116173120729,
|
| 389 |
+
"grad_norm": 0.03409272059798241,
|
| 390 |
+
"learning_rate": 1.5698014969085583e-05,
|
| 391 |
+
"loss": 0.0399,
|
| 392 |
+
"step": 2650
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.5375854214123006,
|
| 396 |
+
"grad_norm": 3.5108702182769775,
|
| 397 |
+
"learning_rate": 1.5616661243084933e-05,
|
| 398 |
+
"loss": 0.0424,
|
| 399 |
+
"step": 2700
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.5660592255125285,
|
| 403 |
+
"grad_norm": 3.5875349044799805,
|
| 404 |
+
"learning_rate": 1.5535307517084283e-05,
|
| 405 |
+
"loss": 0.0497,
|
| 406 |
+
"step": 2750
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.5945330296127562,
|
| 410 |
+
"grad_norm": 0.10498251020908356,
|
| 411 |
+
"learning_rate": 1.5453953791083632e-05,
|
| 412 |
+
"loss": 0.0461,
|
| 413 |
+
"step": 2800
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.6230068337129842,
|
| 417 |
+
"grad_norm": 7.618963718414307,
|
| 418 |
+
"learning_rate": 1.5374227139602996e-05,
|
| 419 |
+
"loss": 0.0394,
|
| 420 |
+
"step": 2850
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 1.6514806378132119,
|
| 424 |
+
"grad_norm": 3.3920793533325195,
|
| 425 |
+
"learning_rate": 1.5292873413602346e-05,
|
| 426 |
+
"loss": 0.0339,
|
| 427 |
+
"step": 2900
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 1.6799544419134396,
|
| 431 |
+
"grad_norm": 0.04736129939556122,
|
| 432 |
+
"learning_rate": 1.5211519687601692e-05,
|
| 433 |
+
"loss": 0.0358,
|
| 434 |
+
"step": 2950
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.7084282460136673,
|
| 438 |
+
"grad_norm": 0.17463508248329163,
|
| 439 |
+
"learning_rate": 1.5130165961601042e-05,
|
| 440 |
+
"loss": 0.034,
|
| 441 |
+
"step": 3000
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 1.7369020501138952,
|
| 445 |
+
"grad_norm": 0.058148205280303955,
|
| 446 |
+
"learning_rate": 1.5048812235600392e-05,
|
| 447 |
+
"loss": 0.0317,
|
| 448 |
+
"step": 3050
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 1.7653758542141231,
|
| 452 |
+
"grad_norm": 3.5262815952301025,
|
| 453 |
+
"learning_rate": 1.4967458509599741e-05,
|
| 454 |
+
"loss": 0.0477,
|
| 455 |
+
"step": 3100
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 1.7938496583143508,
|
| 459 |
+
"grad_norm": 2.971937894821167,
|
| 460 |
+
"learning_rate": 1.488610478359909e-05,
|
| 461 |
+
"loss": 0.0364,
|
| 462 |
+
"step": 3150
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 1.8223234624145785,
|
| 466 |
+
"grad_norm": 0.05181724950671196,
|
| 467 |
+
"learning_rate": 1.4804751057598439e-05,
|
| 468 |
+
"loss": 0.0394,
|
| 469 |
+
"step": 3200
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 1.8507972665148062,
|
| 473 |
+
"grad_norm": 0.2712370455265045,
|
| 474 |
+
"learning_rate": 1.4723397331597789e-05,
|
| 475 |
+
"loss": 0.0456,
|
| 476 |
+
"step": 3250
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.8792710706150342,
|
| 480 |
+
"grad_norm": 1.608622431755066,
|
| 481 |
+
"learning_rate": 1.4642043605597138e-05,
|
| 482 |
+
"loss": 0.05,
|
| 483 |
+
"step": 3300
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 1.907744874715262,
|
| 487 |
+
"grad_norm": 5.2423529624938965,
|
| 488 |
+
"learning_rate": 1.4560689879596486e-05,
|
| 489 |
+
"loss": 0.0371,
|
| 490 |
+
"step": 3350
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 1.9362186788154898,
|
| 494 |
+
"grad_norm": 0.04786211624741554,
|
| 495 |
+
"learning_rate": 1.4479336153595836e-05,
|
| 496 |
+
"loss": 0.0329,
|
| 497 |
+
"step": 3400
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 1.9646924829157175,
|
| 501 |
+
"grad_norm": 0.5264931917190552,
|
| 502 |
+
"learning_rate": 1.4397982427595185e-05,
|
| 503 |
+
"loss": 0.0379,
|
| 504 |
+
"step": 3450
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 1.9931662870159452,
|
| 508 |
+
"grad_norm": 0.236006498336792,
|
| 509 |
+
"learning_rate": 1.4316628701594535e-05,
|
| 510 |
+
"loss": 0.0253,
|
| 511 |
+
"step": 3500
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 2.0,
|
| 515 |
+
"eval_f1": 0.9406438631790744,
|
| 516 |
+
"eval_loss": 0.05223705992102623,
|
| 517 |
+
"eval_precision": 0.9371867691279653,
|
| 518 |
+
"eval_recall": 0.9441265567149109,
|
| 519 |
+
"eval_runtime": 3.7109,
|
| 520 |
+
"eval_samples_per_second": 875.791,
|
| 521 |
+
"eval_steps_per_second": 109.676,
|
| 522 |
+
"step": 3512
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 2.021640091116173,
|
| 526 |
+
"grad_norm": 4.225963115692139,
|
| 527 |
+
"learning_rate": 1.4235274975593883e-05,
|
| 528 |
+
"loss": 0.0373,
|
| 529 |
+
"step": 3550
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 2.050113895216401,
|
| 533 |
+
"grad_norm": 0.15388111770153046,
|
| 534 |
+
"learning_rate": 1.4153921249593233e-05,
|
| 535 |
+
"loss": 0.0258,
|
| 536 |
+
"step": 3600
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.078587699316629,
|
| 540 |
+
"grad_norm": 6.3911895751953125,
|
| 541 |
+
"learning_rate": 1.4072567523592582e-05,
|
| 542 |
+
"loss": 0.0217,
|
| 543 |
+
"step": 3650
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 2.1070615034168565,
|
| 547 |
+
"grad_norm": 0.03075530007481575,
|
| 548 |
+
"learning_rate": 1.3991213797591932e-05,
|
| 549 |
+
"loss": 0.019,
|
| 550 |
+
"step": 3700
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 2.135535307517084,
|
| 554 |
+
"grad_norm": 0.16612432897090912,
|
| 555 |
+
"learning_rate": 1.3909860071591278e-05,
|
| 556 |
+
"loss": 0.0175,
|
| 557 |
+
"step": 3750
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 2.164009111617312,
|
| 561 |
+
"grad_norm": 0.14543022215366364,
|
| 562 |
+
"learning_rate": 1.382850634559063e-05,
|
| 563 |
+
"loss": 0.0289,
|
| 564 |
+
"step": 3800
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 2.19248291571754,
|
| 568 |
+
"grad_norm": 2.9880359172821045,
|
| 569 |
+
"learning_rate": 1.374715261958998e-05,
|
| 570 |
+
"loss": 0.0208,
|
| 571 |
+
"step": 3850
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 2.2209567198177678,
|
| 575 |
+
"grad_norm": 9.878133773803711,
|
| 576 |
+
"learning_rate": 1.3665798893589329e-05,
|
| 577 |
+
"loss": 0.021,
|
| 578 |
+
"step": 3900
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 2.2494305239179955,
|
| 582 |
+
"grad_norm": 9.684176445007324,
|
| 583 |
+
"learning_rate": 1.3584445167588675e-05,
|
| 584 |
+
"loss": 0.0267,
|
| 585 |
+
"step": 3950
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 2.277904328018223,
|
| 589 |
+
"grad_norm": 33.15859603881836,
|
| 590 |
+
"learning_rate": 1.3503091441588025e-05,
|
| 591 |
+
"loss": 0.0295,
|
| 592 |
+
"step": 4000
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 2.306378132118451,
|
| 596 |
+
"grad_norm": 2.559544086456299,
|
| 597 |
+
"learning_rate": 1.3421737715587374e-05,
|
| 598 |
+
"loss": 0.0292,
|
| 599 |
+
"step": 4050
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 2.334851936218679,
|
| 603 |
+
"grad_norm": 0.07091552764177322,
|
| 604 |
+
"learning_rate": 1.3340383989586724e-05,
|
| 605 |
+
"loss": 0.0262,
|
| 606 |
+
"step": 4100
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.3633257403189067,
|
| 610 |
+
"grad_norm": 1.8119585514068604,
|
| 611 |
+
"learning_rate": 1.3259030263586072e-05,
|
| 612 |
+
"loss": 0.03,
|
| 613 |
+
"step": 4150
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 2.3917995444191344,
|
| 617 |
+
"grad_norm": 0.1476636379957199,
|
| 618 |
+
"learning_rate": 1.3177676537585422e-05,
|
| 619 |
+
"loss": 0.0196,
|
| 620 |
+
"step": 4200
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.420273348519362,
|
| 624 |
+
"grad_norm": 0.04482650011777878,
|
| 625 |
+
"learning_rate": 1.3096322811584771e-05,
|
| 626 |
+
"loss": 0.0263,
|
| 627 |
+
"step": 4250
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 2.44874715261959,
|
| 631 |
+
"grad_norm": 0.09201560169458389,
|
| 632 |
+
"learning_rate": 1.3014969085584121e-05,
|
| 633 |
+
"loss": 0.0172,
|
| 634 |
+
"step": 4300
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 2.477220956719818,
|
| 638 |
+
"grad_norm": 0.02215876244008541,
|
| 639 |
+
"learning_rate": 1.2933615359583469e-05,
|
| 640 |
+
"loss": 0.0149,
|
| 641 |
+
"step": 4350
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 2.5056947608200457,
|
| 645 |
+
"grad_norm": 0.6993932127952576,
|
| 646 |
+
"learning_rate": 1.2852261633582819e-05,
|
| 647 |
+
"loss": 0.0331,
|
| 648 |
+
"step": 4400
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 2.5341685649202734,
|
| 652 |
+
"grad_norm": 0.030453965067863464,
|
| 653 |
+
"learning_rate": 1.2770907907582168e-05,
|
| 654 |
+
"loss": 0.0307,
|
| 655 |
+
"step": 4450
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 2.562642369020501,
|
| 659 |
+
"grad_norm": 0.22211593389511108,
|
| 660 |
+
"learning_rate": 1.2689554181581518e-05,
|
| 661 |
+
"loss": 0.0183,
|
| 662 |
+
"step": 4500
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 2.591116173120729,
|
| 666 |
+
"grad_norm": 1.2842738628387451,
|
| 667 |
+
"learning_rate": 1.2608200455580866e-05,
|
| 668 |
+
"loss": 0.0226,
|
| 669 |
+
"step": 4550
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 2.619589977220957,
|
| 673 |
+
"grad_norm": 3.9048423767089844,
|
| 674 |
+
"learning_rate": 1.2526846729580216e-05,
|
| 675 |
+
"loss": 0.0239,
|
| 676 |
+
"step": 4600
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 2.6480637813211843,
|
| 680 |
+
"grad_norm": 0.9887784719467163,
|
| 681 |
+
"learning_rate": 1.2445493003579565e-05,
|
| 682 |
+
"loss": 0.0271,
|
| 683 |
+
"step": 4650
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 2.6765375854214124,
|
| 687 |
+
"grad_norm": 4.759728908538818,
|
| 688 |
+
"learning_rate": 1.2364139277578915e-05,
|
| 689 |
+
"loss": 0.0201,
|
| 690 |
+
"step": 4700
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 2.70501138952164,
|
| 694 |
+
"grad_norm": 0.2671191096305847,
|
| 695 |
+
"learning_rate": 1.2282785551578263e-05,
|
| 696 |
+
"loss": 0.0212,
|
| 697 |
+
"step": 4750
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 2.733485193621868,
|
| 701 |
+
"grad_norm": 0.036626849323511124,
|
| 702 |
+
"learning_rate": 1.2201431825577612e-05,
|
| 703 |
+
"loss": 0.0155,
|
| 704 |
+
"step": 4800
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 2.7619589977220955,
|
| 708 |
+
"grad_norm": 6.877408027648926,
|
| 709 |
+
"learning_rate": 1.2120078099576962e-05,
|
| 710 |
+
"loss": 0.0266,
|
| 711 |
+
"step": 4850
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 2.7904328018223232,
|
| 715 |
+
"grad_norm": 4.725896835327148,
|
| 716 |
+
"learning_rate": 1.2038724373576312e-05,
|
| 717 |
+
"loss": 0.0163,
|
| 718 |
+
"step": 4900
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 2.8189066059225514,
|
| 722 |
+
"grad_norm": 0.11965059489011765,
|
| 723 |
+
"learning_rate": 1.195737064757566e-05,
|
| 724 |
+
"loss": 0.0246,
|
| 725 |
+
"step": 4950
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 2.847380410022779,
|
| 729 |
+
"grad_norm": 0.6634079813957214,
|
| 730 |
+
"learning_rate": 1.187601692157501e-05,
|
| 731 |
+
"loss": 0.0304,
|
| 732 |
+
"step": 5000
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 2.875854214123007,
|
| 736 |
+
"grad_norm": 3.952694892883301,
|
| 737 |
+
"learning_rate": 1.1794663195574359e-05,
|
| 738 |
+
"loss": 0.0183,
|
| 739 |
+
"step": 5050
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 2.9043280182232345,
|
| 743 |
+
"grad_norm": 19.788942337036133,
|
| 744 |
+
"learning_rate": 1.1713309469573709e-05,
|
| 745 |
+
"loss": 0.0193,
|
| 746 |
+
"step": 5100
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 2.932801822323462,
|
| 750 |
+
"grad_norm": 4.989261627197266,
|
| 751 |
+
"learning_rate": 1.1631955743573055e-05,
|
| 752 |
+
"loss": 0.023,
|
| 753 |
+
"step": 5150
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 2.9612756264236904,
|
| 757 |
+
"grad_norm": 9.020853996276855,
|
| 758 |
+
"learning_rate": 1.1550602017572406e-05,
|
| 759 |
+
"loss": 0.0231,
|
| 760 |
+
"step": 5200
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 2.989749430523918,
|
| 764 |
+
"grad_norm": 0.9151140451431274,
|
| 765 |
+
"learning_rate": 1.1469248291571756e-05,
|
| 766 |
+
"loss": 0.0314,
|
| 767 |
+
"step": 5250
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 3.0,
|
| 771 |
+
"eval_f1": 0.9470420646891236,
|
| 772 |
+
"eval_loss": 0.04870199039578438,
|
| 773 |
+
"eval_precision": 0.943090787716956,
|
| 774 |
+
"eval_recall": 0.9510265903736116,
|
| 775 |
+
"eval_runtime": 3.7494,
|
| 776 |
+
"eval_samples_per_second": 866.805,
|
| 777 |
+
"eval_steps_per_second": 108.551,
|
| 778 |
+
"step": 5268
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 3.0182232346241458,
|
| 782 |
+
"grad_norm": 8.62992000579834,
|
| 783 |
+
"learning_rate": 1.1387894565571106e-05,
|
| 784 |
+
"loss": 0.0194,
|
| 785 |
+
"step": 5300
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.0466970387243735,
|
| 789 |
+
"grad_norm": 0.5615554451942444,
|
| 790 |
+
"learning_rate": 1.1306540839570452e-05,
|
| 791 |
+
"loss": 0.0114,
|
| 792 |
+
"step": 5350
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.075170842824601,
|
| 796 |
+
"grad_norm": 2.5735936164855957,
|
| 797 |
+
"learning_rate": 1.1225187113569801e-05,
|
| 798 |
+
"loss": 0.0163,
|
| 799 |
+
"step": 5400
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 3.1036446469248293,
|
| 803 |
+
"grad_norm": 6.504094123840332,
|
| 804 |
+
"learning_rate": 1.1143833387569151e-05,
|
| 805 |
+
"loss": 0.0135,
|
| 806 |
+
"step": 5450
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 3.132118451025057,
|
| 810 |
+
"grad_norm": 14.477537155151367,
|
| 811 |
+
"learning_rate": 1.10624796615685e-05,
|
| 812 |
+
"loss": 0.0129,
|
| 813 |
+
"step": 5500
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 3.1605922551252847,
|
| 817 |
+
"grad_norm": 0.04929841682314873,
|
| 818 |
+
"learning_rate": 1.0981125935567849e-05,
|
| 819 |
+
"loss": 0.0263,
|
| 820 |
+
"step": 5550
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 3.1890660592255125,
|
| 824 |
+
"grad_norm": 0.08875144273042679,
|
| 825 |
+
"learning_rate": 1.0899772209567198e-05,
|
| 826 |
+
"loss": 0.0134,
|
| 827 |
+
"step": 5600
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 3.21753986332574,
|
| 831 |
+
"grad_norm": 0.6238455772399902,
|
| 832 |
+
"learning_rate": 1.0818418483566548e-05,
|
| 833 |
+
"loss": 0.0107,
|
| 834 |
+
"step": 5650
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 3.2460136674259683,
|
| 838 |
+
"grad_norm": 0.004948179703205824,
|
| 839 |
+
"learning_rate": 1.0737064757565898e-05,
|
| 840 |
+
"loss": 0.0167,
|
| 841 |
+
"step": 5700
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 3.274487471526196,
|
| 845 |
+
"grad_norm": 0.017031218856573105,
|
| 846 |
+
"learning_rate": 1.0655711031565246e-05,
|
| 847 |
+
"loss": 0.0085,
|
| 848 |
+
"step": 5750
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 3.3029612756264237,
|
| 852 |
+
"grad_norm": 2.9840469360351562,
|
| 853 |
+
"learning_rate": 1.0574357305564595e-05,
|
| 854 |
+
"loss": 0.0158,
|
| 855 |
+
"step": 5800
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 3.3314350797266514,
|
| 859 |
+
"grad_norm": 1.7477540969848633,
|
| 860 |
+
"learning_rate": 1.0493003579563945e-05,
|
| 861 |
+
"loss": 0.0157,
|
| 862 |
+
"step": 5850
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 3.359908883826879,
|
| 866 |
+
"grad_norm": 0.03963543102145195,
|
| 867 |
+
"learning_rate": 1.0411649853563295e-05,
|
| 868 |
+
"loss": 0.0201,
|
| 869 |
+
"step": 5900
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 3.3883826879271073,
|
| 873 |
+
"grad_norm": 0.16669808328151703,
|
| 874 |
+
"learning_rate": 1.0330296127562643e-05,
|
| 875 |
+
"loss": 0.0159,
|
| 876 |
+
"step": 5950
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 3.416856492027335,
|
| 880 |
+
"grad_norm": 0.01953568309545517,
|
| 881 |
+
"learning_rate": 1.0248942401561992e-05,
|
| 882 |
+
"loss": 0.0117,
|
| 883 |
+
"step": 6000
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 3.4453302961275627,
|
| 887 |
+
"grad_norm": 0.07708246260881424,
|
| 888 |
+
"learning_rate": 1.0167588675561342e-05,
|
| 889 |
+
"loss": 0.0117,
|
| 890 |
+
"step": 6050
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 3.4738041002277904,
|
| 894 |
+
"grad_norm": 2.423590898513794,
|
| 895 |
+
"learning_rate": 1.0086234949560691e-05,
|
| 896 |
+
"loss": 0.0111,
|
| 897 |
+
"step": 6100
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 3.502277904328018,
|
| 901 |
+
"grad_norm": 0.006337775848805904,
|
| 902 |
+
"learning_rate": 1.000488122356004e-05,
|
| 903 |
+
"loss": 0.0052,
|
| 904 |
+
"step": 6150
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 3.5307517084282463,
|
| 908 |
+
"grad_norm": 0.01373753696680069,
|
| 909 |
+
"learning_rate": 9.923527497559389e-06,
|
| 910 |
+
"loss": 0.009,
|
| 911 |
+
"step": 6200
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 3.559225512528474,
|
| 915 |
+
"grad_norm": 0.11260247975587845,
|
| 916 |
+
"learning_rate": 9.842173771558739e-06,
|
| 917 |
+
"loss": 0.0208,
|
| 918 |
+
"step": 6250
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 3.5876993166287017,
|
| 922 |
+
"grad_norm": 1.240822196006775,
|
| 923 |
+
"learning_rate": 9.760820045558087e-06,
|
| 924 |
+
"loss": 0.0157,
|
| 925 |
+
"step": 6300
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 3.6161731207289294,
|
| 929 |
+
"grad_norm": 0.05271293595433235,
|
| 930 |
+
"learning_rate": 9.679466319557436e-06,
|
| 931 |
+
"loss": 0.0175,
|
| 932 |
+
"step": 6350
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 3.644646924829157,
|
| 936 |
+
"grad_norm": 24.546663284301758,
|
| 937 |
+
"learning_rate": 9.598112593556786e-06,
|
| 938 |
+
"loss": 0.015,
|
| 939 |
+
"step": 6400
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 3.6731207289293852,
|
| 943 |
+
"grad_norm": 5.120173931121826,
|
| 944 |
+
"learning_rate": 9.516758867556136e-06,
|
| 945 |
+
"loss": 0.017,
|
| 946 |
+
"step": 6450
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 3.7015945330296125,
|
| 950 |
+
"grad_norm": 1.1922008991241455,
|
| 951 |
+
"learning_rate": 9.435405141555484e-06,
|
| 952 |
+
"loss": 0.0194,
|
| 953 |
+
"step": 6500
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 3.7300683371298406,
|
| 957 |
+
"grad_norm": 0.039439987391233444,
|
| 958 |
+
"learning_rate": 9.354051415554833e-06,
|
| 959 |
+
"loss": 0.0309,
|
| 960 |
+
"step": 6550
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 3.7585421412300684,
|
| 964 |
+
"grad_norm": 0.04055279493331909,
|
| 965 |
+
"learning_rate": 9.272697689554181e-06,
|
| 966 |
+
"loss": 0.0195,
|
| 967 |
+
"step": 6600
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 3.787015945330296,
|
| 971 |
+
"grad_norm": 0.5971085429191589,
|
| 972 |
+
"learning_rate": 9.191343963553533e-06,
|
| 973 |
+
"loss": 0.0156,
|
| 974 |
+
"step": 6650
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 3.8154897494305238,
|
| 978 |
+
"grad_norm": 0.2191866636276245,
|
| 979 |
+
"learning_rate": 9.10999023755288e-06,
|
| 980 |
+
"loss": 0.0174,
|
| 981 |
+
"step": 6700
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 3.8439635535307515,
|
| 985 |
+
"grad_norm": 0.033559828996658325,
|
| 986 |
+
"learning_rate": 9.02863651155223e-06,
|
| 987 |
+
"loss": 0.0149,
|
| 988 |
+
"step": 6750
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 3.8724373576309796,
|
| 992 |
+
"grad_norm": 3.7128634452819824,
|
| 993 |
+
"learning_rate": 8.947282785551578e-06,
|
| 994 |
+
"loss": 0.0221,
|
| 995 |
+
"step": 6800
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 3.9009111617312073,
|
| 999 |
+
"grad_norm": 0.08547580987215042,
|
| 1000 |
+
"learning_rate": 8.865929059550928e-06,
|
| 1001 |
+
"loss": 0.0153,
|
| 1002 |
+
"step": 6850
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 3.929384965831435,
|
| 1006 |
+
"grad_norm": 0.058124080300331116,
|
| 1007 |
+
"learning_rate": 8.784575333550277e-06,
|
| 1008 |
+
"loss": 0.0161,
|
| 1009 |
+
"step": 6900
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 3.9578587699316627,
|
| 1013 |
+
"grad_norm": 0.015579139813780785,
|
| 1014 |
+
"learning_rate": 8.703221607549627e-06,
|
| 1015 |
+
"loss": 0.011,
|
| 1016 |
+
"step": 6950
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 3.9863325740318905,
|
| 1020 |
+
"grad_norm": 0.12764935195446014,
|
| 1021 |
+
"learning_rate": 8.621867881548975e-06,
|
| 1022 |
+
"loss": 0.0166,
|
| 1023 |
+
"step": 7000
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 4.0,
|
| 1027 |
+
"eval_f1": 0.9525329748802823,
|
| 1028 |
+
"eval_loss": 0.04983380436897278,
|
| 1029 |
+
"eval_precision": 0.9510149303808086,
|
| 1030 |
+
"eval_recall": 0.9540558734432851,
|
| 1031 |
+
"eval_runtime": 3.7495,
|
| 1032 |
+
"eval_samples_per_second": 866.79,
|
| 1033 |
+
"eval_steps_per_second": 108.549,
|
| 1034 |
+
"step": 7024
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 4.014806378132119,
|
| 1038 |
+
"grad_norm": 0.04521024227142334,
|
| 1039 |
+
"learning_rate": 8.540514155548325e-06,
|
| 1040 |
+
"loss": 0.0102,
|
| 1041 |
+
"step": 7050
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 4.043280182232346,
|
| 1045 |
+
"grad_norm": 0.015535669401288033,
|
| 1046 |
+
"learning_rate": 8.459160429547674e-06,
|
| 1047 |
+
"loss": 0.0068,
|
| 1048 |
+
"step": 7100
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 4.071753986332574,
|
| 1052 |
+
"grad_norm": 12.566010475158691,
|
| 1053 |
+
"learning_rate": 8.377806703547024e-06,
|
| 1054 |
+
"loss": 0.0151,
|
| 1055 |
+
"step": 7150
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 4.100227790432802,
|
| 1059 |
+
"grad_norm": 0.027761396020650864,
|
| 1060 |
+
"learning_rate": 8.296452977546372e-06,
|
| 1061 |
+
"loss": 0.0053,
|
| 1062 |
+
"step": 7200
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 4.128701594533029,
|
| 1066 |
+
"grad_norm": 1.8846938610076904,
|
| 1067 |
+
"learning_rate": 8.215099251545722e-06,
|
| 1068 |
+
"loss": 0.0077,
|
| 1069 |
+
"step": 7250
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 4.157175398633258,
|
| 1073 |
+
"grad_norm": 0.005013479385524988,
|
| 1074 |
+
"learning_rate": 8.13374552554507e-06,
|
| 1075 |
+
"loss": 0.0073,
|
| 1076 |
+
"step": 7300
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 4.185649202733485,
|
| 1080 |
+
"grad_norm": 0.32696235179901123,
|
| 1081 |
+
"learning_rate": 8.052391799544421e-06,
|
| 1082 |
+
"loss": 0.0092,
|
| 1083 |
+
"step": 7350
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 4.214123006833713,
|
| 1087 |
+
"grad_norm": 0.014357910491526127,
|
| 1088 |
+
"learning_rate": 7.971038073543769e-06,
|
| 1089 |
+
"loss": 0.0164,
|
| 1090 |
+
"step": 7400
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 4.242596810933941,
|
| 1094 |
+
"grad_norm": 0.010845329612493515,
|
| 1095 |
+
"learning_rate": 7.889684347543118e-06,
|
| 1096 |
+
"loss": 0.0123,
|
| 1097 |
+
"step": 7450
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 4.271070615034168,
|
| 1101 |
+
"grad_norm": 0.6969354748725891,
|
| 1102 |
+
"learning_rate": 7.808330621542466e-06,
|
| 1103 |
+
"loss": 0.0103,
|
| 1104 |
+
"step": 7500
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 4.2995444191343966,
|
| 1108 |
+
"grad_norm": 0.3575742840766907,
|
| 1109 |
+
"learning_rate": 7.726976895541816e-06,
|
| 1110 |
+
"loss": 0.0161,
|
| 1111 |
+
"step": 7550
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 4.328018223234624,
|
| 1115 |
+
"grad_norm": 0.008076228201389313,
|
| 1116 |
+
"learning_rate": 7.645623169541166e-06,
|
| 1117 |
+
"loss": 0.0122,
|
| 1118 |
+
"step": 7600
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 4.356492027334852,
|
| 1122 |
+
"grad_norm": 0.029101597145199776,
|
| 1123 |
+
"learning_rate": 7.564269443540515e-06,
|
| 1124 |
+
"loss": 0.0071,
|
| 1125 |
+
"step": 7650
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 4.38496583143508,
|
| 1129 |
+
"grad_norm": 0.06841142475605011,
|
| 1130 |
+
"learning_rate": 7.482915717539863e-06,
|
| 1131 |
+
"loss": 0.0091,
|
| 1132 |
+
"step": 7700
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 4.413439635535307,
|
| 1136 |
+
"grad_norm": 0.04133535176515579,
|
| 1137 |
+
"learning_rate": 7.401561991539213e-06,
|
| 1138 |
+
"loss": 0.0073,
|
| 1139 |
+
"step": 7750
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 4.4419134396355355,
|
| 1143 |
+
"grad_norm": 0.03571132943034172,
|
| 1144 |
+
"learning_rate": 7.320208265538562e-06,
|
| 1145 |
+
"loss": 0.0131,
|
| 1146 |
+
"step": 7800
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 4.470387243735763,
|
| 1150 |
+
"grad_norm": 2.8286924362182617,
|
| 1151 |
+
"learning_rate": 7.2388545395379114e-06,
|
| 1152 |
+
"loss": 0.0043,
|
| 1153 |
+
"step": 7850
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 4.498861047835991,
|
| 1157 |
+
"grad_norm": 1.3785158395767212,
|
| 1158 |
+
"learning_rate": 7.15750081353726e-06,
|
| 1159 |
+
"loss": 0.0103,
|
| 1160 |
+
"step": 7900
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 4.527334851936219,
|
| 1164 |
+
"grad_norm": 0.05979786813259125,
|
| 1165 |
+
"learning_rate": 7.07614708753661e-06,
|
| 1166 |
+
"loss": 0.006,
|
| 1167 |
+
"step": 7950
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 4.555808656036446,
|
| 1171 |
+
"grad_norm": 1.3434393405914307,
|
| 1172 |
+
"learning_rate": 6.994793361535959e-06,
|
| 1173 |
+
"loss": 0.0148,
|
| 1174 |
+
"step": 8000
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 4.5842824601366745,
|
| 1178 |
+
"grad_norm": 0.019605603069067,
|
| 1179 |
+
"learning_rate": 6.913439635535308e-06,
|
| 1180 |
+
"loss": 0.0068,
|
| 1181 |
+
"step": 8050
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 4.612756264236902,
|
| 1185 |
+
"grad_norm": 0.0028891051188111305,
|
| 1186 |
+
"learning_rate": 6.832085909534657e-06,
|
| 1187 |
+
"loss": 0.0075,
|
| 1188 |
+
"step": 8100
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 4.64123006833713,
|
| 1192 |
+
"grad_norm": 0.014856363646686077,
|
| 1193 |
+
"learning_rate": 6.750732183534007e-06,
|
| 1194 |
+
"loss": 0.0073,
|
| 1195 |
+
"step": 8150
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 4.669703872437358,
|
| 1199 |
+
"grad_norm": 0.0037149768322706223,
|
| 1200 |
+
"learning_rate": 6.669378457533356e-06,
|
| 1201 |
+
"loss": 0.0066,
|
| 1202 |
+
"step": 8200
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 4.698177676537585,
|
| 1206 |
+
"grad_norm": 0.008534993045032024,
|
| 1207 |
+
"learning_rate": 6.588024731532705e-06,
|
| 1208 |
+
"loss": 0.0121,
|
| 1209 |
+
"step": 8250
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 4.7266514806378135,
|
| 1213 |
+
"grad_norm": 0.013920712284743786,
|
| 1214 |
+
"learning_rate": 6.506671005532054e-06,
|
| 1215 |
+
"loss": 0.013,
|
| 1216 |
+
"step": 8300
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 4.755125284738041,
|
| 1220 |
+
"grad_norm": 0.10121840238571167,
|
| 1221 |
+
"learning_rate": 6.425317279531404e-06,
|
| 1222 |
+
"loss": 0.0087,
|
| 1223 |
+
"step": 8350
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 4.783599088838269,
|
| 1227 |
+
"grad_norm": 0.07937771081924438,
|
| 1228 |
+
"learning_rate": 6.343963553530752e-06,
|
| 1229 |
+
"loss": 0.0086,
|
| 1230 |
+
"step": 8400
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 4.812072892938497,
|
| 1234 |
+
"grad_norm": 0.07674901187419891,
|
| 1235 |
+
"learning_rate": 6.262609827530101e-06,
|
| 1236 |
+
"loss": 0.0052,
|
| 1237 |
+
"step": 8450
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 4.840546697038724,
|
| 1241 |
+
"grad_norm": 0.06683178246021271,
|
| 1242 |
+
"learning_rate": 6.18125610152945e-06,
|
| 1243 |
+
"loss": 0.015,
|
| 1244 |
+
"step": 8500
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 4.8690205011389525,
|
| 1248 |
+
"grad_norm": 0.08861212432384491,
|
| 1249 |
+
"learning_rate": 6.0999023755288e-06,
|
| 1250 |
+
"loss": 0.0109,
|
| 1251 |
+
"step": 8550
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 4.89749430523918,
|
| 1255 |
+
"grad_norm": 0.006198623217642307,
|
| 1256 |
+
"learning_rate": 6.018548649528149e-06,
|
| 1257 |
+
"loss": 0.0181,
|
| 1258 |
+
"step": 8600
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 4.925968109339408,
|
| 1262 |
+
"grad_norm": 0.009029284119606018,
|
| 1263 |
+
"learning_rate": 5.937194923527498e-06,
|
| 1264 |
+
"loss": 0.0043,
|
| 1265 |
+
"step": 8650
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 4.954441913439636,
|
| 1269 |
+
"grad_norm": 1.1225602626800537,
|
| 1270 |
+
"learning_rate": 5.855841197526847e-06,
|
| 1271 |
+
"loss": 0.0121,
|
| 1272 |
+
"step": 8700
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 4.982915717539863,
|
| 1276 |
+
"grad_norm": 0.01218325924128294,
|
| 1277 |
+
"learning_rate": 5.774487471526197e-06,
|
| 1278 |
+
"loss": 0.0115,
|
| 1279 |
+
"step": 8750
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 5.0,
|
| 1283 |
+
"eval_f1": 0.9533444816053511,
|
| 1284 |
+
"eval_loss": 0.04722925275564194,
|
| 1285 |
+
"eval_precision": 0.9473246925889,
|
| 1286 |
+
"eval_recall": 0.9594412655671491,
|
| 1287 |
+
"eval_runtime": 3.7358,
|
| 1288 |
+
"eval_samples_per_second": 869.968,
|
| 1289 |
+
"eval_steps_per_second": 108.947,
|
| 1290 |
+
"step": 8780
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.011389521640091,
|
| 1294 |
+
"grad_norm": 0.8161097764968872,
|
| 1295 |
+
"learning_rate": 5.6931337455255455e-06,
|
| 1296 |
+
"loss": 0.0127,
|
| 1297 |
+
"step": 8800
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.039863325740319,
|
| 1301 |
+
"grad_norm": 0.048089127987623215,
|
| 1302 |
+
"learning_rate": 5.611780019524895e-06,
|
| 1303 |
+
"loss": 0.0116,
|
| 1304 |
+
"step": 8850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 5.068337129840547,
|
| 1308 |
+
"grad_norm": 1.7576072216033936,
|
| 1309 |
+
"learning_rate": 5.530426293524244e-06,
|
| 1310 |
+
"loss": 0.0093,
|
| 1311 |
+
"step": 8900
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 5.096810933940774,
|
| 1315 |
+
"grad_norm": 0.07270015776157379,
|
| 1316 |
+
"learning_rate": 5.449072567523594e-06,
|
| 1317 |
+
"loss": 0.0043,
|
| 1318 |
+
"step": 8950
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 5.125284738041002,
|
| 1322 |
+
"grad_norm": 0.2580782175064087,
|
| 1323 |
+
"learning_rate": 5.367718841522942e-06,
|
| 1324 |
+
"loss": 0.0058,
|
| 1325 |
+
"step": 9000
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 5.15375854214123,
|
| 1329 |
+
"grad_norm": 0.12649740278720856,
|
| 1330 |
+
"learning_rate": 5.286365115522292e-06,
|
| 1331 |
+
"loss": 0.0033,
|
| 1332 |
+
"step": 9050
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 5.182232346241458,
|
| 1336 |
+
"grad_norm": 0.06850716471672058,
|
| 1337 |
+
"learning_rate": 5.20501138952164e-06,
|
| 1338 |
+
"loss": 0.003,
|
| 1339 |
+
"step": 9100
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 5.210706150341686,
|
| 1343 |
+
"grad_norm": 0.01197959017008543,
|
| 1344 |
+
"learning_rate": 5.12365766352099e-06,
|
| 1345 |
+
"loss": 0.0029,
|
| 1346 |
+
"step": 9150
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 5.239179954441913,
|
| 1350 |
+
"grad_norm": 0.10818086564540863,
|
| 1351 |
+
"learning_rate": 5.0423039375203385e-06,
|
| 1352 |
+
"loss": 0.0025,
|
| 1353 |
+
"step": 9200
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 5.267653758542141,
|
| 1357 |
+
"grad_norm": 0.035460665822029114,
|
| 1358 |
+
"learning_rate": 4.960950211519688e-06,
|
| 1359 |
+
"loss": 0.0047,
|
| 1360 |
+
"step": 9250
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 5.296127562642369,
|
| 1364 |
+
"grad_norm": 4.532562255859375,
|
| 1365 |
+
"learning_rate": 4.879596485519037e-06,
|
| 1366 |
+
"loss": 0.0086,
|
| 1367 |
+
"step": 9300
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 5.324601366742597,
|
| 1371 |
+
"grad_norm": 0.013236219063401222,
|
| 1372 |
+
"learning_rate": 4.7982427595183866e-06,
|
| 1373 |
+
"loss": 0.0027,
|
| 1374 |
+
"step": 9350
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 5.353075170842825,
|
| 1378 |
+
"grad_norm": 0.018059909343719482,
|
| 1379 |
+
"learning_rate": 4.716889033517735e-06,
|
| 1380 |
+
"loss": 0.0065,
|
| 1381 |
+
"step": 9400
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 5.381548974943052,
|
| 1385 |
+
"grad_norm": 0.019096272066235542,
|
| 1386 |
+
"learning_rate": 4.635535307517084e-06,
|
| 1387 |
+
"loss": 0.0053,
|
| 1388 |
+
"step": 9450
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 5.41002277904328,
|
| 1392 |
+
"grad_norm": 0.10389436781406403,
|
| 1393 |
+
"learning_rate": 4.554181581516434e-06,
|
| 1394 |
+
"loss": 0.0029,
|
| 1395 |
+
"step": 9500
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 5.438496583143508,
|
| 1399 |
+
"grad_norm": 0.0029252381063997746,
|
| 1400 |
+
"learning_rate": 4.472827855515783e-06,
|
| 1401 |
+
"loss": 0.0045,
|
| 1402 |
+
"step": 9550
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 5.466970387243736,
|
| 1406 |
+
"grad_norm": 0.004542892333120108,
|
| 1407 |
+
"learning_rate": 4.391474129515132e-06,
|
| 1408 |
+
"loss": 0.0102,
|
| 1409 |
+
"step": 9600
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 5.495444191343964,
|
| 1413 |
+
"grad_norm": 0.130916565656662,
|
| 1414 |
+
"learning_rate": 4.310120403514481e-06,
|
| 1415 |
+
"loss": 0.0064,
|
| 1416 |
+
"step": 9650
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 5.523917995444191,
|
| 1420 |
+
"grad_norm": 0.008013393729925156,
|
| 1421 |
+
"learning_rate": 4.22876667751383e-06,
|
| 1422 |
+
"loss": 0.005,
|
| 1423 |
+
"step": 9700
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 5.552391799544419,
|
| 1427 |
+
"grad_norm": 0.012782514095306396,
|
| 1428 |
+
"learning_rate": 4.1474129515131795e-06,
|
| 1429 |
+
"loss": 0.0122,
|
| 1430 |
+
"step": 9750
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 5.5808656036446465,
|
| 1434 |
+
"grad_norm": 0.9617053866386414,
|
| 1435 |
+
"learning_rate": 4.066059225512528e-06,
|
| 1436 |
+
"loss": 0.0063,
|
| 1437 |
+
"step": 9800
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 5.609339407744875,
|
| 1441 |
+
"grad_norm": 0.014953136444091797,
|
| 1442 |
+
"learning_rate": 3.984705499511878e-06,
|
| 1443 |
+
"loss": 0.0046,
|
| 1444 |
+
"step": 9850
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 5.637813211845103,
|
| 1448 |
+
"grad_norm": 1.0275633335113525,
|
| 1449 |
+
"learning_rate": 3.903351773511227e-06,
|
| 1450 |
+
"loss": 0.0075,
|
| 1451 |
+
"step": 9900
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 5.66628701594533,
|
| 1455 |
+
"grad_norm": 0.007999264635145664,
|
| 1456 |
+
"learning_rate": 3.8219980475105764e-06,
|
| 1457 |
+
"loss": 0.0031,
|
| 1458 |
+
"step": 9950
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 5.694760820045558,
|
| 1462 |
+
"grad_norm": 0.008241960778832436,
|
| 1463 |
+
"learning_rate": 3.7406443215099252e-06,
|
| 1464 |
+
"loss": 0.0065,
|
| 1465 |
+
"step": 10000
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 5.723234624145785,
|
| 1469 |
+
"grad_norm": 2.578386068344116,
|
| 1470 |
+
"learning_rate": 3.6592905955092745e-06,
|
| 1471 |
+
"loss": 0.0057,
|
| 1472 |
+
"step": 10050
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 5.751708428246014,
|
| 1476 |
+
"grad_norm": 2.8855443000793457,
|
| 1477 |
+
"learning_rate": 3.5779368695086237e-06,
|
| 1478 |
+
"loss": 0.0038,
|
| 1479 |
+
"step": 10100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 5.780182232346242,
|
| 1483 |
+
"grad_norm": 0.016262667253613472,
|
| 1484 |
+
"learning_rate": 3.496583143507973e-06,
|
| 1485 |
+
"loss": 0.0128,
|
| 1486 |
+
"step": 10150
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 5.808656036446469,
|
| 1490 |
+
"grad_norm": 0.045168060809373856,
|
| 1491 |
+
"learning_rate": 3.415229417507322e-06,
|
| 1492 |
+
"loss": 0.0036,
|
| 1493 |
+
"step": 10200
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 5.837129840546697,
|
| 1497 |
+
"grad_norm": 0.1192622259259224,
|
| 1498 |
+
"learning_rate": 3.3355027660266842e-06,
|
| 1499 |
+
"loss": 0.0111,
|
| 1500 |
+
"step": 10250
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 5.865603644646924,
|
| 1504 |
+
"grad_norm": 0.08544190227985382,
|
| 1505 |
+
"learning_rate": 3.2541490400260335e-06,
|
| 1506 |
+
"loss": 0.0051,
|
| 1507 |
+
"step": 10300
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 5.894077448747153,
|
| 1511 |
+
"grad_norm": 3.493161678314209,
|
| 1512 |
+
"learning_rate": 3.1727953140253827e-06,
|
| 1513 |
+
"loss": 0.0054,
|
| 1514 |
+
"step": 10350
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 5.922551252847381,
|
| 1518 |
+
"grad_norm": 0.022789066657423973,
|
| 1519 |
+
"learning_rate": 3.0914415880247315e-06,
|
| 1520 |
+
"loss": 0.0033,
|
| 1521 |
+
"step": 10400
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 5.951025056947608,
|
| 1525 |
+
"grad_norm": 12.131625175476074,
|
| 1526 |
+
"learning_rate": 3.0100878620240807e-06,
|
| 1527 |
+
"loss": 0.0023,
|
| 1528 |
+
"step": 10450
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 5.979498861047836,
|
| 1532 |
+
"grad_norm": 0.03329641371965408,
|
| 1533 |
+
"learning_rate": 2.92873413602343e-06,
|
| 1534 |
+
"loss": 0.0058,
|
| 1535 |
+
"step": 10500
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 6.0,
|
| 1539 |
+
"eval_f1": 0.9557566616390145,
|
| 1540 |
+
"eval_loss": 0.05410688370466232,
|
| 1541 |
+
"eval_precision": 0.9517690253671562,
|
| 1542 |
+
"eval_recall": 0.9597778525748906,
|
| 1543 |
+
"eval_runtime": 3.6572,
|
| 1544 |
+
"eval_samples_per_second": 888.662,
|
| 1545 |
+
"eval_steps_per_second": 111.288,
|
| 1546 |
+
"step": 10536
|
| 1547 |
+
}
|
| 1548 |
+
],
|
| 1549 |
+
"logging_steps": 50,
|
| 1550 |
+
"max_steps": 12292,
|
| 1551 |
+
"num_input_tokens_seen": 0,
|
| 1552 |
+
"num_train_epochs": 7,
|
| 1553 |
+
"save_steps": 500,
|
| 1554 |
+
"stateful_callbacks": {
|
| 1555 |
+
"TrainerControl": {
|
| 1556 |
+
"args": {
|
| 1557 |
+
"should_epoch_stop": false,
|
| 1558 |
+
"should_evaluate": false,
|
| 1559 |
+
"should_log": false,
|
| 1560 |
+
"should_save": true,
|
| 1561 |
+
"should_training_stop": false
|
| 1562 |
+
},
|
| 1563 |
+
"attributes": {}
|
| 1564 |
+
}
|
| 1565 |
+
},
|
| 1566 |
+
"total_flos": 2072296586745846.0,
|
| 1567 |
+
"train_batch_size": 8,
|
| 1568 |
+
"trial_name": null,
|
| 1569 |
+
"trial_params": null
|
| 1570 |
+
}
|
checkpoint-10536/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:15e08f60c5260f286c88c6054e532ee0950cb5316081a3a51d07b57f6a266f71
|
| 3 |
+
size 5368
|
checkpoint-12292/config.json
ADDED
|
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"XLMRobertaForTokenClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "O",
|
| 14 |
+
"1": "B-PER",
|
| 15 |
+
"2": "I-PER",
|
| 16 |
+
"3": "B-ORG",
|
| 17 |
+
"4": "I-ORG",
|
| 18 |
+
"5": "B-LOC",
|
| 19 |
+
"6": "I-LOC",
|
| 20 |
+
"7": "B-MISC",
|
| 21 |
+
"8": "I-MISC"
|
| 22 |
+
},
|
| 23 |
+
"initializer_range": 0.02,
|
| 24 |
+
"intermediate_size": 3072,
|
| 25 |
+
"label2id": {
|
| 26 |
+
"B-LOC": 5,
|
| 27 |
+
"B-MISC": 7,
|
| 28 |
+
"B-ORG": 3,
|
| 29 |
+
"B-PER": 1,
|
| 30 |
+
"I-LOC": 6,
|
| 31 |
+
"I-MISC": 8,
|
| 32 |
+
"I-ORG": 4,
|
| 33 |
+
"I-PER": 2,
|
| 34 |
+
"O": 0
|
| 35 |
+
},
|
| 36 |
+
"layer_norm_eps": 1e-05,
|
| 37 |
+
"max_position_embeddings": 514,
|
| 38 |
+
"model_type": "xlm-roberta",
|
| 39 |
+
"num_attention_heads": 12,
|
| 40 |
+
"num_hidden_layers": 12,
|
| 41 |
+
"output_past": true,
|
| 42 |
+
"pad_token_id": 1,
|
| 43 |
+
"position_embedding_type": "absolute",
|
| 44 |
+
"torch_dtype": "float32",
|
| 45 |
+
"transformers_version": "4.51.3",
|
| 46 |
+
"type_vocab_size": 1,
|
| 47 |
+
"use_cache": true,
|
| 48 |
+
"vocab_size": 250002
|
| 49 |
+
}
|
checkpoint-12292/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9f3aa01fa5d49602728eca2190fb610b286d4348c8e470e053f83959dfd91049
|
| 3 |
+
size 1109863956
|
checkpoint-12292/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:328145ba117ef4052abfc88cfdc5ed7cb46c854f712b8a60beb15d66d8807396
|
| 3 |
+
size 2219846202
|
checkpoint-12292/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ab0a0c2e05c9b19ae2fb7f37ade343bd0d98b553cbd00cc47aa20772252004c6
|
| 3 |
+
size 14244
|
checkpoint-12292/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6509928afa01d668bf56b569c4f6a2e6c2fde9763d17681c72b55f19ab92eb20
|
| 3 |
+
size 988
|
checkpoint-12292/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8f42e34b774989a80cf2635a24272d37af726ec5ad847c7648079f6697e6b0d3
|
| 3 |
+
size 1064
|
checkpoint-12292/special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": {
|
| 6 |
+
"content": "<mask>",
|
| 7 |
+
"lstrip": true,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
checkpoint-12292/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3a56def25aa40facc030ea8b0b87f3688e4b3c39eb8b45d5702b3a1300fe2a20
|
| 3 |
+
size 17082734
|
checkpoint-12292/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"250001": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": true,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 512,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizer",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
checkpoint-12292/trainer_state.json
ADDED
|
@@ -0,0 +1,1826 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 12292,
|
| 3 |
+
"best_metric": 0.9563834240267894,
|
| 4 |
+
"best_model_checkpoint": "C:\\Users\\Shara\\projects\\models\\run_16-lr_2e-05-acc_1-wd_0.01-bs_8-ep_7\\checkpoint-12292",
|
| 5 |
+
"epoch": 7.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 12292,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.02847380410022779,
|
| 14 |
+
"grad_norm": 4.8302321434021,
|
| 15 |
+
"learning_rate": 1.992840872111943e-05,
|
| 16 |
+
"loss": 0.9994,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.05694760820045558,
|
| 21 |
+
"grad_norm": 3.4978649616241455,
|
| 22 |
+
"learning_rate": 1.9847054995118778e-05,
|
| 23 |
+
"loss": 0.3939,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.08542141230068337,
|
| 28 |
+
"grad_norm": 8.353672981262207,
|
| 29 |
+
"learning_rate": 1.9765701269118128e-05,
|
| 30 |
+
"loss": 0.2882,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.11389521640091116,
|
| 35 |
+
"grad_norm": 9.822871208190918,
|
| 36 |
+
"learning_rate": 1.9684347543117477e-05,
|
| 37 |
+
"loss": 0.1728,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.14236902050113895,
|
| 42 |
+
"grad_norm": 3.1411221027374268,
|
| 43 |
+
"learning_rate": 1.9602993817116827e-05,
|
| 44 |
+
"loss": 0.1251,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.17084282460136674,
|
| 49 |
+
"grad_norm": 8.493424415588379,
|
| 50 |
+
"learning_rate": 1.9521640091116173e-05,
|
| 51 |
+
"loss": 0.1405,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.19931662870159453,
|
| 56 |
+
"grad_norm": 4.757007122039795,
|
| 57 |
+
"learning_rate": 1.9440286365115523e-05,
|
| 58 |
+
"loss": 0.1431,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.22779043280182232,
|
| 63 |
+
"grad_norm": 4.339049816131592,
|
| 64 |
+
"learning_rate": 1.9358932639114873e-05,
|
| 65 |
+
"loss": 0.0912,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.25626423690205014,
|
| 70 |
+
"grad_norm": 14.540772438049316,
|
| 71 |
+
"learning_rate": 1.9277578913114222e-05,
|
| 72 |
+
"loss": 0.1048,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.2847380410022779,
|
| 77 |
+
"grad_norm": 10.503073692321777,
|
| 78 |
+
"learning_rate": 1.9196225187113572e-05,
|
| 79 |
+
"loss": 0.1224,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.3132118451025057,
|
| 84 |
+
"grad_norm": 2.738131284713745,
|
| 85 |
+
"learning_rate": 1.911487146111292e-05,
|
| 86 |
+
"loss": 0.1025,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.3416856492027335,
|
| 91 |
+
"grad_norm": 6.681102752685547,
|
| 92 |
+
"learning_rate": 1.903351773511227e-05,
|
| 93 |
+
"loss": 0.0869,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.3701594533029613,
|
| 98 |
+
"grad_norm": 1.4330339431762695,
|
| 99 |
+
"learning_rate": 1.895216400911162e-05,
|
| 100 |
+
"loss": 0.052,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.39863325740318906,
|
| 105 |
+
"grad_norm": 2.581470012664795,
|
| 106 |
+
"learning_rate": 1.8870810283110967e-05,
|
| 107 |
+
"loss": 0.0808,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.4271070615034169,
|
| 112 |
+
"grad_norm": 5.825100898742676,
|
| 113 |
+
"learning_rate": 1.8789456557110317e-05,
|
| 114 |
+
"loss": 0.0716,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.45558086560364464,
|
| 119 |
+
"grad_norm": 1.74166738986969,
|
| 120 |
+
"learning_rate": 1.8708102831109666e-05,
|
| 121 |
+
"loss": 0.0806,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.48405466970387245,
|
| 126 |
+
"grad_norm": 7.3097405433654785,
|
| 127 |
+
"learning_rate": 1.8626749105109016e-05,
|
| 128 |
+
"loss": 0.0666,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.5125284738041003,
|
| 133 |
+
"grad_norm": 2.220766067504883,
|
| 134 |
+
"learning_rate": 1.8545395379108362e-05,
|
| 135 |
+
"loss": 0.0768,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.541002277904328,
|
| 140 |
+
"grad_norm": 4.0566511154174805,
|
| 141 |
+
"learning_rate": 1.8464041653107715e-05,
|
| 142 |
+
"loss": 0.0707,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.5694760820045558,
|
| 147 |
+
"grad_norm": 0.2833240330219269,
|
| 148 |
+
"learning_rate": 1.8382687927107065e-05,
|
| 149 |
+
"loss": 0.0858,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.5979498861047836,
|
| 154 |
+
"grad_norm": 21.131481170654297,
|
| 155 |
+
"learning_rate": 1.8301334201106415e-05,
|
| 156 |
+
"loss": 0.0595,
|
| 157 |
+
"step": 1050
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.6264236902050114,
|
| 161 |
+
"grad_norm": 27.655433654785156,
|
| 162 |
+
"learning_rate": 1.821998047510576e-05,
|
| 163 |
+
"loss": 0.0466,
|
| 164 |
+
"step": 1100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.6548974943052391,
|
| 168 |
+
"grad_norm": 3.4721717834472656,
|
| 169 |
+
"learning_rate": 1.813862674910511e-05,
|
| 170 |
+
"loss": 0.0776,
|
| 171 |
+
"step": 1150
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.683371298405467,
|
| 175 |
+
"grad_norm": 2.4811441898345947,
|
| 176 |
+
"learning_rate": 1.805727302310446e-05,
|
| 177 |
+
"loss": 0.0647,
|
| 178 |
+
"step": 1200
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.7118451025056948,
|
| 182 |
+
"grad_norm": 6.169066429138184,
|
| 183 |
+
"learning_rate": 1.797591929710381e-05,
|
| 184 |
+
"loss": 0.0724,
|
| 185 |
+
"step": 1250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.7403189066059226,
|
| 189 |
+
"grad_norm": 8.984748840332031,
|
| 190 |
+
"learning_rate": 1.7894565571103156e-05,
|
| 191 |
+
"loss": 0.0838,
|
| 192 |
+
"step": 1300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.7687927107061503,
|
| 196 |
+
"grad_norm": 5.944505214691162,
|
| 197 |
+
"learning_rate": 1.7813211845102506e-05,
|
| 198 |
+
"loss": 0.0683,
|
| 199 |
+
"step": 1350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.7972665148063781,
|
| 203 |
+
"grad_norm": 0.28812381625175476,
|
| 204 |
+
"learning_rate": 1.7731858119101856e-05,
|
| 205 |
+
"loss": 0.0614,
|
| 206 |
+
"step": 1400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.8257403189066059,
|
| 210 |
+
"grad_norm": 6.176011085510254,
|
| 211 |
+
"learning_rate": 1.7650504393101205e-05,
|
| 212 |
+
"loss": 0.0615,
|
| 213 |
+
"step": 1450
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.8542141230068337,
|
| 217 |
+
"grad_norm": 2.5244405269622803,
|
| 218 |
+
"learning_rate": 1.7569150667100555e-05,
|
| 219 |
+
"loss": 0.0714,
|
| 220 |
+
"step": 1500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.8826879271070615,
|
| 224 |
+
"grad_norm": 3.3508074283599854,
|
| 225 |
+
"learning_rate": 1.7487796941099904e-05,
|
| 226 |
+
"loss": 0.0563,
|
| 227 |
+
"step": 1550
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.9111617312072893,
|
| 231 |
+
"grad_norm": 9.980842590332031,
|
| 232 |
+
"learning_rate": 1.7406443215099254e-05,
|
| 233 |
+
"loss": 0.0647,
|
| 234 |
+
"step": 1600
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.9396355353075171,
|
| 238 |
+
"grad_norm": 3.6959774494171143,
|
| 239 |
+
"learning_rate": 1.7325089489098604e-05,
|
| 240 |
+
"loss": 0.0682,
|
| 241 |
+
"step": 1650
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.9681093394077449,
|
| 245 |
+
"grad_norm": 0.626620888710022,
|
| 246 |
+
"learning_rate": 1.724373576309795e-05,
|
| 247 |
+
"loss": 0.0499,
|
| 248 |
+
"step": 1700
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.9965831435079726,
|
| 252 |
+
"grad_norm": 7.145672798156738,
|
| 253 |
+
"learning_rate": 1.71623820370973e-05,
|
| 254 |
+
"loss": 0.0569,
|
| 255 |
+
"step": 1750
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.0,
|
| 259 |
+
"eval_f1": 0.9242246437552389,
|
| 260 |
+
"eval_loss": 0.05600914731621742,
|
| 261 |
+
"eval_precision": 0.9206746826987308,
|
| 262 |
+
"eval_recall": 0.927802086839448,
|
| 263 |
+
"eval_runtime": 3.7538,
|
| 264 |
+
"eval_samples_per_second": 865.798,
|
| 265 |
+
"eval_steps_per_second": 108.425,
|
| 266 |
+
"step": 1756
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.0250569476082005,
|
| 270 |
+
"grad_norm": 4.280463695526123,
|
| 271 |
+
"learning_rate": 1.708102831109665e-05,
|
| 272 |
+
"loss": 0.0673,
|
| 273 |
+
"step": 1800
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.0535307517084282,
|
| 277 |
+
"grad_norm": 0.06692535430192947,
|
| 278 |
+
"learning_rate": 1.6999674585096e-05,
|
| 279 |
+
"loss": 0.0403,
|
| 280 |
+
"step": 1850
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 1.082004555808656,
|
| 284 |
+
"grad_norm": 0.03789375349879265,
|
| 285 |
+
"learning_rate": 1.691832085909535e-05,
|
| 286 |
+
"loss": 0.0496,
|
| 287 |
+
"step": 1900
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.1104783599088839,
|
| 291 |
+
"grad_norm": 3.698348045349121,
|
| 292 |
+
"learning_rate": 1.6836967133094698e-05,
|
| 293 |
+
"loss": 0.0381,
|
| 294 |
+
"step": 1950
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.1389521640091116,
|
| 298 |
+
"grad_norm": 2.753032684326172,
|
| 299 |
+
"learning_rate": 1.6755613407094048e-05,
|
| 300 |
+
"loss": 0.0432,
|
| 301 |
+
"step": 2000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.1674259681093395,
|
| 305 |
+
"grad_norm": 0.06336130946874619,
|
| 306 |
+
"learning_rate": 1.6674259681093398e-05,
|
| 307 |
+
"loss": 0.0482,
|
| 308 |
+
"step": 2050
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.1958997722095672,
|
| 312 |
+
"grad_norm": 3.9503254890441895,
|
| 313 |
+
"learning_rate": 1.6592905955092744e-05,
|
| 314 |
+
"loss": 0.0436,
|
| 315 |
+
"step": 2100
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.224373576309795,
|
| 319 |
+
"grad_norm": 0.08679741621017456,
|
| 320 |
+
"learning_rate": 1.6511552229092093e-05,
|
| 321 |
+
"loss": 0.0346,
|
| 322 |
+
"step": 2150
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 1.2528473804100229,
|
| 326 |
+
"grad_norm": 2.1995127201080322,
|
| 327 |
+
"learning_rate": 1.6430198503091443e-05,
|
| 328 |
+
"loss": 0.0312,
|
| 329 |
+
"step": 2200
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.2813211845102506,
|
| 333 |
+
"grad_norm": 5.7709808349609375,
|
| 334 |
+
"learning_rate": 1.634884477709079e-05,
|
| 335 |
+
"loss": 0.032,
|
| 336 |
+
"step": 2250
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 1.3097949886104785,
|
| 340 |
+
"grad_norm": 0.2031829059123993,
|
| 341 |
+
"learning_rate": 1.626749105109014e-05,
|
| 342 |
+
"loss": 0.053,
|
| 343 |
+
"step": 2300
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 1.3382687927107062,
|
| 347 |
+
"grad_norm": 4.292157173156738,
|
| 348 |
+
"learning_rate": 1.6186137325089492e-05,
|
| 349 |
+
"loss": 0.0413,
|
| 350 |
+
"step": 2350
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.366742596810934,
|
| 354 |
+
"grad_norm": 0.07821401953697205,
|
| 355 |
+
"learning_rate": 1.6104783599088842e-05,
|
| 356 |
+
"loss": 0.0304,
|
| 357 |
+
"step": 2400
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.3952164009111616,
|
| 361 |
+
"grad_norm": 0.5930206179618835,
|
| 362 |
+
"learning_rate": 1.6023429873088188e-05,
|
| 363 |
+
"loss": 0.0525,
|
| 364 |
+
"step": 2450
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 1.4236902050113895,
|
| 368 |
+
"grad_norm": 1.6114712953567505,
|
| 369 |
+
"learning_rate": 1.5942076147087538e-05,
|
| 370 |
+
"loss": 0.0391,
|
| 371 |
+
"step": 2500
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.4521640091116172,
|
| 375 |
+
"grad_norm": 0.28632065653800964,
|
| 376 |
+
"learning_rate": 1.5860722421086887e-05,
|
| 377 |
+
"loss": 0.0491,
|
| 378 |
+
"step": 2550
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 1.4806378132118452,
|
| 382 |
+
"grad_norm": 1.0017669200897217,
|
| 383 |
+
"learning_rate": 1.5779368695086237e-05,
|
| 384 |
+
"loss": 0.0395,
|
| 385 |
+
"step": 2600
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 1.5091116173120729,
|
| 389 |
+
"grad_norm": 0.03409272059798241,
|
| 390 |
+
"learning_rate": 1.5698014969085583e-05,
|
| 391 |
+
"loss": 0.0399,
|
| 392 |
+
"step": 2650
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.5375854214123006,
|
| 396 |
+
"grad_norm": 3.5108702182769775,
|
| 397 |
+
"learning_rate": 1.5616661243084933e-05,
|
| 398 |
+
"loss": 0.0424,
|
| 399 |
+
"step": 2700
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.5660592255125285,
|
| 403 |
+
"grad_norm": 3.5875349044799805,
|
| 404 |
+
"learning_rate": 1.5535307517084283e-05,
|
| 405 |
+
"loss": 0.0497,
|
| 406 |
+
"step": 2750
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.5945330296127562,
|
| 410 |
+
"grad_norm": 0.10498251020908356,
|
| 411 |
+
"learning_rate": 1.5453953791083632e-05,
|
| 412 |
+
"loss": 0.0461,
|
| 413 |
+
"step": 2800
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.6230068337129842,
|
| 417 |
+
"grad_norm": 7.618963718414307,
|
| 418 |
+
"learning_rate": 1.5374227139602996e-05,
|
| 419 |
+
"loss": 0.0394,
|
| 420 |
+
"step": 2850
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 1.6514806378132119,
|
| 424 |
+
"grad_norm": 3.3920793533325195,
|
| 425 |
+
"learning_rate": 1.5292873413602346e-05,
|
| 426 |
+
"loss": 0.0339,
|
| 427 |
+
"step": 2900
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 1.6799544419134396,
|
| 431 |
+
"grad_norm": 0.04736129939556122,
|
| 432 |
+
"learning_rate": 1.5211519687601692e-05,
|
| 433 |
+
"loss": 0.0358,
|
| 434 |
+
"step": 2950
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.7084282460136673,
|
| 438 |
+
"grad_norm": 0.17463508248329163,
|
| 439 |
+
"learning_rate": 1.5130165961601042e-05,
|
| 440 |
+
"loss": 0.034,
|
| 441 |
+
"step": 3000
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 1.7369020501138952,
|
| 445 |
+
"grad_norm": 0.058148205280303955,
|
| 446 |
+
"learning_rate": 1.5048812235600392e-05,
|
| 447 |
+
"loss": 0.0317,
|
| 448 |
+
"step": 3050
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 1.7653758542141231,
|
| 452 |
+
"grad_norm": 3.5262815952301025,
|
| 453 |
+
"learning_rate": 1.4967458509599741e-05,
|
| 454 |
+
"loss": 0.0477,
|
| 455 |
+
"step": 3100
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 1.7938496583143508,
|
| 459 |
+
"grad_norm": 2.971937894821167,
|
| 460 |
+
"learning_rate": 1.488610478359909e-05,
|
| 461 |
+
"loss": 0.0364,
|
| 462 |
+
"step": 3150
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 1.8223234624145785,
|
| 466 |
+
"grad_norm": 0.05181724950671196,
|
| 467 |
+
"learning_rate": 1.4804751057598439e-05,
|
| 468 |
+
"loss": 0.0394,
|
| 469 |
+
"step": 3200
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 1.8507972665148062,
|
| 473 |
+
"grad_norm": 0.2712370455265045,
|
| 474 |
+
"learning_rate": 1.4723397331597789e-05,
|
| 475 |
+
"loss": 0.0456,
|
| 476 |
+
"step": 3250
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.8792710706150342,
|
| 480 |
+
"grad_norm": 1.608622431755066,
|
| 481 |
+
"learning_rate": 1.4642043605597138e-05,
|
| 482 |
+
"loss": 0.05,
|
| 483 |
+
"step": 3300
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 1.907744874715262,
|
| 487 |
+
"grad_norm": 5.2423529624938965,
|
| 488 |
+
"learning_rate": 1.4560689879596486e-05,
|
| 489 |
+
"loss": 0.0371,
|
| 490 |
+
"step": 3350
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 1.9362186788154898,
|
| 494 |
+
"grad_norm": 0.04786211624741554,
|
| 495 |
+
"learning_rate": 1.4479336153595836e-05,
|
| 496 |
+
"loss": 0.0329,
|
| 497 |
+
"step": 3400
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 1.9646924829157175,
|
| 501 |
+
"grad_norm": 0.5264931917190552,
|
| 502 |
+
"learning_rate": 1.4397982427595185e-05,
|
| 503 |
+
"loss": 0.0379,
|
| 504 |
+
"step": 3450
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 1.9931662870159452,
|
| 508 |
+
"grad_norm": 0.236006498336792,
|
| 509 |
+
"learning_rate": 1.4316628701594535e-05,
|
| 510 |
+
"loss": 0.0253,
|
| 511 |
+
"step": 3500
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 2.0,
|
| 515 |
+
"eval_f1": 0.9406438631790744,
|
| 516 |
+
"eval_loss": 0.05223705992102623,
|
| 517 |
+
"eval_precision": 0.9371867691279653,
|
| 518 |
+
"eval_recall": 0.9441265567149109,
|
| 519 |
+
"eval_runtime": 3.7109,
|
| 520 |
+
"eval_samples_per_second": 875.791,
|
| 521 |
+
"eval_steps_per_second": 109.676,
|
| 522 |
+
"step": 3512
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 2.021640091116173,
|
| 526 |
+
"grad_norm": 4.225963115692139,
|
| 527 |
+
"learning_rate": 1.4235274975593883e-05,
|
| 528 |
+
"loss": 0.0373,
|
| 529 |
+
"step": 3550
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 2.050113895216401,
|
| 533 |
+
"grad_norm": 0.15388111770153046,
|
| 534 |
+
"learning_rate": 1.4153921249593233e-05,
|
| 535 |
+
"loss": 0.0258,
|
| 536 |
+
"step": 3600
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.078587699316629,
|
| 540 |
+
"grad_norm": 6.3911895751953125,
|
| 541 |
+
"learning_rate": 1.4072567523592582e-05,
|
| 542 |
+
"loss": 0.0217,
|
| 543 |
+
"step": 3650
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 2.1070615034168565,
|
| 547 |
+
"grad_norm": 0.03075530007481575,
|
| 548 |
+
"learning_rate": 1.3991213797591932e-05,
|
| 549 |
+
"loss": 0.019,
|
| 550 |
+
"step": 3700
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 2.135535307517084,
|
| 554 |
+
"grad_norm": 0.16612432897090912,
|
| 555 |
+
"learning_rate": 1.3909860071591278e-05,
|
| 556 |
+
"loss": 0.0175,
|
| 557 |
+
"step": 3750
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 2.164009111617312,
|
| 561 |
+
"grad_norm": 0.14543022215366364,
|
| 562 |
+
"learning_rate": 1.382850634559063e-05,
|
| 563 |
+
"loss": 0.0289,
|
| 564 |
+
"step": 3800
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 2.19248291571754,
|
| 568 |
+
"grad_norm": 2.9880359172821045,
|
| 569 |
+
"learning_rate": 1.374715261958998e-05,
|
| 570 |
+
"loss": 0.0208,
|
| 571 |
+
"step": 3850
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 2.2209567198177678,
|
| 575 |
+
"grad_norm": 9.878133773803711,
|
| 576 |
+
"learning_rate": 1.3665798893589329e-05,
|
| 577 |
+
"loss": 0.021,
|
| 578 |
+
"step": 3900
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 2.2494305239179955,
|
| 582 |
+
"grad_norm": 9.684176445007324,
|
| 583 |
+
"learning_rate": 1.3584445167588675e-05,
|
| 584 |
+
"loss": 0.0267,
|
| 585 |
+
"step": 3950
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 2.277904328018223,
|
| 589 |
+
"grad_norm": 33.15859603881836,
|
| 590 |
+
"learning_rate": 1.3503091441588025e-05,
|
| 591 |
+
"loss": 0.0295,
|
| 592 |
+
"step": 4000
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 2.306378132118451,
|
| 596 |
+
"grad_norm": 2.559544086456299,
|
| 597 |
+
"learning_rate": 1.3421737715587374e-05,
|
| 598 |
+
"loss": 0.0292,
|
| 599 |
+
"step": 4050
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 2.334851936218679,
|
| 603 |
+
"grad_norm": 0.07091552764177322,
|
| 604 |
+
"learning_rate": 1.3340383989586724e-05,
|
| 605 |
+
"loss": 0.0262,
|
| 606 |
+
"step": 4100
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.3633257403189067,
|
| 610 |
+
"grad_norm": 1.8119585514068604,
|
| 611 |
+
"learning_rate": 1.3259030263586072e-05,
|
| 612 |
+
"loss": 0.03,
|
| 613 |
+
"step": 4150
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 2.3917995444191344,
|
| 617 |
+
"grad_norm": 0.1476636379957199,
|
| 618 |
+
"learning_rate": 1.3177676537585422e-05,
|
| 619 |
+
"loss": 0.0196,
|
| 620 |
+
"step": 4200
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.420273348519362,
|
| 624 |
+
"grad_norm": 0.04482650011777878,
|
| 625 |
+
"learning_rate": 1.3096322811584771e-05,
|
| 626 |
+
"loss": 0.0263,
|
| 627 |
+
"step": 4250
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 2.44874715261959,
|
| 631 |
+
"grad_norm": 0.09201560169458389,
|
| 632 |
+
"learning_rate": 1.3014969085584121e-05,
|
| 633 |
+
"loss": 0.0172,
|
| 634 |
+
"step": 4300
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 2.477220956719818,
|
| 638 |
+
"grad_norm": 0.02215876244008541,
|
| 639 |
+
"learning_rate": 1.2933615359583469e-05,
|
| 640 |
+
"loss": 0.0149,
|
| 641 |
+
"step": 4350
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 2.5056947608200457,
|
| 645 |
+
"grad_norm": 0.6993932127952576,
|
| 646 |
+
"learning_rate": 1.2852261633582819e-05,
|
| 647 |
+
"loss": 0.0331,
|
| 648 |
+
"step": 4400
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 2.5341685649202734,
|
| 652 |
+
"grad_norm": 0.030453965067863464,
|
| 653 |
+
"learning_rate": 1.2770907907582168e-05,
|
| 654 |
+
"loss": 0.0307,
|
| 655 |
+
"step": 4450
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 2.562642369020501,
|
| 659 |
+
"grad_norm": 0.22211593389511108,
|
| 660 |
+
"learning_rate": 1.2689554181581518e-05,
|
| 661 |
+
"loss": 0.0183,
|
| 662 |
+
"step": 4500
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 2.591116173120729,
|
| 666 |
+
"grad_norm": 1.2842738628387451,
|
| 667 |
+
"learning_rate": 1.2608200455580866e-05,
|
| 668 |
+
"loss": 0.0226,
|
| 669 |
+
"step": 4550
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 2.619589977220957,
|
| 673 |
+
"grad_norm": 3.9048423767089844,
|
| 674 |
+
"learning_rate": 1.2526846729580216e-05,
|
| 675 |
+
"loss": 0.0239,
|
| 676 |
+
"step": 4600
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 2.6480637813211843,
|
| 680 |
+
"grad_norm": 0.9887784719467163,
|
| 681 |
+
"learning_rate": 1.2445493003579565e-05,
|
| 682 |
+
"loss": 0.0271,
|
| 683 |
+
"step": 4650
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 2.6765375854214124,
|
| 687 |
+
"grad_norm": 4.759728908538818,
|
| 688 |
+
"learning_rate": 1.2364139277578915e-05,
|
| 689 |
+
"loss": 0.0201,
|
| 690 |
+
"step": 4700
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 2.70501138952164,
|
| 694 |
+
"grad_norm": 0.2671191096305847,
|
| 695 |
+
"learning_rate": 1.2282785551578263e-05,
|
| 696 |
+
"loss": 0.0212,
|
| 697 |
+
"step": 4750
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 2.733485193621868,
|
| 701 |
+
"grad_norm": 0.036626849323511124,
|
| 702 |
+
"learning_rate": 1.2201431825577612e-05,
|
| 703 |
+
"loss": 0.0155,
|
| 704 |
+
"step": 4800
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 2.7619589977220955,
|
| 708 |
+
"grad_norm": 6.877408027648926,
|
| 709 |
+
"learning_rate": 1.2120078099576962e-05,
|
| 710 |
+
"loss": 0.0266,
|
| 711 |
+
"step": 4850
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 2.7904328018223232,
|
| 715 |
+
"grad_norm": 4.725896835327148,
|
| 716 |
+
"learning_rate": 1.2038724373576312e-05,
|
| 717 |
+
"loss": 0.0163,
|
| 718 |
+
"step": 4900
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 2.8189066059225514,
|
| 722 |
+
"grad_norm": 0.11965059489011765,
|
| 723 |
+
"learning_rate": 1.195737064757566e-05,
|
| 724 |
+
"loss": 0.0246,
|
| 725 |
+
"step": 4950
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 2.847380410022779,
|
| 729 |
+
"grad_norm": 0.6634079813957214,
|
| 730 |
+
"learning_rate": 1.187601692157501e-05,
|
| 731 |
+
"loss": 0.0304,
|
| 732 |
+
"step": 5000
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 2.875854214123007,
|
| 736 |
+
"grad_norm": 3.952694892883301,
|
| 737 |
+
"learning_rate": 1.1794663195574359e-05,
|
| 738 |
+
"loss": 0.0183,
|
| 739 |
+
"step": 5050
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 2.9043280182232345,
|
| 743 |
+
"grad_norm": 19.788942337036133,
|
| 744 |
+
"learning_rate": 1.1713309469573709e-05,
|
| 745 |
+
"loss": 0.0193,
|
| 746 |
+
"step": 5100
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 2.932801822323462,
|
| 750 |
+
"grad_norm": 4.989261627197266,
|
| 751 |
+
"learning_rate": 1.1631955743573055e-05,
|
| 752 |
+
"loss": 0.023,
|
| 753 |
+
"step": 5150
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 2.9612756264236904,
|
| 757 |
+
"grad_norm": 9.020853996276855,
|
| 758 |
+
"learning_rate": 1.1550602017572406e-05,
|
| 759 |
+
"loss": 0.0231,
|
| 760 |
+
"step": 5200
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 2.989749430523918,
|
| 764 |
+
"grad_norm": 0.9151140451431274,
|
| 765 |
+
"learning_rate": 1.1469248291571756e-05,
|
| 766 |
+
"loss": 0.0314,
|
| 767 |
+
"step": 5250
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 3.0,
|
| 771 |
+
"eval_f1": 0.9470420646891236,
|
| 772 |
+
"eval_loss": 0.04870199039578438,
|
| 773 |
+
"eval_precision": 0.943090787716956,
|
| 774 |
+
"eval_recall": 0.9510265903736116,
|
| 775 |
+
"eval_runtime": 3.7494,
|
| 776 |
+
"eval_samples_per_second": 866.805,
|
| 777 |
+
"eval_steps_per_second": 108.551,
|
| 778 |
+
"step": 5268
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 3.0182232346241458,
|
| 782 |
+
"grad_norm": 8.62992000579834,
|
| 783 |
+
"learning_rate": 1.1387894565571106e-05,
|
| 784 |
+
"loss": 0.0194,
|
| 785 |
+
"step": 5300
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.0466970387243735,
|
| 789 |
+
"grad_norm": 0.5615554451942444,
|
| 790 |
+
"learning_rate": 1.1306540839570452e-05,
|
| 791 |
+
"loss": 0.0114,
|
| 792 |
+
"step": 5350
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.075170842824601,
|
| 796 |
+
"grad_norm": 2.5735936164855957,
|
| 797 |
+
"learning_rate": 1.1225187113569801e-05,
|
| 798 |
+
"loss": 0.0163,
|
| 799 |
+
"step": 5400
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 3.1036446469248293,
|
| 803 |
+
"grad_norm": 6.504094123840332,
|
| 804 |
+
"learning_rate": 1.1143833387569151e-05,
|
| 805 |
+
"loss": 0.0135,
|
| 806 |
+
"step": 5450
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 3.132118451025057,
|
| 810 |
+
"grad_norm": 14.477537155151367,
|
| 811 |
+
"learning_rate": 1.10624796615685e-05,
|
| 812 |
+
"loss": 0.0129,
|
| 813 |
+
"step": 5500
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 3.1605922551252847,
|
| 817 |
+
"grad_norm": 0.04929841682314873,
|
| 818 |
+
"learning_rate": 1.0981125935567849e-05,
|
| 819 |
+
"loss": 0.0263,
|
| 820 |
+
"step": 5550
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 3.1890660592255125,
|
| 824 |
+
"grad_norm": 0.08875144273042679,
|
| 825 |
+
"learning_rate": 1.0899772209567198e-05,
|
| 826 |
+
"loss": 0.0134,
|
| 827 |
+
"step": 5600
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 3.21753986332574,
|
| 831 |
+
"grad_norm": 0.6238455772399902,
|
| 832 |
+
"learning_rate": 1.0818418483566548e-05,
|
| 833 |
+
"loss": 0.0107,
|
| 834 |
+
"step": 5650
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 3.2460136674259683,
|
| 838 |
+
"grad_norm": 0.004948179703205824,
|
| 839 |
+
"learning_rate": 1.0737064757565898e-05,
|
| 840 |
+
"loss": 0.0167,
|
| 841 |
+
"step": 5700
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 3.274487471526196,
|
| 845 |
+
"grad_norm": 0.017031218856573105,
|
| 846 |
+
"learning_rate": 1.0655711031565246e-05,
|
| 847 |
+
"loss": 0.0085,
|
| 848 |
+
"step": 5750
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 3.3029612756264237,
|
| 852 |
+
"grad_norm": 2.9840469360351562,
|
| 853 |
+
"learning_rate": 1.0574357305564595e-05,
|
| 854 |
+
"loss": 0.0158,
|
| 855 |
+
"step": 5800
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 3.3314350797266514,
|
| 859 |
+
"grad_norm": 1.7477540969848633,
|
| 860 |
+
"learning_rate": 1.0493003579563945e-05,
|
| 861 |
+
"loss": 0.0157,
|
| 862 |
+
"step": 5850
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 3.359908883826879,
|
| 866 |
+
"grad_norm": 0.03963543102145195,
|
| 867 |
+
"learning_rate": 1.0411649853563295e-05,
|
| 868 |
+
"loss": 0.0201,
|
| 869 |
+
"step": 5900
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 3.3883826879271073,
|
| 873 |
+
"grad_norm": 0.16669808328151703,
|
| 874 |
+
"learning_rate": 1.0330296127562643e-05,
|
| 875 |
+
"loss": 0.0159,
|
| 876 |
+
"step": 5950
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 3.416856492027335,
|
| 880 |
+
"grad_norm": 0.01953568309545517,
|
| 881 |
+
"learning_rate": 1.0248942401561992e-05,
|
| 882 |
+
"loss": 0.0117,
|
| 883 |
+
"step": 6000
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 3.4453302961275627,
|
| 887 |
+
"grad_norm": 0.07708246260881424,
|
| 888 |
+
"learning_rate": 1.0167588675561342e-05,
|
| 889 |
+
"loss": 0.0117,
|
| 890 |
+
"step": 6050
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 3.4738041002277904,
|
| 894 |
+
"grad_norm": 2.423590898513794,
|
| 895 |
+
"learning_rate": 1.0086234949560691e-05,
|
| 896 |
+
"loss": 0.0111,
|
| 897 |
+
"step": 6100
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 3.502277904328018,
|
| 901 |
+
"grad_norm": 0.006337775848805904,
|
| 902 |
+
"learning_rate": 1.000488122356004e-05,
|
| 903 |
+
"loss": 0.0052,
|
| 904 |
+
"step": 6150
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 3.5307517084282463,
|
| 908 |
+
"grad_norm": 0.01373753696680069,
|
| 909 |
+
"learning_rate": 9.923527497559389e-06,
|
| 910 |
+
"loss": 0.009,
|
| 911 |
+
"step": 6200
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 3.559225512528474,
|
| 915 |
+
"grad_norm": 0.11260247975587845,
|
| 916 |
+
"learning_rate": 9.842173771558739e-06,
|
| 917 |
+
"loss": 0.0208,
|
| 918 |
+
"step": 6250
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 3.5876993166287017,
|
| 922 |
+
"grad_norm": 1.240822196006775,
|
| 923 |
+
"learning_rate": 9.760820045558087e-06,
|
| 924 |
+
"loss": 0.0157,
|
| 925 |
+
"step": 6300
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 3.6161731207289294,
|
| 929 |
+
"grad_norm": 0.05271293595433235,
|
| 930 |
+
"learning_rate": 9.679466319557436e-06,
|
| 931 |
+
"loss": 0.0175,
|
| 932 |
+
"step": 6350
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 3.644646924829157,
|
| 936 |
+
"grad_norm": 24.546663284301758,
|
| 937 |
+
"learning_rate": 9.598112593556786e-06,
|
| 938 |
+
"loss": 0.015,
|
| 939 |
+
"step": 6400
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 3.6731207289293852,
|
| 943 |
+
"grad_norm": 5.120173931121826,
|
| 944 |
+
"learning_rate": 9.516758867556136e-06,
|
| 945 |
+
"loss": 0.017,
|
| 946 |
+
"step": 6450
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 3.7015945330296125,
|
| 950 |
+
"grad_norm": 1.1922008991241455,
|
| 951 |
+
"learning_rate": 9.435405141555484e-06,
|
| 952 |
+
"loss": 0.0194,
|
| 953 |
+
"step": 6500
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 3.7300683371298406,
|
| 957 |
+
"grad_norm": 0.039439987391233444,
|
| 958 |
+
"learning_rate": 9.354051415554833e-06,
|
| 959 |
+
"loss": 0.0309,
|
| 960 |
+
"step": 6550
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 3.7585421412300684,
|
| 964 |
+
"grad_norm": 0.04055279493331909,
|
| 965 |
+
"learning_rate": 9.272697689554181e-06,
|
| 966 |
+
"loss": 0.0195,
|
| 967 |
+
"step": 6600
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 3.787015945330296,
|
| 971 |
+
"grad_norm": 0.5971085429191589,
|
| 972 |
+
"learning_rate": 9.191343963553533e-06,
|
| 973 |
+
"loss": 0.0156,
|
| 974 |
+
"step": 6650
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 3.8154897494305238,
|
| 978 |
+
"grad_norm": 0.2191866636276245,
|
| 979 |
+
"learning_rate": 9.10999023755288e-06,
|
| 980 |
+
"loss": 0.0174,
|
| 981 |
+
"step": 6700
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 3.8439635535307515,
|
| 985 |
+
"grad_norm": 0.033559828996658325,
|
| 986 |
+
"learning_rate": 9.02863651155223e-06,
|
| 987 |
+
"loss": 0.0149,
|
| 988 |
+
"step": 6750
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 3.8724373576309796,
|
| 992 |
+
"grad_norm": 3.7128634452819824,
|
| 993 |
+
"learning_rate": 8.947282785551578e-06,
|
| 994 |
+
"loss": 0.0221,
|
| 995 |
+
"step": 6800
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 3.9009111617312073,
|
| 999 |
+
"grad_norm": 0.08547580987215042,
|
| 1000 |
+
"learning_rate": 8.865929059550928e-06,
|
| 1001 |
+
"loss": 0.0153,
|
| 1002 |
+
"step": 6850
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 3.929384965831435,
|
| 1006 |
+
"grad_norm": 0.058124080300331116,
|
| 1007 |
+
"learning_rate": 8.784575333550277e-06,
|
| 1008 |
+
"loss": 0.0161,
|
| 1009 |
+
"step": 6900
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 3.9578587699316627,
|
| 1013 |
+
"grad_norm": 0.015579139813780785,
|
| 1014 |
+
"learning_rate": 8.703221607549627e-06,
|
| 1015 |
+
"loss": 0.011,
|
| 1016 |
+
"step": 6950
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 3.9863325740318905,
|
| 1020 |
+
"grad_norm": 0.12764935195446014,
|
| 1021 |
+
"learning_rate": 8.621867881548975e-06,
|
| 1022 |
+
"loss": 0.0166,
|
| 1023 |
+
"step": 7000
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 4.0,
|
| 1027 |
+
"eval_f1": 0.9525329748802823,
|
| 1028 |
+
"eval_loss": 0.04983380436897278,
|
| 1029 |
+
"eval_precision": 0.9510149303808086,
|
| 1030 |
+
"eval_recall": 0.9540558734432851,
|
| 1031 |
+
"eval_runtime": 3.7495,
|
| 1032 |
+
"eval_samples_per_second": 866.79,
|
| 1033 |
+
"eval_steps_per_second": 108.549,
|
| 1034 |
+
"step": 7024
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 4.014806378132119,
|
| 1038 |
+
"grad_norm": 0.04521024227142334,
|
| 1039 |
+
"learning_rate": 8.540514155548325e-06,
|
| 1040 |
+
"loss": 0.0102,
|
| 1041 |
+
"step": 7050
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 4.043280182232346,
|
| 1045 |
+
"grad_norm": 0.015535669401288033,
|
| 1046 |
+
"learning_rate": 8.459160429547674e-06,
|
| 1047 |
+
"loss": 0.0068,
|
| 1048 |
+
"step": 7100
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 4.071753986332574,
|
| 1052 |
+
"grad_norm": 12.566010475158691,
|
| 1053 |
+
"learning_rate": 8.377806703547024e-06,
|
| 1054 |
+
"loss": 0.0151,
|
| 1055 |
+
"step": 7150
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 4.100227790432802,
|
| 1059 |
+
"grad_norm": 0.027761396020650864,
|
| 1060 |
+
"learning_rate": 8.296452977546372e-06,
|
| 1061 |
+
"loss": 0.0053,
|
| 1062 |
+
"step": 7200
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 4.128701594533029,
|
| 1066 |
+
"grad_norm": 1.8846938610076904,
|
| 1067 |
+
"learning_rate": 8.215099251545722e-06,
|
| 1068 |
+
"loss": 0.0077,
|
| 1069 |
+
"step": 7250
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 4.157175398633258,
|
| 1073 |
+
"grad_norm": 0.005013479385524988,
|
| 1074 |
+
"learning_rate": 8.13374552554507e-06,
|
| 1075 |
+
"loss": 0.0073,
|
| 1076 |
+
"step": 7300
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 4.185649202733485,
|
| 1080 |
+
"grad_norm": 0.32696235179901123,
|
| 1081 |
+
"learning_rate": 8.052391799544421e-06,
|
| 1082 |
+
"loss": 0.0092,
|
| 1083 |
+
"step": 7350
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 4.214123006833713,
|
| 1087 |
+
"grad_norm": 0.014357910491526127,
|
| 1088 |
+
"learning_rate": 7.971038073543769e-06,
|
| 1089 |
+
"loss": 0.0164,
|
| 1090 |
+
"step": 7400
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 4.242596810933941,
|
| 1094 |
+
"grad_norm": 0.010845329612493515,
|
| 1095 |
+
"learning_rate": 7.889684347543118e-06,
|
| 1096 |
+
"loss": 0.0123,
|
| 1097 |
+
"step": 7450
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 4.271070615034168,
|
| 1101 |
+
"grad_norm": 0.6969354748725891,
|
| 1102 |
+
"learning_rate": 7.808330621542466e-06,
|
| 1103 |
+
"loss": 0.0103,
|
| 1104 |
+
"step": 7500
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 4.2995444191343966,
|
| 1108 |
+
"grad_norm": 0.3575742840766907,
|
| 1109 |
+
"learning_rate": 7.726976895541816e-06,
|
| 1110 |
+
"loss": 0.0161,
|
| 1111 |
+
"step": 7550
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 4.328018223234624,
|
| 1115 |
+
"grad_norm": 0.008076228201389313,
|
| 1116 |
+
"learning_rate": 7.645623169541166e-06,
|
| 1117 |
+
"loss": 0.0122,
|
| 1118 |
+
"step": 7600
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 4.356492027334852,
|
| 1122 |
+
"grad_norm": 0.029101597145199776,
|
| 1123 |
+
"learning_rate": 7.564269443540515e-06,
|
| 1124 |
+
"loss": 0.0071,
|
| 1125 |
+
"step": 7650
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 4.38496583143508,
|
| 1129 |
+
"grad_norm": 0.06841142475605011,
|
| 1130 |
+
"learning_rate": 7.482915717539863e-06,
|
| 1131 |
+
"loss": 0.0091,
|
| 1132 |
+
"step": 7700
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 4.413439635535307,
|
| 1136 |
+
"grad_norm": 0.04133535176515579,
|
| 1137 |
+
"learning_rate": 7.401561991539213e-06,
|
| 1138 |
+
"loss": 0.0073,
|
| 1139 |
+
"step": 7750
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 4.4419134396355355,
|
| 1143 |
+
"grad_norm": 0.03571132943034172,
|
| 1144 |
+
"learning_rate": 7.320208265538562e-06,
|
| 1145 |
+
"loss": 0.0131,
|
| 1146 |
+
"step": 7800
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 4.470387243735763,
|
| 1150 |
+
"grad_norm": 2.8286924362182617,
|
| 1151 |
+
"learning_rate": 7.2388545395379114e-06,
|
| 1152 |
+
"loss": 0.0043,
|
| 1153 |
+
"step": 7850
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 4.498861047835991,
|
| 1157 |
+
"grad_norm": 1.3785158395767212,
|
| 1158 |
+
"learning_rate": 7.15750081353726e-06,
|
| 1159 |
+
"loss": 0.0103,
|
| 1160 |
+
"step": 7900
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 4.527334851936219,
|
| 1164 |
+
"grad_norm": 0.05979786813259125,
|
| 1165 |
+
"learning_rate": 7.07614708753661e-06,
|
| 1166 |
+
"loss": 0.006,
|
| 1167 |
+
"step": 7950
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 4.555808656036446,
|
| 1171 |
+
"grad_norm": 1.3434393405914307,
|
| 1172 |
+
"learning_rate": 6.994793361535959e-06,
|
| 1173 |
+
"loss": 0.0148,
|
| 1174 |
+
"step": 8000
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 4.5842824601366745,
|
| 1178 |
+
"grad_norm": 0.019605603069067,
|
| 1179 |
+
"learning_rate": 6.913439635535308e-06,
|
| 1180 |
+
"loss": 0.0068,
|
| 1181 |
+
"step": 8050
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 4.612756264236902,
|
| 1185 |
+
"grad_norm": 0.0028891051188111305,
|
| 1186 |
+
"learning_rate": 6.832085909534657e-06,
|
| 1187 |
+
"loss": 0.0075,
|
| 1188 |
+
"step": 8100
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 4.64123006833713,
|
| 1192 |
+
"grad_norm": 0.014856363646686077,
|
| 1193 |
+
"learning_rate": 6.750732183534007e-06,
|
| 1194 |
+
"loss": 0.0073,
|
| 1195 |
+
"step": 8150
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 4.669703872437358,
|
| 1199 |
+
"grad_norm": 0.0037149768322706223,
|
| 1200 |
+
"learning_rate": 6.669378457533356e-06,
|
| 1201 |
+
"loss": 0.0066,
|
| 1202 |
+
"step": 8200
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 4.698177676537585,
|
| 1206 |
+
"grad_norm": 0.008534993045032024,
|
| 1207 |
+
"learning_rate": 6.588024731532705e-06,
|
| 1208 |
+
"loss": 0.0121,
|
| 1209 |
+
"step": 8250
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 4.7266514806378135,
|
| 1213 |
+
"grad_norm": 0.013920712284743786,
|
| 1214 |
+
"learning_rate": 6.506671005532054e-06,
|
| 1215 |
+
"loss": 0.013,
|
| 1216 |
+
"step": 8300
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 4.755125284738041,
|
| 1220 |
+
"grad_norm": 0.10121840238571167,
|
| 1221 |
+
"learning_rate": 6.425317279531404e-06,
|
| 1222 |
+
"loss": 0.0087,
|
| 1223 |
+
"step": 8350
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 4.783599088838269,
|
| 1227 |
+
"grad_norm": 0.07937771081924438,
|
| 1228 |
+
"learning_rate": 6.343963553530752e-06,
|
| 1229 |
+
"loss": 0.0086,
|
| 1230 |
+
"step": 8400
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 4.812072892938497,
|
| 1234 |
+
"grad_norm": 0.07674901187419891,
|
| 1235 |
+
"learning_rate": 6.262609827530101e-06,
|
| 1236 |
+
"loss": 0.0052,
|
| 1237 |
+
"step": 8450
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 4.840546697038724,
|
| 1241 |
+
"grad_norm": 0.06683178246021271,
|
| 1242 |
+
"learning_rate": 6.18125610152945e-06,
|
| 1243 |
+
"loss": 0.015,
|
| 1244 |
+
"step": 8500
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 4.8690205011389525,
|
| 1248 |
+
"grad_norm": 0.08861212432384491,
|
| 1249 |
+
"learning_rate": 6.0999023755288e-06,
|
| 1250 |
+
"loss": 0.0109,
|
| 1251 |
+
"step": 8550
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 4.89749430523918,
|
| 1255 |
+
"grad_norm": 0.006198623217642307,
|
| 1256 |
+
"learning_rate": 6.018548649528149e-06,
|
| 1257 |
+
"loss": 0.0181,
|
| 1258 |
+
"step": 8600
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 4.925968109339408,
|
| 1262 |
+
"grad_norm": 0.009029284119606018,
|
| 1263 |
+
"learning_rate": 5.937194923527498e-06,
|
| 1264 |
+
"loss": 0.0043,
|
| 1265 |
+
"step": 8650
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 4.954441913439636,
|
| 1269 |
+
"grad_norm": 1.1225602626800537,
|
| 1270 |
+
"learning_rate": 5.855841197526847e-06,
|
| 1271 |
+
"loss": 0.0121,
|
| 1272 |
+
"step": 8700
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 4.982915717539863,
|
| 1276 |
+
"grad_norm": 0.01218325924128294,
|
| 1277 |
+
"learning_rate": 5.774487471526197e-06,
|
| 1278 |
+
"loss": 0.0115,
|
| 1279 |
+
"step": 8750
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 5.0,
|
| 1283 |
+
"eval_f1": 0.9533444816053511,
|
| 1284 |
+
"eval_loss": 0.04722925275564194,
|
| 1285 |
+
"eval_precision": 0.9473246925889,
|
| 1286 |
+
"eval_recall": 0.9594412655671491,
|
| 1287 |
+
"eval_runtime": 3.7358,
|
| 1288 |
+
"eval_samples_per_second": 869.968,
|
| 1289 |
+
"eval_steps_per_second": 108.947,
|
| 1290 |
+
"step": 8780
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.011389521640091,
|
| 1294 |
+
"grad_norm": 0.8161097764968872,
|
| 1295 |
+
"learning_rate": 5.6931337455255455e-06,
|
| 1296 |
+
"loss": 0.0127,
|
| 1297 |
+
"step": 8800
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.039863325740319,
|
| 1301 |
+
"grad_norm": 0.048089127987623215,
|
| 1302 |
+
"learning_rate": 5.611780019524895e-06,
|
| 1303 |
+
"loss": 0.0116,
|
| 1304 |
+
"step": 8850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 5.068337129840547,
|
| 1308 |
+
"grad_norm": 1.7576072216033936,
|
| 1309 |
+
"learning_rate": 5.530426293524244e-06,
|
| 1310 |
+
"loss": 0.0093,
|
| 1311 |
+
"step": 8900
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 5.096810933940774,
|
| 1315 |
+
"grad_norm": 0.07270015776157379,
|
| 1316 |
+
"learning_rate": 5.449072567523594e-06,
|
| 1317 |
+
"loss": 0.0043,
|
| 1318 |
+
"step": 8950
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 5.125284738041002,
|
| 1322 |
+
"grad_norm": 0.2580782175064087,
|
| 1323 |
+
"learning_rate": 5.367718841522942e-06,
|
| 1324 |
+
"loss": 0.0058,
|
| 1325 |
+
"step": 9000
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 5.15375854214123,
|
| 1329 |
+
"grad_norm": 0.12649740278720856,
|
| 1330 |
+
"learning_rate": 5.286365115522292e-06,
|
| 1331 |
+
"loss": 0.0033,
|
| 1332 |
+
"step": 9050
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 5.182232346241458,
|
| 1336 |
+
"grad_norm": 0.06850716471672058,
|
| 1337 |
+
"learning_rate": 5.20501138952164e-06,
|
| 1338 |
+
"loss": 0.003,
|
| 1339 |
+
"step": 9100
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 5.210706150341686,
|
| 1343 |
+
"grad_norm": 0.01197959017008543,
|
| 1344 |
+
"learning_rate": 5.12365766352099e-06,
|
| 1345 |
+
"loss": 0.0029,
|
| 1346 |
+
"step": 9150
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 5.239179954441913,
|
| 1350 |
+
"grad_norm": 0.10818086564540863,
|
| 1351 |
+
"learning_rate": 5.0423039375203385e-06,
|
| 1352 |
+
"loss": 0.0025,
|
| 1353 |
+
"step": 9200
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 5.267653758542141,
|
| 1357 |
+
"grad_norm": 0.035460665822029114,
|
| 1358 |
+
"learning_rate": 4.960950211519688e-06,
|
| 1359 |
+
"loss": 0.0047,
|
| 1360 |
+
"step": 9250
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 5.296127562642369,
|
| 1364 |
+
"grad_norm": 4.532562255859375,
|
| 1365 |
+
"learning_rate": 4.879596485519037e-06,
|
| 1366 |
+
"loss": 0.0086,
|
| 1367 |
+
"step": 9300
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 5.324601366742597,
|
| 1371 |
+
"grad_norm": 0.013236219063401222,
|
| 1372 |
+
"learning_rate": 4.7982427595183866e-06,
|
| 1373 |
+
"loss": 0.0027,
|
| 1374 |
+
"step": 9350
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 5.353075170842825,
|
| 1378 |
+
"grad_norm": 0.018059909343719482,
|
| 1379 |
+
"learning_rate": 4.716889033517735e-06,
|
| 1380 |
+
"loss": 0.0065,
|
| 1381 |
+
"step": 9400
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 5.381548974943052,
|
| 1385 |
+
"grad_norm": 0.019096272066235542,
|
| 1386 |
+
"learning_rate": 4.635535307517084e-06,
|
| 1387 |
+
"loss": 0.0053,
|
| 1388 |
+
"step": 9450
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 5.41002277904328,
|
| 1392 |
+
"grad_norm": 0.10389436781406403,
|
| 1393 |
+
"learning_rate": 4.554181581516434e-06,
|
| 1394 |
+
"loss": 0.0029,
|
| 1395 |
+
"step": 9500
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 5.438496583143508,
|
| 1399 |
+
"grad_norm": 0.0029252381063997746,
|
| 1400 |
+
"learning_rate": 4.472827855515783e-06,
|
| 1401 |
+
"loss": 0.0045,
|
| 1402 |
+
"step": 9550
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 5.466970387243736,
|
| 1406 |
+
"grad_norm": 0.004542892333120108,
|
| 1407 |
+
"learning_rate": 4.391474129515132e-06,
|
| 1408 |
+
"loss": 0.0102,
|
| 1409 |
+
"step": 9600
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 5.495444191343964,
|
| 1413 |
+
"grad_norm": 0.130916565656662,
|
| 1414 |
+
"learning_rate": 4.310120403514481e-06,
|
| 1415 |
+
"loss": 0.0064,
|
| 1416 |
+
"step": 9650
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 5.523917995444191,
|
| 1420 |
+
"grad_norm": 0.008013393729925156,
|
| 1421 |
+
"learning_rate": 4.22876667751383e-06,
|
| 1422 |
+
"loss": 0.005,
|
| 1423 |
+
"step": 9700
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 5.552391799544419,
|
| 1427 |
+
"grad_norm": 0.012782514095306396,
|
| 1428 |
+
"learning_rate": 4.1474129515131795e-06,
|
| 1429 |
+
"loss": 0.0122,
|
| 1430 |
+
"step": 9750
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 5.5808656036446465,
|
| 1434 |
+
"grad_norm": 0.9617053866386414,
|
| 1435 |
+
"learning_rate": 4.066059225512528e-06,
|
| 1436 |
+
"loss": 0.0063,
|
| 1437 |
+
"step": 9800
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 5.609339407744875,
|
| 1441 |
+
"grad_norm": 0.014953136444091797,
|
| 1442 |
+
"learning_rate": 3.984705499511878e-06,
|
| 1443 |
+
"loss": 0.0046,
|
| 1444 |
+
"step": 9850
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 5.637813211845103,
|
| 1448 |
+
"grad_norm": 1.0275633335113525,
|
| 1449 |
+
"learning_rate": 3.903351773511227e-06,
|
| 1450 |
+
"loss": 0.0075,
|
| 1451 |
+
"step": 9900
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 5.66628701594533,
|
| 1455 |
+
"grad_norm": 0.007999264635145664,
|
| 1456 |
+
"learning_rate": 3.8219980475105764e-06,
|
| 1457 |
+
"loss": 0.0031,
|
| 1458 |
+
"step": 9950
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 5.694760820045558,
|
| 1462 |
+
"grad_norm": 0.008241960778832436,
|
| 1463 |
+
"learning_rate": 3.7406443215099252e-06,
|
| 1464 |
+
"loss": 0.0065,
|
| 1465 |
+
"step": 10000
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 5.723234624145785,
|
| 1469 |
+
"grad_norm": 2.578386068344116,
|
| 1470 |
+
"learning_rate": 3.6592905955092745e-06,
|
| 1471 |
+
"loss": 0.0057,
|
| 1472 |
+
"step": 10050
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 5.751708428246014,
|
| 1476 |
+
"grad_norm": 2.8855443000793457,
|
| 1477 |
+
"learning_rate": 3.5779368695086237e-06,
|
| 1478 |
+
"loss": 0.0038,
|
| 1479 |
+
"step": 10100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 5.780182232346242,
|
| 1483 |
+
"grad_norm": 0.016262667253613472,
|
| 1484 |
+
"learning_rate": 3.496583143507973e-06,
|
| 1485 |
+
"loss": 0.0128,
|
| 1486 |
+
"step": 10150
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 5.808656036446469,
|
| 1490 |
+
"grad_norm": 0.045168060809373856,
|
| 1491 |
+
"learning_rate": 3.415229417507322e-06,
|
| 1492 |
+
"loss": 0.0036,
|
| 1493 |
+
"step": 10200
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 5.837129840546697,
|
| 1497 |
+
"grad_norm": 0.1192622259259224,
|
| 1498 |
+
"learning_rate": 3.3355027660266842e-06,
|
| 1499 |
+
"loss": 0.0111,
|
| 1500 |
+
"step": 10250
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 5.865603644646924,
|
| 1504 |
+
"grad_norm": 0.08544190227985382,
|
| 1505 |
+
"learning_rate": 3.2541490400260335e-06,
|
| 1506 |
+
"loss": 0.0051,
|
| 1507 |
+
"step": 10300
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 5.894077448747153,
|
| 1511 |
+
"grad_norm": 3.493161678314209,
|
| 1512 |
+
"learning_rate": 3.1727953140253827e-06,
|
| 1513 |
+
"loss": 0.0054,
|
| 1514 |
+
"step": 10350
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 5.922551252847381,
|
| 1518 |
+
"grad_norm": 0.022789066657423973,
|
| 1519 |
+
"learning_rate": 3.0914415880247315e-06,
|
| 1520 |
+
"loss": 0.0033,
|
| 1521 |
+
"step": 10400
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 5.951025056947608,
|
| 1525 |
+
"grad_norm": 12.131625175476074,
|
| 1526 |
+
"learning_rate": 3.0100878620240807e-06,
|
| 1527 |
+
"loss": 0.0023,
|
| 1528 |
+
"step": 10450
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 5.979498861047836,
|
| 1532 |
+
"grad_norm": 0.03329641371965408,
|
| 1533 |
+
"learning_rate": 2.92873413602343e-06,
|
| 1534 |
+
"loss": 0.0058,
|
| 1535 |
+
"step": 10500
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 6.0,
|
| 1539 |
+
"eval_f1": 0.9557566616390145,
|
| 1540 |
+
"eval_loss": 0.05410688370466232,
|
| 1541 |
+
"eval_precision": 0.9517690253671562,
|
| 1542 |
+
"eval_recall": 0.9597778525748906,
|
| 1543 |
+
"eval_runtime": 3.6572,
|
| 1544 |
+
"eval_samples_per_second": 888.662,
|
| 1545 |
+
"eval_steps_per_second": 111.288,
|
| 1546 |
+
"step": 10536
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 6.007972665148063,
|
| 1550 |
+
"grad_norm": 0.008557640947401524,
|
| 1551 |
+
"learning_rate": 2.847380410022779e-06,
|
| 1552 |
+
"loss": 0.0152,
|
| 1553 |
+
"step": 10550
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 6.0364464692482915,
|
| 1557 |
+
"grad_norm": 0.08105529844760895,
|
| 1558 |
+
"learning_rate": 2.7660266840221284e-06,
|
| 1559 |
+
"loss": 0.0025,
|
| 1560 |
+
"step": 10600
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 6.06492027334852,
|
| 1564 |
+
"grad_norm": 1.100066065788269,
|
| 1565 |
+
"learning_rate": 2.6846729580214776e-06,
|
| 1566 |
+
"loss": 0.0015,
|
| 1567 |
+
"step": 10650
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 6.093394077448747,
|
| 1571 |
+
"grad_norm": 1.96909761428833,
|
| 1572 |
+
"learning_rate": 2.603319232020827e-06,
|
| 1573 |
+
"loss": 0.0021,
|
| 1574 |
+
"step": 10700
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 6.121867881548975,
|
| 1578 |
+
"grad_norm": 0.00583766121417284,
|
| 1579 |
+
"learning_rate": 2.5219655060201757e-06,
|
| 1580 |
+
"loss": 0.0127,
|
| 1581 |
+
"step": 10750
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 6.150341685649202,
|
| 1585 |
+
"grad_norm": 0.026862677186727524,
|
| 1586 |
+
"learning_rate": 2.440611780019525e-06,
|
| 1587 |
+
"loss": 0.004,
|
| 1588 |
+
"step": 10800
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 6.1788154897494305,
|
| 1592 |
+
"grad_norm": 0.010042566806077957,
|
| 1593 |
+
"learning_rate": 2.359258054018874e-06,
|
| 1594 |
+
"loss": 0.0033,
|
| 1595 |
+
"step": 10850
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 6.207289293849659,
|
| 1599 |
+
"grad_norm": 0.8956929445266724,
|
| 1600 |
+
"learning_rate": 2.2779043280182233e-06,
|
| 1601 |
+
"loss": 0.0031,
|
| 1602 |
+
"step": 10900
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 6.235763097949886,
|
| 1606 |
+
"grad_norm": 0.009118441492319107,
|
| 1607 |
+
"learning_rate": 2.1965506020175726e-06,
|
| 1608 |
+
"loss": 0.0039,
|
| 1609 |
+
"step": 10950
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 6.264236902050114,
|
| 1613 |
+
"grad_norm": 0.22793921828269958,
|
| 1614 |
+
"learning_rate": 2.115196876016922e-06,
|
| 1615 |
+
"loss": 0.0013,
|
| 1616 |
+
"step": 11000
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 6.292710706150341,
|
| 1620 |
+
"grad_norm": 0.015608682297170162,
|
| 1621 |
+
"learning_rate": 2.033843150016271e-06,
|
| 1622 |
+
"loss": 0.0021,
|
| 1623 |
+
"step": 11050
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 6.3211845102505695,
|
| 1627 |
+
"grad_norm": 0.004031027667224407,
|
| 1628 |
+
"learning_rate": 1.95248942401562e-06,
|
| 1629 |
+
"loss": 0.0011,
|
| 1630 |
+
"step": 11100
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 6.349658314350798,
|
| 1634 |
+
"grad_norm": 0.008949857205152512,
|
| 1635 |
+
"learning_rate": 1.8711356980149693e-06,
|
| 1636 |
+
"loss": 0.0049,
|
| 1637 |
+
"step": 11150
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 6.378132118451025,
|
| 1641 |
+
"grad_norm": 0.018670039251446724,
|
| 1642 |
+
"learning_rate": 1.7897819720143183e-06,
|
| 1643 |
+
"loss": 0.003,
|
| 1644 |
+
"step": 11200
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 6.406605922551253,
|
| 1648 |
+
"grad_norm": 0.032393742352724075,
|
| 1649 |
+
"learning_rate": 1.7084282460136675e-06,
|
| 1650 |
+
"loss": 0.0051,
|
| 1651 |
+
"step": 11250
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 6.43507972665148,
|
| 1655 |
+
"grad_norm": 0.11851054430007935,
|
| 1656 |
+
"learning_rate": 1.6270745200130167e-06,
|
| 1657 |
+
"loss": 0.0046,
|
| 1658 |
+
"step": 11300
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 6.4635535307517085,
|
| 1662 |
+
"grad_norm": 0.09247086197137833,
|
| 1663 |
+
"learning_rate": 1.5457207940123657e-06,
|
| 1664 |
+
"loss": 0.0049,
|
| 1665 |
+
"step": 11350
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 6.492027334851937,
|
| 1669 |
+
"grad_norm": 0.002555207349359989,
|
| 1670 |
+
"learning_rate": 1.464367068011715e-06,
|
| 1671 |
+
"loss": 0.0026,
|
| 1672 |
+
"step": 11400
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 6.520501138952164,
|
| 1676 |
+
"grad_norm": 3.7294840812683105,
|
| 1677 |
+
"learning_rate": 1.3830133420110642e-06,
|
| 1678 |
+
"loss": 0.0014,
|
| 1679 |
+
"step": 11450
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 6.548974943052392,
|
| 1683 |
+
"grad_norm": 0.00703430688008666,
|
| 1684 |
+
"learning_rate": 1.3016596160104134e-06,
|
| 1685 |
+
"loss": 0.0035,
|
| 1686 |
+
"step": 11500
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 6.577448747152619,
|
| 1690 |
+
"grad_norm": 12.061240196228027,
|
| 1691 |
+
"learning_rate": 1.2203058900097624e-06,
|
| 1692 |
+
"loss": 0.0078,
|
| 1693 |
+
"step": 11550
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 6.605922551252847,
|
| 1697 |
+
"grad_norm": 0.011964640580117702,
|
| 1698 |
+
"learning_rate": 1.1389521640091117e-06,
|
| 1699 |
+
"loss": 0.0024,
|
| 1700 |
+
"step": 11600
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 6.634396355353076,
|
| 1704 |
+
"grad_norm": 0.014603933319449425,
|
| 1705 |
+
"learning_rate": 1.057598438008461e-06,
|
| 1706 |
+
"loss": 0.0072,
|
| 1707 |
+
"step": 11650
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 6.662870159453303,
|
| 1711 |
+
"grad_norm": 0.026642296463251114,
|
| 1712 |
+
"learning_rate": 9.7624471200781e-07,
|
| 1713 |
+
"loss": 0.0007,
|
| 1714 |
+
"step": 11700
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 6.691343963553531,
|
| 1718 |
+
"grad_norm": 3.3646230697631836,
|
| 1719 |
+
"learning_rate": 8.948909860071591e-07,
|
| 1720 |
+
"loss": 0.0046,
|
| 1721 |
+
"step": 11750
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 6.719817767653758,
|
| 1725 |
+
"grad_norm": 0.027053840458393097,
|
| 1726 |
+
"learning_rate": 8.135372600065084e-07,
|
| 1727 |
+
"loss": 0.0031,
|
| 1728 |
+
"step": 11800
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 6.748291571753986,
|
| 1732 |
+
"grad_norm": 2.6146697998046875,
|
| 1733 |
+
"learning_rate": 7.321835340058575e-07,
|
| 1734 |
+
"loss": 0.003,
|
| 1735 |
+
"step": 11850
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 6.776765375854215,
|
| 1739 |
+
"grad_norm": 8.964512825012207,
|
| 1740 |
+
"learning_rate": 6.508298080052067e-07,
|
| 1741 |
+
"loss": 0.0069,
|
| 1742 |
+
"step": 11900
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 6.805239179954442,
|
| 1746 |
+
"grad_norm": 0.3086203932762146,
|
| 1747 |
+
"learning_rate": 5.694760820045558e-07,
|
| 1748 |
+
"loss": 0.0035,
|
| 1749 |
+
"step": 11950
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 6.83371298405467,
|
| 1753 |
+
"grad_norm": 0.018831729888916016,
|
| 1754 |
+
"learning_rate": 4.88122356003905e-07,
|
| 1755 |
+
"loss": 0.0033,
|
| 1756 |
+
"step": 12000
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 6.862186788154897,
|
| 1760 |
+
"grad_norm": 0.0018483272287994623,
|
| 1761 |
+
"learning_rate": 4.067686300032542e-07,
|
| 1762 |
+
"loss": 0.0022,
|
| 1763 |
+
"step": 12050
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 6.890660592255125,
|
| 1767 |
+
"grad_norm": 0.003242627950385213,
|
| 1768 |
+
"learning_rate": 3.2541490400260336e-07,
|
| 1769 |
+
"loss": 0.0043,
|
| 1770 |
+
"step": 12100
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 6.9191343963553535,
|
| 1774 |
+
"grad_norm": 0.002707740291953087,
|
| 1775 |
+
"learning_rate": 2.440611780019525e-07,
|
| 1776 |
+
"loss": 0.0017,
|
| 1777 |
+
"step": 12150
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 6.947608200455581,
|
| 1781 |
+
"grad_norm": 0.21824024617671967,
|
| 1782 |
+
"learning_rate": 1.6270745200130168e-07,
|
| 1783 |
+
"loss": 0.0069,
|
| 1784 |
+
"step": 12200
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 6.976082004555809,
|
| 1788 |
+
"grad_norm": 0.008130647242069244,
|
| 1789 |
+
"learning_rate": 8.135372600065084e-08,
|
| 1790 |
+
"loss": 0.0047,
|
| 1791 |
+
"step": 12250
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 7.0,
|
| 1795 |
+
"eval_f1": 0.9563834240267894,
|
| 1796 |
+
"eval_loss": 0.052909377962350845,
|
| 1797 |
+
"eval_precision": 0.9515242378810594,
|
| 1798 |
+
"eval_recall": 0.9612924941097274,
|
| 1799 |
+
"eval_runtime": 3.7426,
|
| 1800 |
+
"eval_samples_per_second": 868.392,
|
| 1801 |
+
"eval_steps_per_second": 108.749,
|
| 1802 |
+
"step": 12292
|
| 1803 |
+
}
|
| 1804 |
+
],
|
| 1805 |
+
"logging_steps": 50,
|
| 1806 |
+
"max_steps": 12292,
|
| 1807 |
+
"num_input_tokens_seen": 0,
|
| 1808 |
+
"num_train_epochs": 7,
|
| 1809 |
+
"save_steps": 500,
|
| 1810 |
+
"stateful_callbacks": {
|
| 1811 |
+
"TrainerControl": {
|
| 1812 |
+
"args": {
|
| 1813 |
+
"should_epoch_stop": false,
|
| 1814 |
+
"should_evaluate": false,
|
| 1815 |
+
"should_log": false,
|
| 1816 |
+
"should_save": true,
|
| 1817 |
+
"should_training_stop": true
|
| 1818 |
+
},
|
| 1819 |
+
"attributes": {}
|
| 1820 |
+
}
|
| 1821 |
+
},
|
| 1822 |
+
"total_flos": 2418024281348142.0,
|
| 1823 |
+
"train_batch_size": 8,
|
| 1824 |
+
"trial_name": null,
|
| 1825 |
+
"trial_params": null
|
| 1826 |
+
}
|
checkpoint-12292/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:15e08f60c5260f286c88c6054e532ee0950cb5316081a3a51d07b57f6a266f71
|
| 3 |
+
size 5368
|
checkpoint-1756/config.json
ADDED
|
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"XLMRobertaForTokenClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "O",
|
| 14 |
+
"1": "B-PER",
|
| 15 |
+
"2": "I-PER",
|
| 16 |
+
"3": "B-ORG",
|
| 17 |
+
"4": "I-ORG",
|
| 18 |
+
"5": "B-LOC",
|
| 19 |
+
"6": "I-LOC",
|
| 20 |
+
"7": "B-MISC",
|
| 21 |
+
"8": "I-MISC"
|
| 22 |
+
},
|
| 23 |
+
"initializer_range": 0.02,
|
| 24 |
+
"intermediate_size": 3072,
|
| 25 |
+
"label2id": {
|
| 26 |
+
"B-LOC": 5,
|
| 27 |
+
"B-MISC": 7,
|
| 28 |
+
"B-ORG": 3,
|
| 29 |
+
"B-PER": 1,
|
| 30 |
+
"I-LOC": 6,
|
| 31 |
+
"I-MISC": 8,
|
| 32 |
+
"I-ORG": 4,
|
| 33 |
+
"I-PER": 2,
|
| 34 |
+
"O": 0
|
| 35 |
+
},
|
| 36 |
+
"layer_norm_eps": 1e-05,
|
| 37 |
+
"max_position_embeddings": 514,
|
| 38 |
+
"model_type": "xlm-roberta",
|
| 39 |
+
"num_attention_heads": 12,
|
| 40 |
+
"num_hidden_layers": 12,
|
| 41 |
+
"output_past": true,
|
| 42 |
+
"pad_token_id": 1,
|
| 43 |
+
"position_embedding_type": "absolute",
|
| 44 |
+
"torch_dtype": "float32",
|
| 45 |
+
"transformers_version": "4.51.3",
|
| 46 |
+
"type_vocab_size": 1,
|
| 47 |
+
"use_cache": true,
|
| 48 |
+
"vocab_size": 250002
|
| 49 |
+
}
|
checkpoint-1756/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b877b1e180a68933eb90e92f3711348a19892e9e1c0e85f1c0c0ac31c900843d
|
| 3 |
+
size 1109863956
|
checkpoint-1756/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2787580e6c0160af6a4dc99a4e9ce84ae377ab3d6296a6416012dc8df5bfbdf
|
| 3 |
+
size 2219846202
|
checkpoint-1756/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8c3bcda270a9bbdc01c80b4119b83fd5b30aca083ec25b4059eed0da02a64608
|
| 3 |
+
size 14244
|
checkpoint-1756/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d81f85e057ecf3b4e4481a5b7b37c25917c66b4c3f8f132282d8a277499dbbb3
|
| 3 |
+
size 988
|
checkpoint-1756/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9284fedcb9e147844c4eb634bfce0b54551832557512f611ba8e58c50ee1d23f
|
| 3 |
+
size 1064
|
checkpoint-1756/special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": {
|
| 6 |
+
"content": "<mask>",
|
| 7 |
+
"lstrip": true,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
checkpoint-1756/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3a56def25aa40facc030ea8b0b87f3688e4b3c39eb8b45d5702b3a1300fe2a20
|
| 3 |
+
size 17082734
|
checkpoint-1756/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"250001": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": true,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 512,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizer",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
checkpoint-1756/trainer_state.json
ADDED
|
@@ -0,0 +1,290 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1756,
|
| 3 |
+
"best_metric": 0.9242246437552389,
|
| 4 |
+
"best_model_checkpoint": "C:\\Users\\Shara\\projects\\models\\run_16-lr_2e-05-acc_1-wd_0.01-bs_8-ep_7\\checkpoint-1756",
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1756,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.02847380410022779,
|
| 14 |
+
"grad_norm": 4.8302321434021,
|
| 15 |
+
"learning_rate": 1.992840872111943e-05,
|
| 16 |
+
"loss": 0.9994,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.05694760820045558,
|
| 21 |
+
"grad_norm": 3.4978649616241455,
|
| 22 |
+
"learning_rate": 1.9847054995118778e-05,
|
| 23 |
+
"loss": 0.3939,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.08542141230068337,
|
| 28 |
+
"grad_norm": 8.353672981262207,
|
| 29 |
+
"learning_rate": 1.9765701269118128e-05,
|
| 30 |
+
"loss": 0.2882,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.11389521640091116,
|
| 35 |
+
"grad_norm": 9.822871208190918,
|
| 36 |
+
"learning_rate": 1.9684347543117477e-05,
|
| 37 |
+
"loss": 0.1728,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.14236902050113895,
|
| 42 |
+
"grad_norm": 3.1411221027374268,
|
| 43 |
+
"learning_rate": 1.9602993817116827e-05,
|
| 44 |
+
"loss": 0.1251,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.17084282460136674,
|
| 49 |
+
"grad_norm": 8.493424415588379,
|
| 50 |
+
"learning_rate": 1.9521640091116173e-05,
|
| 51 |
+
"loss": 0.1405,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.19931662870159453,
|
| 56 |
+
"grad_norm": 4.757007122039795,
|
| 57 |
+
"learning_rate": 1.9440286365115523e-05,
|
| 58 |
+
"loss": 0.1431,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.22779043280182232,
|
| 63 |
+
"grad_norm": 4.339049816131592,
|
| 64 |
+
"learning_rate": 1.9358932639114873e-05,
|
| 65 |
+
"loss": 0.0912,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.25626423690205014,
|
| 70 |
+
"grad_norm": 14.540772438049316,
|
| 71 |
+
"learning_rate": 1.9277578913114222e-05,
|
| 72 |
+
"loss": 0.1048,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.2847380410022779,
|
| 77 |
+
"grad_norm": 10.503073692321777,
|
| 78 |
+
"learning_rate": 1.9196225187113572e-05,
|
| 79 |
+
"loss": 0.1224,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.3132118451025057,
|
| 84 |
+
"grad_norm": 2.738131284713745,
|
| 85 |
+
"learning_rate": 1.911487146111292e-05,
|
| 86 |
+
"loss": 0.1025,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.3416856492027335,
|
| 91 |
+
"grad_norm": 6.681102752685547,
|
| 92 |
+
"learning_rate": 1.903351773511227e-05,
|
| 93 |
+
"loss": 0.0869,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.3701594533029613,
|
| 98 |
+
"grad_norm": 1.4330339431762695,
|
| 99 |
+
"learning_rate": 1.895216400911162e-05,
|
| 100 |
+
"loss": 0.052,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.39863325740318906,
|
| 105 |
+
"grad_norm": 2.581470012664795,
|
| 106 |
+
"learning_rate": 1.8870810283110967e-05,
|
| 107 |
+
"loss": 0.0808,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.4271070615034169,
|
| 112 |
+
"grad_norm": 5.825100898742676,
|
| 113 |
+
"learning_rate": 1.8789456557110317e-05,
|
| 114 |
+
"loss": 0.0716,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.45558086560364464,
|
| 119 |
+
"grad_norm": 1.74166738986969,
|
| 120 |
+
"learning_rate": 1.8708102831109666e-05,
|
| 121 |
+
"loss": 0.0806,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.48405466970387245,
|
| 126 |
+
"grad_norm": 7.3097405433654785,
|
| 127 |
+
"learning_rate": 1.8626749105109016e-05,
|
| 128 |
+
"loss": 0.0666,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.5125284738041003,
|
| 133 |
+
"grad_norm": 2.220766067504883,
|
| 134 |
+
"learning_rate": 1.8545395379108362e-05,
|
| 135 |
+
"loss": 0.0768,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.541002277904328,
|
| 140 |
+
"grad_norm": 4.0566511154174805,
|
| 141 |
+
"learning_rate": 1.8464041653107715e-05,
|
| 142 |
+
"loss": 0.0707,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.5694760820045558,
|
| 147 |
+
"grad_norm": 0.2833240330219269,
|
| 148 |
+
"learning_rate": 1.8382687927107065e-05,
|
| 149 |
+
"loss": 0.0858,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.5979498861047836,
|
| 154 |
+
"grad_norm": 21.131481170654297,
|
| 155 |
+
"learning_rate": 1.8301334201106415e-05,
|
| 156 |
+
"loss": 0.0595,
|
| 157 |
+
"step": 1050
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.6264236902050114,
|
| 161 |
+
"grad_norm": 27.655433654785156,
|
| 162 |
+
"learning_rate": 1.821998047510576e-05,
|
| 163 |
+
"loss": 0.0466,
|
| 164 |
+
"step": 1100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.6548974943052391,
|
| 168 |
+
"grad_norm": 3.4721717834472656,
|
| 169 |
+
"learning_rate": 1.813862674910511e-05,
|
| 170 |
+
"loss": 0.0776,
|
| 171 |
+
"step": 1150
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.683371298405467,
|
| 175 |
+
"grad_norm": 2.4811441898345947,
|
| 176 |
+
"learning_rate": 1.805727302310446e-05,
|
| 177 |
+
"loss": 0.0647,
|
| 178 |
+
"step": 1200
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.7118451025056948,
|
| 182 |
+
"grad_norm": 6.169066429138184,
|
| 183 |
+
"learning_rate": 1.797591929710381e-05,
|
| 184 |
+
"loss": 0.0724,
|
| 185 |
+
"step": 1250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.7403189066059226,
|
| 189 |
+
"grad_norm": 8.984748840332031,
|
| 190 |
+
"learning_rate": 1.7894565571103156e-05,
|
| 191 |
+
"loss": 0.0838,
|
| 192 |
+
"step": 1300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.7687927107061503,
|
| 196 |
+
"grad_norm": 5.944505214691162,
|
| 197 |
+
"learning_rate": 1.7813211845102506e-05,
|
| 198 |
+
"loss": 0.0683,
|
| 199 |
+
"step": 1350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.7972665148063781,
|
| 203 |
+
"grad_norm": 0.28812381625175476,
|
| 204 |
+
"learning_rate": 1.7731858119101856e-05,
|
| 205 |
+
"loss": 0.0614,
|
| 206 |
+
"step": 1400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.8257403189066059,
|
| 210 |
+
"grad_norm": 6.176011085510254,
|
| 211 |
+
"learning_rate": 1.7650504393101205e-05,
|
| 212 |
+
"loss": 0.0615,
|
| 213 |
+
"step": 1450
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.8542141230068337,
|
| 217 |
+
"grad_norm": 2.5244405269622803,
|
| 218 |
+
"learning_rate": 1.7569150667100555e-05,
|
| 219 |
+
"loss": 0.0714,
|
| 220 |
+
"step": 1500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.8826879271070615,
|
| 224 |
+
"grad_norm": 3.3508074283599854,
|
| 225 |
+
"learning_rate": 1.7487796941099904e-05,
|
| 226 |
+
"loss": 0.0563,
|
| 227 |
+
"step": 1550
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.9111617312072893,
|
| 231 |
+
"grad_norm": 9.980842590332031,
|
| 232 |
+
"learning_rate": 1.7406443215099254e-05,
|
| 233 |
+
"loss": 0.0647,
|
| 234 |
+
"step": 1600
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.9396355353075171,
|
| 238 |
+
"grad_norm": 3.6959774494171143,
|
| 239 |
+
"learning_rate": 1.7325089489098604e-05,
|
| 240 |
+
"loss": 0.0682,
|
| 241 |
+
"step": 1650
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.9681093394077449,
|
| 245 |
+
"grad_norm": 0.626620888710022,
|
| 246 |
+
"learning_rate": 1.724373576309795e-05,
|
| 247 |
+
"loss": 0.0499,
|
| 248 |
+
"step": 1700
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.9965831435079726,
|
| 252 |
+
"grad_norm": 7.145672798156738,
|
| 253 |
+
"learning_rate": 1.71623820370973e-05,
|
| 254 |
+
"loss": 0.0569,
|
| 255 |
+
"step": 1750
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.0,
|
| 259 |
+
"eval_f1": 0.9242246437552389,
|
| 260 |
+
"eval_loss": 0.05600914731621742,
|
| 261 |
+
"eval_precision": 0.9206746826987308,
|
| 262 |
+
"eval_recall": 0.927802086839448,
|
| 263 |
+
"eval_runtime": 3.7538,
|
| 264 |
+
"eval_samples_per_second": 865.798,
|
| 265 |
+
"eval_steps_per_second": 108.425,
|
| 266 |
+
"step": 1756
|
| 267 |
+
}
|
| 268 |
+
],
|
| 269 |
+
"logging_steps": 50,
|
| 270 |
+
"max_steps": 12292,
|
| 271 |
+
"num_input_tokens_seen": 0,
|
| 272 |
+
"num_train_epochs": 7,
|
| 273 |
+
"save_steps": 500,
|
| 274 |
+
"stateful_callbacks": {
|
| 275 |
+
"TrainerControl": {
|
| 276 |
+
"args": {
|
| 277 |
+
"should_epoch_stop": false,
|
| 278 |
+
"should_evaluate": false,
|
| 279 |
+
"should_log": false,
|
| 280 |
+
"should_save": true,
|
| 281 |
+
"should_training_stop": false
|
| 282 |
+
},
|
| 283 |
+
"attributes": {}
|
| 284 |
+
}
|
| 285 |
+
},
|
| 286 |
+
"total_flos": 344842189594686.0,
|
| 287 |
+
"train_batch_size": 8,
|
| 288 |
+
"trial_name": null,
|
| 289 |
+
"trial_params": null
|
| 290 |
+
}
|
checkpoint-1756/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:15e08f60c5260f286c88c6054e532ee0950cb5316081a3a51d07b57f6a266f71
|
| 3 |
+
size 5368
|
checkpoint-3512/config.json
ADDED
|
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"XLMRobertaForTokenClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "O",
|
| 14 |
+
"1": "B-PER",
|
| 15 |
+
"2": "I-PER",
|
| 16 |
+
"3": "B-ORG",
|
| 17 |
+
"4": "I-ORG",
|
| 18 |
+
"5": "B-LOC",
|
| 19 |
+
"6": "I-LOC",
|
| 20 |
+
"7": "B-MISC",
|
| 21 |
+
"8": "I-MISC"
|
| 22 |
+
},
|
| 23 |
+
"initializer_range": 0.02,
|
| 24 |
+
"intermediate_size": 3072,
|
| 25 |
+
"label2id": {
|
| 26 |
+
"B-LOC": 5,
|
| 27 |
+
"B-MISC": 7,
|
| 28 |
+
"B-ORG": 3,
|
| 29 |
+
"B-PER": 1,
|
| 30 |
+
"I-LOC": 6,
|
| 31 |
+
"I-MISC": 8,
|
| 32 |
+
"I-ORG": 4,
|
| 33 |
+
"I-PER": 2,
|
| 34 |
+
"O": 0
|
| 35 |
+
},
|
| 36 |
+
"layer_norm_eps": 1e-05,
|
| 37 |
+
"max_position_embeddings": 514,
|
| 38 |
+
"model_type": "xlm-roberta",
|
| 39 |
+
"num_attention_heads": 12,
|
| 40 |
+
"num_hidden_layers": 12,
|
| 41 |
+
"output_past": true,
|
| 42 |
+
"pad_token_id": 1,
|
| 43 |
+
"position_embedding_type": "absolute",
|
| 44 |
+
"torch_dtype": "float32",
|
| 45 |
+
"transformers_version": "4.51.3",
|
| 46 |
+
"type_vocab_size": 1,
|
| 47 |
+
"use_cache": true,
|
| 48 |
+
"vocab_size": 250002
|
| 49 |
+
}
|
checkpoint-3512/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ddbcca6c4d038a5447fa9bf8ac31d433d751923346e36fedb24008cd64521f50
|
| 3 |
+
size 1109863956
|
checkpoint-3512/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f038c2d0551d74e2318c48245eab9294694194b187f0a6c8635336b32a4a3845
|
| 3 |
+
size 2219846202
|
checkpoint-3512/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0da10cefa2cb8a2be8114fc291a5c55cb96e1047f75b823fc7fb82422f79ce2f
|
| 3 |
+
size 14244
|
checkpoint-3512/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:34bcbd0d9500d8ef0ef7916a5d9e57b44de8cfee243837f363bfae8f0865c9c6
|
| 3 |
+
size 988
|
checkpoint-3512/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cbe09b89711b4ddfda2986487d23873a7e0ee48f39bf17155748cec806279077
|
| 3 |
+
size 1064
|
checkpoint-3512/special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": {
|
| 6 |
+
"content": "<mask>",
|
| 7 |
+
"lstrip": true,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
checkpoint-3512/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3a56def25aa40facc030ea8b0b87f3688e4b3c39eb8b45d5702b3a1300fe2a20
|
| 3 |
+
size 17082734
|
checkpoint-3512/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"250001": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": true,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 512,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizer",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
checkpoint-3512/trainer_state.json
ADDED
|
@@ -0,0 +1,546 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 3512,
|
| 3 |
+
"best_metric": 0.9406438631790744,
|
| 4 |
+
"best_model_checkpoint": "C:\\Users\\Shara\\projects\\models\\run_16-lr_2e-05-acc_1-wd_0.01-bs_8-ep_7\\checkpoint-3512",
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3512,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.02847380410022779,
|
| 14 |
+
"grad_norm": 4.8302321434021,
|
| 15 |
+
"learning_rate": 1.992840872111943e-05,
|
| 16 |
+
"loss": 0.9994,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.05694760820045558,
|
| 21 |
+
"grad_norm": 3.4978649616241455,
|
| 22 |
+
"learning_rate": 1.9847054995118778e-05,
|
| 23 |
+
"loss": 0.3939,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.08542141230068337,
|
| 28 |
+
"grad_norm": 8.353672981262207,
|
| 29 |
+
"learning_rate": 1.9765701269118128e-05,
|
| 30 |
+
"loss": 0.2882,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.11389521640091116,
|
| 35 |
+
"grad_norm": 9.822871208190918,
|
| 36 |
+
"learning_rate": 1.9684347543117477e-05,
|
| 37 |
+
"loss": 0.1728,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.14236902050113895,
|
| 42 |
+
"grad_norm": 3.1411221027374268,
|
| 43 |
+
"learning_rate": 1.9602993817116827e-05,
|
| 44 |
+
"loss": 0.1251,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.17084282460136674,
|
| 49 |
+
"grad_norm": 8.493424415588379,
|
| 50 |
+
"learning_rate": 1.9521640091116173e-05,
|
| 51 |
+
"loss": 0.1405,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.19931662870159453,
|
| 56 |
+
"grad_norm": 4.757007122039795,
|
| 57 |
+
"learning_rate": 1.9440286365115523e-05,
|
| 58 |
+
"loss": 0.1431,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.22779043280182232,
|
| 63 |
+
"grad_norm": 4.339049816131592,
|
| 64 |
+
"learning_rate": 1.9358932639114873e-05,
|
| 65 |
+
"loss": 0.0912,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.25626423690205014,
|
| 70 |
+
"grad_norm": 14.540772438049316,
|
| 71 |
+
"learning_rate": 1.9277578913114222e-05,
|
| 72 |
+
"loss": 0.1048,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.2847380410022779,
|
| 77 |
+
"grad_norm": 10.503073692321777,
|
| 78 |
+
"learning_rate": 1.9196225187113572e-05,
|
| 79 |
+
"loss": 0.1224,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.3132118451025057,
|
| 84 |
+
"grad_norm": 2.738131284713745,
|
| 85 |
+
"learning_rate": 1.911487146111292e-05,
|
| 86 |
+
"loss": 0.1025,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.3416856492027335,
|
| 91 |
+
"grad_norm": 6.681102752685547,
|
| 92 |
+
"learning_rate": 1.903351773511227e-05,
|
| 93 |
+
"loss": 0.0869,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.3701594533029613,
|
| 98 |
+
"grad_norm": 1.4330339431762695,
|
| 99 |
+
"learning_rate": 1.895216400911162e-05,
|
| 100 |
+
"loss": 0.052,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.39863325740318906,
|
| 105 |
+
"grad_norm": 2.581470012664795,
|
| 106 |
+
"learning_rate": 1.8870810283110967e-05,
|
| 107 |
+
"loss": 0.0808,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.4271070615034169,
|
| 112 |
+
"grad_norm": 5.825100898742676,
|
| 113 |
+
"learning_rate": 1.8789456557110317e-05,
|
| 114 |
+
"loss": 0.0716,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.45558086560364464,
|
| 119 |
+
"grad_norm": 1.74166738986969,
|
| 120 |
+
"learning_rate": 1.8708102831109666e-05,
|
| 121 |
+
"loss": 0.0806,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.48405466970387245,
|
| 126 |
+
"grad_norm": 7.3097405433654785,
|
| 127 |
+
"learning_rate": 1.8626749105109016e-05,
|
| 128 |
+
"loss": 0.0666,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.5125284738041003,
|
| 133 |
+
"grad_norm": 2.220766067504883,
|
| 134 |
+
"learning_rate": 1.8545395379108362e-05,
|
| 135 |
+
"loss": 0.0768,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.541002277904328,
|
| 140 |
+
"grad_norm": 4.0566511154174805,
|
| 141 |
+
"learning_rate": 1.8464041653107715e-05,
|
| 142 |
+
"loss": 0.0707,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.5694760820045558,
|
| 147 |
+
"grad_norm": 0.2833240330219269,
|
| 148 |
+
"learning_rate": 1.8382687927107065e-05,
|
| 149 |
+
"loss": 0.0858,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.5979498861047836,
|
| 154 |
+
"grad_norm": 21.131481170654297,
|
| 155 |
+
"learning_rate": 1.8301334201106415e-05,
|
| 156 |
+
"loss": 0.0595,
|
| 157 |
+
"step": 1050
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.6264236902050114,
|
| 161 |
+
"grad_norm": 27.655433654785156,
|
| 162 |
+
"learning_rate": 1.821998047510576e-05,
|
| 163 |
+
"loss": 0.0466,
|
| 164 |
+
"step": 1100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.6548974943052391,
|
| 168 |
+
"grad_norm": 3.4721717834472656,
|
| 169 |
+
"learning_rate": 1.813862674910511e-05,
|
| 170 |
+
"loss": 0.0776,
|
| 171 |
+
"step": 1150
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.683371298405467,
|
| 175 |
+
"grad_norm": 2.4811441898345947,
|
| 176 |
+
"learning_rate": 1.805727302310446e-05,
|
| 177 |
+
"loss": 0.0647,
|
| 178 |
+
"step": 1200
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.7118451025056948,
|
| 182 |
+
"grad_norm": 6.169066429138184,
|
| 183 |
+
"learning_rate": 1.797591929710381e-05,
|
| 184 |
+
"loss": 0.0724,
|
| 185 |
+
"step": 1250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.7403189066059226,
|
| 189 |
+
"grad_norm": 8.984748840332031,
|
| 190 |
+
"learning_rate": 1.7894565571103156e-05,
|
| 191 |
+
"loss": 0.0838,
|
| 192 |
+
"step": 1300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.7687927107061503,
|
| 196 |
+
"grad_norm": 5.944505214691162,
|
| 197 |
+
"learning_rate": 1.7813211845102506e-05,
|
| 198 |
+
"loss": 0.0683,
|
| 199 |
+
"step": 1350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.7972665148063781,
|
| 203 |
+
"grad_norm": 0.28812381625175476,
|
| 204 |
+
"learning_rate": 1.7731858119101856e-05,
|
| 205 |
+
"loss": 0.0614,
|
| 206 |
+
"step": 1400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.8257403189066059,
|
| 210 |
+
"grad_norm": 6.176011085510254,
|
| 211 |
+
"learning_rate": 1.7650504393101205e-05,
|
| 212 |
+
"loss": 0.0615,
|
| 213 |
+
"step": 1450
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.8542141230068337,
|
| 217 |
+
"grad_norm": 2.5244405269622803,
|
| 218 |
+
"learning_rate": 1.7569150667100555e-05,
|
| 219 |
+
"loss": 0.0714,
|
| 220 |
+
"step": 1500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.8826879271070615,
|
| 224 |
+
"grad_norm": 3.3508074283599854,
|
| 225 |
+
"learning_rate": 1.7487796941099904e-05,
|
| 226 |
+
"loss": 0.0563,
|
| 227 |
+
"step": 1550
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.9111617312072893,
|
| 231 |
+
"grad_norm": 9.980842590332031,
|
| 232 |
+
"learning_rate": 1.7406443215099254e-05,
|
| 233 |
+
"loss": 0.0647,
|
| 234 |
+
"step": 1600
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.9396355353075171,
|
| 238 |
+
"grad_norm": 3.6959774494171143,
|
| 239 |
+
"learning_rate": 1.7325089489098604e-05,
|
| 240 |
+
"loss": 0.0682,
|
| 241 |
+
"step": 1650
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.9681093394077449,
|
| 245 |
+
"grad_norm": 0.626620888710022,
|
| 246 |
+
"learning_rate": 1.724373576309795e-05,
|
| 247 |
+
"loss": 0.0499,
|
| 248 |
+
"step": 1700
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.9965831435079726,
|
| 252 |
+
"grad_norm": 7.145672798156738,
|
| 253 |
+
"learning_rate": 1.71623820370973e-05,
|
| 254 |
+
"loss": 0.0569,
|
| 255 |
+
"step": 1750
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.0,
|
| 259 |
+
"eval_f1": 0.9242246437552389,
|
| 260 |
+
"eval_loss": 0.05600914731621742,
|
| 261 |
+
"eval_precision": 0.9206746826987308,
|
| 262 |
+
"eval_recall": 0.927802086839448,
|
| 263 |
+
"eval_runtime": 3.7538,
|
| 264 |
+
"eval_samples_per_second": 865.798,
|
| 265 |
+
"eval_steps_per_second": 108.425,
|
| 266 |
+
"step": 1756
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.0250569476082005,
|
| 270 |
+
"grad_norm": 4.280463695526123,
|
| 271 |
+
"learning_rate": 1.708102831109665e-05,
|
| 272 |
+
"loss": 0.0673,
|
| 273 |
+
"step": 1800
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.0535307517084282,
|
| 277 |
+
"grad_norm": 0.06692535430192947,
|
| 278 |
+
"learning_rate": 1.6999674585096e-05,
|
| 279 |
+
"loss": 0.0403,
|
| 280 |
+
"step": 1850
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 1.082004555808656,
|
| 284 |
+
"grad_norm": 0.03789375349879265,
|
| 285 |
+
"learning_rate": 1.691832085909535e-05,
|
| 286 |
+
"loss": 0.0496,
|
| 287 |
+
"step": 1900
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.1104783599088839,
|
| 291 |
+
"grad_norm": 3.698348045349121,
|
| 292 |
+
"learning_rate": 1.6836967133094698e-05,
|
| 293 |
+
"loss": 0.0381,
|
| 294 |
+
"step": 1950
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.1389521640091116,
|
| 298 |
+
"grad_norm": 2.753032684326172,
|
| 299 |
+
"learning_rate": 1.6755613407094048e-05,
|
| 300 |
+
"loss": 0.0432,
|
| 301 |
+
"step": 2000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.1674259681093395,
|
| 305 |
+
"grad_norm": 0.06336130946874619,
|
| 306 |
+
"learning_rate": 1.6674259681093398e-05,
|
| 307 |
+
"loss": 0.0482,
|
| 308 |
+
"step": 2050
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.1958997722095672,
|
| 312 |
+
"grad_norm": 3.9503254890441895,
|
| 313 |
+
"learning_rate": 1.6592905955092744e-05,
|
| 314 |
+
"loss": 0.0436,
|
| 315 |
+
"step": 2100
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.224373576309795,
|
| 319 |
+
"grad_norm": 0.08679741621017456,
|
| 320 |
+
"learning_rate": 1.6511552229092093e-05,
|
| 321 |
+
"loss": 0.0346,
|
| 322 |
+
"step": 2150
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 1.2528473804100229,
|
| 326 |
+
"grad_norm": 2.1995127201080322,
|
| 327 |
+
"learning_rate": 1.6430198503091443e-05,
|
| 328 |
+
"loss": 0.0312,
|
| 329 |
+
"step": 2200
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.2813211845102506,
|
| 333 |
+
"grad_norm": 5.7709808349609375,
|
| 334 |
+
"learning_rate": 1.634884477709079e-05,
|
| 335 |
+
"loss": 0.032,
|
| 336 |
+
"step": 2250
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 1.3097949886104785,
|
| 340 |
+
"grad_norm": 0.2031829059123993,
|
| 341 |
+
"learning_rate": 1.626749105109014e-05,
|
| 342 |
+
"loss": 0.053,
|
| 343 |
+
"step": 2300
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 1.3382687927107062,
|
| 347 |
+
"grad_norm": 4.292157173156738,
|
| 348 |
+
"learning_rate": 1.6186137325089492e-05,
|
| 349 |
+
"loss": 0.0413,
|
| 350 |
+
"step": 2350
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.366742596810934,
|
| 354 |
+
"grad_norm": 0.07821401953697205,
|
| 355 |
+
"learning_rate": 1.6104783599088842e-05,
|
| 356 |
+
"loss": 0.0304,
|
| 357 |
+
"step": 2400
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.3952164009111616,
|
| 361 |
+
"grad_norm": 0.5930206179618835,
|
| 362 |
+
"learning_rate": 1.6023429873088188e-05,
|
| 363 |
+
"loss": 0.0525,
|
| 364 |
+
"step": 2450
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 1.4236902050113895,
|
| 368 |
+
"grad_norm": 1.6114712953567505,
|
| 369 |
+
"learning_rate": 1.5942076147087538e-05,
|
| 370 |
+
"loss": 0.0391,
|
| 371 |
+
"step": 2500
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.4521640091116172,
|
| 375 |
+
"grad_norm": 0.28632065653800964,
|
| 376 |
+
"learning_rate": 1.5860722421086887e-05,
|
| 377 |
+
"loss": 0.0491,
|
| 378 |
+
"step": 2550
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 1.4806378132118452,
|
| 382 |
+
"grad_norm": 1.0017669200897217,
|
| 383 |
+
"learning_rate": 1.5779368695086237e-05,
|
| 384 |
+
"loss": 0.0395,
|
| 385 |
+
"step": 2600
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 1.5091116173120729,
|
| 389 |
+
"grad_norm": 0.03409272059798241,
|
| 390 |
+
"learning_rate": 1.5698014969085583e-05,
|
| 391 |
+
"loss": 0.0399,
|
| 392 |
+
"step": 2650
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.5375854214123006,
|
| 396 |
+
"grad_norm": 3.5108702182769775,
|
| 397 |
+
"learning_rate": 1.5616661243084933e-05,
|
| 398 |
+
"loss": 0.0424,
|
| 399 |
+
"step": 2700
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.5660592255125285,
|
| 403 |
+
"grad_norm": 3.5875349044799805,
|
| 404 |
+
"learning_rate": 1.5535307517084283e-05,
|
| 405 |
+
"loss": 0.0497,
|
| 406 |
+
"step": 2750
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.5945330296127562,
|
| 410 |
+
"grad_norm": 0.10498251020908356,
|
| 411 |
+
"learning_rate": 1.5453953791083632e-05,
|
| 412 |
+
"loss": 0.0461,
|
| 413 |
+
"step": 2800
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.6230068337129842,
|
| 417 |
+
"grad_norm": 7.618963718414307,
|
| 418 |
+
"learning_rate": 1.5374227139602996e-05,
|
| 419 |
+
"loss": 0.0394,
|
| 420 |
+
"step": 2850
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 1.6514806378132119,
|
| 424 |
+
"grad_norm": 3.3920793533325195,
|
| 425 |
+
"learning_rate": 1.5292873413602346e-05,
|
| 426 |
+
"loss": 0.0339,
|
| 427 |
+
"step": 2900
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 1.6799544419134396,
|
| 431 |
+
"grad_norm": 0.04736129939556122,
|
| 432 |
+
"learning_rate": 1.5211519687601692e-05,
|
| 433 |
+
"loss": 0.0358,
|
| 434 |
+
"step": 2950
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.7084282460136673,
|
| 438 |
+
"grad_norm": 0.17463508248329163,
|
| 439 |
+
"learning_rate": 1.5130165961601042e-05,
|
| 440 |
+
"loss": 0.034,
|
| 441 |
+
"step": 3000
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 1.7369020501138952,
|
| 445 |
+
"grad_norm": 0.058148205280303955,
|
| 446 |
+
"learning_rate": 1.5048812235600392e-05,
|
| 447 |
+
"loss": 0.0317,
|
| 448 |
+
"step": 3050
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 1.7653758542141231,
|
| 452 |
+
"grad_norm": 3.5262815952301025,
|
| 453 |
+
"learning_rate": 1.4967458509599741e-05,
|
| 454 |
+
"loss": 0.0477,
|
| 455 |
+
"step": 3100
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 1.7938496583143508,
|
| 459 |
+
"grad_norm": 2.971937894821167,
|
| 460 |
+
"learning_rate": 1.488610478359909e-05,
|
| 461 |
+
"loss": 0.0364,
|
| 462 |
+
"step": 3150
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 1.8223234624145785,
|
| 466 |
+
"grad_norm": 0.05181724950671196,
|
| 467 |
+
"learning_rate": 1.4804751057598439e-05,
|
| 468 |
+
"loss": 0.0394,
|
| 469 |
+
"step": 3200
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 1.8507972665148062,
|
| 473 |
+
"grad_norm": 0.2712370455265045,
|
| 474 |
+
"learning_rate": 1.4723397331597789e-05,
|
| 475 |
+
"loss": 0.0456,
|
| 476 |
+
"step": 3250
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.8792710706150342,
|
| 480 |
+
"grad_norm": 1.608622431755066,
|
| 481 |
+
"learning_rate": 1.4642043605597138e-05,
|
| 482 |
+
"loss": 0.05,
|
| 483 |
+
"step": 3300
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 1.907744874715262,
|
| 487 |
+
"grad_norm": 5.2423529624938965,
|
| 488 |
+
"learning_rate": 1.4560689879596486e-05,
|
| 489 |
+
"loss": 0.0371,
|
| 490 |
+
"step": 3350
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 1.9362186788154898,
|
| 494 |
+
"grad_norm": 0.04786211624741554,
|
| 495 |
+
"learning_rate": 1.4479336153595836e-05,
|
| 496 |
+
"loss": 0.0329,
|
| 497 |
+
"step": 3400
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 1.9646924829157175,
|
| 501 |
+
"grad_norm": 0.5264931917190552,
|
| 502 |
+
"learning_rate": 1.4397982427595185e-05,
|
| 503 |
+
"loss": 0.0379,
|
| 504 |
+
"step": 3450
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 1.9931662870159452,
|
| 508 |
+
"grad_norm": 0.236006498336792,
|
| 509 |
+
"learning_rate": 1.4316628701594535e-05,
|
| 510 |
+
"loss": 0.0253,
|
| 511 |
+
"step": 3500
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 2.0,
|
| 515 |
+
"eval_f1": 0.9406438631790744,
|
| 516 |
+
"eval_loss": 0.05223705992102623,
|
| 517 |
+
"eval_precision": 0.9371867691279653,
|
| 518 |
+
"eval_recall": 0.9441265567149109,
|
| 519 |
+
"eval_runtime": 3.7109,
|
| 520 |
+
"eval_samples_per_second": 875.791,
|
| 521 |
+
"eval_steps_per_second": 109.676,
|
| 522 |
+
"step": 3512
|
| 523 |
+
}
|
| 524 |
+
],
|
| 525 |
+
"logging_steps": 50,
|
| 526 |
+
"max_steps": 12292,
|
| 527 |
+
"num_input_tokens_seen": 0,
|
| 528 |
+
"num_train_epochs": 7,
|
| 529 |
+
"save_steps": 500,
|
| 530 |
+
"stateful_callbacks": {
|
| 531 |
+
"TrainerControl": {
|
| 532 |
+
"args": {
|
| 533 |
+
"should_epoch_stop": false,
|
| 534 |
+
"should_evaluate": false,
|
| 535 |
+
"should_log": false,
|
| 536 |
+
"should_save": true,
|
| 537 |
+
"should_training_stop": false
|
| 538 |
+
},
|
| 539 |
+
"attributes": {}
|
| 540 |
+
}
|
| 541 |
+
},
|
| 542 |
+
"total_flos": 688219085312226.0,
|
| 543 |
+
"train_batch_size": 8,
|
| 544 |
+
"trial_name": null,
|
| 545 |
+
"trial_params": null
|
| 546 |
+
}
|
checkpoint-3512/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:15e08f60c5260f286c88c6054e532ee0950cb5316081a3a51d07b57f6a266f71
|
| 3 |
+
size 5368
|
checkpoint-5268/config.json
ADDED
|
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"XLMRobertaForTokenClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "O",
|
| 14 |
+
"1": "B-PER",
|
| 15 |
+
"2": "I-PER",
|
| 16 |
+
"3": "B-ORG",
|
| 17 |
+
"4": "I-ORG",
|
| 18 |
+
"5": "B-LOC",
|
| 19 |
+
"6": "I-LOC",
|
| 20 |
+
"7": "B-MISC",
|
| 21 |
+
"8": "I-MISC"
|
| 22 |
+
},
|
| 23 |
+
"initializer_range": 0.02,
|
| 24 |
+
"intermediate_size": 3072,
|
| 25 |
+
"label2id": {
|
| 26 |
+
"B-LOC": 5,
|
| 27 |
+
"B-MISC": 7,
|
| 28 |
+
"B-ORG": 3,
|
| 29 |
+
"B-PER": 1,
|
| 30 |
+
"I-LOC": 6,
|
| 31 |
+
"I-MISC": 8,
|
| 32 |
+
"I-ORG": 4,
|
| 33 |
+
"I-PER": 2,
|
| 34 |
+
"O": 0
|
| 35 |
+
},
|
| 36 |
+
"layer_norm_eps": 1e-05,
|
| 37 |
+
"max_position_embeddings": 514,
|
| 38 |
+
"model_type": "xlm-roberta",
|
| 39 |
+
"num_attention_heads": 12,
|
| 40 |
+
"num_hidden_layers": 12,
|
| 41 |
+
"output_past": true,
|
| 42 |
+
"pad_token_id": 1,
|
| 43 |
+
"position_embedding_type": "absolute",
|
| 44 |
+
"torch_dtype": "float32",
|
| 45 |
+
"transformers_version": "4.51.3",
|
| 46 |
+
"type_vocab_size": 1,
|
| 47 |
+
"use_cache": true,
|
| 48 |
+
"vocab_size": 250002
|
| 49 |
+
}
|
checkpoint-5268/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6b786779ce4d72c2e34b9d4fc244b422821fb82aa9e9496830a4f8a217e8633f
|
| 3 |
+
size 1109863956
|
checkpoint-5268/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:05d411b3ef7b4d64bdb57d539abb0b4f6f1d8c12b5b8383a2c868aac083b2256
|
| 3 |
+
size 2219846202
|
checkpoint-5268/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:773a0f008a7607f68cb3da99962235e5d2984b9344734097cf3b25025846c9bc
|
| 3 |
+
size 14244
|
checkpoint-5268/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5a3533b1927eb7f3b5998624c4eb9b6aa2b02868d4fa466b42a5c27cecc77725
|
| 3 |
+
size 988
|