Commit
•
6266a47
1
Parent(s):
35cc98c
Training in progress, step 1000
Browse files- added_tokens.json +23 -0
- config.json +92 -0
- model.safetensors +3 -0
- preprocessor_config.json +19 -0
- runs/Apr06_03-25-03_DESKTOP-H822ES0/events.out.tfevents.1712384713.DESKTOP-H822ES0 +3 -0
- runs/Apr06_04-21-31_DESKTOP-H822ES0/events.out.tfevents.1712388101.DESKTOP-H822ES0 +3 -0
- runs/Apr06_19-15-16_DESKTOP-H822ES0/events.out.tfevents.1712441723.DESKTOP-H822ES0 +3 -0
- runs/Apr06_19-41-18_DESKTOP-H822ES0/events.out.tfevents.1712443287.DESKTOP-H822ES0 +3 -0
- runs/Apr07_04-01-26_DESKTOP-H822ES0/events.out.tfevents.1712473299.DESKTOP-H822ES0 +3 -0
- runs/Apr07_04-11-24_DESKTOP-H822ES0/events.out.tfevents.1712473888.DESKTOP-H822ES0 +3 -0
- special_tokens_map.json +13 -0
- spm_char.model +3 -0
- tmp-checkpoint-7000/added_tokens.json +23 -0
- tmp-checkpoint-7000/config.json +92 -0
- tmp-checkpoint-7000/generation_config.json +9 -0
- tmp-checkpoint-7000/model.safetensors +3 -0
- tmp-checkpoint-7000/optimizer.pt +3 -0
- tmp-checkpoint-7000/preprocessor_config.json +19 -0
- tmp-checkpoint-7000/rng_state.pth +3 -0
- tmp-checkpoint-7000/scheduler.pt +3 -0
- tmp-checkpoint-7000/special_tokens_map.json +13 -0
- tmp-checkpoint-7000/spm_char.model +3 -0
- tmp-checkpoint-7000/tokenizer_config.json +215 -0
- tmp-checkpoint-7000/trainer_state.json +2037 -0
- tmp-checkpoint-7000/training_args.bin +3 -0
- tokenizer_config.json +215 -0
- training_args.bin +3 -0
added_tokens.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<ctc_blank>": 80,
|
| 3 |
+
"<mask>": 79,
|
| 4 |
+
"À": 89,
|
| 5 |
+
"Á": 90,
|
| 6 |
+
"Ã": 91,
|
| 7 |
+
"Ç": 99,
|
| 8 |
+
"É": 92,
|
| 9 |
+
"Ê": 93,
|
| 10 |
+
"Í": 94,
|
| 11 |
+
"Ó": 95,
|
| 12 |
+
"Ô": 96,
|
| 13 |
+
"Ú": 97,
|
| 14 |
+
"à": 81,
|
| 15 |
+
"á": 82,
|
| 16 |
+
"â": 84,
|
| 17 |
+
"ã": 83,
|
| 18 |
+
"ç": 98,
|
| 19 |
+
"í": 85,
|
| 20 |
+
"ó": 86,
|
| 21 |
+
"ô": 87,
|
| 22 |
+
"ú": 88
|
| 23 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,92 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "microsoft/speecht5_tts",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"apply_spec_augment": true,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"SpeechT5ForTextToSpeech"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 0,
|
| 10 |
+
"conv_bias": false,
|
| 11 |
+
"conv_dim": [
|
| 12 |
+
512,
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512
|
| 19 |
+
],
|
| 20 |
+
"conv_kernel": [
|
| 21 |
+
10,
|
| 22 |
+
3,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
2,
|
| 27 |
+
2
|
| 28 |
+
],
|
| 29 |
+
"conv_stride": [
|
| 30 |
+
5,
|
| 31 |
+
2,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2
|
| 37 |
+
],
|
| 38 |
+
"decoder_attention_heads": 12,
|
| 39 |
+
"decoder_ffn_dim": 3072,
|
| 40 |
+
"decoder_layerdrop": 0.1,
|
| 41 |
+
"decoder_layers": 6,
|
| 42 |
+
"decoder_start_token_id": 2,
|
| 43 |
+
"encoder_attention_heads": 12,
|
| 44 |
+
"encoder_ffn_dim": 3072,
|
| 45 |
+
"encoder_layerdrop": 0.1,
|
| 46 |
+
"encoder_layers": 12,
|
| 47 |
+
"encoder_max_relative_position": 160,
|
| 48 |
+
"eos_token_id": 2,
|
| 49 |
+
"feat_extract_activation": "gelu",
|
| 50 |
+
"feat_extract_norm": "group",
|
| 51 |
+
"feat_proj_dropout": 0.0,
|
| 52 |
+
"guided_attention_loss_num_heads": 2,
|
| 53 |
+
"guided_attention_loss_scale": 10.0,
|
| 54 |
+
"guided_attention_loss_sigma": 0.4,
|
| 55 |
+
"hidden_act": "gelu",
|
| 56 |
+
"hidden_dropout": 0.1,
|
| 57 |
+
"hidden_size": 768,
|
| 58 |
+
"initializer_range": 0.02,
|
| 59 |
+
"is_encoder_decoder": true,
|
| 60 |
+
"layer_norm_eps": 1e-05,
|
| 61 |
+
"mask_feature_length": 10,
|
| 62 |
+
"mask_feature_min_masks": 0,
|
| 63 |
+
"mask_feature_prob": 0.0,
|
| 64 |
+
"mask_time_length": 10,
|
| 65 |
+
"mask_time_min_masks": 2,
|
| 66 |
+
"mask_time_prob": 0.05,
|
| 67 |
+
"max_length": 1876,
|
| 68 |
+
"max_speech_positions": 1876,
|
| 69 |
+
"max_text_positions": 600,
|
| 70 |
+
"model_type": "speecht5",
|
| 71 |
+
"num_conv_pos_embedding_groups": 16,
|
| 72 |
+
"num_conv_pos_embeddings": 128,
|
| 73 |
+
"num_feat_extract_layers": 7,
|
| 74 |
+
"num_mel_bins": 80,
|
| 75 |
+
"pad_token_id": 1,
|
| 76 |
+
"positional_dropout": 0.1,
|
| 77 |
+
"reduction_factor": 2,
|
| 78 |
+
"scale_embedding": false,
|
| 79 |
+
"speaker_embedding_dim": 512,
|
| 80 |
+
"speech_decoder_postnet_dropout": 0.5,
|
| 81 |
+
"speech_decoder_postnet_kernel": 5,
|
| 82 |
+
"speech_decoder_postnet_layers": 5,
|
| 83 |
+
"speech_decoder_postnet_units": 256,
|
| 84 |
+
"speech_decoder_prenet_dropout": 0.5,
|
| 85 |
+
"speech_decoder_prenet_layers": 2,
|
| 86 |
+
"speech_decoder_prenet_units": 256,
|
| 87 |
+
"torch_dtype": "float32",
|
| 88 |
+
"transformers_version": "4.39.0.dev0",
|
| 89 |
+
"use_cache": false,
|
| 90 |
+
"use_guided_attention_loss": true,
|
| 91 |
+
"vocab_size": 104
|
| 92 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:26d627f2fdeec2b0508b308c06324593e939f3488de37c19909afa74c30b22c1
|
| 3 |
+
size 577859976
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_normalize": false,
|
| 3 |
+
"feature_extractor_type": "SpeechT5FeatureExtractor",
|
| 4 |
+
"feature_size": 1,
|
| 5 |
+
"fmax": 7600,
|
| 6 |
+
"fmin": 80,
|
| 7 |
+
"frame_signal_scale": 1.0,
|
| 8 |
+
"hop_length": 16,
|
| 9 |
+
"mel_floor": 1e-10,
|
| 10 |
+
"num_mel_bins": 80,
|
| 11 |
+
"padding_side": "right",
|
| 12 |
+
"padding_value": 0.0,
|
| 13 |
+
"processor_class": "SpeechT5Processor",
|
| 14 |
+
"reduction_factor": 2,
|
| 15 |
+
"return_attention_mask": true,
|
| 16 |
+
"sampling_rate": 16000,
|
| 17 |
+
"win_function": "hann_window",
|
| 18 |
+
"win_length": 64
|
| 19 |
+
}
|
runs/Apr06_03-25-03_DESKTOP-H822ES0/events.out.tfevents.1712384713.DESKTOP-H822ES0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:57857a60c06f3ae6236e49141b5ff80583c65479c7bc0c50cbd4a18a77e484fc
|
| 3 |
+
size 6218
|
runs/Apr06_04-21-31_DESKTOP-H822ES0/events.out.tfevents.1712388101.DESKTOP-H822ES0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3e74ca32533a0f26eb530deb35ae23dc8a12b8cf202887857ea456558d24352f
|
| 3 |
+
size 6218
|
runs/Apr06_19-15-16_DESKTOP-H822ES0/events.out.tfevents.1712441723.DESKTOP-H822ES0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f139434745b8eb46425fbe8c5b2779c75c7fbf4c041c6b842db54762890c8b00
|
| 3 |
+
size 6218
|
runs/Apr06_19-41-18_DESKTOP-H822ES0/events.out.tfevents.1712443287.DESKTOP-H822ES0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4006a286b47218ea91a3fcf17b96366c33831378dd316fb4308aabaca9b1b7f
|
| 3 |
+
size 134312
|
runs/Apr07_04-01-26_DESKTOP-H822ES0/events.out.tfevents.1712473299.DESKTOP-H822ES0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:37cbdb2cb534644b409aeb412681d502c33f8852d826ca1788e1d0eda9a6686a
|
| 3 |
+
size 12398
|
runs/Apr07_04-11-24_DESKTOP-H822ES0/events.out.tfevents.1712473888.DESKTOP-H822ES0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ab5536e76de691bd015ca55fe158e4dfd6a21ce266d57b2670c37c1c9bb86b24
|
| 3 |
+
size 14639
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"eos_token": "</s>",
|
| 4 |
+
"mask_token": {
|
| 5 |
+
"content": "<mask>",
|
| 6 |
+
"lstrip": true,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false
|
| 10 |
+
},
|
| 11 |
+
"pad_token": "<pad>",
|
| 12 |
+
"unk_token": "<unk>"
|
| 13 |
+
}
|
spm_char.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7fcc48f3e225f627b1641db410ceb0c8649bd2b0c982e150b03f8be3728ab560
|
| 3 |
+
size 238473
|
tmp-checkpoint-7000/added_tokens.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<ctc_blank>": 80,
|
| 3 |
+
"<mask>": 79,
|
| 4 |
+
"À": 89,
|
| 5 |
+
"Á": 90,
|
| 6 |
+
"Ã": 91,
|
| 7 |
+
"Ç": 99,
|
| 8 |
+
"É": 92,
|
| 9 |
+
"Ê": 93,
|
| 10 |
+
"Í": 94,
|
| 11 |
+
"Ó": 95,
|
| 12 |
+
"Ô": 96,
|
| 13 |
+
"Ú": 97,
|
| 14 |
+
"à": 81,
|
| 15 |
+
"á": 82,
|
| 16 |
+
"â": 84,
|
| 17 |
+
"ã": 83,
|
| 18 |
+
"ç": 98,
|
| 19 |
+
"í": 85,
|
| 20 |
+
"ó": 86,
|
| 21 |
+
"ô": 87,
|
| 22 |
+
"ú": 88
|
| 23 |
+
}
|
tmp-checkpoint-7000/config.json
ADDED
|
@@ -0,0 +1,92 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "microsoft/speecht5_tts",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"apply_spec_augment": true,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"SpeechT5ForTextToSpeech"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 0,
|
| 10 |
+
"conv_bias": false,
|
| 11 |
+
"conv_dim": [
|
| 12 |
+
512,
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512
|
| 19 |
+
],
|
| 20 |
+
"conv_kernel": [
|
| 21 |
+
10,
|
| 22 |
+
3,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
2,
|
| 27 |
+
2
|
| 28 |
+
],
|
| 29 |
+
"conv_stride": [
|
| 30 |
+
5,
|
| 31 |
+
2,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2
|
| 37 |
+
],
|
| 38 |
+
"decoder_attention_heads": 12,
|
| 39 |
+
"decoder_ffn_dim": 3072,
|
| 40 |
+
"decoder_layerdrop": 0.1,
|
| 41 |
+
"decoder_layers": 6,
|
| 42 |
+
"decoder_start_token_id": 2,
|
| 43 |
+
"encoder_attention_heads": 12,
|
| 44 |
+
"encoder_ffn_dim": 3072,
|
| 45 |
+
"encoder_layerdrop": 0.1,
|
| 46 |
+
"encoder_layers": 12,
|
| 47 |
+
"encoder_max_relative_position": 160,
|
| 48 |
+
"eos_token_id": 2,
|
| 49 |
+
"feat_extract_activation": "gelu",
|
| 50 |
+
"feat_extract_norm": "group",
|
| 51 |
+
"feat_proj_dropout": 0.0,
|
| 52 |
+
"guided_attention_loss_num_heads": 2,
|
| 53 |
+
"guided_attention_loss_scale": 10.0,
|
| 54 |
+
"guided_attention_loss_sigma": 0.4,
|
| 55 |
+
"hidden_act": "gelu",
|
| 56 |
+
"hidden_dropout": 0.1,
|
| 57 |
+
"hidden_size": 768,
|
| 58 |
+
"initializer_range": 0.02,
|
| 59 |
+
"is_encoder_decoder": true,
|
| 60 |
+
"layer_norm_eps": 1e-05,
|
| 61 |
+
"mask_feature_length": 10,
|
| 62 |
+
"mask_feature_min_masks": 0,
|
| 63 |
+
"mask_feature_prob": 0.0,
|
| 64 |
+
"mask_time_length": 10,
|
| 65 |
+
"mask_time_min_masks": 2,
|
| 66 |
+
"mask_time_prob": 0.05,
|
| 67 |
+
"max_length": 1876,
|
| 68 |
+
"max_speech_positions": 1876,
|
| 69 |
+
"max_text_positions": 600,
|
| 70 |
+
"model_type": "speecht5",
|
| 71 |
+
"num_conv_pos_embedding_groups": 16,
|
| 72 |
+
"num_conv_pos_embeddings": 128,
|
| 73 |
+
"num_feat_extract_layers": 7,
|
| 74 |
+
"num_mel_bins": 80,
|
| 75 |
+
"pad_token_id": 1,
|
| 76 |
+
"positional_dropout": 0.1,
|
| 77 |
+
"reduction_factor": 2,
|
| 78 |
+
"scale_embedding": false,
|
| 79 |
+
"speaker_embedding_dim": 512,
|
| 80 |
+
"speech_decoder_postnet_dropout": 0.5,
|
| 81 |
+
"speech_decoder_postnet_kernel": 5,
|
| 82 |
+
"speech_decoder_postnet_layers": 5,
|
| 83 |
+
"speech_decoder_postnet_units": 256,
|
| 84 |
+
"speech_decoder_prenet_dropout": 0.5,
|
| 85 |
+
"speech_decoder_prenet_layers": 2,
|
| 86 |
+
"speech_decoder_prenet_units": 256,
|
| 87 |
+
"torch_dtype": "float32",
|
| 88 |
+
"transformers_version": "4.39.0.dev0",
|
| 89 |
+
"use_cache": false,
|
| 90 |
+
"use_guided_attention_loss": true,
|
| 91 |
+
"vocab_size": 104
|
| 92 |
+
}
|
tmp-checkpoint-7000/generation_config.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"decoder_start_token_id": 2,
|
| 5 |
+
"eos_token_id": 2,
|
| 6 |
+
"max_length": 1876,
|
| 7 |
+
"pad_token_id": 1,
|
| 8 |
+
"transformers_version": "4.39.0.dev0"
|
| 9 |
+
}
|
tmp-checkpoint-7000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7a6470c523168c4aaa944dbbb67844e7ebf65adb63aac0196d4d6ba2faafe66d
|
| 3 |
+
size 577859976
|
tmp-checkpoint-7000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eeca09e53602e812c50987efd688a2469ccca75f0d40ae053a66f271da38b24e
|
| 3 |
+
size 1155913545
|
tmp-checkpoint-7000/preprocessor_config.json
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_normalize": false,
|
| 3 |
+
"feature_extractor_type": "SpeechT5FeatureExtractor",
|
| 4 |
+
"feature_size": 1,
|
| 5 |
+
"fmax": 7600,
|
| 6 |
+
"fmin": 80,
|
| 7 |
+
"frame_signal_scale": 1.0,
|
| 8 |
+
"hop_length": 16,
|
| 9 |
+
"mel_floor": 1e-10,
|
| 10 |
+
"num_mel_bins": 80,
|
| 11 |
+
"padding_side": "right",
|
| 12 |
+
"padding_value": 0.0,
|
| 13 |
+
"processor_class": "SpeechT5Processor",
|
| 14 |
+
"reduction_factor": 2,
|
| 15 |
+
"return_attention_mask": true,
|
| 16 |
+
"sampling_rate": 16000,
|
| 17 |
+
"win_function": "hann_window",
|
| 18 |
+
"win_length": 64
|
| 19 |
+
}
|
tmp-checkpoint-7000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fedd897c5a9892ab446397dec3be6ef5dd31e4752f118311bff5e7f3d61072f1
|
| 3 |
+
size 14244
|
tmp-checkpoint-7000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:042f0e893988566988d5c33ca650bf0ec42341b2d800d4e8f29c5b53fd60c115
|
| 3 |
+
size 1064
|
tmp-checkpoint-7000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"eos_token": "</s>",
|
| 4 |
+
"mask_token": {
|
| 5 |
+
"content": "<mask>",
|
| 6 |
+
"lstrip": true,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false
|
| 10 |
+
},
|
| 11 |
+
"pad_token": "<pad>",
|
| 12 |
+
"unk_token": "<unk>"
|
| 13 |
+
}
|
tmp-checkpoint-7000/spm_char.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7fcc48f3e225f627b1641db410ceb0c8649bd2b0c982e150b03f8be3728ab560
|
| 3 |
+
size 238473
|
tmp-checkpoint-7000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,215 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"79": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": true,
|
| 38 |
+
"normalized": true,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
},
|
| 43 |
+
"80": {
|
| 44 |
+
"content": "<ctc_blank>",
|
| 45 |
+
"lstrip": false,
|
| 46 |
+
"normalized": true,
|
| 47 |
+
"rstrip": false,
|
| 48 |
+
"single_word": false,
|
| 49 |
+
"special": false
|
| 50 |
+
},
|
| 51 |
+
"81": {
|
| 52 |
+
"content": "à",
|
| 53 |
+
"lstrip": false,
|
| 54 |
+
"normalized": true,
|
| 55 |
+
"rstrip": false,
|
| 56 |
+
"single_word": false,
|
| 57 |
+
"special": false
|
| 58 |
+
},
|
| 59 |
+
"82": {
|
| 60 |
+
"content": "á",
|
| 61 |
+
"lstrip": false,
|
| 62 |
+
"normalized": true,
|
| 63 |
+
"rstrip": false,
|
| 64 |
+
"single_word": false,
|
| 65 |
+
"special": false
|
| 66 |
+
},
|
| 67 |
+
"83": {
|
| 68 |
+
"content": "ã",
|
| 69 |
+
"lstrip": false,
|
| 70 |
+
"normalized": true,
|
| 71 |
+
"rstrip": false,
|
| 72 |
+
"single_word": false,
|
| 73 |
+
"special": false
|
| 74 |
+
},
|
| 75 |
+
"84": {
|
| 76 |
+
"content": "â",
|
| 77 |
+
"lstrip": false,
|
| 78 |
+
"normalized": true,
|
| 79 |
+
"rstrip": false,
|
| 80 |
+
"single_word": false,
|
| 81 |
+
"special": false
|
| 82 |
+
},
|
| 83 |
+
"85": {
|
| 84 |
+
"content": "í",
|
| 85 |
+
"lstrip": false,
|
| 86 |
+
"normalized": true,
|
| 87 |
+
"rstrip": false,
|
| 88 |
+
"single_word": false,
|
| 89 |
+
"special": false
|
| 90 |
+
},
|
| 91 |
+
"86": {
|
| 92 |
+
"content": "ó",
|
| 93 |
+
"lstrip": false,
|
| 94 |
+
"normalized": true,
|
| 95 |
+
"rstrip": false,
|
| 96 |
+
"single_word": false,
|
| 97 |
+
"special": false
|
| 98 |
+
},
|
| 99 |
+
"87": {
|
| 100 |
+
"content": "ô",
|
| 101 |
+
"lstrip": false,
|
| 102 |
+
"normalized": true,
|
| 103 |
+
"rstrip": false,
|
| 104 |
+
"single_word": false,
|
| 105 |
+
"special": false
|
| 106 |
+
},
|
| 107 |
+
"88": {
|
| 108 |
+
"content": "ú",
|
| 109 |
+
"lstrip": false,
|
| 110 |
+
"normalized": true,
|
| 111 |
+
"rstrip": false,
|
| 112 |
+
"single_word": false,
|
| 113 |
+
"special": false
|
| 114 |
+
},
|
| 115 |
+
"89": {
|
| 116 |
+
"content": "À",
|
| 117 |
+
"lstrip": false,
|
| 118 |
+
"normalized": true,
|
| 119 |
+
"rstrip": false,
|
| 120 |
+
"single_word": false,
|
| 121 |
+
"special": false
|
| 122 |
+
},
|
| 123 |
+
"90": {
|
| 124 |
+
"content": "Á",
|
| 125 |
+
"lstrip": false,
|
| 126 |
+
"normalized": true,
|
| 127 |
+
"rstrip": false,
|
| 128 |
+
"single_word": false,
|
| 129 |
+
"special": false
|
| 130 |
+
},
|
| 131 |
+
"91": {
|
| 132 |
+
"content": "Ã",
|
| 133 |
+
"lstrip": false,
|
| 134 |
+
"normalized": true,
|
| 135 |
+
"rstrip": false,
|
| 136 |
+
"single_word": false,
|
| 137 |
+
"special": false
|
| 138 |
+
},
|
| 139 |
+
"92": {
|
| 140 |
+
"content": "É",
|
| 141 |
+
"lstrip": false,
|
| 142 |
+
"normalized": true,
|
| 143 |
+
"rstrip": false,
|
| 144 |
+
"single_word": false,
|
| 145 |
+
"special": false
|
| 146 |
+
},
|
| 147 |
+
"93": {
|
| 148 |
+
"content": "Ê",
|
| 149 |
+
"lstrip": false,
|
| 150 |
+
"normalized": true,
|
| 151 |
+
"rstrip": false,
|
| 152 |
+
"single_word": false,
|
| 153 |
+
"special": false
|
| 154 |
+
},
|
| 155 |
+
"94": {
|
| 156 |
+
"content": "Í",
|
| 157 |
+
"lstrip": false,
|
| 158 |
+
"normalized": true,
|
| 159 |
+
"rstrip": false,
|
| 160 |
+
"single_word": false,
|
| 161 |
+
"special": false
|
| 162 |
+
},
|
| 163 |
+
"95": {
|
| 164 |
+
"content": "Ó",
|
| 165 |
+
"lstrip": false,
|
| 166 |
+
"normalized": true,
|
| 167 |
+
"rstrip": false,
|
| 168 |
+
"single_word": false,
|
| 169 |
+
"special": false
|
| 170 |
+
},
|
| 171 |
+
"96": {
|
| 172 |
+
"content": "Ô",
|
| 173 |
+
"lstrip": false,
|
| 174 |
+
"normalized": true,
|
| 175 |
+
"rstrip": false,
|
| 176 |
+
"single_word": false,
|
| 177 |
+
"special": false
|
| 178 |
+
},
|
| 179 |
+
"97": {
|
| 180 |
+
"content": "Ú",
|
| 181 |
+
"lstrip": false,
|
| 182 |
+
"normalized": true,
|
| 183 |
+
"rstrip": false,
|
| 184 |
+
"single_word": false,
|
| 185 |
+
"special": false
|
| 186 |
+
},
|
| 187 |
+
"98": {
|
| 188 |
+
"content": "ç",
|
| 189 |
+
"lstrip": false,
|
| 190 |
+
"normalized": true,
|
| 191 |
+
"rstrip": false,
|
| 192 |
+
"single_word": false,
|
| 193 |
+
"special": false
|
| 194 |
+
},
|
| 195 |
+
"99": {
|
| 196 |
+
"content": "Ç",
|
| 197 |
+
"lstrip": false,
|
| 198 |
+
"normalized": true,
|
| 199 |
+
"rstrip": false,
|
| 200 |
+
"single_word": false,
|
| 201 |
+
"special": false
|
| 202 |
+
}
|
| 203 |
+
},
|
| 204 |
+
"bos_token": "<s>",
|
| 205 |
+
"clean_up_tokenization_spaces": true,
|
| 206 |
+
"eos_token": "</s>",
|
| 207 |
+
"mask_token": "<mask>",
|
| 208 |
+
"model_max_length": 600,
|
| 209 |
+
"normalize": false,
|
| 210 |
+
"pad_token": "<pad>",
|
| 211 |
+
"processor_class": "SpeechT5Processor",
|
| 212 |
+
"sp_model_kwargs": {},
|
| 213 |
+
"tokenizer_class": "SpeechT5Tokenizer",
|
| 214 |
+
"unk_token": "<unk>"
|
| 215 |
+
}
|
tmp-checkpoint-7000/trainer_state.json
ADDED
|
@@ -0,0 +1,2037 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.4256468117237091,
|
| 3 |
+
"best_model_checkpoint": "guilhermebastos96/speecht5_allvoices\\checkpoint-5000",
|
| 4 |
+
"epoch": 4.851425106124924,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 7000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.02,
|
| 13 |
+
"grad_norm": 1.8868106603622437,
|
| 14 |
+
"learning_rate": 5.000000000000001e-07,
|
| 15 |
+
"loss": 0.4682,
|
| 16 |
+
"step": 25
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.03,
|
| 20 |
+
"grad_norm": 4.702238082885742,
|
| 21 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 22 |
+
"loss": 0.4709,
|
| 23 |
+
"step": 50
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.05,
|
| 27 |
+
"grad_norm": 2.156214714050293,
|
| 28 |
+
"learning_rate": 1.5e-06,
|
| 29 |
+
"loss": 0.4661,
|
| 30 |
+
"step": 75
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.07,
|
| 34 |
+
"grad_norm": 1.8122159242630005,
|
| 35 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 36 |
+
"loss": 0.4658,
|
| 37 |
+
"step": 100
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.09,
|
| 41 |
+
"grad_norm": 2.852905511856079,
|
| 42 |
+
"learning_rate": 2.5e-06,
|
| 43 |
+
"loss": 0.4688,
|
| 44 |
+
"step": 125
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.1,
|
| 48 |
+
"grad_norm": 1.4833136796951294,
|
| 49 |
+
"learning_rate": 3e-06,
|
| 50 |
+
"loss": 0.4658,
|
| 51 |
+
"step": 150
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12,
|
| 55 |
+
"grad_norm": 2.180389881134033,
|
| 56 |
+
"learning_rate": 3.5e-06,
|
| 57 |
+
"loss": 0.4766,
|
| 58 |
+
"step": 175
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.14,
|
| 62 |
+
"grad_norm": 2.425173044204712,
|
| 63 |
+
"learning_rate": 4.000000000000001e-06,
|
| 64 |
+
"loss": 0.4625,
|
| 65 |
+
"step": 200
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.16,
|
| 69 |
+
"grad_norm": 2.724726438522339,
|
| 70 |
+
"learning_rate": 4.5e-06,
|
| 71 |
+
"loss": 0.4695,
|
| 72 |
+
"step": 225
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.17,
|
| 76 |
+
"grad_norm": 2.528960943222046,
|
| 77 |
+
"learning_rate": 5e-06,
|
| 78 |
+
"loss": 0.4734,
|
| 79 |
+
"step": 250
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.19,
|
| 83 |
+
"grad_norm": 1.560817003250122,
|
| 84 |
+
"learning_rate": 5.500000000000001e-06,
|
| 85 |
+
"loss": 0.469,
|
| 86 |
+
"step": 275
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.21,
|
| 90 |
+
"grad_norm": 2.050774097442627,
|
| 91 |
+
"learning_rate": 6e-06,
|
| 92 |
+
"loss": 0.4735,
|
| 93 |
+
"step": 300
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.23,
|
| 97 |
+
"grad_norm": 2.356326103210449,
|
| 98 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 99 |
+
"loss": 0.4711,
|
| 100 |
+
"step": 325
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.24,
|
| 104 |
+
"grad_norm": 2.561927318572998,
|
| 105 |
+
"learning_rate": 7e-06,
|
| 106 |
+
"loss": 0.4709,
|
| 107 |
+
"step": 350
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.26,
|
| 111 |
+
"grad_norm": 3.2564566135406494,
|
| 112 |
+
"learning_rate": 7.500000000000001e-06,
|
| 113 |
+
"loss": 0.464,
|
| 114 |
+
"step": 375
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.28,
|
| 118 |
+
"grad_norm": 2.8264591693878174,
|
| 119 |
+
"learning_rate": 8.000000000000001e-06,
|
| 120 |
+
"loss": 0.4728,
|
| 121 |
+
"step": 400
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.29,
|
| 125 |
+
"grad_norm": 2.646620035171509,
|
| 126 |
+
"learning_rate": 8.5e-06,
|
| 127 |
+
"loss": 0.4672,
|
| 128 |
+
"step": 425
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.31,
|
| 132 |
+
"grad_norm": 1.601816177368164,
|
| 133 |
+
"learning_rate": 9e-06,
|
| 134 |
+
"loss": 0.4644,
|
| 135 |
+
"step": 450
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.33,
|
| 139 |
+
"grad_norm": 3.687861919403076,
|
| 140 |
+
"learning_rate": 9.5e-06,
|
| 141 |
+
"loss": 0.4687,
|
| 142 |
+
"step": 475
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.35,
|
| 146 |
+
"grad_norm": 2.237541675567627,
|
| 147 |
+
"learning_rate": 1e-05,
|
| 148 |
+
"loss": 0.4721,
|
| 149 |
+
"step": 500
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.36,
|
| 153 |
+
"grad_norm": 3.4831230640411377,
|
| 154 |
+
"learning_rate": 9.973684210526316e-06,
|
| 155 |
+
"loss": 0.4674,
|
| 156 |
+
"step": 525
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.38,
|
| 160 |
+
"grad_norm": 2.717695951461792,
|
| 161 |
+
"learning_rate": 9.947368421052632e-06,
|
| 162 |
+
"loss": 0.4669,
|
| 163 |
+
"step": 550
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.4,
|
| 167 |
+
"grad_norm": 2.5196590423583984,
|
| 168 |
+
"learning_rate": 9.921052631578947e-06,
|
| 169 |
+
"loss": 0.4664,
|
| 170 |
+
"step": 575
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.42,
|
| 174 |
+
"grad_norm": 2.3250813484191895,
|
| 175 |
+
"learning_rate": 9.894736842105264e-06,
|
| 176 |
+
"loss": 0.4741,
|
| 177 |
+
"step": 600
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.43,
|
| 181 |
+
"grad_norm": 1.8072372674942017,
|
| 182 |
+
"learning_rate": 9.868421052631579e-06,
|
| 183 |
+
"loss": 0.4696,
|
| 184 |
+
"step": 625
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.45,
|
| 188 |
+
"grad_norm": 2.5572052001953125,
|
| 189 |
+
"learning_rate": 9.842105263157896e-06,
|
| 190 |
+
"loss": 0.4742,
|
| 191 |
+
"step": 650
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.47,
|
| 195 |
+
"grad_norm": 2.830040454864502,
|
| 196 |
+
"learning_rate": 9.815789473684212e-06,
|
| 197 |
+
"loss": 0.4679,
|
| 198 |
+
"step": 675
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.49,
|
| 202 |
+
"grad_norm": 3.8328497409820557,
|
| 203 |
+
"learning_rate": 9.789473684210527e-06,
|
| 204 |
+
"loss": 0.4736,
|
| 205 |
+
"step": 700
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.5,
|
| 209 |
+
"grad_norm": 3.366887331008911,
|
| 210 |
+
"learning_rate": 9.763157894736844e-06,
|
| 211 |
+
"loss": 0.4767,
|
| 212 |
+
"step": 725
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.52,
|
| 216 |
+
"grad_norm": 4.468636512756348,
|
| 217 |
+
"learning_rate": 9.736842105263159e-06,
|
| 218 |
+
"loss": 0.4699,
|
| 219 |
+
"step": 750
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.54,
|
| 223 |
+
"grad_norm": 2.1232364177703857,
|
| 224 |
+
"learning_rate": 9.710526315789474e-06,
|
| 225 |
+
"loss": 0.4687,
|
| 226 |
+
"step": 775
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.55,
|
| 230 |
+
"grad_norm": 2.8201797008514404,
|
| 231 |
+
"learning_rate": 9.68421052631579e-06,
|
| 232 |
+
"loss": 0.474,
|
| 233 |
+
"step": 800
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.57,
|
| 237 |
+
"grad_norm": 2.4909231662750244,
|
| 238 |
+
"learning_rate": 9.657894736842106e-06,
|
| 239 |
+
"loss": 0.4697,
|
| 240 |
+
"step": 825
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.59,
|
| 244 |
+
"grad_norm": 2.9799981117248535,
|
| 245 |
+
"learning_rate": 9.631578947368422e-06,
|
| 246 |
+
"loss": 0.4766,
|
| 247 |
+
"step": 850
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.61,
|
| 251 |
+
"grad_norm": 3.508791446685791,
|
| 252 |
+
"learning_rate": 9.605263157894737e-06,
|
| 253 |
+
"loss": 0.4632,
|
| 254 |
+
"step": 875
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.62,
|
| 258 |
+
"grad_norm": 2.9218318462371826,
|
| 259 |
+
"learning_rate": 9.578947368421054e-06,
|
| 260 |
+
"loss": 0.4687,
|
| 261 |
+
"step": 900
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.64,
|
| 265 |
+
"grad_norm": 2.421876907348633,
|
| 266 |
+
"learning_rate": 9.552631578947369e-06,
|
| 267 |
+
"loss": 0.4595,
|
| 268 |
+
"step": 925
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.66,
|
| 272 |
+
"grad_norm": 2.8650059700012207,
|
| 273 |
+
"learning_rate": 9.526315789473684e-06,
|
| 274 |
+
"loss": 0.4606,
|
| 275 |
+
"step": 950
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.68,
|
| 279 |
+
"grad_norm": 3.76725697517395,
|
| 280 |
+
"learning_rate": 9.5e-06,
|
| 281 |
+
"loss": 0.4683,
|
| 282 |
+
"step": 975
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.69,
|
| 286 |
+
"grad_norm": 3.538356304168701,
|
| 287 |
+
"learning_rate": 9.473684210526315e-06,
|
| 288 |
+
"loss": 0.4711,
|
| 289 |
+
"step": 1000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.69,
|
| 293 |
+
"eval_loss": 0.4389101266860962,
|
| 294 |
+
"eval_runtime": 131.3338,
|
| 295 |
+
"eval_samples_per_second": 39.068,
|
| 296 |
+
"eval_steps_per_second": 19.538,
|
| 297 |
+
"step": 1000
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.71,
|
| 301 |
+
"grad_norm": 2.725700855255127,
|
| 302 |
+
"learning_rate": 9.447368421052632e-06,
|
| 303 |
+
"loss": 0.4671,
|
| 304 |
+
"step": 1025
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.73,
|
| 308 |
+
"grad_norm": 2.324768304824829,
|
| 309 |
+
"learning_rate": 9.421052631578949e-06,
|
| 310 |
+
"loss": 0.4746,
|
| 311 |
+
"step": 1050
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.75,
|
| 315 |
+
"grad_norm": 2.6872613430023193,
|
| 316 |
+
"learning_rate": 9.394736842105264e-06,
|
| 317 |
+
"loss": 0.4724,
|
| 318 |
+
"step": 1075
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.76,
|
| 322 |
+
"grad_norm": 2.945366382598877,
|
| 323 |
+
"learning_rate": 9.36842105263158e-06,
|
| 324 |
+
"loss": 0.4755,
|
| 325 |
+
"step": 1100
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.78,
|
| 329 |
+
"grad_norm": 2.513762950897217,
|
| 330 |
+
"learning_rate": 9.342105263157895e-06,
|
| 331 |
+
"loss": 0.4661,
|
| 332 |
+
"step": 1125
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.8,
|
| 336 |
+
"grad_norm": 5.170629024505615,
|
| 337 |
+
"learning_rate": 9.315789473684212e-06,
|
| 338 |
+
"loss": 0.473,
|
| 339 |
+
"step": 1150
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.81,
|
| 343 |
+
"grad_norm": 3.1704061031341553,
|
| 344 |
+
"learning_rate": 9.289473684210527e-06,
|
| 345 |
+
"loss": 0.4641,
|
| 346 |
+
"step": 1175
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.83,
|
| 350 |
+
"grad_norm": 1.7679084539413452,
|
| 351 |
+
"learning_rate": 9.263157894736842e-06,
|
| 352 |
+
"loss": 0.4638,
|
| 353 |
+
"step": 1200
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.85,
|
| 357 |
+
"grad_norm": 4.245815753936768,
|
| 358 |
+
"learning_rate": 9.236842105263159e-06,
|
| 359 |
+
"loss": 0.4736,
|
| 360 |
+
"step": 1225
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.87,
|
| 364 |
+
"grad_norm": 2.433870315551758,
|
| 365 |
+
"learning_rate": 9.210526315789474e-06,
|
| 366 |
+
"loss": 0.4634,
|
| 367 |
+
"step": 1250
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.88,
|
| 371 |
+
"grad_norm": 2.9975013732910156,
|
| 372 |
+
"learning_rate": 9.18421052631579e-06,
|
| 373 |
+
"loss": 0.4715,
|
| 374 |
+
"step": 1275
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.9,
|
| 378 |
+
"grad_norm": 3.1186320781707764,
|
| 379 |
+
"learning_rate": 9.157894736842105e-06,
|
| 380 |
+
"loss": 0.4669,
|
| 381 |
+
"step": 1300
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.92,
|
| 385 |
+
"grad_norm": 3.856799840927124,
|
| 386 |
+
"learning_rate": 9.131578947368422e-06,
|
| 387 |
+
"loss": 0.4707,
|
| 388 |
+
"step": 1325
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.94,
|
| 392 |
+
"grad_norm": 3.04708194732666,
|
| 393 |
+
"learning_rate": 9.105263157894739e-06,
|
| 394 |
+
"loss": 0.4685,
|
| 395 |
+
"step": 1350
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.95,
|
| 399 |
+
"grad_norm": 1.939920425415039,
|
| 400 |
+
"learning_rate": 9.078947368421054e-06,
|
| 401 |
+
"loss": 0.4616,
|
| 402 |
+
"step": 1375
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.97,
|
| 406 |
+
"grad_norm": 2.5178582668304443,
|
| 407 |
+
"learning_rate": 9.05263157894737e-06,
|
| 408 |
+
"loss": 0.4701,
|
| 409 |
+
"step": 1400
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.99,
|
| 413 |
+
"grad_norm": 3.2123663425445557,
|
| 414 |
+
"learning_rate": 9.026315789473685e-06,
|
| 415 |
+
"loss": 0.4708,
|
| 416 |
+
"step": 1425
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.0,
|
| 420 |
+
"grad_norm": 2.4378151893615723,
|
| 421 |
+
"learning_rate": 9e-06,
|
| 422 |
+
"loss": 0.4572,
|
| 423 |
+
"step": 1450
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.02,
|
| 427 |
+
"grad_norm": 3.701718330383301,
|
| 428 |
+
"learning_rate": 8.973684210526317e-06,
|
| 429 |
+
"loss": 0.467,
|
| 430 |
+
"step": 1475
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.04,
|
| 434 |
+
"grad_norm": 2.7244162559509277,
|
| 435 |
+
"learning_rate": 8.947368421052632e-06,
|
| 436 |
+
"loss": 0.4642,
|
| 437 |
+
"step": 1500
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.06,
|
| 441 |
+
"grad_norm": 3.887258291244507,
|
| 442 |
+
"learning_rate": 8.921052631578949e-06,
|
| 443 |
+
"loss": 0.4687,
|
| 444 |
+
"step": 1525
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.07,
|
| 448 |
+
"grad_norm": 9.12584400177002,
|
| 449 |
+
"learning_rate": 8.894736842105264e-06,
|
| 450 |
+
"loss": 0.4651,
|
| 451 |
+
"step": 1550
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.09,
|
| 455 |
+
"grad_norm": 2.1384148597717285,
|
| 456 |
+
"learning_rate": 8.86842105263158e-06,
|
| 457 |
+
"loss": 0.4649,
|
| 458 |
+
"step": 1575
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.11,
|
| 462 |
+
"grad_norm": 1.476319432258606,
|
| 463 |
+
"learning_rate": 8.842105263157895e-06,
|
| 464 |
+
"loss": 0.4729,
|
| 465 |
+
"step": 1600
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.13,
|
| 469 |
+
"grad_norm": 2.3843348026275635,
|
| 470 |
+
"learning_rate": 8.81578947368421e-06,
|
| 471 |
+
"loss": 0.465,
|
| 472 |
+
"step": 1625
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.14,
|
| 476 |
+
"grad_norm": 2.3237972259521484,
|
| 477 |
+
"learning_rate": 8.789473684210527e-06,
|
| 478 |
+
"loss": 0.4696,
|
| 479 |
+
"step": 1650
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.16,
|
| 483 |
+
"grad_norm": 2.5998685359954834,
|
| 484 |
+
"learning_rate": 8.763157894736842e-06,
|
| 485 |
+
"loss": 0.4673,
|
| 486 |
+
"step": 1675
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.18,
|
| 490 |
+
"grad_norm": 2.7698214054107666,
|
| 491 |
+
"learning_rate": 8.736842105263158e-06,
|
| 492 |
+
"loss": 0.4721,
|
| 493 |
+
"step": 1700
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.2,
|
| 497 |
+
"grad_norm": 2.030637741088867,
|
| 498 |
+
"learning_rate": 8.710526315789475e-06,
|
| 499 |
+
"loss": 0.4578,
|
| 500 |
+
"step": 1725
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.21,
|
| 504 |
+
"grad_norm": 1.9983584880828857,
|
| 505 |
+
"learning_rate": 8.68421052631579e-06,
|
| 506 |
+
"loss": 0.4629,
|
| 507 |
+
"step": 1750
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.23,
|
| 511 |
+
"grad_norm": 2.231105327606201,
|
| 512 |
+
"learning_rate": 8.657894736842107e-06,
|
| 513 |
+
"loss": 0.4617,
|
| 514 |
+
"step": 1775
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.25,
|
| 518 |
+
"grad_norm": 2.0504438877105713,
|
| 519 |
+
"learning_rate": 8.63263157894737e-06,
|
| 520 |
+
"loss": 0.4675,
|
| 521 |
+
"step": 1800
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.26,
|
| 525 |
+
"grad_norm": 1.8002873659133911,
|
| 526 |
+
"learning_rate": 8.606315789473684e-06,
|
| 527 |
+
"loss": 0.469,
|
| 528 |
+
"step": 1825
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.28,
|
| 532 |
+
"grad_norm": 2.8942711353302,
|
| 533 |
+
"learning_rate": 8.580000000000001e-06,
|
| 534 |
+
"loss": 0.4668,
|
| 535 |
+
"step": 1850
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.3,
|
| 539 |
+
"grad_norm": 3.3724169731140137,
|
| 540 |
+
"learning_rate": 8.553684210526316e-06,
|
| 541 |
+
"loss": 0.467,
|
| 542 |
+
"step": 1875
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.32,
|
| 546 |
+
"grad_norm": 8.831938743591309,
|
| 547 |
+
"learning_rate": 8.527368421052632e-06,
|
| 548 |
+
"loss": 0.4659,
|
| 549 |
+
"step": 1900
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.33,
|
| 553 |
+
"grad_norm": 2.8694584369659424,
|
| 554 |
+
"learning_rate": 8.501052631578947e-06,
|
| 555 |
+
"loss": 0.4727,
|
| 556 |
+
"step": 1925
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.35,
|
| 560 |
+
"grad_norm": 3.35543155670166,
|
| 561 |
+
"learning_rate": 8.474736842105264e-06,
|
| 562 |
+
"loss": 0.4629,
|
| 563 |
+
"step": 1950
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.37,
|
| 567 |
+
"grad_norm": 2.251941204071045,
|
| 568 |
+
"learning_rate": 8.44842105263158e-06,
|
| 569 |
+
"loss": 0.46,
|
| 570 |
+
"step": 1975
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.39,
|
| 574 |
+
"grad_norm": 1.5079283714294434,
|
| 575 |
+
"learning_rate": 8.422105263157896e-06,
|
| 576 |
+
"loss": 0.4669,
|
| 577 |
+
"step": 2000
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.39,
|
| 581 |
+
"eval_loss": 0.43367624282836914,
|
| 582 |
+
"eval_runtime": 130.0756,
|
| 583 |
+
"eval_samples_per_second": 39.446,
|
| 584 |
+
"eval_steps_per_second": 19.727,
|
| 585 |
+
"step": 2000
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.4,
|
| 589 |
+
"grad_norm": 2.216298818588257,
|
| 590 |
+
"learning_rate": 8.395789473684212e-06,
|
| 591 |
+
"loss": 0.4584,
|
| 592 |
+
"step": 2025
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.42,
|
| 596 |
+
"grad_norm": 2.3113179206848145,
|
| 597 |
+
"learning_rate": 8.369473684210527e-06,
|
| 598 |
+
"loss": 0.4678,
|
| 599 |
+
"step": 2050
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.44,
|
| 603 |
+
"grad_norm": 2.3224334716796875,
|
| 604 |
+
"learning_rate": 8.343157894736842e-06,
|
| 605 |
+
"loss": 0.4565,
|
| 606 |
+
"step": 2075
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.46,
|
| 610 |
+
"grad_norm": 3.6630237102508545,
|
| 611 |
+
"learning_rate": 8.316842105263159e-06,
|
| 612 |
+
"loss": 0.4652,
|
| 613 |
+
"step": 2100
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.47,
|
| 617 |
+
"grad_norm": 2.178736448287964,
|
| 618 |
+
"learning_rate": 8.290526315789474e-06,
|
| 619 |
+
"loss": 0.4615,
|
| 620 |
+
"step": 2125
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.49,
|
| 624 |
+
"grad_norm": 2.6931562423706055,
|
| 625 |
+
"learning_rate": 8.26421052631579e-06,
|
| 626 |
+
"loss": 0.4649,
|
| 627 |
+
"step": 2150
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 1.51,
|
| 631 |
+
"grad_norm": 3.884937286376953,
|
| 632 |
+
"learning_rate": 8.237894736842106e-06,
|
| 633 |
+
"loss": 0.4665,
|
| 634 |
+
"step": 2175
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.52,
|
| 638 |
+
"grad_norm": 2.139063835144043,
|
| 639 |
+
"learning_rate": 8.211578947368422e-06,
|
| 640 |
+
"loss": 0.4629,
|
| 641 |
+
"step": 2200
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.54,
|
| 645 |
+
"grad_norm": 1.8598359823226929,
|
| 646 |
+
"learning_rate": 8.185263157894737e-06,
|
| 647 |
+
"loss": 0.4681,
|
| 648 |
+
"step": 2225
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.56,
|
| 652 |
+
"grad_norm": 2.011505126953125,
|
| 653 |
+
"learning_rate": 8.158947368421052e-06,
|
| 654 |
+
"loss": 0.4639,
|
| 655 |
+
"step": 2250
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.58,
|
| 659 |
+
"grad_norm": 2.256636381149292,
|
| 660 |
+
"learning_rate": 8.132631578947369e-06,
|
| 661 |
+
"loss": 0.4686,
|
| 662 |
+
"step": 2275
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.59,
|
| 666 |
+
"grad_norm": 2.452007293701172,
|
| 667 |
+
"learning_rate": 8.106315789473684e-06,
|
| 668 |
+
"loss": 0.4655,
|
| 669 |
+
"step": 2300
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 1.61,
|
| 673 |
+
"grad_norm": 2.439833164215088,
|
| 674 |
+
"learning_rate": 8.08e-06,
|
| 675 |
+
"loss": 0.4635,
|
| 676 |
+
"step": 2325
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.63,
|
| 680 |
+
"grad_norm": 3.101173162460327,
|
| 681 |
+
"learning_rate": 8.053684210526317e-06,
|
| 682 |
+
"loss": 0.4616,
|
| 683 |
+
"step": 2350
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.65,
|
| 687 |
+
"grad_norm": 3.1077115535736084,
|
| 688 |
+
"learning_rate": 8.027368421052632e-06,
|
| 689 |
+
"loss": 0.4645,
|
| 690 |
+
"step": 2375
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.66,
|
| 694 |
+
"grad_norm": 2.397768020629883,
|
| 695 |
+
"learning_rate": 8.001052631578949e-06,
|
| 696 |
+
"loss": 0.4551,
|
| 697 |
+
"step": 2400
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 1.68,
|
| 701 |
+
"grad_norm": 1.6911340951919556,
|
| 702 |
+
"learning_rate": 7.974736842105264e-06,
|
| 703 |
+
"loss": 0.4693,
|
| 704 |
+
"step": 2425
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.7,
|
| 708 |
+
"grad_norm": 4.298758506774902,
|
| 709 |
+
"learning_rate": 7.94842105263158e-06,
|
| 710 |
+
"loss": 0.463,
|
| 711 |
+
"step": 2450
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.72,
|
| 715 |
+
"grad_norm": 2.41021728515625,
|
| 716 |
+
"learning_rate": 7.922105263157895e-06,
|
| 717 |
+
"loss": 0.4597,
|
| 718 |
+
"step": 2475
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.73,
|
| 722 |
+
"grad_norm": 2.7337491512298584,
|
| 723 |
+
"learning_rate": 7.89578947368421e-06,
|
| 724 |
+
"loss": 0.4572,
|
| 725 |
+
"step": 2500
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.75,
|
| 729 |
+
"grad_norm": 1.6781957149505615,
|
| 730 |
+
"learning_rate": 7.869473684210527e-06,
|
| 731 |
+
"loss": 0.4588,
|
| 732 |
+
"step": 2525
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.77,
|
| 736 |
+
"grad_norm": 2.573647975921631,
|
| 737 |
+
"learning_rate": 7.843157894736842e-06,
|
| 738 |
+
"loss": 0.4662,
|
| 739 |
+
"step": 2550
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.78,
|
| 743 |
+
"grad_norm": 2.0476982593536377,
|
| 744 |
+
"learning_rate": 7.816842105263159e-06,
|
| 745 |
+
"loss": 0.4592,
|
| 746 |
+
"step": 2575
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.8,
|
| 750 |
+
"grad_norm": 1.947978138923645,
|
| 751 |
+
"learning_rate": 7.790526315789474e-06,
|
| 752 |
+
"loss": 0.4629,
|
| 753 |
+
"step": 2600
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.82,
|
| 757 |
+
"grad_norm": 1.7212955951690674,
|
| 758 |
+
"learning_rate": 7.76421052631579e-06,
|
| 759 |
+
"loss": 0.4593,
|
| 760 |
+
"step": 2625
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.84,
|
| 764 |
+
"grad_norm": 2.7885591983795166,
|
| 765 |
+
"learning_rate": 7.737894736842105e-06,
|
| 766 |
+
"loss": 0.469,
|
| 767 |
+
"step": 2650
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.85,
|
| 771 |
+
"grad_norm": 2.266343116760254,
|
| 772 |
+
"learning_rate": 7.711578947368422e-06,
|
| 773 |
+
"loss": 0.4541,
|
| 774 |
+
"step": 2675
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.87,
|
| 778 |
+
"grad_norm": 2.3520565032958984,
|
| 779 |
+
"learning_rate": 7.685263157894739e-06,
|
| 780 |
+
"loss": 0.4608,
|
| 781 |
+
"step": 2700
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.89,
|
| 785 |
+
"grad_norm": 2.0906155109405518,
|
| 786 |
+
"learning_rate": 7.658947368421054e-06,
|
| 787 |
+
"loss": 0.4585,
|
| 788 |
+
"step": 2725
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.91,
|
| 792 |
+
"grad_norm": 2.396610736846924,
|
| 793 |
+
"learning_rate": 7.632631578947369e-06,
|
| 794 |
+
"loss": 0.4607,
|
| 795 |
+
"step": 2750
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.92,
|
| 799 |
+
"grad_norm": 2.4076027870178223,
|
| 800 |
+
"learning_rate": 7.606315789473685e-06,
|
| 801 |
+
"loss": 0.4622,
|
| 802 |
+
"step": 2775
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.94,
|
| 806 |
+
"grad_norm": 5.321292400360107,
|
| 807 |
+
"learning_rate": 7.58e-06,
|
| 808 |
+
"loss": 0.4594,
|
| 809 |
+
"step": 2800
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.96,
|
| 813 |
+
"grad_norm": 4.2904253005981445,
|
| 814 |
+
"learning_rate": 7.553684210526316e-06,
|
| 815 |
+
"loss": 0.4596,
|
| 816 |
+
"step": 2825
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.98,
|
| 820 |
+
"grad_norm": 3.1583385467529297,
|
| 821 |
+
"learning_rate": 7.527368421052632e-06,
|
| 822 |
+
"loss": 0.4703,
|
| 823 |
+
"step": 2850
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1.99,
|
| 827 |
+
"grad_norm": 2.3681633472442627,
|
| 828 |
+
"learning_rate": 7.501052631578948e-06,
|
| 829 |
+
"loss": 0.4618,
|
| 830 |
+
"step": 2875
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 2.01,
|
| 834 |
+
"grad_norm": 2.9283578395843506,
|
| 835 |
+
"learning_rate": 7.4747368421052635e-06,
|
| 836 |
+
"loss": 0.461,
|
| 837 |
+
"step": 2900
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 2.03,
|
| 841 |
+
"grad_norm": 2.1495721340179443,
|
| 842 |
+
"learning_rate": 7.448421052631579e-06,
|
| 843 |
+
"loss": 0.4591,
|
| 844 |
+
"step": 2925
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 2.04,
|
| 848 |
+
"grad_norm": 2.898043394088745,
|
| 849 |
+
"learning_rate": 7.422105263157895e-06,
|
| 850 |
+
"loss": 0.4597,
|
| 851 |
+
"step": 2950
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 2.06,
|
| 855 |
+
"grad_norm": 2.5551648139953613,
|
| 856 |
+
"learning_rate": 7.395789473684211e-06,
|
| 857 |
+
"loss": 0.4598,
|
| 858 |
+
"step": 2975
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 2.08,
|
| 862 |
+
"grad_norm": 1.92994225025177,
|
| 863 |
+
"learning_rate": 7.369473684210528e-06,
|
| 864 |
+
"loss": 0.4647,
|
| 865 |
+
"step": 3000
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 2.08,
|
| 869 |
+
"eval_loss": 0.4316975176334381,
|
| 870 |
+
"eval_runtime": 128.6261,
|
| 871 |
+
"eval_samples_per_second": 39.891,
|
| 872 |
+
"eval_steps_per_second": 19.949,
|
| 873 |
+
"step": 3000
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 2.1,
|
| 877 |
+
"grad_norm": 2.059903383255005,
|
| 878 |
+
"learning_rate": 7.3431578947368435e-06,
|
| 879 |
+
"loss": 0.4634,
|
| 880 |
+
"step": 3025
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 2.11,
|
| 884 |
+
"grad_norm": 1.8484073877334595,
|
| 885 |
+
"learning_rate": 7.3168421052631585e-06,
|
| 886 |
+
"loss": 0.4622,
|
| 887 |
+
"step": 3050
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 2.13,
|
| 891 |
+
"grad_norm": 2.5850799083709717,
|
| 892 |
+
"learning_rate": 7.290526315789474e-06,
|
| 893 |
+
"loss": 0.4479,
|
| 894 |
+
"step": 3075
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 2.15,
|
| 898 |
+
"grad_norm": 3.510633945465088,
|
| 899 |
+
"learning_rate": 7.26421052631579e-06,
|
| 900 |
+
"loss": 0.4585,
|
| 901 |
+
"step": 3100
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 2.17,
|
| 905 |
+
"grad_norm": 3.163177490234375,
|
| 906 |
+
"learning_rate": 7.237894736842106e-06,
|
| 907 |
+
"loss": 0.4682,
|
| 908 |
+
"step": 3125
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 2.18,
|
| 912 |
+
"grad_norm": 3.7852284908294678,
|
| 913 |
+
"learning_rate": 7.211578947368422e-06,
|
| 914 |
+
"loss": 0.4552,
|
| 915 |
+
"step": 3150
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 2.2,
|
| 919 |
+
"grad_norm": 1.6458024978637695,
|
| 920 |
+
"learning_rate": 7.1852631578947375e-06,
|
| 921 |
+
"loss": 0.4623,
|
| 922 |
+
"step": 3175
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 2.22,
|
| 926 |
+
"grad_norm": 2.3463425636291504,
|
| 927 |
+
"learning_rate": 7.158947368421053e-06,
|
| 928 |
+
"loss": 0.4645,
|
| 929 |
+
"step": 3200
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 2.24,
|
| 933 |
+
"grad_norm": 3.481792688369751,
|
| 934 |
+
"learning_rate": 7.132631578947369e-06,
|
| 935 |
+
"loss": 0.4562,
|
| 936 |
+
"step": 3225
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 2.25,
|
| 940 |
+
"grad_norm": 2.2543387413024902,
|
| 941 |
+
"learning_rate": 7.106315789473684e-06,
|
| 942 |
+
"loss": 0.4577,
|
| 943 |
+
"step": 3250
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 2.27,
|
| 947 |
+
"grad_norm": 3.5041863918304443,
|
| 948 |
+
"learning_rate": 7.08e-06,
|
| 949 |
+
"loss": 0.4614,
|
| 950 |
+
"step": 3275
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 2.29,
|
| 954 |
+
"grad_norm": 1.6661858558654785,
|
| 955 |
+
"learning_rate": 7.053684210526316e-06,
|
| 956 |
+
"loss": 0.4611,
|
| 957 |
+
"step": 3300
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 2.3,
|
| 961 |
+
"grad_norm": 1.6855933666229248,
|
| 962 |
+
"learning_rate": 7.027368421052632e-06,
|
| 963 |
+
"loss": 0.4638,
|
| 964 |
+
"step": 3325
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 2.32,
|
| 968 |
+
"grad_norm": 1.8656867742538452,
|
| 969 |
+
"learning_rate": 7.001052631578948e-06,
|
| 970 |
+
"loss": 0.4633,
|
| 971 |
+
"step": 3350
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 2.34,
|
| 975 |
+
"grad_norm": 1.9860706329345703,
|
| 976 |
+
"learning_rate": 6.974736842105264e-06,
|
| 977 |
+
"loss": 0.4609,
|
| 978 |
+
"step": 3375
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 2.36,
|
| 982 |
+
"grad_norm": 3.0542991161346436,
|
| 983 |
+
"learning_rate": 6.94842105263158e-06,
|
| 984 |
+
"loss": 0.4532,
|
| 985 |
+
"step": 3400
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 2.37,
|
| 989 |
+
"grad_norm": 2.680880308151245,
|
| 990 |
+
"learning_rate": 6.922105263157896e-06,
|
| 991 |
+
"loss": 0.4561,
|
| 992 |
+
"step": 3425
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 2.39,
|
| 996 |
+
"grad_norm": 2.62613582611084,
|
| 997 |
+
"learning_rate": 6.8957894736842116e-06,
|
| 998 |
+
"loss": 0.4646,
|
| 999 |
+
"step": 3450
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 2.41,
|
| 1003 |
+
"grad_norm": 2.2299606800079346,
|
| 1004 |
+
"learning_rate": 6.869473684210527e-06,
|
| 1005 |
+
"loss": 0.4591,
|
| 1006 |
+
"step": 3475
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 2.43,
|
| 1010 |
+
"grad_norm": 3.0914087295532227,
|
| 1011 |
+
"learning_rate": 6.843157894736842e-06,
|
| 1012 |
+
"loss": 0.4625,
|
| 1013 |
+
"step": 3500
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 2.44,
|
| 1017 |
+
"grad_norm": 2.1178531646728516,
|
| 1018 |
+
"learning_rate": 6.816842105263158e-06,
|
| 1019 |
+
"loss": 0.4627,
|
| 1020 |
+
"step": 3525
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 2.46,
|
| 1024 |
+
"grad_norm": 2.3285298347473145,
|
| 1025 |
+
"learning_rate": 6.790526315789474e-06,
|
| 1026 |
+
"loss": 0.454,
|
| 1027 |
+
"step": 3550
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 2.48,
|
| 1031 |
+
"grad_norm": 1.802120327949524,
|
| 1032 |
+
"learning_rate": 6.76421052631579e-06,
|
| 1033 |
+
"loss": 0.4637,
|
| 1034 |
+
"step": 3575
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 2.5,
|
| 1038 |
+
"grad_norm": 3.2270822525024414,
|
| 1039 |
+
"learning_rate": 6.737894736842106e-06,
|
| 1040 |
+
"loss": 0.4633,
|
| 1041 |
+
"step": 3600
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 2.51,
|
| 1045 |
+
"grad_norm": 2.21866512298584,
|
| 1046 |
+
"learning_rate": 6.7115789473684214e-06,
|
| 1047 |
+
"loss": 0.4627,
|
| 1048 |
+
"step": 3625
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 2.53,
|
| 1052 |
+
"grad_norm": 3.067976951599121,
|
| 1053 |
+
"learning_rate": 6.685263157894737e-06,
|
| 1054 |
+
"loss": 0.4609,
|
| 1055 |
+
"step": 3650
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 2.55,
|
| 1059 |
+
"grad_norm": 2.3835887908935547,
|
| 1060 |
+
"learning_rate": 6.658947368421054e-06,
|
| 1061 |
+
"loss": 0.4547,
|
| 1062 |
+
"step": 3675
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 2.56,
|
| 1066 |
+
"grad_norm": 3.4934396743774414,
|
| 1067 |
+
"learning_rate": 6.63263157894737e-06,
|
| 1068 |
+
"loss": 0.4513,
|
| 1069 |
+
"step": 3700
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 2.58,
|
| 1073 |
+
"grad_norm": 2.1658859252929688,
|
| 1074 |
+
"learning_rate": 6.606315789473685e-06,
|
| 1075 |
+
"loss": 0.4568,
|
| 1076 |
+
"step": 3725
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 2.6,
|
| 1080 |
+
"grad_norm": 1.9029380083084106,
|
| 1081 |
+
"learning_rate": 6.5800000000000005e-06,
|
| 1082 |
+
"loss": 0.4554,
|
| 1083 |
+
"step": 3750
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 2.62,
|
| 1087 |
+
"grad_norm": 1.7163124084472656,
|
| 1088 |
+
"learning_rate": 6.553684210526316e-06,
|
| 1089 |
+
"loss": 0.4619,
|
| 1090 |
+
"step": 3775
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 2.63,
|
| 1094 |
+
"grad_norm": 4.556973457336426,
|
| 1095 |
+
"learning_rate": 6.527368421052632e-06,
|
| 1096 |
+
"loss": 0.4598,
|
| 1097 |
+
"step": 3800
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 2.65,
|
| 1101 |
+
"grad_norm": 2.304401397705078,
|
| 1102 |
+
"learning_rate": 6.501052631578948e-06,
|
| 1103 |
+
"loss": 0.4613,
|
| 1104 |
+
"step": 3825
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 2.67,
|
| 1108 |
+
"grad_norm": 2.5970966815948486,
|
| 1109 |
+
"learning_rate": 6.474736842105264e-06,
|
| 1110 |
+
"loss": 0.4617,
|
| 1111 |
+
"step": 3850
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 2.69,
|
| 1115 |
+
"grad_norm": 1.3288301229476929,
|
| 1116 |
+
"learning_rate": 6.44842105263158e-06,
|
| 1117 |
+
"loss": 0.4555,
|
| 1118 |
+
"step": 3875
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 2.7,
|
| 1122 |
+
"grad_norm": 1.891694188117981,
|
| 1123 |
+
"learning_rate": 6.4221052631578954e-06,
|
| 1124 |
+
"loss": 0.4534,
|
| 1125 |
+
"step": 3900
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 2.72,
|
| 1129 |
+
"grad_norm": 2.017850637435913,
|
| 1130 |
+
"learning_rate": 6.39578947368421e-06,
|
| 1131 |
+
"loss": 0.4577,
|
| 1132 |
+
"step": 3925
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 2.74,
|
| 1136 |
+
"grad_norm": 3.5200083255767822,
|
| 1137 |
+
"learning_rate": 6.369473684210526e-06,
|
| 1138 |
+
"loss": 0.4496,
|
| 1139 |
+
"step": 3950
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 2.75,
|
| 1143 |
+
"grad_norm": 2.430833339691162,
|
| 1144 |
+
"learning_rate": 6.343157894736842e-06,
|
| 1145 |
+
"loss": 0.4471,
|
| 1146 |
+
"step": 3975
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 2.77,
|
| 1150 |
+
"grad_norm": 2.285391092300415,
|
| 1151 |
+
"learning_rate": 6.316842105263158e-06,
|
| 1152 |
+
"loss": 0.4596,
|
| 1153 |
+
"step": 4000
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 2.77,
|
| 1157 |
+
"eval_loss": 0.4282490015029907,
|
| 1158 |
+
"eval_runtime": 127.4696,
|
| 1159 |
+
"eval_samples_per_second": 40.253,
|
| 1160 |
+
"eval_steps_per_second": 20.13,
|
| 1161 |
+
"step": 4000
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 2.79,
|
| 1165 |
+
"grad_norm": Infinity,
|
| 1166 |
+
"learning_rate": 6.291578947368422e-06,
|
| 1167 |
+
"loss": 0.4625,
|
| 1168 |
+
"step": 4025
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 2.81,
|
| 1172 |
+
"grad_norm": 2.3426144123077393,
|
| 1173 |
+
"learning_rate": 6.265263157894738e-06,
|
| 1174 |
+
"loss": 0.461,
|
| 1175 |
+
"step": 4050
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 2.82,
|
| 1179 |
+
"grad_norm": 2.6181225776672363,
|
| 1180 |
+
"learning_rate": 6.238947368421054e-06,
|
| 1181 |
+
"loss": 0.4601,
|
| 1182 |
+
"step": 4075
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 2.84,
|
| 1186 |
+
"grad_norm": 4.169466018676758,
|
| 1187 |
+
"learning_rate": 6.212631578947369e-06,
|
| 1188 |
+
"loss": 0.4582,
|
| 1189 |
+
"step": 4100
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 2.86,
|
| 1193 |
+
"grad_norm": 2.541663408279419,
|
| 1194 |
+
"learning_rate": 6.1863157894736845e-06,
|
| 1195 |
+
"loss": 0.462,
|
| 1196 |
+
"step": 4125
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.88,
|
| 1200 |
+
"grad_norm": 6.109897613525391,
|
| 1201 |
+
"learning_rate": 6.16e-06,
|
| 1202 |
+
"loss": 0.4574,
|
| 1203 |
+
"step": 4150
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 2.89,
|
| 1207 |
+
"grad_norm": 2.4529943466186523,
|
| 1208 |
+
"learning_rate": 6.133684210526316e-06,
|
| 1209 |
+
"loss": 0.4622,
|
| 1210 |
+
"step": 4175
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 2.91,
|
| 1214 |
+
"grad_norm": 2.091356039047241,
|
| 1215 |
+
"learning_rate": 6.107368421052632e-06,
|
| 1216 |
+
"loss": 0.4624,
|
| 1217 |
+
"step": 4200
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 2.93,
|
| 1221 |
+
"grad_norm": 2.5841152667999268,
|
| 1222 |
+
"learning_rate": 6.081052631578948e-06,
|
| 1223 |
+
"loss": 0.4573,
|
| 1224 |
+
"step": 4225
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 2.95,
|
| 1228 |
+
"grad_norm": 2.812778949737549,
|
| 1229 |
+
"learning_rate": 6.054736842105264e-06,
|
| 1230 |
+
"loss": 0.4529,
|
| 1231 |
+
"step": 4250
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 2.96,
|
| 1235 |
+
"grad_norm": 2.070568084716797,
|
| 1236 |
+
"learning_rate": 6.0284210526315786e-06,
|
| 1237 |
+
"loss": 0.4621,
|
| 1238 |
+
"step": 4275
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 2.98,
|
| 1242 |
+
"grad_norm": 2.89155650138855,
|
| 1243 |
+
"learning_rate": 6.002105263157896e-06,
|
| 1244 |
+
"loss": 0.4534,
|
| 1245 |
+
"step": 4300
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 3.0,
|
| 1249 |
+
"grad_norm": 1.926348328590393,
|
| 1250 |
+
"learning_rate": 5.975789473684212e-06,
|
| 1251 |
+
"loss": 0.4558,
|
| 1252 |
+
"step": 4325
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 3.01,
|
| 1256 |
+
"grad_norm": 2.3729255199432373,
|
| 1257 |
+
"learning_rate": 5.949473684210527e-06,
|
| 1258 |
+
"loss": 0.4569,
|
| 1259 |
+
"step": 4350
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 3.03,
|
| 1263 |
+
"grad_norm": 2.6525228023529053,
|
| 1264 |
+
"learning_rate": 5.923157894736843e-06,
|
| 1265 |
+
"loss": 0.4593,
|
| 1266 |
+
"step": 4375
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 3.05,
|
| 1270 |
+
"grad_norm": 2.258516311645508,
|
| 1271 |
+
"learning_rate": 5.8968421052631585e-06,
|
| 1272 |
+
"loss": 0.4563,
|
| 1273 |
+
"step": 4400
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 3.07,
|
| 1277 |
+
"grad_norm": 3.1756558418273926,
|
| 1278 |
+
"learning_rate": 5.870526315789474e-06,
|
| 1279 |
+
"loss": 0.4565,
|
| 1280 |
+
"step": 4425
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 3.08,
|
| 1284 |
+
"grad_norm": 2.4328861236572266,
|
| 1285 |
+
"learning_rate": 5.84421052631579e-06,
|
| 1286 |
+
"loss": 0.457,
|
| 1287 |
+
"step": 4450
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 3.1,
|
| 1291 |
+
"grad_norm": 2.02143931388855,
|
| 1292 |
+
"learning_rate": 5.817894736842106e-06,
|
| 1293 |
+
"loss": 0.4579,
|
| 1294 |
+
"step": 4475
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 3.12,
|
| 1298 |
+
"grad_norm": 1.9647924900054932,
|
| 1299 |
+
"learning_rate": 5.791578947368422e-06,
|
| 1300 |
+
"loss": 0.4577,
|
| 1301 |
+
"step": 4500
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 3.14,
|
| 1305 |
+
"grad_norm": 3.5917584896087646,
|
| 1306 |
+
"learning_rate": 5.765263157894737e-06,
|
| 1307 |
+
"loss": 0.4486,
|
| 1308 |
+
"step": 4525
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 3.15,
|
| 1312 |
+
"grad_norm": 3.461472749710083,
|
| 1313 |
+
"learning_rate": 5.7389473684210526e-06,
|
| 1314 |
+
"loss": 0.4603,
|
| 1315 |
+
"step": 4550
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 3.17,
|
| 1319 |
+
"grad_norm": 2.0499308109283447,
|
| 1320 |
+
"learning_rate": 5.712631578947368e-06,
|
| 1321 |
+
"loss": 0.4572,
|
| 1322 |
+
"step": 4575
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 3.19,
|
| 1326 |
+
"grad_norm": 2.1889617443084717,
|
| 1327 |
+
"learning_rate": 5.686315789473684e-06,
|
| 1328 |
+
"loss": 0.4567,
|
| 1329 |
+
"step": 4600
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 3.21,
|
| 1333 |
+
"grad_norm": 1.788960576057434,
|
| 1334 |
+
"learning_rate": 5.66e-06,
|
| 1335 |
+
"loss": 0.4523,
|
| 1336 |
+
"step": 4625
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 3.22,
|
| 1340 |
+
"grad_norm": 3.5286545753479004,
|
| 1341 |
+
"learning_rate": 5.633684210526317e-06,
|
| 1342 |
+
"loss": 0.4568,
|
| 1343 |
+
"step": 4650
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 3.24,
|
| 1347 |
+
"grad_norm": 1.9870589971542358,
|
| 1348 |
+
"learning_rate": 5.6073684210526325e-06,
|
| 1349 |
+
"loss": 0.4586,
|
| 1350 |
+
"step": 4675
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 3.26,
|
| 1354 |
+
"grad_norm": 3.326669216156006,
|
| 1355 |
+
"learning_rate": 5.581052631578948e-06,
|
| 1356 |
+
"loss": 0.4629,
|
| 1357 |
+
"step": 4700
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 3.27,
|
| 1361 |
+
"grad_norm": 1.8193014860153198,
|
| 1362 |
+
"learning_rate": 5.554736842105264e-06,
|
| 1363 |
+
"loss": 0.4508,
|
| 1364 |
+
"step": 4725
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 3.29,
|
| 1368 |
+
"grad_norm": 2.358961343765259,
|
| 1369 |
+
"learning_rate": 5.52842105263158e-06,
|
| 1370 |
+
"loss": 0.4572,
|
| 1371 |
+
"step": 4750
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 3.31,
|
| 1375 |
+
"grad_norm": 2.7740235328674316,
|
| 1376 |
+
"learning_rate": 5.502105263157895e-06,
|
| 1377 |
+
"loss": 0.4644,
|
| 1378 |
+
"step": 4775
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 3.33,
|
| 1382 |
+
"grad_norm": 4.845912933349609,
|
| 1383 |
+
"learning_rate": 5.475789473684211e-06,
|
| 1384 |
+
"loss": 0.4558,
|
| 1385 |
+
"step": 4800
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 3.34,
|
| 1389 |
+
"grad_norm": 2.390437364578247,
|
| 1390 |
+
"learning_rate": 5.4494736842105266e-06,
|
| 1391 |
+
"loss": 0.4605,
|
| 1392 |
+
"step": 4825
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 3.36,
|
| 1396 |
+
"grad_norm": 2.571261405944824,
|
| 1397 |
+
"learning_rate": 5.423157894736842e-06,
|
| 1398 |
+
"loss": 0.4562,
|
| 1399 |
+
"step": 4850
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 3.38,
|
| 1403 |
+
"grad_norm": 2.0128092765808105,
|
| 1404 |
+
"learning_rate": 5.396842105263158e-06,
|
| 1405 |
+
"loss": 0.46,
|
| 1406 |
+
"step": 4875
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 3.4,
|
| 1410 |
+
"grad_norm": 2.0535905361175537,
|
| 1411 |
+
"learning_rate": 5.370526315789474e-06,
|
| 1412 |
+
"loss": 0.4505,
|
| 1413 |
+
"step": 4900
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 3.41,
|
| 1417 |
+
"grad_norm": 10.284537315368652,
|
| 1418 |
+
"learning_rate": 5.34421052631579e-06,
|
| 1419 |
+
"loss": 0.4598,
|
| 1420 |
+
"step": 4925
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 3.43,
|
| 1424 |
+
"grad_norm": 1.640753984451294,
|
| 1425 |
+
"learning_rate": 5.317894736842105e-06,
|
| 1426 |
+
"loss": 0.4558,
|
| 1427 |
+
"step": 4950
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 3.45,
|
| 1431 |
+
"grad_norm": 3.6907098293304443,
|
| 1432 |
+
"learning_rate": 5.291578947368422e-06,
|
| 1433 |
+
"loss": 0.4521,
|
| 1434 |
+
"step": 4975
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 3.47,
|
| 1438 |
+
"grad_norm": 3.541050672531128,
|
| 1439 |
+
"learning_rate": 5.265263157894738e-06,
|
| 1440 |
+
"loss": 0.4528,
|
| 1441 |
+
"step": 5000
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 3.47,
|
| 1445 |
+
"eval_loss": 0.4256468117237091,
|
| 1446 |
+
"eval_runtime": 131.0319,
|
| 1447 |
+
"eval_samples_per_second": 39.158,
|
| 1448 |
+
"eval_steps_per_second": 19.583,
|
| 1449 |
+
"step": 5000
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 3.48,
|
| 1453 |
+
"grad_norm": 3.7902073860168457,
|
| 1454 |
+
"learning_rate": 5.238947368421053e-06,
|
| 1455 |
+
"loss": 0.4571,
|
| 1456 |
+
"step": 5025
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 3.5,
|
| 1460 |
+
"grad_norm": 4.066291332244873,
|
| 1461 |
+
"learning_rate": 5.212631578947369e-06,
|
| 1462 |
+
"loss": 0.4512,
|
| 1463 |
+
"step": 5050
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 3.52,
|
| 1467 |
+
"grad_norm": 2.505599021911621,
|
| 1468 |
+
"learning_rate": 5.186315789473685e-06,
|
| 1469 |
+
"loss": 0.4468,
|
| 1470 |
+
"step": 5075
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 3.53,
|
| 1474 |
+
"grad_norm": 1.8621853590011597,
|
| 1475 |
+
"learning_rate": 5.1600000000000006e-06,
|
| 1476 |
+
"loss": 0.4523,
|
| 1477 |
+
"step": 5100
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 3.55,
|
| 1481 |
+
"grad_norm": 2.166684865951538,
|
| 1482 |
+
"learning_rate": 5.133684210526316e-06,
|
| 1483 |
+
"loss": 0.4545,
|
| 1484 |
+
"step": 5125
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 3.57,
|
| 1488 |
+
"grad_norm": 2.113370656967163,
|
| 1489 |
+
"learning_rate": 5.107368421052632e-06,
|
| 1490 |
+
"loss": 0.4564,
|
| 1491 |
+
"step": 5150
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 3.59,
|
| 1495 |
+
"grad_norm": 1.8998031616210938,
|
| 1496 |
+
"learning_rate": 5.081052631578948e-06,
|
| 1497 |
+
"loss": 0.4527,
|
| 1498 |
+
"step": 5175
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 3.6,
|
| 1502 |
+
"grad_norm": 1.5921194553375244,
|
| 1503 |
+
"learning_rate": 5.054736842105263e-06,
|
| 1504 |
+
"loss": 0.4562,
|
| 1505 |
+
"step": 5200
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 3.62,
|
| 1509 |
+
"grad_norm": 6.103330612182617,
|
| 1510 |
+
"learning_rate": 5.028421052631579e-06,
|
| 1511 |
+
"loss": 0.4549,
|
| 1512 |
+
"step": 5225
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 3.64,
|
| 1516 |
+
"grad_norm": 3.6465394496917725,
|
| 1517 |
+
"learning_rate": 5.002105263157895e-06,
|
| 1518 |
+
"loss": 0.4576,
|
| 1519 |
+
"step": 5250
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 3.66,
|
| 1523 |
+
"grad_norm": 2.6215322017669678,
|
| 1524 |
+
"learning_rate": 4.975789473684211e-06,
|
| 1525 |
+
"loss": 0.4587,
|
| 1526 |
+
"step": 5275
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 3.67,
|
| 1530 |
+
"grad_norm": 2.8484139442443848,
|
| 1531 |
+
"learning_rate": 4.949473684210527e-06,
|
| 1532 |
+
"loss": 0.4524,
|
| 1533 |
+
"step": 5300
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 3.69,
|
| 1537 |
+
"grad_norm": 2.0593345165252686,
|
| 1538 |
+
"learning_rate": 4.923157894736842e-06,
|
| 1539 |
+
"loss": 0.4486,
|
| 1540 |
+
"step": 5325
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 3.71,
|
| 1544 |
+
"grad_norm": 2.749562978744507,
|
| 1545 |
+
"learning_rate": 4.896842105263158e-06,
|
| 1546 |
+
"loss": 0.4556,
|
| 1547 |
+
"step": 5350
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 3.73,
|
| 1551 |
+
"grad_norm": 5.627275466918945,
|
| 1552 |
+
"learning_rate": 4.870526315789474e-06,
|
| 1553 |
+
"loss": 0.4535,
|
| 1554 |
+
"step": 5375
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 3.74,
|
| 1558 |
+
"grad_norm": 4.1247992515563965,
|
| 1559 |
+
"learning_rate": 4.845263157894737e-06,
|
| 1560 |
+
"loss": 0.4585,
|
| 1561 |
+
"step": 5400
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 3.76,
|
| 1565 |
+
"grad_norm": 1.740677833557129,
|
| 1566 |
+
"learning_rate": 4.818947368421053e-06,
|
| 1567 |
+
"loss": 0.4613,
|
| 1568 |
+
"step": 5425
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 3.78,
|
| 1572 |
+
"grad_norm": 2.885836362838745,
|
| 1573 |
+
"learning_rate": 4.792631578947369e-06,
|
| 1574 |
+
"loss": 0.4509,
|
| 1575 |
+
"step": 5450
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 3.79,
|
| 1579 |
+
"grad_norm": 2.179293394088745,
|
| 1580 |
+
"learning_rate": 4.7663157894736845e-06,
|
| 1581 |
+
"loss": 0.4626,
|
| 1582 |
+
"step": 5475
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 3.81,
|
| 1586 |
+
"grad_norm": 2.766608953475952,
|
| 1587 |
+
"learning_rate": 4.74e-06,
|
| 1588 |
+
"loss": 0.4532,
|
| 1589 |
+
"step": 5500
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 3.83,
|
| 1593 |
+
"grad_norm": 2.0567455291748047,
|
| 1594 |
+
"learning_rate": 4.713684210526316e-06,
|
| 1595 |
+
"loss": 0.456,
|
| 1596 |
+
"step": 5525
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 3.85,
|
| 1600 |
+
"grad_norm": 3.076902151107788,
|
| 1601 |
+
"learning_rate": 4.687368421052632e-06,
|
| 1602 |
+
"loss": 0.4456,
|
| 1603 |
+
"step": 5550
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 3.86,
|
| 1607 |
+
"grad_norm": 2.915499687194824,
|
| 1608 |
+
"learning_rate": 4.661052631578948e-06,
|
| 1609 |
+
"loss": 0.4504,
|
| 1610 |
+
"step": 5575
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 3.88,
|
| 1614 |
+
"grad_norm": 3.0495471954345703,
|
| 1615 |
+
"learning_rate": 4.634736842105264e-06,
|
| 1616 |
+
"loss": 0.4539,
|
| 1617 |
+
"step": 5600
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 3.9,
|
| 1621 |
+
"grad_norm": 1.6944961547851562,
|
| 1622 |
+
"learning_rate": 4.6084210526315794e-06,
|
| 1623 |
+
"loss": 0.4541,
|
| 1624 |
+
"step": 5625
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 3.92,
|
| 1628 |
+
"grad_norm": 2.135935068130493,
|
| 1629 |
+
"learning_rate": 4.582105263157895e-06,
|
| 1630 |
+
"loss": 0.4525,
|
| 1631 |
+
"step": 5650
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 3.93,
|
| 1635 |
+
"grad_norm": 6.483078956604004,
|
| 1636 |
+
"learning_rate": 4.55578947368421e-06,
|
| 1637 |
+
"loss": 0.4535,
|
| 1638 |
+
"step": 5675
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 3.95,
|
| 1642 |
+
"grad_norm": 2.811405658721924,
|
| 1643 |
+
"learning_rate": 4.529473684210527e-06,
|
| 1644 |
+
"loss": 0.4538,
|
| 1645 |
+
"step": 5700
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 3.97,
|
| 1649 |
+
"grad_norm": 3.5844473838806152,
|
| 1650 |
+
"learning_rate": 4.503157894736843e-06,
|
| 1651 |
+
"loss": 0.4556,
|
| 1652 |
+
"step": 5725
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 3.99,
|
| 1656 |
+
"grad_norm": 2.2661666870117188,
|
| 1657 |
+
"learning_rate": 4.4768421052631585e-06,
|
| 1658 |
+
"loss": 0.4585,
|
| 1659 |
+
"step": 5750
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 4.0,
|
| 1663 |
+
"grad_norm": 2.6553196907043457,
|
| 1664 |
+
"learning_rate": 4.450526315789474e-06,
|
| 1665 |
+
"loss": 0.441,
|
| 1666 |
+
"step": 5775
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 4.02,
|
| 1670 |
+
"grad_norm": 2.779641628265381,
|
| 1671 |
+
"learning_rate": 4.424210526315789e-06,
|
| 1672 |
+
"loss": 0.4552,
|
| 1673 |
+
"step": 5800
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 4.04,
|
| 1677 |
+
"grad_norm": 11.688698768615723,
|
| 1678 |
+
"learning_rate": 4.397894736842105e-06,
|
| 1679 |
+
"loss": 0.4496,
|
| 1680 |
+
"step": 5825
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 4.05,
|
| 1684 |
+
"grad_norm": 1.6974663734436035,
|
| 1685 |
+
"learning_rate": 4.371578947368421e-06,
|
| 1686 |
+
"loss": 0.4552,
|
| 1687 |
+
"step": 5850
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 4.07,
|
| 1691 |
+
"grad_norm": 3.191411256790161,
|
| 1692 |
+
"learning_rate": 4.345263157894738e-06,
|
| 1693 |
+
"loss": 0.4522,
|
| 1694 |
+
"step": 5875
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 4.09,
|
| 1698 |
+
"grad_norm": 2.5754339694976807,
|
| 1699 |
+
"learning_rate": 4.3189473684210535e-06,
|
| 1700 |
+
"loss": 0.4565,
|
| 1701 |
+
"step": 5900
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 4.11,
|
| 1705 |
+
"grad_norm": 2.6307075023651123,
|
| 1706 |
+
"learning_rate": 4.2926315789473684e-06,
|
| 1707 |
+
"loss": 0.4446,
|
| 1708 |
+
"step": 5925
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 4.12,
|
| 1712 |
+
"grad_norm": 2.7968878746032715,
|
| 1713 |
+
"learning_rate": 4.266315789473684e-06,
|
| 1714 |
+
"loss": 0.4522,
|
| 1715 |
+
"step": 5950
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 4.14,
|
| 1719 |
+
"grad_norm": 4.277090549468994,
|
| 1720 |
+
"learning_rate": 4.24e-06,
|
| 1721 |
+
"loss": 0.454,
|
| 1722 |
+
"step": 5975
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 4.16,
|
| 1726 |
+
"grad_norm": 4.684088706970215,
|
| 1727 |
+
"learning_rate": 4.213684210526316e-06,
|
| 1728 |
+
"loss": 0.4564,
|
| 1729 |
+
"step": 6000
|
| 1730 |
+
},
|
| 1731 |
+
{
|
| 1732 |
+
"epoch": 4.16,
|
| 1733 |
+
"eval_loss": 0.4263022243976593,
|
| 1734 |
+
"eval_runtime": 129.2969,
|
| 1735 |
+
"eval_samples_per_second": 39.684,
|
| 1736 |
+
"eval_steps_per_second": 19.846,
|
| 1737 |
+
"step": 6000
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 4.18,
|
| 1741 |
+
"grad_norm": 2.7073895931243896,
|
| 1742 |
+
"learning_rate": 4.1873684210526325e-06,
|
| 1743 |
+
"loss": 0.4527,
|
| 1744 |
+
"step": 6025
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 4.19,
|
| 1748 |
+
"grad_norm": 1.904488444328308,
|
| 1749 |
+
"learning_rate": 4.1610526315789475e-06,
|
| 1750 |
+
"loss": 0.454,
|
| 1751 |
+
"step": 6050
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 4.21,
|
| 1755 |
+
"grad_norm": 2.016899347305298,
|
| 1756 |
+
"learning_rate": 4.134736842105263e-06,
|
| 1757 |
+
"loss": 0.4482,
|
| 1758 |
+
"step": 6075
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 4.23,
|
| 1762 |
+
"grad_norm": 2.5908937454223633,
|
| 1763 |
+
"learning_rate": 4.108421052631579e-06,
|
| 1764 |
+
"loss": 0.4545,
|
| 1765 |
+
"step": 6100
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 4.24,
|
| 1769 |
+
"grad_norm": 2.05924129486084,
|
| 1770 |
+
"learning_rate": 4.082105263157895e-06,
|
| 1771 |
+
"loss": 0.4561,
|
| 1772 |
+
"step": 6125
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 4.26,
|
| 1776 |
+
"grad_norm": 1.9701108932495117,
|
| 1777 |
+
"learning_rate": 4.055789473684211e-06,
|
| 1778 |
+
"loss": 0.4505,
|
| 1779 |
+
"step": 6150
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 4.28,
|
| 1783 |
+
"grad_norm": 2.8021397590637207,
|
| 1784 |
+
"learning_rate": 4.029473684210527e-06,
|
| 1785 |
+
"loss": 0.4512,
|
| 1786 |
+
"step": 6175
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 4.3,
|
| 1790 |
+
"grad_norm": 1.7641443014144897,
|
| 1791 |
+
"learning_rate": 4.0031578947368424e-06,
|
| 1792 |
+
"loss": 0.4492,
|
| 1793 |
+
"step": 6200
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 4.31,
|
| 1797 |
+
"grad_norm": 2.424809217453003,
|
| 1798 |
+
"learning_rate": 3.976842105263158e-06,
|
| 1799 |
+
"loss": 0.4591,
|
| 1800 |
+
"step": 6225
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 4.33,
|
| 1804 |
+
"grad_norm": 3.036320686340332,
|
| 1805 |
+
"learning_rate": 3.950526315789474e-06,
|
| 1806 |
+
"loss": 0.448,
|
| 1807 |
+
"step": 6250
|
| 1808 |
+
},
|
| 1809 |
+
{
|
| 1810 |
+
"epoch": 4.35,
|
| 1811 |
+
"grad_norm": 2.072425127029419,
|
| 1812 |
+
"learning_rate": 3.92421052631579e-06,
|
| 1813 |
+
"loss": 0.4459,
|
| 1814 |
+
"step": 6275
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 4.37,
|
| 1818 |
+
"grad_norm": 14.710461616516113,
|
| 1819 |
+
"learning_rate": 3.897894736842106e-06,
|
| 1820 |
+
"loss": 0.4552,
|
| 1821 |
+
"step": 6300
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 4.38,
|
| 1825 |
+
"grad_norm": 2.3576061725616455,
|
| 1826 |
+
"learning_rate": 3.8715789473684215e-06,
|
| 1827 |
+
"loss": 0.4528,
|
| 1828 |
+
"step": 6325
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 4.4,
|
| 1832 |
+
"grad_norm": 1.984506368637085,
|
| 1833 |
+
"learning_rate": 3.845263157894737e-06,
|
| 1834 |
+
"loss": 0.4501,
|
| 1835 |
+
"step": 6350
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 4.42,
|
| 1839 |
+
"grad_norm": 2.9575440883636475,
|
| 1840 |
+
"learning_rate": 3.818947368421053e-06,
|
| 1841 |
+
"loss": 0.4505,
|
| 1842 |
+
"step": 6375
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 4.44,
|
| 1846 |
+
"grad_norm": 2.4890646934509277,
|
| 1847 |
+
"learning_rate": 3.792631578947369e-06,
|
| 1848 |
+
"loss": 0.4546,
|
| 1849 |
+
"step": 6400
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 4.45,
|
| 1853 |
+
"grad_norm": 1.8990737199783325,
|
| 1854 |
+
"learning_rate": 3.766315789473685e-06,
|
| 1855 |
+
"loss": 0.4534,
|
| 1856 |
+
"step": 6425
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 4.47,
|
| 1860 |
+
"grad_norm": 3.2801966667175293,
|
| 1861 |
+
"learning_rate": 3.74e-06,
|
| 1862 |
+
"loss": 0.4517,
|
| 1863 |
+
"step": 6450
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 4.49,
|
| 1867 |
+
"grad_norm": 2.2833831310272217,
|
| 1868 |
+
"learning_rate": 3.713684210526316e-06,
|
| 1869 |
+
"loss": 0.4501,
|
| 1870 |
+
"step": 6475
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 4.5,
|
| 1874 |
+
"grad_norm": 3.27669620513916,
|
| 1875 |
+
"learning_rate": 3.687368421052632e-06,
|
| 1876 |
+
"loss": 0.4596,
|
| 1877 |
+
"step": 6500
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 4.52,
|
| 1881 |
+
"grad_norm": 3.1152820587158203,
|
| 1882 |
+
"learning_rate": 3.6610526315789472e-06,
|
| 1883 |
+
"loss": 0.4577,
|
| 1884 |
+
"step": 6525
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 4.54,
|
| 1888 |
+
"grad_norm": 3.170578718185425,
|
| 1889 |
+
"learning_rate": 3.6347368421052635e-06,
|
| 1890 |
+
"loss": 0.4542,
|
| 1891 |
+
"step": 6550
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 4.56,
|
| 1895 |
+
"grad_norm": 4.5797295570373535,
|
| 1896 |
+
"learning_rate": 3.6084210526315793e-06,
|
| 1897 |
+
"loss": 0.4467,
|
| 1898 |
+
"step": 6575
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 4.57,
|
| 1902 |
+
"grad_norm": 2.258434295654297,
|
| 1903 |
+
"learning_rate": 3.582105263157895e-06,
|
| 1904 |
+
"loss": 0.4511,
|
| 1905 |
+
"step": 6600
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 4.59,
|
| 1909 |
+
"grad_norm": 2.859511137008667,
|
| 1910 |
+
"learning_rate": 3.555789473684211e-06,
|
| 1911 |
+
"loss": 0.449,
|
| 1912 |
+
"step": 6625
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 4.61,
|
| 1916 |
+
"grad_norm": 3.1564037799835205,
|
| 1917 |
+
"learning_rate": 3.5294736842105263e-06,
|
| 1918 |
+
"loss": 0.4488,
|
| 1919 |
+
"step": 6650
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 4.63,
|
| 1923 |
+
"grad_norm": 2.4222614765167236,
|
| 1924 |
+
"learning_rate": 3.503157894736842e-06,
|
| 1925 |
+
"loss": 0.4505,
|
| 1926 |
+
"step": 6675
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 4.64,
|
| 1930 |
+
"grad_norm": 6.269469261169434,
|
| 1931 |
+
"learning_rate": 3.476842105263158e-06,
|
| 1932 |
+
"loss": 0.4512,
|
| 1933 |
+
"step": 6700
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 4.66,
|
| 1937 |
+
"grad_norm": 2.1644115447998047,
|
| 1938 |
+
"learning_rate": 3.450526315789474e-06,
|
| 1939 |
+
"loss": 0.4533,
|
| 1940 |
+
"step": 6725
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 4.68,
|
| 1944 |
+
"grad_norm": 4.101073741912842,
|
| 1945 |
+
"learning_rate": 3.42421052631579e-06,
|
| 1946 |
+
"loss": 0.4538,
|
| 1947 |
+
"step": 6750
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 4.7,
|
| 1951 |
+
"grad_norm": 2.309544563293457,
|
| 1952 |
+
"learning_rate": 3.3978947368421054e-06,
|
| 1953 |
+
"loss": 0.4518,
|
| 1954 |
+
"step": 6775
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 4.71,
|
| 1958 |
+
"grad_norm": 2.7543396949768066,
|
| 1959 |
+
"learning_rate": 3.3715789473684212e-06,
|
| 1960 |
+
"loss": 0.4488,
|
| 1961 |
+
"step": 6800
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 4.73,
|
| 1965 |
+
"grad_norm": 1.9031325578689575,
|
| 1966 |
+
"learning_rate": 3.345263157894737e-06,
|
| 1967 |
+
"loss": 0.4572,
|
| 1968 |
+
"step": 6825
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 4.75,
|
| 1972 |
+
"grad_norm": 1.8639992475509644,
|
| 1973 |
+
"learning_rate": 3.318947368421053e-06,
|
| 1974 |
+
"loss": 0.4547,
|
| 1975 |
+
"step": 6850
|
| 1976 |
+
},
|
| 1977 |
+
{
|
| 1978 |
+
"epoch": 4.76,
|
| 1979 |
+
"grad_norm": 2.6859822273254395,
|
| 1980 |
+
"learning_rate": 3.292631578947369e-06,
|
| 1981 |
+
"loss": 0.4504,
|
| 1982 |
+
"step": 6875
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 4.78,
|
| 1986 |
+
"grad_norm": 2.0838868618011475,
|
| 1987 |
+
"learning_rate": 3.2663157894736845e-06,
|
| 1988 |
+
"loss": 0.4503,
|
| 1989 |
+
"step": 6900
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 4.8,
|
| 1993 |
+
"grad_norm": 1.9140487909317017,
|
| 1994 |
+
"learning_rate": 3.2400000000000003e-06,
|
| 1995 |
+
"loss": 0.4508,
|
| 1996 |
+
"step": 6925
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 4.82,
|
| 2000 |
+
"grad_norm": 2.2218894958496094,
|
| 2001 |
+
"learning_rate": 3.213684210526316e-06,
|
| 2002 |
+
"loss": 0.4542,
|
| 2003 |
+
"step": 6950
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 4.83,
|
| 2007 |
+
"grad_norm": 1.9918988943099976,
|
| 2008 |
+
"learning_rate": 3.187368421052632e-06,
|
| 2009 |
+
"loss": 0.4488,
|
| 2010 |
+
"step": 6975
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 4.85,
|
| 2014 |
+
"grad_norm": 2.9417710304260254,
|
| 2015 |
+
"learning_rate": 3.1610526315789474e-06,
|
| 2016 |
+
"loss": 0.4529,
|
| 2017 |
+
"step": 7000
|
| 2018 |
+
},
|
| 2019 |
+
{
|
| 2020 |
+
"epoch": 4.85,
|
| 2021 |
+
"eval_loss": 0.42716261744499207,
|
| 2022 |
+
"eval_runtime": 125.3767,
|
| 2023 |
+
"eval_samples_per_second": 40.925,
|
| 2024 |
+
"eval_steps_per_second": 20.466,
|
| 2025 |
+
"step": 7000
|
| 2026 |
+
}
|
| 2027 |
+
],
|
| 2028 |
+
"logging_steps": 25,
|
| 2029 |
+
"max_steps": 10000,
|
| 2030 |
+
"num_input_tokens_seen": 0,
|
| 2031 |
+
"num_train_epochs": 7,
|
| 2032 |
+
"save_steps": 1000,
|
| 2033 |
+
"total_flos": 2.5358032943217216e+16,
|
| 2034 |
+
"train_batch_size": 4,
|
| 2035 |
+
"trial_name": null,
|
| 2036 |
+
"trial_params": null
|
| 2037 |
+
}
|
tmp-checkpoint-7000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac83b20df9e27290a55d6c6236d5e8bc328003655e506e4e96e514a11c16a73d
|
| 3 |
+
size 5112
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,215 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"79": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": true,
|
| 38 |
+
"normalized": true,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
},
|
| 43 |
+
"80": {
|
| 44 |
+
"content": "<ctc_blank>",
|
| 45 |
+
"lstrip": false,
|
| 46 |
+
"normalized": true,
|
| 47 |
+
"rstrip": false,
|
| 48 |
+
"single_word": false,
|
| 49 |
+
"special": false
|
| 50 |
+
},
|
| 51 |
+
"81": {
|
| 52 |
+
"content": "à",
|
| 53 |
+
"lstrip": false,
|
| 54 |
+
"normalized": true,
|
| 55 |
+
"rstrip": false,
|
| 56 |
+
"single_word": false,
|
| 57 |
+
"special": false
|
| 58 |
+
},
|
| 59 |
+
"82": {
|
| 60 |
+
"content": "á",
|
| 61 |
+
"lstrip": false,
|
| 62 |
+
"normalized": true,
|
| 63 |
+
"rstrip": false,
|
| 64 |
+
"single_word": false,
|
| 65 |
+
"special": false
|
| 66 |
+
},
|
| 67 |
+
"83": {
|
| 68 |
+
"content": "ã",
|
| 69 |
+
"lstrip": false,
|
| 70 |
+
"normalized": true,
|
| 71 |
+
"rstrip": false,
|
| 72 |
+
"single_word": false,
|
| 73 |
+
"special": false
|
| 74 |
+
},
|
| 75 |
+
"84": {
|
| 76 |
+
"content": "â",
|
| 77 |
+
"lstrip": false,
|
| 78 |
+
"normalized": true,
|
| 79 |
+
"rstrip": false,
|
| 80 |
+
"single_word": false,
|
| 81 |
+
"special": false
|
| 82 |
+
},
|
| 83 |
+
"85": {
|
| 84 |
+
"content": "í",
|
| 85 |
+
"lstrip": false,
|
| 86 |
+
"normalized": true,
|
| 87 |
+
"rstrip": false,
|
| 88 |
+
"single_word": false,
|
| 89 |
+
"special": false
|
| 90 |
+
},
|
| 91 |
+
"86": {
|
| 92 |
+
"content": "ó",
|
| 93 |
+
"lstrip": false,
|
| 94 |
+
"normalized": true,
|
| 95 |
+
"rstrip": false,
|
| 96 |
+
"single_word": false,
|
| 97 |
+
"special": false
|
| 98 |
+
},
|
| 99 |
+
"87": {
|
| 100 |
+
"content": "ô",
|
| 101 |
+
"lstrip": false,
|
| 102 |
+
"normalized": true,
|
| 103 |
+
"rstrip": false,
|
| 104 |
+
"single_word": false,
|
| 105 |
+
"special": false
|
| 106 |
+
},
|
| 107 |
+
"88": {
|
| 108 |
+
"content": "ú",
|
| 109 |
+
"lstrip": false,
|
| 110 |
+
"normalized": true,
|
| 111 |
+
"rstrip": false,
|
| 112 |
+
"single_word": false,
|
| 113 |
+
"special": false
|
| 114 |
+
},
|
| 115 |
+
"89": {
|
| 116 |
+
"content": "À",
|
| 117 |
+
"lstrip": false,
|
| 118 |
+
"normalized": true,
|
| 119 |
+
"rstrip": false,
|
| 120 |
+
"single_word": false,
|
| 121 |
+
"special": false
|
| 122 |
+
},
|
| 123 |
+
"90": {
|
| 124 |
+
"content": "Á",
|
| 125 |
+
"lstrip": false,
|
| 126 |
+
"normalized": true,
|
| 127 |
+
"rstrip": false,
|
| 128 |
+
"single_word": false,
|
| 129 |
+
"special": false
|
| 130 |
+
},
|
| 131 |
+
"91": {
|
| 132 |
+
"content": "Ã",
|
| 133 |
+
"lstrip": false,
|
| 134 |
+
"normalized": true,
|
| 135 |
+
"rstrip": false,
|
| 136 |
+
"single_word": false,
|
| 137 |
+
"special": false
|
| 138 |
+
},
|
| 139 |
+
"92": {
|
| 140 |
+
"content": "É",
|
| 141 |
+
"lstrip": false,
|
| 142 |
+
"normalized": true,
|
| 143 |
+
"rstrip": false,
|
| 144 |
+
"single_word": false,
|
| 145 |
+
"special": false
|
| 146 |
+
},
|
| 147 |
+
"93": {
|
| 148 |
+
"content": "Ê",
|
| 149 |
+
"lstrip": false,
|
| 150 |
+
"normalized": true,
|
| 151 |
+
"rstrip": false,
|
| 152 |
+
"single_word": false,
|
| 153 |
+
"special": false
|
| 154 |
+
},
|
| 155 |
+
"94": {
|
| 156 |
+
"content": "Í",
|
| 157 |
+
"lstrip": false,
|
| 158 |
+
"normalized": true,
|
| 159 |
+
"rstrip": false,
|
| 160 |
+
"single_word": false,
|
| 161 |
+
"special": false
|
| 162 |
+
},
|
| 163 |
+
"95": {
|
| 164 |
+
"content": "Ó",
|
| 165 |
+
"lstrip": false,
|
| 166 |
+
"normalized": true,
|
| 167 |
+
"rstrip": false,
|
| 168 |
+
"single_word": false,
|
| 169 |
+
"special": false
|
| 170 |
+
},
|
| 171 |
+
"96": {
|
| 172 |
+
"content": "Ô",
|
| 173 |
+
"lstrip": false,
|
| 174 |
+
"normalized": true,
|
| 175 |
+
"rstrip": false,
|
| 176 |
+
"single_word": false,
|
| 177 |
+
"special": false
|
| 178 |
+
},
|
| 179 |
+
"97": {
|
| 180 |
+
"content": "Ú",
|
| 181 |
+
"lstrip": false,
|
| 182 |
+
"normalized": true,
|
| 183 |
+
"rstrip": false,
|
| 184 |
+
"single_word": false,
|
| 185 |
+
"special": false
|
| 186 |
+
},
|
| 187 |
+
"98": {
|
| 188 |
+
"content": "ç",
|
| 189 |
+
"lstrip": false,
|
| 190 |
+
"normalized": true,
|
| 191 |
+
"rstrip": false,
|
| 192 |
+
"single_word": false,
|
| 193 |
+
"special": false
|
| 194 |
+
},
|
| 195 |
+
"99": {
|
| 196 |
+
"content": "Ç",
|
| 197 |
+
"lstrip": false,
|
| 198 |
+
"normalized": true,
|
| 199 |
+
"rstrip": false,
|
| 200 |
+
"single_word": false,
|
| 201 |
+
"special": false
|
| 202 |
+
}
|
| 203 |
+
},
|
| 204 |
+
"bos_token": "<s>",
|
| 205 |
+
"clean_up_tokenization_spaces": true,
|
| 206 |
+
"eos_token": "</s>",
|
| 207 |
+
"mask_token": "<mask>",
|
| 208 |
+
"model_max_length": 600,
|
| 209 |
+
"normalize": false,
|
| 210 |
+
"pad_token": "<pad>",
|
| 211 |
+
"processor_class": "SpeechT5Processor",
|
| 212 |
+
"sp_model_kwargs": {},
|
| 213 |
+
"tokenizer_class": "SpeechT5Tokenizer",
|
| 214 |
+
"unk_token": "<unk>"
|
| 215 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0013f387492a9d599f92d36e33fa5a88cabd87184b8111202bcd7971ff8a3ada
|
| 3 |
+
size 5112
|