Upload folder using huggingface_hub
Browse files- distilbert_ranking.onnx +1 -1
- distilbert_ranking/checkpoint-3500/config.json +25 -0
- distilbert_ranking/checkpoint-3500/model.safetensors +3 -0
- distilbert_ranking/checkpoint-3500/optimizer.pt +3 -0
- distilbert_ranking/checkpoint-3500/rng_state.pth +3 -0
- distilbert_ranking/checkpoint-3500/scheduler.pt +3 -0
- distilbert_ranking/checkpoint-3500/special_tokens_map.json +7 -0
- distilbert_ranking/checkpoint-3500/tokenizer_config.json +57 -0
- distilbert_ranking/checkpoint-3500/trainer_state.json +2547 -0
- distilbert_ranking/checkpoint-3500/training_args.bin +3 -0
- distilbert_ranking/checkpoint-3500/vocab.txt +0 -0
- distilbert_ranking/checkpoint-3940/config.json +25 -0
- distilbert_ranking/checkpoint-3940/model.safetensors +3 -0
- distilbert_ranking/checkpoint-3940/optimizer.pt +3 -0
- distilbert_ranking/checkpoint-3940/rng_state.pth +3 -0
- distilbert_ranking/checkpoint-3940/scheduler.pt +3 -0
- distilbert_ranking/checkpoint-3940/special_tokens_map.json +7 -0
- distilbert_ranking/checkpoint-3940/tokenizer_config.json +57 -0
- distilbert_ranking/checkpoint-3940/trainer_state.json +2863 -0
- distilbert_ranking/checkpoint-3940/training_args.bin +3 -0
- distilbert_ranking/checkpoint-3940/vocab.txt +0 -0
- distilbert_ranking_final/model.safetensors +1 -1
- distilbert_ranking_final/training_args.bin +1 -1
- logs/events.out.tfevents.1726931188.88832d4e4248.36.0 +3 -0
distilbert_ranking.onnx
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 267918628
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:035a6396973b071c3117b8a6c540a0e124dc5b443fd316afe97f4ea4fc5a44f6
|
| 3 |
size 267918628
|
distilbert_ranking/checkpoint-3500/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "distilbert-base-uncased",
|
| 3 |
+
"activation": "gelu",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"DistilBertForSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_dropout": 0.1,
|
| 8 |
+
"dim": 768,
|
| 9 |
+
"dropout": 0.1,
|
| 10 |
+
"hidden_dim": 3072,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"max_position_embeddings": 512,
|
| 13 |
+
"model_type": "distilbert",
|
| 14 |
+
"n_heads": 12,
|
| 15 |
+
"n_layers": 6,
|
| 16 |
+
"pad_token_id": 0,
|
| 17 |
+
"problem_type": "single_label_classification",
|
| 18 |
+
"qa_dropout": 0.1,
|
| 19 |
+
"seq_classif_dropout": 0.2,
|
| 20 |
+
"sinusoidal_pos_embds": false,
|
| 21 |
+
"tie_weights_": true,
|
| 22 |
+
"torch_dtype": "float32",
|
| 23 |
+
"transformers_version": "4.44.0",
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
distilbert_ranking/checkpoint-3500/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:45436594dbc9259a00e4ae5f8f20bcc57140caff928c3438465695c609cf775e
|
| 3 |
+
size 267832560
|
distilbert_ranking/checkpoint-3500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:513f45967a65197d4b32ed72d6d471be96399f2136b16b62bb00337fae90aefd
|
| 3 |
+
size 535727290
|
distilbert_ranking/checkpoint-3500/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:719e104db35e32b713094916a14de6afb115ab0286453c8e1018611840e62b1c
|
| 3 |
+
size 14244
|
distilbert_ranking/checkpoint-3500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2fec5ec49252a2e86f8ab1f30dd99086eacecb046d760410207f3c90e1e7b393
|
| 3 |
+
size 1064
|
distilbert_ranking/checkpoint-3500/special_tokens_map.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": "[CLS]",
|
| 3 |
+
"mask_token": "[MASK]",
|
| 4 |
+
"pad_token": "[PAD]",
|
| 5 |
+
"sep_token": "[SEP]",
|
| 6 |
+
"unk_token": "[UNK]"
|
| 7 |
+
}
|
distilbert_ranking/checkpoint-3500/tokenizer_config.json
ADDED
|
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"100": {
|
| 12 |
+
"content": "[UNK]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"101": {
|
| 20 |
+
"content": "[CLS]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"102": {
|
| 28 |
+
"content": "[SEP]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"103": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"clean_up_tokenization_spaces": true,
|
| 45 |
+
"cls_token": "[CLS]",
|
| 46 |
+
"do_basic_tokenize": true,
|
| 47 |
+
"do_lower_case": true,
|
| 48 |
+
"mask_token": "[MASK]",
|
| 49 |
+
"model_max_length": 512,
|
| 50 |
+
"never_split": null,
|
| 51 |
+
"pad_token": "[PAD]",
|
| 52 |
+
"sep_token": "[SEP]",
|
| 53 |
+
"strip_accents": null,
|
| 54 |
+
"tokenize_chinese_chars": true,
|
| 55 |
+
"tokenizer_class": "DistilBertTokenizer",
|
| 56 |
+
"unk_token": "[UNK]"
|
| 57 |
+
}
|
distilbert_ranking/checkpoint-3500/trainer_state.json
ADDED
|
@@ -0,0 +1,2547 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 8.883248730964468,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.025380710659898477,
|
| 13 |
+
"grad_norm": 1.7737329006195068,
|
| 14 |
+
"learning_rate": 4.9873096446700515e-05,
|
| 15 |
+
"loss": 0.4425,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.050761421319796954,
|
| 20 |
+
"grad_norm": 1.6950019598007202,
|
| 21 |
+
"learning_rate": 4.9746192893401014e-05,
|
| 22 |
+
"loss": 0.5594,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.07614213197969544,
|
| 27 |
+
"grad_norm": 3.7862815856933594,
|
| 28 |
+
"learning_rate": 4.961928934010153e-05,
|
| 29 |
+
"loss": 0.5435,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.10152284263959391,
|
| 34 |
+
"grad_norm": 2.450183868408203,
|
| 35 |
+
"learning_rate": 4.949238578680203e-05,
|
| 36 |
+
"loss": 0.5429,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.12690355329949238,
|
| 41 |
+
"grad_norm": 1.6361172199249268,
|
| 42 |
+
"learning_rate": 4.936548223350254e-05,
|
| 43 |
+
"loss": 0.6185,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.15228426395939088,
|
| 48 |
+
"grad_norm": 1.3755431175231934,
|
| 49 |
+
"learning_rate": 4.9238578680203045e-05,
|
| 50 |
+
"loss": 0.5633,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.17766497461928935,
|
| 55 |
+
"grad_norm": 2.1979947090148926,
|
| 56 |
+
"learning_rate": 4.911167512690356e-05,
|
| 57 |
+
"loss": 0.5031,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.20304568527918782,
|
| 62 |
+
"grad_norm": 1.493660569190979,
|
| 63 |
+
"learning_rate": 4.8984771573604064e-05,
|
| 64 |
+
"loss": 0.5942,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.22842639593908629,
|
| 69 |
+
"grad_norm": 1.642001748085022,
|
| 70 |
+
"learning_rate": 4.885786802030457e-05,
|
| 71 |
+
"loss": 0.5513,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.25380710659898476,
|
| 76 |
+
"grad_norm": 4.11527681350708,
|
| 77 |
+
"learning_rate": 4.873096446700508e-05,
|
| 78 |
+
"loss": 0.5544,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.27918781725888325,
|
| 83 |
+
"grad_norm": 1.966141700744629,
|
| 84 |
+
"learning_rate": 4.860406091370558e-05,
|
| 85 |
+
"loss": 0.5311,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.30456852791878175,
|
| 90 |
+
"grad_norm": 1.490822196006775,
|
| 91 |
+
"learning_rate": 4.8477157360406095e-05,
|
| 92 |
+
"loss": 0.4546,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.3299492385786802,
|
| 97 |
+
"grad_norm": 2.575047254562378,
|
| 98 |
+
"learning_rate": 4.83502538071066e-05,
|
| 99 |
+
"loss": 0.4532,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.3553299492385787,
|
| 104 |
+
"grad_norm": 1.9474763870239258,
|
| 105 |
+
"learning_rate": 4.822335025380711e-05,
|
| 106 |
+
"loss": 0.4499,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.38071065989847713,
|
| 111 |
+
"grad_norm": 2.9648075103759766,
|
| 112 |
+
"learning_rate": 4.809644670050762e-05,
|
| 113 |
+
"loss": 0.5425,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.40609137055837563,
|
| 118 |
+
"grad_norm": 1.9473344087600708,
|
| 119 |
+
"learning_rate": 4.7969543147208126e-05,
|
| 120 |
+
"loss": 0.529,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.43147208121827413,
|
| 125 |
+
"grad_norm": 2.3414478302001953,
|
| 126 |
+
"learning_rate": 4.784263959390863e-05,
|
| 127 |
+
"loss": 0.5281,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.45685279187817257,
|
| 132 |
+
"grad_norm": 1.2157188653945923,
|
| 133 |
+
"learning_rate": 4.771573604060914e-05,
|
| 134 |
+
"loss": 0.4963,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.48223350253807107,
|
| 139 |
+
"grad_norm": 3.3916754722595215,
|
| 140 |
+
"learning_rate": 4.758883248730965e-05,
|
| 141 |
+
"loss": 0.4245,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.5076142131979695,
|
| 146 |
+
"grad_norm": 3.030148983001709,
|
| 147 |
+
"learning_rate": 4.746192893401015e-05,
|
| 148 |
+
"loss": 0.3352,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.5329949238578681,
|
| 153 |
+
"grad_norm": 2.1008453369140625,
|
| 154 |
+
"learning_rate": 4.733502538071066e-05,
|
| 155 |
+
"loss": 0.6109,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.5583756345177665,
|
| 160 |
+
"grad_norm": 1.6099292039871216,
|
| 161 |
+
"learning_rate": 4.7208121827411175e-05,
|
| 162 |
+
"loss": 0.5327,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.583756345177665,
|
| 167 |
+
"grad_norm": 1.7559535503387451,
|
| 168 |
+
"learning_rate": 4.7081218274111674e-05,
|
| 169 |
+
"loss": 0.4475,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.6091370558375635,
|
| 174 |
+
"grad_norm": 1.5413782596588135,
|
| 175 |
+
"learning_rate": 4.695431472081219e-05,
|
| 176 |
+
"loss": 0.534,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.6345177664974619,
|
| 181 |
+
"grad_norm": 1.810226321220398,
|
| 182 |
+
"learning_rate": 4.682741116751269e-05,
|
| 183 |
+
"loss": 0.5739,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.6598984771573604,
|
| 188 |
+
"grad_norm": 3.676072120666504,
|
| 189 |
+
"learning_rate": 4.67005076142132e-05,
|
| 190 |
+
"loss": 0.4702,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.6852791878172588,
|
| 195 |
+
"grad_norm": 2.691065788269043,
|
| 196 |
+
"learning_rate": 4.6573604060913705e-05,
|
| 197 |
+
"loss": 0.5858,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.7106598984771574,
|
| 202 |
+
"grad_norm": 1.8071280717849731,
|
| 203 |
+
"learning_rate": 4.644670050761422e-05,
|
| 204 |
+
"loss": 0.5294,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.7360406091370558,
|
| 209 |
+
"grad_norm": 4.747298240661621,
|
| 210 |
+
"learning_rate": 4.631979695431472e-05,
|
| 211 |
+
"loss": 0.4718,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.7614213197969543,
|
| 216 |
+
"grad_norm": 2.4223101139068604,
|
| 217 |
+
"learning_rate": 4.619289340101523e-05,
|
| 218 |
+
"loss": 0.4712,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.7868020304568528,
|
| 223 |
+
"grad_norm": 2.2267863750457764,
|
| 224 |
+
"learning_rate": 4.606598984771574e-05,
|
| 225 |
+
"loss": 0.5829,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.8121827411167513,
|
| 230 |
+
"grad_norm": 2.0561490058898926,
|
| 231 |
+
"learning_rate": 4.593908629441624e-05,
|
| 232 |
+
"loss": 0.4033,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.8375634517766497,
|
| 237 |
+
"grad_norm": 3.0562074184417725,
|
| 238 |
+
"learning_rate": 4.5812182741116755e-05,
|
| 239 |
+
"loss": 0.6535,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.8629441624365483,
|
| 244 |
+
"grad_norm": 3.355173110961914,
|
| 245 |
+
"learning_rate": 4.568527918781726e-05,
|
| 246 |
+
"loss": 0.4635,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.8883248730964467,
|
| 251 |
+
"grad_norm": 4.357395172119141,
|
| 252 |
+
"learning_rate": 4.555837563451777e-05,
|
| 253 |
+
"loss": 0.4995,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.9137055837563451,
|
| 258 |
+
"grad_norm": 1.0308293104171753,
|
| 259 |
+
"learning_rate": 4.543147208121827e-05,
|
| 260 |
+
"loss": 0.5207,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.9390862944162437,
|
| 265 |
+
"grad_norm": 1.9433645009994507,
|
| 266 |
+
"learning_rate": 4.5304568527918786e-05,
|
| 267 |
+
"loss": 0.2991,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.9644670050761421,
|
| 272 |
+
"grad_norm": 3.8325693607330322,
|
| 273 |
+
"learning_rate": 4.517766497461929e-05,
|
| 274 |
+
"loss": 0.6014,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.9898477157360406,
|
| 279 |
+
"grad_norm": 1.0364559888839722,
|
| 280 |
+
"learning_rate": 4.50507614213198e-05,
|
| 281 |
+
"loss": 0.3693,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.0,
|
| 286 |
+
"eval_loss": 0.49558719992637634,
|
| 287 |
+
"eval_runtime": 8.8144,
|
| 288 |
+
"eval_samples_per_second": 89.4,
|
| 289 |
+
"eval_steps_per_second": 11.232,
|
| 290 |
+
"step": 394
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.015228426395939,
|
| 294 |
+
"grad_norm": 2.267185926437378,
|
| 295 |
+
"learning_rate": 4.492385786802031e-05,
|
| 296 |
+
"loss": 0.5173,
|
| 297 |
+
"step": 400
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.0406091370558375,
|
| 301 |
+
"grad_norm": 1.7508792877197266,
|
| 302 |
+
"learning_rate": 4.479695431472081e-05,
|
| 303 |
+
"loss": 0.4722,
|
| 304 |
+
"step": 410
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 1.0659898477157361,
|
| 308 |
+
"grad_norm": 1.0362215042114258,
|
| 309 |
+
"learning_rate": 4.467005076142132e-05,
|
| 310 |
+
"loss": 0.4053,
|
| 311 |
+
"step": 420
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 1.0913705583756346,
|
| 315 |
+
"grad_norm": 1.5984660387039185,
|
| 316 |
+
"learning_rate": 4.454314720812183e-05,
|
| 317 |
+
"loss": 0.6716,
|
| 318 |
+
"step": 430
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 1.116751269035533,
|
| 322 |
+
"grad_norm": 2.6108462810516357,
|
| 323 |
+
"learning_rate": 4.4416243654822335e-05,
|
| 324 |
+
"loss": 0.4829,
|
| 325 |
+
"step": 440
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 1.1421319796954315,
|
| 329 |
+
"grad_norm": 6.843586444854736,
|
| 330 |
+
"learning_rate": 4.428934010152285e-05,
|
| 331 |
+
"loss": 0.4716,
|
| 332 |
+
"step": 450
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.16751269035533,
|
| 336 |
+
"grad_norm": 5.383293151855469,
|
| 337 |
+
"learning_rate": 4.416243654822335e-05,
|
| 338 |
+
"loss": 0.6196,
|
| 339 |
+
"step": 460
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.1928934010152283,
|
| 343 |
+
"grad_norm": 1.4819399118423462,
|
| 344 |
+
"learning_rate": 4.403553299492386e-05,
|
| 345 |
+
"loss": 0.4594,
|
| 346 |
+
"step": 470
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 1.218274111675127,
|
| 350 |
+
"grad_norm": 3.468229055404663,
|
| 351 |
+
"learning_rate": 4.3908629441624365e-05,
|
| 352 |
+
"loss": 0.4838,
|
| 353 |
+
"step": 480
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.2436548223350254,
|
| 357 |
+
"grad_norm": 3.6312994956970215,
|
| 358 |
+
"learning_rate": 4.378172588832488e-05,
|
| 359 |
+
"loss": 0.4686,
|
| 360 |
+
"step": 490
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 1.2690355329949239,
|
| 364 |
+
"grad_norm": 1.4617408514022827,
|
| 365 |
+
"learning_rate": 4.365482233502538e-05,
|
| 366 |
+
"loss": 0.4094,
|
| 367 |
+
"step": 500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 1.2944162436548223,
|
| 371 |
+
"grad_norm": 0.9887295365333557,
|
| 372 |
+
"learning_rate": 4.352791878172589e-05,
|
| 373 |
+
"loss": 0.5219,
|
| 374 |
+
"step": 510
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.3197969543147208,
|
| 378 |
+
"grad_norm": 1.6651699542999268,
|
| 379 |
+
"learning_rate": 4.34010152284264e-05,
|
| 380 |
+
"loss": 0.6159,
|
| 381 |
+
"step": 520
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.3451776649746192,
|
| 385 |
+
"grad_norm": 2.0405914783477783,
|
| 386 |
+
"learning_rate": 4.32741116751269e-05,
|
| 387 |
+
"loss": 0.41,
|
| 388 |
+
"step": 530
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.3705583756345177,
|
| 392 |
+
"grad_norm": 4.893404483795166,
|
| 393 |
+
"learning_rate": 4.3147208121827415e-05,
|
| 394 |
+
"loss": 0.5679,
|
| 395 |
+
"step": 540
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.3959390862944163,
|
| 399 |
+
"grad_norm": 2.044116735458374,
|
| 400 |
+
"learning_rate": 4.302030456852792e-05,
|
| 401 |
+
"loss": 0.4693,
|
| 402 |
+
"step": 550
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.4213197969543148,
|
| 406 |
+
"grad_norm": 2.040766477584839,
|
| 407 |
+
"learning_rate": 4.289340101522843e-05,
|
| 408 |
+
"loss": 0.4736,
|
| 409 |
+
"step": 560
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.4467005076142132,
|
| 413 |
+
"grad_norm": 1.883665919303894,
|
| 414 |
+
"learning_rate": 4.276649746192893e-05,
|
| 415 |
+
"loss": 0.6475,
|
| 416 |
+
"step": 570
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.4720812182741116,
|
| 420 |
+
"grad_norm": 1.6452559232711792,
|
| 421 |
+
"learning_rate": 4.2639593908629446e-05,
|
| 422 |
+
"loss": 0.5092,
|
| 423 |
+
"step": 580
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.49746192893401,
|
| 427 |
+
"grad_norm": 2.059413194656372,
|
| 428 |
+
"learning_rate": 4.251269035532995e-05,
|
| 429 |
+
"loss": 0.5469,
|
| 430 |
+
"step": 590
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.5228426395939088,
|
| 434 |
+
"grad_norm": 5.789554595947266,
|
| 435 |
+
"learning_rate": 4.238578680203046e-05,
|
| 436 |
+
"loss": 0.4548,
|
| 437 |
+
"step": 600
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.548223350253807,
|
| 441 |
+
"grad_norm": 1.7683823108673096,
|
| 442 |
+
"learning_rate": 4.225888324873097e-05,
|
| 443 |
+
"loss": 0.3807,
|
| 444 |
+
"step": 610
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.5736040609137056,
|
| 448 |
+
"grad_norm": 1.6267879009246826,
|
| 449 |
+
"learning_rate": 4.213197969543147e-05,
|
| 450 |
+
"loss": 0.3814,
|
| 451 |
+
"step": 620
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.598984771573604,
|
| 455 |
+
"grad_norm": 2.024305582046509,
|
| 456 |
+
"learning_rate": 4.200507614213198e-05,
|
| 457 |
+
"loss": 0.5454,
|
| 458 |
+
"step": 630
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.6243654822335025,
|
| 462 |
+
"grad_norm": 2.527076482772827,
|
| 463 |
+
"learning_rate": 4.187817258883249e-05,
|
| 464 |
+
"loss": 0.5996,
|
| 465 |
+
"step": 640
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.649746192893401,
|
| 469 |
+
"grad_norm": 2.8863742351531982,
|
| 470 |
+
"learning_rate": 4.1751269035532995e-05,
|
| 471 |
+
"loss": 0.4548,
|
| 472 |
+
"step": 650
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.6751269035532994,
|
| 476 |
+
"grad_norm": 1.8288835287094116,
|
| 477 |
+
"learning_rate": 4.162436548223351e-05,
|
| 478 |
+
"loss": 0.3666,
|
| 479 |
+
"step": 660
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.700507614213198,
|
| 483 |
+
"grad_norm": 2.004929304122925,
|
| 484 |
+
"learning_rate": 4.1497461928934013e-05,
|
| 485 |
+
"loss": 0.4643,
|
| 486 |
+
"step": 670
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.7258883248730963,
|
| 490 |
+
"grad_norm": 2.2006640434265137,
|
| 491 |
+
"learning_rate": 4.137055837563452e-05,
|
| 492 |
+
"loss": 0.5934,
|
| 493 |
+
"step": 680
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.751269035532995,
|
| 497 |
+
"grad_norm": 1.2337054014205933,
|
| 498 |
+
"learning_rate": 4.1243654822335025e-05,
|
| 499 |
+
"loss": 0.5315,
|
| 500 |
+
"step": 690
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.7766497461928934,
|
| 504 |
+
"grad_norm": 1.163268804550171,
|
| 505 |
+
"learning_rate": 4.111675126903554e-05,
|
| 506 |
+
"loss": 0.4484,
|
| 507 |
+
"step": 700
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.8020304568527918,
|
| 511 |
+
"grad_norm": 1.2907366752624512,
|
| 512 |
+
"learning_rate": 4.098984771573604e-05,
|
| 513 |
+
"loss": 0.4027,
|
| 514 |
+
"step": 710
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.8274111675126905,
|
| 518 |
+
"grad_norm": 1.8477003574371338,
|
| 519 |
+
"learning_rate": 4.086294416243655e-05,
|
| 520 |
+
"loss": 0.503,
|
| 521 |
+
"step": 720
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.8527918781725887,
|
| 525 |
+
"grad_norm": 2.8408010005950928,
|
| 526 |
+
"learning_rate": 4.073604060913706e-05,
|
| 527 |
+
"loss": 0.5602,
|
| 528 |
+
"step": 730
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.8781725888324874,
|
| 532 |
+
"grad_norm": 3.5004968643188477,
|
| 533 |
+
"learning_rate": 4.060913705583756e-05,
|
| 534 |
+
"loss": 0.5681,
|
| 535 |
+
"step": 740
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.9035532994923858,
|
| 539 |
+
"grad_norm": 1.7411813735961914,
|
| 540 |
+
"learning_rate": 4.0482233502538075e-05,
|
| 541 |
+
"loss": 0.4501,
|
| 542 |
+
"step": 750
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.9289340101522843,
|
| 546 |
+
"grad_norm": 4.301488876342773,
|
| 547 |
+
"learning_rate": 4.035532994923858e-05,
|
| 548 |
+
"loss": 0.5262,
|
| 549 |
+
"step": 760
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.9543147208121827,
|
| 553 |
+
"grad_norm": 0.8759877681732178,
|
| 554 |
+
"learning_rate": 4.022842639593909e-05,
|
| 555 |
+
"loss": 0.4527,
|
| 556 |
+
"step": 770
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.9796954314720812,
|
| 560 |
+
"grad_norm": 2.361971855163574,
|
| 561 |
+
"learning_rate": 4.010152284263959e-05,
|
| 562 |
+
"loss": 0.487,
|
| 563 |
+
"step": 780
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 2.0,
|
| 567 |
+
"eval_loss": 0.4712839722633362,
|
| 568 |
+
"eval_runtime": 8.7867,
|
| 569 |
+
"eval_samples_per_second": 89.681,
|
| 570 |
+
"eval_steps_per_second": 11.267,
|
| 571 |
+
"step": 788
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 2.00507614213198,
|
| 575 |
+
"grad_norm": 2.517071485519409,
|
| 576 |
+
"learning_rate": 3.9974619289340106e-05,
|
| 577 |
+
"loss": 0.4159,
|
| 578 |
+
"step": 790
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 2.030456852791878,
|
| 582 |
+
"grad_norm": 1.2710137367248535,
|
| 583 |
+
"learning_rate": 3.9847715736040605e-05,
|
| 584 |
+
"loss": 0.4777,
|
| 585 |
+
"step": 800
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 2.0558375634517767,
|
| 589 |
+
"grad_norm": 3.463624954223633,
|
| 590 |
+
"learning_rate": 3.972081218274112e-05,
|
| 591 |
+
"loss": 0.4792,
|
| 592 |
+
"step": 810
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 2.081218274111675,
|
| 596 |
+
"grad_norm": 1.5746160745620728,
|
| 597 |
+
"learning_rate": 3.959390862944163e-05,
|
| 598 |
+
"loss": 0.4457,
|
| 599 |
+
"step": 820
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 2.1065989847715736,
|
| 603 |
+
"grad_norm": 1.3191742897033691,
|
| 604 |
+
"learning_rate": 3.946700507614213e-05,
|
| 605 |
+
"loss": 0.4615,
|
| 606 |
+
"step": 830
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.1319796954314723,
|
| 610 |
+
"grad_norm": 4.860599040985107,
|
| 611 |
+
"learning_rate": 3.934010152284264e-05,
|
| 612 |
+
"loss": 0.4459,
|
| 613 |
+
"step": 840
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 2.1573604060913705,
|
| 617 |
+
"grad_norm": 2.636868715286255,
|
| 618 |
+
"learning_rate": 3.921319796954315e-05,
|
| 619 |
+
"loss": 0.537,
|
| 620 |
+
"step": 850
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.182741116751269,
|
| 624 |
+
"grad_norm": 1.672642469406128,
|
| 625 |
+
"learning_rate": 3.9086294416243655e-05,
|
| 626 |
+
"loss": 0.4848,
|
| 627 |
+
"step": 860
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 2.2081218274111674,
|
| 631 |
+
"grad_norm": 8.49184799194336,
|
| 632 |
+
"learning_rate": 3.895939086294416e-05,
|
| 633 |
+
"loss": 0.4826,
|
| 634 |
+
"step": 870
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 2.233502538071066,
|
| 638 |
+
"grad_norm": 1.6219210624694824,
|
| 639 |
+
"learning_rate": 3.8832487309644673e-05,
|
| 640 |
+
"loss": 0.3403,
|
| 641 |
+
"step": 880
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 2.2588832487309647,
|
| 645 |
+
"grad_norm": 1.356152892112732,
|
| 646 |
+
"learning_rate": 3.870558375634518e-05,
|
| 647 |
+
"loss": 0.5415,
|
| 648 |
+
"step": 890
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 2.284263959390863,
|
| 652 |
+
"grad_norm": 3.455899238586426,
|
| 653 |
+
"learning_rate": 3.8578680203045685e-05,
|
| 654 |
+
"loss": 0.4958,
|
| 655 |
+
"step": 900
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 2.3096446700507616,
|
| 659 |
+
"grad_norm": 3.994067430496216,
|
| 660 |
+
"learning_rate": 3.84517766497462e-05,
|
| 661 |
+
"loss": 0.4324,
|
| 662 |
+
"step": 910
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 2.33502538071066,
|
| 666 |
+
"grad_norm": 5.774569034576416,
|
| 667 |
+
"learning_rate": 3.83248730964467e-05,
|
| 668 |
+
"loss": 0.3871,
|
| 669 |
+
"step": 920
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 2.3604060913705585,
|
| 673 |
+
"grad_norm": 4.524348258972168,
|
| 674 |
+
"learning_rate": 3.819796954314721e-05,
|
| 675 |
+
"loss": 0.5642,
|
| 676 |
+
"step": 930
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 2.3857868020304567,
|
| 680 |
+
"grad_norm": 1.5101829767227173,
|
| 681 |
+
"learning_rate": 3.8071065989847716e-05,
|
| 682 |
+
"loss": 0.5396,
|
| 683 |
+
"step": 940
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 2.4111675126903553,
|
| 687 |
+
"grad_norm": 1.4824628829956055,
|
| 688 |
+
"learning_rate": 3.794416243654822e-05,
|
| 689 |
+
"loss": 0.5178,
|
| 690 |
+
"step": 950
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 2.436548223350254,
|
| 694 |
+
"grad_norm": 1.3847397565841675,
|
| 695 |
+
"learning_rate": 3.7817258883248735e-05,
|
| 696 |
+
"loss": 0.3747,
|
| 697 |
+
"step": 960
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 2.4619289340101522,
|
| 701 |
+
"grad_norm": 2.138828992843628,
|
| 702 |
+
"learning_rate": 3.769035532994924e-05,
|
| 703 |
+
"loss": 0.5756,
|
| 704 |
+
"step": 970
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 2.487309644670051,
|
| 708 |
+
"grad_norm": 2.0886220932006836,
|
| 709 |
+
"learning_rate": 3.756345177664975e-05,
|
| 710 |
+
"loss": 0.5496,
|
| 711 |
+
"step": 980
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 2.512690355329949,
|
| 715 |
+
"grad_norm": 2.8054535388946533,
|
| 716 |
+
"learning_rate": 3.743654822335025e-05,
|
| 717 |
+
"loss": 0.5984,
|
| 718 |
+
"step": 990
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 2.5380710659898478,
|
| 722 |
+
"grad_norm": 3.6035268306732178,
|
| 723 |
+
"learning_rate": 3.7309644670050766e-05,
|
| 724 |
+
"loss": 0.4475,
|
| 725 |
+
"step": 1000
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 2.563451776649746,
|
| 729 |
+
"grad_norm": 2.931821823120117,
|
| 730 |
+
"learning_rate": 3.7182741116751265e-05,
|
| 731 |
+
"loss": 0.5179,
|
| 732 |
+
"step": 1010
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 2.5888324873096447,
|
| 736 |
+
"grad_norm": 1.8696964979171753,
|
| 737 |
+
"learning_rate": 3.705583756345178e-05,
|
| 738 |
+
"loss": 0.6541,
|
| 739 |
+
"step": 1020
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 2.6142131979695433,
|
| 743 |
+
"grad_norm": 1.36885666847229,
|
| 744 |
+
"learning_rate": 3.692893401015229e-05,
|
| 745 |
+
"loss": 0.3136,
|
| 746 |
+
"step": 1030
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 2.6395939086294415,
|
| 750 |
+
"grad_norm": 3.520601511001587,
|
| 751 |
+
"learning_rate": 3.680203045685279e-05,
|
| 752 |
+
"loss": 0.4434,
|
| 753 |
+
"step": 1040
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 2.66497461928934,
|
| 757 |
+
"grad_norm": 3.95267915725708,
|
| 758 |
+
"learning_rate": 3.66751269035533e-05,
|
| 759 |
+
"loss": 0.5012,
|
| 760 |
+
"step": 1050
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 2.6903553299492384,
|
| 764 |
+
"grad_norm": 2.064523935317993,
|
| 765 |
+
"learning_rate": 3.654822335025381e-05,
|
| 766 |
+
"loss": 0.3251,
|
| 767 |
+
"step": 1060
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 2.715736040609137,
|
| 771 |
+
"grad_norm": 3.3240392208099365,
|
| 772 |
+
"learning_rate": 3.6421319796954315e-05,
|
| 773 |
+
"loss": 0.7704,
|
| 774 |
+
"step": 1070
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 2.7411167512690353,
|
| 778 |
+
"grad_norm": 3.2816076278686523,
|
| 779 |
+
"learning_rate": 3.629441624365482e-05,
|
| 780 |
+
"loss": 0.4753,
|
| 781 |
+
"step": 1080
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 2.766497461928934,
|
| 785 |
+
"grad_norm": 2.7639636993408203,
|
| 786 |
+
"learning_rate": 3.6167512690355334e-05,
|
| 787 |
+
"loss": 0.4443,
|
| 788 |
+
"step": 1090
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 2.7918781725888326,
|
| 792 |
+
"grad_norm": 2.5042450428009033,
|
| 793 |
+
"learning_rate": 3.604060913705584e-05,
|
| 794 |
+
"loss": 0.4979,
|
| 795 |
+
"step": 1100
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 2.817258883248731,
|
| 799 |
+
"grad_norm": 2.695377826690674,
|
| 800 |
+
"learning_rate": 3.5913705583756346e-05,
|
| 801 |
+
"loss": 0.4835,
|
| 802 |
+
"step": 1110
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.8426395939086295,
|
| 806 |
+
"grad_norm": 2.2236552238464355,
|
| 807 |
+
"learning_rate": 3.578680203045686e-05,
|
| 808 |
+
"loss": 0.5162,
|
| 809 |
+
"step": 1120
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 2.868020304568528,
|
| 813 |
+
"grad_norm": 5.009507179260254,
|
| 814 |
+
"learning_rate": 3.565989847715736e-05,
|
| 815 |
+
"loss": 0.4046,
|
| 816 |
+
"step": 1130
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 2.8934010152284264,
|
| 820 |
+
"grad_norm": 2.037398338317871,
|
| 821 |
+
"learning_rate": 3.553299492385787e-05,
|
| 822 |
+
"loss": 0.4502,
|
| 823 |
+
"step": 1140
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 2.9187817258883246,
|
| 827 |
+
"grad_norm": 5.31472635269165,
|
| 828 |
+
"learning_rate": 3.5406091370558376e-05,
|
| 829 |
+
"loss": 0.6371,
|
| 830 |
+
"step": 1150
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 2.9441624365482233,
|
| 834 |
+
"grad_norm": 1.666244387626648,
|
| 835 |
+
"learning_rate": 3.527918781725888e-05,
|
| 836 |
+
"loss": 0.4088,
|
| 837 |
+
"step": 1160
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 2.969543147208122,
|
| 841 |
+
"grad_norm": 4.006988525390625,
|
| 842 |
+
"learning_rate": 3.5152284263959395e-05,
|
| 843 |
+
"loss": 0.4483,
|
| 844 |
+
"step": 1170
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 2.99492385786802,
|
| 848 |
+
"grad_norm": 2.429619312286377,
|
| 849 |
+
"learning_rate": 3.50253807106599e-05,
|
| 850 |
+
"loss": 0.4617,
|
| 851 |
+
"step": 1180
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 3.0,
|
| 855 |
+
"eval_loss": 0.5104432106018066,
|
| 856 |
+
"eval_runtime": 8.8793,
|
| 857 |
+
"eval_samples_per_second": 88.746,
|
| 858 |
+
"eval_steps_per_second": 11.15,
|
| 859 |
+
"step": 1182
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 3.020304568527919,
|
| 863 |
+
"grad_norm": 1.9103385210037231,
|
| 864 |
+
"learning_rate": 3.489847715736041e-05,
|
| 865 |
+
"loss": 0.461,
|
| 866 |
+
"step": 1190
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 3.045685279187817,
|
| 870 |
+
"grad_norm": 2.838453769683838,
|
| 871 |
+
"learning_rate": 3.477157360406091e-05,
|
| 872 |
+
"loss": 0.4661,
|
| 873 |
+
"step": 1200
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 3.0710659898477157,
|
| 877 |
+
"grad_norm": 2.706493616104126,
|
| 878 |
+
"learning_rate": 3.4644670050761426e-05,
|
| 879 |
+
"loss": 0.4335,
|
| 880 |
+
"step": 1210
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 3.0964467005076144,
|
| 884 |
+
"grad_norm": 4.179074287414551,
|
| 885 |
+
"learning_rate": 3.451776649746193e-05,
|
| 886 |
+
"loss": 0.3517,
|
| 887 |
+
"step": 1220
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 3.1218274111675126,
|
| 891 |
+
"grad_norm": 1.1146340370178223,
|
| 892 |
+
"learning_rate": 3.439086294416244e-05,
|
| 893 |
+
"loss": 0.2261,
|
| 894 |
+
"step": 1230
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 3.1472081218274113,
|
| 898 |
+
"grad_norm": 2.5486109256744385,
|
| 899 |
+
"learning_rate": 3.4263959390862944e-05,
|
| 900 |
+
"loss": 0.3664,
|
| 901 |
+
"step": 1240
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 3.1725888324873095,
|
| 905 |
+
"grad_norm": 2.886986017227173,
|
| 906 |
+
"learning_rate": 3.413705583756345e-05,
|
| 907 |
+
"loss": 0.3861,
|
| 908 |
+
"step": 1250
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 3.197969543147208,
|
| 912 |
+
"grad_norm": 3.5457489490509033,
|
| 913 |
+
"learning_rate": 3.401015228426396e-05,
|
| 914 |
+
"loss": 0.5043,
|
| 915 |
+
"step": 1260
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 3.223350253807107,
|
| 919 |
+
"grad_norm": 4.039999485015869,
|
| 920 |
+
"learning_rate": 3.388324873096447e-05,
|
| 921 |
+
"loss": 0.3362,
|
| 922 |
+
"step": 1270
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 3.248730964467005,
|
| 926 |
+
"grad_norm": 4.554147720336914,
|
| 927 |
+
"learning_rate": 3.3756345177664975e-05,
|
| 928 |
+
"loss": 0.4709,
|
| 929 |
+
"step": 1280
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 3.2741116751269037,
|
| 933 |
+
"grad_norm": 2.4214282035827637,
|
| 934 |
+
"learning_rate": 3.362944162436548e-05,
|
| 935 |
+
"loss": 0.3041,
|
| 936 |
+
"step": 1290
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 3.299492385786802,
|
| 940 |
+
"grad_norm": 4.587721824645996,
|
| 941 |
+
"learning_rate": 3.3502538071065994e-05,
|
| 942 |
+
"loss": 0.6426,
|
| 943 |
+
"step": 1300
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 3.3248730964467006,
|
| 947 |
+
"grad_norm": 2.0584123134613037,
|
| 948 |
+
"learning_rate": 3.33756345177665e-05,
|
| 949 |
+
"loss": 0.3861,
|
| 950 |
+
"step": 1310
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 3.350253807106599,
|
| 954 |
+
"grad_norm": 2.152904510498047,
|
| 955 |
+
"learning_rate": 3.3248730964467006e-05,
|
| 956 |
+
"loss": 0.434,
|
| 957 |
+
"step": 1320
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 3.3756345177664975,
|
| 961 |
+
"grad_norm": 3.3086836338043213,
|
| 962 |
+
"learning_rate": 3.312182741116752e-05,
|
| 963 |
+
"loss": 0.406,
|
| 964 |
+
"step": 1330
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 3.401015228426396,
|
| 968 |
+
"grad_norm": 2.4695167541503906,
|
| 969 |
+
"learning_rate": 3.299492385786802e-05,
|
| 970 |
+
"loss": 0.4226,
|
| 971 |
+
"step": 1340
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 3.4263959390862944,
|
| 975 |
+
"grad_norm": 2.482093334197998,
|
| 976 |
+
"learning_rate": 3.286802030456853e-05,
|
| 977 |
+
"loss": 0.335,
|
| 978 |
+
"step": 1350
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 3.451776649746193,
|
| 982 |
+
"grad_norm": 6.9000959396362305,
|
| 983 |
+
"learning_rate": 3.2741116751269036e-05,
|
| 984 |
+
"loss": 0.3506,
|
| 985 |
+
"step": 1360
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 3.4771573604060912,
|
| 989 |
+
"grad_norm": 2.6196579933166504,
|
| 990 |
+
"learning_rate": 3.261421319796954e-05,
|
| 991 |
+
"loss": 0.3377,
|
| 992 |
+
"step": 1370
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 3.50253807106599,
|
| 996 |
+
"grad_norm": 1.8537606000900269,
|
| 997 |
+
"learning_rate": 3.248730964467005e-05,
|
| 998 |
+
"loss": 0.399,
|
| 999 |
+
"step": 1380
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 3.527918781725888,
|
| 1003 |
+
"grad_norm": 7.022123336791992,
|
| 1004 |
+
"learning_rate": 3.236040609137056e-05,
|
| 1005 |
+
"loss": 0.4255,
|
| 1006 |
+
"step": 1390
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 3.553299492385787,
|
| 1010 |
+
"grad_norm": 3.112264394760132,
|
| 1011 |
+
"learning_rate": 3.223350253807107e-05,
|
| 1012 |
+
"loss": 0.399,
|
| 1013 |
+
"step": 1400
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 3.5786802030456855,
|
| 1017 |
+
"grad_norm": 4.884408473968506,
|
| 1018 |
+
"learning_rate": 3.210659898477157e-05,
|
| 1019 |
+
"loss": 0.3533,
|
| 1020 |
+
"step": 1410
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 3.6040609137055837,
|
| 1024 |
+
"grad_norm": 3.7339515686035156,
|
| 1025 |
+
"learning_rate": 3.1979695431472086e-05,
|
| 1026 |
+
"loss": 0.4329,
|
| 1027 |
+
"step": 1420
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 3.6294416243654823,
|
| 1031 |
+
"grad_norm": 11.797500610351562,
|
| 1032 |
+
"learning_rate": 3.185279187817259e-05,
|
| 1033 |
+
"loss": 0.4496,
|
| 1034 |
+
"step": 1430
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 3.6548223350253806,
|
| 1038 |
+
"grad_norm": 2.509971857070923,
|
| 1039 |
+
"learning_rate": 3.17258883248731e-05,
|
| 1040 |
+
"loss": 0.3758,
|
| 1041 |
+
"step": 1440
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 3.6802030456852792,
|
| 1045 |
+
"grad_norm": 2.554452419281006,
|
| 1046 |
+
"learning_rate": 3.1598984771573604e-05,
|
| 1047 |
+
"loss": 0.2789,
|
| 1048 |
+
"step": 1450
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 3.7055837563451774,
|
| 1052 |
+
"grad_norm": 13.046540260314941,
|
| 1053 |
+
"learning_rate": 3.147208121827411e-05,
|
| 1054 |
+
"loss": 0.52,
|
| 1055 |
+
"step": 1460
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 3.730964467005076,
|
| 1059 |
+
"grad_norm": 6.821984767913818,
|
| 1060 |
+
"learning_rate": 3.134517766497462e-05,
|
| 1061 |
+
"loss": 0.4285,
|
| 1062 |
+
"step": 1470
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 3.7563451776649748,
|
| 1066 |
+
"grad_norm": 2.692779064178467,
|
| 1067 |
+
"learning_rate": 3.121827411167513e-05,
|
| 1068 |
+
"loss": 0.2977,
|
| 1069 |
+
"step": 1480
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 3.781725888324873,
|
| 1073 |
+
"grad_norm": 3.510523796081543,
|
| 1074 |
+
"learning_rate": 3.1091370558375635e-05,
|
| 1075 |
+
"loss": 0.4887,
|
| 1076 |
+
"step": 1490
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 3.8071065989847717,
|
| 1080 |
+
"grad_norm": 4.374626159667969,
|
| 1081 |
+
"learning_rate": 3.096446700507614e-05,
|
| 1082 |
+
"loss": 0.2093,
|
| 1083 |
+
"step": 1500
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 3.8324873096446703,
|
| 1087 |
+
"grad_norm": 0.6560042500495911,
|
| 1088 |
+
"learning_rate": 3.0837563451776654e-05,
|
| 1089 |
+
"loss": 0.613,
|
| 1090 |
+
"step": 1510
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 3.8578680203045685,
|
| 1094 |
+
"grad_norm": 4.9201483726501465,
|
| 1095 |
+
"learning_rate": 3.071065989847716e-05,
|
| 1096 |
+
"loss": 0.4483,
|
| 1097 |
+
"step": 1520
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 3.8832487309644668,
|
| 1101 |
+
"grad_norm": 2.497208833694458,
|
| 1102 |
+
"learning_rate": 3.0583756345177666e-05,
|
| 1103 |
+
"loss": 0.4434,
|
| 1104 |
+
"step": 1530
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 3.9086294416243654,
|
| 1108 |
+
"grad_norm": 5.912675380706787,
|
| 1109 |
+
"learning_rate": 3.0456852791878175e-05,
|
| 1110 |
+
"loss": 0.4259,
|
| 1111 |
+
"step": 1540
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 3.934010152284264,
|
| 1115 |
+
"grad_norm": 1.8950650691986084,
|
| 1116 |
+
"learning_rate": 3.032994923857868e-05,
|
| 1117 |
+
"loss": 0.4649,
|
| 1118 |
+
"step": 1550
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 3.9593908629441623,
|
| 1122 |
+
"grad_norm": 15.265602111816406,
|
| 1123 |
+
"learning_rate": 3.020304568527919e-05,
|
| 1124 |
+
"loss": 0.4176,
|
| 1125 |
+
"step": 1560
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 3.984771573604061,
|
| 1129 |
+
"grad_norm": 6.062721252441406,
|
| 1130 |
+
"learning_rate": 3.0076142131979696e-05,
|
| 1131 |
+
"loss": 0.3854,
|
| 1132 |
+
"step": 1570
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 4.0,
|
| 1136 |
+
"eval_loss": 0.5223032832145691,
|
| 1137 |
+
"eval_runtime": 8.7536,
|
| 1138 |
+
"eval_samples_per_second": 90.02,
|
| 1139 |
+
"eval_steps_per_second": 11.31,
|
| 1140 |
+
"step": 1576
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 4.01015228426396,
|
| 1144 |
+
"grad_norm": 4.975412368774414,
|
| 1145 |
+
"learning_rate": 2.9949238578680206e-05,
|
| 1146 |
+
"loss": 0.4798,
|
| 1147 |
+
"step": 1580
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 4.035532994923858,
|
| 1151 |
+
"grad_norm": 5.022281646728516,
|
| 1152 |
+
"learning_rate": 2.982233502538071e-05,
|
| 1153 |
+
"loss": 0.3322,
|
| 1154 |
+
"step": 1590
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 4.060913705583756,
|
| 1158 |
+
"grad_norm": 5.229846000671387,
|
| 1159 |
+
"learning_rate": 2.969543147208122e-05,
|
| 1160 |
+
"loss": 0.2691,
|
| 1161 |
+
"step": 1600
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 4.086294416243655,
|
| 1165 |
+
"grad_norm": 21.07980728149414,
|
| 1166 |
+
"learning_rate": 2.956852791878173e-05,
|
| 1167 |
+
"loss": 0.285,
|
| 1168 |
+
"step": 1610
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 4.111675126903553,
|
| 1172 |
+
"grad_norm": 28.23366355895996,
|
| 1173 |
+
"learning_rate": 2.9441624365482233e-05,
|
| 1174 |
+
"loss": 0.1946,
|
| 1175 |
+
"step": 1620
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 4.137055837563452,
|
| 1179 |
+
"grad_norm": 0.49203088879585266,
|
| 1180 |
+
"learning_rate": 2.9314720812182743e-05,
|
| 1181 |
+
"loss": 0.2176,
|
| 1182 |
+
"step": 1630
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 4.16243654822335,
|
| 1186 |
+
"grad_norm": 6.196091175079346,
|
| 1187 |
+
"learning_rate": 2.918781725888325e-05,
|
| 1188 |
+
"loss": 0.1592,
|
| 1189 |
+
"step": 1640
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 4.187817258883249,
|
| 1193 |
+
"grad_norm": 0.43765848875045776,
|
| 1194 |
+
"learning_rate": 2.9060913705583758e-05,
|
| 1195 |
+
"loss": 0.3764,
|
| 1196 |
+
"step": 1650
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 4.213197969543147,
|
| 1200 |
+
"grad_norm": 4.543421745300293,
|
| 1201 |
+
"learning_rate": 2.8934010152284264e-05,
|
| 1202 |
+
"loss": 0.3371,
|
| 1203 |
+
"step": 1660
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 4.238578680203045,
|
| 1207 |
+
"grad_norm": 2.611579418182373,
|
| 1208 |
+
"learning_rate": 2.8807106598984774e-05,
|
| 1209 |
+
"loss": 0.3052,
|
| 1210 |
+
"step": 1670
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 4.2639593908629445,
|
| 1214 |
+
"grad_norm": 3.3673095703125,
|
| 1215 |
+
"learning_rate": 2.8680203045685283e-05,
|
| 1216 |
+
"loss": 0.1776,
|
| 1217 |
+
"step": 1680
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 4.289340101522843,
|
| 1221 |
+
"grad_norm": 0.7387822270393372,
|
| 1222 |
+
"learning_rate": 2.855329949238579e-05,
|
| 1223 |
+
"loss": 0.3827,
|
| 1224 |
+
"step": 1690
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 4.314720812182741,
|
| 1228 |
+
"grad_norm": 16.637300491333008,
|
| 1229 |
+
"learning_rate": 2.84263959390863e-05,
|
| 1230 |
+
"loss": 0.3993,
|
| 1231 |
+
"step": 1700
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 4.340101522842639,
|
| 1235 |
+
"grad_norm": 6.28633975982666,
|
| 1236 |
+
"learning_rate": 2.82994923857868e-05,
|
| 1237 |
+
"loss": 0.3733,
|
| 1238 |
+
"step": 1710
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 4.365482233502538,
|
| 1242 |
+
"grad_norm": 3.3295745849609375,
|
| 1243 |
+
"learning_rate": 2.8172588832487314e-05,
|
| 1244 |
+
"loss": 0.1166,
|
| 1245 |
+
"step": 1720
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 4.3908629441624365,
|
| 1249 |
+
"grad_norm": 14.652837753295898,
|
| 1250 |
+
"learning_rate": 2.8045685279187816e-05,
|
| 1251 |
+
"loss": 0.3326,
|
| 1252 |
+
"step": 1730
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 4.416243654822335,
|
| 1256 |
+
"grad_norm": 31.232574462890625,
|
| 1257 |
+
"learning_rate": 2.7918781725888326e-05,
|
| 1258 |
+
"loss": 0.2127,
|
| 1259 |
+
"step": 1740
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 4.441624365482234,
|
| 1263 |
+
"grad_norm": 12.528191566467285,
|
| 1264 |
+
"learning_rate": 2.7791878172588832e-05,
|
| 1265 |
+
"loss": 0.4061,
|
| 1266 |
+
"step": 1750
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 4.467005076142132,
|
| 1270 |
+
"grad_norm": 1.7689141035079956,
|
| 1271 |
+
"learning_rate": 2.766497461928934e-05,
|
| 1272 |
+
"loss": 0.3366,
|
| 1273 |
+
"step": 1760
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 4.49238578680203,
|
| 1277 |
+
"grad_norm": 4.724018096923828,
|
| 1278 |
+
"learning_rate": 2.753807106598985e-05,
|
| 1279 |
+
"loss": 0.2359,
|
| 1280 |
+
"step": 1770
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 4.517766497461929,
|
| 1284 |
+
"grad_norm": 1.561919093132019,
|
| 1285 |
+
"learning_rate": 2.7411167512690357e-05,
|
| 1286 |
+
"loss": 0.3106,
|
| 1287 |
+
"step": 1780
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 4.543147208121828,
|
| 1291 |
+
"grad_norm": 5.754579067230225,
|
| 1292 |
+
"learning_rate": 2.7284263959390866e-05,
|
| 1293 |
+
"loss": 0.2982,
|
| 1294 |
+
"step": 1790
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 4.568527918781726,
|
| 1298 |
+
"grad_norm": 1.181515097618103,
|
| 1299 |
+
"learning_rate": 2.715736040609137e-05,
|
| 1300 |
+
"loss": 0.2384,
|
| 1301 |
+
"step": 1800
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 4.593908629441624,
|
| 1305 |
+
"grad_norm": 7.012800693511963,
|
| 1306 |
+
"learning_rate": 2.703045685279188e-05,
|
| 1307 |
+
"loss": 0.394,
|
| 1308 |
+
"step": 1810
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 4.619289340101523,
|
| 1312 |
+
"grad_norm": 0.49557140469551086,
|
| 1313 |
+
"learning_rate": 2.6903553299492384e-05,
|
| 1314 |
+
"loss": 0.2869,
|
| 1315 |
+
"step": 1820
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 4.644670050761421,
|
| 1319 |
+
"grad_norm": 3.6110830307006836,
|
| 1320 |
+
"learning_rate": 2.6776649746192893e-05,
|
| 1321 |
+
"loss": 0.3724,
|
| 1322 |
+
"step": 1830
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 4.67005076142132,
|
| 1326 |
+
"grad_norm": 3.3351943492889404,
|
| 1327 |
+
"learning_rate": 2.6649746192893406e-05,
|
| 1328 |
+
"loss": 0.2966,
|
| 1329 |
+
"step": 1840
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 4.695431472081218,
|
| 1333 |
+
"grad_norm": 1.342678427696228,
|
| 1334 |
+
"learning_rate": 2.652284263959391e-05,
|
| 1335 |
+
"loss": 0.3826,
|
| 1336 |
+
"step": 1850
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 4.720812182741117,
|
| 1340 |
+
"grad_norm": 0.4344848096370697,
|
| 1341 |
+
"learning_rate": 2.6395939086294418e-05,
|
| 1342 |
+
"loss": 0.1962,
|
| 1343 |
+
"step": 1860
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 4.746192893401015,
|
| 1347 |
+
"grad_norm": 56.233985900878906,
|
| 1348 |
+
"learning_rate": 2.6269035532994924e-05,
|
| 1349 |
+
"loss": 0.3758,
|
| 1350 |
+
"step": 1870
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 4.771573604060913,
|
| 1354 |
+
"grad_norm": 10.647058486938477,
|
| 1355 |
+
"learning_rate": 2.6142131979695434e-05,
|
| 1356 |
+
"loss": 0.356,
|
| 1357 |
+
"step": 1880
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 4.7969543147208125,
|
| 1361 |
+
"grad_norm": 0.9422564506530762,
|
| 1362 |
+
"learning_rate": 2.6015228426395936e-05,
|
| 1363 |
+
"loss": 0.2271,
|
| 1364 |
+
"step": 1890
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 4.822335025380711,
|
| 1368 |
+
"grad_norm": 0.22354674339294434,
|
| 1369 |
+
"learning_rate": 2.588832487309645e-05,
|
| 1370 |
+
"loss": 0.1374,
|
| 1371 |
+
"step": 1900
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 4.847715736040609,
|
| 1375 |
+
"grad_norm": 8.779372215270996,
|
| 1376 |
+
"learning_rate": 2.576142131979696e-05,
|
| 1377 |
+
"loss": 0.329,
|
| 1378 |
+
"step": 1910
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 4.873096446700508,
|
| 1382 |
+
"grad_norm": 41.76162338256836,
|
| 1383 |
+
"learning_rate": 2.563451776649746e-05,
|
| 1384 |
+
"loss": 0.5914,
|
| 1385 |
+
"step": 1920
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 4.898477157360406,
|
| 1389 |
+
"grad_norm": 26.878578186035156,
|
| 1390 |
+
"learning_rate": 2.5507614213197974e-05,
|
| 1391 |
+
"loss": 0.5029,
|
| 1392 |
+
"step": 1930
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 4.9238578680203045,
|
| 1396 |
+
"grad_norm": 0.41534122824668884,
|
| 1397 |
+
"learning_rate": 2.5380710659898476e-05,
|
| 1398 |
+
"loss": 0.1997,
|
| 1399 |
+
"step": 1940
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 4.949238578680203,
|
| 1403 |
+
"grad_norm": 4.708491325378418,
|
| 1404 |
+
"learning_rate": 2.5253807106598986e-05,
|
| 1405 |
+
"loss": 0.3183,
|
| 1406 |
+
"step": 1950
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 4.974619289340102,
|
| 1410 |
+
"grad_norm": 0.31694722175598145,
|
| 1411 |
+
"learning_rate": 2.5126903553299492e-05,
|
| 1412 |
+
"loss": 0.3627,
|
| 1413 |
+
"step": 1960
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 5.0,
|
| 1417 |
+
"grad_norm": 2.7903828620910645,
|
| 1418 |
+
"learning_rate": 2.5e-05,
|
| 1419 |
+
"loss": 0.454,
|
| 1420 |
+
"step": 1970
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 5.0,
|
| 1424 |
+
"eval_loss": 0.6968420743942261,
|
| 1425 |
+
"eval_runtime": 8.7659,
|
| 1426 |
+
"eval_samples_per_second": 89.894,
|
| 1427 |
+
"eval_steps_per_second": 11.294,
|
| 1428 |
+
"step": 1970
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 5.025380710659898,
|
| 1432 |
+
"grad_norm": 0.33573099970817566,
|
| 1433 |
+
"learning_rate": 2.4873096446700507e-05,
|
| 1434 |
+
"loss": 0.0922,
|
| 1435 |
+
"step": 1980
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 5.050761421319797,
|
| 1439 |
+
"grad_norm": 0.38098305463790894,
|
| 1440 |
+
"learning_rate": 2.4746192893401017e-05,
|
| 1441 |
+
"loss": 0.1296,
|
| 1442 |
+
"step": 1990
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 5.0761421319796955,
|
| 1446 |
+
"grad_norm": 6.538486480712891,
|
| 1447 |
+
"learning_rate": 2.4619289340101523e-05,
|
| 1448 |
+
"loss": 0.2388,
|
| 1449 |
+
"step": 2000
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 5.101522842639594,
|
| 1453 |
+
"grad_norm": 0.3389257788658142,
|
| 1454 |
+
"learning_rate": 2.4492385786802032e-05,
|
| 1455 |
+
"loss": 0.1525,
|
| 1456 |
+
"step": 2010
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 5.126903553299492,
|
| 1460 |
+
"grad_norm": 29.134950637817383,
|
| 1461 |
+
"learning_rate": 2.436548223350254e-05,
|
| 1462 |
+
"loss": 0.2023,
|
| 1463 |
+
"step": 2020
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 5.152284263959391,
|
| 1467 |
+
"grad_norm": 2.5168628692626953,
|
| 1468 |
+
"learning_rate": 2.4238578680203047e-05,
|
| 1469 |
+
"loss": 0.1126,
|
| 1470 |
+
"step": 2030
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 5.177664974619289,
|
| 1474 |
+
"grad_norm": 0.4162174463272095,
|
| 1475 |
+
"learning_rate": 2.4111675126903553e-05,
|
| 1476 |
+
"loss": 0.1466,
|
| 1477 |
+
"step": 2040
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 5.2030456852791875,
|
| 1481 |
+
"grad_norm": 0.08261796087026596,
|
| 1482 |
+
"learning_rate": 2.3984771573604063e-05,
|
| 1483 |
+
"loss": 0.1032,
|
| 1484 |
+
"step": 2050
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 5.228426395939087,
|
| 1488 |
+
"grad_norm": 0.13249030709266663,
|
| 1489 |
+
"learning_rate": 2.385786802030457e-05,
|
| 1490 |
+
"loss": 0.0117,
|
| 1491 |
+
"step": 2060
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 5.253807106598985,
|
| 1495 |
+
"grad_norm": 54.72965621948242,
|
| 1496 |
+
"learning_rate": 2.3730964467005075e-05,
|
| 1497 |
+
"loss": 0.1383,
|
| 1498 |
+
"step": 2070
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 5.279187817258883,
|
| 1502 |
+
"grad_norm": 10.03510856628418,
|
| 1503 |
+
"learning_rate": 2.3604060913705588e-05,
|
| 1504 |
+
"loss": 0.3034,
|
| 1505 |
+
"step": 2080
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 5.304568527918782,
|
| 1509 |
+
"grad_norm": 3.790813446044922,
|
| 1510 |
+
"learning_rate": 2.3477157360406094e-05,
|
| 1511 |
+
"loss": 0.0487,
|
| 1512 |
+
"step": 2090
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 5.32994923857868,
|
| 1516 |
+
"grad_norm": 10.79232120513916,
|
| 1517 |
+
"learning_rate": 2.33502538071066e-05,
|
| 1518 |
+
"loss": 0.1897,
|
| 1519 |
+
"step": 2100
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 5.355329949238579,
|
| 1523 |
+
"grad_norm": 3.8837833404541016,
|
| 1524 |
+
"learning_rate": 2.322335025380711e-05,
|
| 1525 |
+
"loss": 0.5641,
|
| 1526 |
+
"step": 2110
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 5.380710659898477,
|
| 1530 |
+
"grad_norm": 0.5432278513908386,
|
| 1531 |
+
"learning_rate": 2.3096446700507615e-05,
|
| 1532 |
+
"loss": 0.3232,
|
| 1533 |
+
"step": 2120
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 5.406091370558376,
|
| 1537 |
+
"grad_norm": 2.8115179538726807,
|
| 1538 |
+
"learning_rate": 2.296954314720812e-05,
|
| 1539 |
+
"loss": 0.2467,
|
| 1540 |
+
"step": 2130
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 5.431472081218274,
|
| 1544 |
+
"grad_norm": 0.23173533380031586,
|
| 1545 |
+
"learning_rate": 2.284263959390863e-05,
|
| 1546 |
+
"loss": 0.0366,
|
| 1547 |
+
"step": 2140
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 5.456852791878172,
|
| 1551 |
+
"grad_norm": 0.22747744619846344,
|
| 1552 |
+
"learning_rate": 2.2715736040609136e-05,
|
| 1553 |
+
"loss": 0.2833,
|
| 1554 |
+
"step": 2150
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 5.482233502538071,
|
| 1558 |
+
"grad_norm": 0.18748463690280914,
|
| 1559 |
+
"learning_rate": 2.2588832487309646e-05,
|
| 1560 |
+
"loss": 0.2338,
|
| 1561 |
+
"step": 2160
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 5.50761421319797,
|
| 1565 |
+
"grad_norm": 0.2156940996646881,
|
| 1566 |
+
"learning_rate": 2.2461928934010155e-05,
|
| 1567 |
+
"loss": 0.0311,
|
| 1568 |
+
"step": 2170
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 5.532994923857868,
|
| 1572 |
+
"grad_norm": 26.930912017822266,
|
| 1573 |
+
"learning_rate": 2.233502538071066e-05,
|
| 1574 |
+
"loss": 0.2466,
|
| 1575 |
+
"step": 2180
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 5.558375634517766,
|
| 1579 |
+
"grad_norm": 23.53707504272461,
|
| 1580 |
+
"learning_rate": 2.2208121827411167e-05,
|
| 1581 |
+
"loss": 0.2277,
|
| 1582 |
+
"step": 2190
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 5.583756345177665,
|
| 1586 |
+
"grad_norm": 6.032989025115967,
|
| 1587 |
+
"learning_rate": 2.2081218274111677e-05,
|
| 1588 |
+
"loss": 0.1784,
|
| 1589 |
+
"step": 2200
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 5.6091370558375635,
|
| 1593 |
+
"grad_norm": 15.71194076538086,
|
| 1594 |
+
"learning_rate": 2.1954314720812183e-05,
|
| 1595 |
+
"loss": 0.2382,
|
| 1596 |
+
"step": 2210
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 5.634517766497462,
|
| 1600 |
+
"grad_norm": 6.757187366485596,
|
| 1601 |
+
"learning_rate": 2.182741116751269e-05,
|
| 1602 |
+
"loss": 0.2233,
|
| 1603 |
+
"step": 2220
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 5.659898477157361,
|
| 1607 |
+
"grad_norm": 6.697112560272217,
|
| 1608 |
+
"learning_rate": 2.17005076142132e-05,
|
| 1609 |
+
"loss": 0.2094,
|
| 1610 |
+
"step": 2230
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 5.685279187817259,
|
| 1614 |
+
"grad_norm": 12.990015983581543,
|
| 1615 |
+
"learning_rate": 2.1573604060913707e-05,
|
| 1616 |
+
"loss": 0.167,
|
| 1617 |
+
"step": 2240
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 5.710659898477157,
|
| 1621 |
+
"grad_norm": 0.234835684299469,
|
| 1622 |
+
"learning_rate": 2.1446700507614213e-05,
|
| 1623 |
+
"loss": 0.1446,
|
| 1624 |
+
"step": 2250
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 5.7360406091370555,
|
| 1628 |
+
"grad_norm": 6.382123947143555,
|
| 1629 |
+
"learning_rate": 2.1319796954314723e-05,
|
| 1630 |
+
"loss": 0.2736,
|
| 1631 |
+
"step": 2260
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 5.761421319796955,
|
| 1635 |
+
"grad_norm": 0.26427528262138367,
|
| 1636 |
+
"learning_rate": 2.119289340101523e-05,
|
| 1637 |
+
"loss": 0.1443,
|
| 1638 |
+
"step": 2270
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 5.786802030456853,
|
| 1642 |
+
"grad_norm": 0.2004345953464508,
|
| 1643 |
+
"learning_rate": 2.1065989847715735e-05,
|
| 1644 |
+
"loss": 0.1685,
|
| 1645 |
+
"step": 2280
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 5.812182741116751,
|
| 1649 |
+
"grad_norm": 0.21756359934806824,
|
| 1650 |
+
"learning_rate": 2.0939086294416244e-05,
|
| 1651 |
+
"loss": 0.385,
|
| 1652 |
+
"step": 2290
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 5.837563451776649,
|
| 1656 |
+
"grad_norm": 5.085987567901611,
|
| 1657 |
+
"learning_rate": 2.0812182741116754e-05,
|
| 1658 |
+
"loss": 0.233,
|
| 1659 |
+
"step": 2300
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 5.862944162436548,
|
| 1663 |
+
"grad_norm": 0.2375006377696991,
|
| 1664 |
+
"learning_rate": 2.068527918781726e-05,
|
| 1665 |
+
"loss": 0.2106,
|
| 1666 |
+
"step": 2310
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 5.888324873096447,
|
| 1670 |
+
"grad_norm": 11.780427932739258,
|
| 1671 |
+
"learning_rate": 2.055837563451777e-05,
|
| 1672 |
+
"loss": 0.2142,
|
| 1673 |
+
"step": 2320
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 5.913705583756345,
|
| 1677 |
+
"grad_norm": 1.2173970937728882,
|
| 1678 |
+
"learning_rate": 2.0431472081218275e-05,
|
| 1679 |
+
"loss": 0.0754,
|
| 1680 |
+
"step": 2330
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 5.939086294416244,
|
| 1684 |
+
"grad_norm": 6.028761386871338,
|
| 1685 |
+
"learning_rate": 2.030456852791878e-05,
|
| 1686 |
+
"loss": 0.4157,
|
| 1687 |
+
"step": 2340
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 5.964467005076142,
|
| 1691 |
+
"grad_norm": 19.581605911254883,
|
| 1692 |
+
"learning_rate": 2.017766497461929e-05,
|
| 1693 |
+
"loss": 0.1866,
|
| 1694 |
+
"step": 2350
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 5.98984771573604,
|
| 1698 |
+
"grad_norm": 0.22685207426548004,
|
| 1699 |
+
"learning_rate": 2.0050761421319797e-05,
|
| 1700 |
+
"loss": 0.2823,
|
| 1701 |
+
"step": 2360
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 6.0,
|
| 1705 |
+
"eval_loss": 0.9028552770614624,
|
| 1706 |
+
"eval_runtime": 8.7725,
|
| 1707 |
+
"eval_samples_per_second": 89.826,
|
| 1708 |
+
"eval_steps_per_second": 11.285,
|
| 1709 |
+
"step": 2364
|
| 1710 |
+
},
|
| 1711 |
+
{
|
| 1712 |
+
"epoch": 6.0152284263959395,
|
| 1713 |
+
"grad_norm": 0.280078649520874,
|
| 1714 |
+
"learning_rate": 1.9923857868020303e-05,
|
| 1715 |
+
"loss": 0.1616,
|
| 1716 |
+
"step": 2370
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 6.040609137055838,
|
| 1720 |
+
"grad_norm": 0.17733518779277802,
|
| 1721 |
+
"learning_rate": 1.9796954314720815e-05,
|
| 1722 |
+
"loss": 0.0129,
|
| 1723 |
+
"step": 2380
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"epoch": 6.065989847715736,
|
| 1727 |
+
"grad_norm": 4.652949333190918,
|
| 1728 |
+
"learning_rate": 1.967005076142132e-05,
|
| 1729 |
+
"loss": 0.2607,
|
| 1730 |
+
"step": 2390
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 6.091370558375634,
|
| 1734 |
+
"grad_norm": 17.740243911743164,
|
| 1735 |
+
"learning_rate": 1.9543147208121827e-05,
|
| 1736 |
+
"loss": 0.2297,
|
| 1737 |
+
"step": 2400
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 6.116751269035533,
|
| 1741 |
+
"grad_norm": 59.82321548461914,
|
| 1742 |
+
"learning_rate": 1.9416243654822337e-05,
|
| 1743 |
+
"loss": 0.2063,
|
| 1744 |
+
"step": 2410
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 6.1421319796954315,
|
| 1748 |
+
"grad_norm": 0.1367008090019226,
|
| 1749 |
+
"learning_rate": 1.9289340101522843e-05,
|
| 1750 |
+
"loss": 0.1211,
|
| 1751 |
+
"step": 2420
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 6.16751269035533,
|
| 1755 |
+
"grad_norm": 0.14475467801094055,
|
| 1756 |
+
"learning_rate": 1.916243654822335e-05,
|
| 1757 |
+
"loss": 0.0078,
|
| 1758 |
+
"step": 2430
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 6.192893401015229,
|
| 1762 |
+
"grad_norm": 14.927183151245117,
|
| 1763 |
+
"learning_rate": 1.9035532994923858e-05,
|
| 1764 |
+
"loss": 0.1697,
|
| 1765 |
+
"step": 2440
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 6.218274111675127,
|
| 1769 |
+
"grad_norm": 0.11621695011854172,
|
| 1770 |
+
"learning_rate": 1.8908629441624368e-05,
|
| 1771 |
+
"loss": 0.0255,
|
| 1772 |
+
"step": 2450
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 6.243654822335025,
|
| 1776 |
+
"grad_norm": 0.12786021828651428,
|
| 1777 |
+
"learning_rate": 1.8781725888324874e-05,
|
| 1778 |
+
"loss": 0.008,
|
| 1779 |
+
"step": 2460
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 6.269035532994923,
|
| 1783 |
+
"grad_norm": 0.07750383019447327,
|
| 1784 |
+
"learning_rate": 1.8654822335025383e-05,
|
| 1785 |
+
"loss": 0.21,
|
| 1786 |
+
"step": 2470
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 6.2944162436548226,
|
| 1790 |
+
"grad_norm": 0.11915043741464615,
|
| 1791 |
+
"learning_rate": 1.852791878172589e-05,
|
| 1792 |
+
"loss": 0.0661,
|
| 1793 |
+
"step": 2480
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 6.319796954314721,
|
| 1797 |
+
"grad_norm": 0.09122921526432037,
|
| 1798 |
+
"learning_rate": 1.8401015228426395e-05,
|
| 1799 |
+
"loss": 0.0079,
|
| 1800 |
+
"step": 2490
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 6.345177664974619,
|
| 1804 |
+
"grad_norm": 0.08648554980754852,
|
| 1805 |
+
"learning_rate": 1.8274111675126904e-05,
|
| 1806 |
+
"loss": 0.121,
|
| 1807 |
+
"step": 2500
|
| 1808 |
+
},
|
| 1809 |
+
{
|
| 1810 |
+
"epoch": 6.370558375634518,
|
| 1811 |
+
"grad_norm": 0.13753150403499603,
|
| 1812 |
+
"learning_rate": 1.814720812182741e-05,
|
| 1813 |
+
"loss": 0.0717,
|
| 1814 |
+
"step": 2510
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 6.395939086294416,
|
| 1818 |
+
"grad_norm": 20.555654525756836,
|
| 1819 |
+
"learning_rate": 1.802030456852792e-05,
|
| 1820 |
+
"loss": 0.0759,
|
| 1821 |
+
"step": 2520
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 6.4213197969543145,
|
| 1825 |
+
"grad_norm": 0.07365628331899643,
|
| 1826 |
+
"learning_rate": 1.789340101522843e-05,
|
| 1827 |
+
"loss": 0.0801,
|
| 1828 |
+
"step": 2530
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 6.446700507614214,
|
| 1832 |
+
"grad_norm": 0.05513671413064003,
|
| 1833 |
+
"learning_rate": 1.7766497461928935e-05,
|
| 1834 |
+
"loss": 0.0839,
|
| 1835 |
+
"step": 2540
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 6.472081218274112,
|
| 1839 |
+
"grad_norm": 1.4962226152420044,
|
| 1840 |
+
"learning_rate": 1.763959390862944e-05,
|
| 1841 |
+
"loss": 0.1288,
|
| 1842 |
+
"step": 2550
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 6.49746192893401,
|
| 1846 |
+
"grad_norm": 10.59736442565918,
|
| 1847 |
+
"learning_rate": 1.751269035532995e-05,
|
| 1848 |
+
"loss": 0.1318,
|
| 1849 |
+
"step": 2560
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 6.522842639593908,
|
| 1853 |
+
"grad_norm": 0.06739044189453125,
|
| 1854 |
+
"learning_rate": 1.7385786802030457e-05,
|
| 1855 |
+
"loss": 0.1288,
|
| 1856 |
+
"step": 2570
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 6.548223350253807,
|
| 1860 |
+
"grad_norm": 3.2351067066192627,
|
| 1861 |
+
"learning_rate": 1.7258883248730966e-05,
|
| 1862 |
+
"loss": 0.0801,
|
| 1863 |
+
"step": 2580
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 6.573604060913706,
|
| 1867 |
+
"grad_norm": 0.10322804003953934,
|
| 1868 |
+
"learning_rate": 1.7131979695431472e-05,
|
| 1869 |
+
"loss": 0.0687,
|
| 1870 |
+
"step": 2590
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 6.598984771573604,
|
| 1874 |
+
"grad_norm": 1.834286093711853,
|
| 1875 |
+
"learning_rate": 1.700507614213198e-05,
|
| 1876 |
+
"loss": 0.1842,
|
| 1877 |
+
"step": 2600
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 6.624365482233502,
|
| 1881 |
+
"grad_norm": 0.10404256731271744,
|
| 1882 |
+
"learning_rate": 1.6878172588832487e-05,
|
| 1883 |
+
"loss": 0.2148,
|
| 1884 |
+
"step": 2610
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 6.649746192893401,
|
| 1888 |
+
"grad_norm": 0.11388051509857178,
|
| 1889 |
+
"learning_rate": 1.6751269035532997e-05,
|
| 1890 |
+
"loss": 0.2577,
|
| 1891 |
+
"step": 2620
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 6.675126903553299,
|
| 1895 |
+
"grad_norm": 0.1220388114452362,
|
| 1896 |
+
"learning_rate": 1.6624365482233503e-05,
|
| 1897 |
+
"loss": 0.0425,
|
| 1898 |
+
"step": 2630
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 6.700507614213198,
|
| 1902 |
+
"grad_norm": 18.549720764160156,
|
| 1903 |
+
"learning_rate": 1.649746192893401e-05,
|
| 1904 |
+
"loss": 0.2702,
|
| 1905 |
+
"step": 2640
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 6.725888324873097,
|
| 1909 |
+
"grad_norm": 0.12684045732021332,
|
| 1910 |
+
"learning_rate": 1.6370558375634518e-05,
|
| 1911 |
+
"loss": 0.334,
|
| 1912 |
+
"step": 2650
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 6.751269035532995,
|
| 1916 |
+
"grad_norm": 0.27108675241470337,
|
| 1917 |
+
"learning_rate": 1.6243654822335024e-05,
|
| 1918 |
+
"loss": 0.1592,
|
| 1919 |
+
"step": 2660
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 6.776649746192893,
|
| 1923 |
+
"grad_norm": 0.17042884230613708,
|
| 1924 |
+
"learning_rate": 1.6116751269035534e-05,
|
| 1925 |
+
"loss": 0.0726,
|
| 1926 |
+
"step": 2670
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 6.802030456852792,
|
| 1930 |
+
"grad_norm": 0.11051168292760849,
|
| 1931 |
+
"learning_rate": 1.5989847715736043e-05,
|
| 1932 |
+
"loss": 0.0702,
|
| 1933 |
+
"step": 2680
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 6.8274111675126905,
|
| 1937 |
+
"grad_norm": 0.11623559892177582,
|
| 1938 |
+
"learning_rate": 1.586294416243655e-05,
|
| 1939 |
+
"loss": 0.1979,
|
| 1940 |
+
"step": 2690
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 6.852791878172589,
|
| 1944 |
+
"grad_norm": 0.16744251549243927,
|
| 1945 |
+
"learning_rate": 1.5736040609137055e-05,
|
| 1946 |
+
"loss": 0.1624,
|
| 1947 |
+
"step": 2700
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 6.878172588832487,
|
| 1951 |
+
"grad_norm": 62.56240463256836,
|
| 1952 |
+
"learning_rate": 1.5609137055837564e-05,
|
| 1953 |
+
"loss": 0.1029,
|
| 1954 |
+
"step": 2710
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 6.903553299492386,
|
| 1958 |
+
"grad_norm": 0.11622487008571625,
|
| 1959 |
+
"learning_rate": 1.548223350253807e-05,
|
| 1960 |
+
"loss": 0.0093,
|
| 1961 |
+
"step": 2720
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 6.928934010152284,
|
| 1965 |
+
"grad_norm": 0.09292230755090714,
|
| 1966 |
+
"learning_rate": 1.535532994923858e-05,
|
| 1967 |
+
"loss": 0.1308,
|
| 1968 |
+
"step": 2730
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 6.9543147208121825,
|
| 1972 |
+
"grad_norm": 0.20956498384475708,
|
| 1973 |
+
"learning_rate": 1.5228426395939088e-05,
|
| 1974 |
+
"loss": 0.0052,
|
| 1975 |
+
"step": 2740
|
| 1976 |
+
},
|
| 1977 |
+
{
|
| 1978 |
+
"epoch": 6.979695431472082,
|
| 1979 |
+
"grad_norm": 44.860416412353516,
|
| 1980 |
+
"learning_rate": 1.5101522842639595e-05,
|
| 1981 |
+
"loss": 0.2451,
|
| 1982 |
+
"step": 2750
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 7.0,
|
| 1986 |
+
"eval_loss": 1.0761854648590088,
|
| 1987 |
+
"eval_runtime": 8.8065,
|
| 1988 |
+
"eval_samples_per_second": 89.479,
|
| 1989 |
+
"eval_steps_per_second": 11.242,
|
| 1990 |
+
"step": 2758
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 7.00507614213198,
|
| 1994 |
+
"grad_norm": 0.09350736439228058,
|
| 1995 |
+
"learning_rate": 1.4974619289340103e-05,
|
| 1996 |
+
"loss": 0.0638,
|
| 1997 |
+
"step": 2760
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 7.030456852791878,
|
| 2001 |
+
"grad_norm": 3.844224214553833,
|
| 2002 |
+
"learning_rate": 1.484771573604061e-05,
|
| 2003 |
+
"loss": 0.0735,
|
| 2004 |
+
"step": 2770
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 7.055837563451776,
|
| 2008 |
+
"grad_norm": 0.09096001088619232,
|
| 2009 |
+
"learning_rate": 1.4720812182741117e-05,
|
| 2010 |
+
"loss": 0.0736,
|
| 2011 |
+
"step": 2780
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 7.081218274111675,
|
| 2015 |
+
"grad_norm": 0.06552541255950928,
|
| 2016 |
+
"learning_rate": 1.4593908629441624e-05,
|
| 2017 |
+
"loss": 0.0525,
|
| 2018 |
+
"step": 2790
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 7.106598984771574,
|
| 2022 |
+
"grad_norm": 0.10838861763477325,
|
| 2023 |
+
"learning_rate": 1.4467005076142132e-05,
|
| 2024 |
+
"loss": 0.4002,
|
| 2025 |
+
"step": 2800
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 7.131979695431472,
|
| 2029 |
+
"grad_norm": 0.12170158326625824,
|
| 2030 |
+
"learning_rate": 1.4340101522842641e-05,
|
| 2031 |
+
"loss": 0.1041,
|
| 2032 |
+
"step": 2810
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 7.157360406091371,
|
| 2036 |
+
"grad_norm": 0.15080027282238007,
|
| 2037 |
+
"learning_rate": 1.421319796954315e-05,
|
| 2038 |
+
"loss": 0.085,
|
| 2039 |
+
"step": 2820
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 7.182741116751269,
|
| 2043 |
+
"grad_norm": 0.13942450284957886,
|
| 2044 |
+
"learning_rate": 1.4086294416243657e-05,
|
| 2045 |
+
"loss": 0.007,
|
| 2046 |
+
"step": 2830
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 7.208121827411167,
|
| 2050 |
+
"grad_norm": 0.09515988826751709,
|
| 2051 |
+
"learning_rate": 1.3959390862944163e-05,
|
| 2052 |
+
"loss": 0.1357,
|
| 2053 |
+
"step": 2840
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 7.233502538071066,
|
| 2057 |
+
"grad_norm": 0.09349621832370758,
|
| 2058 |
+
"learning_rate": 1.383248730964467e-05,
|
| 2059 |
+
"loss": 0.0644,
|
| 2060 |
+
"step": 2850
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 7.258883248730965,
|
| 2064 |
+
"grad_norm": 0.25502246618270874,
|
| 2065 |
+
"learning_rate": 1.3705583756345178e-05,
|
| 2066 |
+
"loss": 0.0059,
|
| 2067 |
+
"step": 2860
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 7.284263959390863,
|
| 2071 |
+
"grad_norm": 5.170173168182373,
|
| 2072 |
+
"learning_rate": 1.3578680203045684e-05,
|
| 2073 |
+
"loss": 0.072,
|
| 2074 |
+
"step": 2870
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 7.309644670050761,
|
| 2078 |
+
"grad_norm": 0.09840302914381027,
|
| 2079 |
+
"learning_rate": 1.3451776649746192e-05,
|
| 2080 |
+
"loss": 0.1308,
|
| 2081 |
+
"step": 2880
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 7.33502538071066,
|
| 2085 |
+
"grad_norm": 0.07067205011844635,
|
| 2086 |
+
"learning_rate": 1.3324873096446703e-05,
|
| 2087 |
+
"loss": 0.1064,
|
| 2088 |
+
"step": 2890
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 7.3604060913705585,
|
| 2092 |
+
"grad_norm": 2.9675955772399902,
|
| 2093 |
+
"learning_rate": 1.3197969543147209e-05,
|
| 2094 |
+
"loss": 0.1342,
|
| 2095 |
+
"step": 2900
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 7.385786802030457,
|
| 2099 |
+
"grad_norm": 0.07539036124944687,
|
| 2100 |
+
"learning_rate": 1.3071065989847717e-05,
|
| 2101 |
+
"loss": 0.0642,
|
| 2102 |
+
"step": 2910
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 7.411167512690355,
|
| 2106 |
+
"grad_norm": 0.19838696718215942,
|
| 2107 |
+
"learning_rate": 1.2944162436548224e-05,
|
| 2108 |
+
"loss": 0.2019,
|
| 2109 |
+
"step": 2920
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 7.436548223350254,
|
| 2113 |
+
"grad_norm": 0.08878432959318161,
|
| 2114 |
+
"learning_rate": 1.281725888324873e-05,
|
| 2115 |
+
"loss": 0.104,
|
| 2116 |
+
"step": 2930
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 7.461928934010152,
|
| 2120 |
+
"grad_norm": 0.09144201874732971,
|
| 2121 |
+
"learning_rate": 1.2690355329949238e-05,
|
| 2122 |
+
"loss": 0.0046,
|
| 2123 |
+
"step": 2940
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 7.4873096446700504,
|
| 2127 |
+
"grad_norm": 0.0494062639772892,
|
| 2128 |
+
"learning_rate": 1.2563451776649746e-05,
|
| 2129 |
+
"loss": 0.0705,
|
| 2130 |
+
"step": 2950
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 7.5126903553299496,
|
| 2134 |
+
"grad_norm": 0.1255367249250412,
|
| 2135 |
+
"learning_rate": 1.2436548223350254e-05,
|
| 2136 |
+
"loss": 0.0382,
|
| 2137 |
+
"step": 2960
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 7.538071065989848,
|
| 2141 |
+
"grad_norm": 0.06985324621200562,
|
| 2142 |
+
"learning_rate": 1.2309644670050761e-05,
|
| 2143 |
+
"loss": 0.0863,
|
| 2144 |
+
"step": 2970
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 7.563451776649746,
|
| 2148 |
+
"grad_norm": 0.08622120320796967,
|
| 2149 |
+
"learning_rate": 1.218274111675127e-05,
|
| 2150 |
+
"loss": 0.0729,
|
| 2151 |
+
"step": 2980
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 7.588832487309645,
|
| 2155 |
+
"grad_norm": 0.07268553227186203,
|
| 2156 |
+
"learning_rate": 1.2055837563451777e-05,
|
| 2157 |
+
"loss": 0.1981,
|
| 2158 |
+
"step": 2990
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 7.614213197969543,
|
| 2162 |
+
"grad_norm": 5.232954025268555,
|
| 2163 |
+
"learning_rate": 1.1928934010152284e-05,
|
| 2164 |
+
"loss": 0.0885,
|
| 2165 |
+
"step": 3000
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 7.6395939086294415,
|
| 2169 |
+
"grad_norm": 0.07025603950023651,
|
| 2170 |
+
"learning_rate": 1.1802030456852794e-05,
|
| 2171 |
+
"loss": 0.1437,
|
| 2172 |
+
"step": 3010
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 7.66497461928934,
|
| 2176 |
+
"grad_norm": 3.0778110027313232,
|
| 2177 |
+
"learning_rate": 1.16751269035533e-05,
|
| 2178 |
+
"loss": 0.2441,
|
| 2179 |
+
"step": 3020
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 7.690355329949239,
|
| 2183 |
+
"grad_norm": 0.1191609650850296,
|
| 2184 |
+
"learning_rate": 1.1548223350253808e-05,
|
| 2185 |
+
"loss": 0.0717,
|
| 2186 |
+
"step": 3030
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 7.715736040609137,
|
| 2190 |
+
"grad_norm": 0.07436536252498627,
|
| 2191 |
+
"learning_rate": 1.1421319796954315e-05,
|
| 2192 |
+
"loss": 0.0636,
|
| 2193 |
+
"step": 3040
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 7.741116751269035,
|
| 2197 |
+
"grad_norm": 0.08062940835952759,
|
| 2198 |
+
"learning_rate": 1.1294416243654823e-05,
|
| 2199 |
+
"loss": 0.0918,
|
| 2200 |
+
"step": 3050
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 7.7664974619289335,
|
| 2204 |
+
"grad_norm": 0.06816007196903229,
|
| 2205 |
+
"learning_rate": 1.116751269035533e-05,
|
| 2206 |
+
"loss": 0.1376,
|
| 2207 |
+
"step": 3060
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 7.791878172588833,
|
| 2211 |
+
"grad_norm": 2.068460464477539,
|
| 2212 |
+
"learning_rate": 1.1040609137055838e-05,
|
| 2213 |
+
"loss": 0.0642,
|
| 2214 |
+
"step": 3070
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 7.817258883248731,
|
| 2218 |
+
"grad_norm": 0.1358051747083664,
|
| 2219 |
+
"learning_rate": 1.0913705583756344e-05,
|
| 2220 |
+
"loss": 0.1455,
|
| 2221 |
+
"step": 3080
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 7.842639593908629,
|
| 2225 |
+
"grad_norm": 0.07467884570360184,
|
| 2226 |
+
"learning_rate": 1.0786802030456854e-05,
|
| 2227 |
+
"loss": 0.0041,
|
| 2228 |
+
"step": 3090
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 7.868020304568528,
|
| 2232 |
+
"grad_norm": 0.08706346899271011,
|
| 2233 |
+
"learning_rate": 1.0659898477157361e-05,
|
| 2234 |
+
"loss": 0.0838,
|
| 2235 |
+
"step": 3100
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 7.893401015228426,
|
| 2239 |
+
"grad_norm": 0.07726430892944336,
|
| 2240 |
+
"learning_rate": 1.0532994923857867e-05,
|
| 2241 |
+
"loss": 0.0689,
|
| 2242 |
+
"step": 3110
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 7.918781725888325,
|
| 2246 |
+
"grad_norm": 0.09187603741884232,
|
| 2247 |
+
"learning_rate": 1.0406091370558377e-05,
|
| 2248 |
+
"loss": 0.0785,
|
| 2249 |
+
"step": 3120
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 7.944162436548224,
|
| 2253 |
+
"grad_norm": 0.1465052366256714,
|
| 2254 |
+
"learning_rate": 1.0279187817258885e-05,
|
| 2255 |
+
"loss": 0.0778,
|
| 2256 |
+
"step": 3130
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 7.969543147208122,
|
| 2260 |
+
"grad_norm": 0.06538953632116318,
|
| 2261 |
+
"learning_rate": 1.015228426395939e-05,
|
| 2262 |
+
"loss": 0.0577,
|
| 2263 |
+
"step": 3140
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 7.99492385786802,
|
| 2267 |
+
"grad_norm": 0.17358700931072235,
|
| 2268 |
+
"learning_rate": 1.0025380710659898e-05,
|
| 2269 |
+
"loss": 0.0041,
|
| 2270 |
+
"step": 3150
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 8.0,
|
| 2274 |
+
"eval_loss": 1.1998591423034668,
|
| 2275 |
+
"eval_runtime": 8.6883,
|
| 2276 |
+
"eval_samples_per_second": 90.697,
|
| 2277 |
+
"eval_steps_per_second": 11.395,
|
| 2278 |
+
"step": 3152
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 8.02030456852792,
|
| 2282 |
+
"grad_norm": 0.05773229897022247,
|
| 2283 |
+
"learning_rate": 9.898477157360408e-06,
|
| 2284 |
+
"loss": 0.0856,
|
| 2285 |
+
"step": 3160
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 8.045685279187817,
|
| 2289 |
+
"grad_norm": 0.07206518203020096,
|
| 2290 |
+
"learning_rate": 9.771573604060914e-06,
|
| 2291 |
+
"loss": 0.1201,
|
| 2292 |
+
"step": 3170
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 8.071065989847716,
|
| 2296 |
+
"grad_norm": 0.07798884809017181,
|
| 2297 |
+
"learning_rate": 9.644670050761421e-06,
|
| 2298 |
+
"loss": 0.069,
|
| 2299 |
+
"step": 3180
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 8.096446700507615,
|
| 2303 |
+
"grad_norm": 0.07664274424314499,
|
| 2304 |
+
"learning_rate": 9.517766497461929e-06,
|
| 2305 |
+
"loss": 0.0036,
|
| 2306 |
+
"step": 3190
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 8.121827411167512,
|
| 2310 |
+
"grad_norm": 0.050856612622737885,
|
| 2311 |
+
"learning_rate": 9.390862944162437e-06,
|
| 2312 |
+
"loss": 0.003,
|
| 2313 |
+
"step": 3200
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 8.147208121827411,
|
| 2317 |
+
"grad_norm": 0.0529194138944149,
|
| 2318 |
+
"learning_rate": 9.263959390862944e-06,
|
| 2319 |
+
"loss": 0.1412,
|
| 2320 |
+
"step": 3210
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 8.17258883248731,
|
| 2324 |
+
"grad_norm": 13.00708293914795,
|
| 2325 |
+
"learning_rate": 9.137055837563452e-06,
|
| 2326 |
+
"loss": 0.0513,
|
| 2327 |
+
"step": 3220
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 8.197969543147208,
|
| 2331 |
+
"grad_norm": 0.06696105003356934,
|
| 2332 |
+
"learning_rate": 9.01015228426396e-06,
|
| 2333 |
+
"loss": 0.0704,
|
| 2334 |
+
"step": 3230
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 8.223350253807107,
|
| 2338 |
+
"grad_norm": 0.06790643185377121,
|
| 2339 |
+
"learning_rate": 8.883248730964468e-06,
|
| 2340 |
+
"loss": 0.1119,
|
| 2341 |
+
"step": 3240
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 8.248730964467006,
|
| 2345 |
+
"grad_norm": 0.06100468337535858,
|
| 2346 |
+
"learning_rate": 8.756345177664975e-06,
|
| 2347 |
+
"loss": 0.0032,
|
| 2348 |
+
"step": 3250
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 8.274111675126903,
|
| 2352 |
+
"grad_norm": 8.70279312133789,
|
| 2353 |
+
"learning_rate": 8.629441624365483e-06,
|
| 2354 |
+
"loss": 0.0811,
|
| 2355 |
+
"step": 3260
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 8.299492385786802,
|
| 2359 |
+
"grad_norm": 0.09535681456327438,
|
| 2360 |
+
"learning_rate": 8.50253807106599e-06,
|
| 2361 |
+
"loss": 0.0795,
|
| 2362 |
+
"step": 3270
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 8.3248730964467,
|
| 2366 |
+
"grad_norm": 0.07176022976636887,
|
| 2367 |
+
"learning_rate": 8.375634517766498e-06,
|
| 2368 |
+
"loss": 0.0893,
|
| 2369 |
+
"step": 3280
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 8.350253807106599,
|
| 2373 |
+
"grad_norm": 0.051213160157203674,
|
| 2374 |
+
"learning_rate": 8.248730964467004e-06,
|
| 2375 |
+
"loss": 0.0701,
|
| 2376 |
+
"step": 3290
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 8.375634517766498,
|
| 2380 |
+
"grad_norm": 0.09963097423315048,
|
| 2381 |
+
"learning_rate": 8.121827411167512e-06,
|
| 2382 |
+
"loss": 0.0745,
|
| 2383 |
+
"step": 3300
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 8.401015228426395,
|
| 2387 |
+
"grad_norm": 0.0691935271024704,
|
| 2388 |
+
"learning_rate": 7.994923857868022e-06,
|
| 2389 |
+
"loss": 0.0884,
|
| 2390 |
+
"step": 3310
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 8.426395939086294,
|
| 2394 |
+
"grad_norm": 0.16761644184589386,
|
| 2395 |
+
"learning_rate": 7.868020304568528e-06,
|
| 2396 |
+
"loss": 0.0043,
|
| 2397 |
+
"step": 3320
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 8.451776649746193,
|
| 2401 |
+
"grad_norm": 0.04886673018336296,
|
| 2402 |
+
"learning_rate": 7.741116751269035e-06,
|
| 2403 |
+
"loss": 0.0037,
|
| 2404 |
+
"step": 3330
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 8.47715736040609,
|
| 2408 |
+
"grad_norm": 0.06593126058578491,
|
| 2409 |
+
"learning_rate": 7.614213197969544e-06,
|
| 2410 |
+
"loss": 0.0029,
|
| 2411 |
+
"step": 3340
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 8.50253807106599,
|
| 2415 |
+
"grad_norm": 8.315119743347168,
|
| 2416 |
+
"learning_rate": 7.4873096446700515e-06,
|
| 2417 |
+
"loss": 0.1359,
|
| 2418 |
+
"step": 3350
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 8.527918781725889,
|
| 2422 |
+
"grad_norm": 0.19459432363510132,
|
| 2423 |
+
"learning_rate": 7.360406091370558e-06,
|
| 2424 |
+
"loss": 0.0771,
|
| 2425 |
+
"step": 3360
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 8.553299492385786,
|
| 2429 |
+
"grad_norm": 0.09421133995056152,
|
| 2430 |
+
"learning_rate": 7.233502538071066e-06,
|
| 2431 |
+
"loss": 0.003,
|
| 2432 |
+
"step": 3370
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 8.578680203045685,
|
| 2436 |
+
"grad_norm": 0.04559624567627907,
|
| 2437 |
+
"learning_rate": 7.106598984771575e-06,
|
| 2438 |
+
"loss": 0.0647,
|
| 2439 |
+
"step": 3380
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 8.604060913705585,
|
| 2443 |
+
"grad_norm": 3.285724639892578,
|
| 2444 |
+
"learning_rate": 6.9796954314720814e-06,
|
| 2445 |
+
"loss": 0.0744,
|
| 2446 |
+
"step": 3390
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 8.629441624365482,
|
| 2450 |
+
"grad_norm": 0.0715055987238884,
|
| 2451 |
+
"learning_rate": 6.852791878172589e-06,
|
| 2452 |
+
"loss": 0.0696,
|
| 2453 |
+
"step": 3400
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 8.654822335025381,
|
| 2457 |
+
"grad_norm": 0.05040270462632179,
|
| 2458 |
+
"learning_rate": 6.725888324873096e-06,
|
| 2459 |
+
"loss": 0.0028,
|
| 2460 |
+
"step": 3410
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 8.680203045685278,
|
| 2464 |
+
"grad_norm": 0.05061913654208183,
|
| 2465 |
+
"learning_rate": 6.5989847715736045e-06,
|
| 2466 |
+
"loss": 0.1397,
|
| 2467 |
+
"step": 3420
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 8.705583756345177,
|
| 2471 |
+
"grad_norm": 0.06146248057484627,
|
| 2472 |
+
"learning_rate": 6.472081218274112e-06,
|
| 2473 |
+
"loss": 0.0757,
|
| 2474 |
+
"step": 3430
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 8.730964467005077,
|
| 2478 |
+
"grad_norm": 0.0398724228143692,
|
| 2479 |
+
"learning_rate": 6.345177664974619e-06,
|
| 2480 |
+
"loss": 0.0888,
|
| 2481 |
+
"step": 3440
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 8.756345177664974,
|
| 2485 |
+
"grad_norm": 0.058130159974098206,
|
| 2486 |
+
"learning_rate": 6.218274111675127e-06,
|
| 2487 |
+
"loss": 0.0714,
|
| 2488 |
+
"step": 3450
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 8.781725888324873,
|
| 2492 |
+
"grad_norm": 0.06121835485100746,
|
| 2493 |
+
"learning_rate": 6.091370558375635e-06,
|
| 2494 |
+
"loss": 0.0329,
|
| 2495 |
+
"step": 3460
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 8.807106598984772,
|
| 2499 |
+
"grad_norm": 0.048920340836048126,
|
| 2500 |
+
"learning_rate": 5.964467005076142e-06,
|
| 2501 |
+
"loss": 0.0032,
|
| 2502 |
+
"step": 3470
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 8.83248730964467,
|
| 2506 |
+
"grad_norm": 0.05234299972653389,
|
| 2507 |
+
"learning_rate": 5.83756345177665e-06,
|
| 2508 |
+
"loss": 0.07,
|
| 2509 |
+
"step": 3480
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 8.857868020304569,
|
| 2513 |
+
"grad_norm": 0.04120004549622536,
|
| 2514 |
+
"learning_rate": 5.710659898477158e-06,
|
| 2515 |
+
"loss": 0.152,
|
| 2516 |
+
"step": 3490
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 8.883248730964468,
|
| 2520 |
+
"grad_norm": 0.0480622835457325,
|
| 2521 |
+
"learning_rate": 5.583756345177665e-06,
|
| 2522 |
+
"loss": 0.162,
|
| 2523 |
+
"step": 3500
|
| 2524 |
+
}
|
| 2525 |
+
],
|
| 2526 |
+
"logging_steps": 10,
|
| 2527 |
+
"max_steps": 3940,
|
| 2528 |
+
"num_input_tokens_seen": 0,
|
| 2529 |
+
"num_train_epochs": 10,
|
| 2530 |
+
"save_steps": 500,
|
| 2531 |
+
"stateful_callbacks": {
|
| 2532 |
+
"TrainerControl": {
|
| 2533 |
+
"args": {
|
| 2534 |
+
"should_epoch_stop": false,
|
| 2535 |
+
"should_evaluate": false,
|
| 2536 |
+
"should_log": false,
|
| 2537 |
+
"should_save": true,
|
| 2538 |
+
"should_training_stop": false
|
| 2539 |
+
},
|
| 2540 |
+
"attributes": {}
|
| 2541 |
+
}
|
| 2542 |
+
},
|
| 2543 |
+
"total_flos": 3709087162368000.0,
|
| 2544 |
+
"train_batch_size": 8,
|
| 2545 |
+
"trial_name": null,
|
| 2546 |
+
"trial_params": null
|
| 2547 |
+
}
|
distilbert_ranking/checkpoint-3500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5fa327595511a6fa6b860f7ffa79847603a3d8d181ec87c4527ae8ccaceb8863
|
| 3 |
+
size 5176
|
distilbert_ranking/checkpoint-3500/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
distilbert_ranking/checkpoint-3940/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "distilbert-base-uncased",
|
| 3 |
+
"activation": "gelu",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"DistilBertForSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_dropout": 0.1,
|
| 8 |
+
"dim": 768,
|
| 9 |
+
"dropout": 0.1,
|
| 10 |
+
"hidden_dim": 3072,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"max_position_embeddings": 512,
|
| 13 |
+
"model_type": "distilbert",
|
| 14 |
+
"n_heads": 12,
|
| 15 |
+
"n_layers": 6,
|
| 16 |
+
"pad_token_id": 0,
|
| 17 |
+
"problem_type": "single_label_classification",
|
| 18 |
+
"qa_dropout": 0.1,
|
| 19 |
+
"seq_classif_dropout": 0.2,
|
| 20 |
+
"sinusoidal_pos_embds": false,
|
| 21 |
+
"tie_weights_": true,
|
| 22 |
+
"torch_dtype": "float32",
|
| 23 |
+
"transformers_version": "4.44.0",
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
distilbert_ranking/checkpoint-3940/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c1659eb788caa8eed53779b80179c9fa1813a2fd2205143c5bccc4710177c68e
|
| 3 |
+
size 267832560
|
distilbert_ranking/checkpoint-3940/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:238b2efb74c8dec48037c0369e770277f2d5dd9d9b687d31dc400d44ba0a6b73
|
| 3 |
+
size 535727290
|
distilbert_ranking/checkpoint-3940/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f8f7b762b9891586998e49fa2ae8ef030ed0f63860940f80f9f970f469db8caa
|
| 3 |
+
size 14244
|
distilbert_ranking/checkpoint-3940/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b1ff4d7f36e7b5a3cd09b7c781d38a7fda66bc90a4f56407f69050973e8bd376
|
| 3 |
+
size 1064
|
distilbert_ranking/checkpoint-3940/special_tokens_map.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": "[CLS]",
|
| 3 |
+
"mask_token": "[MASK]",
|
| 4 |
+
"pad_token": "[PAD]",
|
| 5 |
+
"sep_token": "[SEP]",
|
| 6 |
+
"unk_token": "[UNK]"
|
| 7 |
+
}
|
distilbert_ranking/checkpoint-3940/tokenizer_config.json
ADDED
|
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"100": {
|
| 12 |
+
"content": "[UNK]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"101": {
|
| 20 |
+
"content": "[CLS]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"102": {
|
| 28 |
+
"content": "[SEP]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"103": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"clean_up_tokenization_spaces": true,
|
| 45 |
+
"cls_token": "[CLS]",
|
| 46 |
+
"do_basic_tokenize": true,
|
| 47 |
+
"do_lower_case": true,
|
| 48 |
+
"mask_token": "[MASK]",
|
| 49 |
+
"model_max_length": 512,
|
| 50 |
+
"never_split": null,
|
| 51 |
+
"pad_token": "[PAD]",
|
| 52 |
+
"sep_token": "[SEP]",
|
| 53 |
+
"strip_accents": null,
|
| 54 |
+
"tokenize_chinese_chars": true,
|
| 55 |
+
"tokenizer_class": "DistilBertTokenizer",
|
| 56 |
+
"unk_token": "[UNK]"
|
| 57 |
+
}
|
distilbert_ranking/checkpoint-3940/trainer_state.json
ADDED
|
@@ -0,0 +1,2863 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3940,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.025380710659898477,
|
| 13 |
+
"grad_norm": 1.7737329006195068,
|
| 14 |
+
"learning_rate": 4.9873096446700515e-05,
|
| 15 |
+
"loss": 0.4425,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.050761421319796954,
|
| 20 |
+
"grad_norm": 1.6950019598007202,
|
| 21 |
+
"learning_rate": 4.9746192893401014e-05,
|
| 22 |
+
"loss": 0.5594,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.07614213197969544,
|
| 27 |
+
"grad_norm": 3.7862815856933594,
|
| 28 |
+
"learning_rate": 4.961928934010153e-05,
|
| 29 |
+
"loss": 0.5435,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.10152284263959391,
|
| 34 |
+
"grad_norm": 2.450183868408203,
|
| 35 |
+
"learning_rate": 4.949238578680203e-05,
|
| 36 |
+
"loss": 0.5429,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.12690355329949238,
|
| 41 |
+
"grad_norm": 1.6361172199249268,
|
| 42 |
+
"learning_rate": 4.936548223350254e-05,
|
| 43 |
+
"loss": 0.6185,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.15228426395939088,
|
| 48 |
+
"grad_norm": 1.3755431175231934,
|
| 49 |
+
"learning_rate": 4.9238578680203045e-05,
|
| 50 |
+
"loss": 0.5633,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.17766497461928935,
|
| 55 |
+
"grad_norm": 2.1979947090148926,
|
| 56 |
+
"learning_rate": 4.911167512690356e-05,
|
| 57 |
+
"loss": 0.5031,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.20304568527918782,
|
| 62 |
+
"grad_norm": 1.493660569190979,
|
| 63 |
+
"learning_rate": 4.8984771573604064e-05,
|
| 64 |
+
"loss": 0.5942,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.22842639593908629,
|
| 69 |
+
"grad_norm": 1.642001748085022,
|
| 70 |
+
"learning_rate": 4.885786802030457e-05,
|
| 71 |
+
"loss": 0.5513,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.25380710659898476,
|
| 76 |
+
"grad_norm": 4.11527681350708,
|
| 77 |
+
"learning_rate": 4.873096446700508e-05,
|
| 78 |
+
"loss": 0.5544,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.27918781725888325,
|
| 83 |
+
"grad_norm": 1.966141700744629,
|
| 84 |
+
"learning_rate": 4.860406091370558e-05,
|
| 85 |
+
"loss": 0.5311,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.30456852791878175,
|
| 90 |
+
"grad_norm": 1.490822196006775,
|
| 91 |
+
"learning_rate": 4.8477157360406095e-05,
|
| 92 |
+
"loss": 0.4546,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.3299492385786802,
|
| 97 |
+
"grad_norm": 2.575047254562378,
|
| 98 |
+
"learning_rate": 4.83502538071066e-05,
|
| 99 |
+
"loss": 0.4532,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.3553299492385787,
|
| 104 |
+
"grad_norm": 1.9474763870239258,
|
| 105 |
+
"learning_rate": 4.822335025380711e-05,
|
| 106 |
+
"loss": 0.4499,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.38071065989847713,
|
| 111 |
+
"grad_norm": 2.9648075103759766,
|
| 112 |
+
"learning_rate": 4.809644670050762e-05,
|
| 113 |
+
"loss": 0.5425,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.40609137055837563,
|
| 118 |
+
"grad_norm": 1.9473344087600708,
|
| 119 |
+
"learning_rate": 4.7969543147208126e-05,
|
| 120 |
+
"loss": 0.529,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.43147208121827413,
|
| 125 |
+
"grad_norm": 2.3414478302001953,
|
| 126 |
+
"learning_rate": 4.784263959390863e-05,
|
| 127 |
+
"loss": 0.5281,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.45685279187817257,
|
| 132 |
+
"grad_norm": 1.2157188653945923,
|
| 133 |
+
"learning_rate": 4.771573604060914e-05,
|
| 134 |
+
"loss": 0.4963,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.48223350253807107,
|
| 139 |
+
"grad_norm": 3.3916754722595215,
|
| 140 |
+
"learning_rate": 4.758883248730965e-05,
|
| 141 |
+
"loss": 0.4245,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.5076142131979695,
|
| 146 |
+
"grad_norm": 3.030148983001709,
|
| 147 |
+
"learning_rate": 4.746192893401015e-05,
|
| 148 |
+
"loss": 0.3352,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.5329949238578681,
|
| 153 |
+
"grad_norm": 2.1008453369140625,
|
| 154 |
+
"learning_rate": 4.733502538071066e-05,
|
| 155 |
+
"loss": 0.6109,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.5583756345177665,
|
| 160 |
+
"grad_norm": 1.6099292039871216,
|
| 161 |
+
"learning_rate": 4.7208121827411175e-05,
|
| 162 |
+
"loss": 0.5327,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.583756345177665,
|
| 167 |
+
"grad_norm": 1.7559535503387451,
|
| 168 |
+
"learning_rate": 4.7081218274111674e-05,
|
| 169 |
+
"loss": 0.4475,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.6091370558375635,
|
| 174 |
+
"grad_norm": 1.5413782596588135,
|
| 175 |
+
"learning_rate": 4.695431472081219e-05,
|
| 176 |
+
"loss": 0.534,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.6345177664974619,
|
| 181 |
+
"grad_norm": 1.810226321220398,
|
| 182 |
+
"learning_rate": 4.682741116751269e-05,
|
| 183 |
+
"loss": 0.5739,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.6598984771573604,
|
| 188 |
+
"grad_norm": 3.676072120666504,
|
| 189 |
+
"learning_rate": 4.67005076142132e-05,
|
| 190 |
+
"loss": 0.4702,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.6852791878172588,
|
| 195 |
+
"grad_norm": 2.691065788269043,
|
| 196 |
+
"learning_rate": 4.6573604060913705e-05,
|
| 197 |
+
"loss": 0.5858,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.7106598984771574,
|
| 202 |
+
"grad_norm": 1.8071280717849731,
|
| 203 |
+
"learning_rate": 4.644670050761422e-05,
|
| 204 |
+
"loss": 0.5294,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.7360406091370558,
|
| 209 |
+
"grad_norm": 4.747298240661621,
|
| 210 |
+
"learning_rate": 4.631979695431472e-05,
|
| 211 |
+
"loss": 0.4718,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.7614213197969543,
|
| 216 |
+
"grad_norm": 2.4223101139068604,
|
| 217 |
+
"learning_rate": 4.619289340101523e-05,
|
| 218 |
+
"loss": 0.4712,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.7868020304568528,
|
| 223 |
+
"grad_norm": 2.2267863750457764,
|
| 224 |
+
"learning_rate": 4.606598984771574e-05,
|
| 225 |
+
"loss": 0.5829,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.8121827411167513,
|
| 230 |
+
"grad_norm": 2.0561490058898926,
|
| 231 |
+
"learning_rate": 4.593908629441624e-05,
|
| 232 |
+
"loss": 0.4033,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.8375634517766497,
|
| 237 |
+
"grad_norm": 3.0562074184417725,
|
| 238 |
+
"learning_rate": 4.5812182741116755e-05,
|
| 239 |
+
"loss": 0.6535,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.8629441624365483,
|
| 244 |
+
"grad_norm": 3.355173110961914,
|
| 245 |
+
"learning_rate": 4.568527918781726e-05,
|
| 246 |
+
"loss": 0.4635,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.8883248730964467,
|
| 251 |
+
"grad_norm": 4.357395172119141,
|
| 252 |
+
"learning_rate": 4.555837563451777e-05,
|
| 253 |
+
"loss": 0.4995,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.9137055837563451,
|
| 258 |
+
"grad_norm": 1.0308293104171753,
|
| 259 |
+
"learning_rate": 4.543147208121827e-05,
|
| 260 |
+
"loss": 0.5207,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.9390862944162437,
|
| 265 |
+
"grad_norm": 1.9433645009994507,
|
| 266 |
+
"learning_rate": 4.5304568527918786e-05,
|
| 267 |
+
"loss": 0.2991,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.9644670050761421,
|
| 272 |
+
"grad_norm": 3.8325693607330322,
|
| 273 |
+
"learning_rate": 4.517766497461929e-05,
|
| 274 |
+
"loss": 0.6014,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.9898477157360406,
|
| 279 |
+
"grad_norm": 1.0364559888839722,
|
| 280 |
+
"learning_rate": 4.50507614213198e-05,
|
| 281 |
+
"loss": 0.3693,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.0,
|
| 286 |
+
"eval_loss": 0.49558719992637634,
|
| 287 |
+
"eval_runtime": 8.8144,
|
| 288 |
+
"eval_samples_per_second": 89.4,
|
| 289 |
+
"eval_steps_per_second": 11.232,
|
| 290 |
+
"step": 394
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.015228426395939,
|
| 294 |
+
"grad_norm": 2.267185926437378,
|
| 295 |
+
"learning_rate": 4.492385786802031e-05,
|
| 296 |
+
"loss": 0.5173,
|
| 297 |
+
"step": 400
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.0406091370558375,
|
| 301 |
+
"grad_norm": 1.7508792877197266,
|
| 302 |
+
"learning_rate": 4.479695431472081e-05,
|
| 303 |
+
"loss": 0.4722,
|
| 304 |
+
"step": 410
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 1.0659898477157361,
|
| 308 |
+
"grad_norm": 1.0362215042114258,
|
| 309 |
+
"learning_rate": 4.467005076142132e-05,
|
| 310 |
+
"loss": 0.4053,
|
| 311 |
+
"step": 420
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 1.0913705583756346,
|
| 315 |
+
"grad_norm": 1.5984660387039185,
|
| 316 |
+
"learning_rate": 4.454314720812183e-05,
|
| 317 |
+
"loss": 0.6716,
|
| 318 |
+
"step": 430
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 1.116751269035533,
|
| 322 |
+
"grad_norm": 2.6108462810516357,
|
| 323 |
+
"learning_rate": 4.4416243654822335e-05,
|
| 324 |
+
"loss": 0.4829,
|
| 325 |
+
"step": 440
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 1.1421319796954315,
|
| 329 |
+
"grad_norm": 6.843586444854736,
|
| 330 |
+
"learning_rate": 4.428934010152285e-05,
|
| 331 |
+
"loss": 0.4716,
|
| 332 |
+
"step": 450
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.16751269035533,
|
| 336 |
+
"grad_norm": 5.383293151855469,
|
| 337 |
+
"learning_rate": 4.416243654822335e-05,
|
| 338 |
+
"loss": 0.6196,
|
| 339 |
+
"step": 460
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.1928934010152283,
|
| 343 |
+
"grad_norm": 1.4819399118423462,
|
| 344 |
+
"learning_rate": 4.403553299492386e-05,
|
| 345 |
+
"loss": 0.4594,
|
| 346 |
+
"step": 470
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 1.218274111675127,
|
| 350 |
+
"grad_norm": 3.468229055404663,
|
| 351 |
+
"learning_rate": 4.3908629441624365e-05,
|
| 352 |
+
"loss": 0.4838,
|
| 353 |
+
"step": 480
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.2436548223350254,
|
| 357 |
+
"grad_norm": 3.6312994956970215,
|
| 358 |
+
"learning_rate": 4.378172588832488e-05,
|
| 359 |
+
"loss": 0.4686,
|
| 360 |
+
"step": 490
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 1.2690355329949239,
|
| 364 |
+
"grad_norm": 1.4617408514022827,
|
| 365 |
+
"learning_rate": 4.365482233502538e-05,
|
| 366 |
+
"loss": 0.4094,
|
| 367 |
+
"step": 500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 1.2944162436548223,
|
| 371 |
+
"grad_norm": 0.9887295365333557,
|
| 372 |
+
"learning_rate": 4.352791878172589e-05,
|
| 373 |
+
"loss": 0.5219,
|
| 374 |
+
"step": 510
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.3197969543147208,
|
| 378 |
+
"grad_norm": 1.6651699542999268,
|
| 379 |
+
"learning_rate": 4.34010152284264e-05,
|
| 380 |
+
"loss": 0.6159,
|
| 381 |
+
"step": 520
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.3451776649746192,
|
| 385 |
+
"grad_norm": 2.0405914783477783,
|
| 386 |
+
"learning_rate": 4.32741116751269e-05,
|
| 387 |
+
"loss": 0.41,
|
| 388 |
+
"step": 530
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.3705583756345177,
|
| 392 |
+
"grad_norm": 4.893404483795166,
|
| 393 |
+
"learning_rate": 4.3147208121827415e-05,
|
| 394 |
+
"loss": 0.5679,
|
| 395 |
+
"step": 540
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.3959390862944163,
|
| 399 |
+
"grad_norm": 2.044116735458374,
|
| 400 |
+
"learning_rate": 4.302030456852792e-05,
|
| 401 |
+
"loss": 0.4693,
|
| 402 |
+
"step": 550
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.4213197969543148,
|
| 406 |
+
"grad_norm": 2.040766477584839,
|
| 407 |
+
"learning_rate": 4.289340101522843e-05,
|
| 408 |
+
"loss": 0.4736,
|
| 409 |
+
"step": 560
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.4467005076142132,
|
| 413 |
+
"grad_norm": 1.883665919303894,
|
| 414 |
+
"learning_rate": 4.276649746192893e-05,
|
| 415 |
+
"loss": 0.6475,
|
| 416 |
+
"step": 570
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.4720812182741116,
|
| 420 |
+
"grad_norm": 1.6452559232711792,
|
| 421 |
+
"learning_rate": 4.2639593908629446e-05,
|
| 422 |
+
"loss": 0.5092,
|
| 423 |
+
"step": 580
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.49746192893401,
|
| 427 |
+
"grad_norm": 2.059413194656372,
|
| 428 |
+
"learning_rate": 4.251269035532995e-05,
|
| 429 |
+
"loss": 0.5469,
|
| 430 |
+
"step": 590
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.5228426395939088,
|
| 434 |
+
"grad_norm": 5.789554595947266,
|
| 435 |
+
"learning_rate": 4.238578680203046e-05,
|
| 436 |
+
"loss": 0.4548,
|
| 437 |
+
"step": 600
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.548223350253807,
|
| 441 |
+
"grad_norm": 1.7683823108673096,
|
| 442 |
+
"learning_rate": 4.225888324873097e-05,
|
| 443 |
+
"loss": 0.3807,
|
| 444 |
+
"step": 610
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.5736040609137056,
|
| 448 |
+
"grad_norm": 1.6267879009246826,
|
| 449 |
+
"learning_rate": 4.213197969543147e-05,
|
| 450 |
+
"loss": 0.3814,
|
| 451 |
+
"step": 620
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.598984771573604,
|
| 455 |
+
"grad_norm": 2.024305582046509,
|
| 456 |
+
"learning_rate": 4.200507614213198e-05,
|
| 457 |
+
"loss": 0.5454,
|
| 458 |
+
"step": 630
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.6243654822335025,
|
| 462 |
+
"grad_norm": 2.527076482772827,
|
| 463 |
+
"learning_rate": 4.187817258883249e-05,
|
| 464 |
+
"loss": 0.5996,
|
| 465 |
+
"step": 640
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.649746192893401,
|
| 469 |
+
"grad_norm": 2.8863742351531982,
|
| 470 |
+
"learning_rate": 4.1751269035532995e-05,
|
| 471 |
+
"loss": 0.4548,
|
| 472 |
+
"step": 650
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.6751269035532994,
|
| 476 |
+
"grad_norm": 1.8288835287094116,
|
| 477 |
+
"learning_rate": 4.162436548223351e-05,
|
| 478 |
+
"loss": 0.3666,
|
| 479 |
+
"step": 660
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.700507614213198,
|
| 483 |
+
"grad_norm": 2.004929304122925,
|
| 484 |
+
"learning_rate": 4.1497461928934013e-05,
|
| 485 |
+
"loss": 0.4643,
|
| 486 |
+
"step": 670
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.7258883248730963,
|
| 490 |
+
"grad_norm": 2.2006640434265137,
|
| 491 |
+
"learning_rate": 4.137055837563452e-05,
|
| 492 |
+
"loss": 0.5934,
|
| 493 |
+
"step": 680
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.751269035532995,
|
| 497 |
+
"grad_norm": 1.2337054014205933,
|
| 498 |
+
"learning_rate": 4.1243654822335025e-05,
|
| 499 |
+
"loss": 0.5315,
|
| 500 |
+
"step": 690
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.7766497461928934,
|
| 504 |
+
"grad_norm": 1.163268804550171,
|
| 505 |
+
"learning_rate": 4.111675126903554e-05,
|
| 506 |
+
"loss": 0.4484,
|
| 507 |
+
"step": 700
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.8020304568527918,
|
| 511 |
+
"grad_norm": 1.2907366752624512,
|
| 512 |
+
"learning_rate": 4.098984771573604e-05,
|
| 513 |
+
"loss": 0.4027,
|
| 514 |
+
"step": 710
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.8274111675126905,
|
| 518 |
+
"grad_norm": 1.8477003574371338,
|
| 519 |
+
"learning_rate": 4.086294416243655e-05,
|
| 520 |
+
"loss": 0.503,
|
| 521 |
+
"step": 720
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.8527918781725887,
|
| 525 |
+
"grad_norm": 2.8408010005950928,
|
| 526 |
+
"learning_rate": 4.073604060913706e-05,
|
| 527 |
+
"loss": 0.5602,
|
| 528 |
+
"step": 730
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.8781725888324874,
|
| 532 |
+
"grad_norm": 3.5004968643188477,
|
| 533 |
+
"learning_rate": 4.060913705583756e-05,
|
| 534 |
+
"loss": 0.5681,
|
| 535 |
+
"step": 740
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.9035532994923858,
|
| 539 |
+
"grad_norm": 1.7411813735961914,
|
| 540 |
+
"learning_rate": 4.0482233502538075e-05,
|
| 541 |
+
"loss": 0.4501,
|
| 542 |
+
"step": 750
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.9289340101522843,
|
| 546 |
+
"grad_norm": 4.301488876342773,
|
| 547 |
+
"learning_rate": 4.035532994923858e-05,
|
| 548 |
+
"loss": 0.5262,
|
| 549 |
+
"step": 760
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.9543147208121827,
|
| 553 |
+
"grad_norm": 0.8759877681732178,
|
| 554 |
+
"learning_rate": 4.022842639593909e-05,
|
| 555 |
+
"loss": 0.4527,
|
| 556 |
+
"step": 770
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.9796954314720812,
|
| 560 |
+
"grad_norm": 2.361971855163574,
|
| 561 |
+
"learning_rate": 4.010152284263959e-05,
|
| 562 |
+
"loss": 0.487,
|
| 563 |
+
"step": 780
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 2.0,
|
| 567 |
+
"eval_loss": 0.4712839722633362,
|
| 568 |
+
"eval_runtime": 8.7867,
|
| 569 |
+
"eval_samples_per_second": 89.681,
|
| 570 |
+
"eval_steps_per_second": 11.267,
|
| 571 |
+
"step": 788
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 2.00507614213198,
|
| 575 |
+
"grad_norm": 2.517071485519409,
|
| 576 |
+
"learning_rate": 3.9974619289340106e-05,
|
| 577 |
+
"loss": 0.4159,
|
| 578 |
+
"step": 790
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 2.030456852791878,
|
| 582 |
+
"grad_norm": 1.2710137367248535,
|
| 583 |
+
"learning_rate": 3.9847715736040605e-05,
|
| 584 |
+
"loss": 0.4777,
|
| 585 |
+
"step": 800
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 2.0558375634517767,
|
| 589 |
+
"grad_norm": 3.463624954223633,
|
| 590 |
+
"learning_rate": 3.972081218274112e-05,
|
| 591 |
+
"loss": 0.4792,
|
| 592 |
+
"step": 810
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 2.081218274111675,
|
| 596 |
+
"grad_norm": 1.5746160745620728,
|
| 597 |
+
"learning_rate": 3.959390862944163e-05,
|
| 598 |
+
"loss": 0.4457,
|
| 599 |
+
"step": 820
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 2.1065989847715736,
|
| 603 |
+
"grad_norm": 1.3191742897033691,
|
| 604 |
+
"learning_rate": 3.946700507614213e-05,
|
| 605 |
+
"loss": 0.4615,
|
| 606 |
+
"step": 830
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.1319796954314723,
|
| 610 |
+
"grad_norm": 4.860599040985107,
|
| 611 |
+
"learning_rate": 3.934010152284264e-05,
|
| 612 |
+
"loss": 0.4459,
|
| 613 |
+
"step": 840
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 2.1573604060913705,
|
| 617 |
+
"grad_norm": 2.636868715286255,
|
| 618 |
+
"learning_rate": 3.921319796954315e-05,
|
| 619 |
+
"loss": 0.537,
|
| 620 |
+
"step": 850
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.182741116751269,
|
| 624 |
+
"grad_norm": 1.672642469406128,
|
| 625 |
+
"learning_rate": 3.9086294416243655e-05,
|
| 626 |
+
"loss": 0.4848,
|
| 627 |
+
"step": 860
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 2.2081218274111674,
|
| 631 |
+
"grad_norm": 8.49184799194336,
|
| 632 |
+
"learning_rate": 3.895939086294416e-05,
|
| 633 |
+
"loss": 0.4826,
|
| 634 |
+
"step": 870
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 2.233502538071066,
|
| 638 |
+
"grad_norm": 1.6219210624694824,
|
| 639 |
+
"learning_rate": 3.8832487309644673e-05,
|
| 640 |
+
"loss": 0.3403,
|
| 641 |
+
"step": 880
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 2.2588832487309647,
|
| 645 |
+
"grad_norm": 1.356152892112732,
|
| 646 |
+
"learning_rate": 3.870558375634518e-05,
|
| 647 |
+
"loss": 0.5415,
|
| 648 |
+
"step": 890
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 2.284263959390863,
|
| 652 |
+
"grad_norm": 3.455899238586426,
|
| 653 |
+
"learning_rate": 3.8578680203045685e-05,
|
| 654 |
+
"loss": 0.4958,
|
| 655 |
+
"step": 900
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 2.3096446700507616,
|
| 659 |
+
"grad_norm": 3.994067430496216,
|
| 660 |
+
"learning_rate": 3.84517766497462e-05,
|
| 661 |
+
"loss": 0.4324,
|
| 662 |
+
"step": 910
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 2.33502538071066,
|
| 666 |
+
"grad_norm": 5.774569034576416,
|
| 667 |
+
"learning_rate": 3.83248730964467e-05,
|
| 668 |
+
"loss": 0.3871,
|
| 669 |
+
"step": 920
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 2.3604060913705585,
|
| 673 |
+
"grad_norm": 4.524348258972168,
|
| 674 |
+
"learning_rate": 3.819796954314721e-05,
|
| 675 |
+
"loss": 0.5642,
|
| 676 |
+
"step": 930
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 2.3857868020304567,
|
| 680 |
+
"grad_norm": 1.5101829767227173,
|
| 681 |
+
"learning_rate": 3.8071065989847716e-05,
|
| 682 |
+
"loss": 0.5396,
|
| 683 |
+
"step": 940
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 2.4111675126903553,
|
| 687 |
+
"grad_norm": 1.4824628829956055,
|
| 688 |
+
"learning_rate": 3.794416243654822e-05,
|
| 689 |
+
"loss": 0.5178,
|
| 690 |
+
"step": 950
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 2.436548223350254,
|
| 694 |
+
"grad_norm": 1.3847397565841675,
|
| 695 |
+
"learning_rate": 3.7817258883248735e-05,
|
| 696 |
+
"loss": 0.3747,
|
| 697 |
+
"step": 960
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 2.4619289340101522,
|
| 701 |
+
"grad_norm": 2.138828992843628,
|
| 702 |
+
"learning_rate": 3.769035532994924e-05,
|
| 703 |
+
"loss": 0.5756,
|
| 704 |
+
"step": 970
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 2.487309644670051,
|
| 708 |
+
"grad_norm": 2.0886220932006836,
|
| 709 |
+
"learning_rate": 3.756345177664975e-05,
|
| 710 |
+
"loss": 0.5496,
|
| 711 |
+
"step": 980
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 2.512690355329949,
|
| 715 |
+
"grad_norm": 2.8054535388946533,
|
| 716 |
+
"learning_rate": 3.743654822335025e-05,
|
| 717 |
+
"loss": 0.5984,
|
| 718 |
+
"step": 990
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 2.5380710659898478,
|
| 722 |
+
"grad_norm": 3.6035268306732178,
|
| 723 |
+
"learning_rate": 3.7309644670050766e-05,
|
| 724 |
+
"loss": 0.4475,
|
| 725 |
+
"step": 1000
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 2.563451776649746,
|
| 729 |
+
"grad_norm": 2.931821823120117,
|
| 730 |
+
"learning_rate": 3.7182741116751265e-05,
|
| 731 |
+
"loss": 0.5179,
|
| 732 |
+
"step": 1010
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 2.5888324873096447,
|
| 736 |
+
"grad_norm": 1.8696964979171753,
|
| 737 |
+
"learning_rate": 3.705583756345178e-05,
|
| 738 |
+
"loss": 0.6541,
|
| 739 |
+
"step": 1020
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 2.6142131979695433,
|
| 743 |
+
"grad_norm": 1.36885666847229,
|
| 744 |
+
"learning_rate": 3.692893401015229e-05,
|
| 745 |
+
"loss": 0.3136,
|
| 746 |
+
"step": 1030
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 2.6395939086294415,
|
| 750 |
+
"grad_norm": 3.520601511001587,
|
| 751 |
+
"learning_rate": 3.680203045685279e-05,
|
| 752 |
+
"loss": 0.4434,
|
| 753 |
+
"step": 1040
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 2.66497461928934,
|
| 757 |
+
"grad_norm": 3.95267915725708,
|
| 758 |
+
"learning_rate": 3.66751269035533e-05,
|
| 759 |
+
"loss": 0.5012,
|
| 760 |
+
"step": 1050
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 2.6903553299492384,
|
| 764 |
+
"grad_norm": 2.064523935317993,
|
| 765 |
+
"learning_rate": 3.654822335025381e-05,
|
| 766 |
+
"loss": 0.3251,
|
| 767 |
+
"step": 1060
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 2.715736040609137,
|
| 771 |
+
"grad_norm": 3.3240392208099365,
|
| 772 |
+
"learning_rate": 3.6421319796954315e-05,
|
| 773 |
+
"loss": 0.7704,
|
| 774 |
+
"step": 1070
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 2.7411167512690353,
|
| 778 |
+
"grad_norm": 3.2816076278686523,
|
| 779 |
+
"learning_rate": 3.629441624365482e-05,
|
| 780 |
+
"loss": 0.4753,
|
| 781 |
+
"step": 1080
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 2.766497461928934,
|
| 785 |
+
"grad_norm": 2.7639636993408203,
|
| 786 |
+
"learning_rate": 3.6167512690355334e-05,
|
| 787 |
+
"loss": 0.4443,
|
| 788 |
+
"step": 1090
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 2.7918781725888326,
|
| 792 |
+
"grad_norm": 2.5042450428009033,
|
| 793 |
+
"learning_rate": 3.604060913705584e-05,
|
| 794 |
+
"loss": 0.4979,
|
| 795 |
+
"step": 1100
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 2.817258883248731,
|
| 799 |
+
"grad_norm": 2.695377826690674,
|
| 800 |
+
"learning_rate": 3.5913705583756346e-05,
|
| 801 |
+
"loss": 0.4835,
|
| 802 |
+
"step": 1110
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.8426395939086295,
|
| 806 |
+
"grad_norm": 2.2236552238464355,
|
| 807 |
+
"learning_rate": 3.578680203045686e-05,
|
| 808 |
+
"loss": 0.5162,
|
| 809 |
+
"step": 1120
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 2.868020304568528,
|
| 813 |
+
"grad_norm": 5.009507179260254,
|
| 814 |
+
"learning_rate": 3.565989847715736e-05,
|
| 815 |
+
"loss": 0.4046,
|
| 816 |
+
"step": 1130
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 2.8934010152284264,
|
| 820 |
+
"grad_norm": 2.037398338317871,
|
| 821 |
+
"learning_rate": 3.553299492385787e-05,
|
| 822 |
+
"loss": 0.4502,
|
| 823 |
+
"step": 1140
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 2.9187817258883246,
|
| 827 |
+
"grad_norm": 5.31472635269165,
|
| 828 |
+
"learning_rate": 3.5406091370558376e-05,
|
| 829 |
+
"loss": 0.6371,
|
| 830 |
+
"step": 1150
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 2.9441624365482233,
|
| 834 |
+
"grad_norm": 1.666244387626648,
|
| 835 |
+
"learning_rate": 3.527918781725888e-05,
|
| 836 |
+
"loss": 0.4088,
|
| 837 |
+
"step": 1160
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 2.969543147208122,
|
| 841 |
+
"grad_norm": 4.006988525390625,
|
| 842 |
+
"learning_rate": 3.5152284263959395e-05,
|
| 843 |
+
"loss": 0.4483,
|
| 844 |
+
"step": 1170
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 2.99492385786802,
|
| 848 |
+
"grad_norm": 2.429619312286377,
|
| 849 |
+
"learning_rate": 3.50253807106599e-05,
|
| 850 |
+
"loss": 0.4617,
|
| 851 |
+
"step": 1180
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 3.0,
|
| 855 |
+
"eval_loss": 0.5104432106018066,
|
| 856 |
+
"eval_runtime": 8.8793,
|
| 857 |
+
"eval_samples_per_second": 88.746,
|
| 858 |
+
"eval_steps_per_second": 11.15,
|
| 859 |
+
"step": 1182
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 3.020304568527919,
|
| 863 |
+
"grad_norm": 1.9103385210037231,
|
| 864 |
+
"learning_rate": 3.489847715736041e-05,
|
| 865 |
+
"loss": 0.461,
|
| 866 |
+
"step": 1190
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 3.045685279187817,
|
| 870 |
+
"grad_norm": 2.838453769683838,
|
| 871 |
+
"learning_rate": 3.477157360406091e-05,
|
| 872 |
+
"loss": 0.4661,
|
| 873 |
+
"step": 1200
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 3.0710659898477157,
|
| 877 |
+
"grad_norm": 2.706493616104126,
|
| 878 |
+
"learning_rate": 3.4644670050761426e-05,
|
| 879 |
+
"loss": 0.4335,
|
| 880 |
+
"step": 1210
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 3.0964467005076144,
|
| 884 |
+
"grad_norm": 4.179074287414551,
|
| 885 |
+
"learning_rate": 3.451776649746193e-05,
|
| 886 |
+
"loss": 0.3517,
|
| 887 |
+
"step": 1220
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 3.1218274111675126,
|
| 891 |
+
"grad_norm": 1.1146340370178223,
|
| 892 |
+
"learning_rate": 3.439086294416244e-05,
|
| 893 |
+
"loss": 0.2261,
|
| 894 |
+
"step": 1230
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 3.1472081218274113,
|
| 898 |
+
"grad_norm": 2.5486109256744385,
|
| 899 |
+
"learning_rate": 3.4263959390862944e-05,
|
| 900 |
+
"loss": 0.3664,
|
| 901 |
+
"step": 1240
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 3.1725888324873095,
|
| 905 |
+
"grad_norm": 2.886986017227173,
|
| 906 |
+
"learning_rate": 3.413705583756345e-05,
|
| 907 |
+
"loss": 0.3861,
|
| 908 |
+
"step": 1250
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 3.197969543147208,
|
| 912 |
+
"grad_norm": 3.5457489490509033,
|
| 913 |
+
"learning_rate": 3.401015228426396e-05,
|
| 914 |
+
"loss": 0.5043,
|
| 915 |
+
"step": 1260
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 3.223350253807107,
|
| 919 |
+
"grad_norm": 4.039999485015869,
|
| 920 |
+
"learning_rate": 3.388324873096447e-05,
|
| 921 |
+
"loss": 0.3362,
|
| 922 |
+
"step": 1270
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 3.248730964467005,
|
| 926 |
+
"grad_norm": 4.554147720336914,
|
| 927 |
+
"learning_rate": 3.3756345177664975e-05,
|
| 928 |
+
"loss": 0.4709,
|
| 929 |
+
"step": 1280
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 3.2741116751269037,
|
| 933 |
+
"grad_norm": 2.4214282035827637,
|
| 934 |
+
"learning_rate": 3.362944162436548e-05,
|
| 935 |
+
"loss": 0.3041,
|
| 936 |
+
"step": 1290
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 3.299492385786802,
|
| 940 |
+
"grad_norm": 4.587721824645996,
|
| 941 |
+
"learning_rate": 3.3502538071065994e-05,
|
| 942 |
+
"loss": 0.6426,
|
| 943 |
+
"step": 1300
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 3.3248730964467006,
|
| 947 |
+
"grad_norm": 2.0584123134613037,
|
| 948 |
+
"learning_rate": 3.33756345177665e-05,
|
| 949 |
+
"loss": 0.3861,
|
| 950 |
+
"step": 1310
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 3.350253807106599,
|
| 954 |
+
"grad_norm": 2.152904510498047,
|
| 955 |
+
"learning_rate": 3.3248730964467006e-05,
|
| 956 |
+
"loss": 0.434,
|
| 957 |
+
"step": 1320
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 3.3756345177664975,
|
| 961 |
+
"grad_norm": 3.3086836338043213,
|
| 962 |
+
"learning_rate": 3.312182741116752e-05,
|
| 963 |
+
"loss": 0.406,
|
| 964 |
+
"step": 1330
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 3.401015228426396,
|
| 968 |
+
"grad_norm": 2.4695167541503906,
|
| 969 |
+
"learning_rate": 3.299492385786802e-05,
|
| 970 |
+
"loss": 0.4226,
|
| 971 |
+
"step": 1340
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 3.4263959390862944,
|
| 975 |
+
"grad_norm": 2.482093334197998,
|
| 976 |
+
"learning_rate": 3.286802030456853e-05,
|
| 977 |
+
"loss": 0.335,
|
| 978 |
+
"step": 1350
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 3.451776649746193,
|
| 982 |
+
"grad_norm": 6.9000959396362305,
|
| 983 |
+
"learning_rate": 3.2741116751269036e-05,
|
| 984 |
+
"loss": 0.3506,
|
| 985 |
+
"step": 1360
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 3.4771573604060912,
|
| 989 |
+
"grad_norm": 2.6196579933166504,
|
| 990 |
+
"learning_rate": 3.261421319796954e-05,
|
| 991 |
+
"loss": 0.3377,
|
| 992 |
+
"step": 1370
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 3.50253807106599,
|
| 996 |
+
"grad_norm": 1.8537606000900269,
|
| 997 |
+
"learning_rate": 3.248730964467005e-05,
|
| 998 |
+
"loss": 0.399,
|
| 999 |
+
"step": 1380
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 3.527918781725888,
|
| 1003 |
+
"grad_norm": 7.022123336791992,
|
| 1004 |
+
"learning_rate": 3.236040609137056e-05,
|
| 1005 |
+
"loss": 0.4255,
|
| 1006 |
+
"step": 1390
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 3.553299492385787,
|
| 1010 |
+
"grad_norm": 3.112264394760132,
|
| 1011 |
+
"learning_rate": 3.223350253807107e-05,
|
| 1012 |
+
"loss": 0.399,
|
| 1013 |
+
"step": 1400
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 3.5786802030456855,
|
| 1017 |
+
"grad_norm": 4.884408473968506,
|
| 1018 |
+
"learning_rate": 3.210659898477157e-05,
|
| 1019 |
+
"loss": 0.3533,
|
| 1020 |
+
"step": 1410
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 3.6040609137055837,
|
| 1024 |
+
"grad_norm": 3.7339515686035156,
|
| 1025 |
+
"learning_rate": 3.1979695431472086e-05,
|
| 1026 |
+
"loss": 0.4329,
|
| 1027 |
+
"step": 1420
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 3.6294416243654823,
|
| 1031 |
+
"grad_norm": 11.797500610351562,
|
| 1032 |
+
"learning_rate": 3.185279187817259e-05,
|
| 1033 |
+
"loss": 0.4496,
|
| 1034 |
+
"step": 1430
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 3.6548223350253806,
|
| 1038 |
+
"grad_norm": 2.509971857070923,
|
| 1039 |
+
"learning_rate": 3.17258883248731e-05,
|
| 1040 |
+
"loss": 0.3758,
|
| 1041 |
+
"step": 1440
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 3.6802030456852792,
|
| 1045 |
+
"grad_norm": 2.554452419281006,
|
| 1046 |
+
"learning_rate": 3.1598984771573604e-05,
|
| 1047 |
+
"loss": 0.2789,
|
| 1048 |
+
"step": 1450
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 3.7055837563451774,
|
| 1052 |
+
"grad_norm": 13.046540260314941,
|
| 1053 |
+
"learning_rate": 3.147208121827411e-05,
|
| 1054 |
+
"loss": 0.52,
|
| 1055 |
+
"step": 1460
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 3.730964467005076,
|
| 1059 |
+
"grad_norm": 6.821984767913818,
|
| 1060 |
+
"learning_rate": 3.134517766497462e-05,
|
| 1061 |
+
"loss": 0.4285,
|
| 1062 |
+
"step": 1470
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 3.7563451776649748,
|
| 1066 |
+
"grad_norm": 2.692779064178467,
|
| 1067 |
+
"learning_rate": 3.121827411167513e-05,
|
| 1068 |
+
"loss": 0.2977,
|
| 1069 |
+
"step": 1480
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 3.781725888324873,
|
| 1073 |
+
"grad_norm": 3.510523796081543,
|
| 1074 |
+
"learning_rate": 3.1091370558375635e-05,
|
| 1075 |
+
"loss": 0.4887,
|
| 1076 |
+
"step": 1490
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 3.8071065989847717,
|
| 1080 |
+
"grad_norm": 4.374626159667969,
|
| 1081 |
+
"learning_rate": 3.096446700507614e-05,
|
| 1082 |
+
"loss": 0.2093,
|
| 1083 |
+
"step": 1500
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 3.8324873096446703,
|
| 1087 |
+
"grad_norm": 0.6560042500495911,
|
| 1088 |
+
"learning_rate": 3.0837563451776654e-05,
|
| 1089 |
+
"loss": 0.613,
|
| 1090 |
+
"step": 1510
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 3.8578680203045685,
|
| 1094 |
+
"grad_norm": 4.9201483726501465,
|
| 1095 |
+
"learning_rate": 3.071065989847716e-05,
|
| 1096 |
+
"loss": 0.4483,
|
| 1097 |
+
"step": 1520
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 3.8832487309644668,
|
| 1101 |
+
"grad_norm": 2.497208833694458,
|
| 1102 |
+
"learning_rate": 3.0583756345177666e-05,
|
| 1103 |
+
"loss": 0.4434,
|
| 1104 |
+
"step": 1530
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 3.9086294416243654,
|
| 1108 |
+
"grad_norm": 5.912675380706787,
|
| 1109 |
+
"learning_rate": 3.0456852791878175e-05,
|
| 1110 |
+
"loss": 0.4259,
|
| 1111 |
+
"step": 1540
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 3.934010152284264,
|
| 1115 |
+
"grad_norm": 1.8950650691986084,
|
| 1116 |
+
"learning_rate": 3.032994923857868e-05,
|
| 1117 |
+
"loss": 0.4649,
|
| 1118 |
+
"step": 1550
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 3.9593908629441623,
|
| 1122 |
+
"grad_norm": 15.265602111816406,
|
| 1123 |
+
"learning_rate": 3.020304568527919e-05,
|
| 1124 |
+
"loss": 0.4176,
|
| 1125 |
+
"step": 1560
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 3.984771573604061,
|
| 1129 |
+
"grad_norm": 6.062721252441406,
|
| 1130 |
+
"learning_rate": 3.0076142131979696e-05,
|
| 1131 |
+
"loss": 0.3854,
|
| 1132 |
+
"step": 1570
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 4.0,
|
| 1136 |
+
"eval_loss": 0.5223032832145691,
|
| 1137 |
+
"eval_runtime": 8.7536,
|
| 1138 |
+
"eval_samples_per_second": 90.02,
|
| 1139 |
+
"eval_steps_per_second": 11.31,
|
| 1140 |
+
"step": 1576
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 4.01015228426396,
|
| 1144 |
+
"grad_norm": 4.975412368774414,
|
| 1145 |
+
"learning_rate": 2.9949238578680206e-05,
|
| 1146 |
+
"loss": 0.4798,
|
| 1147 |
+
"step": 1580
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 4.035532994923858,
|
| 1151 |
+
"grad_norm": 5.022281646728516,
|
| 1152 |
+
"learning_rate": 2.982233502538071e-05,
|
| 1153 |
+
"loss": 0.3322,
|
| 1154 |
+
"step": 1590
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 4.060913705583756,
|
| 1158 |
+
"grad_norm": 5.229846000671387,
|
| 1159 |
+
"learning_rate": 2.969543147208122e-05,
|
| 1160 |
+
"loss": 0.2691,
|
| 1161 |
+
"step": 1600
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 4.086294416243655,
|
| 1165 |
+
"grad_norm": 21.07980728149414,
|
| 1166 |
+
"learning_rate": 2.956852791878173e-05,
|
| 1167 |
+
"loss": 0.285,
|
| 1168 |
+
"step": 1610
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 4.111675126903553,
|
| 1172 |
+
"grad_norm": 28.23366355895996,
|
| 1173 |
+
"learning_rate": 2.9441624365482233e-05,
|
| 1174 |
+
"loss": 0.1946,
|
| 1175 |
+
"step": 1620
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 4.137055837563452,
|
| 1179 |
+
"grad_norm": 0.49203088879585266,
|
| 1180 |
+
"learning_rate": 2.9314720812182743e-05,
|
| 1181 |
+
"loss": 0.2176,
|
| 1182 |
+
"step": 1630
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 4.16243654822335,
|
| 1186 |
+
"grad_norm": 6.196091175079346,
|
| 1187 |
+
"learning_rate": 2.918781725888325e-05,
|
| 1188 |
+
"loss": 0.1592,
|
| 1189 |
+
"step": 1640
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 4.187817258883249,
|
| 1193 |
+
"grad_norm": 0.43765848875045776,
|
| 1194 |
+
"learning_rate": 2.9060913705583758e-05,
|
| 1195 |
+
"loss": 0.3764,
|
| 1196 |
+
"step": 1650
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 4.213197969543147,
|
| 1200 |
+
"grad_norm": 4.543421745300293,
|
| 1201 |
+
"learning_rate": 2.8934010152284264e-05,
|
| 1202 |
+
"loss": 0.3371,
|
| 1203 |
+
"step": 1660
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 4.238578680203045,
|
| 1207 |
+
"grad_norm": 2.611579418182373,
|
| 1208 |
+
"learning_rate": 2.8807106598984774e-05,
|
| 1209 |
+
"loss": 0.3052,
|
| 1210 |
+
"step": 1670
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 4.2639593908629445,
|
| 1214 |
+
"grad_norm": 3.3673095703125,
|
| 1215 |
+
"learning_rate": 2.8680203045685283e-05,
|
| 1216 |
+
"loss": 0.1776,
|
| 1217 |
+
"step": 1680
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 4.289340101522843,
|
| 1221 |
+
"grad_norm": 0.7387822270393372,
|
| 1222 |
+
"learning_rate": 2.855329949238579e-05,
|
| 1223 |
+
"loss": 0.3827,
|
| 1224 |
+
"step": 1690
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 4.314720812182741,
|
| 1228 |
+
"grad_norm": 16.637300491333008,
|
| 1229 |
+
"learning_rate": 2.84263959390863e-05,
|
| 1230 |
+
"loss": 0.3993,
|
| 1231 |
+
"step": 1700
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 4.340101522842639,
|
| 1235 |
+
"grad_norm": 6.28633975982666,
|
| 1236 |
+
"learning_rate": 2.82994923857868e-05,
|
| 1237 |
+
"loss": 0.3733,
|
| 1238 |
+
"step": 1710
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 4.365482233502538,
|
| 1242 |
+
"grad_norm": 3.3295745849609375,
|
| 1243 |
+
"learning_rate": 2.8172588832487314e-05,
|
| 1244 |
+
"loss": 0.1166,
|
| 1245 |
+
"step": 1720
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 4.3908629441624365,
|
| 1249 |
+
"grad_norm": 14.652837753295898,
|
| 1250 |
+
"learning_rate": 2.8045685279187816e-05,
|
| 1251 |
+
"loss": 0.3326,
|
| 1252 |
+
"step": 1730
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 4.416243654822335,
|
| 1256 |
+
"grad_norm": 31.232574462890625,
|
| 1257 |
+
"learning_rate": 2.7918781725888326e-05,
|
| 1258 |
+
"loss": 0.2127,
|
| 1259 |
+
"step": 1740
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 4.441624365482234,
|
| 1263 |
+
"grad_norm": 12.528191566467285,
|
| 1264 |
+
"learning_rate": 2.7791878172588832e-05,
|
| 1265 |
+
"loss": 0.4061,
|
| 1266 |
+
"step": 1750
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 4.467005076142132,
|
| 1270 |
+
"grad_norm": 1.7689141035079956,
|
| 1271 |
+
"learning_rate": 2.766497461928934e-05,
|
| 1272 |
+
"loss": 0.3366,
|
| 1273 |
+
"step": 1760
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 4.49238578680203,
|
| 1277 |
+
"grad_norm": 4.724018096923828,
|
| 1278 |
+
"learning_rate": 2.753807106598985e-05,
|
| 1279 |
+
"loss": 0.2359,
|
| 1280 |
+
"step": 1770
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 4.517766497461929,
|
| 1284 |
+
"grad_norm": 1.561919093132019,
|
| 1285 |
+
"learning_rate": 2.7411167512690357e-05,
|
| 1286 |
+
"loss": 0.3106,
|
| 1287 |
+
"step": 1780
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 4.543147208121828,
|
| 1291 |
+
"grad_norm": 5.754579067230225,
|
| 1292 |
+
"learning_rate": 2.7284263959390866e-05,
|
| 1293 |
+
"loss": 0.2982,
|
| 1294 |
+
"step": 1790
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 4.568527918781726,
|
| 1298 |
+
"grad_norm": 1.181515097618103,
|
| 1299 |
+
"learning_rate": 2.715736040609137e-05,
|
| 1300 |
+
"loss": 0.2384,
|
| 1301 |
+
"step": 1800
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 4.593908629441624,
|
| 1305 |
+
"grad_norm": 7.012800693511963,
|
| 1306 |
+
"learning_rate": 2.703045685279188e-05,
|
| 1307 |
+
"loss": 0.394,
|
| 1308 |
+
"step": 1810
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 4.619289340101523,
|
| 1312 |
+
"grad_norm": 0.49557140469551086,
|
| 1313 |
+
"learning_rate": 2.6903553299492384e-05,
|
| 1314 |
+
"loss": 0.2869,
|
| 1315 |
+
"step": 1820
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 4.644670050761421,
|
| 1319 |
+
"grad_norm": 3.6110830307006836,
|
| 1320 |
+
"learning_rate": 2.6776649746192893e-05,
|
| 1321 |
+
"loss": 0.3724,
|
| 1322 |
+
"step": 1830
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 4.67005076142132,
|
| 1326 |
+
"grad_norm": 3.3351943492889404,
|
| 1327 |
+
"learning_rate": 2.6649746192893406e-05,
|
| 1328 |
+
"loss": 0.2966,
|
| 1329 |
+
"step": 1840
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 4.695431472081218,
|
| 1333 |
+
"grad_norm": 1.342678427696228,
|
| 1334 |
+
"learning_rate": 2.652284263959391e-05,
|
| 1335 |
+
"loss": 0.3826,
|
| 1336 |
+
"step": 1850
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 4.720812182741117,
|
| 1340 |
+
"grad_norm": 0.4344848096370697,
|
| 1341 |
+
"learning_rate": 2.6395939086294418e-05,
|
| 1342 |
+
"loss": 0.1962,
|
| 1343 |
+
"step": 1860
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 4.746192893401015,
|
| 1347 |
+
"grad_norm": 56.233985900878906,
|
| 1348 |
+
"learning_rate": 2.6269035532994924e-05,
|
| 1349 |
+
"loss": 0.3758,
|
| 1350 |
+
"step": 1870
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 4.771573604060913,
|
| 1354 |
+
"grad_norm": 10.647058486938477,
|
| 1355 |
+
"learning_rate": 2.6142131979695434e-05,
|
| 1356 |
+
"loss": 0.356,
|
| 1357 |
+
"step": 1880
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 4.7969543147208125,
|
| 1361 |
+
"grad_norm": 0.9422564506530762,
|
| 1362 |
+
"learning_rate": 2.6015228426395936e-05,
|
| 1363 |
+
"loss": 0.2271,
|
| 1364 |
+
"step": 1890
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 4.822335025380711,
|
| 1368 |
+
"grad_norm": 0.22354674339294434,
|
| 1369 |
+
"learning_rate": 2.588832487309645e-05,
|
| 1370 |
+
"loss": 0.1374,
|
| 1371 |
+
"step": 1900
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 4.847715736040609,
|
| 1375 |
+
"grad_norm": 8.779372215270996,
|
| 1376 |
+
"learning_rate": 2.576142131979696e-05,
|
| 1377 |
+
"loss": 0.329,
|
| 1378 |
+
"step": 1910
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 4.873096446700508,
|
| 1382 |
+
"grad_norm": 41.76162338256836,
|
| 1383 |
+
"learning_rate": 2.563451776649746e-05,
|
| 1384 |
+
"loss": 0.5914,
|
| 1385 |
+
"step": 1920
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 4.898477157360406,
|
| 1389 |
+
"grad_norm": 26.878578186035156,
|
| 1390 |
+
"learning_rate": 2.5507614213197974e-05,
|
| 1391 |
+
"loss": 0.5029,
|
| 1392 |
+
"step": 1930
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 4.9238578680203045,
|
| 1396 |
+
"grad_norm": 0.41534122824668884,
|
| 1397 |
+
"learning_rate": 2.5380710659898476e-05,
|
| 1398 |
+
"loss": 0.1997,
|
| 1399 |
+
"step": 1940
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 4.949238578680203,
|
| 1403 |
+
"grad_norm": 4.708491325378418,
|
| 1404 |
+
"learning_rate": 2.5253807106598986e-05,
|
| 1405 |
+
"loss": 0.3183,
|
| 1406 |
+
"step": 1950
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 4.974619289340102,
|
| 1410 |
+
"grad_norm": 0.31694722175598145,
|
| 1411 |
+
"learning_rate": 2.5126903553299492e-05,
|
| 1412 |
+
"loss": 0.3627,
|
| 1413 |
+
"step": 1960
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 5.0,
|
| 1417 |
+
"grad_norm": 2.7903828620910645,
|
| 1418 |
+
"learning_rate": 2.5e-05,
|
| 1419 |
+
"loss": 0.454,
|
| 1420 |
+
"step": 1970
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 5.0,
|
| 1424 |
+
"eval_loss": 0.6968420743942261,
|
| 1425 |
+
"eval_runtime": 8.7659,
|
| 1426 |
+
"eval_samples_per_second": 89.894,
|
| 1427 |
+
"eval_steps_per_second": 11.294,
|
| 1428 |
+
"step": 1970
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 5.025380710659898,
|
| 1432 |
+
"grad_norm": 0.33573099970817566,
|
| 1433 |
+
"learning_rate": 2.4873096446700507e-05,
|
| 1434 |
+
"loss": 0.0922,
|
| 1435 |
+
"step": 1980
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 5.050761421319797,
|
| 1439 |
+
"grad_norm": 0.38098305463790894,
|
| 1440 |
+
"learning_rate": 2.4746192893401017e-05,
|
| 1441 |
+
"loss": 0.1296,
|
| 1442 |
+
"step": 1990
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 5.0761421319796955,
|
| 1446 |
+
"grad_norm": 6.538486480712891,
|
| 1447 |
+
"learning_rate": 2.4619289340101523e-05,
|
| 1448 |
+
"loss": 0.2388,
|
| 1449 |
+
"step": 2000
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 5.101522842639594,
|
| 1453 |
+
"grad_norm": 0.3389257788658142,
|
| 1454 |
+
"learning_rate": 2.4492385786802032e-05,
|
| 1455 |
+
"loss": 0.1525,
|
| 1456 |
+
"step": 2010
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 5.126903553299492,
|
| 1460 |
+
"grad_norm": 29.134950637817383,
|
| 1461 |
+
"learning_rate": 2.436548223350254e-05,
|
| 1462 |
+
"loss": 0.2023,
|
| 1463 |
+
"step": 2020
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 5.152284263959391,
|
| 1467 |
+
"grad_norm": 2.5168628692626953,
|
| 1468 |
+
"learning_rate": 2.4238578680203047e-05,
|
| 1469 |
+
"loss": 0.1126,
|
| 1470 |
+
"step": 2030
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 5.177664974619289,
|
| 1474 |
+
"grad_norm": 0.4162174463272095,
|
| 1475 |
+
"learning_rate": 2.4111675126903553e-05,
|
| 1476 |
+
"loss": 0.1466,
|
| 1477 |
+
"step": 2040
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 5.2030456852791875,
|
| 1481 |
+
"grad_norm": 0.08261796087026596,
|
| 1482 |
+
"learning_rate": 2.3984771573604063e-05,
|
| 1483 |
+
"loss": 0.1032,
|
| 1484 |
+
"step": 2050
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 5.228426395939087,
|
| 1488 |
+
"grad_norm": 0.13249030709266663,
|
| 1489 |
+
"learning_rate": 2.385786802030457e-05,
|
| 1490 |
+
"loss": 0.0117,
|
| 1491 |
+
"step": 2060
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 5.253807106598985,
|
| 1495 |
+
"grad_norm": 54.72965621948242,
|
| 1496 |
+
"learning_rate": 2.3730964467005075e-05,
|
| 1497 |
+
"loss": 0.1383,
|
| 1498 |
+
"step": 2070
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 5.279187817258883,
|
| 1502 |
+
"grad_norm": 10.03510856628418,
|
| 1503 |
+
"learning_rate": 2.3604060913705588e-05,
|
| 1504 |
+
"loss": 0.3034,
|
| 1505 |
+
"step": 2080
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 5.304568527918782,
|
| 1509 |
+
"grad_norm": 3.790813446044922,
|
| 1510 |
+
"learning_rate": 2.3477157360406094e-05,
|
| 1511 |
+
"loss": 0.0487,
|
| 1512 |
+
"step": 2090
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 5.32994923857868,
|
| 1516 |
+
"grad_norm": 10.79232120513916,
|
| 1517 |
+
"learning_rate": 2.33502538071066e-05,
|
| 1518 |
+
"loss": 0.1897,
|
| 1519 |
+
"step": 2100
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 5.355329949238579,
|
| 1523 |
+
"grad_norm": 3.8837833404541016,
|
| 1524 |
+
"learning_rate": 2.322335025380711e-05,
|
| 1525 |
+
"loss": 0.5641,
|
| 1526 |
+
"step": 2110
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 5.380710659898477,
|
| 1530 |
+
"grad_norm": 0.5432278513908386,
|
| 1531 |
+
"learning_rate": 2.3096446700507615e-05,
|
| 1532 |
+
"loss": 0.3232,
|
| 1533 |
+
"step": 2120
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 5.406091370558376,
|
| 1537 |
+
"grad_norm": 2.8115179538726807,
|
| 1538 |
+
"learning_rate": 2.296954314720812e-05,
|
| 1539 |
+
"loss": 0.2467,
|
| 1540 |
+
"step": 2130
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 5.431472081218274,
|
| 1544 |
+
"grad_norm": 0.23173533380031586,
|
| 1545 |
+
"learning_rate": 2.284263959390863e-05,
|
| 1546 |
+
"loss": 0.0366,
|
| 1547 |
+
"step": 2140
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 5.456852791878172,
|
| 1551 |
+
"grad_norm": 0.22747744619846344,
|
| 1552 |
+
"learning_rate": 2.2715736040609136e-05,
|
| 1553 |
+
"loss": 0.2833,
|
| 1554 |
+
"step": 2150
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 5.482233502538071,
|
| 1558 |
+
"grad_norm": 0.18748463690280914,
|
| 1559 |
+
"learning_rate": 2.2588832487309646e-05,
|
| 1560 |
+
"loss": 0.2338,
|
| 1561 |
+
"step": 2160
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 5.50761421319797,
|
| 1565 |
+
"grad_norm": 0.2156940996646881,
|
| 1566 |
+
"learning_rate": 2.2461928934010155e-05,
|
| 1567 |
+
"loss": 0.0311,
|
| 1568 |
+
"step": 2170
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 5.532994923857868,
|
| 1572 |
+
"grad_norm": 26.930912017822266,
|
| 1573 |
+
"learning_rate": 2.233502538071066e-05,
|
| 1574 |
+
"loss": 0.2466,
|
| 1575 |
+
"step": 2180
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 5.558375634517766,
|
| 1579 |
+
"grad_norm": 23.53707504272461,
|
| 1580 |
+
"learning_rate": 2.2208121827411167e-05,
|
| 1581 |
+
"loss": 0.2277,
|
| 1582 |
+
"step": 2190
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 5.583756345177665,
|
| 1586 |
+
"grad_norm": 6.032989025115967,
|
| 1587 |
+
"learning_rate": 2.2081218274111677e-05,
|
| 1588 |
+
"loss": 0.1784,
|
| 1589 |
+
"step": 2200
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 5.6091370558375635,
|
| 1593 |
+
"grad_norm": 15.71194076538086,
|
| 1594 |
+
"learning_rate": 2.1954314720812183e-05,
|
| 1595 |
+
"loss": 0.2382,
|
| 1596 |
+
"step": 2210
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 5.634517766497462,
|
| 1600 |
+
"grad_norm": 6.757187366485596,
|
| 1601 |
+
"learning_rate": 2.182741116751269e-05,
|
| 1602 |
+
"loss": 0.2233,
|
| 1603 |
+
"step": 2220
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 5.659898477157361,
|
| 1607 |
+
"grad_norm": 6.697112560272217,
|
| 1608 |
+
"learning_rate": 2.17005076142132e-05,
|
| 1609 |
+
"loss": 0.2094,
|
| 1610 |
+
"step": 2230
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 5.685279187817259,
|
| 1614 |
+
"grad_norm": 12.990015983581543,
|
| 1615 |
+
"learning_rate": 2.1573604060913707e-05,
|
| 1616 |
+
"loss": 0.167,
|
| 1617 |
+
"step": 2240
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 5.710659898477157,
|
| 1621 |
+
"grad_norm": 0.234835684299469,
|
| 1622 |
+
"learning_rate": 2.1446700507614213e-05,
|
| 1623 |
+
"loss": 0.1446,
|
| 1624 |
+
"step": 2250
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 5.7360406091370555,
|
| 1628 |
+
"grad_norm": 6.382123947143555,
|
| 1629 |
+
"learning_rate": 2.1319796954314723e-05,
|
| 1630 |
+
"loss": 0.2736,
|
| 1631 |
+
"step": 2260
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 5.761421319796955,
|
| 1635 |
+
"grad_norm": 0.26427528262138367,
|
| 1636 |
+
"learning_rate": 2.119289340101523e-05,
|
| 1637 |
+
"loss": 0.1443,
|
| 1638 |
+
"step": 2270
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 5.786802030456853,
|
| 1642 |
+
"grad_norm": 0.2004345953464508,
|
| 1643 |
+
"learning_rate": 2.1065989847715735e-05,
|
| 1644 |
+
"loss": 0.1685,
|
| 1645 |
+
"step": 2280
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 5.812182741116751,
|
| 1649 |
+
"grad_norm": 0.21756359934806824,
|
| 1650 |
+
"learning_rate": 2.0939086294416244e-05,
|
| 1651 |
+
"loss": 0.385,
|
| 1652 |
+
"step": 2290
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 5.837563451776649,
|
| 1656 |
+
"grad_norm": 5.085987567901611,
|
| 1657 |
+
"learning_rate": 2.0812182741116754e-05,
|
| 1658 |
+
"loss": 0.233,
|
| 1659 |
+
"step": 2300
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 5.862944162436548,
|
| 1663 |
+
"grad_norm": 0.2375006377696991,
|
| 1664 |
+
"learning_rate": 2.068527918781726e-05,
|
| 1665 |
+
"loss": 0.2106,
|
| 1666 |
+
"step": 2310
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 5.888324873096447,
|
| 1670 |
+
"grad_norm": 11.780427932739258,
|
| 1671 |
+
"learning_rate": 2.055837563451777e-05,
|
| 1672 |
+
"loss": 0.2142,
|
| 1673 |
+
"step": 2320
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 5.913705583756345,
|
| 1677 |
+
"grad_norm": 1.2173970937728882,
|
| 1678 |
+
"learning_rate": 2.0431472081218275e-05,
|
| 1679 |
+
"loss": 0.0754,
|
| 1680 |
+
"step": 2330
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 5.939086294416244,
|
| 1684 |
+
"grad_norm": 6.028761386871338,
|
| 1685 |
+
"learning_rate": 2.030456852791878e-05,
|
| 1686 |
+
"loss": 0.4157,
|
| 1687 |
+
"step": 2340
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 5.964467005076142,
|
| 1691 |
+
"grad_norm": 19.581605911254883,
|
| 1692 |
+
"learning_rate": 2.017766497461929e-05,
|
| 1693 |
+
"loss": 0.1866,
|
| 1694 |
+
"step": 2350
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 5.98984771573604,
|
| 1698 |
+
"grad_norm": 0.22685207426548004,
|
| 1699 |
+
"learning_rate": 2.0050761421319797e-05,
|
| 1700 |
+
"loss": 0.2823,
|
| 1701 |
+
"step": 2360
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 6.0,
|
| 1705 |
+
"eval_loss": 0.9028552770614624,
|
| 1706 |
+
"eval_runtime": 8.7725,
|
| 1707 |
+
"eval_samples_per_second": 89.826,
|
| 1708 |
+
"eval_steps_per_second": 11.285,
|
| 1709 |
+
"step": 2364
|
| 1710 |
+
},
|
| 1711 |
+
{
|
| 1712 |
+
"epoch": 6.0152284263959395,
|
| 1713 |
+
"grad_norm": 0.280078649520874,
|
| 1714 |
+
"learning_rate": 1.9923857868020303e-05,
|
| 1715 |
+
"loss": 0.1616,
|
| 1716 |
+
"step": 2370
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 6.040609137055838,
|
| 1720 |
+
"grad_norm": 0.17733518779277802,
|
| 1721 |
+
"learning_rate": 1.9796954314720815e-05,
|
| 1722 |
+
"loss": 0.0129,
|
| 1723 |
+
"step": 2380
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"epoch": 6.065989847715736,
|
| 1727 |
+
"grad_norm": 4.652949333190918,
|
| 1728 |
+
"learning_rate": 1.967005076142132e-05,
|
| 1729 |
+
"loss": 0.2607,
|
| 1730 |
+
"step": 2390
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 6.091370558375634,
|
| 1734 |
+
"grad_norm": 17.740243911743164,
|
| 1735 |
+
"learning_rate": 1.9543147208121827e-05,
|
| 1736 |
+
"loss": 0.2297,
|
| 1737 |
+
"step": 2400
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 6.116751269035533,
|
| 1741 |
+
"grad_norm": 59.82321548461914,
|
| 1742 |
+
"learning_rate": 1.9416243654822337e-05,
|
| 1743 |
+
"loss": 0.2063,
|
| 1744 |
+
"step": 2410
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 6.1421319796954315,
|
| 1748 |
+
"grad_norm": 0.1367008090019226,
|
| 1749 |
+
"learning_rate": 1.9289340101522843e-05,
|
| 1750 |
+
"loss": 0.1211,
|
| 1751 |
+
"step": 2420
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 6.16751269035533,
|
| 1755 |
+
"grad_norm": 0.14475467801094055,
|
| 1756 |
+
"learning_rate": 1.916243654822335e-05,
|
| 1757 |
+
"loss": 0.0078,
|
| 1758 |
+
"step": 2430
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 6.192893401015229,
|
| 1762 |
+
"grad_norm": 14.927183151245117,
|
| 1763 |
+
"learning_rate": 1.9035532994923858e-05,
|
| 1764 |
+
"loss": 0.1697,
|
| 1765 |
+
"step": 2440
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 6.218274111675127,
|
| 1769 |
+
"grad_norm": 0.11621695011854172,
|
| 1770 |
+
"learning_rate": 1.8908629441624368e-05,
|
| 1771 |
+
"loss": 0.0255,
|
| 1772 |
+
"step": 2450
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 6.243654822335025,
|
| 1776 |
+
"grad_norm": 0.12786021828651428,
|
| 1777 |
+
"learning_rate": 1.8781725888324874e-05,
|
| 1778 |
+
"loss": 0.008,
|
| 1779 |
+
"step": 2460
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 6.269035532994923,
|
| 1783 |
+
"grad_norm": 0.07750383019447327,
|
| 1784 |
+
"learning_rate": 1.8654822335025383e-05,
|
| 1785 |
+
"loss": 0.21,
|
| 1786 |
+
"step": 2470
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 6.2944162436548226,
|
| 1790 |
+
"grad_norm": 0.11915043741464615,
|
| 1791 |
+
"learning_rate": 1.852791878172589e-05,
|
| 1792 |
+
"loss": 0.0661,
|
| 1793 |
+
"step": 2480
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 6.319796954314721,
|
| 1797 |
+
"grad_norm": 0.09122921526432037,
|
| 1798 |
+
"learning_rate": 1.8401015228426395e-05,
|
| 1799 |
+
"loss": 0.0079,
|
| 1800 |
+
"step": 2490
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 6.345177664974619,
|
| 1804 |
+
"grad_norm": 0.08648554980754852,
|
| 1805 |
+
"learning_rate": 1.8274111675126904e-05,
|
| 1806 |
+
"loss": 0.121,
|
| 1807 |
+
"step": 2500
|
| 1808 |
+
},
|
| 1809 |
+
{
|
| 1810 |
+
"epoch": 6.370558375634518,
|
| 1811 |
+
"grad_norm": 0.13753150403499603,
|
| 1812 |
+
"learning_rate": 1.814720812182741e-05,
|
| 1813 |
+
"loss": 0.0717,
|
| 1814 |
+
"step": 2510
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 6.395939086294416,
|
| 1818 |
+
"grad_norm": 20.555654525756836,
|
| 1819 |
+
"learning_rate": 1.802030456852792e-05,
|
| 1820 |
+
"loss": 0.0759,
|
| 1821 |
+
"step": 2520
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 6.4213197969543145,
|
| 1825 |
+
"grad_norm": 0.07365628331899643,
|
| 1826 |
+
"learning_rate": 1.789340101522843e-05,
|
| 1827 |
+
"loss": 0.0801,
|
| 1828 |
+
"step": 2530
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 6.446700507614214,
|
| 1832 |
+
"grad_norm": 0.05513671413064003,
|
| 1833 |
+
"learning_rate": 1.7766497461928935e-05,
|
| 1834 |
+
"loss": 0.0839,
|
| 1835 |
+
"step": 2540
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 6.472081218274112,
|
| 1839 |
+
"grad_norm": 1.4962226152420044,
|
| 1840 |
+
"learning_rate": 1.763959390862944e-05,
|
| 1841 |
+
"loss": 0.1288,
|
| 1842 |
+
"step": 2550
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 6.49746192893401,
|
| 1846 |
+
"grad_norm": 10.59736442565918,
|
| 1847 |
+
"learning_rate": 1.751269035532995e-05,
|
| 1848 |
+
"loss": 0.1318,
|
| 1849 |
+
"step": 2560
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 6.522842639593908,
|
| 1853 |
+
"grad_norm": 0.06739044189453125,
|
| 1854 |
+
"learning_rate": 1.7385786802030457e-05,
|
| 1855 |
+
"loss": 0.1288,
|
| 1856 |
+
"step": 2570
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 6.548223350253807,
|
| 1860 |
+
"grad_norm": 3.2351067066192627,
|
| 1861 |
+
"learning_rate": 1.7258883248730966e-05,
|
| 1862 |
+
"loss": 0.0801,
|
| 1863 |
+
"step": 2580
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 6.573604060913706,
|
| 1867 |
+
"grad_norm": 0.10322804003953934,
|
| 1868 |
+
"learning_rate": 1.7131979695431472e-05,
|
| 1869 |
+
"loss": 0.0687,
|
| 1870 |
+
"step": 2590
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 6.598984771573604,
|
| 1874 |
+
"grad_norm": 1.834286093711853,
|
| 1875 |
+
"learning_rate": 1.700507614213198e-05,
|
| 1876 |
+
"loss": 0.1842,
|
| 1877 |
+
"step": 2600
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 6.624365482233502,
|
| 1881 |
+
"grad_norm": 0.10404256731271744,
|
| 1882 |
+
"learning_rate": 1.6878172588832487e-05,
|
| 1883 |
+
"loss": 0.2148,
|
| 1884 |
+
"step": 2610
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 6.649746192893401,
|
| 1888 |
+
"grad_norm": 0.11388051509857178,
|
| 1889 |
+
"learning_rate": 1.6751269035532997e-05,
|
| 1890 |
+
"loss": 0.2577,
|
| 1891 |
+
"step": 2620
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 6.675126903553299,
|
| 1895 |
+
"grad_norm": 0.1220388114452362,
|
| 1896 |
+
"learning_rate": 1.6624365482233503e-05,
|
| 1897 |
+
"loss": 0.0425,
|
| 1898 |
+
"step": 2630
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 6.700507614213198,
|
| 1902 |
+
"grad_norm": 18.549720764160156,
|
| 1903 |
+
"learning_rate": 1.649746192893401e-05,
|
| 1904 |
+
"loss": 0.2702,
|
| 1905 |
+
"step": 2640
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 6.725888324873097,
|
| 1909 |
+
"grad_norm": 0.12684045732021332,
|
| 1910 |
+
"learning_rate": 1.6370558375634518e-05,
|
| 1911 |
+
"loss": 0.334,
|
| 1912 |
+
"step": 2650
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 6.751269035532995,
|
| 1916 |
+
"grad_norm": 0.27108675241470337,
|
| 1917 |
+
"learning_rate": 1.6243654822335024e-05,
|
| 1918 |
+
"loss": 0.1592,
|
| 1919 |
+
"step": 2660
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 6.776649746192893,
|
| 1923 |
+
"grad_norm": 0.17042884230613708,
|
| 1924 |
+
"learning_rate": 1.6116751269035534e-05,
|
| 1925 |
+
"loss": 0.0726,
|
| 1926 |
+
"step": 2670
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 6.802030456852792,
|
| 1930 |
+
"grad_norm": 0.11051168292760849,
|
| 1931 |
+
"learning_rate": 1.5989847715736043e-05,
|
| 1932 |
+
"loss": 0.0702,
|
| 1933 |
+
"step": 2680
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 6.8274111675126905,
|
| 1937 |
+
"grad_norm": 0.11623559892177582,
|
| 1938 |
+
"learning_rate": 1.586294416243655e-05,
|
| 1939 |
+
"loss": 0.1979,
|
| 1940 |
+
"step": 2690
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 6.852791878172589,
|
| 1944 |
+
"grad_norm": 0.16744251549243927,
|
| 1945 |
+
"learning_rate": 1.5736040609137055e-05,
|
| 1946 |
+
"loss": 0.1624,
|
| 1947 |
+
"step": 2700
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 6.878172588832487,
|
| 1951 |
+
"grad_norm": 62.56240463256836,
|
| 1952 |
+
"learning_rate": 1.5609137055837564e-05,
|
| 1953 |
+
"loss": 0.1029,
|
| 1954 |
+
"step": 2710
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 6.903553299492386,
|
| 1958 |
+
"grad_norm": 0.11622487008571625,
|
| 1959 |
+
"learning_rate": 1.548223350253807e-05,
|
| 1960 |
+
"loss": 0.0093,
|
| 1961 |
+
"step": 2720
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 6.928934010152284,
|
| 1965 |
+
"grad_norm": 0.09292230755090714,
|
| 1966 |
+
"learning_rate": 1.535532994923858e-05,
|
| 1967 |
+
"loss": 0.1308,
|
| 1968 |
+
"step": 2730
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 6.9543147208121825,
|
| 1972 |
+
"grad_norm": 0.20956498384475708,
|
| 1973 |
+
"learning_rate": 1.5228426395939088e-05,
|
| 1974 |
+
"loss": 0.0052,
|
| 1975 |
+
"step": 2740
|
| 1976 |
+
},
|
| 1977 |
+
{
|
| 1978 |
+
"epoch": 6.979695431472082,
|
| 1979 |
+
"grad_norm": 44.860416412353516,
|
| 1980 |
+
"learning_rate": 1.5101522842639595e-05,
|
| 1981 |
+
"loss": 0.2451,
|
| 1982 |
+
"step": 2750
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 7.0,
|
| 1986 |
+
"eval_loss": 1.0761854648590088,
|
| 1987 |
+
"eval_runtime": 8.8065,
|
| 1988 |
+
"eval_samples_per_second": 89.479,
|
| 1989 |
+
"eval_steps_per_second": 11.242,
|
| 1990 |
+
"step": 2758
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 7.00507614213198,
|
| 1994 |
+
"grad_norm": 0.09350736439228058,
|
| 1995 |
+
"learning_rate": 1.4974619289340103e-05,
|
| 1996 |
+
"loss": 0.0638,
|
| 1997 |
+
"step": 2760
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 7.030456852791878,
|
| 2001 |
+
"grad_norm": 3.844224214553833,
|
| 2002 |
+
"learning_rate": 1.484771573604061e-05,
|
| 2003 |
+
"loss": 0.0735,
|
| 2004 |
+
"step": 2770
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 7.055837563451776,
|
| 2008 |
+
"grad_norm": 0.09096001088619232,
|
| 2009 |
+
"learning_rate": 1.4720812182741117e-05,
|
| 2010 |
+
"loss": 0.0736,
|
| 2011 |
+
"step": 2780
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 7.081218274111675,
|
| 2015 |
+
"grad_norm": 0.06552541255950928,
|
| 2016 |
+
"learning_rate": 1.4593908629441624e-05,
|
| 2017 |
+
"loss": 0.0525,
|
| 2018 |
+
"step": 2790
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 7.106598984771574,
|
| 2022 |
+
"grad_norm": 0.10838861763477325,
|
| 2023 |
+
"learning_rate": 1.4467005076142132e-05,
|
| 2024 |
+
"loss": 0.4002,
|
| 2025 |
+
"step": 2800
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 7.131979695431472,
|
| 2029 |
+
"grad_norm": 0.12170158326625824,
|
| 2030 |
+
"learning_rate": 1.4340101522842641e-05,
|
| 2031 |
+
"loss": 0.1041,
|
| 2032 |
+
"step": 2810
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 7.157360406091371,
|
| 2036 |
+
"grad_norm": 0.15080027282238007,
|
| 2037 |
+
"learning_rate": 1.421319796954315e-05,
|
| 2038 |
+
"loss": 0.085,
|
| 2039 |
+
"step": 2820
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 7.182741116751269,
|
| 2043 |
+
"grad_norm": 0.13942450284957886,
|
| 2044 |
+
"learning_rate": 1.4086294416243657e-05,
|
| 2045 |
+
"loss": 0.007,
|
| 2046 |
+
"step": 2830
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 7.208121827411167,
|
| 2050 |
+
"grad_norm": 0.09515988826751709,
|
| 2051 |
+
"learning_rate": 1.3959390862944163e-05,
|
| 2052 |
+
"loss": 0.1357,
|
| 2053 |
+
"step": 2840
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 7.233502538071066,
|
| 2057 |
+
"grad_norm": 0.09349621832370758,
|
| 2058 |
+
"learning_rate": 1.383248730964467e-05,
|
| 2059 |
+
"loss": 0.0644,
|
| 2060 |
+
"step": 2850
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 7.258883248730965,
|
| 2064 |
+
"grad_norm": 0.25502246618270874,
|
| 2065 |
+
"learning_rate": 1.3705583756345178e-05,
|
| 2066 |
+
"loss": 0.0059,
|
| 2067 |
+
"step": 2860
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 7.284263959390863,
|
| 2071 |
+
"grad_norm": 5.170173168182373,
|
| 2072 |
+
"learning_rate": 1.3578680203045684e-05,
|
| 2073 |
+
"loss": 0.072,
|
| 2074 |
+
"step": 2870
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 7.309644670050761,
|
| 2078 |
+
"grad_norm": 0.09840302914381027,
|
| 2079 |
+
"learning_rate": 1.3451776649746192e-05,
|
| 2080 |
+
"loss": 0.1308,
|
| 2081 |
+
"step": 2880
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 7.33502538071066,
|
| 2085 |
+
"grad_norm": 0.07067205011844635,
|
| 2086 |
+
"learning_rate": 1.3324873096446703e-05,
|
| 2087 |
+
"loss": 0.1064,
|
| 2088 |
+
"step": 2890
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 7.3604060913705585,
|
| 2092 |
+
"grad_norm": 2.9675955772399902,
|
| 2093 |
+
"learning_rate": 1.3197969543147209e-05,
|
| 2094 |
+
"loss": 0.1342,
|
| 2095 |
+
"step": 2900
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 7.385786802030457,
|
| 2099 |
+
"grad_norm": 0.07539036124944687,
|
| 2100 |
+
"learning_rate": 1.3071065989847717e-05,
|
| 2101 |
+
"loss": 0.0642,
|
| 2102 |
+
"step": 2910
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 7.411167512690355,
|
| 2106 |
+
"grad_norm": 0.19838696718215942,
|
| 2107 |
+
"learning_rate": 1.2944162436548224e-05,
|
| 2108 |
+
"loss": 0.2019,
|
| 2109 |
+
"step": 2920
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 7.436548223350254,
|
| 2113 |
+
"grad_norm": 0.08878432959318161,
|
| 2114 |
+
"learning_rate": 1.281725888324873e-05,
|
| 2115 |
+
"loss": 0.104,
|
| 2116 |
+
"step": 2930
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 7.461928934010152,
|
| 2120 |
+
"grad_norm": 0.09144201874732971,
|
| 2121 |
+
"learning_rate": 1.2690355329949238e-05,
|
| 2122 |
+
"loss": 0.0046,
|
| 2123 |
+
"step": 2940
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 7.4873096446700504,
|
| 2127 |
+
"grad_norm": 0.0494062639772892,
|
| 2128 |
+
"learning_rate": 1.2563451776649746e-05,
|
| 2129 |
+
"loss": 0.0705,
|
| 2130 |
+
"step": 2950
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 7.5126903553299496,
|
| 2134 |
+
"grad_norm": 0.1255367249250412,
|
| 2135 |
+
"learning_rate": 1.2436548223350254e-05,
|
| 2136 |
+
"loss": 0.0382,
|
| 2137 |
+
"step": 2960
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 7.538071065989848,
|
| 2141 |
+
"grad_norm": 0.06985324621200562,
|
| 2142 |
+
"learning_rate": 1.2309644670050761e-05,
|
| 2143 |
+
"loss": 0.0863,
|
| 2144 |
+
"step": 2970
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 7.563451776649746,
|
| 2148 |
+
"grad_norm": 0.08622120320796967,
|
| 2149 |
+
"learning_rate": 1.218274111675127e-05,
|
| 2150 |
+
"loss": 0.0729,
|
| 2151 |
+
"step": 2980
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 7.588832487309645,
|
| 2155 |
+
"grad_norm": 0.07268553227186203,
|
| 2156 |
+
"learning_rate": 1.2055837563451777e-05,
|
| 2157 |
+
"loss": 0.1981,
|
| 2158 |
+
"step": 2990
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 7.614213197969543,
|
| 2162 |
+
"grad_norm": 5.232954025268555,
|
| 2163 |
+
"learning_rate": 1.1928934010152284e-05,
|
| 2164 |
+
"loss": 0.0885,
|
| 2165 |
+
"step": 3000
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 7.6395939086294415,
|
| 2169 |
+
"grad_norm": 0.07025603950023651,
|
| 2170 |
+
"learning_rate": 1.1802030456852794e-05,
|
| 2171 |
+
"loss": 0.1437,
|
| 2172 |
+
"step": 3010
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 7.66497461928934,
|
| 2176 |
+
"grad_norm": 3.0778110027313232,
|
| 2177 |
+
"learning_rate": 1.16751269035533e-05,
|
| 2178 |
+
"loss": 0.2441,
|
| 2179 |
+
"step": 3020
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 7.690355329949239,
|
| 2183 |
+
"grad_norm": 0.1191609650850296,
|
| 2184 |
+
"learning_rate": 1.1548223350253808e-05,
|
| 2185 |
+
"loss": 0.0717,
|
| 2186 |
+
"step": 3030
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 7.715736040609137,
|
| 2190 |
+
"grad_norm": 0.07436536252498627,
|
| 2191 |
+
"learning_rate": 1.1421319796954315e-05,
|
| 2192 |
+
"loss": 0.0636,
|
| 2193 |
+
"step": 3040
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 7.741116751269035,
|
| 2197 |
+
"grad_norm": 0.08062940835952759,
|
| 2198 |
+
"learning_rate": 1.1294416243654823e-05,
|
| 2199 |
+
"loss": 0.0918,
|
| 2200 |
+
"step": 3050
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 7.7664974619289335,
|
| 2204 |
+
"grad_norm": 0.06816007196903229,
|
| 2205 |
+
"learning_rate": 1.116751269035533e-05,
|
| 2206 |
+
"loss": 0.1376,
|
| 2207 |
+
"step": 3060
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 7.791878172588833,
|
| 2211 |
+
"grad_norm": 2.068460464477539,
|
| 2212 |
+
"learning_rate": 1.1040609137055838e-05,
|
| 2213 |
+
"loss": 0.0642,
|
| 2214 |
+
"step": 3070
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 7.817258883248731,
|
| 2218 |
+
"grad_norm": 0.1358051747083664,
|
| 2219 |
+
"learning_rate": 1.0913705583756344e-05,
|
| 2220 |
+
"loss": 0.1455,
|
| 2221 |
+
"step": 3080
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 7.842639593908629,
|
| 2225 |
+
"grad_norm": 0.07467884570360184,
|
| 2226 |
+
"learning_rate": 1.0786802030456854e-05,
|
| 2227 |
+
"loss": 0.0041,
|
| 2228 |
+
"step": 3090
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 7.868020304568528,
|
| 2232 |
+
"grad_norm": 0.08706346899271011,
|
| 2233 |
+
"learning_rate": 1.0659898477157361e-05,
|
| 2234 |
+
"loss": 0.0838,
|
| 2235 |
+
"step": 3100
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 7.893401015228426,
|
| 2239 |
+
"grad_norm": 0.07726430892944336,
|
| 2240 |
+
"learning_rate": 1.0532994923857867e-05,
|
| 2241 |
+
"loss": 0.0689,
|
| 2242 |
+
"step": 3110
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 7.918781725888325,
|
| 2246 |
+
"grad_norm": 0.09187603741884232,
|
| 2247 |
+
"learning_rate": 1.0406091370558377e-05,
|
| 2248 |
+
"loss": 0.0785,
|
| 2249 |
+
"step": 3120
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 7.944162436548224,
|
| 2253 |
+
"grad_norm": 0.1465052366256714,
|
| 2254 |
+
"learning_rate": 1.0279187817258885e-05,
|
| 2255 |
+
"loss": 0.0778,
|
| 2256 |
+
"step": 3130
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 7.969543147208122,
|
| 2260 |
+
"grad_norm": 0.06538953632116318,
|
| 2261 |
+
"learning_rate": 1.015228426395939e-05,
|
| 2262 |
+
"loss": 0.0577,
|
| 2263 |
+
"step": 3140
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 7.99492385786802,
|
| 2267 |
+
"grad_norm": 0.17358700931072235,
|
| 2268 |
+
"learning_rate": 1.0025380710659898e-05,
|
| 2269 |
+
"loss": 0.0041,
|
| 2270 |
+
"step": 3150
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 8.0,
|
| 2274 |
+
"eval_loss": 1.1998591423034668,
|
| 2275 |
+
"eval_runtime": 8.6883,
|
| 2276 |
+
"eval_samples_per_second": 90.697,
|
| 2277 |
+
"eval_steps_per_second": 11.395,
|
| 2278 |
+
"step": 3152
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 8.02030456852792,
|
| 2282 |
+
"grad_norm": 0.05773229897022247,
|
| 2283 |
+
"learning_rate": 9.898477157360408e-06,
|
| 2284 |
+
"loss": 0.0856,
|
| 2285 |
+
"step": 3160
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 8.045685279187817,
|
| 2289 |
+
"grad_norm": 0.07206518203020096,
|
| 2290 |
+
"learning_rate": 9.771573604060914e-06,
|
| 2291 |
+
"loss": 0.1201,
|
| 2292 |
+
"step": 3170
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 8.071065989847716,
|
| 2296 |
+
"grad_norm": 0.07798884809017181,
|
| 2297 |
+
"learning_rate": 9.644670050761421e-06,
|
| 2298 |
+
"loss": 0.069,
|
| 2299 |
+
"step": 3180
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 8.096446700507615,
|
| 2303 |
+
"grad_norm": 0.07664274424314499,
|
| 2304 |
+
"learning_rate": 9.517766497461929e-06,
|
| 2305 |
+
"loss": 0.0036,
|
| 2306 |
+
"step": 3190
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 8.121827411167512,
|
| 2310 |
+
"grad_norm": 0.050856612622737885,
|
| 2311 |
+
"learning_rate": 9.390862944162437e-06,
|
| 2312 |
+
"loss": 0.003,
|
| 2313 |
+
"step": 3200
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 8.147208121827411,
|
| 2317 |
+
"grad_norm": 0.0529194138944149,
|
| 2318 |
+
"learning_rate": 9.263959390862944e-06,
|
| 2319 |
+
"loss": 0.1412,
|
| 2320 |
+
"step": 3210
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 8.17258883248731,
|
| 2324 |
+
"grad_norm": 13.00708293914795,
|
| 2325 |
+
"learning_rate": 9.137055837563452e-06,
|
| 2326 |
+
"loss": 0.0513,
|
| 2327 |
+
"step": 3220
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 8.197969543147208,
|
| 2331 |
+
"grad_norm": 0.06696105003356934,
|
| 2332 |
+
"learning_rate": 9.01015228426396e-06,
|
| 2333 |
+
"loss": 0.0704,
|
| 2334 |
+
"step": 3230
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 8.223350253807107,
|
| 2338 |
+
"grad_norm": 0.06790643185377121,
|
| 2339 |
+
"learning_rate": 8.883248730964468e-06,
|
| 2340 |
+
"loss": 0.1119,
|
| 2341 |
+
"step": 3240
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 8.248730964467006,
|
| 2345 |
+
"grad_norm": 0.06100468337535858,
|
| 2346 |
+
"learning_rate": 8.756345177664975e-06,
|
| 2347 |
+
"loss": 0.0032,
|
| 2348 |
+
"step": 3250
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 8.274111675126903,
|
| 2352 |
+
"grad_norm": 8.70279312133789,
|
| 2353 |
+
"learning_rate": 8.629441624365483e-06,
|
| 2354 |
+
"loss": 0.0811,
|
| 2355 |
+
"step": 3260
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 8.299492385786802,
|
| 2359 |
+
"grad_norm": 0.09535681456327438,
|
| 2360 |
+
"learning_rate": 8.50253807106599e-06,
|
| 2361 |
+
"loss": 0.0795,
|
| 2362 |
+
"step": 3270
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 8.3248730964467,
|
| 2366 |
+
"grad_norm": 0.07176022976636887,
|
| 2367 |
+
"learning_rate": 8.375634517766498e-06,
|
| 2368 |
+
"loss": 0.0893,
|
| 2369 |
+
"step": 3280
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 8.350253807106599,
|
| 2373 |
+
"grad_norm": 0.051213160157203674,
|
| 2374 |
+
"learning_rate": 8.248730964467004e-06,
|
| 2375 |
+
"loss": 0.0701,
|
| 2376 |
+
"step": 3290
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 8.375634517766498,
|
| 2380 |
+
"grad_norm": 0.09963097423315048,
|
| 2381 |
+
"learning_rate": 8.121827411167512e-06,
|
| 2382 |
+
"loss": 0.0745,
|
| 2383 |
+
"step": 3300
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 8.401015228426395,
|
| 2387 |
+
"grad_norm": 0.0691935271024704,
|
| 2388 |
+
"learning_rate": 7.994923857868022e-06,
|
| 2389 |
+
"loss": 0.0884,
|
| 2390 |
+
"step": 3310
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 8.426395939086294,
|
| 2394 |
+
"grad_norm": 0.16761644184589386,
|
| 2395 |
+
"learning_rate": 7.868020304568528e-06,
|
| 2396 |
+
"loss": 0.0043,
|
| 2397 |
+
"step": 3320
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 8.451776649746193,
|
| 2401 |
+
"grad_norm": 0.04886673018336296,
|
| 2402 |
+
"learning_rate": 7.741116751269035e-06,
|
| 2403 |
+
"loss": 0.0037,
|
| 2404 |
+
"step": 3330
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 8.47715736040609,
|
| 2408 |
+
"grad_norm": 0.06593126058578491,
|
| 2409 |
+
"learning_rate": 7.614213197969544e-06,
|
| 2410 |
+
"loss": 0.0029,
|
| 2411 |
+
"step": 3340
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 8.50253807106599,
|
| 2415 |
+
"grad_norm": 8.315119743347168,
|
| 2416 |
+
"learning_rate": 7.4873096446700515e-06,
|
| 2417 |
+
"loss": 0.1359,
|
| 2418 |
+
"step": 3350
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 8.527918781725889,
|
| 2422 |
+
"grad_norm": 0.19459432363510132,
|
| 2423 |
+
"learning_rate": 7.360406091370558e-06,
|
| 2424 |
+
"loss": 0.0771,
|
| 2425 |
+
"step": 3360
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 8.553299492385786,
|
| 2429 |
+
"grad_norm": 0.09421133995056152,
|
| 2430 |
+
"learning_rate": 7.233502538071066e-06,
|
| 2431 |
+
"loss": 0.003,
|
| 2432 |
+
"step": 3370
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 8.578680203045685,
|
| 2436 |
+
"grad_norm": 0.04559624567627907,
|
| 2437 |
+
"learning_rate": 7.106598984771575e-06,
|
| 2438 |
+
"loss": 0.0647,
|
| 2439 |
+
"step": 3380
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 8.604060913705585,
|
| 2443 |
+
"grad_norm": 3.285724639892578,
|
| 2444 |
+
"learning_rate": 6.9796954314720814e-06,
|
| 2445 |
+
"loss": 0.0744,
|
| 2446 |
+
"step": 3390
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 8.629441624365482,
|
| 2450 |
+
"grad_norm": 0.0715055987238884,
|
| 2451 |
+
"learning_rate": 6.852791878172589e-06,
|
| 2452 |
+
"loss": 0.0696,
|
| 2453 |
+
"step": 3400
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 8.654822335025381,
|
| 2457 |
+
"grad_norm": 0.05040270462632179,
|
| 2458 |
+
"learning_rate": 6.725888324873096e-06,
|
| 2459 |
+
"loss": 0.0028,
|
| 2460 |
+
"step": 3410
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 8.680203045685278,
|
| 2464 |
+
"grad_norm": 0.05061913654208183,
|
| 2465 |
+
"learning_rate": 6.5989847715736045e-06,
|
| 2466 |
+
"loss": 0.1397,
|
| 2467 |
+
"step": 3420
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 8.705583756345177,
|
| 2471 |
+
"grad_norm": 0.06146248057484627,
|
| 2472 |
+
"learning_rate": 6.472081218274112e-06,
|
| 2473 |
+
"loss": 0.0757,
|
| 2474 |
+
"step": 3430
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 8.730964467005077,
|
| 2478 |
+
"grad_norm": 0.0398724228143692,
|
| 2479 |
+
"learning_rate": 6.345177664974619e-06,
|
| 2480 |
+
"loss": 0.0888,
|
| 2481 |
+
"step": 3440
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 8.756345177664974,
|
| 2485 |
+
"grad_norm": 0.058130159974098206,
|
| 2486 |
+
"learning_rate": 6.218274111675127e-06,
|
| 2487 |
+
"loss": 0.0714,
|
| 2488 |
+
"step": 3450
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 8.781725888324873,
|
| 2492 |
+
"grad_norm": 0.06121835485100746,
|
| 2493 |
+
"learning_rate": 6.091370558375635e-06,
|
| 2494 |
+
"loss": 0.0329,
|
| 2495 |
+
"step": 3460
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 8.807106598984772,
|
| 2499 |
+
"grad_norm": 0.048920340836048126,
|
| 2500 |
+
"learning_rate": 5.964467005076142e-06,
|
| 2501 |
+
"loss": 0.0032,
|
| 2502 |
+
"step": 3470
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 8.83248730964467,
|
| 2506 |
+
"grad_norm": 0.05234299972653389,
|
| 2507 |
+
"learning_rate": 5.83756345177665e-06,
|
| 2508 |
+
"loss": 0.07,
|
| 2509 |
+
"step": 3480
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 8.857868020304569,
|
| 2513 |
+
"grad_norm": 0.04120004549622536,
|
| 2514 |
+
"learning_rate": 5.710659898477158e-06,
|
| 2515 |
+
"loss": 0.152,
|
| 2516 |
+
"step": 3490
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 8.883248730964468,
|
| 2520 |
+
"grad_norm": 0.0480622835457325,
|
| 2521 |
+
"learning_rate": 5.583756345177665e-06,
|
| 2522 |
+
"loss": 0.162,
|
| 2523 |
+
"step": 3500
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 8.908629441624365,
|
| 2527 |
+
"grad_norm": 0.09636899083852768,
|
| 2528 |
+
"learning_rate": 5.456852791878172e-06,
|
| 2529 |
+
"loss": 0.0632,
|
| 2530 |
+
"step": 3510
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 8.934010152284264,
|
| 2534 |
+
"grad_norm": 11.911639213562012,
|
| 2535 |
+
"learning_rate": 5.329949238578681e-06,
|
| 2536 |
+
"loss": 0.0646,
|
| 2537 |
+
"step": 3520
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 8.959390862944163,
|
| 2541 |
+
"grad_norm": 109.4986801147461,
|
| 2542 |
+
"learning_rate": 5.203045685279188e-06,
|
| 2543 |
+
"loss": 0.1313,
|
| 2544 |
+
"step": 3530
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 8.98477157360406,
|
| 2548 |
+
"grad_norm": 0.07897670567035675,
|
| 2549 |
+
"learning_rate": 5.076142131979695e-06,
|
| 2550 |
+
"loss": 0.1454,
|
| 2551 |
+
"step": 3540
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 9.0,
|
| 2555 |
+
"eval_loss": 1.2512013912200928,
|
| 2556 |
+
"eval_runtime": 8.7832,
|
| 2557 |
+
"eval_samples_per_second": 89.717,
|
| 2558 |
+
"eval_steps_per_second": 11.272,
|
| 2559 |
+
"step": 3546
|
| 2560 |
+
},
|
| 2561 |
+
{
|
| 2562 |
+
"epoch": 9.01015228426396,
|
| 2563 |
+
"grad_norm": 0.08339915424585342,
|
| 2564 |
+
"learning_rate": 4.949238578680204e-06,
|
| 2565 |
+
"loss": 0.0042,
|
| 2566 |
+
"step": 3550
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"epoch": 9.035532994923859,
|
| 2570 |
+
"grad_norm": 0.06098608672618866,
|
| 2571 |
+
"learning_rate": 4.822335025380711e-06,
|
| 2572 |
+
"loss": 0.0843,
|
| 2573 |
+
"step": 3560
|
| 2574 |
+
},
|
| 2575 |
+
{
|
| 2576 |
+
"epoch": 9.060913705583756,
|
| 2577 |
+
"grad_norm": 0.09945400804281235,
|
| 2578 |
+
"learning_rate": 4.695431472081218e-06,
|
| 2579 |
+
"loss": 0.146,
|
| 2580 |
+
"step": 3570
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 9.086294416243655,
|
| 2584 |
+
"grad_norm": 0.4852578043937683,
|
| 2585 |
+
"learning_rate": 4.568527918781726e-06,
|
| 2586 |
+
"loss": 0.0034,
|
| 2587 |
+
"step": 3580
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 9.111675126903553,
|
| 2591 |
+
"grad_norm": 0.048197049647569656,
|
| 2592 |
+
"learning_rate": 4.441624365482234e-06,
|
| 2593 |
+
"loss": 0.0031,
|
| 2594 |
+
"step": 3590
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 9.137055837563452,
|
| 2598 |
+
"grad_norm": 0.07233395427465439,
|
| 2599 |
+
"learning_rate": 4.3147208121827415e-06,
|
| 2600 |
+
"loss": 0.0028,
|
| 2601 |
+
"step": 3600
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 9.16243654822335,
|
| 2605 |
+
"grad_norm": 0.05249156430363655,
|
| 2606 |
+
"learning_rate": 4.187817258883249e-06,
|
| 2607 |
+
"loss": 0.003,
|
| 2608 |
+
"step": 3610
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 9.187817258883248,
|
| 2612 |
+
"grad_norm": 0.056859567761421204,
|
| 2613 |
+
"learning_rate": 4.060913705583756e-06,
|
| 2614 |
+
"loss": 0.1712,
|
| 2615 |
+
"step": 3620
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 9.213197969543147,
|
| 2619 |
+
"grad_norm": 0.048108216375112534,
|
| 2620 |
+
"learning_rate": 3.934010152284264e-06,
|
| 2621 |
+
"loss": 0.046,
|
| 2622 |
+
"step": 3630
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"epoch": 9.238578680203046,
|
| 2626 |
+
"grad_norm": 3.2927534580230713,
|
| 2627 |
+
"learning_rate": 3.807106598984772e-06,
|
| 2628 |
+
"loss": 0.1724,
|
| 2629 |
+
"step": 3640
|
| 2630 |
+
},
|
| 2631 |
+
{
|
| 2632 |
+
"epoch": 9.263959390862944,
|
| 2633 |
+
"grad_norm": 0.05313184857368469,
|
| 2634 |
+
"learning_rate": 3.680203045685279e-06,
|
| 2635 |
+
"loss": 0.1504,
|
| 2636 |
+
"step": 3650
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 9.289340101522843,
|
| 2640 |
+
"grad_norm": 0.0587027408182621,
|
| 2641 |
+
"learning_rate": 3.5532994923857873e-06,
|
| 2642 |
+
"loss": 0.0811,
|
| 2643 |
+
"step": 3660
|
| 2644 |
+
},
|
| 2645 |
+
{
|
| 2646 |
+
"epoch": 9.314720812182742,
|
| 2647 |
+
"grad_norm": 0.0536118745803833,
|
| 2648 |
+
"learning_rate": 3.4263959390862946e-06,
|
| 2649 |
+
"loss": 0.0613,
|
| 2650 |
+
"step": 3670
|
| 2651 |
+
},
|
| 2652 |
+
{
|
| 2653 |
+
"epoch": 9.34010152284264,
|
| 2654 |
+
"grad_norm": 0.05343451723456383,
|
| 2655 |
+
"learning_rate": 3.2994923857868023e-06,
|
| 2656 |
+
"loss": 0.0028,
|
| 2657 |
+
"step": 3680
|
| 2658 |
+
},
|
| 2659 |
+
{
|
| 2660 |
+
"epoch": 9.365482233502538,
|
| 2661 |
+
"grad_norm": 0.0781240463256836,
|
| 2662 |
+
"learning_rate": 3.1725888324873095e-06,
|
| 2663 |
+
"loss": 0.0034,
|
| 2664 |
+
"step": 3690
|
| 2665 |
+
},
|
| 2666 |
+
{
|
| 2667 |
+
"epoch": 9.390862944162437,
|
| 2668 |
+
"grad_norm": 3.497049331665039,
|
| 2669 |
+
"learning_rate": 3.0456852791878177e-06,
|
| 2670 |
+
"loss": 0.2281,
|
| 2671 |
+
"step": 3700
|
| 2672 |
+
},
|
| 2673 |
+
{
|
| 2674 |
+
"epoch": 9.416243654822335,
|
| 2675 |
+
"grad_norm": 0.06697162985801697,
|
| 2676 |
+
"learning_rate": 2.918781725888325e-06,
|
| 2677 |
+
"loss": 0.0031,
|
| 2678 |
+
"step": 3710
|
| 2679 |
+
},
|
| 2680 |
+
{
|
| 2681 |
+
"epoch": 9.441624365482234,
|
| 2682 |
+
"grad_norm": 0.09822709858417511,
|
| 2683 |
+
"learning_rate": 2.7918781725888327e-06,
|
| 2684 |
+
"loss": 0.0186,
|
| 2685 |
+
"step": 3720
|
| 2686 |
+
},
|
| 2687 |
+
{
|
| 2688 |
+
"epoch": 9.467005076142131,
|
| 2689 |
+
"grad_norm": 0.0815950334072113,
|
| 2690 |
+
"learning_rate": 2.6649746192893404e-06,
|
| 2691 |
+
"loss": 0.1364,
|
| 2692 |
+
"step": 3730
|
| 2693 |
+
},
|
| 2694 |
+
{
|
| 2695 |
+
"epoch": 9.49238578680203,
|
| 2696 |
+
"grad_norm": 0.12073977291584015,
|
| 2697 |
+
"learning_rate": 2.5380710659898476e-06,
|
| 2698 |
+
"loss": 0.1374,
|
| 2699 |
+
"step": 3740
|
| 2700 |
+
},
|
| 2701 |
+
{
|
| 2702 |
+
"epoch": 9.51776649746193,
|
| 2703 |
+
"grad_norm": 0.07973968237638474,
|
| 2704 |
+
"learning_rate": 2.4111675126903553e-06,
|
| 2705 |
+
"loss": 0.0035,
|
| 2706 |
+
"step": 3750
|
| 2707 |
+
},
|
| 2708 |
+
{
|
| 2709 |
+
"epoch": 9.543147208121827,
|
| 2710 |
+
"grad_norm": 0.06170722842216492,
|
| 2711 |
+
"learning_rate": 2.284263959390863e-06,
|
| 2712 |
+
"loss": 0.0126,
|
| 2713 |
+
"step": 3760
|
| 2714 |
+
},
|
| 2715 |
+
{
|
| 2716 |
+
"epoch": 9.568527918781726,
|
| 2717 |
+
"grad_norm": 0.05730574578046799,
|
| 2718 |
+
"learning_rate": 2.1573604060913707e-06,
|
| 2719 |
+
"loss": 0.0032,
|
| 2720 |
+
"step": 3770
|
| 2721 |
+
},
|
| 2722 |
+
{
|
| 2723 |
+
"epoch": 9.593908629441625,
|
| 2724 |
+
"grad_norm": 0.0643545612692833,
|
| 2725 |
+
"learning_rate": 2.030456852791878e-06,
|
| 2726 |
+
"loss": 0.0552,
|
| 2727 |
+
"step": 3780
|
| 2728 |
+
},
|
| 2729 |
+
{
|
| 2730 |
+
"epoch": 9.619289340101522,
|
| 2731 |
+
"grad_norm": 0.051492851227521896,
|
| 2732 |
+
"learning_rate": 1.903553299492386e-06,
|
| 2733 |
+
"loss": 0.0026,
|
| 2734 |
+
"step": 3790
|
| 2735 |
+
},
|
| 2736 |
+
{
|
| 2737 |
+
"epoch": 9.644670050761421,
|
| 2738 |
+
"grad_norm": 0.05664081871509552,
|
| 2739 |
+
"learning_rate": 1.7766497461928936e-06,
|
| 2740 |
+
"loss": 0.0646,
|
| 2741 |
+
"step": 3800
|
| 2742 |
+
},
|
| 2743 |
+
{
|
| 2744 |
+
"epoch": 9.67005076142132,
|
| 2745 |
+
"grad_norm": 0.05502878874540329,
|
| 2746 |
+
"learning_rate": 1.6497461928934011e-06,
|
| 2747 |
+
"loss": 0.0031,
|
| 2748 |
+
"step": 3810
|
| 2749 |
+
},
|
| 2750 |
+
{
|
| 2751 |
+
"epoch": 9.695431472081218,
|
| 2752 |
+
"grad_norm": 0.3916303813457489,
|
| 2753 |
+
"learning_rate": 1.5228426395939088e-06,
|
| 2754 |
+
"loss": 0.0031,
|
| 2755 |
+
"step": 3820
|
| 2756 |
+
},
|
| 2757 |
+
{
|
| 2758 |
+
"epoch": 9.720812182741117,
|
| 2759 |
+
"grad_norm": 0.054553914815187454,
|
| 2760 |
+
"learning_rate": 1.3959390862944163e-06,
|
| 2761 |
+
"loss": 0.0842,
|
| 2762 |
+
"step": 3830
|
| 2763 |
+
},
|
| 2764 |
+
{
|
| 2765 |
+
"epoch": 9.746192893401016,
|
| 2766 |
+
"grad_norm": 0.07038414478302002,
|
| 2767 |
+
"learning_rate": 1.2690355329949238e-06,
|
| 2768 |
+
"loss": 0.0026,
|
| 2769 |
+
"step": 3840
|
| 2770 |
+
},
|
| 2771 |
+
{
|
| 2772 |
+
"epoch": 9.771573604060913,
|
| 2773 |
+
"grad_norm": 0.05792733281850815,
|
| 2774 |
+
"learning_rate": 1.1421319796954315e-06,
|
| 2775 |
+
"loss": 0.0025,
|
| 2776 |
+
"step": 3850
|
| 2777 |
+
},
|
| 2778 |
+
{
|
| 2779 |
+
"epoch": 9.796954314720812,
|
| 2780 |
+
"grad_norm": 0.07036729902029037,
|
| 2781 |
+
"learning_rate": 1.015228426395939e-06,
|
| 2782 |
+
"loss": 0.0039,
|
| 2783 |
+
"step": 3860
|
| 2784 |
+
},
|
| 2785 |
+
{
|
| 2786 |
+
"epoch": 9.82233502538071,
|
| 2787 |
+
"grad_norm": 0.05665631592273712,
|
| 2788 |
+
"learning_rate": 8.883248730964468e-07,
|
| 2789 |
+
"loss": 0.0852,
|
| 2790 |
+
"step": 3870
|
| 2791 |
+
},
|
| 2792 |
+
{
|
| 2793 |
+
"epoch": 9.847715736040609,
|
| 2794 |
+
"grad_norm": 0.0733490139245987,
|
| 2795 |
+
"learning_rate": 7.614213197969544e-07,
|
| 2796 |
+
"loss": 0.0025,
|
| 2797 |
+
"step": 3880
|
| 2798 |
+
},
|
| 2799 |
+
{
|
| 2800 |
+
"epoch": 9.873096446700508,
|
| 2801 |
+
"grad_norm": 0.05341969057917595,
|
| 2802 |
+
"learning_rate": 6.345177664974619e-07,
|
| 2803 |
+
"loss": 0.0595,
|
| 2804 |
+
"step": 3890
|
| 2805 |
+
},
|
| 2806 |
+
{
|
| 2807 |
+
"epoch": 9.898477157360405,
|
| 2808 |
+
"grad_norm": 0.06003400310873985,
|
| 2809 |
+
"learning_rate": 5.076142131979695e-07,
|
| 2810 |
+
"loss": 0.1296,
|
| 2811 |
+
"step": 3900
|
| 2812 |
+
},
|
| 2813 |
+
{
|
| 2814 |
+
"epoch": 9.923857868020304,
|
| 2815 |
+
"grad_norm": 0.18207885324954987,
|
| 2816 |
+
"learning_rate": 3.807106598984772e-07,
|
| 2817 |
+
"loss": 0.0575,
|
| 2818 |
+
"step": 3910
|
| 2819 |
+
},
|
| 2820 |
+
{
|
| 2821 |
+
"epoch": 9.949238578680204,
|
| 2822 |
+
"grad_norm": 5.820472240447998,
|
| 2823 |
+
"learning_rate": 2.5380710659898475e-07,
|
| 2824 |
+
"loss": 0.0699,
|
| 2825 |
+
"step": 3920
|
| 2826 |
+
},
|
| 2827 |
+
{
|
| 2828 |
+
"epoch": 9.974619289340101,
|
| 2829 |
+
"grad_norm": 0.10407544672489166,
|
| 2830 |
+
"learning_rate": 1.2690355329949238e-07,
|
| 2831 |
+
"loss": 0.105,
|
| 2832 |
+
"step": 3930
|
| 2833 |
+
},
|
| 2834 |
+
{
|
| 2835 |
+
"epoch": 10.0,
|
| 2836 |
+
"grad_norm": 0.062188029289245605,
|
| 2837 |
+
"learning_rate": 0.0,
|
| 2838 |
+
"loss": 0.003,
|
| 2839 |
+
"step": 3940
|
| 2840 |
+
}
|
| 2841 |
+
],
|
| 2842 |
+
"logging_steps": 10,
|
| 2843 |
+
"max_steps": 3940,
|
| 2844 |
+
"num_input_tokens_seen": 0,
|
| 2845 |
+
"num_train_epochs": 10,
|
| 2846 |
+
"save_steps": 500,
|
| 2847 |
+
"stateful_callbacks": {
|
| 2848 |
+
"TrainerControl": {
|
| 2849 |
+
"args": {
|
| 2850 |
+
"should_epoch_stop": false,
|
| 2851 |
+
"should_evaluate": false,
|
| 2852 |
+
"should_log": false,
|
| 2853 |
+
"should_save": true,
|
| 2854 |
+
"should_training_stop": true
|
| 2855 |
+
},
|
| 2856 |
+
"attributes": {}
|
| 2857 |
+
}
|
| 2858 |
+
},
|
| 2859 |
+
"total_flos": 4175372405637120.0,
|
| 2860 |
+
"train_batch_size": 8,
|
| 2861 |
+
"trial_name": null,
|
| 2862 |
+
"trial_params": null
|
| 2863 |
+
}
|
distilbert_ranking/checkpoint-3940/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5fa327595511a6fa6b860f7ffa79847603a3d8d181ec87c4527ae8ccaceb8863
|
| 3 |
+
size 5176
|
distilbert_ranking/checkpoint-3940/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
distilbert_ranking_final/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 267832560
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c1659eb788caa8eed53779b80179c9fa1813a2fd2205143c5bccc4710177c68e
|
| 3 |
size 267832560
|
distilbert_ranking_final/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5fa327595511a6fa6b860f7ffa79847603a3d8d181ec87c4527ae8ccaceb8863
|
| 3 |
size 5176
|
logs/events.out.tfevents.1726931188.88832d4e4248.36.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ed3676e6a2066d21ab1b9251363757f5464a61330c988c8766d081bdb869315e
|
| 3 |
+
size 90901
|