Upload folder using huggingface_hub
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- .gitattributes +14 -0
- checkpoint-11948/config.json +41 -0
- checkpoint-11948/generation_config.json +16 -0
- checkpoint-11948/model.safetensors +3 -0
- checkpoint-11948/optimizer.pt +3 -0
- checkpoint-11948/rng_state.pth +3 -0
- checkpoint-11948/scaler.pt +3 -0
- checkpoint-11948/scheduler.pt +3 -0
- checkpoint-11948/source.spm +3 -0
- checkpoint-11948/special_tokens_map.json +23 -0
- checkpoint-11948/target.spm +3 -0
- checkpoint-11948/tokenizer_config.json +40 -0
- checkpoint-11948/trainer_state.json +1731 -0
- checkpoint-11948/training_args.bin +3 -0
- checkpoint-11948/vocab.json +0 -0
- checkpoint-14935/config.json +41 -0
- checkpoint-14935/generation_config.json +16 -0
- checkpoint-14935/model.safetensors +3 -0
- checkpoint-14935/optimizer.pt +3 -0
- checkpoint-14935/rng_state.pth +3 -0
- checkpoint-14935/scaler.pt +3 -0
- checkpoint-14935/scheduler.pt +3 -0
- checkpoint-14935/source.spm +3 -0
- checkpoint-14935/special_tokens_map.json +23 -0
- checkpoint-14935/target.spm +3 -0
- checkpoint-14935/tokenizer_config.json +40 -0
- checkpoint-14935/trainer_state.json +2159 -0
- checkpoint-14935/training_args.bin +3 -0
- checkpoint-14935/vocab.json +0 -0
- checkpoint-17922/config.json +41 -0
- checkpoint-17922/generation_config.json +16 -0
- checkpoint-17922/model.safetensors +3 -0
- checkpoint-17922/optimizer.pt +3 -0
- checkpoint-17922/rng_state.pth +3 -0
- checkpoint-17922/scaler.pt +3 -0
- checkpoint-17922/scheduler.pt +3 -0
- checkpoint-17922/source.spm +3 -0
- checkpoint-17922/special_tokens_map.json +23 -0
- checkpoint-17922/target.spm +3 -0
- checkpoint-17922/tokenizer_config.json +40 -0
- checkpoint-17922/trainer_state.json +2587 -0
- checkpoint-17922/training_args.bin +3 -0
- checkpoint-17922/vocab.json +0 -0
- checkpoint-2987/config.json +41 -0
- checkpoint-2987/generation_config.json +16 -0
- checkpoint-2987/model.safetensors +3 -0
- checkpoint-2987/optimizer.pt +3 -0
- checkpoint-2987/rng_state.pth +3 -0
- checkpoint-2987/scaler.pt +3 -0
- checkpoint-2987/scheduler.pt +3 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,17 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
checkpoint-11948/source.spm filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
checkpoint-11948/target.spm filter=lfs diff=lfs merge=lfs -text
|
| 38 |
+
checkpoint-14935/source.spm filter=lfs diff=lfs merge=lfs -text
|
| 39 |
+
checkpoint-14935/target.spm filter=lfs diff=lfs merge=lfs -text
|
| 40 |
+
checkpoint-17922/source.spm filter=lfs diff=lfs merge=lfs -text
|
| 41 |
+
checkpoint-17922/target.spm filter=lfs diff=lfs merge=lfs -text
|
| 42 |
+
checkpoint-2987/source.spm filter=lfs diff=lfs merge=lfs -text
|
| 43 |
+
checkpoint-2987/target.spm filter=lfs diff=lfs merge=lfs -text
|
| 44 |
+
checkpoint-5974/source.spm filter=lfs diff=lfs merge=lfs -text
|
| 45 |
+
checkpoint-5974/target.spm filter=lfs diff=lfs merge=lfs -text
|
| 46 |
+
checkpoint-8961/source.spm filter=lfs diff=lfs merge=lfs -text
|
| 47 |
+
checkpoint-8961/target.spm filter=lfs diff=lfs merge=lfs -text
|
| 48 |
+
source.spm filter=lfs diff=lfs merge=lfs -text
|
| 49 |
+
target.spm filter=lfs diff=lfs merge=lfs -text
|
checkpoint-11948/config.json
ADDED
|
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ckartal/english-to-turkish-finetuned-model",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"activation_function": "swish",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"MarianMTModel"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"bos_token_id": 0,
|
| 10 |
+
"classifier_dropout": 0.0,
|
| 11 |
+
"d_model": 512,
|
| 12 |
+
"decoder_attention_heads": 8,
|
| 13 |
+
"decoder_ffn_dim": 2048,
|
| 14 |
+
"decoder_layerdrop": 0.0,
|
| 15 |
+
"decoder_layers": 6,
|
| 16 |
+
"decoder_start_token_id": 59993,
|
| 17 |
+
"decoder_vocab_size": 59994,
|
| 18 |
+
"dropout": 0.1,
|
| 19 |
+
"encoder_attention_heads": 8,
|
| 20 |
+
"encoder_ffn_dim": 2048,
|
| 21 |
+
"encoder_layerdrop": 0.0,
|
| 22 |
+
"encoder_layers": 6,
|
| 23 |
+
"eos_token_id": 0,
|
| 24 |
+
"forced_eos_token_id": 0,
|
| 25 |
+
"init_std": 0.02,
|
| 26 |
+
"is_encoder_decoder": true,
|
| 27 |
+
"max_length": null,
|
| 28 |
+
"max_position_embeddings": 512,
|
| 29 |
+
"model_type": "marian",
|
| 30 |
+
"normalize_embedding": false,
|
| 31 |
+
"num_beams": null,
|
| 32 |
+
"num_hidden_layers": 6,
|
| 33 |
+
"pad_token_id": 59993,
|
| 34 |
+
"scale_embedding": true,
|
| 35 |
+
"share_encoder_decoder_embeddings": true,
|
| 36 |
+
"static_position_embeddings": true,
|
| 37 |
+
"torch_dtype": "float32",
|
| 38 |
+
"transformers_version": "4.49.0",
|
| 39 |
+
"use_cache": true,
|
| 40 |
+
"vocab_size": 59994
|
| 41 |
+
}
|
checkpoint-11948/generation_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bad_words_ids": [
|
| 3 |
+
[
|
| 4 |
+
59993
|
| 5 |
+
]
|
| 6 |
+
],
|
| 7 |
+
"bos_token_id": 0,
|
| 8 |
+
"decoder_start_token_id": 59993,
|
| 9 |
+
"eos_token_id": 0,
|
| 10 |
+
"forced_eos_token_id": 0,
|
| 11 |
+
"max_length": 512,
|
| 12 |
+
"num_beams": 6,
|
| 13 |
+
"pad_token_id": 59993,
|
| 14 |
+
"renormalize_logits": true,
|
| 15 |
+
"transformers_version": "4.49.0"
|
| 16 |
+
}
|
checkpoint-11948/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:51d1dd137adffe19c3b054ce6fde4add2d4588c755def2d90103ef25413d9ff3
|
| 3 |
+
size 299690728
|
checkpoint-11948/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9fae7f2689749c5d7511fbc91404d7932e246d3bcacbb83d035467ed24c5573f
|
| 3 |
+
size 599054970
|
checkpoint-11948/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e98dff2b36b15ed0da9adbed2868493be995c81afcb89ad8f263069c503c6599
|
| 3 |
+
size 14244
|
checkpoint-11948/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5f824a254b9714a4704b979751121752508a64954c9977c2312faa340ea20ba0
|
| 3 |
+
size 988
|
checkpoint-11948/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:de83e7c49986c79a6b5bc283e32a2889ea6bca6ea6f90dbe4cbd62bdef7dcccd
|
| 3 |
+
size 1064
|
checkpoint-11948/source.spm
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:98eb24f0995a9d5f7cb0fb628c474628b1d2284615e881e857d062c0b651ce10
|
| 3 |
+
size 793920
|
checkpoint-11948/special_tokens_map.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"eos_token": {
|
| 3 |
+
"content": "</s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"pad_token": {
|
| 10 |
+
"content": "<pad>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"unk_token": {
|
| 17 |
+
"content": "<unk>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
}
|
| 23 |
+
}
|
checkpoint-11948/target.spm
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:45cc6000ed513cdca8f80739087fbcbf9933dc50c9ae36c319c9670882f72e1b
|
| 3 |
+
size 837876
|
checkpoint-11948/tokenizer_config.json
ADDED
|
@@ -0,0 +1,40 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "</s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<unk>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"59993": {
|
| 20 |
+
"content": "<pad>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
}
|
| 27 |
+
},
|
| 28 |
+
"clean_up_tokenization_spaces": true,
|
| 29 |
+
"eos_token": "</s>",
|
| 30 |
+
"extra_special_tokens": {},
|
| 31 |
+
"model_max_length": 512,
|
| 32 |
+
"pad_token": "<pad>",
|
| 33 |
+
"return_tensors": "pt",
|
| 34 |
+
"separate_vocabs": false,
|
| 35 |
+
"source_lang": "eng",
|
| 36 |
+
"sp_model_kwargs": {},
|
| 37 |
+
"target_lang": "tur",
|
| 38 |
+
"tokenizer_class": "MarianTokenizer",
|
| 39 |
+
"unk_token": "<unk>"
|
| 40 |
+
}
|
checkpoint-11948/trainer_state.json
ADDED
|
@@ -0,0 +1,1731 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 11948,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.016739203213927016,
|
| 13 |
+
"grad_norm": 0.439177542924881,
|
| 14 |
+
"learning_rate": 4.987445597589555e-05,
|
| 15 |
+
"loss": 1.5222,
|
| 16 |
+
"step": 50
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.03347840642785403,
|
| 20 |
+
"grad_norm": 0.4248828887939453,
|
| 21 |
+
"learning_rate": 4.973496261577949e-05,
|
| 22 |
+
"loss": 0.1842,
|
| 23 |
+
"step": 100
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.05021760964178105,
|
| 27 |
+
"grad_norm": 0.3019009232521057,
|
| 28 |
+
"learning_rate": 4.9595469255663436e-05,
|
| 29 |
+
"loss": 0.1471,
|
| 30 |
+
"step": 150
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.06695681285570806,
|
| 34 |
+
"grad_norm": 0.2518245577812195,
|
| 35 |
+
"learning_rate": 4.9455975895547376e-05,
|
| 36 |
+
"loss": 0.1306,
|
| 37 |
+
"step": 200
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.08369601606963509,
|
| 41 |
+
"grad_norm": 0.3660012185573578,
|
| 42 |
+
"learning_rate": 4.931648253543131e-05,
|
| 43 |
+
"loss": 0.1276,
|
| 44 |
+
"step": 250
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.1004352192835621,
|
| 48 |
+
"grad_norm": 0.32854148745536804,
|
| 49 |
+
"learning_rate": 4.917698917531526e-05,
|
| 50 |
+
"loss": 0.1048,
|
| 51 |
+
"step": 300
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.11717442249748912,
|
| 55 |
+
"grad_norm": 0.24879515171051025,
|
| 56 |
+
"learning_rate": 4.90374958151992e-05,
|
| 57 |
+
"loss": 0.1054,
|
| 58 |
+
"step": 350
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.13391362571141613,
|
| 62 |
+
"grad_norm": 0.36416563391685486,
|
| 63 |
+
"learning_rate": 4.889800245508314e-05,
|
| 64 |
+
"loss": 0.0985,
|
| 65 |
+
"step": 400
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.15065282892534315,
|
| 69 |
+
"grad_norm": 0.33641186356544495,
|
| 70 |
+
"learning_rate": 4.875850909496708e-05,
|
| 71 |
+
"loss": 0.1044,
|
| 72 |
+
"step": 450
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.16739203213927017,
|
| 76 |
+
"grad_norm": 0.32909244298934937,
|
| 77 |
+
"learning_rate": 4.861901573485103e-05,
|
| 78 |
+
"loss": 0.1089,
|
| 79 |
+
"step": 500
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.1841312353531972,
|
| 83 |
+
"grad_norm": 0.36060285568237305,
|
| 84 |
+
"learning_rate": 4.847952237473497e-05,
|
| 85 |
+
"loss": 0.09,
|
| 86 |
+
"step": 550
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.2008704385671242,
|
| 90 |
+
"grad_norm": 0.2510785758495331,
|
| 91 |
+
"learning_rate": 4.83400290146189e-05,
|
| 92 |
+
"loss": 0.0884,
|
| 93 |
+
"step": 600
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.21760964178105122,
|
| 97 |
+
"grad_norm": 0.22478719055652618,
|
| 98 |
+
"learning_rate": 4.820053565450285e-05,
|
| 99 |
+
"loss": 0.0866,
|
| 100 |
+
"step": 650
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.23434884499497824,
|
| 104 |
+
"grad_norm": 0.37321263551712036,
|
| 105 |
+
"learning_rate": 4.806104229438679e-05,
|
| 106 |
+
"loss": 0.0884,
|
| 107 |
+
"step": 700
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.25108804820890523,
|
| 111 |
+
"grad_norm": 0.2660929262638092,
|
| 112 |
+
"learning_rate": 4.792154893427073e-05,
|
| 113 |
+
"loss": 0.0819,
|
| 114 |
+
"step": 750
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.26782725142283226,
|
| 118 |
+
"grad_norm": 0.2338525801897049,
|
| 119 |
+
"learning_rate": 4.778205557415467e-05,
|
| 120 |
+
"loss": 0.0845,
|
| 121 |
+
"step": 800
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.2845664546367593,
|
| 125 |
+
"grad_norm": 0.308557391166687,
|
| 126 |
+
"learning_rate": 4.764256221403862e-05,
|
| 127 |
+
"loss": 0.0815,
|
| 128 |
+
"step": 850
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.3013056578506863,
|
| 132 |
+
"grad_norm": 0.27098262310028076,
|
| 133 |
+
"learning_rate": 4.750306885392255e-05,
|
| 134 |
+
"loss": 0.0833,
|
| 135 |
+
"step": 900
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.3180448610646133,
|
| 139 |
+
"grad_norm": 0.23054952919483185,
|
| 140 |
+
"learning_rate": 4.736357549380649e-05,
|
| 141 |
+
"loss": 0.0806,
|
| 142 |
+
"step": 950
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.33478406427854035,
|
| 146 |
+
"grad_norm": 0.21355900168418884,
|
| 147 |
+
"learning_rate": 4.722408213369044e-05,
|
| 148 |
+
"loss": 0.073,
|
| 149 |
+
"step": 1000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.3515232674924674,
|
| 153 |
+
"grad_norm": 0.20395708084106445,
|
| 154 |
+
"learning_rate": 4.708458877357438e-05,
|
| 155 |
+
"loss": 0.0775,
|
| 156 |
+
"step": 1050
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.3682624707063944,
|
| 160 |
+
"grad_norm": 0.21063613891601562,
|
| 161 |
+
"learning_rate": 4.694509541345832e-05,
|
| 162 |
+
"loss": 0.0789,
|
| 163 |
+
"step": 1100
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.3850016739203214,
|
| 167 |
+
"grad_norm": 0.20589284598827362,
|
| 168 |
+
"learning_rate": 4.680560205334226e-05,
|
| 169 |
+
"loss": 0.0809,
|
| 170 |
+
"step": 1150
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.4017408771342484,
|
| 174 |
+
"grad_norm": 0.27975228428840637,
|
| 175 |
+
"learning_rate": 4.666610869322621e-05,
|
| 176 |
+
"loss": 0.078,
|
| 177 |
+
"step": 1200
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.4184800803481754,
|
| 181 |
+
"grad_norm": 0.2529745399951935,
|
| 182 |
+
"learning_rate": 4.6526615333110144e-05,
|
| 183 |
+
"loss": 0.0704,
|
| 184 |
+
"step": 1250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.43521928356210243,
|
| 188 |
+
"grad_norm": 0.2205154448747635,
|
| 189 |
+
"learning_rate": 4.6387121972994084e-05,
|
| 190 |
+
"loss": 0.0733,
|
| 191 |
+
"step": 1300
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.45195848677602946,
|
| 195 |
+
"grad_norm": 0.2254629135131836,
|
| 196 |
+
"learning_rate": 4.624762861287803e-05,
|
| 197 |
+
"loss": 0.0751,
|
| 198 |
+
"step": 1350
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.4686976899899565,
|
| 202 |
+
"grad_norm": 0.17614957690238953,
|
| 203 |
+
"learning_rate": 4.610813525276197e-05,
|
| 204 |
+
"loss": 0.0747,
|
| 205 |
+
"step": 1400
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.4854368932038835,
|
| 209 |
+
"grad_norm": 0.15940478444099426,
|
| 210 |
+
"learning_rate": 4.596864189264591e-05,
|
| 211 |
+
"loss": 0.0698,
|
| 212 |
+
"step": 1450
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5021760964178105,
|
| 216 |
+
"grad_norm": 0.1869521141052246,
|
| 217 |
+
"learning_rate": 4.5829148532529854e-05,
|
| 218 |
+
"loss": 0.0721,
|
| 219 |
+
"step": 1500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.5189152996317375,
|
| 223 |
+
"grad_norm": 0.36063650250434875,
|
| 224 |
+
"learning_rate": 4.5689655172413794e-05,
|
| 225 |
+
"loss": 0.0706,
|
| 226 |
+
"step": 1550
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.5356545028456645,
|
| 230 |
+
"grad_norm": 0.16967014968395233,
|
| 231 |
+
"learning_rate": 4.5550161812297735e-05,
|
| 232 |
+
"loss": 0.0759,
|
| 233 |
+
"step": 1600
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.5523937060595916,
|
| 237 |
+
"grad_norm": 0.29293423891067505,
|
| 238 |
+
"learning_rate": 4.5410668452181676e-05,
|
| 239 |
+
"loss": 0.0711,
|
| 240 |
+
"step": 1650
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.5691329092735186,
|
| 244 |
+
"grad_norm": 0.3034748136997223,
|
| 245 |
+
"learning_rate": 4.527117509206562e-05,
|
| 246 |
+
"loss": 0.067,
|
| 247 |
+
"step": 1700
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.5858721124874456,
|
| 251 |
+
"grad_norm": 0.1974593997001648,
|
| 252 |
+
"learning_rate": 4.513168173194956e-05,
|
| 253 |
+
"loss": 0.0701,
|
| 254 |
+
"step": 1750
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.6026113157013726,
|
| 258 |
+
"grad_norm": 0.18101799488067627,
|
| 259 |
+
"learning_rate": 4.4992188371833505e-05,
|
| 260 |
+
"loss": 0.0717,
|
| 261 |
+
"step": 1800
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.6193505189152997,
|
| 265 |
+
"grad_norm": 0.14422941207885742,
|
| 266 |
+
"learning_rate": 4.4852695011717445e-05,
|
| 267 |
+
"loss": 0.0686,
|
| 268 |
+
"step": 1850
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.6360897221292267,
|
| 272 |
+
"grad_norm": 0.28663551807403564,
|
| 273 |
+
"learning_rate": 4.4713201651601386e-05,
|
| 274 |
+
"loss": 0.0646,
|
| 275 |
+
"step": 1900
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.6528289253431536,
|
| 279 |
+
"grad_norm": 0.23879379034042358,
|
| 280 |
+
"learning_rate": 4.4573708291485327e-05,
|
| 281 |
+
"loss": 0.0684,
|
| 282 |
+
"step": 1950
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.6695681285570807,
|
| 286 |
+
"grad_norm": 0.21389362215995789,
|
| 287 |
+
"learning_rate": 4.443421493136927e-05,
|
| 288 |
+
"loss": 0.066,
|
| 289 |
+
"step": 2000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.6863073317710077,
|
| 293 |
+
"grad_norm": 0.26841893792152405,
|
| 294 |
+
"learning_rate": 4.4294721571253215e-05,
|
| 295 |
+
"loss": 0.0717,
|
| 296 |
+
"step": 2050
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.7030465349849347,
|
| 300 |
+
"grad_norm": 0.240205317735672,
|
| 301 |
+
"learning_rate": 4.415522821113715e-05,
|
| 302 |
+
"loss": 0.0697,
|
| 303 |
+
"step": 2100
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.7197857381988617,
|
| 307 |
+
"grad_norm": 0.28098127245903015,
|
| 308 |
+
"learning_rate": 4.4015734851021096e-05,
|
| 309 |
+
"loss": 0.0713,
|
| 310 |
+
"step": 2150
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.7365249414127888,
|
| 314 |
+
"grad_norm": 0.23308847844600677,
|
| 315 |
+
"learning_rate": 4.3876241490905037e-05,
|
| 316 |
+
"loss": 0.0667,
|
| 317 |
+
"step": 2200
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.7532641446267158,
|
| 321 |
+
"grad_norm": 0.22748568654060364,
|
| 322 |
+
"learning_rate": 4.373674813078898e-05,
|
| 323 |
+
"loss": 0.0605,
|
| 324 |
+
"step": 2250
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.7700033478406428,
|
| 328 |
+
"grad_norm": 0.3932187259197235,
|
| 329 |
+
"learning_rate": 4.359725477067292e-05,
|
| 330 |
+
"loss": 0.0676,
|
| 331 |
+
"step": 2300
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.7867425510545698,
|
| 335 |
+
"grad_norm": 0.23918767273426056,
|
| 336 |
+
"learning_rate": 4.345776141055686e-05,
|
| 337 |
+
"loss": 0.0624,
|
| 338 |
+
"step": 2350
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.8034817542684968,
|
| 342 |
+
"grad_norm": 0.3068426549434662,
|
| 343 |
+
"learning_rate": 4.33182680504408e-05,
|
| 344 |
+
"loss": 0.0664,
|
| 345 |
+
"step": 2400
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.8202209574824239,
|
| 349 |
+
"grad_norm": 0.17977873980998993,
|
| 350 |
+
"learning_rate": 4.317877469032474e-05,
|
| 351 |
+
"loss": 0.0726,
|
| 352 |
+
"step": 2450
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.8369601606963508,
|
| 356 |
+
"grad_norm": 0.16876642405986786,
|
| 357 |
+
"learning_rate": 4.303928133020869e-05,
|
| 358 |
+
"loss": 0.0639,
|
| 359 |
+
"step": 2500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.8536993639102779,
|
| 363 |
+
"grad_norm": 0.17980250716209412,
|
| 364 |
+
"learning_rate": 4.289978797009263e-05,
|
| 365 |
+
"loss": 0.0701,
|
| 366 |
+
"step": 2550
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.8704385671242049,
|
| 370 |
+
"grad_norm": 0.1711459904909134,
|
| 371 |
+
"learning_rate": 4.276029460997656e-05,
|
| 372 |
+
"loss": 0.063,
|
| 373 |
+
"step": 2600
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.8871777703381319,
|
| 377 |
+
"grad_norm": 0.443228542804718,
|
| 378 |
+
"learning_rate": 4.262080124986051e-05,
|
| 379 |
+
"loss": 0.0675,
|
| 380 |
+
"step": 2650
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.9039169735520589,
|
| 384 |
+
"grad_norm": 0.2098589390516281,
|
| 385 |
+
"learning_rate": 4.248130788974445e-05,
|
| 386 |
+
"loss": 0.062,
|
| 387 |
+
"step": 2700
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.920656176765986,
|
| 391 |
+
"grad_norm": 0.3022039234638214,
|
| 392 |
+
"learning_rate": 4.234181452962839e-05,
|
| 393 |
+
"loss": 0.07,
|
| 394 |
+
"step": 2750
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.937395379979913,
|
| 398 |
+
"grad_norm": 0.19368910789489746,
|
| 399 |
+
"learning_rate": 4.220232116951233e-05,
|
| 400 |
+
"loss": 0.0621,
|
| 401 |
+
"step": 2800
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.9541345831938399,
|
| 405 |
+
"grad_norm": 0.18753108382225037,
|
| 406 |
+
"learning_rate": 4.206282780939628e-05,
|
| 407 |
+
"loss": 0.0631,
|
| 408 |
+
"step": 2850
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.970873786407767,
|
| 412 |
+
"grad_norm": 0.15517786145210266,
|
| 413 |
+
"learning_rate": 4.192333444928022e-05,
|
| 414 |
+
"loss": 0.0641,
|
| 415 |
+
"step": 2900
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.987612989621694,
|
| 419 |
+
"grad_norm": 0.11765792220830917,
|
| 420 |
+
"learning_rate": 4.178384108916415e-05,
|
| 421 |
+
"loss": 0.0612,
|
| 422 |
+
"step": 2950
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.0,
|
| 426 |
+
"eval_loss": 0.05521286651492119,
|
| 427 |
+
"eval_runtime": 50.415,
|
| 428 |
+
"eval_samples_per_second": 236.993,
|
| 429 |
+
"eval_steps_per_second": 14.817,
|
| 430 |
+
"step": 2987
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.004352192835621,
|
| 434 |
+
"grad_norm": 0.2691793739795685,
|
| 435 |
+
"learning_rate": 4.16443477290481e-05,
|
| 436 |
+
"loss": 0.059,
|
| 437 |
+
"step": 3000
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.021091396049548,
|
| 441 |
+
"grad_norm": 0.394694060087204,
|
| 442 |
+
"learning_rate": 4.150485436893204e-05,
|
| 443 |
+
"loss": 0.0566,
|
| 444 |
+
"step": 3050
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.037830599263475,
|
| 448 |
+
"grad_norm": 0.19438503682613373,
|
| 449 |
+
"learning_rate": 4.136536100881598e-05,
|
| 450 |
+
"loss": 0.0591,
|
| 451 |
+
"step": 3100
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.0545698024774022,
|
| 455 |
+
"grad_norm": 0.21350933611392975,
|
| 456 |
+
"learning_rate": 4.122586764869992e-05,
|
| 457 |
+
"loss": 0.0509,
|
| 458 |
+
"step": 3150
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.071309005691329,
|
| 462 |
+
"grad_norm": 0.26747575402259827,
|
| 463 |
+
"learning_rate": 4.108637428858387e-05,
|
| 464 |
+
"loss": 0.0589,
|
| 465 |
+
"step": 3200
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.088048208905256,
|
| 469 |
+
"grad_norm": 0.31256961822509766,
|
| 470 |
+
"learning_rate": 4.0946880928467804e-05,
|
| 471 |
+
"loss": 0.0602,
|
| 472 |
+
"step": 3250
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.1047874121191832,
|
| 476 |
+
"grad_norm": 0.18631280958652496,
|
| 477 |
+
"learning_rate": 4.0807387568351745e-05,
|
| 478 |
+
"loss": 0.0547,
|
| 479 |
+
"step": 3300
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.12152661533311,
|
| 483 |
+
"grad_norm": 0.18677473068237305,
|
| 484 |
+
"learning_rate": 4.066789420823569e-05,
|
| 485 |
+
"loss": 0.0543,
|
| 486 |
+
"step": 3350
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.1382658185470371,
|
| 490 |
+
"grad_norm": 0.24535444378852844,
|
| 491 |
+
"learning_rate": 4.052840084811963e-05,
|
| 492 |
+
"loss": 0.0583,
|
| 493 |
+
"step": 3400
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.1550050217609642,
|
| 497 |
+
"grad_norm": 0.1752105951309204,
|
| 498 |
+
"learning_rate": 4.038890748800357e-05,
|
| 499 |
+
"loss": 0.0504,
|
| 500 |
+
"step": 3450
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.1717442249748913,
|
| 504 |
+
"grad_norm": 0.14743360877037048,
|
| 505 |
+
"learning_rate": 4.0249414127887514e-05,
|
| 506 |
+
"loss": 0.055,
|
| 507 |
+
"step": 3500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.1884834281888181,
|
| 511 |
+
"grad_norm": 0.11535945534706116,
|
| 512 |
+
"learning_rate": 4.010992076777146e-05,
|
| 513 |
+
"loss": 0.0552,
|
| 514 |
+
"step": 3550
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.2052226314027452,
|
| 518 |
+
"grad_norm": 0.26563358306884766,
|
| 519 |
+
"learning_rate": 3.9970427407655395e-05,
|
| 520 |
+
"loss": 0.0552,
|
| 521 |
+
"step": 3600
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.2219618346166723,
|
| 525 |
+
"grad_norm": 0.15104246139526367,
|
| 526 |
+
"learning_rate": 3.9830934047539336e-05,
|
| 527 |
+
"loss": 0.0575,
|
| 528 |
+
"step": 3650
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.2387010378305994,
|
| 532 |
+
"grad_norm": 0.2198421210050583,
|
| 533 |
+
"learning_rate": 3.9691440687423283e-05,
|
| 534 |
+
"loss": 0.0567,
|
| 535 |
+
"step": 3700
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.2554402410445262,
|
| 539 |
+
"grad_norm": 0.20177733898162842,
|
| 540 |
+
"learning_rate": 3.955194732730722e-05,
|
| 541 |
+
"loss": 0.0556,
|
| 542 |
+
"step": 3750
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.2721794442584533,
|
| 546 |
+
"grad_norm": 0.36604830622673035,
|
| 547 |
+
"learning_rate": 3.9412453967191165e-05,
|
| 548 |
+
"loss": 0.0569,
|
| 549 |
+
"step": 3800
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.2889186474723804,
|
| 553 |
+
"grad_norm": 0.18883727490901947,
|
| 554 |
+
"learning_rate": 3.9272960607075105e-05,
|
| 555 |
+
"loss": 0.0595,
|
| 556 |
+
"step": 3850
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.3056578506863072,
|
| 560 |
+
"grad_norm": 0.14828617870807648,
|
| 561 |
+
"learning_rate": 3.9133467246959046e-05,
|
| 562 |
+
"loss": 0.0548,
|
| 563 |
+
"step": 3900
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.3223970539002343,
|
| 567 |
+
"grad_norm": 0.19220437109470367,
|
| 568 |
+
"learning_rate": 3.899397388684299e-05,
|
| 569 |
+
"loss": 0.053,
|
| 570 |
+
"step": 3950
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.3391362571141614,
|
| 574 |
+
"grad_norm": 0.16049669682979584,
|
| 575 |
+
"learning_rate": 3.885448052672693e-05,
|
| 576 |
+
"loss": 0.0581,
|
| 577 |
+
"step": 4000
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.3558754603280883,
|
| 581 |
+
"grad_norm": 0.22821515798568726,
|
| 582 |
+
"learning_rate": 3.8714987166610875e-05,
|
| 583 |
+
"loss": 0.0518,
|
| 584 |
+
"step": 4050
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.3726146635420153,
|
| 588 |
+
"grad_norm": 0.1879580318927765,
|
| 589 |
+
"learning_rate": 3.857549380649481e-05,
|
| 590 |
+
"loss": 0.0574,
|
| 591 |
+
"step": 4100
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.3893538667559424,
|
| 595 |
+
"grad_norm": 0.16026251018047333,
|
| 596 |
+
"learning_rate": 3.8436000446378756e-05,
|
| 597 |
+
"loss": 0.063,
|
| 598 |
+
"step": 4150
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.4060930699698695,
|
| 602 |
+
"grad_norm": 0.26868143677711487,
|
| 603 |
+
"learning_rate": 3.82965070862627e-05,
|
| 604 |
+
"loss": 0.0571,
|
| 605 |
+
"step": 4200
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.4228322731837966,
|
| 609 |
+
"grad_norm": 0.2529687285423279,
|
| 610 |
+
"learning_rate": 3.815701372614664e-05,
|
| 611 |
+
"loss": 0.0528,
|
| 612 |
+
"step": 4250
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.4395714763977234,
|
| 616 |
+
"grad_norm": 0.19138221442699432,
|
| 617 |
+
"learning_rate": 3.801752036603058e-05,
|
| 618 |
+
"loss": 0.0584,
|
| 619 |
+
"step": 4300
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.4563106796116505,
|
| 623 |
+
"grad_norm": 0.16359661519527435,
|
| 624 |
+
"learning_rate": 3.787802700591452e-05,
|
| 625 |
+
"loss": 0.0539,
|
| 626 |
+
"step": 4350
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.4730498828255776,
|
| 630 |
+
"grad_norm": 0.1373494267463684,
|
| 631 |
+
"learning_rate": 3.7738533645798466e-05,
|
| 632 |
+
"loss": 0.0557,
|
| 633 |
+
"step": 4400
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.4897890860395044,
|
| 637 |
+
"grad_norm": 0.15695162117481232,
|
| 638 |
+
"learning_rate": 3.75990402856824e-05,
|
| 639 |
+
"loss": 0.0491,
|
| 640 |
+
"step": 4450
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.5065282892534315,
|
| 644 |
+
"grad_norm": 0.18462614715099335,
|
| 645 |
+
"learning_rate": 3.745954692556635e-05,
|
| 646 |
+
"loss": 0.0495,
|
| 647 |
+
"step": 4500
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.5232674924673586,
|
| 651 |
+
"grad_norm": 0.27876704931259155,
|
| 652 |
+
"learning_rate": 3.732005356545029e-05,
|
| 653 |
+
"loss": 0.0523,
|
| 654 |
+
"step": 4550
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.5400066956812855,
|
| 658 |
+
"grad_norm": 0.30491840839385986,
|
| 659 |
+
"learning_rate": 3.718056020533423e-05,
|
| 660 |
+
"loss": 0.0564,
|
| 661 |
+
"step": 4600
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.5567458988952128,
|
| 665 |
+
"grad_norm": 0.18721336126327515,
|
| 666 |
+
"learning_rate": 3.704106684521817e-05,
|
| 667 |
+
"loss": 0.0524,
|
| 668 |
+
"step": 4650
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.5734851021091396,
|
| 672 |
+
"grad_norm": 0.21216215193271637,
|
| 673 |
+
"learning_rate": 3.690157348510211e-05,
|
| 674 |
+
"loss": 0.0521,
|
| 675 |
+
"step": 4700
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.5902243053230665,
|
| 679 |
+
"grad_norm": 0.1368396282196045,
|
| 680 |
+
"learning_rate": 3.676208012498605e-05,
|
| 681 |
+
"loss": 0.056,
|
| 682 |
+
"step": 4750
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.6069635085369938,
|
| 686 |
+
"grad_norm": 0.13692086935043335,
|
| 687 |
+
"learning_rate": 3.662258676486999e-05,
|
| 688 |
+
"loss": 0.0443,
|
| 689 |
+
"step": 4800
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.6237027117509206,
|
| 693 |
+
"grad_norm": 0.11640128493309021,
|
| 694 |
+
"learning_rate": 3.648309340475394e-05,
|
| 695 |
+
"loss": 0.0488,
|
| 696 |
+
"step": 4850
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.6404419149648477,
|
| 700 |
+
"grad_norm": 0.19953882694244385,
|
| 701 |
+
"learning_rate": 3.634360004463788e-05,
|
| 702 |
+
"loss": 0.0553,
|
| 703 |
+
"step": 4900
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.6571811181787748,
|
| 707 |
+
"grad_norm": 0.1966984122991562,
|
| 708 |
+
"learning_rate": 3.6204106684521813e-05,
|
| 709 |
+
"loss": 0.0536,
|
| 710 |
+
"step": 4950
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.6739203213927016,
|
| 714 |
+
"grad_norm": 0.2324533313512802,
|
| 715 |
+
"learning_rate": 3.606461332440576e-05,
|
| 716 |
+
"loss": 0.0493,
|
| 717 |
+
"step": 5000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.6906595246066287,
|
| 721 |
+
"grad_norm": 0.16217607259750366,
|
| 722 |
+
"learning_rate": 3.59251199642897e-05,
|
| 723 |
+
"loss": 0.0503,
|
| 724 |
+
"step": 5050
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.7073987278205558,
|
| 728 |
+
"grad_norm": 0.23949602246284485,
|
| 729 |
+
"learning_rate": 3.578562660417364e-05,
|
| 730 |
+
"loss": 0.0556,
|
| 731 |
+
"step": 5100
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.7241379310344827,
|
| 735 |
+
"grad_norm": 0.21387897431850433,
|
| 736 |
+
"learning_rate": 3.564613324405758e-05,
|
| 737 |
+
"loss": 0.0548,
|
| 738 |
+
"step": 5150
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.7408771342484097,
|
| 742 |
+
"grad_norm": 0.2055111676454544,
|
| 743 |
+
"learning_rate": 3.550663988394153e-05,
|
| 744 |
+
"loss": 0.06,
|
| 745 |
+
"step": 5200
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.7576163374623368,
|
| 749 |
+
"grad_norm": 0.20280921459197998,
|
| 750 |
+
"learning_rate": 3.5367146523825464e-05,
|
| 751 |
+
"loss": 0.0508,
|
| 752 |
+
"step": 5250
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.7743555406762637,
|
| 756 |
+
"grad_norm": 0.14165103435516357,
|
| 757 |
+
"learning_rate": 3.5227653163709405e-05,
|
| 758 |
+
"loss": 0.0581,
|
| 759 |
+
"step": 5300
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.791094743890191,
|
| 763 |
+
"grad_norm": 0.18099863827228546,
|
| 764 |
+
"learning_rate": 3.508815980359335e-05,
|
| 765 |
+
"loss": 0.0562,
|
| 766 |
+
"step": 5350
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.8078339471041178,
|
| 770 |
+
"grad_norm": 0.21743184328079224,
|
| 771 |
+
"learning_rate": 3.494866644347729e-05,
|
| 772 |
+
"loss": 0.0498,
|
| 773 |
+
"step": 5400
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.824573150318045,
|
| 777 |
+
"grad_norm": 0.20934534072875977,
|
| 778 |
+
"learning_rate": 3.4809173083361234e-05,
|
| 779 |
+
"loss": 0.0549,
|
| 780 |
+
"step": 5450
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.841312353531972,
|
| 784 |
+
"grad_norm": 0.1582174152135849,
|
| 785 |
+
"learning_rate": 3.4669679723245174e-05,
|
| 786 |
+
"loss": 0.0556,
|
| 787 |
+
"step": 5500
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.8580515567458988,
|
| 791 |
+
"grad_norm": 0.1624903827905655,
|
| 792 |
+
"learning_rate": 3.453018636312912e-05,
|
| 793 |
+
"loss": 0.0516,
|
| 794 |
+
"step": 5550
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.874790759959826,
|
| 798 |
+
"grad_norm": 0.16255798935890198,
|
| 799 |
+
"learning_rate": 3.4390693003013056e-05,
|
| 800 |
+
"loss": 0.0542,
|
| 801 |
+
"step": 5600
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.891529963173753,
|
| 805 |
+
"grad_norm": 0.1269742250442505,
|
| 806 |
+
"learning_rate": 3.4251199642896996e-05,
|
| 807 |
+
"loss": 0.0565,
|
| 808 |
+
"step": 5650
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.9082691663876798,
|
| 812 |
+
"grad_norm": 0.15966229140758514,
|
| 813 |
+
"learning_rate": 3.4111706282780944e-05,
|
| 814 |
+
"loss": 0.0538,
|
| 815 |
+
"step": 5700
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.925008369601607,
|
| 819 |
+
"grad_norm": 0.21506330370903015,
|
| 820 |
+
"learning_rate": 3.3972212922664884e-05,
|
| 821 |
+
"loss": 0.0505,
|
| 822 |
+
"step": 5750
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.941747572815534,
|
| 826 |
+
"grad_norm": 0.2145415097475052,
|
| 827 |
+
"learning_rate": 3.3832719562548825e-05,
|
| 828 |
+
"loss": 0.0521,
|
| 829 |
+
"step": 5800
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.9584867760294609,
|
| 833 |
+
"grad_norm": 0.10960496962070465,
|
| 834 |
+
"learning_rate": 3.3693226202432766e-05,
|
| 835 |
+
"loss": 0.0513,
|
| 836 |
+
"step": 5850
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.9752259792433882,
|
| 840 |
+
"grad_norm": 0.13635843992233276,
|
| 841 |
+
"learning_rate": 3.355373284231671e-05,
|
| 842 |
+
"loss": 0.0499,
|
| 843 |
+
"step": 5900
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.991965182457315,
|
| 847 |
+
"grad_norm": 0.1542210429906845,
|
| 848 |
+
"learning_rate": 3.341423948220065e-05,
|
| 849 |
+
"loss": 0.0556,
|
| 850 |
+
"step": 5950
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 2.0,
|
| 854 |
+
"eval_loss": 0.04946442320942879,
|
| 855 |
+
"eval_runtime": 55.6114,
|
| 856 |
+
"eval_samples_per_second": 214.848,
|
| 857 |
+
"eval_steps_per_second": 13.432,
|
| 858 |
+
"step": 5974
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 2.008704385671242,
|
| 862 |
+
"grad_norm": 0.1718842089176178,
|
| 863 |
+
"learning_rate": 3.327474612208459e-05,
|
| 864 |
+
"loss": 0.0503,
|
| 865 |
+
"step": 6000
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 2.025443588885169,
|
| 869 |
+
"grad_norm": 0.1528020203113556,
|
| 870 |
+
"learning_rate": 3.3135252761968535e-05,
|
| 871 |
+
"loss": 0.0479,
|
| 872 |
+
"step": 6050
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 2.042182792099096,
|
| 876 |
+
"grad_norm": 0.19148772954940796,
|
| 877 |
+
"learning_rate": 3.299575940185247e-05,
|
| 878 |
+
"loss": 0.0443,
|
| 879 |
+
"step": 6100
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 2.058921995313023,
|
| 883 |
+
"grad_norm": 0.18125496804714203,
|
| 884 |
+
"learning_rate": 3.2856266041736416e-05,
|
| 885 |
+
"loss": 0.0473,
|
| 886 |
+
"step": 6150
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 2.07566119852695,
|
| 890 |
+
"grad_norm": 0.20772996544837952,
|
| 891 |
+
"learning_rate": 3.271677268162036e-05,
|
| 892 |
+
"loss": 0.0539,
|
| 893 |
+
"step": 6200
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 2.092400401740877,
|
| 897 |
+
"grad_norm": 0.2518468201160431,
|
| 898 |
+
"learning_rate": 3.25772793215043e-05,
|
| 899 |
+
"loss": 0.0468,
|
| 900 |
+
"step": 6250
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 2.1091396049548043,
|
| 904 |
+
"grad_norm": 0.1350301206111908,
|
| 905 |
+
"learning_rate": 3.243778596138824e-05,
|
| 906 |
+
"loss": 0.0431,
|
| 907 |
+
"step": 6300
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 2.125878808168731,
|
| 911 |
+
"grad_norm": 0.19141735136508942,
|
| 912 |
+
"learning_rate": 3.229829260127218e-05,
|
| 913 |
+
"loss": 0.048,
|
| 914 |
+
"step": 6350
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 2.142618011382658,
|
| 918 |
+
"grad_norm": 0.2404586374759674,
|
| 919 |
+
"learning_rate": 3.2158799241156126e-05,
|
| 920 |
+
"loss": 0.0516,
|
| 921 |
+
"step": 6400
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 2.1593572145965854,
|
| 925 |
+
"grad_norm": 0.21710112690925598,
|
| 926 |
+
"learning_rate": 3.201930588104006e-05,
|
| 927 |
+
"loss": 0.0472,
|
| 928 |
+
"step": 6450
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 2.176096417810512,
|
| 932 |
+
"grad_norm": 0.14395031332969666,
|
| 933 |
+
"learning_rate": 3.187981252092401e-05,
|
| 934 |
+
"loss": 0.0438,
|
| 935 |
+
"step": 6500
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 2.192835621024439,
|
| 939 |
+
"grad_norm": 0.20882932841777802,
|
| 940 |
+
"learning_rate": 3.174031916080795e-05,
|
| 941 |
+
"loss": 0.0488,
|
| 942 |
+
"step": 6550
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 2.2095748242383664,
|
| 946 |
+
"grad_norm": 0.13824905455112457,
|
| 947 |
+
"learning_rate": 3.160082580069189e-05,
|
| 948 |
+
"loss": 0.0464,
|
| 949 |
+
"step": 6600
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 2.2263140274522932,
|
| 953 |
+
"grad_norm": 0.1783577799797058,
|
| 954 |
+
"learning_rate": 3.146133244057583e-05,
|
| 955 |
+
"loss": 0.0459,
|
| 956 |
+
"step": 6650
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 2.24305323066622,
|
| 960 |
+
"grad_norm": 0.22147531807422638,
|
| 961 |
+
"learning_rate": 3.132183908045977e-05,
|
| 962 |
+
"loss": 0.0476,
|
| 963 |
+
"step": 6700
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 2.2597924338801474,
|
| 967 |
+
"grad_norm": 0.17393821477890015,
|
| 968 |
+
"learning_rate": 3.118234572034371e-05,
|
| 969 |
+
"loss": 0.0436,
|
| 970 |
+
"step": 6750
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 2.2765316370940742,
|
| 974 |
+
"grad_norm": 0.15850785374641418,
|
| 975 |
+
"learning_rate": 3.104285236022765e-05,
|
| 976 |
+
"loss": 0.0476,
|
| 977 |
+
"step": 6800
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 2.2932708403080015,
|
| 981 |
+
"grad_norm": 0.16232182085514069,
|
| 982 |
+
"learning_rate": 3.09033590001116e-05,
|
| 983 |
+
"loss": 0.0473,
|
| 984 |
+
"step": 6850
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 2.3100100435219284,
|
| 988 |
+
"grad_norm": 0.1816001981496811,
|
| 989 |
+
"learning_rate": 3.076386563999554e-05,
|
| 990 |
+
"loss": 0.0427,
|
| 991 |
+
"step": 6900
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 2.3267492467358553,
|
| 995 |
+
"grad_norm": 0.13417834043502808,
|
| 996 |
+
"learning_rate": 3.062437227987948e-05,
|
| 997 |
+
"loss": 0.0448,
|
| 998 |
+
"step": 6950
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 2.3434884499497826,
|
| 1002 |
+
"grad_norm": 0.12576530873775482,
|
| 1003 |
+
"learning_rate": 3.048487891976342e-05,
|
| 1004 |
+
"loss": 0.0453,
|
| 1005 |
+
"step": 7000
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 2.3602276531637094,
|
| 1009 |
+
"grad_norm": 0.33120718598365784,
|
| 1010 |
+
"learning_rate": 3.0345385559647362e-05,
|
| 1011 |
+
"loss": 0.0462,
|
| 1012 |
+
"step": 7050
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 2.3769668563776363,
|
| 1016 |
+
"grad_norm": 0.22310969233512878,
|
| 1017 |
+
"learning_rate": 3.0205892199531306e-05,
|
| 1018 |
+
"loss": 0.0475,
|
| 1019 |
+
"step": 7100
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 2.3937060595915636,
|
| 1023 |
+
"grad_norm": 0.18150626122951508,
|
| 1024 |
+
"learning_rate": 3.0066398839415243e-05,
|
| 1025 |
+
"loss": 0.0489,
|
| 1026 |
+
"step": 7150
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 2.4104452628054904,
|
| 1030 |
+
"grad_norm": 0.28730452060699463,
|
| 1031 |
+
"learning_rate": 2.9926905479299187e-05,
|
| 1032 |
+
"loss": 0.0536,
|
| 1033 |
+
"step": 7200
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 2.4271844660194173,
|
| 1037 |
+
"grad_norm": 0.1918480098247528,
|
| 1038 |
+
"learning_rate": 2.9787412119183128e-05,
|
| 1039 |
+
"loss": 0.0426,
|
| 1040 |
+
"step": 7250
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 2.4439236692333446,
|
| 1044 |
+
"grad_norm": 0.16158398985862732,
|
| 1045 |
+
"learning_rate": 2.964791875906707e-05,
|
| 1046 |
+
"loss": 0.0458,
|
| 1047 |
+
"step": 7300
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 2.4606628724472714,
|
| 1051 |
+
"grad_norm": 0.27141231298446655,
|
| 1052 |
+
"learning_rate": 2.9508425398951012e-05,
|
| 1053 |
+
"loss": 0.0454,
|
| 1054 |
+
"step": 7350
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 2.4774020756611987,
|
| 1058 |
+
"grad_norm": 0.1777345836162567,
|
| 1059 |
+
"learning_rate": 2.936893203883495e-05,
|
| 1060 |
+
"loss": 0.0435,
|
| 1061 |
+
"step": 7400
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 2.4941412788751256,
|
| 1065 |
+
"grad_norm": 0.14735421538352966,
|
| 1066 |
+
"learning_rate": 2.9229438678718897e-05,
|
| 1067 |
+
"loss": 0.0489,
|
| 1068 |
+
"step": 7450
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 2.5108804820890525,
|
| 1072 |
+
"grad_norm": 0.1486055999994278,
|
| 1073 |
+
"learning_rate": 2.9089945318602834e-05,
|
| 1074 |
+
"loss": 0.0477,
|
| 1075 |
+
"step": 7500
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 2.5276196853029793,
|
| 1079 |
+
"grad_norm": 0.17078754305839539,
|
| 1080 |
+
"learning_rate": 2.895045195848678e-05,
|
| 1081 |
+
"loss": 0.0444,
|
| 1082 |
+
"step": 7550
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 2.5443588885169066,
|
| 1086 |
+
"grad_norm": 0.19276435673236847,
|
| 1087 |
+
"learning_rate": 2.881095859837072e-05,
|
| 1088 |
+
"loss": 0.0486,
|
| 1089 |
+
"step": 7600
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 2.5610980917308335,
|
| 1093 |
+
"grad_norm": 0.21209606528282166,
|
| 1094 |
+
"learning_rate": 2.8671465238254656e-05,
|
| 1095 |
+
"loss": 0.0497,
|
| 1096 |
+
"step": 7650
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 2.5778372949447608,
|
| 1100 |
+
"grad_norm": 0.21018877625465393,
|
| 1101 |
+
"learning_rate": 2.8531971878138604e-05,
|
| 1102 |
+
"loss": 0.0441,
|
| 1103 |
+
"step": 7700
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 2.5945764981586876,
|
| 1107 |
+
"grad_norm": 0.15666617453098297,
|
| 1108 |
+
"learning_rate": 2.839247851802254e-05,
|
| 1109 |
+
"loss": 0.0467,
|
| 1110 |
+
"step": 7750
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 2.6113157013726145,
|
| 1114 |
+
"grad_norm": 0.1940685212612152,
|
| 1115 |
+
"learning_rate": 2.8252985157906485e-05,
|
| 1116 |
+
"loss": 0.0523,
|
| 1117 |
+
"step": 7800
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 2.628054904586542,
|
| 1121 |
+
"grad_norm": 0.28480586409568787,
|
| 1122 |
+
"learning_rate": 2.8113491797790426e-05,
|
| 1123 |
+
"loss": 0.0481,
|
| 1124 |
+
"step": 7850
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 2.6447941078004686,
|
| 1128 |
+
"grad_norm": 0.2223973125219345,
|
| 1129 |
+
"learning_rate": 2.797399843767437e-05,
|
| 1130 |
+
"loss": 0.0432,
|
| 1131 |
+
"step": 7900
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 2.661533311014396,
|
| 1135 |
+
"grad_norm": 0.15986157953739166,
|
| 1136 |
+
"learning_rate": 2.783450507755831e-05,
|
| 1137 |
+
"loss": 0.0454,
|
| 1138 |
+
"step": 7950
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 2.678272514228323,
|
| 1142 |
+
"grad_norm": 0.1384258270263672,
|
| 1143 |
+
"learning_rate": 2.7695011717442248e-05,
|
| 1144 |
+
"loss": 0.0477,
|
| 1145 |
+
"step": 8000
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 2.6950117174422497,
|
| 1149 |
+
"grad_norm": 0.1721869707107544,
|
| 1150 |
+
"learning_rate": 2.7555518357326192e-05,
|
| 1151 |
+
"loss": 0.0453,
|
| 1152 |
+
"step": 8050
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 2.7117509206561765,
|
| 1156 |
+
"grad_norm": 0.20737840235233307,
|
| 1157 |
+
"learning_rate": 2.7416024997210132e-05,
|
| 1158 |
+
"loss": 0.0504,
|
| 1159 |
+
"step": 8100
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 2.728490123870104,
|
| 1163 |
+
"grad_norm": 0.18823584914207458,
|
| 1164 |
+
"learning_rate": 2.7276531637094077e-05,
|
| 1165 |
+
"loss": 0.0453,
|
| 1166 |
+
"step": 8150
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 2.7452293270840307,
|
| 1170 |
+
"grad_norm": 0.13201962411403656,
|
| 1171 |
+
"learning_rate": 2.7137038276978017e-05,
|
| 1172 |
+
"loss": 0.0433,
|
| 1173 |
+
"step": 8200
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 2.761968530297958,
|
| 1177 |
+
"grad_norm": 0.1443973183631897,
|
| 1178 |
+
"learning_rate": 2.699754491686196e-05,
|
| 1179 |
+
"loss": 0.0486,
|
| 1180 |
+
"step": 8250
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 2.778707733511885,
|
| 1184 |
+
"grad_norm": 0.29314514994621277,
|
| 1185 |
+
"learning_rate": 2.68580515567459e-05,
|
| 1186 |
+
"loss": 0.05,
|
| 1187 |
+
"step": 8300
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 2.7954469367258117,
|
| 1191 |
+
"grad_norm": 0.14852124452590942,
|
| 1192 |
+
"learning_rate": 2.671855819662984e-05,
|
| 1193 |
+
"loss": 0.0495,
|
| 1194 |
+
"step": 8350
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 2.812186139939739,
|
| 1198 |
+
"grad_norm": 0.19024662673473358,
|
| 1199 |
+
"learning_rate": 2.6579064836513783e-05,
|
| 1200 |
+
"loss": 0.0508,
|
| 1201 |
+
"step": 8400
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 2.828925343153666,
|
| 1205 |
+
"grad_norm": 0.1745578795671463,
|
| 1206 |
+
"learning_rate": 2.6439571476397724e-05,
|
| 1207 |
+
"loss": 0.0443,
|
| 1208 |
+
"step": 8450
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 2.845664546367593,
|
| 1212 |
+
"grad_norm": 0.18390017747879028,
|
| 1213 |
+
"learning_rate": 2.6300078116281668e-05,
|
| 1214 |
+
"loss": 0.0468,
|
| 1215 |
+
"step": 8500
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 2.86240374958152,
|
| 1219 |
+
"grad_norm": 0.22483347356319427,
|
| 1220 |
+
"learning_rate": 2.616058475616561e-05,
|
| 1221 |
+
"loss": 0.0467,
|
| 1222 |
+
"step": 8550
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 2.879142952795447,
|
| 1226 |
+
"grad_norm": 0.18160563707351685,
|
| 1227 |
+
"learning_rate": 2.6021091396049553e-05,
|
| 1228 |
+
"loss": 0.0441,
|
| 1229 |
+
"step": 8600
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 2.8958821560093737,
|
| 1233 |
+
"grad_norm": 0.13408955931663513,
|
| 1234 |
+
"learning_rate": 2.588159803593349e-05,
|
| 1235 |
+
"loss": 0.0446,
|
| 1236 |
+
"step": 8650
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 2.912621359223301,
|
| 1240 |
+
"grad_norm": 0.16038326919078827,
|
| 1241 |
+
"learning_rate": 2.574210467581743e-05,
|
| 1242 |
+
"loss": 0.0456,
|
| 1243 |
+
"step": 8700
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 2.929360562437228,
|
| 1247 |
+
"grad_norm": 0.22738413512706757,
|
| 1248 |
+
"learning_rate": 2.5602611315701375e-05,
|
| 1249 |
+
"loss": 0.0479,
|
| 1250 |
+
"step": 8750
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 2.946099765651155,
|
| 1254 |
+
"grad_norm": 0.20327210426330566,
|
| 1255 |
+
"learning_rate": 2.5463117955585315e-05,
|
| 1256 |
+
"loss": 0.0511,
|
| 1257 |
+
"step": 8800
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 2.962838968865082,
|
| 1261 |
+
"grad_norm": 0.15756353735923767,
|
| 1262 |
+
"learning_rate": 2.532362459546926e-05,
|
| 1263 |
+
"loss": 0.0426,
|
| 1264 |
+
"step": 8850
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 2.979578172079009,
|
| 1268 |
+
"grad_norm": 0.1305045783519745,
|
| 1269 |
+
"learning_rate": 2.5184131235353197e-05,
|
| 1270 |
+
"loss": 0.0442,
|
| 1271 |
+
"step": 8900
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 2.996317375292936,
|
| 1275 |
+
"grad_norm": 0.1610562801361084,
|
| 1276 |
+
"learning_rate": 2.5044637875237144e-05,
|
| 1277 |
+
"loss": 0.0467,
|
| 1278 |
+
"step": 8950
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 3.0,
|
| 1282 |
+
"eval_loss": 0.04702676460146904,
|
| 1283 |
+
"eval_runtime": 52.8989,
|
| 1284 |
+
"eval_samples_per_second": 225.865,
|
| 1285 |
+
"eval_steps_per_second": 14.121,
|
| 1286 |
+
"step": 8961
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 3.013056578506863,
|
| 1290 |
+
"grad_norm": 0.2042045295238495,
|
| 1291 |
+
"learning_rate": 2.490514451512108e-05,
|
| 1292 |
+
"loss": 0.0457,
|
| 1293 |
+
"step": 9000
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 3.02979578172079,
|
| 1297 |
+
"grad_norm": 0.27092283964157104,
|
| 1298 |
+
"learning_rate": 2.4765651155005022e-05,
|
| 1299 |
+
"loss": 0.0437,
|
| 1300 |
+
"step": 9050
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 3.046534984934717,
|
| 1304 |
+
"grad_norm": 0.18729308247566223,
|
| 1305 |
+
"learning_rate": 2.4626157794888966e-05,
|
| 1306 |
+
"loss": 0.039,
|
| 1307 |
+
"step": 9100
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 3.063274188148644,
|
| 1311 |
+
"grad_norm": 0.18845289945602417,
|
| 1312 |
+
"learning_rate": 2.4486664434772903e-05,
|
| 1313 |
+
"loss": 0.0422,
|
| 1314 |
+
"step": 9150
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 3.0800133913625714,
|
| 1318 |
+
"grad_norm": 0.17593874037265778,
|
| 1319 |
+
"learning_rate": 2.4347171074656847e-05,
|
| 1320 |
+
"loss": 0.0384,
|
| 1321 |
+
"step": 9200
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 3.096752594576498,
|
| 1325 |
+
"grad_norm": 0.23149123787879944,
|
| 1326 |
+
"learning_rate": 2.4207677714540788e-05,
|
| 1327 |
+
"loss": 0.0443,
|
| 1328 |
+
"step": 9250
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 3.113491797790425,
|
| 1332 |
+
"grad_norm": 0.1968075931072235,
|
| 1333 |
+
"learning_rate": 2.4068184354424732e-05,
|
| 1334 |
+
"loss": 0.0402,
|
| 1335 |
+
"step": 9300
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 3.1302310010043524,
|
| 1339 |
+
"grad_norm": 0.16893354058265686,
|
| 1340 |
+
"learning_rate": 2.3928690994308673e-05,
|
| 1341 |
+
"loss": 0.043,
|
| 1342 |
+
"step": 9350
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 3.146970204218279,
|
| 1346 |
+
"grad_norm": 0.161103293299675,
|
| 1347 |
+
"learning_rate": 2.3789197634192613e-05,
|
| 1348 |
+
"loss": 0.0414,
|
| 1349 |
+
"step": 9400
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 3.163709407432206,
|
| 1353 |
+
"grad_norm": 0.22712625563144684,
|
| 1354 |
+
"learning_rate": 2.3649704274076557e-05,
|
| 1355 |
+
"loss": 0.0426,
|
| 1356 |
+
"step": 9450
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 3.1804486106461334,
|
| 1360 |
+
"grad_norm": 0.12811464071273804,
|
| 1361 |
+
"learning_rate": 2.3510210913960495e-05,
|
| 1362 |
+
"loss": 0.0406,
|
| 1363 |
+
"step": 9500
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 3.1971878138600602,
|
| 1367 |
+
"grad_norm": 0.16762731969356537,
|
| 1368 |
+
"learning_rate": 2.337071755384444e-05,
|
| 1369 |
+
"loss": 0.0417,
|
| 1370 |
+
"step": 9550
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 3.213927017073987,
|
| 1374 |
+
"grad_norm": 0.14003124833106995,
|
| 1375 |
+
"learning_rate": 2.323122419372838e-05,
|
| 1376 |
+
"loss": 0.0453,
|
| 1377 |
+
"step": 9600
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 3.2306662202879144,
|
| 1381 |
+
"grad_norm": 0.16891060769557953,
|
| 1382 |
+
"learning_rate": 2.309173083361232e-05,
|
| 1383 |
+
"loss": 0.0445,
|
| 1384 |
+
"step": 9650
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 3.2474054235018412,
|
| 1388 |
+
"grad_norm": 0.16900208592414856,
|
| 1389 |
+
"learning_rate": 2.2952237473496264e-05,
|
| 1390 |
+
"loss": 0.039,
|
| 1391 |
+
"step": 9700
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 3.264144626715768,
|
| 1395 |
+
"grad_norm": 0.17154955863952637,
|
| 1396 |
+
"learning_rate": 2.2812744113380205e-05,
|
| 1397 |
+
"loss": 0.0403,
|
| 1398 |
+
"step": 9750
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 3.2808838299296954,
|
| 1402 |
+
"grad_norm": 0.1620296835899353,
|
| 1403 |
+
"learning_rate": 2.2673250753264145e-05,
|
| 1404 |
+
"loss": 0.0406,
|
| 1405 |
+
"step": 9800
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 3.2976230331436223,
|
| 1409 |
+
"grad_norm": 0.14487063884735107,
|
| 1410 |
+
"learning_rate": 2.2533757393148086e-05,
|
| 1411 |
+
"loss": 0.0467,
|
| 1412 |
+
"step": 9850
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 3.3143622363575496,
|
| 1416 |
+
"grad_norm": 0.13799156248569489,
|
| 1417 |
+
"learning_rate": 2.2394264033032027e-05,
|
| 1418 |
+
"loss": 0.0433,
|
| 1419 |
+
"step": 9900
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 3.3311014395714764,
|
| 1423 |
+
"grad_norm": 0.1507265418767929,
|
| 1424 |
+
"learning_rate": 2.225477067291597e-05,
|
| 1425 |
+
"loss": 0.0446,
|
| 1426 |
+
"step": 9950
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 3.3478406427854033,
|
| 1430 |
+
"grad_norm": 0.1605840027332306,
|
| 1431 |
+
"learning_rate": 2.211527731279991e-05,
|
| 1432 |
+
"loss": 0.0415,
|
| 1433 |
+
"step": 10000
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 3.3645798459993306,
|
| 1437 |
+
"grad_norm": 0.11910756677389145,
|
| 1438 |
+
"learning_rate": 2.1975783952683855e-05,
|
| 1439 |
+
"loss": 0.0453,
|
| 1440 |
+
"step": 10050
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 3.3813190492132574,
|
| 1444 |
+
"grad_norm": 0.15115414559841156,
|
| 1445 |
+
"learning_rate": 2.1836290592567796e-05,
|
| 1446 |
+
"loss": 0.0395,
|
| 1447 |
+
"step": 10100
|
| 1448 |
+
},
|
| 1449 |
+
{
|
| 1450 |
+
"epoch": 3.3980582524271843,
|
| 1451 |
+
"grad_norm": 0.2044568508863449,
|
| 1452 |
+
"learning_rate": 2.1696797232451737e-05,
|
| 1453 |
+
"loss": 0.0473,
|
| 1454 |
+
"step": 10150
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 3.4147974556411116,
|
| 1458 |
+
"grad_norm": 0.18123552203178406,
|
| 1459 |
+
"learning_rate": 2.1557303872335677e-05,
|
| 1460 |
+
"loss": 0.0411,
|
| 1461 |
+
"step": 10200
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 3.4315366588550384,
|
| 1465 |
+
"grad_norm": 0.6018120646476746,
|
| 1466 |
+
"learning_rate": 2.1417810512219618e-05,
|
| 1467 |
+
"loss": 0.0413,
|
| 1468 |
+
"step": 10250
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 3.4482758620689653,
|
| 1472 |
+
"grad_norm": 0.22490189969539642,
|
| 1473 |
+
"learning_rate": 2.1278317152103562e-05,
|
| 1474 |
+
"loss": 0.0437,
|
| 1475 |
+
"step": 10300
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 3.4650150652828926,
|
| 1479 |
+
"grad_norm": 0.1784990429878235,
|
| 1480 |
+
"learning_rate": 2.1138823791987503e-05,
|
| 1481 |
+
"loss": 0.0453,
|
| 1482 |
+
"step": 10350
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 3.4817542684968195,
|
| 1486 |
+
"grad_norm": 0.15248402953147888,
|
| 1487 |
+
"learning_rate": 2.0999330431871443e-05,
|
| 1488 |
+
"loss": 0.0456,
|
| 1489 |
+
"step": 10400
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 3.4984934717107468,
|
| 1493 |
+
"grad_norm": 0.15146291255950928,
|
| 1494 |
+
"learning_rate": 2.0859837071755387e-05,
|
| 1495 |
+
"loss": 0.0393,
|
| 1496 |
+
"step": 10450
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 3.5152326749246736,
|
| 1500 |
+
"grad_norm": 0.1662750393152237,
|
| 1501 |
+
"learning_rate": 2.0720343711639325e-05,
|
| 1502 |
+
"loss": 0.0408,
|
| 1503 |
+
"step": 10500
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 3.5319718781386005,
|
| 1507 |
+
"grad_norm": 0.1244506984949112,
|
| 1508 |
+
"learning_rate": 2.058085035152327e-05,
|
| 1509 |
+
"loss": 0.0426,
|
| 1510 |
+
"step": 10550
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 3.5487110813525278,
|
| 1514 |
+
"grad_norm": 0.1180344969034195,
|
| 1515 |
+
"learning_rate": 2.044135699140721e-05,
|
| 1516 |
+
"loss": 0.0434,
|
| 1517 |
+
"step": 10600
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 3.5654502845664546,
|
| 1521 |
+
"grad_norm": 0.15951013565063477,
|
| 1522 |
+
"learning_rate": 2.030186363129115e-05,
|
| 1523 |
+
"loss": 0.0387,
|
| 1524 |
+
"step": 10650
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 3.582189487780382,
|
| 1528 |
+
"grad_norm": 0.16064217686653137,
|
| 1529 |
+
"learning_rate": 2.0162370271175094e-05,
|
| 1530 |
+
"loss": 0.0445,
|
| 1531 |
+
"step": 10700
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 3.598928690994309,
|
| 1535 |
+
"grad_norm": 0.17813698947429657,
|
| 1536 |
+
"learning_rate": 2.0022876911059035e-05,
|
| 1537 |
+
"loss": 0.0407,
|
| 1538 |
+
"step": 10750
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 3.6156678942082356,
|
| 1542 |
+
"grad_norm": 0.1256450116634369,
|
| 1543 |
+
"learning_rate": 1.988338355094298e-05,
|
| 1544 |
+
"loss": 0.0462,
|
| 1545 |
+
"step": 10800
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 3.6324070974221625,
|
| 1549 |
+
"grad_norm": 0.14016403257846832,
|
| 1550 |
+
"learning_rate": 1.9743890190826916e-05,
|
| 1551 |
+
"loss": 0.0456,
|
| 1552 |
+
"step": 10850
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 3.64914630063609,
|
| 1556 |
+
"grad_norm": 0.1396850347518921,
|
| 1557 |
+
"learning_rate": 1.9604396830710857e-05,
|
| 1558 |
+
"loss": 0.0403,
|
| 1559 |
+
"step": 10900
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 3.6658855038500167,
|
| 1563 |
+
"grad_norm": 0.17943057417869568,
|
| 1564 |
+
"learning_rate": 1.94649034705948e-05,
|
| 1565 |
+
"loss": 0.0417,
|
| 1566 |
+
"step": 10950
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 3.682624707063944,
|
| 1570 |
+
"grad_norm": 0.14947953820228577,
|
| 1571 |
+
"learning_rate": 1.932541011047874e-05,
|
| 1572 |
+
"loss": 0.0421,
|
| 1573 |
+
"step": 11000
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 3.699363910277871,
|
| 1577 |
+
"grad_norm": 0.12628613412380219,
|
| 1578 |
+
"learning_rate": 1.9185916750362685e-05,
|
| 1579 |
+
"loss": 0.0435,
|
| 1580 |
+
"step": 11050
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 3.7161031134917977,
|
| 1584 |
+
"grad_norm": 0.2205984890460968,
|
| 1585 |
+
"learning_rate": 1.9046423390246626e-05,
|
| 1586 |
+
"loss": 0.0396,
|
| 1587 |
+
"step": 11100
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 3.732842316705725,
|
| 1591 |
+
"grad_norm": 0.13236357271671295,
|
| 1592 |
+
"learning_rate": 1.8906930030130567e-05,
|
| 1593 |
+
"loss": 0.0405,
|
| 1594 |
+
"step": 11150
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 3.749581519919652,
|
| 1598 |
+
"grad_norm": 0.15023528039455414,
|
| 1599 |
+
"learning_rate": 1.8767436670014507e-05,
|
| 1600 |
+
"loss": 0.0434,
|
| 1601 |
+
"step": 11200
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 3.7663207231335787,
|
| 1605 |
+
"grad_norm": 0.1427326649427414,
|
| 1606 |
+
"learning_rate": 1.8627943309898448e-05,
|
| 1607 |
+
"loss": 0.0437,
|
| 1608 |
+
"step": 11250
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 3.783059926347506,
|
| 1612 |
+
"grad_norm": 0.1890624761581421,
|
| 1613 |
+
"learning_rate": 1.8488449949782392e-05,
|
| 1614 |
+
"loss": 0.0408,
|
| 1615 |
+
"step": 11300
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 3.799799129561433,
|
| 1619 |
+
"grad_norm": 0.27970972657203674,
|
| 1620 |
+
"learning_rate": 1.8348956589666333e-05,
|
| 1621 |
+
"loss": 0.0441,
|
| 1622 |
+
"step": 11350
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 3.8165383327753597,
|
| 1626 |
+
"grad_norm": 0.12823455035686493,
|
| 1627 |
+
"learning_rate": 1.8209463229550273e-05,
|
| 1628 |
+
"loss": 0.0412,
|
| 1629 |
+
"step": 11400
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 3.833277535989287,
|
| 1633 |
+
"grad_norm": 0.1442965269088745,
|
| 1634 |
+
"learning_rate": 1.8069969869434218e-05,
|
| 1635 |
+
"loss": 0.0416,
|
| 1636 |
+
"step": 11450
|
| 1637 |
+
},
|
| 1638 |
+
{
|
| 1639 |
+
"epoch": 3.850016739203214,
|
| 1640 |
+
"grad_norm": 0.13739417493343353,
|
| 1641 |
+
"learning_rate": 1.7930476509318158e-05,
|
| 1642 |
+
"loss": 0.0397,
|
| 1643 |
+
"step": 11500
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 3.866755942417141,
|
| 1647 |
+
"grad_norm": 0.16616705060005188,
|
| 1648 |
+
"learning_rate": 1.77909831492021e-05,
|
| 1649 |
+
"loss": 0.0413,
|
| 1650 |
+
"step": 11550
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 3.883495145631068,
|
| 1654 |
+
"grad_norm": 0.23060384392738342,
|
| 1655 |
+
"learning_rate": 1.765148978908604e-05,
|
| 1656 |
+
"loss": 0.0447,
|
| 1657 |
+
"step": 11600
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 3.900234348844995,
|
| 1661 |
+
"grad_norm": 0.2936810553073883,
|
| 1662 |
+
"learning_rate": 1.751199642896998e-05,
|
| 1663 |
+
"loss": 0.0409,
|
| 1664 |
+
"step": 11650
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 3.9169735520589217,
|
| 1668 |
+
"grad_norm": 0.17367126047611237,
|
| 1669 |
+
"learning_rate": 1.7372503068853924e-05,
|
| 1670 |
+
"loss": 0.0411,
|
| 1671 |
+
"step": 11700
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 3.933712755272849,
|
| 1675 |
+
"grad_norm": 0.14550547301769257,
|
| 1676 |
+
"learning_rate": 1.7233009708737865e-05,
|
| 1677 |
+
"loss": 0.0445,
|
| 1678 |
+
"step": 11750
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 3.950451958486776,
|
| 1682 |
+
"grad_norm": 0.13322454690933228,
|
| 1683 |
+
"learning_rate": 1.709351634862181e-05,
|
| 1684 |
+
"loss": 0.0444,
|
| 1685 |
+
"step": 11800
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 3.967191161700703,
|
| 1689 |
+
"grad_norm": 0.13606959581375122,
|
| 1690 |
+
"learning_rate": 1.6954022988505746e-05,
|
| 1691 |
+
"loss": 0.041,
|
| 1692 |
+
"step": 11850
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 3.98393036491463,
|
| 1696 |
+
"grad_norm": 0.14227426052093506,
|
| 1697 |
+
"learning_rate": 1.681452962838969e-05,
|
| 1698 |
+
"loss": 0.0412,
|
| 1699 |
+
"step": 11900
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 4.0,
|
| 1703 |
+
"eval_loss": 0.045825306326150894,
|
| 1704 |
+
"eval_runtime": 48.668,
|
| 1705 |
+
"eval_samples_per_second": 245.5,
|
| 1706 |
+
"eval_steps_per_second": 15.349,
|
| 1707 |
+
"step": 11948
|
| 1708 |
+
}
|
| 1709 |
+
],
|
| 1710 |
+
"logging_steps": 50,
|
| 1711 |
+
"max_steps": 17922,
|
| 1712 |
+
"num_input_tokens_seen": 0,
|
| 1713 |
+
"num_train_epochs": 6,
|
| 1714 |
+
"save_steps": 500,
|
| 1715 |
+
"stateful_callbacks": {
|
| 1716 |
+
"TrainerControl": {
|
| 1717 |
+
"args": {
|
| 1718 |
+
"should_epoch_stop": false,
|
| 1719 |
+
"should_evaluate": false,
|
| 1720 |
+
"should_log": false,
|
| 1721 |
+
"should_save": true,
|
| 1722 |
+
"should_training_stop": false
|
| 1723 |
+
},
|
| 1724 |
+
"attributes": {}
|
| 1725 |
+
}
|
| 1726 |
+
},
|
| 1727 |
+
"total_flos": 2.5921130506223616e+16,
|
| 1728 |
+
"train_batch_size": 16,
|
| 1729 |
+
"trial_name": null,
|
| 1730 |
+
"trial_params": null
|
| 1731 |
+
}
|
checkpoint-11948/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:37dd546b69fb60d8deb15a8b88e40b23e367c0e9f5a053ea3ae7c730b3874f2e
|
| 3 |
+
size 5304
|
checkpoint-11948/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-14935/config.json
ADDED
|
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ckartal/english-to-turkish-finetuned-model",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"activation_function": "swish",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"MarianMTModel"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"bos_token_id": 0,
|
| 10 |
+
"classifier_dropout": 0.0,
|
| 11 |
+
"d_model": 512,
|
| 12 |
+
"decoder_attention_heads": 8,
|
| 13 |
+
"decoder_ffn_dim": 2048,
|
| 14 |
+
"decoder_layerdrop": 0.0,
|
| 15 |
+
"decoder_layers": 6,
|
| 16 |
+
"decoder_start_token_id": 59993,
|
| 17 |
+
"decoder_vocab_size": 59994,
|
| 18 |
+
"dropout": 0.1,
|
| 19 |
+
"encoder_attention_heads": 8,
|
| 20 |
+
"encoder_ffn_dim": 2048,
|
| 21 |
+
"encoder_layerdrop": 0.0,
|
| 22 |
+
"encoder_layers": 6,
|
| 23 |
+
"eos_token_id": 0,
|
| 24 |
+
"forced_eos_token_id": 0,
|
| 25 |
+
"init_std": 0.02,
|
| 26 |
+
"is_encoder_decoder": true,
|
| 27 |
+
"max_length": null,
|
| 28 |
+
"max_position_embeddings": 512,
|
| 29 |
+
"model_type": "marian",
|
| 30 |
+
"normalize_embedding": false,
|
| 31 |
+
"num_beams": null,
|
| 32 |
+
"num_hidden_layers": 6,
|
| 33 |
+
"pad_token_id": 59993,
|
| 34 |
+
"scale_embedding": true,
|
| 35 |
+
"share_encoder_decoder_embeddings": true,
|
| 36 |
+
"static_position_embeddings": true,
|
| 37 |
+
"torch_dtype": "float32",
|
| 38 |
+
"transformers_version": "4.49.0",
|
| 39 |
+
"use_cache": true,
|
| 40 |
+
"vocab_size": 59994
|
| 41 |
+
}
|
checkpoint-14935/generation_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bad_words_ids": [
|
| 3 |
+
[
|
| 4 |
+
59993
|
| 5 |
+
]
|
| 6 |
+
],
|
| 7 |
+
"bos_token_id": 0,
|
| 8 |
+
"decoder_start_token_id": 59993,
|
| 9 |
+
"eos_token_id": 0,
|
| 10 |
+
"forced_eos_token_id": 0,
|
| 11 |
+
"max_length": 512,
|
| 12 |
+
"num_beams": 6,
|
| 13 |
+
"pad_token_id": 59993,
|
| 14 |
+
"renormalize_logits": true,
|
| 15 |
+
"transformers_version": "4.49.0"
|
| 16 |
+
}
|
checkpoint-14935/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:10b7619757ec37001c328bb68f33749e3183758158980f8c3a5c346dcb866279
|
| 3 |
+
size 299690728
|
checkpoint-14935/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ade4432f6ff505cb413003c71d8fa4ceedcadd0040b992e0df392dbd10889c3d
|
| 3 |
+
size 599054970
|
checkpoint-14935/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3f721aaec70579761bcfa3d2d4441178be84c6f3dcc5f6d933e54b21fe8e6cc2
|
| 3 |
+
size 14244
|
checkpoint-14935/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:36a823fbff578c8a82c44346b99397ed351d5e6783ee511d8e35285b13133caf
|
| 3 |
+
size 988
|
checkpoint-14935/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f321649b2aa5ecd287cb6a64d13837fccb0d3045d069aabc11f66246e4800051
|
| 3 |
+
size 1064
|
checkpoint-14935/source.spm
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:98eb24f0995a9d5f7cb0fb628c474628b1d2284615e881e857d062c0b651ce10
|
| 3 |
+
size 793920
|
checkpoint-14935/special_tokens_map.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"eos_token": {
|
| 3 |
+
"content": "</s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"pad_token": {
|
| 10 |
+
"content": "<pad>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"unk_token": {
|
| 17 |
+
"content": "<unk>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
}
|
| 23 |
+
}
|
checkpoint-14935/target.spm
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:45cc6000ed513cdca8f80739087fbcbf9933dc50c9ae36c319c9670882f72e1b
|
| 3 |
+
size 837876
|
checkpoint-14935/tokenizer_config.json
ADDED
|
@@ -0,0 +1,40 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "</s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<unk>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"59993": {
|
| 20 |
+
"content": "<pad>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
}
|
| 27 |
+
},
|
| 28 |
+
"clean_up_tokenization_spaces": true,
|
| 29 |
+
"eos_token": "</s>",
|
| 30 |
+
"extra_special_tokens": {},
|
| 31 |
+
"model_max_length": 512,
|
| 32 |
+
"pad_token": "<pad>",
|
| 33 |
+
"return_tensors": "pt",
|
| 34 |
+
"separate_vocabs": false,
|
| 35 |
+
"source_lang": "eng",
|
| 36 |
+
"sp_model_kwargs": {},
|
| 37 |
+
"target_lang": "tur",
|
| 38 |
+
"tokenizer_class": "MarianTokenizer",
|
| 39 |
+
"unk_token": "<unk>"
|
| 40 |
+
}
|
checkpoint-14935/trainer_state.json
ADDED
|
@@ -0,0 +1,2159 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 14935,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.016739203213927016,
|
| 13 |
+
"grad_norm": 0.439177542924881,
|
| 14 |
+
"learning_rate": 4.987445597589555e-05,
|
| 15 |
+
"loss": 1.5222,
|
| 16 |
+
"step": 50
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.03347840642785403,
|
| 20 |
+
"grad_norm": 0.4248828887939453,
|
| 21 |
+
"learning_rate": 4.973496261577949e-05,
|
| 22 |
+
"loss": 0.1842,
|
| 23 |
+
"step": 100
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.05021760964178105,
|
| 27 |
+
"grad_norm": 0.3019009232521057,
|
| 28 |
+
"learning_rate": 4.9595469255663436e-05,
|
| 29 |
+
"loss": 0.1471,
|
| 30 |
+
"step": 150
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.06695681285570806,
|
| 34 |
+
"grad_norm": 0.2518245577812195,
|
| 35 |
+
"learning_rate": 4.9455975895547376e-05,
|
| 36 |
+
"loss": 0.1306,
|
| 37 |
+
"step": 200
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.08369601606963509,
|
| 41 |
+
"grad_norm": 0.3660012185573578,
|
| 42 |
+
"learning_rate": 4.931648253543131e-05,
|
| 43 |
+
"loss": 0.1276,
|
| 44 |
+
"step": 250
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.1004352192835621,
|
| 48 |
+
"grad_norm": 0.32854148745536804,
|
| 49 |
+
"learning_rate": 4.917698917531526e-05,
|
| 50 |
+
"loss": 0.1048,
|
| 51 |
+
"step": 300
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.11717442249748912,
|
| 55 |
+
"grad_norm": 0.24879515171051025,
|
| 56 |
+
"learning_rate": 4.90374958151992e-05,
|
| 57 |
+
"loss": 0.1054,
|
| 58 |
+
"step": 350
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.13391362571141613,
|
| 62 |
+
"grad_norm": 0.36416563391685486,
|
| 63 |
+
"learning_rate": 4.889800245508314e-05,
|
| 64 |
+
"loss": 0.0985,
|
| 65 |
+
"step": 400
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.15065282892534315,
|
| 69 |
+
"grad_norm": 0.33641186356544495,
|
| 70 |
+
"learning_rate": 4.875850909496708e-05,
|
| 71 |
+
"loss": 0.1044,
|
| 72 |
+
"step": 450
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.16739203213927017,
|
| 76 |
+
"grad_norm": 0.32909244298934937,
|
| 77 |
+
"learning_rate": 4.861901573485103e-05,
|
| 78 |
+
"loss": 0.1089,
|
| 79 |
+
"step": 500
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.1841312353531972,
|
| 83 |
+
"grad_norm": 0.36060285568237305,
|
| 84 |
+
"learning_rate": 4.847952237473497e-05,
|
| 85 |
+
"loss": 0.09,
|
| 86 |
+
"step": 550
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.2008704385671242,
|
| 90 |
+
"grad_norm": 0.2510785758495331,
|
| 91 |
+
"learning_rate": 4.83400290146189e-05,
|
| 92 |
+
"loss": 0.0884,
|
| 93 |
+
"step": 600
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.21760964178105122,
|
| 97 |
+
"grad_norm": 0.22478719055652618,
|
| 98 |
+
"learning_rate": 4.820053565450285e-05,
|
| 99 |
+
"loss": 0.0866,
|
| 100 |
+
"step": 650
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.23434884499497824,
|
| 104 |
+
"grad_norm": 0.37321263551712036,
|
| 105 |
+
"learning_rate": 4.806104229438679e-05,
|
| 106 |
+
"loss": 0.0884,
|
| 107 |
+
"step": 700
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.25108804820890523,
|
| 111 |
+
"grad_norm": 0.2660929262638092,
|
| 112 |
+
"learning_rate": 4.792154893427073e-05,
|
| 113 |
+
"loss": 0.0819,
|
| 114 |
+
"step": 750
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.26782725142283226,
|
| 118 |
+
"grad_norm": 0.2338525801897049,
|
| 119 |
+
"learning_rate": 4.778205557415467e-05,
|
| 120 |
+
"loss": 0.0845,
|
| 121 |
+
"step": 800
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.2845664546367593,
|
| 125 |
+
"grad_norm": 0.308557391166687,
|
| 126 |
+
"learning_rate": 4.764256221403862e-05,
|
| 127 |
+
"loss": 0.0815,
|
| 128 |
+
"step": 850
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.3013056578506863,
|
| 132 |
+
"grad_norm": 0.27098262310028076,
|
| 133 |
+
"learning_rate": 4.750306885392255e-05,
|
| 134 |
+
"loss": 0.0833,
|
| 135 |
+
"step": 900
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.3180448610646133,
|
| 139 |
+
"grad_norm": 0.23054952919483185,
|
| 140 |
+
"learning_rate": 4.736357549380649e-05,
|
| 141 |
+
"loss": 0.0806,
|
| 142 |
+
"step": 950
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.33478406427854035,
|
| 146 |
+
"grad_norm": 0.21355900168418884,
|
| 147 |
+
"learning_rate": 4.722408213369044e-05,
|
| 148 |
+
"loss": 0.073,
|
| 149 |
+
"step": 1000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.3515232674924674,
|
| 153 |
+
"grad_norm": 0.20395708084106445,
|
| 154 |
+
"learning_rate": 4.708458877357438e-05,
|
| 155 |
+
"loss": 0.0775,
|
| 156 |
+
"step": 1050
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.3682624707063944,
|
| 160 |
+
"grad_norm": 0.21063613891601562,
|
| 161 |
+
"learning_rate": 4.694509541345832e-05,
|
| 162 |
+
"loss": 0.0789,
|
| 163 |
+
"step": 1100
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.3850016739203214,
|
| 167 |
+
"grad_norm": 0.20589284598827362,
|
| 168 |
+
"learning_rate": 4.680560205334226e-05,
|
| 169 |
+
"loss": 0.0809,
|
| 170 |
+
"step": 1150
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.4017408771342484,
|
| 174 |
+
"grad_norm": 0.27975228428840637,
|
| 175 |
+
"learning_rate": 4.666610869322621e-05,
|
| 176 |
+
"loss": 0.078,
|
| 177 |
+
"step": 1200
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.4184800803481754,
|
| 181 |
+
"grad_norm": 0.2529745399951935,
|
| 182 |
+
"learning_rate": 4.6526615333110144e-05,
|
| 183 |
+
"loss": 0.0704,
|
| 184 |
+
"step": 1250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.43521928356210243,
|
| 188 |
+
"grad_norm": 0.2205154448747635,
|
| 189 |
+
"learning_rate": 4.6387121972994084e-05,
|
| 190 |
+
"loss": 0.0733,
|
| 191 |
+
"step": 1300
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.45195848677602946,
|
| 195 |
+
"grad_norm": 0.2254629135131836,
|
| 196 |
+
"learning_rate": 4.624762861287803e-05,
|
| 197 |
+
"loss": 0.0751,
|
| 198 |
+
"step": 1350
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.4686976899899565,
|
| 202 |
+
"grad_norm": 0.17614957690238953,
|
| 203 |
+
"learning_rate": 4.610813525276197e-05,
|
| 204 |
+
"loss": 0.0747,
|
| 205 |
+
"step": 1400
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.4854368932038835,
|
| 209 |
+
"grad_norm": 0.15940478444099426,
|
| 210 |
+
"learning_rate": 4.596864189264591e-05,
|
| 211 |
+
"loss": 0.0698,
|
| 212 |
+
"step": 1450
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5021760964178105,
|
| 216 |
+
"grad_norm": 0.1869521141052246,
|
| 217 |
+
"learning_rate": 4.5829148532529854e-05,
|
| 218 |
+
"loss": 0.0721,
|
| 219 |
+
"step": 1500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.5189152996317375,
|
| 223 |
+
"grad_norm": 0.36063650250434875,
|
| 224 |
+
"learning_rate": 4.5689655172413794e-05,
|
| 225 |
+
"loss": 0.0706,
|
| 226 |
+
"step": 1550
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.5356545028456645,
|
| 230 |
+
"grad_norm": 0.16967014968395233,
|
| 231 |
+
"learning_rate": 4.5550161812297735e-05,
|
| 232 |
+
"loss": 0.0759,
|
| 233 |
+
"step": 1600
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.5523937060595916,
|
| 237 |
+
"grad_norm": 0.29293423891067505,
|
| 238 |
+
"learning_rate": 4.5410668452181676e-05,
|
| 239 |
+
"loss": 0.0711,
|
| 240 |
+
"step": 1650
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.5691329092735186,
|
| 244 |
+
"grad_norm": 0.3034748136997223,
|
| 245 |
+
"learning_rate": 4.527117509206562e-05,
|
| 246 |
+
"loss": 0.067,
|
| 247 |
+
"step": 1700
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.5858721124874456,
|
| 251 |
+
"grad_norm": 0.1974593997001648,
|
| 252 |
+
"learning_rate": 4.513168173194956e-05,
|
| 253 |
+
"loss": 0.0701,
|
| 254 |
+
"step": 1750
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.6026113157013726,
|
| 258 |
+
"grad_norm": 0.18101799488067627,
|
| 259 |
+
"learning_rate": 4.4992188371833505e-05,
|
| 260 |
+
"loss": 0.0717,
|
| 261 |
+
"step": 1800
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.6193505189152997,
|
| 265 |
+
"grad_norm": 0.14422941207885742,
|
| 266 |
+
"learning_rate": 4.4852695011717445e-05,
|
| 267 |
+
"loss": 0.0686,
|
| 268 |
+
"step": 1850
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.6360897221292267,
|
| 272 |
+
"grad_norm": 0.28663551807403564,
|
| 273 |
+
"learning_rate": 4.4713201651601386e-05,
|
| 274 |
+
"loss": 0.0646,
|
| 275 |
+
"step": 1900
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.6528289253431536,
|
| 279 |
+
"grad_norm": 0.23879379034042358,
|
| 280 |
+
"learning_rate": 4.4573708291485327e-05,
|
| 281 |
+
"loss": 0.0684,
|
| 282 |
+
"step": 1950
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.6695681285570807,
|
| 286 |
+
"grad_norm": 0.21389362215995789,
|
| 287 |
+
"learning_rate": 4.443421493136927e-05,
|
| 288 |
+
"loss": 0.066,
|
| 289 |
+
"step": 2000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.6863073317710077,
|
| 293 |
+
"grad_norm": 0.26841893792152405,
|
| 294 |
+
"learning_rate": 4.4294721571253215e-05,
|
| 295 |
+
"loss": 0.0717,
|
| 296 |
+
"step": 2050
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.7030465349849347,
|
| 300 |
+
"grad_norm": 0.240205317735672,
|
| 301 |
+
"learning_rate": 4.415522821113715e-05,
|
| 302 |
+
"loss": 0.0697,
|
| 303 |
+
"step": 2100
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.7197857381988617,
|
| 307 |
+
"grad_norm": 0.28098127245903015,
|
| 308 |
+
"learning_rate": 4.4015734851021096e-05,
|
| 309 |
+
"loss": 0.0713,
|
| 310 |
+
"step": 2150
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.7365249414127888,
|
| 314 |
+
"grad_norm": 0.23308847844600677,
|
| 315 |
+
"learning_rate": 4.3876241490905037e-05,
|
| 316 |
+
"loss": 0.0667,
|
| 317 |
+
"step": 2200
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.7532641446267158,
|
| 321 |
+
"grad_norm": 0.22748568654060364,
|
| 322 |
+
"learning_rate": 4.373674813078898e-05,
|
| 323 |
+
"loss": 0.0605,
|
| 324 |
+
"step": 2250
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.7700033478406428,
|
| 328 |
+
"grad_norm": 0.3932187259197235,
|
| 329 |
+
"learning_rate": 4.359725477067292e-05,
|
| 330 |
+
"loss": 0.0676,
|
| 331 |
+
"step": 2300
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.7867425510545698,
|
| 335 |
+
"grad_norm": 0.23918767273426056,
|
| 336 |
+
"learning_rate": 4.345776141055686e-05,
|
| 337 |
+
"loss": 0.0624,
|
| 338 |
+
"step": 2350
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.8034817542684968,
|
| 342 |
+
"grad_norm": 0.3068426549434662,
|
| 343 |
+
"learning_rate": 4.33182680504408e-05,
|
| 344 |
+
"loss": 0.0664,
|
| 345 |
+
"step": 2400
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.8202209574824239,
|
| 349 |
+
"grad_norm": 0.17977873980998993,
|
| 350 |
+
"learning_rate": 4.317877469032474e-05,
|
| 351 |
+
"loss": 0.0726,
|
| 352 |
+
"step": 2450
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.8369601606963508,
|
| 356 |
+
"grad_norm": 0.16876642405986786,
|
| 357 |
+
"learning_rate": 4.303928133020869e-05,
|
| 358 |
+
"loss": 0.0639,
|
| 359 |
+
"step": 2500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.8536993639102779,
|
| 363 |
+
"grad_norm": 0.17980250716209412,
|
| 364 |
+
"learning_rate": 4.289978797009263e-05,
|
| 365 |
+
"loss": 0.0701,
|
| 366 |
+
"step": 2550
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.8704385671242049,
|
| 370 |
+
"grad_norm": 0.1711459904909134,
|
| 371 |
+
"learning_rate": 4.276029460997656e-05,
|
| 372 |
+
"loss": 0.063,
|
| 373 |
+
"step": 2600
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.8871777703381319,
|
| 377 |
+
"grad_norm": 0.443228542804718,
|
| 378 |
+
"learning_rate": 4.262080124986051e-05,
|
| 379 |
+
"loss": 0.0675,
|
| 380 |
+
"step": 2650
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.9039169735520589,
|
| 384 |
+
"grad_norm": 0.2098589390516281,
|
| 385 |
+
"learning_rate": 4.248130788974445e-05,
|
| 386 |
+
"loss": 0.062,
|
| 387 |
+
"step": 2700
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.920656176765986,
|
| 391 |
+
"grad_norm": 0.3022039234638214,
|
| 392 |
+
"learning_rate": 4.234181452962839e-05,
|
| 393 |
+
"loss": 0.07,
|
| 394 |
+
"step": 2750
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.937395379979913,
|
| 398 |
+
"grad_norm": 0.19368910789489746,
|
| 399 |
+
"learning_rate": 4.220232116951233e-05,
|
| 400 |
+
"loss": 0.0621,
|
| 401 |
+
"step": 2800
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.9541345831938399,
|
| 405 |
+
"grad_norm": 0.18753108382225037,
|
| 406 |
+
"learning_rate": 4.206282780939628e-05,
|
| 407 |
+
"loss": 0.0631,
|
| 408 |
+
"step": 2850
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.970873786407767,
|
| 412 |
+
"grad_norm": 0.15517786145210266,
|
| 413 |
+
"learning_rate": 4.192333444928022e-05,
|
| 414 |
+
"loss": 0.0641,
|
| 415 |
+
"step": 2900
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.987612989621694,
|
| 419 |
+
"grad_norm": 0.11765792220830917,
|
| 420 |
+
"learning_rate": 4.178384108916415e-05,
|
| 421 |
+
"loss": 0.0612,
|
| 422 |
+
"step": 2950
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.0,
|
| 426 |
+
"eval_loss": 0.05521286651492119,
|
| 427 |
+
"eval_runtime": 50.415,
|
| 428 |
+
"eval_samples_per_second": 236.993,
|
| 429 |
+
"eval_steps_per_second": 14.817,
|
| 430 |
+
"step": 2987
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.004352192835621,
|
| 434 |
+
"grad_norm": 0.2691793739795685,
|
| 435 |
+
"learning_rate": 4.16443477290481e-05,
|
| 436 |
+
"loss": 0.059,
|
| 437 |
+
"step": 3000
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.021091396049548,
|
| 441 |
+
"grad_norm": 0.394694060087204,
|
| 442 |
+
"learning_rate": 4.150485436893204e-05,
|
| 443 |
+
"loss": 0.0566,
|
| 444 |
+
"step": 3050
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.037830599263475,
|
| 448 |
+
"grad_norm": 0.19438503682613373,
|
| 449 |
+
"learning_rate": 4.136536100881598e-05,
|
| 450 |
+
"loss": 0.0591,
|
| 451 |
+
"step": 3100
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.0545698024774022,
|
| 455 |
+
"grad_norm": 0.21350933611392975,
|
| 456 |
+
"learning_rate": 4.122586764869992e-05,
|
| 457 |
+
"loss": 0.0509,
|
| 458 |
+
"step": 3150
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.071309005691329,
|
| 462 |
+
"grad_norm": 0.26747575402259827,
|
| 463 |
+
"learning_rate": 4.108637428858387e-05,
|
| 464 |
+
"loss": 0.0589,
|
| 465 |
+
"step": 3200
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.088048208905256,
|
| 469 |
+
"grad_norm": 0.31256961822509766,
|
| 470 |
+
"learning_rate": 4.0946880928467804e-05,
|
| 471 |
+
"loss": 0.0602,
|
| 472 |
+
"step": 3250
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.1047874121191832,
|
| 476 |
+
"grad_norm": 0.18631280958652496,
|
| 477 |
+
"learning_rate": 4.0807387568351745e-05,
|
| 478 |
+
"loss": 0.0547,
|
| 479 |
+
"step": 3300
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.12152661533311,
|
| 483 |
+
"grad_norm": 0.18677473068237305,
|
| 484 |
+
"learning_rate": 4.066789420823569e-05,
|
| 485 |
+
"loss": 0.0543,
|
| 486 |
+
"step": 3350
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.1382658185470371,
|
| 490 |
+
"grad_norm": 0.24535444378852844,
|
| 491 |
+
"learning_rate": 4.052840084811963e-05,
|
| 492 |
+
"loss": 0.0583,
|
| 493 |
+
"step": 3400
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.1550050217609642,
|
| 497 |
+
"grad_norm": 0.1752105951309204,
|
| 498 |
+
"learning_rate": 4.038890748800357e-05,
|
| 499 |
+
"loss": 0.0504,
|
| 500 |
+
"step": 3450
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.1717442249748913,
|
| 504 |
+
"grad_norm": 0.14743360877037048,
|
| 505 |
+
"learning_rate": 4.0249414127887514e-05,
|
| 506 |
+
"loss": 0.055,
|
| 507 |
+
"step": 3500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.1884834281888181,
|
| 511 |
+
"grad_norm": 0.11535945534706116,
|
| 512 |
+
"learning_rate": 4.010992076777146e-05,
|
| 513 |
+
"loss": 0.0552,
|
| 514 |
+
"step": 3550
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.2052226314027452,
|
| 518 |
+
"grad_norm": 0.26563358306884766,
|
| 519 |
+
"learning_rate": 3.9970427407655395e-05,
|
| 520 |
+
"loss": 0.0552,
|
| 521 |
+
"step": 3600
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.2219618346166723,
|
| 525 |
+
"grad_norm": 0.15104246139526367,
|
| 526 |
+
"learning_rate": 3.9830934047539336e-05,
|
| 527 |
+
"loss": 0.0575,
|
| 528 |
+
"step": 3650
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.2387010378305994,
|
| 532 |
+
"grad_norm": 0.2198421210050583,
|
| 533 |
+
"learning_rate": 3.9691440687423283e-05,
|
| 534 |
+
"loss": 0.0567,
|
| 535 |
+
"step": 3700
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.2554402410445262,
|
| 539 |
+
"grad_norm": 0.20177733898162842,
|
| 540 |
+
"learning_rate": 3.955194732730722e-05,
|
| 541 |
+
"loss": 0.0556,
|
| 542 |
+
"step": 3750
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.2721794442584533,
|
| 546 |
+
"grad_norm": 0.36604830622673035,
|
| 547 |
+
"learning_rate": 3.9412453967191165e-05,
|
| 548 |
+
"loss": 0.0569,
|
| 549 |
+
"step": 3800
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.2889186474723804,
|
| 553 |
+
"grad_norm": 0.18883727490901947,
|
| 554 |
+
"learning_rate": 3.9272960607075105e-05,
|
| 555 |
+
"loss": 0.0595,
|
| 556 |
+
"step": 3850
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.3056578506863072,
|
| 560 |
+
"grad_norm": 0.14828617870807648,
|
| 561 |
+
"learning_rate": 3.9133467246959046e-05,
|
| 562 |
+
"loss": 0.0548,
|
| 563 |
+
"step": 3900
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.3223970539002343,
|
| 567 |
+
"grad_norm": 0.19220437109470367,
|
| 568 |
+
"learning_rate": 3.899397388684299e-05,
|
| 569 |
+
"loss": 0.053,
|
| 570 |
+
"step": 3950
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.3391362571141614,
|
| 574 |
+
"grad_norm": 0.16049669682979584,
|
| 575 |
+
"learning_rate": 3.885448052672693e-05,
|
| 576 |
+
"loss": 0.0581,
|
| 577 |
+
"step": 4000
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.3558754603280883,
|
| 581 |
+
"grad_norm": 0.22821515798568726,
|
| 582 |
+
"learning_rate": 3.8714987166610875e-05,
|
| 583 |
+
"loss": 0.0518,
|
| 584 |
+
"step": 4050
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.3726146635420153,
|
| 588 |
+
"grad_norm": 0.1879580318927765,
|
| 589 |
+
"learning_rate": 3.857549380649481e-05,
|
| 590 |
+
"loss": 0.0574,
|
| 591 |
+
"step": 4100
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.3893538667559424,
|
| 595 |
+
"grad_norm": 0.16026251018047333,
|
| 596 |
+
"learning_rate": 3.8436000446378756e-05,
|
| 597 |
+
"loss": 0.063,
|
| 598 |
+
"step": 4150
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.4060930699698695,
|
| 602 |
+
"grad_norm": 0.26868143677711487,
|
| 603 |
+
"learning_rate": 3.82965070862627e-05,
|
| 604 |
+
"loss": 0.0571,
|
| 605 |
+
"step": 4200
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.4228322731837966,
|
| 609 |
+
"grad_norm": 0.2529687285423279,
|
| 610 |
+
"learning_rate": 3.815701372614664e-05,
|
| 611 |
+
"loss": 0.0528,
|
| 612 |
+
"step": 4250
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.4395714763977234,
|
| 616 |
+
"grad_norm": 0.19138221442699432,
|
| 617 |
+
"learning_rate": 3.801752036603058e-05,
|
| 618 |
+
"loss": 0.0584,
|
| 619 |
+
"step": 4300
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.4563106796116505,
|
| 623 |
+
"grad_norm": 0.16359661519527435,
|
| 624 |
+
"learning_rate": 3.787802700591452e-05,
|
| 625 |
+
"loss": 0.0539,
|
| 626 |
+
"step": 4350
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.4730498828255776,
|
| 630 |
+
"grad_norm": 0.1373494267463684,
|
| 631 |
+
"learning_rate": 3.7738533645798466e-05,
|
| 632 |
+
"loss": 0.0557,
|
| 633 |
+
"step": 4400
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.4897890860395044,
|
| 637 |
+
"grad_norm": 0.15695162117481232,
|
| 638 |
+
"learning_rate": 3.75990402856824e-05,
|
| 639 |
+
"loss": 0.0491,
|
| 640 |
+
"step": 4450
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.5065282892534315,
|
| 644 |
+
"grad_norm": 0.18462614715099335,
|
| 645 |
+
"learning_rate": 3.745954692556635e-05,
|
| 646 |
+
"loss": 0.0495,
|
| 647 |
+
"step": 4500
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.5232674924673586,
|
| 651 |
+
"grad_norm": 0.27876704931259155,
|
| 652 |
+
"learning_rate": 3.732005356545029e-05,
|
| 653 |
+
"loss": 0.0523,
|
| 654 |
+
"step": 4550
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.5400066956812855,
|
| 658 |
+
"grad_norm": 0.30491840839385986,
|
| 659 |
+
"learning_rate": 3.718056020533423e-05,
|
| 660 |
+
"loss": 0.0564,
|
| 661 |
+
"step": 4600
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.5567458988952128,
|
| 665 |
+
"grad_norm": 0.18721336126327515,
|
| 666 |
+
"learning_rate": 3.704106684521817e-05,
|
| 667 |
+
"loss": 0.0524,
|
| 668 |
+
"step": 4650
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.5734851021091396,
|
| 672 |
+
"grad_norm": 0.21216215193271637,
|
| 673 |
+
"learning_rate": 3.690157348510211e-05,
|
| 674 |
+
"loss": 0.0521,
|
| 675 |
+
"step": 4700
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.5902243053230665,
|
| 679 |
+
"grad_norm": 0.1368396282196045,
|
| 680 |
+
"learning_rate": 3.676208012498605e-05,
|
| 681 |
+
"loss": 0.056,
|
| 682 |
+
"step": 4750
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.6069635085369938,
|
| 686 |
+
"grad_norm": 0.13692086935043335,
|
| 687 |
+
"learning_rate": 3.662258676486999e-05,
|
| 688 |
+
"loss": 0.0443,
|
| 689 |
+
"step": 4800
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.6237027117509206,
|
| 693 |
+
"grad_norm": 0.11640128493309021,
|
| 694 |
+
"learning_rate": 3.648309340475394e-05,
|
| 695 |
+
"loss": 0.0488,
|
| 696 |
+
"step": 4850
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.6404419149648477,
|
| 700 |
+
"grad_norm": 0.19953882694244385,
|
| 701 |
+
"learning_rate": 3.634360004463788e-05,
|
| 702 |
+
"loss": 0.0553,
|
| 703 |
+
"step": 4900
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.6571811181787748,
|
| 707 |
+
"grad_norm": 0.1966984122991562,
|
| 708 |
+
"learning_rate": 3.6204106684521813e-05,
|
| 709 |
+
"loss": 0.0536,
|
| 710 |
+
"step": 4950
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.6739203213927016,
|
| 714 |
+
"grad_norm": 0.2324533313512802,
|
| 715 |
+
"learning_rate": 3.606461332440576e-05,
|
| 716 |
+
"loss": 0.0493,
|
| 717 |
+
"step": 5000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.6906595246066287,
|
| 721 |
+
"grad_norm": 0.16217607259750366,
|
| 722 |
+
"learning_rate": 3.59251199642897e-05,
|
| 723 |
+
"loss": 0.0503,
|
| 724 |
+
"step": 5050
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.7073987278205558,
|
| 728 |
+
"grad_norm": 0.23949602246284485,
|
| 729 |
+
"learning_rate": 3.578562660417364e-05,
|
| 730 |
+
"loss": 0.0556,
|
| 731 |
+
"step": 5100
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.7241379310344827,
|
| 735 |
+
"grad_norm": 0.21387897431850433,
|
| 736 |
+
"learning_rate": 3.564613324405758e-05,
|
| 737 |
+
"loss": 0.0548,
|
| 738 |
+
"step": 5150
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.7408771342484097,
|
| 742 |
+
"grad_norm": 0.2055111676454544,
|
| 743 |
+
"learning_rate": 3.550663988394153e-05,
|
| 744 |
+
"loss": 0.06,
|
| 745 |
+
"step": 5200
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.7576163374623368,
|
| 749 |
+
"grad_norm": 0.20280921459197998,
|
| 750 |
+
"learning_rate": 3.5367146523825464e-05,
|
| 751 |
+
"loss": 0.0508,
|
| 752 |
+
"step": 5250
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.7743555406762637,
|
| 756 |
+
"grad_norm": 0.14165103435516357,
|
| 757 |
+
"learning_rate": 3.5227653163709405e-05,
|
| 758 |
+
"loss": 0.0581,
|
| 759 |
+
"step": 5300
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.791094743890191,
|
| 763 |
+
"grad_norm": 0.18099863827228546,
|
| 764 |
+
"learning_rate": 3.508815980359335e-05,
|
| 765 |
+
"loss": 0.0562,
|
| 766 |
+
"step": 5350
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.8078339471041178,
|
| 770 |
+
"grad_norm": 0.21743184328079224,
|
| 771 |
+
"learning_rate": 3.494866644347729e-05,
|
| 772 |
+
"loss": 0.0498,
|
| 773 |
+
"step": 5400
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.824573150318045,
|
| 777 |
+
"grad_norm": 0.20934534072875977,
|
| 778 |
+
"learning_rate": 3.4809173083361234e-05,
|
| 779 |
+
"loss": 0.0549,
|
| 780 |
+
"step": 5450
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.841312353531972,
|
| 784 |
+
"grad_norm": 0.1582174152135849,
|
| 785 |
+
"learning_rate": 3.4669679723245174e-05,
|
| 786 |
+
"loss": 0.0556,
|
| 787 |
+
"step": 5500
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.8580515567458988,
|
| 791 |
+
"grad_norm": 0.1624903827905655,
|
| 792 |
+
"learning_rate": 3.453018636312912e-05,
|
| 793 |
+
"loss": 0.0516,
|
| 794 |
+
"step": 5550
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.874790759959826,
|
| 798 |
+
"grad_norm": 0.16255798935890198,
|
| 799 |
+
"learning_rate": 3.4390693003013056e-05,
|
| 800 |
+
"loss": 0.0542,
|
| 801 |
+
"step": 5600
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.891529963173753,
|
| 805 |
+
"grad_norm": 0.1269742250442505,
|
| 806 |
+
"learning_rate": 3.4251199642896996e-05,
|
| 807 |
+
"loss": 0.0565,
|
| 808 |
+
"step": 5650
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.9082691663876798,
|
| 812 |
+
"grad_norm": 0.15966229140758514,
|
| 813 |
+
"learning_rate": 3.4111706282780944e-05,
|
| 814 |
+
"loss": 0.0538,
|
| 815 |
+
"step": 5700
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.925008369601607,
|
| 819 |
+
"grad_norm": 0.21506330370903015,
|
| 820 |
+
"learning_rate": 3.3972212922664884e-05,
|
| 821 |
+
"loss": 0.0505,
|
| 822 |
+
"step": 5750
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.941747572815534,
|
| 826 |
+
"grad_norm": 0.2145415097475052,
|
| 827 |
+
"learning_rate": 3.3832719562548825e-05,
|
| 828 |
+
"loss": 0.0521,
|
| 829 |
+
"step": 5800
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.9584867760294609,
|
| 833 |
+
"grad_norm": 0.10960496962070465,
|
| 834 |
+
"learning_rate": 3.3693226202432766e-05,
|
| 835 |
+
"loss": 0.0513,
|
| 836 |
+
"step": 5850
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.9752259792433882,
|
| 840 |
+
"grad_norm": 0.13635843992233276,
|
| 841 |
+
"learning_rate": 3.355373284231671e-05,
|
| 842 |
+
"loss": 0.0499,
|
| 843 |
+
"step": 5900
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.991965182457315,
|
| 847 |
+
"grad_norm": 0.1542210429906845,
|
| 848 |
+
"learning_rate": 3.341423948220065e-05,
|
| 849 |
+
"loss": 0.0556,
|
| 850 |
+
"step": 5950
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 2.0,
|
| 854 |
+
"eval_loss": 0.04946442320942879,
|
| 855 |
+
"eval_runtime": 55.6114,
|
| 856 |
+
"eval_samples_per_second": 214.848,
|
| 857 |
+
"eval_steps_per_second": 13.432,
|
| 858 |
+
"step": 5974
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 2.008704385671242,
|
| 862 |
+
"grad_norm": 0.1718842089176178,
|
| 863 |
+
"learning_rate": 3.327474612208459e-05,
|
| 864 |
+
"loss": 0.0503,
|
| 865 |
+
"step": 6000
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 2.025443588885169,
|
| 869 |
+
"grad_norm": 0.1528020203113556,
|
| 870 |
+
"learning_rate": 3.3135252761968535e-05,
|
| 871 |
+
"loss": 0.0479,
|
| 872 |
+
"step": 6050
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 2.042182792099096,
|
| 876 |
+
"grad_norm": 0.19148772954940796,
|
| 877 |
+
"learning_rate": 3.299575940185247e-05,
|
| 878 |
+
"loss": 0.0443,
|
| 879 |
+
"step": 6100
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 2.058921995313023,
|
| 883 |
+
"grad_norm": 0.18125496804714203,
|
| 884 |
+
"learning_rate": 3.2856266041736416e-05,
|
| 885 |
+
"loss": 0.0473,
|
| 886 |
+
"step": 6150
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 2.07566119852695,
|
| 890 |
+
"grad_norm": 0.20772996544837952,
|
| 891 |
+
"learning_rate": 3.271677268162036e-05,
|
| 892 |
+
"loss": 0.0539,
|
| 893 |
+
"step": 6200
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 2.092400401740877,
|
| 897 |
+
"grad_norm": 0.2518468201160431,
|
| 898 |
+
"learning_rate": 3.25772793215043e-05,
|
| 899 |
+
"loss": 0.0468,
|
| 900 |
+
"step": 6250
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 2.1091396049548043,
|
| 904 |
+
"grad_norm": 0.1350301206111908,
|
| 905 |
+
"learning_rate": 3.243778596138824e-05,
|
| 906 |
+
"loss": 0.0431,
|
| 907 |
+
"step": 6300
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 2.125878808168731,
|
| 911 |
+
"grad_norm": 0.19141735136508942,
|
| 912 |
+
"learning_rate": 3.229829260127218e-05,
|
| 913 |
+
"loss": 0.048,
|
| 914 |
+
"step": 6350
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 2.142618011382658,
|
| 918 |
+
"grad_norm": 0.2404586374759674,
|
| 919 |
+
"learning_rate": 3.2158799241156126e-05,
|
| 920 |
+
"loss": 0.0516,
|
| 921 |
+
"step": 6400
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 2.1593572145965854,
|
| 925 |
+
"grad_norm": 0.21710112690925598,
|
| 926 |
+
"learning_rate": 3.201930588104006e-05,
|
| 927 |
+
"loss": 0.0472,
|
| 928 |
+
"step": 6450
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 2.176096417810512,
|
| 932 |
+
"grad_norm": 0.14395031332969666,
|
| 933 |
+
"learning_rate": 3.187981252092401e-05,
|
| 934 |
+
"loss": 0.0438,
|
| 935 |
+
"step": 6500
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 2.192835621024439,
|
| 939 |
+
"grad_norm": 0.20882932841777802,
|
| 940 |
+
"learning_rate": 3.174031916080795e-05,
|
| 941 |
+
"loss": 0.0488,
|
| 942 |
+
"step": 6550
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 2.2095748242383664,
|
| 946 |
+
"grad_norm": 0.13824905455112457,
|
| 947 |
+
"learning_rate": 3.160082580069189e-05,
|
| 948 |
+
"loss": 0.0464,
|
| 949 |
+
"step": 6600
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 2.2263140274522932,
|
| 953 |
+
"grad_norm": 0.1783577799797058,
|
| 954 |
+
"learning_rate": 3.146133244057583e-05,
|
| 955 |
+
"loss": 0.0459,
|
| 956 |
+
"step": 6650
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 2.24305323066622,
|
| 960 |
+
"grad_norm": 0.22147531807422638,
|
| 961 |
+
"learning_rate": 3.132183908045977e-05,
|
| 962 |
+
"loss": 0.0476,
|
| 963 |
+
"step": 6700
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 2.2597924338801474,
|
| 967 |
+
"grad_norm": 0.17393821477890015,
|
| 968 |
+
"learning_rate": 3.118234572034371e-05,
|
| 969 |
+
"loss": 0.0436,
|
| 970 |
+
"step": 6750
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 2.2765316370940742,
|
| 974 |
+
"grad_norm": 0.15850785374641418,
|
| 975 |
+
"learning_rate": 3.104285236022765e-05,
|
| 976 |
+
"loss": 0.0476,
|
| 977 |
+
"step": 6800
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 2.2932708403080015,
|
| 981 |
+
"grad_norm": 0.16232182085514069,
|
| 982 |
+
"learning_rate": 3.09033590001116e-05,
|
| 983 |
+
"loss": 0.0473,
|
| 984 |
+
"step": 6850
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 2.3100100435219284,
|
| 988 |
+
"grad_norm": 0.1816001981496811,
|
| 989 |
+
"learning_rate": 3.076386563999554e-05,
|
| 990 |
+
"loss": 0.0427,
|
| 991 |
+
"step": 6900
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 2.3267492467358553,
|
| 995 |
+
"grad_norm": 0.13417834043502808,
|
| 996 |
+
"learning_rate": 3.062437227987948e-05,
|
| 997 |
+
"loss": 0.0448,
|
| 998 |
+
"step": 6950
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 2.3434884499497826,
|
| 1002 |
+
"grad_norm": 0.12576530873775482,
|
| 1003 |
+
"learning_rate": 3.048487891976342e-05,
|
| 1004 |
+
"loss": 0.0453,
|
| 1005 |
+
"step": 7000
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 2.3602276531637094,
|
| 1009 |
+
"grad_norm": 0.33120718598365784,
|
| 1010 |
+
"learning_rate": 3.0345385559647362e-05,
|
| 1011 |
+
"loss": 0.0462,
|
| 1012 |
+
"step": 7050
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 2.3769668563776363,
|
| 1016 |
+
"grad_norm": 0.22310969233512878,
|
| 1017 |
+
"learning_rate": 3.0205892199531306e-05,
|
| 1018 |
+
"loss": 0.0475,
|
| 1019 |
+
"step": 7100
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 2.3937060595915636,
|
| 1023 |
+
"grad_norm": 0.18150626122951508,
|
| 1024 |
+
"learning_rate": 3.0066398839415243e-05,
|
| 1025 |
+
"loss": 0.0489,
|
| 1026 |
+
"step": 7150
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 2.4104452628054904,
|
| 1030 |
+
"grad_norm": 0.28730452060699463,
|
| 1031 |
+
"learning_rate": 2.9926905479299187e-05,
|
| 1032 |
+
"loss": 0.0536,
|
| 1033 |
+
"step": 7200
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 2.4271844660194173,
|
| 1037 |
+
"grad_norm": 0.1918480098247528,
|
| 1038 |
+
"learning_rate": 2.9787412119183128e-05,
|
| 1039 |
+
"loss": 0.0426,
|
| 1040 |
+
"step": 7250
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 2.4439236692333446,
|
| 1044 |
+
"grad_norm": 0.16158398985862732,
|
| 1045 |
+
"learning_rate": 2.964791875906707e-05,
|
| 1046 |
+
"loss": 0.0458,
|
| 1047 |
+
"step": 7300
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 2.4606628724472714,
|
| 1051 |
+
"grad_norm": 0.27141231298446655,
|
| 1052 |
+
"learning_rate": 2.9508425398951012e-05,
|
| 1053 |
+
"loss": 0.0454,
|
| 1054 |
+
"step": 7350
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 2.4774020756611987,
|
| 1058 |
+
"grad_norm": 0.1777345836162567,
|
| 1059 |
+
"learning_rate": 2.936893203883495e-05,
|
| 1060 |
+
"loss": 0.0435,
|
| 1061 |
+
"step": 7400
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 2.4941412788751256,
|
| 1065 |
+
"grad_norm": 0.14735421538352966,
|
| 1066 |
+
"learning_rate": 2.9229438678718897e-05,
|
| 1067 |
+
"loss": 0.0489,
|
| 1068 |
+
"step": 7450
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 2.5108804820890525,
|
| 1072 |
+
"grad_norm": 0.1486055999994278,
|
| 1073 |
+
"learning_rate": 2.9089945318602834e-05,
|
| 1074 |
+
"loss": 0.0477,
|
| 1075 |
+
"step": 7500
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 2.5276196853029793,
|
| 1079 |
+
"grad_norm": 0.17078754305839539,
|
| 1080 |
+
"learning_rate": 2.895045195848678e-05,
|
| 1081 |
+
"loss": 0.0444,
|
| 1082 |
+
"step": 7550
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 2.5443588885169066,
|
| 1086 |
+
"grad_norm": 0.19276435673236847,
|
| 1087 |
+
"learning_rate": 2.881095859837072e-05,
|
| 1088 |
+
"loss": 0.0486,
|
| 1089 |
+
"step": 7600
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 2.5610980917308335,
|
| 1093 |
+
"grad_norm": 0.21209606528282166,
|
| 1094 |
+
"learning_rate": 2.8671465238254656e-05,
|
| 1095 |
+
"loss": 0.0497,
|
| 1096 |
+
"step": 7650
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 2.5778372949447608,
|
| 1100 |
+
"grad_norm": 0.21018877625465393,
|
| 1101 |
+
"learning_rate": 2.8531971878138604e-05,
|
| 1102 |
+
"loss": 0.0441,
|
| 1103 |
+
"step": 7700
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 2.5945764981586876,
|
| 1107 |
+
"grad_norm": 0.15666617453098297,
|
| 1108 |
+
"learning_rate": 2.839247851802254e-05,
|
| 1109 |
+
"loss": 0.0467,
|
| 1110 |
+
"step": 7750
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 2.6113157013726145,
|
| 1114 |
+
"grad_norm": 0.1940685212612152,
|
| 1115 |
+
"learning_rate": 2.8252985157906485e-05,
|
| 1116 |
+
"loss": 0.0523,
|
| 1117 |
+
"step": 7800
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 2.628054904586542,
|
| 1121 |
+
"grad_norm": 0.28480586409568787,
|
| 1122 |
+
"learning_rate": 2.8113491797790426e-05,
|
| 1123 |
+
"loss": 0.0481,
|
| 1124 |
+
"step": 7850
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 2.6447941078004686,
|
| 1128 |
+
"grad_norm": 0.2223973125219345,
|
| 1129 |
+
"learning_rate": 2.797399843767437e-05,
|
| 1130 |
+
"loss": 0.0432,
|
| 1131 |
+
"step": 7900
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 2.661533311014396,
|
| 1135 |
+
"grad_norm": 0.15986157953739166,
|
| 1136 |
+
"learning_rate": 2.783450507755831e-05,
|
| 1137 |
+
"loss": 0.0454,
|
| 1138 |
+
"step": 7950
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 2.678272514228323,
|
| 1142 |
+
"grad_norm": 0.1384258270263672,
|
| 1143 |
+
"learning_rate": 2.7695011717442248e-05,
|
| 1144 |
+
"loss": 0.0477,
|
| 1145 |
+
"step": 8000
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 2.6950117174422497,
|
| 1149 |
+
"grad_norm": 0.1721869707107544,
|
| 1150 |
+
"learning_rate": 2.7555518357326192e-05,
|
| 1151 |
+
"loss": 0.0453,
|
| 1152 |
+
"step": 8050
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 2.7117509206561765,
|
| 1156 |
+
"grad_norm": 0.20737840235233307,
|
| 1157 |
+
"learning_rate": 2.7416024997210132e-05,
|
| 1158 |
+
"loss": 0.0504,
|
| 1159 |
+
"step": 8100
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 2.728490123870104,
|
| 1163 |
+
"grad_norm": 0.18823584914207458,
|
| 1164 |
+
"learning_rate": 2.7276531637094077e-05,
|
| 1165 |
+
"loss": 0.0453,
|
| 1166 |
+
"step": 8150
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 2.7452293270840307,
|
| 1170 |
+
"grad_norm": 0.13201962411403656,
|
| 1171 |
+
"learning_rate": 2.7137038276978017e-05,
|
| 1172 |
+
"loss": 0.0433,
|
| 1173 |
+
"step": 8200
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 2.761968530297958,
|
| 1177 |
+
"grad_norm": 0.1443973183631897,
|
| 1178 |
+
"learning_rate": 2.699754491686196e-05,
|
| 1179 |
+
"loss": 0.0486,
|
| 1180 |
+
"step": 8250
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 2.778707733511885,
|
| 1184 |
+
"grad_norm": 0.29314514994621277,
|
| 1185 |
+
"learning_rate": 2.68580515567459e-05,
|
| 1186 |
+
"loss": 0.05,
|
| 1187 |
+
"step": 8300
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 2.7954469367258117,
|
| 1191 |
+
"grad_norm": 0.14852124452590942,
|
| 1192 |
+
"learning_rate": 2.671855819662984e-05,
|
| 1193 |
+
"loss": 0.0495,
|
| 1194 |
+
"step": 8350
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 2.812186139939739,
|
| 1198 |
+
"grad_norm": 0.19024662673473358,
|
| 1199 |
+
"learning_rate": 2.6579064836513783e-05,
|
| 1200 |
+
"loss": 0.0508,
|
| 1201 |
+
"step": 8400
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 2.828925343153666,
|
| 1205 |
+
"grad_norm": 0.1745578795671463,
|
| 1206 |
+
"learning_rate": 2.6439571476397724e-05,
|
| 1207 |
+
"loss": 0.0443,
|
| 1208 |
+
"step": 8450
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 2.845664546367593,
|
| 1212 |
+
"grad_norm": 0.18390017747879028,
|
| 1213 |
+
"learning_rate": 2.6300078116281668e-05,
|
| 1214 |
+
"loss": 0.0468,
|
| 1215 |
+
"step": 8500
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 2.86240374958152,
|
| 1219 |
+
"grad_norm": 0.22483347356319427,
|
| 1220 |
+
"learning_rate": 2.616058475616561e-05,
|
| 1221 |
+
"loss": 0.0467,
|
| 1222 |
+
"step": 8550
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 2.879142952795447,
|
| 1226 |
+
"grad_norm": 0.18160563707351685,
|
| 1227 |
+
"learning_rate": 2.6021091396049553e-05,
|
| 1228 |
+
"loss": 0.0441,
|
| 1229 |
+
"step": 8600
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 2.8958821560093737,
|
| 1233 |
+
"grad_norm": 0.13408955931663513,
|
| 1234 |
+
"learning_rate": 2.588159803593349e-05,
|
| 1235 |
+
"loss": 0.0446,
|
| 1236 |
+
"step": 8650
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 2.912621359223301,
|
| 1240 |
+
"grad_norm": 0.16038326919078827,
|
| 1241 |
+
"learning_rate": 2.574210467581743e-05,
|
| 1242 |
+
"loss": 0.0456,
|
| 1243 |
+
"step": 8700
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 2.929360562437228,
|
| 1247 |
+
"grad_norm": 0.22738413512706757,
|
| 1248 |
+
"learning_rate": 2.5602611315701375e-05,
|
| 1249 |
+
"loss": 0.0479,
|
| 1250 |
+
"step": 8750
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 2.946099765651155,
|
| 1254 |
+
"grad_norm": 0.20327210426330566,
|
| 1255 |
+
"learning_rate": 2.5463117955585315e-05,
|
| 1256 |
+
"loss": 0.0511,
|
| 1257 |
+
"step": 8800
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 2.962838968865082,
|
| 1261 |
+
"grad_norm": 0.15756353735923767,
|
| 1262 |
+
"learning_rate": 2.532362459546926e-05,
|
| 1263 |
+
"loss": 0.0426,
|
| 1264 |
+
"step": 8850
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 2.979578172079009,
|
| 1268 |
+
"grad_norm": 0.1305045783519745,
|
| 1269 |
+
"learning_rate": 2.5184131235353197e-05,
|
| 1270 |
+
"loss": 0.0442,
|
| 1271 |
+
"step": 8900
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 2.996317375292936,
|
| 1275 |
+
"grad_norm": 0.1610562801361084,
|
| 1276 |
+
"learning_rate": 2.5044637875237144e-05,
|
| 1277 |
+
"loss": 0.0467,
|
| 1278 |
+
"step": 8950
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 3.0,
|
| 1282 |
+
"eval_loss": 0.04702676460146904,
|
| 1283 |
+
"eval_runtime": 52.8989,
|
| 1284 |
+
"eval_samples_per_second": 225.865,
|
| 1285 |
+
"eval_steps_per_second": 14.121,
|
| 1286 |
+
"step": 8961
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 3.013056578506863,
|
| 1290 |
+
"grad_norm": 0.2042045295238495,
|
| 1291 |
+
"learning_rate": 2.490514451512108e-05,
|
| 1292 |
+
"loss": 0.0457,
|
| 1293 |
+
"step": 9000
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 3.02979578172079,
|
| 1297 |
+
"grad_norm": 0.27092283964157104,
|
| 1298 |
+
"learning_rate": 2.4765651155005022e-05,
|
| 1299 |
+
"loss": 0.0437,
|
| 1300 |
+
"step": 9050
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 3.046534984934717,
|
| 1304 |
+
"grad_norm": 0.18729308247566223,
|
| 1305 |
+
"learning_rate": 2.4626157794888966e-05,
|
| 1306 |
+
"loss": 0.039,
|
| 1307 |
+
"step": 9100
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 3.063274188148644,
|
| 1311 |
+
"grad_norm": 0.18845289945602417,
|
| 1312 |
+
"learning_rate": 2.4486664434772903e-05,
|
| 1313 |
+
"loss": 0.0422,
|
| 1314 |
+
"step": 9150
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 3.0800133913625714,
|
| 1318 |
+
"grad_norm": 0.17593874037265778,
|
| 1319 |
+
"learning_rate": 2.4347171074656847e-05,
|
| 1320 |
+
"loss": 0.0384,
|
| 1321 |
+
"step": 9200
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 3.096752594576498,
|
| 1325 |
+
"grad_norm": 0.23149123787879944,
|
| 1326 |
+
"learning_rate": 2.4207677714540788e-05,
|
| 1327 |
+
"loss": 0.0443,
|
| 1328 |
+
"step": 9250
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 3.113491797790425,
|
| 1332 |
+
"grad_norm": 0.1968075931072235,
|
| 1333 |
+
"learning_rate": 2.4068184354424732e-05,
|
| 1334 |
+
"loss": 0.0402,
|
| 1335 |
+
"step": 9300
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 3.1302310010043524,
|
| 1339 |
+
"grad_norm": 0.16893354058265686,
|
| 1340 |
+
"learning_rate": 2.3928690994308673e-05,
|
| 1341 |
+
"loss": 0.043,
|
| 1342 |
+
"step": 9350
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 3.146970204218279,
|
| 1346 |
+
"grad_norm": 0.161103293299675,
|
| 1347 |
+
"learning_rate": 2.3789197634192613e-05,
|
| 1348 |
+
"loss": 0.0414,
|
| 1349 |
+
"step": 9400
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 3.163709407432206,
|
| 1353 |
+
"grad_norm": 0.22712625563144684,
|
| 1354 |
+
"learning_rate": 2.3649704274076557e-05,
|
| 1355 |
+
"loss": 0.0426,
|
| 1356 |
+
"step": 9450
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 3.1804486106461334,
|
| 1360 |
+
"grad_norm": 0.12811464071273804,
|
| 1361 |
+
"learning_rate": 2.3510210913960495e-05,
|
| 1362 |
+
"loss": 0.0406,
|
| 1363 |
+
"step": 9500
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 3.1971878138600602,
|
| 1367 |
+
"grad_norm": 0.16762731969356537,
|
| 1368 |
+
"learning_rate": 2.337071755384444e-05,
|
| 1369 |
+
"loss": 0.0417,
|
| 1370 |
+
"step": 9550
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 3.213927017073987,
|
| 1374 |
+
"grad_norm": 0.14003124833106995,
|
| 1375 |
+
"learning_rate": 2.323122419372838e-05,
|
| 1376 |
+
"loss": 0.0453,
|
| 1377 |
+
"step": 9600
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 3.2306662202879144,
|
| 1381 |
+
"grad_norm": 0.16891060769557953,
|
| 1382 |
+
"learning_rate": 2.309173083361232e-05,
|
| 1383 |
+
"loss": 0.0445,
|
| 1384 |
+
"step": 9650
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 3.2474054235018412,
|
| 1388 |
+
"grad_norm": 0.16900208592414856,
|
| 1389 |
+
"learning_rate": 2.2952237473496264e-05,
|
| 1390 |
+
"loss": 0.039,
|
| 1391 |
+
"step": 9700
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 3.264144626715768,
|
| 1395 |
+
"grad_norm": 0.17154955863952637,
|
| 1396 |
+
"learning_rate": 2.2812744113380205e-05,
|
| 1397 |
+
"loss": 0.0403,
|
| 1398 |
+
"step": 9750
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 3.2808838299296954,
|
| 1402 |
+
"grad_norm": 0.1620296835899353,
|
| 1403 |
+
"learning_rate": 2.2673250753264145e-05,
|
| 1404 |
+
"loss": 0.0406,
|
| 1405 |
+
"step": 9800
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 3.2976230331436223,
|
| 1409 |
+
"grad_norm": 0.14487063884735107,
|
| 1410 |
+
"learning_rate": 2.2533757393148086e-05,
|
| 1411 |
+
"loss": 0.0467,
|
| 1412 |
+
"step": 9850
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 3.3143622363575496,
|
| 1416 |
+
"grad_norm": 0.13799156248569489,
|
| 1417 |
+
"learning_rate": 2.2394264033032027e-05,
|
| 1418 |
+
"loss": 0.0433,
|
| 1419 |
+
"step": 9900
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 3.3311014395714764,
|
| 1423 |
+
"grad_norm": 0.1507265418767929,
|
| 1424 |
+
"learning_rate": 2.225477067291597e-05,
|
| 1425 |
+
"loss": 0.0446,
|
| 1426 |
+
"step": 9950
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 3.3478406427854033,
|
| 1430 |
+
"grad_norm": 0.1605840027332306,
|
| 1431 |
+
"learning_rate": 2.211527731279991e-05,
|
| 1432 |
+
"loss": 0.0415,
|
| 1433 |
+
"step": 10000
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 3.3645798459993306,
|
| 1437 |
+
"grad_norm": 0.11910756677389145,
|
| 1438 |
+
"learning_rate": 2.1975783952683855e-05,
|
| 1439 |
+
"loss": 0.0453,
|
| 1440 |
+
"step": 10050
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 3.3813190492132574,
|
| 1444 |
+
"grad_norm": 0.15115414559841156,
|
| 1445 |
+
"learning_rate": 2.1836290592567796e-05,
|
| 1446 |
+
"loss": 0.0395,
|
| 1447 |
+
"step": 10100
|
| 1448 |
+
},
|
| 1449 |
+
{
|
| 1450 |
+
"epoch": 3.3980582524271843,
|
| 1451 |
+
"grad_norm": 0.2044568508863449,
|
| 1452 |
+
"learning_rate": 2.1696797232451737e-05,
|
| 1453 |
+
"loss": 0.0473,
|
| 1454 |
+
"step": 10150
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 3.4147974556411116,
|
| 1458 |
+
"grad_norm": 0.18123552203178406,
|
| 1459 |
+
"learning_rate": 2.1557303872335677e-05,
|
| 1460 |
+
"loss": 0.0411,
|
| 1461 |
+
"step": 10200
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 3.4315366588550384,
|
| 1465 |
+
"grad_norm": 0.6018120646476746,
|
| 1466 |
+
"learning_rate": 2.1417810512219618e-05,
|
| 1467 |
+
"loss": 0.0413,
|
| 1468 |
+
"step": 10250
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 3.4482758620689653,
|
| 1472 |
+
"grad_norm": 0.22490189969539642,
|
| 1473 |
+
"learning_rate": 2.1278317152103562e-05,
|
| 1474 |
+
"loss": 0.0437,
|
| 1475 |
+
"step": 10300
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 3.4650150652828926,
|
| 1479 |
+
"grad_norm": 0.1784990429878235,
|
| 1480 |
+
"learning_rate": 2.1138823791987503e-05,
|
| 1481 |
+
"loss": 0.0453,
|
| 1482 |
+
"step": 10350
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 3.4817542684968195,
|
| 1486 |
+
"grad_norm": 0.15248402953147888,
|
| 1487 |
+
"learning_rate": 2.0999330431871443e-05,
|
| 1488 |
+
"loss": 0.0456,
|
| 1489 |
+
"step": 10400
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 3.4984934717107468,
|
| 1493 |
+
"grad_norm": 0.15146291255950928,
|
| 1494 |
+
"learning_rate": 2.0859837071755387e-05,
|
| 1495 |
+
"loss": 0.0393,
|
| 1496 |
+
"step": 10450
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 3.5152326749246736,
|
| 1500 |
+
"grad_norm": 0.1662750393152237,
|
| 1501 |
+
"learning_rate": 2.0720343711639325e-05,
|
| 1502 |
+
"loss": 0.0408,
|
| 1503 |
+
"step": 10500
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 3.5319718781386005,
|
| 1507 |
+
"grad_norm": 0.1244506984949112,
|
| 1508 |
+
"learning_rate": 2.058085035152327e-05,
|
| 1509 |
+
"loss": 0.0426,
|
| 1510 |
+
"step": 10550
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 3.5487110813525278,
|
| 1514 |
+
"grad_norm": 0.1180344969034195,
|
| 1515 |
+
"learning_rate": 2.044135699140721e-05,
|
| 1516 |
+
"loss": 0.0434,
|
| 1517 |
+
"step": 10600
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 3.5654502845664546,
|
| 1521 |
+
"grad_norm": 0.15951013565063477,
|
| 1522 |
+
"learning_rate": 2.030186363129115e-05,
|
| 1523 |
+
"loss": 0.0387,
|
| 1524 |
+
"step": 10650
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 3.582189487780382,
|
| 1528 |
+
"grad_norm": 0.16064217686653137,
|
| 1529 |
+
"learning_rate": 2.0162370271175094e-05,
|
| 1530 |
+
"loss": 0.0445,
|
| 1531 |
+
"step": 10700
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 3.598928690994309,
|
| 1535 |
+
"grad_norm": 0.17813698947429657,
|
| 1536 |
+
"learning_rate": 2.0022876911059035e-05,
|
| 1537 |
+
"loss": 0.0407,
|
| 1538 |
+
"step": 10750
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 3.6156678942082356,
|
| 1542 |
+
"grad_norm": 0.1256450116634369,
|
| 1543 |
+
"learning_rate": 1.988338355094298e-05,
|
| 1544 |
+
"loss": 0.0462,
|
| 1545 |
+
"step": 10800
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 3.6324070974221625,
|
| 1549 |
+
"grad_norm": 0.14016403257846832,
|
| 1550 |
+
"learning_rate": 1.9743890190826916e-05,
|
| 1551 |
+
"loss": 0.0456,
|
| 1552 |
+
"step": 10850
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 3.64914630063609,
|
| 1556 |
+
"grad_norm": 0.1396850347518921,
|
| 1557 |
+
"learning_rate": 1.9604396830710857e-05,
|
| 1558 |
+
"loss": 0.0403,
|
| 1559 |
+
"step": 10900
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 3.6658855038500167,
|
| 1563 |
+
"grad_norm": 0.17943057417869568,
|
| 1564 |
+
"learning_rate": 1.94649034705948e-05,
|
| 1565 |
+
"loss": 0.0417,
|
| 1566 |
+
"step": 10950
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 3.682624707063944,
|
| 1570 |
+
"grad_norm": 0.14947953820228577,
|
| 1571 |
+
"learning_rate": 1.932541011047874e-05,
|
| 1572 |
+
"loss": 0.0421,
|
| 1573 |
+
"step": 11000
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 3.699363910277871,
|
| 1577 |
+
"grad_norm": 0.12628613412380219,
|
| 1578 |
+
"learning_rate": 1.9185916750362685e-05,
|
| 1579 |
+
"loss": 0.0435,
|
| 1580 |
+
"step": 11050
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 3.7161031134917977,
|
| 1584 |
+
"grad_norm": 0.2205984890460968,
|
| 1585 |
+
"learning_rate": 1.9046423390246626e-05,
|
| 1586 |
+
"loss": 0.0396,
|
| 1587 |
+
"step": 11100
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 3.732842316705725,
|
| 1591 |
+
"grad_norm": 0.13236357271671295,
|
| 1592 |
+
"learning_rate": 1.8906930030130567e-05,
|
| 1593 |
+
"loss": 0.0405,
|
| 1594 |
+
"step": 11150
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 3.749581519919652,
|
| 1598 |
+
"grad_norm": 0.15023528039455414,
|
| 1599 |
+
"learning_rate": 1.8767436670014507e-05,
|
| 1600 |
+
"loss": 0.0434,
|
| 1601 |
+
"step": 11200
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 3.7663207231335787,
|
| 1605 |
+
"grad_norm": 0.1427326649427414,
|
| 1606 |
+
"learning_rate": 1.8627943309898448e-05,
|
| 1607 |
+
"loss": 0.0437,
|
| 1608 |
+
"step": 11250
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 3.783059926347506,
|
| 1612 |
+
"grad_norm": 0.1890624761581421,
|
| 1613 |
+
"learning_rate": 1.8488449949782392e-05,
|
| 1614 |
+
"loss": 0.0408,
|
| 1615 |
+
"step": 11300
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 3.799799129561433,
|
| 1619 |
+
"grad_norm": 0.27970972657203674,
|
| 1620 |
+
"learning_rate": 1.8348956589666333e-05,
|
| 1621 |
+
"loss": 0.0441,
|
| 1622 |
+
"step": 11350
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 3.8165383327753597,
|
| 1626 |
+
"grad_norm": 0.12823455035686493,
|
| 1627 |
+
"learning_rate": 1.8209463229550273e-05,
|
| 1628 |
+
"loss": 0.0412,
|
| 1629 |
+
"step": 11400
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 3.833277535989287,
|
| 1633 |
+
"grad_norm": 0.1442965269088745,
|
| 1634 |
+
"learning_rate": 1.8069969869434218e-05,
|
| 1635 |
+
"loss": 0.0416,
|
| 1636 |
+
"step": 11450
|
| 1637 |
+
},
|
| 1638 |
+
{
|
| 1639 |
+
"epoch": 3.850016739203214,
|
| 1640 |
+
"grad_norm": 0.13739417493343353,
|
| 1641 |
+
"learning_rate": 1.7930476509318158e-05,
|
| 1642 |
+
"loss": 0.0397,
|
| 1643 |
+
"step": 11500
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 3.866755942417141,
|
| 1647 |
+
"grad_norm": 0.16616705060005188,
|
| 1648 |
+
"learning_rate": 1.77909831492021e-05,
|
| 1649 |
+
"loss": 0.0413,
|
| 1650 |
+
"step": 11550
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 3.883495145631068,
|
| 1654 |
+
"grad_norm": 0.23060384392738342,
|
| 1655 |
+
"learning_rate": 1.765148978908604e-05,
|
| 1656 |
+
"loss": 0.0447,
|
| 1657 |
+
"step": 11600
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 3.900234348844995,
|
| 1661 |
+
"grad_norm": 0.2936810553073883,
|
| 1662 |
+
"learning_rate": 1.751199642896998e-05,
|
| 1663 |
+
"loss": 0.0409,
|
| 1664 |
+
"step": 11650
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 3.9169735520589217,
|
| 1668 |
+
"grad_norm": 0.17367126047611237,
|
| 1669 |
+
"learning_rate": 1.7372503068853924e-05,
|
| 1670 |
+
"loss": 0.0411,
|
| 1671 |
+
"step": 11700
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 3.933712755272849,
|
| 1675 |
+
"grad_norm": 0.14550547301769257,
|
| 1676 |
+
"learning_rate": 1.7233009708737865e-05,
|
| 1677 |
+
"loss": 0.0445,
|
| 1678 |
+
"step": 11750
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 3.950451958486776,
|
| 1682 |
+
"grad_norm": 0.13322454690933228,
|
| 1683 |
+
"learning_rate": 1.709351634862181e-05,
|
| 1684 |
+
"loss": 0.0444,
|
| 1685 |
+
"step": 11800
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 3.967191161700703,
|
| 1689 |
+
"grad_norm": 0.13606959581375122,
|
| 1690 |
+
"learning_rate": 1.6954022988505746e-05,
|
| 1691 |
+
"loss": 0.041,
|
| 1692 |
+
"step": 11850
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 3.98393036491463,
|
| 1696 |
+
"grad_norm": 0.14227426052093506,
|
| 1697 |
+
"learning_rate": 1.681452962838969e-05,
|
| 1698 |
+
"loss": 0.0412,
|
| 1699 |
+
"step": 11900
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 4.0,
|
| 1703 |
+
"eval_loss": 0.045825306326150894,
|
| 1704 |
+
"eval_runtime": 48.668,
|
| 1705 |
+
"eval_samples_per_second": 245.5,
|
| 1706 |
+
"eval_steps_per_second": 15.349,
|
| 1707 |
+
"step": 11948
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 4.000669568128557,
|
| 1711 |
+
"grad_norm": 0.14429377019405365,
|
| 1712 |
+
"learning_rate": 1.667503626827363e-05,
|
| 1713 |
+
"loss": 0.0398,
|
| 1714 |
+
"step": 11950
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 4.017408771342484,
|
| 1718 |
+
"grad_norm": 0.13842210173606873,
|
| 1719 |
+
"learning_rate": 1.653554290815757e-05,
|
| 1720 |
+
"loss": 0.0365,
|
| 1721 |
+
"step": 12000
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 4.0341479745564115,
|
| 1725 |
+
"grad_norm": 0.2168322205543518,
|
| 1726 |
+
"learning_rate": 1.6396049548041516e-05,
|
| 1727 |
+
"loss": 0.0395,
|
| 1728 |
+
"step": 12050
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 4.050887177770338,
|
| 1732 |
+
"grad_norm": 0.1534542292356491,
|
| 1733 |
+
"learning_rate": 1.6256556187925456e-05,
|
| 1734 |
+
"loss": 0.0377,
|
| 1735 |
+
"step": 12100
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 4.067626380984265,
|
| 1739 |
+
"grad_norm": 0.13067083060741425,
|
| 1740 |
+
"learning_rate": 1.6117062827809397e-05,
|
| 1741 |
+
"loss": 0.042,
|
| 1742 |
+
"step": 12150
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 4.084365584198192,
|
| 1746 |
+
"grad_norm": 0.13599875569343567,
|
| 1747 |
+
"learning_rate": 1.5977569467693338e-05,
|
| 1748 |
+
"loss": 0.0403,
|
| 1749 |
+
"step": 12200
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 4.101104787412119,
|
| 1753 |
+
"grad_norm": 0.23121419548988342,
|
| 1754 |
+
"learning_rate": 1.5838076107577278e-05,
|
| 1755 |
+
"loss": 0.0406,
|
| 1756 |
+
"step": 12250
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 4.117843990626046,
|
| 1760 |
+
"grad_norm": 0.15491856634616852,
|
| 1761 |
+
"learning_rate": 1.5698582747461222e-05,
|
| 1762 |
+
"loss": 0.0381,
|
| 1763 |
+
"step": 12300
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 4.1345831938399735,
|
| 1767 |
+
"grad_norm": 0.13168101012706757,
|
| 1768 |
+
"learning_rate": 1.5559089387345163e-05,
|
| 1769 |
+
"loss": 0.0353,
|
| 1770 |
+
"step": 12350
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 4.1513223970539,
|
| 1774 |
+
"grad_norm": 0.1694163829088211,
|
| 1775 |
+
"learning_rate": 1.5419596027229104e-05,
|
| 1776 |
+
"loss": 0.04,
|
| 1777 |
+
"step": 12400
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 4.168061600267827,
|
| 1781 |
+
"grad_norm": 0.17727704346179962,
|
| 1782 |
+
"learning_rate": 1.5280102667113048e-05,
|
| 1783 |
+
"loss": 0.0364,
|
| 1784 |
+
"step": 12450
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 4.184800803481754,
|
| 1788 |
+
"grad_norm": 0.18948902189731598,
|
| 1789 |
+
"learning_rate": 1.5140609306996988e-05,
|
| 1790 |
+
"loss": 0.0372,
|
| 1791 |
+
"step": 12500
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 4.201540006695681,
|
| 1795 |
+
"grad_norm": 0.2130223959684372,
|
| 1796 |
+
"learning_rate": 1.5001115946880929e-05,
|
| 1797 |
+
"loss": 0.0398,
|
| 1798 |
+
"step": 12550
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 4.218279209909609,
|
| 1802 |
+
"grad_norm": 0.15893810987472534,
|
| 1803 |
+
"learning_rate": 1.486162258676487e-05,
|
| 1804 |
+
"loss": 0.0387,
|
| 1805 |
+
"step": 12600
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 4.2350184131235356,
|
| 1809 |
+
"grad_norm": 0.19536694884300232,
|
| 1810 |
+
"learning_rate": 1.4722129226648812e-05,
|
| 1811 |
+
"loss": 0.041,
|
| 1812 |
+
"step": 12650
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 4.251757616337462,
|
| 1816 |
+
"grad_norm": 0.16755405068397522,
|
| 1817 |
+
"learning_rate": 1.4582635866532754e-05,
|
| 1818 |
+
"loss": 0.0367,
|
| 1819 |
+
"step": 12700
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"epoch": 4.268496819551389,
|
| 1823 |
+
"grad_norm": 0.14820334315299988,
|
| 1824 |
+
"learning_rate": 1.4443142506416697e-05,
|
| 1825 |
+
"loss": 0.0382,
|
| 1826 |
+
"step": 12750
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 4.285236022765316,
|
| 1830 |
+
"grad_norm": 0.21781007945537567,
|
| 1831 |
+
"learning_rate": 1.4303649146300637e-05,
|
| 1832 |
+
"loss": 0.0398,
|
| 1833 |
+
"step": 12800
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 4.301975225979243,
|
| 1837 |
+
"grad_norm": 0.17162373661994934,
|
| 1838 |
+
"learning_rate": 1.4164155786184576e-05,
|
| 1839 |
+
"loss": 0.042,
|
| 1840 |
+
"step": 12850
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 4.318714429193171,
|
| 1844 |
+
"grad_norm": 0.13765838742256165,
|
| 1845 |
+
"learning_rate": 1.4024662426068519e-05,
|
| 1846 |
+
"loss": 0.0378,
|
| 1847 |
+
"step": 12900
|
| 1848 |
+
},
|
| 1849 |
+
{
|
| 1850 |
+
"epoch": 4.335453632407098,
|
| 1851 |
+
"grad_norm": 0.22830092906951904,
|
| 1852 |
+
"learning_rate": 1.3885169065952461e-05,
|
| 1853 |
+
"loss": 0.0387,
|
| 1854 |
+
"step": 12950
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 4.352192835621024,
|
| 1858 |
+
"grad_norm": 0.1682949960231781,
|
| 1859 |
+
"learning_rate": 1.3745675705836403e-05,
|
| 1860 |
+
"loss": 0.0408,
|
| 1861 |
+
"step": 13000
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"epoch": 4.368932038834951,
|
| 1865 |
+
"grad_norm": 0.16153910756111145,
|
| 1866 |
+
"learning_rate": 1.3606182345720344e-05,
|
| 1867 |
+
"loss": 0.0405,
|
| 1868 |
+
"step": 13050
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 4.385671242048878,
|
| 1872 |
+
"grad_norm": 0.18771956861019135,
|
| 1873 |
+
"learning_rate": 1.3466688985604286e-05,
|
| 1874 |
+
"loss": 0.0427,
|
| 1875 |
+
"step": 13100
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 4.402410445262806,
|
| 1879 |
+
"grad_norm": 0.11162823438644409,
|
| 1880 |
+
"learning_rate": 1.3327195625488229e-05,
|
| 1881 |
+
"loss": 0.0399,
|
| 1882 |
+
"step": 13150
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 4.419149648476733,
|
| 1886 |
+
"grad_norm": 0.14466647803783417,
|
| 1887 |
+
"learning_rate": 1.3187702265372168e-05,
|
| 1888 |
+
"loss": 0.0401,
|
| 1889 |
+
"step": 13200
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 4.43588885169066,
|
| 1893 |
+
"grad_norm": 0.14834430813789368,
|
| 1894 |
+
"learning_rate": 1.304820890525611e-05,
|
| 1895 |
+
"loss": 0.0386,
|
| 1896 |
+
"step": 13250
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 4.4526280549045865,
|
| 1900 |
+
"grad_norm": 0.20324522256851196,
|
| 1901 |
+
"learning_rate": 1.2908715545140052e-05,
|
| 1902 |
+
"loss": 0.0379,
|
| 1903 |
+
"step": 13300
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 4.469367258118513,
|
| 1907 |
+
"grad_norm": 0.12625128030776978,
|
| 1908 |
+
"learning_rate": 1.2769222185023993e-05,
|
| 1909 |
+
"loss": 0.0403,
|
| 1910 |
+
"step": 13350
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 4.48610646133244,
|
| 1914 |
+
"grad_norm": 0.11018920689821243,
|
| 1915 |
+
"learning_rate": 1.2629728824907935e-05,
|
| 1916 |
+
"loss": 0.0383,
|
| 1917 |
+
"step": 13400
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 4.502845664546368,
|
| 1921 |
+
"grad_norm": 0.18631067872047424,
|
| 1922 |
+
"learning_rate": 1.2490235464791876e-05,
|
| 1923 |
+
"loss": 0.0424,
|
| 1924 |
+
"step": 13450
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 4.519584867760295,
|
| 1928 |
+
"grad_norm": 0.28578242659568787,
|
| 1929 |
+
"learning_rate": 1.2350742104675818e-05,
|
| 1930 |
+
"loss": 0.0385,
|
| 1931 |
+
"step": 13500
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 4.536324070974222,
|
| 1935 |
+
"grad_norm": 0.1431867629289627,
|
| 1936 |
+
"learning_rate": 1.2211248744559759e-05,
|
| 1937 |
+
"loss": 0.038,
|
| 1938 |
+
"step": 13550
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 4.5530632741881485,
|
| 1942 |
+
"grad_norm": 0.12731611728668213,
|
| 1943 |
+
"learning_rate": 1.2071755384443701e-05,
|
| 1944 |
+
"loss": 0.0389,
|
| 1945 |
+
"step": 13600
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 4.569802477402075,
|
| 1949 |
+
"grad_norm": 0.20958903431892395,
|
| 1950 |
+
"learning_rate": 1.1932262024327642e-05,
|
| 1951 |
+
"loss": 0.0383,
|
| 1952 |
+
"step": 13650
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 4.586541680616003,
|
| 1956 |
+
"grad_norm": 0.21877717971801758,
|
| 1957 |
+
"learning_rate": 1.1792768664211584e-05,
|
| 1958 |
+
"loss": 0.0399,
|
| 1959 |
+
"step": 13700
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 4.60328088382993,
|
| 1963 |
+
"grad_norm": 0.1719764918088913,
|
| 1964 |
+
"learning_rate": 1.1653275304095527e-05,
|
| 1965 |
+
"loss": 0.0355,
|
| 1966 |
+
"step": 13750
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 4.620020087043857,
|
| 1970 |
+
"grad_norm": 0.13044840097427368,
|
| 1971 |
+
"learning_rate": 1.1513781943979467e-05,
|
| 1972 |
+
"loss": 0.0386,
|
| 1973 |
+
"step": 13800
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 4.636759290257784,
|
| 1977 |
+
"grad_norm": 0.20552796125411987,
|
| 1978 |
+
"learning_rate": 1.1374288583863408e-05,
|
| 1979 |
+
"loss": 0.04,
|
| 1980 |
+
"step": 13850
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 4.6534984934717105,
|
| 1984 |
+
"grad_norm": 0.15806210041046143,
|
| 1985 |
+
"learning_rate": 1.123479522374735e-05,
|
| 1986 |
+
"loss": 0.0381,
|
| 1987 |
+
"step": 13900
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 4.670237696685637,
|
| 1991 |
+
"grad_norm": 0.2317190170288086,
|
| 1992 |
+
"learning_rate": 1.1095301863631291e-05,
|
| 1993 |
+
"loss": 0.0394,
|
| 1994 |
+
"step": 13950
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 4.686976899899565,
|
| 1998 |
+
"grad_norm": 0.16183790564537048,
|
| 1999 |
+
"learning_rate": 1.0955808503515233e-05,
|
| 2000 |
+
"loss": 0.0392,
|
| 2001 |
+
"step": 14000
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 4.703716103113492,
|
| 2005 |
+
"grad_norm": 0.14107303321361542,
|
| 2006 |
+
"learning_rate": 1.0816315143399176e-05,
|
| 2007 |
+
"loss": 0.039,
|
| 2008 |
+
"step": 14050
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 4.720455306327419,
|
| 2012 |
+
"grad_norm": 0.21340009570121765,
|
| 2013 |
+
"learning_rate": 1.0676821783283116e-05,
|
| 2014 |
+
"loss": 0.04,
|
| 2015 |
+
"step": 14100
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 4.737194509541346,
|
| 2019 |
+
"grad_norm": 0.1754944771528244,
|
| 2020 |
+
"learning_rate": 1.0537328423167057e-05,
|
| 2021 |
+
"loss": 0.0371,
|
| 2022 |
+
"step": 14150
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 4.7539337127552725,
|
| 2026 |
+
"grad_norm": 0.13280175626277924,
|
| 2027 |
+
"learning_rate": 1.0397835063051e-05,
|
| 2028 |
+
"loss": 0.0358,
|
| 2029 |
+
"step": 14200
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"epoch": 4.7706729159692,
|
| 2033 |
+
"grad_norm": 0.15039420127868652,
|
| 2034 |
+
"learning_rate": 1.0258341702934942e-05,
|
| 2035 |
+
"loss": 0.0438,
|
| 2036 |
+
"step": 14250
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 4.787412119183127,
|
| 2040 |
+
"grad_norm": 0.30474409461021423,
|
| 2041 |
+
"learning_rate": 1.0118848342818882e-05,
|
| 2042 |
+
"loss": 0.0377,
|
| 2043 |
+
"step": 14300
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 4.804151322397054,
|
| 2047 |
+
"grad_norm": 0.2012936919927597,
|
| 2048 |
+
"learning_rate": 9.979354982702823e-06,
|
| 2049 |
+
"loss": 0.0417,
|
| 2050 |
+
"step": 14350
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 4.820890525610981,
|
| 2054 |
+
"grad_norm": 0.192657932639122,
|
| 2055 |
+
"learning_rate": 9.839861622586765e-06,
|
| 2056 |
+
"loss": 0.0392,
|
| 2057 |
+
"step": 14400
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 4.837629728824908,
|
| 2061 |
+
"grad_norm": 0.16141368448734283,
|
| 2062 |
+
"learning_rate": 9.700368262470706e-06,
|
| 2063 |
+
"loss": 0.0378,
|
| 2064 |
+
"step": 14450
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 4.854368932038835,
|
| 2068 |
+
"grad_norm": 0.19473744928836823,
|
| 2069 |
+
"learning_rate": 9.560874902354648e-06,
|
| 2070 |
+
"loss": 0.0368,
|
| 2071 |
+
"step": 14500
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"epoch": 4.871108135252762,
|
| 2075 |
+
"grad_norm": 0.15624327957630157,
|
| 2076 |
+
"learning_rate": 9.42138154223859e-06,
|
| 2077 |
+
"loss": 0.0416,
|
| 2078 |
+
"step": 14550
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 4.887847338466689,
|
| 2082 |
+
"grad_norm": 0.1572103202342987,
|
| 2083 |
+
"learning_rate": 9.281888182122531e-06,
|
| 2084 |
+
"loss": 0.0383,
|
| 2085 |
+
"step": 14600
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 4.904586541680616,
|
| 2089 |
+
"grad_norm": 0.15121281147003174,
|
| 2090 |
+
"learning_rate": 9.142394822006472e-06,
|
| 2091 |
+
"loss": 0.0405,
|
| 2092 |
+
"step": 14650
|
| 2093 |
+
},
|
| 2094 |
+
{
|
| 2095 |
+
"epoch": 4.921325744894543,
|
| 2096 |
+
"grad_norm": 0.1739313155412674,
|
| 2097 |
+
"learning_rate": 9.002901461890415e-06,
|
| 2098 |
+
"loss": 0.0389,
|
| 2099 |
+
"step": 14700
|
| 2100 |
+
},
|
| 2101 |
+
{
|
| 2102 |
+
"epoch": 4.93806494810847,
|
| 2103 |
+
"grad_norm": 0.12826618552207947,
|
| 2104 |
+
"learning_rate": 8.863408101774357e-06,
|
| 2105 |
+
"loss": 0.0385,
|
| 2106 |
+
"step": 14750
|
| 2107 |
+
},
|
| 2108 |
+
{
|
| 2109 |
+
"epoch": 4.9548041513223975,
|
| 2110 |
+
"grad_norm": 0.11419858038425446,
|
| 2111 |
+
"learning_rate": 8.723914741658298e-06,
|
| 2112 |
+
"loss": 0.0387,
|
| 2113 |
+
"step": 14800
|
| 2114 |
+
},
|
| 2115 |
+
{
|
| 2116 |
+
"epoch": 4.971543354536324,
|
| 2117 |
+
"grad_norm": 0.18640589714050293,
|
| 2118 |
+
"learning_rate": 8.584421381542238e-06,
|
| 2119 |
+
"loss": 0.0376,
|
| 2120 |
+
"step": 14850
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 4.988282557750251,
|
| 2124 |
+
"grad_norm": 0.18249401450157166,
|
| 2125 |
+
"learning_rate": 8.44492802142618e-06,
|
| 2126 |
+
"loss": 0.041,
|
| 2127 |
+
"step": 14900
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 5.0,
|
| 2131 |
+
"eval_loss": 0.045213617384433746,
|
| 2132 |
+
"eval_runtime": 49.0497,
|
| 2133 |
+
"eval_samples_per_second": 243.589,
|
| 2134 |
+
"eval_steps_per_second": 15.229,
|
| 2135 |
+
"step": 14935
|
| 2136 |
+
}
|
| 2137 |
+
],
|
| 2138 |
+
"logging_steps": 50,
|
| 2139 |
+
"max_steps": 17922,
|
| 2140 |
+
"num_input_tokens_seen": 0,
|
| 2141 |
+
"num_train_epochs": 6,
|
| 2142 |
+
"save_steps": 500,
|
| 2143 |
+
"stateful_callbacks": {
|
| 2144 |
+
"TrainerControl": {
|
| 2145 |
+
"args": {
|
| 2146 |
+
"should_epoch_stop": false,
|
| 2147 |
+
"should_evaluate": false,
|
| 2148 |
+
"should_log": false,
|
| 2149 |
+
"should_save": true,
|
| 2150 |
+
"should_training_stop": false
|
| 2151 |
+
},
|
| 2152 |
+
"attributes": {}
|
| 2153 |
+
}
|
| 2154 |
+
},
|
| 2155 |
+
"total_flos": 3.240141313277952e+16,
|
| 2156 |
+
"train_batch_size": 16,
|
| 2157 |
+
"trial_name": null,
|
| 2158 |
+
"trial_params": null
|
| 2159 |
+
}
|
checkpoint-14935/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:37dd546b69fb60d8deb15a8b88e40b23e367c0e9f5a053ea3ae7c730b3874f2e
|
| 3 |
+
size 5304
|
checkpoint-14935/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-17922/config.json
ADDED
|
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ckartal/english-to-turkish-finetuned-model",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"activation_function": "swish",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"MarianMTModel"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"bos_token_id": 0,
|
| 10 |
+
"classifier_dropout": 0.0,
|
| 11 |
+
"d_model": 512,
|
| 12 |
+
"decoder_attention_heads": 8,
|
| 13 |
+
"decoder_ffn_dim": 2048,
|
| 14 |
+
"decoder_layerdrop": 0.0,
|
| 15 |
+
"decoder_layers": 6,
|
| 16 |
+
"decoder_start_token_id": 59993,
|
| 17 |
+
"decoder_vocab_size": 59994,
|
| 18 |
+
"dropout": 0.1,
|
| 19 |
+
"encoder_attention_heads": 8,
|
| 20 |
+
"encoder_ffn_dim": 2048,
|
| 21 |
+
"encoder_layerdrop": 0.0,
|
| 22 |
+
"encoder_layers": 6,
|
| 23 |
+
"eos_token_id": 0,
|
| 24 |
+
"forced_eos_token_id": 0,
|
| 25 |
+
"init_std": 0.02,
|
| 26 |
+
"is_encoder_decoder": true,
|
| 27 |
+
"max_length": null,
|
| 28 |
+
"max_position_embeddings": 512,
|
| 29 |
+
"model_type": "marian",
|
| 30 |
+
"normalize_embedding": false,
|
| 31 |
+
"num_beams": null,
|
| 32 |
+
"num_hidden_layers": 6,
|
| 33 |
+
"pad_token_id": 59993,
|
| 34 |
+
"scale_embedding": true,
|
| 35 |
+
"share_encoder_decoder_embeddings": true,
|
| 36 |
+
"static_position_embeddings": true,
|
| 37 |
+
"torch_dtype": "float32",
|
| 38 |
+
"transformers_version": "4.49.0",
|
| 39 |
+
"use_cache": true,
|
| 40 |
+
"vocab_size": 59994
|
| 41 |
+
}
|
checkpoint-17922/generation_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bad_words_ids": [
|
| 3 |
+
[
|
| 4 |
+
59993
|
| 5 |
+
]
|
| 6 |
+
],
|
| 7 |
+
"bos_token_id": 0,
|
| 8 |
+
"decoder_start_token_id": 59993,
|
| 9 |
+
"eos_token_id": 0,
|
| 10 |
+
"forced_eos_token_id": 0,
|
| 11 |
+
"max_length": 512,
|
| 12 |
+
"num_beams": 6,
|
| 13 |
+
"pad_token_id": 59993,
|
| 14 |
+
"renormalize_logits": true,
|
| 15 |
+
"transformers_version": "4.49.0"
|
| 16 |
+
}
|
checkpoint-17922/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0f997b0d1f9ca9a397fb73218159940280ff32f3cbdca06c9d17a7f7d3d66fd6
|
| 3 |
+
size 299690728
|
checkpoint-17922/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:06b55ff1b42484ee0be4850e41eb3646e2b79718e1fa5063f2d0b28829c5a60c
|
| 3 |
+
size 599054970
|
checkpoint-17922/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ced61baf7d63e0cafc1cd86b7b3b037ca4c3dbe957b8fabdf5fca21030341962
|
| 3 |
+
size 14244
|
checkpoint-17922/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d6dcca174cef295bb501b3b96815eb3658373aa4e458716e9734213c4c12acfc
|
| 3 |
+
size 988
|
checkpoint-17922/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a21237281603ebbc5900a536864616af817483454cb0f49aefdabd07813fd8a8
|
| 3 |
+
size 1064
|
checkpoint-17922/source.spm
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:98eb24f0995a9d5f7cb0fb628c474628b1d2284615e881e857d062c0b651ce10
|
| 3 |
+
size 793920
|
checkpoint-17922/special_tokens_map.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"eos_token": {
|
| 3 |
+
"content": "</s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"pad_token": {
|
| 10 |
+
"content": "<pad>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"unk_token": {
|
| 17 |
+
"content": "<unk>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
}
|
| 23 |
+
}
|
checkpoint-17922/target.spm
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:45cc6000ed513cdca8f80739087fbcbf9933dc50c9ae36c319c9670882f72e1b
|
| 3 |
+
size 837876
|
checkpoint-17922/tokenizer_config.json
ADDED
|
@@ -0,0 +1,40 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "</s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<unk>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"59993": {
|
| 20 |
+
"content": "<pad>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
}
|
| 27 |
+
},
|
| 28 |
+
"clean_up_tokenization_spaces": true,
|
| 29 |
+
"eos_token": "</s>",
|
| 30 |
+
"extra_special_tokens": {},
|
| 31 |
+
"model_max_length": 512,
|
| 32 |
+
"pad_token": "<pad>",
|
| 33 |
+
"return_tensors": "pt",
|
| 34 |
+
"separate_vocabs": false,
|
| 35 |
+
"source_lang": "eng",
|
| 36 |
+
"sp_model_kwargs": {},
|
| 37 |
+
"target_lang": "tur",
|
| 38 |
+
"tokenizer_class": "MarianTokenizer",
|
| 39 |
+
"unk_token": "<unk>"
|
| 40 |
+
}
|
checkpoint-17922/trainer_state.json
ADDED
|
@@ -0,0 +1,2587 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 6.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 17922,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.016739203213927016,
|
| 13 |
+
"grad_norm": 0.439177542924881,
|
| 14 |
+
"learning_rate": 4.987445597589555e-05,
|
| 15 |
+
"loss": 1.5222,
|
| 16 |
+
"step": 50
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.03347840642785403,
|
| 20 |
+
"grad_norm": 0.4248828887939453,
|
| 21 |
+
"learning_rate": 4.973496261577949e-05,
|
| 22 |
+
"loss": 0.1842,
|
| 23 |
+
"step": 100
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.05021760964178105,
|
| 27 |
+
"grad_norm": 0.3019009232521057,
|
| 28 |
+
"learning_rate": 4.9595469255663436e-05,
|
| 29 |
+
"loss": 0.1471,
|
| 30 |
+
"step": 150
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.06695681285570806,
|
| 34 |
+
"grad_norm": 0.2518245577812195,
|
| 35 |
+
"learning_rate": 4.9455975895547376e-05,
|
| 36 |
+
"loss": 0.1306,
|
| 37 |
+
"step": 200
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.08369601606963509,
|
| 41 |
+
"grad_norm": 0.3660012185573578,
|
| 42 |
+
"learning_rate": 4.931648253543131e-05,
|
| 43 |
+
"loss": 0.1276,
|
| 44 |
+
"step": 250
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.1004352192835621,
|
| 48 |
+
"grad_norm": 0.32854148745536804,
|
| 49 |
+
"learning_rate": 4.917698917531526e-05,
|
| 50 |
+
"loss": 0.1048,
|
| 51 |
+
"step": 300
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.11717442249748912,
|
| 55 |
+
"grad_norm": 0.24879515171051025,
|
| 56 |
+
"learning_rate": 4.90374958151992e-05,
|
| 57 |
+
"loss": 0.1054,
|
| 58 |
+
"step": 350
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.13391362571141613,
|
| 62 |
+
"grad_norm": 0.36416563391685486,
|
| 63 |
+
"learning_rate": 4.889800245508314e-05,
|
| 64 |
+
"loss": 0.0985,
|
| 65 |
+
"step": 400
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.15065282892534315,
|
| 69 |
+
"grad_norm": 0.33641186356544495,
|
| 70 |
+
"learning_rate": 4.875850909496708e-05,
|
| 71 |
+
"loss": 0.1044,
|
| 72 |
+
"step": 450
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.16739203213927017,
|
| 76 |
+
"grad_norm": 0.32909244298934937,
|
| 77 |
+
"learning_rate": 4.861901573485103e-05,
|
| 78 |
+
"loss": 0.1089,
|
| 79 |
+
"step": 500
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.1841312353531972,
|
| 83 |
+
"grad_norm": 0.36060285568237305,
|
| 84 |
+
"learning_rate": 4.847952237473497e-05,
|
| 85 |
+
"loss": 0.09,
|
| 86 |
+
"step": 550
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.2008704385671242,
|
| 90 |
+
"grad_norm": 0.2510785758495331,
|
| 91 |
+
"learning_rate": 4.83400290146189e-05,
|
| 92 |
+
"loss": 0.0884,
|
| 93 |
+
"step": 600
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.21760964178105122,
|
| 97 |
+
"grad_norm": 0.22478719055652618,
|
| 98 |
+
"learning_rate": 4.820053565450285e-05,
|
| 99 |
+
"loss": 0.0866,
|
| 100 |
+
"step": 650
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.23434884499497824,
|
| 104 |
+
"grad_norm": 0.37321263551712036,
|
| 105 |
+
"learning_rate": 4.806104229438679e-05,
|
| 106 |
+
"loss": 0.0884,
|
| 107 |
+
"step": 700
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.25108804820890523,
|
| 111 |
+
"grad_norm": 0.2660929262638092,
|
| 112 |
+
"learning_rate": 4.792154893427073e-05,
|
| 113 |
+
"loss": 0.0819,
|
| 114 |
+
"step": 750
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.26782725142283226,
|
| 118 |
+
"grad_norm": 0.2338525801897049,
|
| 119 |
+
"learning_rate": 4.778205557415467e-05,
|
| 120 |
+
"loss": 0.0845,
|
| 121 |
+
"step": 800
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.2845664546367593,
|
| 125 |
+
"grad_norm": 0.308557391166687,
|
| 126 |
+
"learning_rate": 4.764256221403862e-05,
|
| 127 |
+
"loss": 0.0815,
|
| 128 |
+
"step": 850
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.3013056578506863,
|
| 132 |
+
"grad_norm": 0.27098262310028076,
|
| 133 |
+
"learning_rate": 4.750306885392255e-05,
|
| 134 |
+
"loss": 0.0833,
|
| 135 |
+
"step": 900
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.3180448610646133,
|
| 139 |
+
"grad_norm": 0.23054952919483185,
|
| 140 |
+
"learning_rate": 4.736357549380649e-05,
|
| 141 |
+
"loss": 0.0806,
|
| 142 |
+
"step": 950
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.33478406427854035,
|
| 146 |
+
"grad_norm": 0.21355900168418884,
|
| 147 |
+
"learning_rate": 4.722408213369044e-05,
|
| 148 |
+
"loss": 0.073,
|
| 149 |
+
"step": 1000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.3515232674924674,
|
| 153 |
+
"grad_norm": 0.20395708084106445,
|
| 154 |
+
"learning_rate": 4.708458877357438e-05,
|
| 155 |
+
"loss": 0.0775,
|
| 156 |
+
"step": 1050
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.3682624707063944,
|
| 160 |
+
"grad_norm": 0.21063613891601562,
|
| 161 |
+
"learning_rate": 4.694509541345832e-05,
|
| 162 |
+
"loss": 0.0789,
|
| 163 |
+
"step": 1100
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.3850016739203214,
|
| 167 |
+
"grad_norm": 0.20589284598827362,
|
| 168 |
+
"learning_rate": 4.680560205334226e-05,
|
| 169 |
+
"loss": 0.0809,
|
| 170 |
+
"step": 1150
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.4017408771342484,
|
| 174 |
+
"grad_norm": 0.27975228428840637,
|
| 175 |
+
"learning_rate": 4.666610869322621e-05,
|
| 176 |
+
"loss": 0.078,
|
| 177 |
+
"step": 1200
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.4184800803481754,
|
| 181 |
+
"grad_norm": 0.2529745399951935,
|
| 182 |
+
"learning_rate": 4.6526615333110144e-05,
|
| 183 |
+
"loss": 0.0704,
|
| 184 |
+
"step": 1250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.43521928356210243,
|
| 188 |
+
"grad_norm": 0.2205154448747635,
|
| 189 |
+
"learning_rate": 4.6387121972994084e-05,
|
| 190 |
+
"loss": 0.0733,
|
| 191 |
+
"step": 1300
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.45195848677602946,
|
| 195 |
+
"grad_norm": 0.2254629135131836,
|
| 196 |
+
"learning_rate": 4.624762861287803e-05,
|
| 197 |
+
"loss": 0.0751,
|
| 198 |
+
"step": 1350
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.4686976899899565,
|
| 202 |
+
"grad_norm": 0.17614957690238953,
|
| 203 |
+
"learning_rate": 4.610813525276197e-05,
|
| 204 |
+
"loss": 0.0747,
|
| 205 |
+
"step": 1400
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.4854368932038835,
|
| 209 |
+
"grad_norm": 0.15940478444099426,
|
| 210 |
+
"learning_rate": 4.596864189264591e-05,
|
| 211 |
+
"loss": 0.0698,
|
| 212 |
+
"step": 1450
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5021760964178105,
|
| 216 |
+
"grad_norm": 0.1869521141052246,
|
| 217 |
+
"learning_rate": 4.5829148532529854e-05,
|
| 218 |
+
"loss": 0.0721,
|
| 219 |
+
"step": 1500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.5189152996317375,
|
| 223 |
+
"grad_norm": 0.36063650250434875,
|
| 224 |
+
"learning_rate": 4.5689655172413794e-05,
|
| 225 |
+
"loss": 0.0706,
|
| 226 |
+
"step": 1550
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.5356545028456645,
|
| 230 |
+
"grad_norm": 0.16967014968395233,
|
| 231 |
+
"learning_rate": 4.5550161812297735e-05,
|
| 232 |
+
"loss": 0.0759,
|
| 233 |
+
"step": 1600
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.5523937060595916,
|
| 237 |
+
"grad_norm": 0.29293423891067505,
|
| 238 |
+
"learning_rate": 4.5410668452181676e-05,
|
| 239 |
+
"loss": 0.0711,
|
| 240 |
+
"step": 1650
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.5691329092735186,
|
| 244 |
+
"grad_norm": 0.3034748136997223,
|
| 245 |
+
"learning_rate": 4.527117509206562e-05,
|
| 246 |
+
"loss": 0.067,
|
| 247 |
+
"step": 1700
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.5858721124874456,
|
| 251 |
+
"grad_norm": 0.1974593997001648,
|
| 252 |
+
"learning_rate": 4.513168173194956e-05,
|
| 253 |
+
"loss": 0.0701,
|
| 254 |
+
"step": 1750
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.6026113157013726,
|
| 258 |
+
"grad_norm": 0.18101799488067627,
|
| 259 |
+
"learning_rate": 4.4992188371833505e-05,
|
| 260 |
+
"loss": 0.0717,
|
| 261 |
+
"step": 1800
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.6193505189152997,
|
| 265 |
+
"grad_norm": 0.14422941207885742,
|
| 266 |
+
"learning_rate": 4.4852695011717445e-05,
|
| 267 |
+
"loss": 0.0686,
|
| 268 |
+
"step": 1850
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.6360897221292267,
|
| 272 |
+
"grad_norm": 0.28663551807403564,
|
| 273 |
+
"learning_rate": 4.4713201651601386e-05,
|
| 274 |
+
"loss": 0.0646,
|
| 275 |
+
"step": 1900
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.6528289253431536,
|
| 279 |
+
"grad_norm": 0.23879379034042358,
|
| 280 |
+
"learning_rate": 4.4573708291485327e-05,
|
| 281 |
+
"loss": 0.0684,
|
| 282 |
+
"step": 1950
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.6695681285570807,
|
| 286 |
+
"grad_norm": 0.21389362215995789,
|
| 287 |
+
"learning_rate": 4.443421493136927e-05,
|
| 288 |
+
"loss": 0.066,
|
| 289 |
+
"step": 2000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.6863073317710077,
|
| 293 |
+
"grad_norm": 0.26841893792152405,
|
| 294 |
+
"learning_rate": 4.4294721571253215e-05,
|
| 295 |
+
"loss": 0.0717,
|
| 296 |
+
"step": 2050
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.7030465349849347,
|
| 300 |
+
"grad_norm": 0.240205317735672,
|
| 301 |
+
"learning_rate": 4.415522821113715e-05,
|
| 302 |
+
"loss": 0.0697,
|
| 303 |
+
"step": 2100
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.7197857381988617,
|
| 307 |
+
"grad_norm": 0.28098127245903015,
|
| 308 |
+
"learning_rate": 4.4015734851021096e-05,
|
| 309 |
+
"loss": 0.0713,
|
| 310 |
+
"step": 2150
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.7365249414127888,
|
| 314 |
+
"grad_norm": 0.23308847844600677,
|
| 315 |
+
"learning_rate": 4.3876241490905037e-05,
|
| 316 |
+
"loss": 0.0667,
|
| 317 |
+
"step": 2200
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.7532641446267158,
|
| 321 |
+
"grad_norm": 0.22748568654060364,
|
| 322 |
+
"learning_rate": 4.373674813078898e-05,
|
| 323 |
+
"loss": 0.0605,
|
| 324 |
+
"step": 2250
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.7700033478406428,
|
| 328 |
+
"grad_norm": 0.3932187259197235,
|
| 329 |
+
"learning_rate": 4.359725477067292e-05,
|
| 330 |
+
"loss": 0.0676,
|
| 331 |
+
"step": 2300
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.7867425510545698,
|
| 335 |
+
"grad_norm": 0.23918767273426056,
|
| 336 |
+
"learning_rate": 4.345776141055686e-05,
|
| 337 |
+
"loss": 0.0624,
|
| 338 |
+
"step": 2350
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.8034817542684968,
|
| 342 |
+
"grad_norm": 0.3068426549434662,
|
| 343 |
+
"learning_rate": 4.33182680504408e-05,
|
| 344 |
+
"loss": 0.0664,
|
| 345 |
+
"step": 2400
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.8202209574824239,
|
| 349 |
+
"grad_norm": 0.17977873980998993,
|
| 350 |
+
"learning_rate": 4.317877469032474e-05,
|
| 351 |
+
"loss": 0.0726,
|
| 352 |
+
"step": 2450
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.8369601606963508,
|
| 356 |
+
"grad_norm": 0.16876642405986786,
|
| 357 |
+
"learning_rate": 4.303928133020869e-05,
|
| 358 |
+
"loss": 0.0639,
|
| 359 |
+
"step": 2500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.8536993639102779,
|
| 363 |
+
"grad_norm": 0.17980250716209412,
|
| 364 |
+
"learning_rate": 4.289978797009263e-05,
|
| 365 |
+
"loss": 0.0701,
|
| 366 |
+
"step": 2550
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.8704385671242049,
|
| 370 |
+
"grad_norm": 0.1711459904909134,
|
| 371 |
+
"learning_rate": 4.276029460997656e-05,
|
| 372 |
+
"loss": 0.063,
|
| 373 |
+
"step": 2600
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.8871777703381319,
|
| 377 |
+
"grad_norm": 0.443228542804718,
|
| 378 |
+
"learning_rate": 4.262080124986051e-05,
|
| 379 |
+
"loss": 0.0675,
|
| 380 |
+
"step": 2650
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.9039169735520589,
|
| 384 |
+
"grad_norm": 0.2098589390516281,
|
| 385 |
+
"learning_rate": 4.248130788974445e-05,
|
| 386 |
+
"loss": 0.062,
|
| 387 |
+
"step": 2700
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.920656176765986,
|
| 391 |
+
"grad_norm": 0.3022039234638214,
|
| 392 |
+
"learning_rate": 4.234181452962839e-05,
|
| 393 |
+
"loss": 0.07,
|
| 394 |
+
"step": 2750
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.937395379979913,
|
| 398 |
+
"grad_norm": 0.19368910789489746,
|
| 399 |
+
"learning_rate": 4.220232116951233e-05,
|
| 400 |
+
"loss": 0.0621,
|
| 401 |
+
"step": 2800
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.9541345831938399,
|
| 405 |
+
"grad_norm": 0.18753108382225037,
|
| 406 |
+
"learning_rate": 4.206282780939628e-05,
|
| 407 |
+
"loss": 0.0631,
|
| 408 |
+
"step": 2850
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.970873786407767,
|
| 412 |
+
"grad_norm": 0.15517786145210266,
|
| 413 |
+
"learning_rate": 4.192333444928022e-05,
|
| 414 |
+
"loss": 0.0641,
|
| 415 |
+
"step": 2900
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.987612989621694,
|
| 419 |
+
"grad_norm": 0.11765792220830917,
|
| 420 |
+
"learning_rate": 4.178384108916415e-05,
|
| 421 |
+
"loss": 0.0612,
|
| 422 |
+
"step": 2950
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.0,
|
| 426 |
+
"eval_loss": 0.05521286651492119,
|
| 427 |
+
"eval_runtime": 50.415,
|
| 428 |
+
"eval_samples_per_second": 236.993,
|
| 429 |
+
"eval_steps_per_second": 14.817,
|
| 430 |
+
"step": 2987
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.004352192835621,
|
| 434 |
+
"grad_norm": 0.2691793739795685,
|
| 435 |
+
"learning_rate": 4.16443477290481e-05,
|
| 436 |
+
"loss": 0.059,
|
| 437 |
+
"step": 3000
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.021091396049548,
|
| 441 |
+
"grad_norm": 0.394694060087204,
|
| 442 |
+
"learning_rate": 4.150485436893204e-05,
|
| 443 |
+
"loss": 0.0566,
|
| 444 |
+
"step": 3050
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.037830599263475,
|
| 448 |
+
"grad_norm": 0.19438503682613373,
|
| 449 |
+
"learning_rate": 4.136536100881598e-05,
|
| 450 |
+
"loss": 0.0591,
|
| 451 |
+
"step": 3100
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.0545698024774022,
|
| 455 |
+
"grad_norm": 0.21350933611392975,
|
| 456 |
+
"learning_rate": 4.122586764869992e-05,
|
| 457 |
+
"loss": 0.0509,
|
| 458 |
+
"step": 3150
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.071309005691329,
|
| 462 |
+
"grad_norm": 0.26747575402259827,
|
| 463 |
+
"learning_rate": 4.108637428858387e-05,
|
| 464 |
+
"loss": 0.0589,
|
| 465 |
+
"step": 3200
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.088048208905256,
|
| 469 |
+
"grad_norm": 0.31256961822509766,
|
| 470 |
+
"learning_rate": 4.0946880928467804e-05,
|
| 471 |
+
"loss": 0.0602,
|
| 472 |
+
"step": 3250
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.1047874121191832,
|
| 476 |
+
"grad_norm": 0.18631280958652496,
|
| 477 |
+
"learning_rate": 4.0807387568351745e-05,
|
| 478 |
+
"loss": 0.0547,
|
| 479 |
+
"step": 3300
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.12152661533311,
|
| 483 |
+
"grad_norm": 0.18677473068237305,
|
| 484 |
+
"learning_rate": 4.066789420823569e-05,
|
| 485 |
+
"loss": 0.0543,
|
| 486 |
+
"step": 3350
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.1382658185470371,
|
| 490 |
+
"grad_norm": 0.24535444378852844,
|
| 491 |
+
"learning_rate": 4.052840084811963e-05,
|
| 492 |
+
"loss": 0.0583,
|
| 493 |
+
"step": 3400
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.1550050217609642,
|
| 497 |
+
"grad_norm": 0.1752105951309204,
|
| 498 |
+
"learning_rate": 4.038890748800357e-05,
|
| 499 |
+
"loss": 0.0504,
|
| 500 |
+
"step": 3450
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.1717442249748913,
|
| 504 |
+
"grad_norm": 0.14743360877037048,
|
| 505 |
+
"learning_rate": 4.0249414127887514e-05,
|
| 506 |
+
"loss": 0.055,
|
| 507 |
+
"step": 3500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.1884834281888181,
|
| 511 |
+
"grad_norm": 0.11535945534706116,
|
| 512 |
+
"learning_rate": 4.010992076777146e-05,
|
| 513 |
+
"loss": 0.0552,
|
| 514 |
+
"step": 3550
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.2052226314027452,
|
| 518 |
+
"grad_norm": 0.26563358306884766,
|
| 519 |
+
"learning_rate": 3.9970427407655395e-05,
|
| 520 |
+
"loss": 0.0552,
|
| 521 |
+
"step": 3600
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.2219618346166723,
|
| 525 |
+
"grad_norm": 0.15104246139526367,
|
| 526 |
+
"learning_rate": 3.9830934047539336e-05,
|
| 527 |
+
"loss": 0.0575,
|
| 528 |
+
"step": 3650
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.2387010378305994,
|
| 532 |
+
"grad_norm": 0.2198421210050583,
|
| 533 |
+
"learning_rate": 3.9691440687423283e-05,
|
| 534 |
+
"loss": 0.0567,
|
| 535 |
+
"step": 3700
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.2554402410445262,
|
| 539 |
+
"grad_norm": 0.20177733898162842,
|
| 540 |
+
"learning_rate": 3.955194732730722e-05,
|
| 541 |
+
"loss": 0.0556,
|
| 542 |
+
"step": 3750
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.2721794442584533,
|
| 546 |
+
"grad_norm": 0.36604830622673035,
|
| 547 |
+
"learning_rate": 3.9412453967191165e-05,
|
| 548 |
+
"loss": 0.0569,
|
| 549 |
+
"step": 3800
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.2889186474723804,
|
| 553 |
+
"grad_norm": 0.18883727490901947,
|
| 554 |
+
"learning_rate": 3.9272960607075105e-05,
|
| 555 |
+
"loss": 0.0595,
|
| 556 |
+
"step": 3850
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.3056578506863072,
|
| 560 |
+
"grad_norm": 0.14828617870807648,
|
| 561 |
+
"learning_rate": 3.9133467246959046e-05,
|
| 562 |
+
"loss": 0.0548,
|
| 563 |
+
"step": 3900
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.3223970539002343,
|
| 567 |
+
"grad_norm": 0.19220437109470367,
|
| 568 |
+
"learning_rate": 3.899397388684299e-05,
|
| 569 |
+
"loss": 0.053,
|
| 570 |
+
"step": 3950
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.3391362571141614,
|
| 574 |
+
"grad_norm": 0.16049669682979584,
|
| 575 |
+
"learning_rate": 3.885448052672693e-05,
|
| 576 |
+
"loss": 0.0581,
|
| 577 |
+
"step": 4000
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.3558754603280883,
|
| 581 |
+
"grad_norm": 0.22821515798568726,
|
| 582 |
+
"learning_rate": 3.8714987166610875e-05,
|
| 583 |
+
"loss": 0.0518,
|
| 584 |
+
"step": 4050
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.3726146635420153,
|
| 588 |
+
"grad_norm": 0.1879580318927765,
|
| 589 |
+
"learning_rate": 3.857549380649481e-05,
|
| 590 |
+
"loss": 0.0574,
|
| 591 |
+
"step": 4100
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.3893538667559424,
|
| 595 |
+
"grad_norm": 0.16026251018047333,
|
| 596 |
+
"learning_rate": 3.8436000446378756e-05,
|
| 597 |
+
"loss": 0.063,
|
| 598 |
+
"step": 4150
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.4060930699698695,
|
| 602 |
+
"grad_norm": 0.26868143677711487,
|
| 603 |
+
"learning_rate": 3.82965070862627e-05,
|
| 604 |
+
"loss": 0.0571,
|
| 605 |
+
"step": 4200
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.4228322731837966,
|
| 609 |
+
"grad_norm": 0.2529687285423279,
|
| 610 |
+
"learning_rate": 3.815701372614664e-05,
|
| 611 |
+
"loss": 0.0528,
|
| 612 |
+
"step": 4250
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.4395714763977234,
|
| 616 |
+
"grad_norm": 0.19138221442699432,
|
| 617 |
+
"learning_rate": 3.801752036603058e-05,
|
| 618 |
+
"loss": 0.0584,
|
| 619 |
+
"step": 4300
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.4563106796116505,
|
| 623 |
+
"grad_norm": 0.16359661519527435,
|
| 624 |
+
"learning_rate": 3.787802700591452e-05,
|
| 625 |
+
"loss": 0.0539,
|
| 626 |
+
"step": 4350
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.4730498828255776,
|
| 630 |
+
"grad_norm": 0.1373494267463684,
|
| 631 |
+
"learning_rate": 3.7738533645798466e-05,
|
| 632 |
+
"loss": 0.0557,
|
| 633 |
+
"step": 4400
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.4897890860395044,
|
| 637 |
+
"grad_norm": 0.15695162117481232,
|
| 638 |
+
"learning_rate": 3.75990402856824e-05,
|
| 639 |
+
"loss": 0.0491,
|
| 640 |
+
"step": 4450
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.5065282892534315,
|
| 644 |
+
"grad_norm": 0.18462614715099335,
|
| 645 |
+
"learning_rate": 3.745954692556635e-05,
|
| 646 |
+
"loss": 0.0495,
|
| 647 |
+
"step": 4500
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.5232674924673586,
|
| 651 |
+
"grad_norm": 0.27876704931259155,
|
| 652 |
+
"learning_rate": 3.732005356545029e-05,
|
| 653 |
+
"loss": 0.0523,
|
| 654 |
+
"step": 4550
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.5400066956812855,
|
| 658 |
+
"grad_norm": 0.30491840839385986,
|
| 659 |
+
"learning_rate": 3.718056020533423e-05,
|
| 660 |
+
"loss": 0.0564,
|
| 661 |
+
"step": 4600
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.5567458988952128,
|
| 665 |
+
"grad_norm": 0.18721336126327515,
|
| 666 |
+
"learning_rate": 3.704106684521817e-05,
|
| 667 |
+
"loss": 0.0524,
|
| 668 |
+
"step": 4650
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.5734851021091396,
|
| 672 |
+
"grad_norm": 0.21216215193271637,
|
| 673 |
+
"learning_rate": 3.690157348510211e-05,
|
| 674 |
+
"loss": 0.0521,
|
| 675 |
+
"step": 4700
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.5902243053230665,
|
| 679 |
+
"grad_norm": 0.1368396282196045,
|
| 680 |
+
"learning_rate": 3.676208012498605e-05,
|
| 681 |
+
"loss": 0.056,
|
| 682 |
+
"step": 4750
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.6069635085369938,
|
| 686 |
+
"grad_norm": 0.13692086935043335,
|
| 687 |
+
"learning_rate": 3.662258676486999e-05,
|
| 688 |
+
"loss": 0.0443,
|
| 689 |
+
"step": 4800
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.6237027117509206,
|
| 693 |
+
"grad_norm": 0.11640128493309021,
|
| 694 |
+
"learning_rate": 3.648309340475394e-05,
|
| 695 |
+
"loss": 0.0488,
|
| 696 |
+
"step": 4850
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.6404419149648477,
|
| 700 |
+
"grad_norm": 0.19953882694244385,
|
| 701 |
+
"learning_rate": 3.634360004463788e-05,
|
| 702 |
+
"loss": 0.0553,
|
| 703 |
+
"step": 4900
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.6571811181787748,
|
| 707 |
+
"grad_norm": 0.1966984122991562,
|
| 708 |
+
"learning_rate": 3.6204106684521813e-05,
|
| 709 |
+
"loss": 0.0536,
|
| 710 |
+
"step": 4950
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.6739203213927016,
|
| 714 |
+
"grad_norm": 0.2324533313512802,
|
| 715 |
+
"learning_rate": 3.606461332440576e-05,
|
| 716 |
+
"loss": 0.0493,
|
| 717 |
+
"step": 5000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.6906595246066287,
|
| 721 |
+
"grad_norm": 0.16217607259750366,
|
| 722 |
+
"learning_rate": 3.59251199642897e-05,
|
| 723 |
+
"loss": 0.0503,
|
| 724 |
+
"step": 5050
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.7073987278205558,
|
| 728 |
+
"grad_norm": 0.23949602246284485,
|
| 729 |
+
"learning_rate": 3.578562660417364e-05,
|
| 730 |
+
"loss": 0.0556,
|
| 731 |
+
"step": 5100
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.7241379310344827,
|
| 735 |
+
"grad_norm": 0.21387897431850433,
|
| 736 |
+
"learning_rate": 3.564613324405758e-05,
|
| 737 |
+
"loss": 0.0548,
|
| 738 |
+
"step": 5150
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.7408771342484097,
|
| 742 |
+
"grad_norm": 0.2055111676454544,
|
| 743 |
+
"learning_rate": 3.550663988394153e-05,
|
| 744 |
+
"loss": 0.06,
|
| 745 |
+
"step": 5200
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.7576163374623368,
|
| 749 |
+
"grad_norm": 0.20280921459197998,
|
| 750 |
+
"learning_rate": 3.5367146523825464e-05,
|
| 751 |
+
"loss": 0.0508,
|
| 752 |
+
"step": 5250
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.7743555406762637,
|
| 756 |
+
"grad_norm": 0.14165103435516357,
|
| 757 |
+
"learning_rate": 3.5227653163709405e-05,
|
| 758 |
+
"loss": 0.0581,
|
| 759 |
+
"step": 5300
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.791094743890191,
|
| 763 |
+
"grad_norm": 0.18099863827228546,
|
| 764 |
+
"learning_rate": 3.508815980359335e-05,
|
| 765 |
+
"loss": 0.0562,
|
| 766 |
+
"step": 5350
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.8078339471041178,
|
| 770 |
+
"grad_norm": 0.21743184328079224,
|
| 771 |
+
"learning_rate": 3.494866644347729e-05,
|
| 772 |
+
"loss": 0.0498,
|
| 773 |
+
"step": 5400
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.824573150318045,
|
| 777 |
+
"grad_norm": 0.20934534072875977,
|
| 778 |
+
"learning_rate": 3.4809173083361234e-05,
|
| 779 |
+
"loss": 0.0549,
|
| 780 |
+
"step": 5450
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.841312353531972,
|
| 784 |
+
"grad_norm": 0.1582174152135849,
|
| 785 |
+
"learning_rate": 3.4669679723245174e-05,
|
| 786 |
+
"loss": 0.0556,
|
| 787 |
+
"step": 5500
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.8580515567458988,
|
| 791 |
+
"grad_norm": 0.1624903827905655,
|
| 792 |
+
"learning_rate": 3.453018636312912e-05,
|
| 793 |
+
"loss": 0.0516,
|
| 794 |
+
"step": 5550
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.874790759959826,
|
| 798 |
+
"grad_norm": 0.16255798935890198,
|
| 799 |
+
"learning_rate": 3.4390693003013056e-05,
|
| 800 |
+
"loss": 0.0542,
|
| 801 |
+
"step": 5600
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.891529963173753,
|
| 805 |
+
"grad_norm": 0.1269742250442505,
|
| 806 |
+
"learning_rate": 3.4251199642896996e-05,
|
| 807 |
+
"loss": 0.0565,
|
| 808 |
+
"step": 5650
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.9082691663876798,
|
| 812 |
+
"grad_norm": 0.15966229140758514,
|
| 813 |
+
"learning_rate": 3.4111706282780944e-05,
|
| 814 |
+
"loss": 0.0538,
|
| 815 |
+
"step": 5700
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.925008369601607,
|
| 819 |
+
"grad_norm": 0.21506330370903015,
|
| 820 |
+
"learning_rate": 3.3972212922664884e-05,
|
| 821 |
+
"loss": 0.0505,
|
| 822 |
+
"step": 5750
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.941747572815534,
|
| 826 |
+
"grad_norm": 0.2145415097475052,
|
| 827 |
+
"learning_rate": 3.3832719562548825e-05,
|
| 828 |
+
"loss": 0.0521,
|
| 829 |
+
"step": 5800
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.9584867760294609,
|
| 833 |
+
"grad_norm": 0.10960496962070465,
|
| 834 |
+
"learning_rate": 3.3693226202432766e-05,
|
| 835 |
+
"loss": 0.0513,
|
| 836 |
+
"step": 5850
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.9752259792433882,
|
| 840 |
+
"grad_norm": 0.13635843992233276,
|
| 841 |
+
"learning_rate": 3.355373284231671e-05,
|
| 842 |
+
"loss": 0.0499,
|
| 843 |
+
"step": 5900
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.991965182457315,
|
| 847 |
+
"grad_norm": 0.1542210429906845,
|
| 848 |
+
"learning_rate": 3.341423948220065e-05,
|
| 849 |
+
"loss": 0.0556,
|
| 850 |
+
"step": 5950
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 2.0,
|
| 854 |
+
"eval_loss": 0.04946442320942879,
|
| 855 |
+
"eval_runtime": 55.6114,
|
| 856 |
+
"eval_samples_per_second": 214.848,
|
| 857 |
+
"eval_steps_per_second": 13.432,
|
| 858 |
+
"step": 5974
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 2.008704385671242,
|
| 862 |
+
"grad_norm": 0.1718842089176178,
|
| 863 |
+
"learning_rate": 3.327474612208459e-05,
|
| 864 |
+
"loss": 0.0503,
|
| 865 |
+
"step": 6000
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 2.025443588885169,
|
| 869 |
+
"grad_norm": 0.1528020203113556,
|
| 870 |
+
"learning_rate": 3.3135252761968535e-05,
|
| 871 |
+
"loss": 0.0479,
|
| 872 |
+
"step": 6050
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 2.042182792099096,
|
| 876 |
+
"grad_norm": 0.19148772954940796,
|
| 877 |
+
"learning_rate": 3.299575940185247e-05,
|
| 878 |
+
"loss": 0.0443,
|
| 879 |
+
"step": 6100
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 2.058921995313023,
|
| 883 |
+
"grad_norm": 0.18125496804714203,
|
| 884 |
+
"learning_rate": 3.2856266041736416e-05,
|
| 885 |
+
"loss": 0.0473,
|
| 886 |
+
"step": 6150
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 2.07566119852695,
|
| 890 |
+
"grad_norm": 0.20772996544837952,
|
| 891 |
+
"learning_rate": 3.271677268162036e-05,
|
| 892 |
+
"loss": 0.0539,
|
| 893 |
+
"step": 6200
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 2.092400401740877,
|
| 897 |
+
"grad_norm": 0.2518468201160431,
|
| 898 |
+
"learning_rate": 3.25772793215043e-05,
|
| 899 |
+
"loss": 0.0468,
|
| 900 |
+
"step": 6250
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 2.1091396049548043,
|
| 904 |
+
"grad_norm": 0.1350301206111908,
|
| 905 |
+
"learning_rate": 3.243778596138824e-05,
|
| 906 |
+
"loss": 0.0431,
|
| 907 |
+
"step": 6300
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 2.125878808168731,
|
| 911 |
+
"grad_norm": 0.19141735136508942,
|
| 912 |
+
"learning_rate": 3.229829260127218e-05,
|
| 913 |
+
"loss": 0.048,
|
| 914 |
+
"step": 6350
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 2.142618011382658,
|
| 918 |
+
"grad_norm": 0.2404586374759674,
|
| 919 |
+
"learning_rate": 3.2158799241156126e-05,
|
| 920 |
+
"loss": 0.0516,
|
| 921 |
+
"step": 6400
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 2.1593572145965854,
|
| 925 |
+
"grad_norm": 0.21710112690925598,
|
| 926 |
+
"learning_rate": 3.201930588104006e-05,
|
| 927 |
+
"loss": 0.0472,
|
| 928 |
+
"step": 6450
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 2.176096417810512,
|
| 932 |
+
"grad_norm": 0.14395031332969666,
|
| 933 |
+
"learning_rate": 3.187981252092401e-05,
|
| 934 |
+
"loss": 0.0438,
|
| 935 |
+
"step": 6500
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 2.192835621024439,
|
| 939 |
+
"grad_norm": 0.20882932841777802,
|
| 940 |
+
"learning_rate": 3.174031916080795e-05,
|
| 941 |
+
"loss": 0.0488,
|
| 942 |
+
"step": 6550
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 2.2095748242383664,
|
| 946 |
+
"grad_norm": 0.13824905455112457,
|
| 947 |
+
"learning_rate": 3.160082580069189e-05,
|
| 948 |
+
"loss": 0.0464,
|
| 949 |
+
"step": 6600
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 2.2263140274522932,
|
| 953 |
+
"grad_norm": 0.1783577799797058,
|
| 954 |
+
"learning_rate": 3.146133244057583e-05,
|
| 955 |
+
"loss": 0.0459,
|
| 956 |
+
"step": 6650
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 2.24305323066622,
|
| 960 |
+
"grad_norm": 0.22147531807422638,
|
| 961 |
+
"learning_rate": 3.132183908045977e-05,
|
| 962 |
+
"loss": 0.0476,
|
| 963 |
+
"step": 6700
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 2.2597924338801474,
|
| 967 |
+
"grad_norm": 0.17393821477890015,
|
| 968 |
+
"learning_rate": 3.118234572034371e-05,
|
| 969 |
+
"loss": 0.0436,
|
| 970 |
+
"step": 6750
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 2.2765316370940742,
|
| 974 |
+
"grad_norm": 0.15850785374641418,
|
| 975 |
+
"learning_rate": 3.104285236022765e-05,
|
| 976 |
+
"loss": 0.0476,
|
| 977 |
+
"step": 6800
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 2.2932708403080015,
|
| 981 |
+
"grad_norm": 0.16232182085514069,
|
| 982 |
+
"learning_rate": 3.09033590001116e-05,
|
| 983 |
+
"loss": 0.0473,
|
| 984 |
+
"step": 6850
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 2.3100100435219284,
|
| 988 |
+
"grad_norm": 0.1816001981496811,
|
| 989 |
+
"learning_rate": 3.076386563999554e-05,
|
| 990 |
+
"loss": 0.0427,
|
| 991 |
+
"step": 6900
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 2.3267492467358553,
|
| 995 |
+
"grad_norm": 0.13417834043502808,
|
| 996 |
+
"learning_rate": 3.062437227987948e-05,
|
| 997 |
+
"loss": 0.0448,
|
| 998 |
+
"step": 6950
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 2.3434884499497826,
|
| 1002 |
+
"grad_norm": 0.12576530873775482,
|
| 1003 |
+
"learning_rate": 3.048487891976342e-05,
|
| 1004 |
+
"loss": 0.0453,
|
| 1005 |
+
"step": 7000
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 2.3602276531637094,
|
| 1009 |
+
"grad_norm": 0.33120718598365784,
|
| 1010 |
+
"learning_rate": 3.0345385559647362e-05,
|
| 1011 |
+
"loss": 0.0462,
|
| 1012 |
+
"step": 7050
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 2.3769668563776363,
|
| 1016 |
+
"grad_norm": 0.22310969233512878,
|
| 1017 |
+
"learning_rate": 3.0205892199531306e-05,
|
| 1018 |
+
"loss": 0.0475,
|
| 1019 |
+
"step": 7100
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 2.3937060595915636,
|
| 1023 |
+
"grad_norm": 0.18150626122951508,
|
| 1024 |
+
"learning_rate": 3.0066398839415243e-05,
|
| 1025 |
+
"loss": 0.0489,
|
| 1026 |
+
"step": 7150
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 2.4104452628054904,
|
| 1030 |
+
"grad_norm": 0.28730452060699463,
|
| 1031 |
+
"learning_rate": 2.9926905479299187e-05,
|
| 1032 |
+
"loss": 0.0536,
|
| 1033 |
+
"step": 7200
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 2.4271844660194173,
|
| 1037 |
+
"grad_norm": 0.1918480098247528,
|
| 1038 |
+
"learning_rate": 2.9787412119183128e-05,
|
| 1039 |
+
"loss": 0.0426,
|
| 1040 |
+
"step": 7250
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 2.4439236692333446,
|
| 1044 |
+
"grad_norm": 0.16158398985862732,
|
| 1045 |
+
"learning_rate": 2.964791875906707e-05,
|
| 1046 |
+
"loss": 0.0458,
|
| 1047 |
+
"step": 7300
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 2.4606628724472714,
|
| 1051 |
+
"grad_norm": 0.27141231298446655,
|
| 1052 |
+
"learning_rate": 2.9508425398951012e-05,
|
| 1053 |
+
"loss": 0.0454,
|
| 1054 |
+
"step": 7350
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 2.4774020756611987,
|
| 1058 |
+
"grad_norm": 0.1777345836162567,
|
| 1059 |
+
"learning_rate": 2.936893203883495e-05,
|
| 1060 |
+
"loss": 0.0435,
|
| 1061 |
+
"step": 7400
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 2.4941412788751256,
|
| 1065 |
+
"grad_norm": 0.14735421538352966,
|
| 1066 |
+
"learning_rate": 2.9229438678718897e-05,
|
| 1067 |
+
"loss": 0.0489,
|
| 1068 |
+
"step": 7450
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 2.5108804820890525,
|
| 1072 |
+
"grad_norm": 0.1486055999994278,
|
| 1073 |
+
"learning_rate": 2.9089945318602834e-05,
|
| 1074 |
+
"loss": 0.0477,
|
| 1075 |
+
"step": 7500
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 2.5276196853029793,
|
| 1079 |
+
"grad_norm": 0.17078754305839539,
|
| 1080 |
+
"learning_rate": 2.895045195848678e-05,
|
| 1081 |
+
"loss": 0.0444,
|
| 1082 |
+
"step": 7550
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 2.5443588885169066,
|
| 1086 |
+
"grad_norm": 0.19276435673236847,
|
| 1087 |
+
"learning_rate": 2.881095859837072e-05,
|
| 1088 |
+
"loss": 0.0486,
|
| 1089 |
+
"step": 7600
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 2.5610980917308335,
|
| 1093 |
+
"grad_norm": 0.21209606528282166,
|
| 1094 |
+
"learning_rate": 2.8671465238254656e-05,
|
| 1095 |
+
"loss": 0.0497,
|
| 1096 |
+
"step": 7650
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 2.5778372949447608,
|
| 1100 |
+
"grad_norm": 0.21018877625465393,
|
| 1101 |
+
"learning_rate": 2.8531971878138604e-05,
|
| 1102 |
+
"loss": 0.0441,
|
| 1103 |
+
"step": 7700
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 2.5945764981586876,
|
| 1107 |
+
"grad_norm": 0.15666617453098297,
|
| 1108 |
+
"learning_rate": 2.839247851802254e-05,
|
| 1109 |
+
"loss": 0.0467,
|
| 1110 |
+
"step": 7750
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 2.6113157013726145,
|
| 1114 |
+
"grad_norm": 0.1940685212612152,
|
| 1115 |
+
"learning_rate": 2.8252985157906485e-05,
|
| 1116 |
+
"loss": 0.0523,
|
| 1117 |
+
"step": 7800
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 2.628054904586542,
|
| 1121 |
+
"grad_norm": 0.28480586409568787,
|
| 1122 |
+
"learning_rate": 2.8113491797790426e-05,
|
| 1123 |
+
"loss": 0.0481,
|
| 1124 |
+
"step": 7850
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 2.6447941078004686,
|
| 1128 |
+
"grad_norm": 0.2223973125219345,
|
| 1129 |
+
"learning_rate": 2.797399843767437e-05,
|
| 1130 |
+
"loss": 0.0432,
|
| 1131 |
+
"step": 7900
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 2.661533311014396,
|
| 1135 |
+
"grad_norm": 0.15986157953739166,
|
| 1136 |
+
"learning_rate": 2.783450507755831e-05,
|
| 1137 |
+
"loss": 0.0454,
|
| 1138 |
+
"step": 7950
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 2.678272514228323,
|
| 1142 |
+
"grad_norm": 0.1384258270263672,
|
| 1143 |
+
"learning_rate": 2.7695011717442248e-05,
|
| 1144 |
+
"loss": 0.0477,
|
| 1145 |
+
"step": 8000
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 2.6950117174422497,
|
| 1149 |
+
"grad_norm": 0.1721869707107544,
|
| 1150 |
+
"learning_rate": 2.7555518357326192e-05,
|
| 1151 |
+
"loss": 0.0453,
|
| 1152 |
+
"step": 8050
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 2.7117509206561765,
|
| 1156 |
+
"grad_norm": 0.20737840235233307,
|
| 1157 |
+
"learning_rate": 2.7416024997210132e-05,
|
| 1158 |
+
"loss": 0.0504,
|
| 1159 |
+
"step": 8100
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 2.728490123870104,
|
| 1163 |
+
"grad_norm": 0.18823584914207458,
|
| 1164 |
+
"learning_rate": 2.7276531637094077e-05,
|
| 1165 |
+
"loss": 0.0453,
|
| 1166 |
+
"step": 8150
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 2.7452293270840307,
|
| 1170 |
+
"grad_norm": 0.13201962411403656,
|
| 1171 |
+
"learning_rate": 2.7137038276978017e-05,
|
| 1172 |
+
"loss": 0.0433,
|
| 1173 |
+
"step": 8200
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 2.761968530297958,
|
| 1177 |
+
"grad_norm": 0.1443973183631897,
|
| 1178 |
+
"learning_rate": 2.699754491686196e-05,
|
| 1179 |
+
"loss": 0.0486,
|
| 1180 |
+
"step": 8250
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 2.778707733511885,
|
| 1184 |
+
"grad_norm": 0.29314514994621277,
|
| 1185 |
+
"learning_rate": 2.68580515567459e-05,
|
| 1186 |
+
"loss": 0.05,
|
| 1187 |
+
"step": 8300
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 2.7954469367258117,
|
| 1191 |
+
"grad_norm": 0.14852124452590942,
|
| 1192 |
+
"learning_rate": 2.671855819662984e-05,
|
| 1193 |
+
"loss": 0.0495,
|
| 1194 |
+
"step": 8350
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 2.812186139939739,
|
| 1198 |
+
"grad_norm": 0.19024662673473358,
|
| 1199 |
+
"learning_rate": 2.6579064836513783e-05,
|
| 1200 |
+
"loss": 0.0508,
|
| 1201 |
+
"step": 8400
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 2.828925343153666,
|
| 1205 |
+
"grad_norm": 0.1745578795671463,
|
| 1206 |
+
"learning_rate": 2.6439571476397724e-05,
|
| 1207 |
+
"loss": 0.0443,
|
| 1208 |
+
"step": 8450
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 2.845664546367593,
|
| 1212 |
+
"grad_norm": 0.18390017747879028,
|
| 1213 |
+
"learning_rate": 2.6300078116281668e-05,
|
| 1214 |
+
"loss": 0.0468,
|
| 1215 |
+
"step": 8500
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 2.86240374958152,
|
| 1219 |
+
"grad_norm": 0.22483347356319427,
|
| 1220 |
+
"learning_rate": 2.616058475616561e-05,
|
| 1221 |
+
"loss": 0.0467,
|
| 1222 |
+
"step": 8550
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 2.879142952795447,
|
| 1226 |
+
"grad_norm": 0.18160563707351685,
|
| 1227 |
+
"learning_rate": 2.6021091396049553e-05,
|
| 1228 |
+
"loss": 0.0441,
|
| 1229 |
+
"step": 8600
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 2.8958821560093737,
|
| 1233 |
+
"grad_norm": 0.13408955931663513,
|
| 1234 |
+
"learning_rate": 2.588159803593349e-05,
|
| 1235 |
+
"loss": 0.0446,
|
| 1236 |
+
"step": 8650
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 2.912621359223301,
|
| 1240 |
+
"grad_norm": 0.16038326919078827,
|
| 1241 |
+
"learning_rate": 2.574210467581743e-05,
|
| 1242 |
+
"loss": 0.0456,
|
| 1243 |
+
"step": 8700
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 2.929360562437228,
|
| 1247 |
+
"grad_norm": 0.22738413512706757,
|
| 1248 |
+
"learning_rate": 2.5602611315701375e-05,
|
| 1249 |
+
"loss": 0.0479,
|
| 1250 |
+
"step": 8750
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 2.946099765651155,
|
| 1254 |
+
"grad_norm": 0.20327210426330566,
|
| 1255 |
+
"learning_rate": 2.5463117955585315e-05,
|
| 1256 |
+
"loss": 0.0511,
|
| 1257 |
+
"step": 8800
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 2.962838968865082,
|
| 1261 |
+
"grad_norm": 0.15756353735923767,
|
| 1262 |
+
"learning_rate": 2.532362459546926e-05,
|
| 1263 |
+
"loss": 0.0426,
|
| 1264 |
+
"step": 8850
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 2.979578172079009,
|
| 1268 |
+
"grad_norm": 0.1305045783519745,
|
| 1269 |
+
"learning_rate": 2.5184131235353197e-05,
|
| 1270 |
+
"loss": 0.0442,
|
| 1271 |
+
"step": 8900
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 2.996317375292936,
|
| 1275 |
+
"grad_norm": 0.1610562801361084,
|
| 1276 |
+
"learning_rate": 2.5044637875237144e-05,
|
| 1277 |
+
"loss": 0.0467,
|
| 1278 |
+
"step": 8950
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 3.0,
|
| 1282 |
+
"eval_loss": 0.04702676460146904,
|
| 1283 |
+
"eval_runtime": 52.8989,
|
| 1284 |
+
"eval_samples_per_second": 225.865,
|
| 1285 |
+
"eval_steps_per_second": 14.121,
|
| 1286 |
+
"step": 8961
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 3.013056578506863,
|
| 1290 |
+
"grad_norm": 0.2042045295238495,
|
| 1291 |
+
"learning_rate": 2.490514451512108e-05,
|
| 1292 |
+
"loss": 0.0457,
|
| 1293 |
+
"step": 9000
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 3.02979578172079,
|
| 1297 |
+
"grad_norm": 0.27092283964157104,
|
| 1298 |
+
"learning_rate": 2.4765651155005022e-05,
|
| 1299 |
+
"loss": 0.0437,
|
| 1300 |
+
"step": 9050
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 3.046534984934717,
|
| 1304 |
+
"grad_norm": 0.18729308247566223,
|
| 1305 |
+
"learning_rate": 2.4626157794888966e-05,
|
| 1306 |
+
"loss": 0.039,
|
| 1307 |
+
"step": 9100
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 3.063274188148644,
|
| 1311 |
+
"grad_norm": 0.18845289945602417,
|
| 1312 |
+
"learning_rate": 2.4486664434772903e-05,
|
| 1313 |
+
"loss": 0.0422,
|
| 1314 |
+
"step": 9150
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 3.0800133913625714,
|
| 1318 |
+
"grad_norm": 0.17593874037265778,
|
| 1319 |
+
"learning_rate": 2.4347171074656847e-05,
|
| 1320 |
+
"loss": 0.0384,
|
| 1321 |
+
"step": 9200
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 3.096752594576498,
|
| 1325 |
+
"grad_norm": 0.23149123787879944,
|
| 1326 |
+
"learning_rate": 2.4207677714540788e-05,
|
| 1327 |
+
"loss": 0.0443,
|
| 1328 |
+
"step": 9250
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 3.113491797790425,
|
| 1332 |
+
"grad_norm": 0.1968075931072235,
|
| 1333 |
+
"learning_rate": 2.4068184354424732e-05,
|
| 1334 |
+
"loss": 0.0402,
|
| 1335 |
+
"step": 9300
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 3.1302310010043524,
|
| 1339 |
+
"grad_norm": 0.16893354058265686,
|
| 1340 |
+
"learning_rate": 2.3928690994308673e-05,
|
| 1341 |
+
"loss": 0.043,
|
| 1342 |
+
"step": 9350
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 3.146970204218279,
|
| 1346 |
+
"grad_norm": 0.161103293299675,
|
| 1347 |
+
"learning_rate": 2.3789197634192613e-05,
|
| 1348 |
+
"loss": 0.0414,
|
| 1349 |
+
"step": 9400
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 3.163709407432206,
|
| 1353 |
+
"grad_norm": 0.22712625563144684,
|
| 1354 |
+
"learning_rate": 2.3649704274076557e-05,
|
| 1355 |
+
"loss": 0.0426,
|
| 1356 |
+
"step": 9450
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 3.1804486106461334,
|
| 1360 |
+
"grad_norm": 0.12811464071273804,
|
| 1361 |
+
"learning_rate": 2.3510210913960495e-05,
|
| 1362 |
+
"loss": 0.0406,
|
| 1363 |
+
"step": 9500
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 3.1971878138600602,
|
| 1367 |
+
"grad_norm": 0.16762731969356537,
|
| 1368 |
+
"learning_rate": 2.337071755384444e-05,
|
| 1369 |
+
"loss": 0.0417,
|
| 1370 |
+
"step": 9550
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 3.213927017073987,
|
| 1374 |
+
"grad_norm": 0.14003124833106995,
|
| 1375 |
+
"learning_rate": 2.323122419372838e-05,
|
| 1376 |
+
"loss": 0.0453,
|
| 1377 |
+
"step": 9600
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 3.2306662202879144,
|
| 1381 |
+
"grad_norm": 0.16891060769557953,
|
| 1382 |
+
"learning_rate": 2.309173083361232e-05,
|
| 1383 |
+
"loss": 0.0445,
|
| 1384 |
+
"step": 9650
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 3.2474054235018412,
|
| 1388 |
+
"grad_norm": 0.16900208592414856,
|
| 1389 |
+
"learning_rate": 2.2952237473496264e-05,
|
| 1390 |
+
"loss": 0.039,
|
| 1391 |
+
"step": 9700
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 3.264144626715768,
|
| 1395 |
+
"grad_norm": 0.17154955863952637,
|
| 1396 |
+
"learning_rate": 2.2812744113380205e-05,
|
| 1397 |
+
"loss": 0.0403,
|
| 1398 |
+
"step": 9750
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 3.2808838299296954,
|
| 1402 |
+
"grad_norm": 0.1620296835899353,
|
| 1403 |
+
"learning_rate": 2.2673250753264145e-05,
|
| 1404 |
+
"loss": 0.0406,
|
| 1405 |
+
"step": 9800
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 3.2976230331436223,
|
| 1409 |
+
"grad_norm": 0.14487063884735107,
|
| 1410 |
+
"learning_rate": 2.2533757393148086e-05,
|
| 1411 |
+
"loss": 0.0467,
|
| 1412 |
+
"step": 9850
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 3.3143622363575496,
|
| 1416 |
+
"grad_norm": 0.13799156248569489,
|
| 1417 |
+
"learning_rate": 2.2394264033032027e-05,
|
| 1418 |
+
"loss": 0.0433,
|
| 1419 |
+
"step": 9900
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 3.3311014395714764,
|
| 1423 |
+
"grad_norm": 0.1507265418767929,
|
| 1424 |
+
"learning_rate": 2.225477067291597e-05,
|
| 1425 |
+
"loss": 0.0446,
|
| 1426 |
+
"step": 9950
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 3.3478406427854033,
|
| 1430 |
+
"grad_norm": 0.1605840027332306,
|
| 1431 |
+
"learning_rate": 2.211527731279991e-05,
|
| 1432 |
+
"loss": 0.0415,
|
| 1433 |
+
"step": 10000
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 3.3645798459993306,
|
| 1437 |
+
"grad_norm": 0.11910756677389145,
|
| 1438 |
+
"learning_rate": 2.1975783952683855e-05,
|
| 1439 |
+
"loss": 0.0453,
|
| 1440 |
+
"step": 10050
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 3.3813190492132574,
|
| 1444 |
+
"grad_norm": 0.15115414559841156,
|
| 1445 |
+
"learning_rate": 2.1836290592567796e-05,
|
| 1446 |
+
"loss": 0.0395,
|
| 1447 |
+
"step": 10100
|
| 1448 |
+
},
|
| 1449 |
+
{
|
| 1450 |
+
"epoch": 3.3980582524271843,
|
| 1451 |
+
"grad_norm": 0.2044568508863449,
|
| 1452 |
+
"learning_rate": 2.1696797232451737e-05,
|
| 1453 |
+
"loss": 0.0473,
|
| 1454 |
+
"step": 10150
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 3.4147974556411116,
|
| 1458 |
+
"grad_norm": 0.18123552203178406,
|
| 1459 |
+
"learning_rate": 2.1557303872335677e-05,
|
| 1460 |
+
"loss": 0.0411,
|
| 1461 |
+
"step": 10200
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 3.4315366588550384,
|
| 1465 |
+
"grad_norm": 0.6018120646476746,
|
| 1466 |
+
"learning_rate": 2.1417810512219618e-05,
|
| 1467 |
+
"loss": 0.0413,
|
| 1468 |
+
"step": 10250
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 3.4482758620689653,
|
| 1472 |
+
"grad_norm": 0.22490189969539642,
|
| 1473 |
+
"learning_rate": 2.1278317152103562e-05,
|
| 1474 |
+
"loss": 0.0437,
|
| 1475 |
+
"step": 10300
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 3.4650150652828926,
|
| 1479 |
+
"grad_norm": 0.1784990429878235,
|
| 1480 |
+
"learning_rate": 2.1138823791987503e-05,
|
| 1481 |
+
"loss": 0.0453,
|
| 1482 |
+
"step": 10350
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 3.4817542684968195,
|
| 1486 |
+
"grad_norm": 0.15248402953147888,
|
| 1487 |
+
"learning_rate": 2.0999330431871443e-05,
|
| 1488 |
+
"loss": 0.0456,
|
| 1489 |
+
"step": 10400
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 3.4984934717107468,
|
| 1493 |
+
"grad_norm": 0.15146291255950928,
|
| 1494 |
+
"learning_rate": 2.0859837071755387e-05,
|
| 1495 |
+
"loss": 0.0393,
|
| 1496 |
+
"step": 10450
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 3.5152326749246736,
|
| 1500 |
+
"grad_norm": 0.1662750393152237,
|
| 1501 |
+
"learning_rate": 2.0720343711639325e-05,
|
| 1502 |
+
"loss": 0.0408,
|
| 1503 |
+
"step": 10500
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 3.5319718781386005,
|
| 1507 |
+
"grad_norm": 0.1244506984949112,
|
| 1508 |
+
"learning_rate": 2.058085035152327e-05,
|
| 1509 |
+
"loss": 0.0426,
|
| 1510 |
+
"step": 10550
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 3.5487110813525278,
|
| 1514 |
+
"grad_norm": 0.1180344969034195,
|
| 1515 |
+
"learning_rate": 2.044135699140721e-05,
|
| 1516 |
+
"loss": 0.0434,
|
| 1517 |
+
"step": 10600
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 3.5654502845664546,
|
| 1521 |
+
"grad_norm": 0.15951013565063477,
|
| 1522 |
+
"learning_rate": 2.030186363129115e-05,
|
| 1523 |
+
"loss": 0.0387,
|
| 1524 |
+
"step": 10650
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 3.582189487780382,
|
| 1528 |
+
"grad_norm": 0.16064217686653137,
|
| 1529 |
+
"learning_rate": 2.0162370271175094e-05,
|
| 1530 |
+
"loss": 0.0445,
|
| 1531 |
+
"step": 10700
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 3.598928690994309,
|
| 1535 |
+
"grad_norm": 0.17813698947429657,
|
| 1536 |
+
"learning_rate": 2.0022876911059035e-05,
|
| 1537 |
+
"loss": 0.0407,
|
| 1538 |
+
"step": 10750
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 3.6156678942082356,
|
| 1542 |
+
"grad_norm": 0.1256450116634369,
|
| 1543 |
+
"learning_rate": 1.988338355094298e-05,
|
| 1544 |
+
"loss": 0.0462,
|
| 1545 |
+
"step": 10800
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 3.6324070974221625,
|
| 1549 |
+
"grad_norm": 0.14016403257846832,
|
| 1550 |
+
"learning_rate": 1.9743890190826916e-05,
|
| 1551 |
+
"loss": 0.0456,
|
| 1552 |
+
"step": 10850
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 3.64914630063609,
|
| 1556 |
+
"grad_norm": 0.1396850347518921,
|
| 1557 |
+
"learning_rate": 1.9604396830710857e-05,
|
| 1558 |
+
"loss": 0.0403,
|
| 1559 |
+
"step": 10900
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 3.6658855038500167,
|
| 1563 |
+
"grad_norm": 0.17943057417869568,
|
| 1564 |
+
"learning_rate": 1.94649034705948e-05,
|
| 1565 |
+
"loss": 0.0417,
|
| 1566 |
+
"step": 10950
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 3.682624707063944,
|
| 1570 |
+
"grad_norm": 0.14947953820228577,
|
| 1571 |
+
"learning_rate": 1.932541011047874e-05,
|
| 1572 |
+
"loss": 0.0421,
|
| 1573 |
+
"step": 11000
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 3.699363910277871,
|
| 1577 |
+
"grad_norm": 0.12628613412380219,
|
| 1578 |
+
"learning_rate": 1.9185916750362685e-05,
|
| 1579 |
+
"loss": 0.0435,
|
| 1580 |
+
"step": 11050
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 3.7161031134917977,
|
| 1584 |
+
"grad_norm": 0.2205984890460968,
|
| 1585 |
+
"learning_rate": 1.9046423390246626e-05,
|
| 1586 |
+
"loss": 0.0396,
|
| 1587 |
+
"step": 11100
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 3.732842316705725,
|
| 1591 |
+
"grad_norm": 0.13236357271671295,
|
| 1592 |
+
"learning_rate": 1.8906930030130567e-05,
|
| 1593 |
+
"loss": 0.0405,
|
| 1594 |
+
"step": 11150
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 3.749581519919652,
|
| 1598 |
+
"grad_norm": 0.15023528039455414,
|
| 1599 |
+
"learning_rate": 1.8767436670014507e-05,
|
| 1600 |
+
"loss": 0.0434,
|
| 1601 |
+
"step": 11200
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 3.7663207231335787,
|
| 1605 |
+
"grad_norm": 0.1427326649427414,
|
| 1606 |
+
"learning_rate": 1.8627943309898448e-05,
|
| 1607 |
+
"loss": 0.0437,
|
| 1608 |
+
"step": 11250
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 3.783059926347506,
|
| 1612 |
+
"grad_norm": 0.1890624761581421,
|
| 1613 |
+
"learning_rate": 1.8488449949782392e-05,
|
| 1614 |
+
"loss": 0.0408,
|
| 1615 |
+
"step": 11300
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 3.799799129561433,
|
| 1619 |
+
"grad_norm": 0.27970972657203674,
|
| 1620 |
+
"learning_rate": 1.8348956589666333e-05,
|
| 1621 |
+
"loss": 0.0441,
|
| 1622 |
+
"step": 11350
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 3.8165383327753597,
|
| 1626 |
+
"grad_norm": 0.12823455035686493,
|
| 1627 |
+
"learning_rate": 1.8209463229550273e-05,
|
| 1628 |
+
"loss": 0.0412,
|
| 1629 |
+
"step": 11400
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 3.833277535989287,
|
| 1633 |
+
"grad_norm": 0.1442965269088745,
|
| 1634 |
+
"learning_rate": 1.8069969869434218e-05,
|
| 1635 |
+
"loss": 0.0416,
|
| 1636 |
+
"step": 11450
|
| 1637 |
+
},
|
| 1638 |
+
{
|
| 1639 |
+
"epoch": 3.850016739203214,
|
| 1640 |
+
"grad_norm": 0.13739417493343353,
|
| 1641 |
+
"learning_rate": 1.7930476509318158e-05,
|
| 1642 |
+
"loss": 0.0397,
|
| 1643 |
+
"step": 11500
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 3.866755942417141,
|
| 1647 |
+
"grad_norm": 0.16616705060005188,
|
| 1648 |
+
"learning_rate": 1.77909831492021e-05,
|
| 1649 |
+
"loss": 0.0413,
|
| 1650 |
+
"step": 11550
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 3.883495145631068,
|
| 1654 |
+
"grad_norm": 0.23060384392738342,
|
| 1655 |
+
"learning_rate": 1.765148978908604e-05,
|
| 1656 |
+
"loss": 0.0447,
|
| 1657 |
+
"step": 11600
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 3.900234348844995,
|
| 1661 |
+
"grad_norm": 0.2936810553073883,
|
| 1662 |
+
"learning_rate": 1.751199642896998e-05,
|
| 1663 |
+
"loss": 0.0409,
|
| 1664 |
+
"step": 11650
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 3.9169735520589217,
|
| 1668 |
+
"grad_norm": 0.17367126047611237,
|
| 1669 |
+
"learning_rate": 1.7372503068853924e-05,
|
| 1670 |
+
"loss": 0.0411,
|
| 1671 |
+
"step": 11700
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 3.933712755272849,
|
| 1675 |
+
"grad_norm": 0.14550547301769257,
|
| 1676 |
+
"learning_rate": 1.7233009708737865e-05,
|
| 1677 |
+
"loss": 0.0445,
|
| 1678 |
+
"step": 11750
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 3.950451958486776,
|
| 1682 |
+
"grad_norm": 0.13322454690933228,
|
| 1683 |
+
"learning_rate": 1.709351634862181e-05,
|
| 1684 |
+
"loss": 0.0444,
|
| 1685 |
+
"step": 11800
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 3.967191161700703,
|
| 1689 |
+
"grad_norm": 0.13606959581375122,
|
| 1690 |
+
"learning_rate": 1.6954022988505746e-05,
|
| 1691 |
+
"loss": 0.041,
|
| 1692 |
+
"step": 11850
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 3.98393036491463,
|
| 1696 |
+
"grad_norm": 0.14227426052093506,
|
| 1697 |
+
"learning_rate": 1.681452962838969e-05,
|
| 1698 |
+
"loss": 0.0412,
|
| 1699 |
+
"step": 11900
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 4.0,
|
| 1703 |
+
"eval_loss": 0.045825306326150894,
|
| 1704 |
+
"eval_runtime": 48.668,
|
| 1705 |
+
"eval_samples_per_second": 245.5,
|
| 1706 |
+
"eval_steps_per_second": 15.349,
|
| 1707 |
+
"step": 11948
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 4.000669568128557,
|
| 1711 |
+
"grad_norm": 0.14429377019405365,
|
| 1712 |
+
"learning_rate": 1.667503626827363e-05,
|
| 1713 |
+
"loss": 0.0398,
|
| 1714 |
+
"step": 11950
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 4.017408771342484,
|
| 1718 |
+
"grad_norm": 0.13842210173606873,
|
| 1719 |
+
"learning_rate": 1.653554290815757e-05,
|
| 1720 |
+
"loss": 0.0365,
|
| 1721 |
+
"step": 12000
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 4.0341479745564115,
|
| 1725 |
+
"grad_norm": 0.2168322205543518,
|
| 1726 |
+
"learning_rate": 1.6396049548041516e-05,
|
| 1727 |
+
"loss": 0.0395,
|
| 1728 |
+
"step": 12050
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 4.050887177770338,
|
| 1732 |
+
"grad_norm": 0.1534542292356491,
|
| 1733 |
+
"learning_rate": 1.6256556187925456e-05,
|
| 1734 |
+
"loss": 0.0377,
|
| 1735 |
+
"step": 12100
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 4.067626380984265,
|
| 1739 |
+
"grad_norm": 0.13067083060741425,
|
| 1740 |
+
"learning_rate": 1.6117062827809397e-05,
|
| 1741 |
+
"loss": 0.042,
|
| 1742 |
+
"step": 12150
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 4.084365584198192,
|
| 1746 |
+
"grad_norm": 0.13599875569343567,
|
| 1747 |
+
"learning_rate": 1.5977569467693338e-05,
|
| 1748 |
+
"loss": 0.0403,
|
| 1749 |
+
"step": 12200
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 4.101104787412119,
|
| 1753 |
+
"grad_norm": 0.23121419548988342,
|
| 1754 |
+
"learning_rate": 1.5838076107577278e-05,
|
| 1755 |
+
"loss": 0.0406,
|
| 1756 |
+
"step": 12250
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 4.117843990626046,
|
| 1760 |
+
"grad_norm": 0.15491856634616852,
|
| 1761 |
+
"learning_rate": 1.5698582747461222e-05,
|
| 1762 |
+
"loss": 0.0381,
|
| 1763 |
+
"step": 12300
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 4.1345831938399735,
|
| 1767 |
+
"grad_norm": 0.13168101012706757,
|
| 1768 |
+
"learning_rate": 1.5559089387345163e-05,
|
| 1769 |
+
"loss": 0.0353,
|
| 1770 |
+
"step": 12350
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 4.1513223970539,
|
| 1774 |
+
"grad_norm": 0.1694163829088211,
|
| 1775 |
+
"learning_rate": 1.5419596027229104e-05,
|
| 1776 |
+
"loss": 0.04,
|
| 1777 |
+
"step": 12400
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 4.168061600267827,
|
| 1781 |
+
"grad_norm": 0.17727704346179962,
|
| 1782 |
+
"learning_rate": 1.5280102667113048e-05,
|
| 1783 |
+
"loss": 0.0364,
|
| 1784 |
+
"step": 12450
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 4.184800803481754,
|
| 1788 |
+
"grad_norm": 0.18948902189731598,
|
| 1789 |
+
"learning_rate": 1.5140609306996988e-05,
|
| 1790 |
+
"loss": 0.0372,
|
| 1791 |
+
"step": 12500
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 4.201540006695681,
|
| 1795 |
+
"grad_norm": 0.2130223959684372,
|
| 1796 |
+
"learning_rate": 1.5001115946880929e-05,
|
| 1797 |
+
"loss": 0.0398,
|
| 1798 |
+
"step": 12550
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 4.218279209909609,
|
| 1802 |
+
"grad_norm": 0.15893810987472534,
|
| 1803 |
+
"learning_rate": 1.486162258676487e-05,
|
| 1804 |
+
"loss": 0.0387,
|
| 1805 |
+
"step": 12600
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 4.2350184131235356,
|
| 1809 |
+
"grad_norm": 0.19536694884300232,
|
| 1810 |
+
"learning_rate": 1.4722129226648812e-05,
|
| 1811 |
+
"loss": 0.041,
|
| 1812 |
+
"step": 12650
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 4.251757616337462,
|
| 1816 |
+
"grad_norm": 0.16755405068397522,
|
| 1817 |
+
"learning_rate": 1.4582635866532754e-05,
|
| 1818 |
+
"loss": 0.0367,
|
| 1819 |
+
"step": 12700
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"epoch": 4.268496819551389,
|
| 1823 |
+
"grad_norm": 0.14820334315299988,
|
| 1824 |
+
"learning_rate": 1.4443142506416697e-05,
|
| 1825 |
+
"loss": 0.0382,
|
| 1826 |
+
"step": 12750
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 4.285236022765316,
|
| 1830 |
+
"grad_norm": 0.21781007945537567,
|
| 1831 |
+
"learning_rate": 1.4303649146300637e-05,
|
| 1832 |
+
"loss": 0.0398,
|
| 1833 |
+
"step": 12800
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 4.301975225979243,
|
| 1837 |
+
"grad_norm": 0.17162373661994934,
|
| 1838 |
+
"learning_rate": 1.4164155786184576e-05,
|
| 1839 |
+
"loss": 0.042,
|
| 1840 |
+
"step": 12850
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 4.318714429193171,
|
| 1844 |
+
"grad_norm": 0.13765838742256165,
|
| 1845 |
+
"learning_rate": 1.4024662426068519e-05,
|
| 1846 |
+
"loss": 0.0378,
|
| 1847 |
+
"step": 12900
|
| 1848 |
+
},
|
| 1849 |
+
{
|
| 1850 |
+
"epoch": 4.335453632407098,
|
| 1851 |
+
"grad_norm": 0.22830092906951904,
|
| 1852 |
+
"learning_rate": 1.3885169065952461e-05,
|
| 1853 |
+
"loss": 0.0387,
|
| 1854 |
+
"step": 12950
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 4.352192835621024,
|
| 1858 |
+
"grad_norm": 0.1682949960231781,
|
| 1859 |
+
"learning_rate": 1.3745675705836403e-05,
|
| 1860 |
+
"loss": 0.0408,
|
| 1861 |
+
"step": 13000
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"epoch": 4.368932038834951,
|
| 1865 |
+
"grad_norm": 0.16153910756111145,
|
| 1866 |
+
"learning_rate": 1.3606182345720344e-05,
|
| 1867 |
+
"loss": 0.0405,
|
| 1868 |
+
"step": 13050
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 4.385671242048878,
|
| 1872 |
+
"grad_norm": 0.18771956861019135,
|
| 1873 |
+
"learning_rate": 1.3466688985604286e-05,
|
| 1874 |
+
"loss": 0.0427,
|
| 1875 |
+
"step": 13100
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 4.402410445262806,
|
| 1879 |
+
"grad_norm": 0.11162823438644409,
|
| 1880 |
+
"learning_rate": 1.3327195625488229e-05,
|
| 1881 |
+
"loss": 0.0399,
|
| 1882 |
+
"step": 13150
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 4.419149648476733,
|
| 1886 |
+
"grad_norm": 0.14466647803783417,
|
| 1887 |
+
"learning_rate": 1.3187702265372168e-05,
|
| 1888 |
+
"loss": 0.0401,
|
| 1889 |
+
"step": 13200
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 4.43588885169066,
|
| 1893 |
+
"grad_norm": 0.14834430813789368,
|
| 1894 |
+
"learning_rate": 1.304820890525611e-05,
|
| 1895 |
+
"loss": 0.0386,
|
| 1896 |
+
"step": 13250
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 4.4526280549045865,
|
| 1900 |
+
"grad_norm": 0.20324522256851196,
|
| 1901 |
+
"learning_rate": 1.2908715545140052e-05,
|
| 1902 |
+
"loss": 0.0379,
|
| 1903 |
+
"step": 13300
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 4.469367258118513,
|
| 1907 |
+
"grad_norm": 0.12625128030776978,
|
| 1908 |
+
"learning_rate": 1.2769222185023993e-05,
|
| 1909 |
+
"loss": 0.0403,
|
| 1910 |
+
"step": 13350
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 4.48610646133244,
|
| 1914 |
+
"grad_norm": 0.11018920689821243,
|
| 1915 |
+
"learning_rate": 1.2629728824907935e-05,
|
| 1916 |
+
"loss": 0.0383,
|
| 1917 |
+
"step": 13400
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 4.502845664546368,
|
| 1921 |
+
"grad_norm": 0.18631067872047424,
|
| 1922 |
+
"learning_rate": 1.2490235464791876e-05,
|
| 1923 |
+
"loss": 0.0424,
|
| 1924 |
+
"step": 13450
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 4.519584867760295,
|
| 1928 |
+
"grad_norm": 0.28578242659568787,
|
| 1929 |
+
"learning_rate": 1.2350742104675818e-05,
|
| 1930 |
+
"loss": 0.0385,
|
| 1931 |
+
"step": 13500
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 4.536324070974222,
|
| 1935 |
+
"grad_norm": 0.1431867629289627,
|
| 1936 |
+
"learning_rate": 1.2211248744559759e-05,
|
| 1937 |
+
"loss": 0.038,
|
| 1938 |
+
"step": 13550
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 4.5530632741881485,
|
| 1942 |
+
"grad_norm": 0.12731611728668213,
|
| 1943 |
+
"learning_rate": 1.2071755384443701e-05,
|
| 1944 |
+
"loss": 0.0389,
|
| 1945 |
+
"step": 13600
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 4.569802477402075,
|
| 1949 |
+
"grad_norm": 0.20958903431892395,
|
| 1950 |
+
"learning_rate": 1.1932262024327642e-05,
|
| 1951 |
+
"loss": 0.0383,
|
| 1952 |
+
"step": 13650
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 4.586541680616003,
|
| 1956 |
+
"grad_norm": 0.21877717971801758,
|
| 1957 |
+
"learning_rate": 1.1792768664211584e-05,
|
| 1958 |
+
"loss": 0.0399,
|
| 1959 |
+
"step": 13700
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 4.60328088382993,
|
| 1963 |
+
"grad_norm": 0.1719764918088913,
|
| 1964 |
+
"learning_rate": 1.1653275304095527e-05,
|
| 1965 |
+
"loss": 0.0355,
|
| 1966 |
+
"step": 13750
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 4.620020087043857,
|
| 1970 |
+
"grad_norm": 0.13044840097427368,
|
| 1971 |
+
"learning_rate": 1.1513781943979467e-05,
|
| 1972 |
+
"loss": 0.0386,
|
| 1973 |
+
"step": 13800
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 4.636759290257784,
|
| 1977 |
+
"grad_norm": 0.20552796125411987,
|
| 1978 |
+
"learning_rate": 1.1374288583863408e-05,
|
| 1979 |
+
"loss": 0.04,
|
| 1980 |
+
"step": 13850
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 4.6534984934717105,
|
| 1984 |
+
"grad_norm": 0.15806210041046143,
|
| 1985 |
+
"learning_rate": 1.123479522374735e-05,
|
| 1986 |
+
"loss": 0.0381,
|
| 1987 |
+
"step": 13900
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 4.670237696685637,
|
| 1991 |
+
"grad_norm": 0.2317190170288086,
|
| 1992 |
+
"learning_rate": 1.1095301863631291e-05,
|
| 1993 |
+
"loss": 0.0394,
|
| 1994 |
+
"step": 13950
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 4.686976899899565,
|
| 1998 |
+
"grad_norm": 0.16183790564537048,
|
| 1999 |
+
"learning_rate": 1.0955808503515233e-05,
|
| 2000 |
+
"loss": 0.0392,
|
| 2001 |
+
"step": 14000
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 4.703716103113492,
|
| 2005 |
+
"grad_norm": 0.14107303321361542,
|
| 2006 |
+
"learning_rate": 1.0816315143399176e-05,
|
| 2007 |
+
"loss": 0.039,
|
| 2008 |
+
"step": 14050
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 4.720455306327419,
|
| 2012 |
+
"grad_norm": 0.21340009570121765,
|
| 2013 |
+
"learning_rate": 1.0676821783283116e-05,
|
| 2014 |
+
"loss": 0.04,
|
| 2015 |
+
"step": 14100
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 4.737194509541346,
|
| 2019 |
+
"grad_norm": 0.1754944771528244,
|
| 2020 |
+
"learning_rate": 1.0537328423167057e-05,
|
| 2021 |
+
"loss": 0.0371,
|
| 2022 |
+
"step": 14150
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 4.7539337127552725,
|
| 2026 |
+
"grad_norm": 0.13280175626277924,
|
| 2027 |
+
"learning_rate": 1.0397835063051e-05,
|
| 2028 |
+
"loss": 0.0358,
|
| 2029 |
+
"step": 14200
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"epoch": 4.7706729159692,
|
| 2033 |
+
"grad_norm": 0.15039420127868652,
|
| 2034 |
+
"learning_rate": 1.0258341702934942e-05,
|
| 2035 |
+
"loss": 0.0438,
|
| 2036 |
+
"step": 14250
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 4.787412119183127,
|
| 2040 |
+
"grad_norm": 0.30474409461021423,
|
| 2041 |
+
"learning_rate": 1.0118848342818882e-05,
|
| 2042 |
+
"loss": 0.0377,
|
| 2043 |
+
"step": 14300
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 4.804151322397054,
|
| 2047 |
+
"grad_norm": 0.2012936919927597,
|
| 2048 |
+
"learning_rate": 9.979354982702823e-06,
|
| 2049 |
+
"loss": 0.0417,
|
| 2050 |
+
"step": 14350
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 4.820890525610981,
|
| 2054 |
+
"grad_norm": 0.192657932639122,
|
| 2055 |
+
"learning_rate": 9.839861622586765e-06,
|
| 2056 |
+
"loss": 0.0392,
|
| 2057 |
+
"step": 14400
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 4.837629728824908,
|
| 2061 |
+
"grad_norm": 0.16141368448734283,
|
| 2062 |
+
"learning_rate": 9.700368262470706e-06,
|
| 2063 |
+
"loss": 0.0378,
|
| 2064 |
+
"step": 14450
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 4.854368932038835,
|
| 2068 |
+
"grad_norm": 0.19473744928836823,
|
| 2069 |
+
"learning_rate": 9.560874902354648e-06,
|
| 2070 |
+
"loss": 0.0368,
|
| 2071 |
+
"step": 14500
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"epoch": 4.871108135252762,
|
| 2075 |
+
"grad_norm": 0.15624327957630157,
|
| 2076 |
+
"learning_rate": 9.42138154223859e-06,
|
| 2077 |
+
"loss": 0.0416,
|
| 2078 |
+
"step": 14550
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 4.887847338466689,
|
| 2082 |
+
"grad_norm": 0.1572103202342987,
|
| 2083 |
+
"learning_rate": 9.281888182122531e-06,
|
| 2084 |
+
"loss": 0.0383,
|
| 2085 |
+
"step": 14600
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 4.904586541680616,
|
| 2089 |
+
"grad_norm": 0.15121281147003174,
|
| 2090 |
+
"learning_rate": 9.142394822006472e-06,
|
| 2091 |
+
"loss": 0.0405,
|
| 2092 |
+
"step": 14650
|
| 2093 |
+
},
|
| 2094 |
+
{
|
| 2095 |
+
"epoch": 4.921325744894543,
|
| 2096 |
+
"grad_norm": 0.1739313155412674,
|
| 2097 |
+
"learning_rate": 9.002901461890415e-06,
|
| 2098 |
+
"loss": 0.0389,
|
| 2099 |
+
"step": 14700
|
| 2100 |
+
},
|
| 2101 |
+
{
|
| 2102 |
+
"epoch": 4.93806494810847,
|
| 2103 |
+
"grad_norm": 0.12826618552207947,
|
| 2104 |
+
"learning_rate": 8.863408101774357e-06,
|
| 2105 |
+
"loss": 0.0385,
|
| 2106 |
+
"step": 14750
|
| 2107 |
+
},
|
| 2108 |
+
{
|
| 2109 |
+
"epoch": 4.9548041513223975,
|
| 2110 |
+
"grad_norm": 0.11419858038425446,
|
| 2111 |
+
"learning_rate": 8.723914741658298e-06,
|
| 2112 |
+
"loss": 0.0387,
|
| 2113 |
+
"step": 14800
|
| 2114 |
+
},
|
| 2115 |
+
{
|
| 2116 |
+
"epoch": 4.971543354536324,
|
| 2117 |
+
"grad_norm": 0.18640589714050293,
|
| 2118 |
+
"learning_rate": 8.584421381542238e-06,
|
| 2119 |
+
"loss": 0.0376,
|
| 2120 |
+
"step": 14850
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 4.988282557750251,
|
| 2124 |
+
"grad_norm": 0.18249401450157166,
|
| 2125 |
+
"learning_rate": 8.44492802142618e-06,
|
| 2126 |
+
"loss": 0.041,
|
| 2127 |
+
"step": 14900
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 5.0,
|
| 2131 |
+
"eval_loss": 0.045213617384433746,
|
| 2132 |
+
"eval_runtime": 49.0497,
|
| 2133 |
+
"eval_samples_per_second": 243.589,
|
| 2134 |
+
"eval_steps_per_second": 15.229,
|
| 2135 |
+
"step": 14935
|
| 2136 |
+
},
|
| 2137 |
+
{
|
| 2138 |
+
"epoch": 5.005021760964178,
|
| 2139 |
+
"grad_norm": 0.25643524527549744,
|
| 2140 |
+
"learning_rate": 8.305434661310121e-06,
|
| 2141 |
+
"loss": 0.0403,
|
| 2142 |
+
"step": 14950
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 5.021760964178105,
|
| 2146 |
+
"grad_norm": 0.16471606492996216,
|
| 2147 |
+
"learning_rate": 8.165941301194064e-06,
|
| 2148 |
+
"loss": 0.0374,
|
| 2149 |
+
"step": 15000
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 5.038500167392032,
|
| 2153 |
+
"grad_norm": 0.14938130974769592,
|
| 2154 |
+
"learning_rate": 8.026447941078006e-06,
|
| 2155 |
+
"loss": 0.0363,
|
| 2156 |
+
"step": 15050
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 5.0552393706059595,
|
| 2160 |
+
"grad_norm": 0.13239839673042297,
|
| 2161 |
+
"learning_rate": 7.886954580961947e-06,
|
| 2162 |
+
"loss": 0.037,
|
| 2163 |
+
"step": 15100
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 5.071978573819886,
|
| 2167 |
+
"grad_norm": 0.17447875440120697,
|
| 2168 |
+
"learning_rate": 7.747461220845887e-06,
|
| 2169 |
+
"loss": 0.0374,
|
| 2170 |
+
"step": 15150
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 5.088717777033813,
|
| 2174 |
+
"grad_norm": 0.15902550518512726,
|
| 2175 |
+
"learning_rate": 7.6079678607298296e-06,
|
| 2176 |
+
"loss": 0.0361,
|
| 2177 |
+
"step": 15200
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 5.10545698024774,
|
| 2181 |
+
"grad_norm": 0.22053466737270355,
|
| 2182 |
+
"learning_rate": 7.468474500613772e-06,
|
| 2183 |
+
"loss": 0.0381,
|
| 2184 |
+
"step": 15250
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 5.122196183461667,
|
| 2188 |
+
"grad_norm": 0.1754140853881836,
|
| 2189 |
+
"learning_rate": 7.328981140497712e-06,
|
| 2190 |
+
"loss": 0.0327,
|
| 2191 |
+
"step": 15300
|
| 2192 |
+
},
|
| 2193 |
+
{
|
| 2194 |
+
"epoch": 5.138935386675595,
|
| 2195 |
+
"grad_norm": 0.15227694809436798,
|
| 2196 |
+
"learning_rate": 7.189487780381654e-06,
|
| 2197 |
+
"loss": 0.0343,
|
| 2198 |
+
"step": 15350
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 5.1556745898895215,
|
| 2202 |
+
"grad_norm": 0.1644590198993683,
|
| 2203 |
+
"learning_rate": 7.049994420265596e-06,
|
| 2204 |
+
"loss": 0.0373,
|
| 2205 |
+
"step": 15400
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 5.172413793103448,
|
| 2209 |
+
"grad_norm": 0.13261474668979645,
|
| 2210 |
+
"learning_rate": 6.910501060149536e-06,
|
| 2211 |
+
"loss": 0.0354,
|
| 2212 |
+
"step": 15450
|
| 2213 |
+
},
|
| 2214 |
+
{
|
| 2215 |
+
"epoch": 5.189152996317375,
|
| 2216 |
+
"grad_norm": 0.16326355934143066,
|
| 2217 |
+
"learning_rate": 6.771007700033479e-06,
|
| 2218 |
+
"loss": 0.0368,
|
| 2219 |
+
"step": 15500
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"epoch": 5.205892199531302,
|
| 2223 |
+
"grad_norm": 0.21868841350078583,
|
| 2224 |
+
"learning_rate": 6.63151433991742e-06,
|
| 2225 |
+
"loss": 0.0384,
|
| 2226 |
+
"step": 15550
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"epoch": 5.222631402745229,
|
| 2230 |
+
"grad_norm": 0.1279917061328888,
|
| 2231 |
+
"learning_rate": 6.4920209798013624e-06,
|
| 2232 |
+
"loss": 0.037,
|
| 2233 |
+
"step": 15600
|
| 2234 |
+
},
|
| 2235 |
+
{
|
| 2236 |
+
"epoch": 5.239370605959157,
|
| 2237 |
+
"grad_norm": 0.14255809783935547,
|
| 2238 |
+
"learning_rate": 6.352527619685303e-06,
|
| 2239 |
+
"loss": 0.0345,
|
| 2240 |
+
"step": 15650
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"epoch": 5.256109809173084,
|
| 2244 |
+
"grad_norm": 0.15950387716293335,
|
| 2245 |
+
"learning_rate": 6.213034259569245e-06,
|
| 2246 |
+
"loss": 0.0389,
|
| 2247 |
+
"step": 15700
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 5.27284901238701,
|
| 2251 |
+
"grad_norm": 0.1789381355047226,
|
| 2252 |
+
"learning_rate": 6.073540899453186e-06,
|
| 2253 |
+
"loss": 0.0368,
|
| 2254 |
+
"step": 15750
|
| 2255 |
+
},
|
| 2256 |
+
{
|
| 2257 |
+
"epoch": 5.289588215600937,
|
| 2258 |
+
"grad_norm": 0.17775952816009521,
|
| 2259 |
+
"learning_rate": 5.934047539337128e-06,
|
| 2260 |
+
"loss": 0.0367,
|
| 2261 |
+
"step": 15800
|
| 2262 |
+
},
|
| 2263 |
+
{
|
| 2264 |
+
"epoch": 5.306327418814864,
|
| 2265 |
+
"grad_norm": 0.16045907139778137,
|
| 2266 |
+
"learning_rate": 5.794554179221069e-06,
|
| 2267 |
+
"loss": 0.0363,
|
| 2268 |
+
"step": 15850
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 5.323066622028792,
|
| 2272 |
+
"grad_norm": 0.14778949320316315,
|
| 2273 |
+
"learning_rate": 5.6550608191050115e-06,
|
| 2274 |
+
"loss": 0.038,
|
| 2275 |
+
"step": 15900
|
| 2276 |
+
},
|
| 2277 |
+
{
|
| 2278 |
+
"epoch": 5.339805825242719,
|
| 2279 |
+
"grad_norm": 0.13788272440433502,
|
| 2280 |
+
"learning_rate": 5.515567458988952e-06,
|
| 2281 |
+
"loss": 0.0336,
|
| 2282 |
+
"step": 15950
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 5.356545028456646,
|
| 2286 |
+
"grad_norm": 0.14058952033519745,
|
| 2287 |
+
"learning_rate": 5.376074098872894e-06,
|
| 2288 |
+
"loss": 0.0371,
|
| 2289 |
+
"step": 16000
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"epoch": 5.3732842316705725,
|
| 2293 |
+
"grad_norm": 0.13372714817523956,
|
| 2294 |
+
"learning_rate": 5.236580738756835e-06,
|
| 2295 |
+
"loss": 0.0388,
|
| 2296 |
+
"step": 16050
|
| 2297 |
+
},
|
| 2298 |
+
{
|
| 2299 |
+
"epoch": 5.390023434884499,
|
| 2300 |
+
"grad_norm": 0.09582552313804626,
|
| 2301 |
+
"learning_rate": 5.097087378640777e-06,
|
| 2302 |
+
"loss": 0.0345,
|
| 2303 |
+
"step": 16100
|
| 2304 |
+
},
|
| 2305 |
+
{
|
| 2306 |
+
"epoch": 5.406762638098426,
|
| 2307 |
+
"grad_norm": 0.14247213304042816,
|
| 2308 |
+
"learning_rate": 4.957594018524719e-06,
|
| 2309 |
+
"loss": 0.0348,
|
| 2310 |
+
"step": 16150
|
| 2311 |
+
},
|
| 2312 |
+
{
|
| 2313 |
+
"epoch": 5.423501841312354,
|
| 2314 |
+
"grad_norm": 0.18378828465938568,
|
| 2315 |
+
"learning_rate": 4.81810065840866e-06,
|
| 2316 |
+
"loss": 0.036,
|
| 2317 |
+
"step": 16200
|
| 2318 |
+
},
|
| 2319 |
+
{
|
| 2320 |
+
"epoch": 5.440241044526281,
|
| 2321 |
+
"grad_norm": 0.14933009445667267,
|
| 2322 |
+
"learning_rate": 4.678607298292601e-06,
|
| 2323 |
+
"loss": 0.0385,
|
| 2324 |
+
"step": 16250
|
| 2325 |
+
},
|
| 2326 |
+
{
|
| 2327 |
+
"epoch": 5.456980247740208,
|
| 2328 |
+
"grad_norm": 0.14775556325912476,
|
| 2329 |
+
"learning_rate": 4.5391139381765435e-06,
|
| 2330 |
+
"loss": 0.0381,
|
| 2331 |
+
"step": 16300
|
| 2332 |
+
},
|
| 2333 |
+
{
|
| 2334 |
+
"epoch": 5.4737194509541345,
|
| 2335 |
+
"grad_norm": 0.17397841811180115,
|
| 2336 |
+
"learning_rate": 4.399620578060484e-06,
|
| 2337 |
+
"loss": 0.0355,
|
| 2338 |
+
"step": 16350
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"epoch": 5.490458654168061,
|
| 2342 |
+
"grad_norm": 0.15232603251934052,
|
| 2343 |
+
"learning_rate": 4.2601272179444265e-06,
|
| 2344 |
+
"loss": 0.0363,
|
| 2345 |
+
"step": 16400
|
| 2346 |
+
},
|
| 2347 |
+
{
|
| 2348 |
+
"epoch": 5.507197857381989,
|
| 2349 |
+
"grad_norm": 0.1729612499475479,
|
| 2350 |
+
"learning_rate": 4.120633857828367e-06,
|
| 2351 |
+
"loss": 0.0356,
|
| 2352 |
+
"step": 16450
|
| 2353 |
+
},
|
| 2354 |
+
{
|
| 2355 |
+
"epoch": 5.523937060595916,
|
| 2356 |
+
"grad_norm": 0.12964119017124176,
|
| 2357 |
+
"learning_rate": 3.9811404977123095e-06,
|
| 2358 |
+
"loss": 0.0393,
|
| 2359 |
+
"step": 16500
|
| 2360 |
+
},
|
| 2361 |
+
{
|
| 2362 |
+
"epoch": 5.540676263809843,
|
| 2363 |
+
"grad_norm": 0.201249361038208,
|
| 2364 |
+
"learning_rate": 3.841647137596251e-06,
|
| 2365 |
+
"loss": 0.0368,
|
| 2366 |
+
"step": 16550
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 5.55741546702377,
|
| 2370 |
+
"grad_norm": 0.14805611968040466,
|
| 2371 |
+
"learning_rate": 3.7021537774801917e-06,
|
| 2372 |
+
"loss": 0.0338,
|
| 2373 |
+
"step": 16600
|
| 2374 |
+
},
|
| 2375 |
+
{
|
| 2376 |
+
"epoch": 5.5741546702376965,
|
| 2377 |
+
"grad_norm": 0.11392233520746231,
|
| 2378 |
+
"learning_rate": 3.5626604173641336e-06,
|
| 2379 |
+
"loss": 0.0351,
|
| 2380 |
+
"step": 16650
|
| 2381 |
+
},
|
| 2382 |
+
{
|
| 2383 |
+
"epoch": 5.590893873451623,
|
| 2384 |
+
"grad_norm": 0.13584397733211517,
|
| 2385 |
+
"learning_rate": 3.423167057248075e-06,
|
| 2386 |
+
"loss": 0.0373,
|
| 2387 |
+
"step": 16700
|
| 2388 |
+
},
|
| 2389 |
+
{
|
| 2390 |
+
"epoch": 5.60763307666555,
|
| 2391 |
+
"grad_norm": 0.17985470592975616,
|
| 2392 |
+
"learning_rate": 3.283673697132017e-06,
|
| 2393 |
+
"loss": 0.0355,
|
| 2394 |
+
"step": 16750
|
| 2395 |
+
},
|
| 2396 |
+
{
|
| 2397 |
+
"epoch": 5.624372279879478,
|
| 2398 |
+
"grad_norm": 0.18598899245262146,
|
| 2399 |
+
"learning_rate": 3.144180337015958e-06,
|
| 2400 |
+
"loss": 0.0409,
|
| 2401 |
+
"step": 16800
|
| 2402 |
+
},
|
| 2403 |
+
{
|
| 2404 |
+
"epoch": 5.641111483093405,
|
| 2405 |
+
"grad_norm": 0.17507706582546234,
|
| 2406 |
+
"learning_rate": 3.0046869768998996e-06,
|
| 2407 |
+
"loss": 0.0354,
|
| 2408 |
+
"step": 16850
|
| 2409 |
+
},
|
| 2410 |
+
{
|
| 2411 |
+
"epoch": 5.657850686307332,
|
| 2412 |
+
"grad_norm": 0.15566672384738922,
|
| 2413 |
+
"learning_rate": 2.865193616783841e-06,
|
| 2414 |
+
"loss": 0.0347,
|
| 2415 |
+
"step": 16900
|
| 2416 |
+
},
|
| 2417 |
+
{
|
| 2418 |
+
"epoch": 5.6745898895212585,
|
| 2419 |
+
"grad_norm": 0.15541379153728485,
|
| 2420 |
+
"learning_rate": 2.7257002566677827e-06,
|
| 2421 |
+
"loss": 0.0398,
|
| 2422 |
+
"step": 16950
|
| 2423 |
+
},
|
| 2424 |
+
{
|
| 2425 |
+
"epoch": 5.691329092735186,
|
| 2426 |
+
"grad_norm": 0.20108754932880402,
|
| 2427 |
+
"learning_rate": 2.586206896551724e-06,
|
| 2428 |
+
"loss": 0.0358,
|
| 2429 |
+
"step": 17000
|
| 2430 |
+
},
|
| 2431 |
+
{
|
| 2432 |
+
"epoch": 5.708068295949113,
|
| 2433 |
+
"grad_norm": 0.1303117424249649,
|
| 2434 |
+
"learning_rate": 2.4467135364356657e-06,
|
| 2435 |
+
"loss": 0.043,
|
| 2436 |
+
"step": 17050
|
| 2437 |
+
},
|
| 2438 |
+
{
|
| 2439 |
+
"epoch": 5.72480749916304,
|
| 2440 |
+
"grad_norm": 0.14489831030368805,
|
| 2441 |
+
"learning_rate": 2.3072201763196076e-06,
|
| 2442 |
+
"loss": 0.0377,
|
| 2443 |
+
"step": 17100
|
| 2444 |
+
},
|
| 2445 |
+
{
|
| 2446 |
+
"epoch": 5.741546702376967,
|
| 2447 |
+
"grad_norm": 0.10605516284704208,
|
| 2448 |
+
"learning_rate": 2.167726816203549e-06,
|
| 2449 |
+
"loss": 0.0348,
|
| 2450 |
+
"step": 17150
|
| 2451 |
+
},
|
| 2452 |
+
{
|
| 2453 |
+
"epoch": 5.758285905590894,
|
| 2454 |
+
"grad_norm": 0.29746726155281067,
|
| 2455 |
+
"learning_rate": 2.02823345608749e-06,
|
| 2456 |
+
"loss": 0.0359,
|
| 2457 |
+
"step": 17200
|
| 2458 |
+
},
|
| 2459 |
+
{
|
| 2460 |
+
"epoch": 5.775025108804821,
|
| 2461 |
+
"grad_norm": 0.14607931673526764,
|
| 2462 |
+
"learning_rate": 1.888740095971432e-06,
|
| 2463 |
+
"loss": 0.0361,
|
| 2464 |
+
"step": 17250
|
| 2465 |
+
},
|
| 2466 |
+
{
|
| 2467 |
+
"epoch": 5.791764312018747,
|
| 2468 |
+
"grad_norm": 0.12281953543424606,
|
| 2469 |
+
"learning_rate": 1.7492467358553734e-06,
|
| 2470 |
+
"loss": 0.0377,
|
| 2471 |
+
"step": 17300
|
| 2472 |
+
},
|
| 2473 |
+
{
|
| 2474 |
+
"epoch": 5.808503515232675,
|
| 2475 |
+
"grad_norm": 0.19870831072330475,
|
| 2476 |
+
"learning_rate": 1.609753375739315e-06,
|
| 2477 |
+
"loss": 0.0383,
|
| 2478 |
+
"step": 17350
|
| 2479 |
+
},
|
| 2480 |
+
{
|
| 2481 |
+
"epoch": 5.825242718446602,
|
| 2482 |
+
"grad_norm": 0.13006938993930817,
|
| 2483 |
+
"learning_rate": 1.4702600156232564e-06,
|
| 2484 |
+
"loss": 0.0381,
|
| 2485 |
+
"step": 17400
|
| 2486 |
+
},
|
| 2487 |
+
{
|
| 2488 |
+
"epoch": 5.841981921660529,
|
| 2489 |
+
"grad_norm": 0.12237653881311417,
|
| 2490 |
+
"learning_rate": 1.330766655507198e-06,
|
| 2491 |
+
"loss": 0.0375,
|
| 2492 |
+
"step": 17450
|
| 2493 |
+
},
|
| 2494 |
+
{
|
| 2495 |
+
"epoch": 5.858721124874456,
|
| 2496 |
+
"grad_norm": 0.17220979928970337,
|
| 2497 |
+
"learning_rate": 1.1912732953911394e-06,
|
| 2498 |
+
"loss": 0.0365,
|
| 2499 |
+
"step": 17500
|
| 2500 |
+
},
|
| 2501 |
+
{
|
| 2502 |
+
"epoch": 5.8754603280883835,
|
| 2503 |
+
"grad_norm": 0.18653325736522675,
|
| 2504 |
+
"learning_rate": 1.051779935275081e-06,
|
| 2505 |
+
"loss": 0.0383,
|
| 2506 |
+
"step": 17550
|
| 2507 |
+
},
|
| 2508 |
+
{
|
| 2509 |
+
"epoch": 5.89219953130231,
|
| 2510 |
+
"grad_norm": 0.11963806301355362,
|
| 2511 |
+
"learning_rate": 9.122865751590225e-07,
|
| 2512 |
+
"loss": 0.0399,
|
| 2513 |
+
"step": 17600
|
| 2514 |
+
},
|
| 2515 |
+
{
|
| 2516 |
+
"epoch": 5.908938734516237,
|
| 2517 |
+
"grad_norm": 0.26904088258743286,
|
| 2518 |
+
"learning_rate": 7.72793215042964e-07,
|
| 2519 |
+
"loss": 0.0405,
|
| 2520 |
+
"step": 17650
|
| 2521 |
+
},
|
| 2522 |
+
{
|
| 2523 |
+
"epoch": 5.925677937730164,
|
| 2524 |
+
"grad_norm": 0.149702250957489,
|
| 2525 |
+
"learning_rate": 6.332998549269055e-07,
|
| 2526 |
+
"loss": 0.0358,
|
| 2527 |
+
"step": 17700
|
| 2528 |
+
},
|
| 2529 |
+
{
|
| 2530 |
+
"epoch": 5.942417140944091,
|
| 2531 |
+
"grad_norm": 0.15700772404670715,
|
| 2532 |
+
"learning_rate": 4.93806494810847e-07,
|
| 2533 |
+
"loss": 0.0374,
|
| 2534 |
+
"step": 17750
|
| 2535 |
+
},
|
| 2536 |
+
{
|
| 2537 |
+
"epoch": 5.959156344158018,
|
| 2538 |
+
"grad_norm": 0.20573197305202484,
|
| 2539 |
+
"learning_rate": 3.5431313469478856e-07,
|
| 2540 |
+
"loss": 0.0362,
|
| 2541 |
+
"step": 17800
|
| 2542 |
+
},
|
| 2543 |
+
{
|
| 2544 |
+
"epoch": 5.975895547371945,
|
| 2545 |
+
"grad_norm": 0.11808827519416809,
|
| 2546 |
+
"learning_rate": 2.1481977457873006e-07,
|
| 2547 |
+
"loss": 0.0371,
|
| 2548 |
+
"step": 17850
|
| 2549 |
+
},
|
| 2550 |
+
{
|
| 2551 |
+
"epoch": 5.992634750585872,
|
| 2552 |
+
"grad_norm": 0.1634693145751953,
|
| 2553 |
+
"learning_rate": 7.532641446267158e-08,
|
| 2554 |
+
"loss": 0.0365,
|
| 2555 |
+
"step": 17900
|
| 2556 |
+
},
|
| 2557 |
+
{
|
| 2558 |
+
"epoch": 6.0,
|
| 2559 |
+
"eval_loss": 0.045013878494501114,
|
| 2560 |
+
"eval_runtime": 49.7786,
|
| 2561 |
+
"eval_samples_per_second": 240.023,
|
| 2562 |
+
"eval_steps_per_second": 15.006,
|
| 2563 |
+
"step": 17922
|
| 2564 |
+
}
|
| 2565 |
+
],
|
| 2566 |
+
"logging_steps": 50,
|
| 2567 |
+
"max_steps": 17922,
|
| 2568 |
+
"num_input_tokens_seen": 0,
|
| 2569 |
+
"num_train_epochs": 6,
|
| 2570 |
+
"save_steps": 500,
|
| 2571 |
+
"stateful_callbacks": {
|
| 2572 |
+
"TrainerControl": {
|
| 2573 |
+
"args": {
|
| 2574 |
+
"should_epoch_stop": false,
|
| 2575 |
+
"should_evaluate": false,
|
| 2576 |
+
"should_log": false,
|
| 2577 |
+
"should_save": true,
|
| 2578 |
+
"should_training_stop": true
|
| 2579 |
+
},
|
| 2580 |
+
"attributes": {}
|
| 2581 |
+
}
|
| 2582 |
+
},
|
| 2583 |
+
"total_flos": 3.888169575933542e+16,
|
| 2584 |
+
"train_batch_size": 16,
|
| 2585 |
+
"trial_name": null,
|
| 2586 |
+
"trial_params": null
|
| 2587 |
+
}
|
checkpoint-17922/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:37dd546b69fb60d8deb15a8b88e40b23e367c0e9f5a053ea3ae7c730b3874f2e
|
| 3 |
+
size 5304
|
checkpoint-17922/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-2987/config.json
ADDED
|
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ckartal/english-to-turkish-finetuned-model",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"activation_function": "swish",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"MarianMTModel"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"bos_token_id": 0,
|
| 10 |
+
"classifier_dropout": 0.0,
|
| 11 |
+
"d_model": 512,
|
| 12 |
+
"decoder_attention_heads": 8,
|
| 13 |
+
"decoder_ffn_dim": 2048,
|
| 14 |
+
"decoder_layerdrop": 0.0,
|
| 15 |
+
"decoder_layers": 6,
|
| 16 |
+
"decoder_start_token_id": 59993,
|
| 17 |
+
"decoder_vocab_size": 59994,
|
| 18 |
+
"dropout": 0.1,
|
| 19 |
+
"encoder_attention_heads": 8,
|
| 20 |
+
"encoder_ffn_dim": 2048,
|
| 21 |
+
"encoder_layerdrop": 0.0,
|
| 22 |
+
"encoder_layers": 6,
|
| 23 |
+
"eos_token_id": 0,
|
| 24 |
+
"forced_eos_token_id": 0,
|
| 25 |
+
"init_std": 0.02,
|
| 26 |
+
"is_encoder_decoder": true,
|
| 27 |
+
"max_length": null,
|
| 28 |
+
"max_position_embeddings": 512,
|
| 29 |
+
"model_type": "marian",
|
| 30 |
+
"normalize_embedding": false,
|
| 31 |
+
"num_beams": null,
|
| 32 |
+
"num_hidden_layers": 6,
|
| 33 |
+
"pad_token_id": 59993,
|
| 34 |
+
"scale_embedding": true,
|
| 35 |
+
"share_encoder_decoder_embeddings": true,
|
| 36 |
+
"static_position_embeddings": true,
|
| 37 |
+
"torch_dtype": "float32",
|
| 38 |
+
"transformers_version": "4.49.0",
|
| 39 |
+
"use_cache": true,
|
| 40 |
+
"vocab_size": 59994
|
| 41 |
+
}
|
checkpoint-2987/generation_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bad_words_ids": [
|
| 3 |
+
[
|
| 4 |
+
59993
|
| 5 |
+
]
|
| 6 |
+
],
|
| 7 |
+
"bos_token_id": 0,
|
| 8 |
+
"decoder_start_token_id": 59993,
|
| 9 |
+
"eos_token_id": 0,
|
| 10 |
+
"forced_eos_token_id": 0,
|
| 11 |
+
"max_length": 512,
|
| 12 |
+
"num_beams": 6,
|
| 13 |
+
"pad_token_id": 59993,
|
| 14 |
+
"renormalize_logits": true,
|
| 15 |
+
"transformers_version": "4.49.0"
|
| 16 |
+
}
|
checkpoint-2987/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:772b21519870015edcff8e2db9a6a4448060eeeb8178debe04f9b0dd8da0ceff
|
| 3 |
+
size 299690728
|
checkpoint-2987/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:06b8b97b6f892dac0c6a91fa2167f3739d8cd231960f03b5887658f9fbe99aa4
|
| 3 |
+
size 599054970
|
checkpoint-2987/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f9ec85fbd9238c9947c10323eb559f724ad5ae320e21f57a8df756ef2b058ef4
|
| 3 |
+
size 14244
|
checkpoint-2987/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:73f80cf9a0e45ee7096ade8909076162b7d66cc8713162644286a9a6524a58c6
|
| 3 |
+
size 988
|
checkpoint-2987/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:48275315da074348c5e3aa2b9e9ce7e52122fa39adbd74e2b655d4d79629ace5
|
| 3 |
+
size 1064
|