b-atmaja
commited on
Commit
•
e55f38f
1
Parent(s):
7b51472
new model from indonesian-nlp
Browse files- all_results.json +14 -0
- config.json +13 -10
- eval_results.json +9 -0
- preprocessor_config.json +1 -0
- special_tokens_map.json +1 -0
- tokenizer_config.json +1 -0
- train_results.json +8 -0
- trainer_state.json +1135 -0
- vocab.json +1 -1
all_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 60.0,
|
| 3 |
+
"eval_loss": 0.3271900713443756,
|
| 4 |
+
"eval_runtime": 107.3558,
|
| 5 |
+
"eval_samples": 1844,
|
| 6 |
+
"eval_samples_per_second": 17.177,
|
| 7 |
+
"eval_steps_per_second": 2.152,
|
| 8 |
+
"eval_wer": 0.3851389147865001,
|
| 9 |
+
"train_loss": 0.8272683016715511,
|
| 10 |
+
"train_runtime": 19767.0967,
|
| 11 |
+
"train_samples": 3965,
|
| 12 |
+
"train_samples_per_second": 12.035,
|
| 13 |
+
"train_steps_per_second": 0.376
|
| 14 |
+
}
|
config.json
CHANGED
|
@@ -1,12 +1,13 @@
|
|
| 1 |
{
|
| 2 |
"_name_or_path": "facebook/wav2vec2-large-xlsr-53",
|
| 3 |
-
"activation_dropout": 0.
|
| 4 |
"apply_spec_augment": true,
|
| 5 |
"architectures": [
|
| 6 |
"Wav2Vec2ForCTC"
|
| 7 |
],
|
| 8 |
-
"attention_dropout": 0.
|
| 9 |
"bos_token_id": 1,
|
|
|
|
| 10 |
"codevector_dim": 768,
|
| 11 |
"contrastive_logits_temperature": 0.1,
|
| 12 |
"conv_bias": true,
|
|
@@ -38,24 +39,24 @@
|
|
| 38 |
2
|
| 39 |
],
|
| 40 |
"ctc_loss_reduction": "mean",
|
| 41 |
-
"ctc_zero_infinity":
|
| 42 |
"diversity_loss_weight": 0.1,
|
| 43 |
"do_stable_layer_norm": true,
|
| 44 |
"eos_token_id": 2,
|
| 45 |
"feat_extract_activation": "gelu",
|
| 46 |
"feat_extract_dropout": 0.0,
|
| 47 |
"feat_extract_norm": "layer",
|
| 48 |
-
"feat_proj_dropout": 0.
|
| 49 |
"feat_quantizer_dropout": 0.0,
|
| 50 |
"final_dropout": 0.0,
|
| 51 |
"gradient_checkpointing": true,
|
| 52 |
"hidden_act": "gelu",
|
| 53 |
-
"hidden_dropout": 0.
|
| 54 |
"hidden_size": 1024,
|
| 55 |
"initializer_range": 0.02,
|
| 56 |
"intermediate_size": 4096,
|
| 57 |
"layer_norm_eps": 1e-05,
|
| 58 |
-
"layerdrop": 0.
|
| 59 |
"mask_channel_length": 10,
|
| 60 |
"mask_channel_min_space": 1,
|
| 61 |
"mask_channel_other": 0.0,
|
|
@@ -66,7 +67,7 @@
|
|
| 66 |
"mask_time_length": 10,
|
| 67 |
"mask_time_min_space": 1,
|
| 68 |
"mask_time_other": 0.0,
|
| 69 |
-
"mask_time_prob": 0.
|
| 70 |
"mask_time_selection": "static",
|
| 71 |
"model_type": "wav2vec2",
|
| 72 |
"num_attention_heads": 16,
|
|
@@ -77,8 +78,10 @@
|
|
| 77 |
"num_feat_extract_layers": 7,
|
| 78 |
"num_hidden_layers": 24,
|
| 79 |
"num_negatives": 100,
|
| 80 |
-
"pad_token_id":
|
| 81 |
"proj_codevector_dim": 768,
|
| 82 |
-
"
|
| 83 |
-
"
|
|
|
|
|
|
|
| 84 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"_name_or_path": "facebook/wav2vec2-large-xlsr-53",
|
| 3 |
+
"activation_dropout": 0.055,
|
| 4 |
"apply_spec_augment": true,
|
| 5 |
"architectures": [
|
| 6 |
"Wav2Vec2ForCTC"
|
| 7 |
],
|
| 8 |
+
"attention_dropout": 0.094,
|
| 9 |
"bos_token_id": 1,
|
| 10 |
+
"classifier_proj_size": 256,
|
| 11 |
"codevector_dim": 768,
|
| 12 |
"contrastive_logits_temperature": 0.1,
|
| 13 |
"conv_bias": true,
|
|
|
|
| 39 |
2
|
| 40 |
],
|
| 41 |
"ctc_loss_reduction": "mean",
|
| 42 |
+
"ctc_zero_infinity": true,
|
| 43 |
"diversity_loss_weight": 0.1,
|
| 44 |
"do_stable_layer_norm": true,
|
| 45 |
"eos_token_id": 2,
|
| 46 |
"feat_extract_activation": "gelu",
|
| 47 |
"feat_extract_dropout": 0.0,
|
| 48 |
"feat_extract_norm": "layer",
|
| 49 |
+
"feat_proj_dropout": 0.04,
|
| 50 |
"feat_quantizer_dropout": 0.0,
|
| 51 |
"final_dropout": 0.0,
|
| 52 |
"gradient_checkpointing": true,
|
| 53 |
"hidden_act": "gelu",
|
| 54 |
+
"hidden_dropout": 0.047,
|
| 55 |
"hidden_size": 1024,
|
| 56 |
"initializer_range": 0.02,
|
| 57 |
"intermediate_size": 4096,
|
| 58 |
"layer_norm_eps": 1e-05,
|
| 59 |
+
"layerdrop": 0.041,
|
| 60 |
"mask_channel_length": 10,
|
| 61 |
"mask_channel_min_space": 1,
|
| 62 |
"mask_channel_other": 0.0,
|
|
|
|
| 67 |
"mask_time_length": 10,
|
| 68 |
"mask_time_min_space": 1,
|
| 69 |
"mask_time_other": 0.0,
|
| 70 |
+
"mask_time_prob": 0.4,
|
| 71 |
"mask_time_selection": "static",
|
| 72 |
"model_type": "wav2vec2",
|
| 73 |
"num_attention_heads": 16,
|
|
|
|
| 78 |
"num_feat_extract_layers": 7,
|
| 79 |
"num_hidden_layers": 24,
|
| 80 |
"num_negatives": 100,
|
| 81 |
+
"pad_token_id": 30,
|
| 82 |
"proj_codevector_dim": 768,
|
| 83 |
+
"torch_dtype": "float32",
|
| 84 |
+
"transformers_version": "4.10.3",
|
| 85 |
+
"use_weighted_layer_sum": false,
|
| 86 |
+
"vocab_size": 31
|
| 87 |
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 60.0,
|
| 3 |
+
"eval_loss": 0.3271900713443756,
|
| 4 |
+
"eval_runtime": 107.3558,
|
| 5 |
+
"eval_samples": 1844,
|
| 6 |
+
"eval_samples_per_second": 17.177,
|
| 7 |
+
"eval_steps_per_second": 2.152,
|
| 8 |
+
"eval_wer": 0.3851389147865001
|
| 9 |
+
}
|
preprocessor_config.json
CHANGED
|
@@ -1,5 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"do_normalize": true,
|
|
|
|
| 3 |
"feature_size": 1,
|
| 4 |
"padding_side": "right",
|
| 5 |
"padding_value": 0.0,
|
|
|
|
| 1 |
{
|
| 2 |
"do_normalize": true,
|
| 3 |
+
"feature_extractor_type": "Wav2Vec2FeatureExtractor",
|
| 4 |
"feature_size": 1,
|
| 5 |
"padding_side": "right",
|
| 6 |
"padding_value": 0.0,
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "[UNK]", "pad_token": "[PAD]"}
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"unk_token": "[UNK]", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "[PAD]", "do_lower_case": false, "word_delimiter_token": "|", "tokenizer_class": "Wav2Vec2CTCTokenizer"}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 60.0,
|
| 3 |
+
"train_loss": 0.8272683016715511,
|
| 4 |
+
"train_runtime": 19767.0967,
|
| 5 |
+
"train_samples": 3965,
|
| 6 |
+
"train_samples_per_second": 12.035,
|
| 7 |
+
"train_steps_per_second": 0.376
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1135 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 60.0,
|
| 5 |
+
"global_step": 7440,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.81,
|
| 12 |
+
"learning_rate": 1.3172043010752688e-05,
|
| 13 |
+
"loss": 13.9663,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.81,
|
| 18 |
+
"eval_loss": 15.575762748718262,
|
| 19 |
+
"eval_runtime": 98.9721,
|
| 20 |
+
"eval_samples_per_second": 18.632,
|
| 21 |
+
"eval_steps_per_second": 2.334,
|
| 22 |
+
"eval_wer": 1.0,
|
| 23 |
+
"step": 100
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 1.61,
|
| 27 |
+
"learning_rate": 2.6344086021505376e-05,
|
| 28 |
+
"loss": 6.7623,
|
| 29 |
+
"step": 200
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 1.61,
|
| 33 |
+
"eval_loss": 3.259247064590454,
|
| 34 |
+
"eval_runtime": 99.4059,
|
| 35 |
+
"eval_samples_per_second": 18.55,
|
| 36 |
+
"eval_steps_per_second": 2.324,
|
| 37 |
+
"eval_wer": 1.0,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 2.42,
|
| 42 |
+
"learning_rate": 3.978494623655914e-05,
|
| 43 |
+
"loss": 3.0242,
|
| 44 |
+
"step": 300
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 2.42,
|
| 48 |
+
"eval_loss": 2.9369125366210938,
|
| 49 |
+
"eval_runtime": 98.1481,
|
| 50 |
+
"eval_samples_per_second": 18.788,
|
| 51 |
+
"eval_steps_per_second": 2.354,
|
| 52 |
+
"eval_wer": 1.0,
|
| 53 |
+
"step": 300
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 3.23,
|
| 57 |
+
"learning_rate": 5.32258064516129e-05,
|
| 58 |
+
"loss": 2.8806,
|
| 59 |
+
"step": 400
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 3.23,
|
| 63 |
+
"eval_loss": 2.8887438774108887,
|
| 64 |
+
"eval_runtime": 97.5774,
|
| 65 |
+
"eval_samples_per_second": 18.898,
|
| 66 |
+
"eval_steps_per_second": 2.367,
|
| 67 |
+
"eval_wer": 1.0,
|
| 68 |
+
"step": 400
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 4.03,
|
| 72 |
+
"learning_rate": 6.666666666666667e-05,
|
| 73 |
+
"loss": 2.8344,
|
| 74 |
+
"step": 500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 4.03,
|
| 78 |
+
"eval_loss": 2.8125975131988525,
|
| 79 |
+
"eval_runtime": 97.5647,
|
| 80 |
+
"eval_samples_per_second": 18.9,
|
| 81 |
+
"eval_steps_per_second": 2.368,
|
| 82 |
+
"eval_wer": 1.0,
|
| 83 |
+
"step": 500
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 4.84,
|
| 87 |
+
"learning_rate": 8.010752688172043e-05,
|
| 88 |
+
"loss": 2.7158,
|
| 89 |
+
"step": 600
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 4.84,
|
| 93 |
+
"eval_loss": 2.304689884185791,
|
| 94 |
+
"eval_runtime": 97.2319,
|
| 95 |
+
"eval_samples_per_second": 18.965,
|
| 96 |
+
"eval_steps_per_second": 2.376,
|
| 97 |
+
"eval_wer": 1.003542793212754,
|
| 98 |
+
"step": 600
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 5.65,
|
| 102 |
+
"learning_rate": 9.35483870967742e-05,
|
| 103 |
+
"loss": 1.7164,
|
| 104 |
+
"step": 700
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 5.65,
|
| 108 |
+
"eval_loss": 0.7104876041412354,
|
| 109 |
+
"eval_runtime": 96.4932,
|
| 110 |
+
"eval_samples_per_second": 19.11,
|
| 111 |
+
"eval_steps_per_second": 2.394,
|
| 112 |
+
"eval_wer": 0.7594629871340668,
|
| 113 |
+
"step": 700
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 6.45,
|
| 117 |
+
"learning_rate": 0.0001,
|
| 118 |
+
"loss": 1.0757,
|
| 119 |
+
"step": 800
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 6.45,
|
| 123 |
+
"eval_loss": 0.5160450339317322,
|
| 124 |
+
"eval_runtime": 97.694,
|
| 125 |
+
"eval_samples_per_second": 18.875,
|
| 126 |
+
"eval_steps_per_second": 2.365,
|
| 127 |
+
"eval_wer": 0.6501025545403692,
|
| 128 |
+
"step": 800
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 7.26,
|
| 132 |
+
"learning_rate": 0.0001,
|
| 133 |
+
"loss": 0.9208,
|
| 134 |
+
"step": 900
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 7.26,
|
| 138 |
+
"eval_loss": 0.43188726902008057,
|
| 139 |
+
"eval_runtime": 101.9111,
|
| 140 |
+
"eval_samples_per_second": 18.094,
|
| 141 |
+
"eval_steps_per_second": 2.267,
|
| 142 |
+
"eval_wer": 0.5842811859034123,
|
| 143 |
+
"step": 900
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 8.06,
|
| 147 |
+
"learning_rate": 0.0001,
|
| 148 |
+
"loss": 0.8012,
|
| 149 |
+
"step": 1000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 8.06,
|
| 153 |
+
"eval_loss": 0.39437660574913025,
|
| 154 |
+
"eval_runtime": 97.2792,
|
| 155 |
+
"eval_samples_per_second": 18.956,
|
| 156 |
+
"eval_steps_per_second": 2.375,
|
| 157 |
+
"eval_wer": 0.5456833861644602,
|
| 158 |
+
"step": 1000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 8.87,
|
| 162 |
+
"learning_rate": 0.0001,
|
| 163 |
+
"loss": 0.7334,
|
| 164 |
+
"step": 1100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 8.87,
|
| 168 |
+
"eval_loss": 0.36806735396385193,
|
| 169 |
+
"eval_runtime": 97.7448,
|
| 170 |
+
"eval_samples_per_second": 18.865,
|
| 171 |
+
"eval_steps_per_second": 2.363,
|
| 172 |
+
"eval_wer": 0.5120268506432967,
|
| 173 |
+
"step": 1100
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 9.68,
|
| 177 |
+
"learning_rate": 0.0001,
|
| 178 |
+
"loss": 0.6839,
|
| 179 |
+
"step": 1200
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 9.68,
|
| 183 |
+
"eval_loss": 0.35295596718788147,
|
| 184 |
+
"eval_runtime": 104.7228,
|
| 185 |
+
"eval_samples_per_second": 17.608,
|
| 186 |
+
"eval_steps_per_second": 2.206,
|
| 187 |
+
"eval_wer": 0.49822860339362296,
|
| 188 |
+
"step": 1200
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 10.48,
|
| 192 |
+
"learning_rate": 0.0001,
|
| 193 |
+
"loss": 0.6599,
|
| 194 |
+
"step": 1300
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 10.48,
|
| 198 |
+
"eval_loss": 0.34405317902565,
|
| 199 |
+
"eval_runtime": 100.6994,
|
| 200 |
+
"eval_samples_per_second": 18.312,
|
| 201 |
+
"eval_steps_per_second": 2.294,
|
| 202 |
+
"eval_wer": 0.49561812418422524,
|
| 203 |
+
"step": 1300
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 11.29,
|
| 207 |
+
"learning_rate": 0.0001,
|
| 208 |
+
"loss": 0.6101,
|
| 209 |
+
"step": 1400
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 11.29,
|
| 213 |
+
"eval_loss": 0.3226686716079712,
|
| 214 |
+
"eval_runtime": 97.8943,
|
| 215 |
+
"eval_samples_per_second": 18.837,
|
| 216 |
+
"eval_steps_per_second": 2.36,
|
| 217 |
+
"eval_wer": 0.4828454223382435,
|
| 218 |
+
"step": 1400
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 12.1,
|
| 222 |
+
"learning_rate": 0.0001,
|
| 223 |
+
"loss": 0.5844,
|
| 224 |
+
"step": 1500
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 12.1,
|
| 228 |
+
"eval_loss": 0.34125906229019165,
|
| 229 |
+
"eval_runtime": 98.6933,
|
| 230 |
+
"eval_samples_per_second": 18.684,
|
| 231 |
+
"eval_steps_per_second": 2.341,
|
| 232 |
+
"eval_wer": 0.47445459630803655,
|
| 233 |
+
"step": 1500
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 12.9,
|
| 237 |
+
"learning_rate": 0.0001,
|
| 238 |
+
"loss": 0.5754,
|
| 239 |
+
"step": 1600
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 12.9,
|
| 243 |
+
"eval_loss": 0.33336642384529114,
|
| 244 |
+
"eval_runtime": 99.0007,
|
| 245 |
+
"eval_samples_per_second": 18.626,
|
| 246 |
+
"eval_steps_per_second": 2.333,
|
| 247 |
+
"eval_wer": 0.46140220026104795,
|
| 248 |
+
"step": 1600
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 13.71,
|
| 252 |
+
"learning_rate": 0.0001,
|
| 253 |
+
"loss": 0.5611,
|
| 254 |
+
"step": 1700
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 13.71,
|
| 258 |
+
"eval_loss": 0.30597391724586487,
|
| 259 |
+
"eval_runtime": 101.2827,
|
| 260 |
+
"eval_samples_per_second": 18.206,
|
| 261 |
+
"eval_steps_per_second": 2.281,
|
| 262 |
+
"eval_wer": 0.4672757784821928,
|
| 263 |
+
"step": 1700
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 14.52,
|
| 267 |
+
"learning_rate": 0.0001,
|
| 268 |
+
"loss": 0.5476,
|
| 269 |
+
"step": 1800
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 14.52,
|
| 273 |
+
"eval_loss": 0.3194342851638794,
|
| 274 |
+
"eval_runtime": 105.7302,
|
| 275 |
+
"eval_samples_per_second": 17.441,
|
| 276 |
+
"eval_steps_per_second": 2.185,
|
| 277 |
+
"eval_wer": 0.44881596121573747,
|
| 278 |
+
"step": 1800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 15.32,
|
| 282 |
+
"learning_rate": 0.0001,
|
| 283 |
+
"loss": 0.526,
|
| 284 |
+
"step": 1900
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 15.32,
|
| 288 |
+
"eval_loss": 0.30019786953926086,
|
| 289 |
+
"eval_runtime": 96.396,
|
| 290 |
+
"eval_samples_per_second": 19.129,
|
| 291 |
+
"eval_steps_per_second": 2.396,
|
| 292 |
+
"eval_wer": 0.4449934738019765,
|
| 293 |
+
"step": 1900
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 16.13,
|
| 297 |
+
"learning_rate": 0.0001,
|
| 298 |
+
"loss": 0.5267,
|
| 299 |
+
"step": 2000
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 16.13,
|
| 303 |
+
"eval_loss": 0.3249629735946655,
|
| 304 |
+
"eval_runtime": 105.1317,
|
| 305 |
+
"eval_samples_per_second": 17.54,
|
| 306 |
+
"eval_steps_per_second": 2.197,
|
| 307 |
+
"eval_wer": 0.43660264777176955,
|
| 308 |
+
"step": 2000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 16.94,
|
| 312 |
+
"learning_rate": 0.0001,
|
| 313 |
+
"loss": 0.49,
|
| 314 |
+
"step": 2100
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 16.94,
|
| 318 |
+
"eval_loss": 0.2999746799468994,
|
| 319 |
+
"eval_runtime": 97.027,
|
| 320 |
+
"eval_samples_per_second": 19.005,
|
| 321 |
+
"eval_steps_per_second": 2.381,
|
| 322 |
+
"eval_wer": 0.4337124743613649,
|
| 323 |
+
"step": 2100
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 17.74,
|
| 327 |
+
"learning_rate": 0.0001,
|
| 328 |
+
"loss": 0.4945,
|
| 329 |
+
"step": 2200
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 17.74,
|
| 333 |
+
"eval_loss": 0.30671271681785583,
|
| 334 |
+
"eval_runtime": 96.8855,
|
| 335 |
+
"eval_samples_per_second": 19.033,
|
| 336 |
+
"eval_steps_per_second": 2.384,
|
| 337 |
+
"eval_wer": 0.43100876375163155,
|
| 338 |
+
"step": 2200
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 18.55,
|
| 342 |
+
"learning_rate": 0.0001,
|
| 343 |
+
"loss": 0.485,
|
| 344 |
+
"step": 2300
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 18.55,
|
| 348 |
+
"eval_loss": 0.3134038746356964,
|
| 349 |
+
"eval_runtime": 98.1951,
|
| 350 |
+
"eval_samples_per_second": 18.779,
|
| 351 |
+
"eval_steps_per_second": 2.352,
|
| 352 |
+
"eval_wer": 0.43706880477344767,
|
| 353 |
+
"step": 2300
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 19.35,
|
| 357 |
+
"learning_rate": 0.0001,
|
| 358 |
+
"loss": 0.486,
|
| 359 |
+
"step": 2400
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 19.35,
|
| 363 |
+
"eval_loss": 0.3182927072048187,
|
| 364 |
+
"eval_runtime": 100.6349,
|
| 365 |
+
"eval_samples_per_second": 18.324,
|
| 366 |
+
"eval_steps_per_second": 2.295,
|
| 367 |
+
"eval_wer": 0.43044937534961775,
|
| 368 |
+
"step": 2400
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 20.16,
|
| 372 |
+
"learning_rate": 0.0001,
|
| 373 |
+
"loss": 0.4617,
|
| 374 |
+
"step": 2500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 20.16,
|
| 378 |
+
"eval_loss": 0.3004864752292633,
|
| 379 |
+
"eval_runtime": 102.1006,
|
| 380 |
+
"eval_samples_per_second": 18.061,
|
| 381 |
+
"eval_steps_per_second": 2.262,
|
| 382 |
+
"eval_wer": 0.4238299459257878,
|
| 383 |
+
"step": 2500
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 20.97,
|
| 387 |
+
"learning_rate": 0.0001,
|
| 388 |
+
"loss": 0.465,
|
| 389 |
+
"step": 2600
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 20.97,
|
| 393 |
+
"eval_loss": 0.313532292842865,
|
| 394 |
+
"eval_runtime": 101.2873,
|
| 395 |
+
"eval_samples_per_second": 18.206,
|
| 396 |
+
"eval_steps_per_second": 2.281,
|
| 397 |
+
"eval_wer": 0.4311019951519672,
|
| 398 |
+
"step": 2600
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 21.77,
|
| 402 |
+
"learning_rate": 0.0001,
|
| 403 |
+
"loss": 0.4414,
|
| 404 |
+
"step": 2700
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 21.77,
|
| 408 |
+
"eval_loss": 0.3118414580821991,
|
| 409 |
+
"eval_runtime": 97.4069,
|
| 410 |
+
"eval_samples_per_second": 18.931,
|
| 411 |
+
"eval_steps_per_second": 2.371,
|
| 412 |
+
"eval_wer": 0.4256013425321648,
|
| 413 |
+
"step": 2700
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 22.58,
|
| 417 |
+
"learning_rate": 0.0001,
|
| 418 |
+
"loss": 0.4424,
|
| 419 |
+
"step": 2800
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 22.58,
|
| 423 |
+
"eval_loss": 0.3086116313934326,
|
| 424 |
+
"eval_runtime": 97.4809,
|
| 425 |
+
"eval_samples_per_second": 18.917,
|
| 426 |
+
"eval_steps_per_second": 2.37,
|
| 427 |
+
"eval_wer": 0.4160917396979303,
|
| 428 |
+
"step": 2800
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 23.39,
|
| 432 |
+
"learning_rate": 0.0001,
|
| 433 |
+
"loss": 0.4352,
|
| 434 |
+
"step": 2900
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 23.39,
|
| 438 |
+
"eval_loss": 0.3213385045528412,
|
| 439 |
+
"eval_runtime": 96.5991,
|
| 440 |
+
"eval_samples_per_second": 19.089,
|
| 441 |
+
"eval_steps_per_second": 2.391,
|
| 442 |
+
"eval_wer": 0.41879545030766363,
|
| 443 |
+
"step": 2900
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 24.19,
|
| 447 |
+
"learning_rate": 0.0001,
|
| 448 |
+
"loss": 0.4324,
|
| 449 |
+
"step": 3000
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 24.19,
|
| 453 |
+
"eval_loss": 0.3011772036552429,
|
| 454 |
+
"eval_runtime": 97.1915,
|
| 455 |
+
"eval_samples_per_second": 18.973,
|
| 456 |
+
"eval_steps_per_second": 2.377,
|
| 457 |
+
"eval_wer": 0.4236434831251165,
|
| 458 |
+
"step": 3000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 25.0,
|
| 462 |
+
"learning_rate": 0.0001,
|
| 463 |
+
"loss": 0.4246,
|
| 464 |
+
"step": 3100
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 25.0,
|
| 468 |
+
"eval_loss": 0.30194172263145447,
|
| 469 |
+
"eval_runtime": 100.0292,
|
| 470 |
+
"eval_samples_per_second": 18.435,
|
| 471 |
+
"eval_steps_per_second": 2.309,
|
| 472 |
+
"eval_wer": 0.4165578966996084,
|
| 473 |
+
"step": 3100
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 25.81,
|
| 477 |
+
"learning_rate": 0.0001,
|
| 478 |
+
"loss": 0.4132,
|
| 479 |
+
"step": 3200
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 25.81,
|
| 483 |
+
"eval_loss": 0.299437552690506,
|
| 484 |
+
"eval_runtime": 97.7952,
|
| 485 |
+
"eval_samples_per_second": 18.856,
|
| 486 |
+
"eval_steps_per_second": 2.362,
|
| 487 |
+
"eval_wer": 0.4160917396979303,
|
| 488 |
+
"step": 3200
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 26.61,
|
| 492 |
+
"learning_rate": 0.0001,
|
| 493 |
+
"loss": 0.4074,
|
| 494 |
+
"step": 3300
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 26.61,
|
| 498 |
+
"eval_loss": 0.31787610054016113,
|
| 499 |
+
"eval_runtime": 97.3925,
|
| 500 |
+
"eval_samples_per_second": 18.934,
|
| 501 |
+
"eval_steps_per_second": 2.372,
|
| 502 |
+
"eval_wer": 0.41795636770464295,
|
| 503 |
+
"step": 3300
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 27.42,
|
| 507 |
+
"learning_rate": 0.0001,
|
| 508 |
+
"loss": 0.4031,
|
| 509 |
+
"step": 3400
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 27.42,
|
| 513 |
+
"eval_loss": 0.296748548746109,
|
| 514 |
+
"eval_runtime": 101.7762,
|
| 515 |
+
"eval_samples_per_second": 18.118,
|
| 516 |
+
"eval_steps_per_second": 2.27,
|
| 517 |
+
"eval_wer": 0.40760768226738764,
|
| 518 |
+
"step": 3400
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 28.23,
|
| 522 |
+
"learning_rate": 0.0001,
|
| 523 |
+
"loss": 0.4082,
|
| 524 |
+
"step": 3500
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 28.23,
|
| 528 |
+
"eval_loss": 0.31329259276390076,
|
| 529 |
+
"eval_runtime": 102.7944,
|
| 530 |
+
"eval_samples_per_second": 17.939,
|
| 531 |
+
"eval_steps_per_second": 2.247,
|
| 532 |
+
"eval_wer": 0.41310833488719,
|
| 533 |
+
"step": 3500
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 29.03,
|
| 537 |
+
"learning_rate": 0.0001,
|
| 538 |
+
"loss": 0.3912,
|
| 539 |
+
"step": 3600
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 29.03,
|
| 543 |
+
"eval_loss": 0.3259478509426117,
|
| 544 |
+
"eval_runtime": 98.5616,
|
| 545 |
+
"eval_samples_per_second": 18.709,
|
| 546 |
+
"eval_steps_per_second": 2.344,
|
| 547 |
+
"eval_wer": 0.41133693828081297,
|
| 548 |
+
"step": 3600
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 29.84,
|
| 552 |
+
"learning_rate": 0.0001,
|
| 553 |
+
"loss": 0.3922,
|
| 554 |
+
"step": 3700
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 29.84,
|
| 558 |
+
"eval_loss": 0.32406386733055115,
|
| 559 |
+
"eval_runtime": 96.8005,
|
| 560 |
+
"eval_samples_per_second": 19.049,
|
| 561 |
+
"eval_steps_per_second": 2.386,
|
| 562 |
+
"eval_wer": 0.41226925228416933,
|
| 563 |
+
"step": 3700
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 30.65,
|
| 567 |
+
"learning_rate": 9.795698924731184e-05,
|
| 568 |
+
"loss": 0.3851,
|
| 569 |
+
"step": 3800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 30.65,
|
| 573 |
+
"eval_loss": 0.29852330684661865,
|
| 574 |
+
"eval_runtime": 105.0804,
|
| 575 |
+
"eval_samples_per_second": 17.548,
|
| 576 |
+
"eval_steps_per_second": 2.198,
|
| 577 |
+
"eval_wer": 0.41133693828081297,
|
| 578 |
+
"step": 3800
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 31.45,
|
| 582 |
+
"learning_rate": 9.526881720430108e-05,
|
| 583 |
+
"loss": 0.3768,
|
| 584 |
+
"step": 3900
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 31.45,
|
| 588 |
+
"eval_loss": 0.3223101794719696,
|
| 589 |
+
"eval_runtime": 97.6458,
|
| 590 |
+
"eval_samples_per_second": 18.885,
|
| 591 |
+
"eval_steps_per_second": 2.366,
|
| 592 |
+
"eval_wer": 0.41040462427745666,
|
| 593 |
+
"step": 3900
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 32.26,
|
| 597 |
+
"learning_rate": 9.258064516129033e-05,
|
| 598 |
+
"loss": 0.3728,
|
| 599 |
+
"step": 4000
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 32.26,
|
| 603 |
+
"eval_loss": 0.3212699890136719,
|
| 604 |
+
"eval_runtime": 100.4336,
|
| 605 |
+
"eval_samples_per_second": 18.36,
|
| 606 |
+
"eval_steps_per_second": 2.3,
|
| 607 |
+
"eval_wer": 0.40770091366772326,
|
| 608 |
+
"step": 4000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 33.06,
|
| 612 |
+
"learning_rate": 8.989247311827958e-05,
|
| 613 |
+
"loss": 0.3646,
|
| 614 |
+
"step": 4100
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 33.06,
|
| 618 |
+
"eval_loss": 0.31770506501197815,
|
| 619 |
+
"eval_runtime": 97.2462,
|
| 620 |
+
"eval_samples_per_second": 18.962,
|
| 621 |
+
"eval_steps_per_second": 2.375,
|
| 622 |
+
"eval_wer": 0.4116166324818199,
|
| 623 |
+
"step": 4100
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 33.87,
|
| 627 |
+
"learning_rate": 8.720430107526883e-05,
|
| 628 |
+
"loss": 0.3681,
|
| 629 |
+
"step": 4200
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 33.87,
|
| 633 |
+
"eval_loss": 0.3010982871055603,
|
| 634 |
+
"eval_runtime": 99.148,
|
| 635 |
+
"eval_samples_per_second": 18.598,
|
| 636 |
+
"eval_steps_per_second": 2.33,
|
| 637 |
+
"eval_wer": 0.4105910870781279,
|
| 638 |
+
"step": 4200
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 34.68,
|
| 642 |
+
"learning_rate": 8.451612903225808e-05,
|
| 643 |
+
"loss": 0.3634,
|
| 644 |
+
"step": 4300
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 34.68,
|
| 648 |
+
"eval_loss": 0.3075733780860901,
|
| 649 |
+
"eval_runtime": 99.8799,
|
| 650 |
+
"eval_samples_per_second": 18.462,
|
| 651 |
+
"eval_steps_per_second": 2.313,
|
| 652 |
+
"eval_wer": 0.4069550624650382,
|
| 653 |
+
"step": 4300
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 35.48,
|
| 657 |
+
"learning_rate": 8.182795698924732e-05,
|
| 658 |
+
"loss": 0.3471,
|
| 659 |
+
"step": 4400
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 35.48,
|
| 663 |
+
"eval_loss": 0.31304171681404114,
|
| 664 |
+
"eval_runtime": 99.6768,
|
| 665 |
+
"eval_samples_per_second": 18.5,
|
| 666 |
+
"eval_steps_per_second": 2.317,
|
| 667 |
+
"eval_wer": 0.40266641804959913,
|
| 668 |
+
"step": 4400
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 36.29,
|
| 672 |
+
"learning_rate": 7.913978494623657e-05,
|
| 673 |
+
"loss": 0.3398,
|
| 674 |
+
"step": 4500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 36.29,
|
| 678 |
+
"eval_loss": 0.306076318025589,
|
| 679 |
+
"eval_runtime": 103.5936,
|
| 680 |
+
"eval_samples_per_second": 17.8,
|
| 681 |
+
"eval_steps_per_second": 2.23,
|
| 682 |
+
"eval_wer": 0.4037851948536267,
|
| 683 |
+
"step": 4500
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 37.1,
|
| 687 |
+
"learning_rate": 7.645161290322582e-05,
|
| 688 |
+
"loss": 0.3394,
|
| 689 |
+
"step": 4600
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 37.1,
|
| 693 |
+
"eval_loss": 0.31049424409866333,
|
| 694 |
+
"eval_runtime": 98.5424,
|
| 695 |
+
"eval_samples_per_second": 18.713,
|
| 696 |
+
"eval_steps_per_second": 2.344,
|
| 697 |
+
"eval_wer": 0.4037851948536267,
|
| 698 |
+
"step": 4600
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 37.9,
|
| 702 |
+
"learning_rate": 7.376344086021507e-05,
|
| 703 |
+
"loss": 0.331,
|
| 704 |
+
"step": 4700
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 37.9,
|
| 708 |
+
"eval_loss": 0.3191221058368683,
|
| 709 |
+
"eval_runtime": 102.0886,
|
| 710 |
+
"eval_samples_per_second": 18.063,
|
| 711 |
+
"eval_steps_per_second": 2.263,
|
| 712 |
+
"eval_wer": 0.3992168562371807,
|
| 713 |
+
"step": 4700
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 38.71,
|
| 717 |
+
"learning_rate": 7.10752688172043e-05,
|
| 718 |
+
"loss": 0.3221,
|
| 719 |
+
"step": 4800
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 38.71,
|
| 723 |
+
"eval_loss": 0.3119024336338043,
|
| 724 |
+
"eval_runtime": 99.689,
|
| 725 |
+
"eval_samples_per_second": 18.498,
|
| 726 |
+
"eval_steps_per_second": 2.317,
|
| 727 |
+
"eval_wer": 0.39958978183852323,
|
| 728 |
+
"step": 4800
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 39.52,
|
| 732 |
+
"learning_rate": 6.838709677419355e-05,
|
| 733 |
+
"loss": 0.3207,
|
| 734 |
+
"step": 4900
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 39.52,
|
| 738 |
+
"eval_loss": 0.3276441991329193,
|
| 739 |
+
"eval_runtime": 96.7884,
|
| 740 |
+
"eval_samples_per_second": 19.052,
|
| 741 |
+
"eval_steps_per_second": 2.387,
|
| 742 |
+
"eval_wer": 0.40313257505127725,
|
| 743 |
+
"step": 4900
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 40.32,
|
| 747 |
+
"learning_rate": 6.56989247311828e-05,
|
| 748 |
+
"loss": 0.3186,
|
| 749 |
+
"step": 5000
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 40.32,
|
| 753 |
+
"eval_loss": 0.3335840106010437,
|
| 754 |
+
"eval_runtime": 98.1942,
|
| 755 |
+
"eval_samples_per_second": 18.779,
|
| 756 |
+
"eval_steps_per_second": 2.352,
|
| 757 |
+
"eval_wer": 0.4032258064516129,
|
| 758 |
+
"step": 5000
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 41.13,
|
| 762 |
+
"learning_rate": 6.301075268817205e-05,
|
| 763 |
+
"loss": 0.3152,
|
| 764 |
+
"step": 5100
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 41.13,
|
| 768 |
+
"eval_loss": 0.3129470944404602,
|
| 769 |
+
"eval_runtime": 99.396,
|
| 770 |
+
"eval_samples_per_second": 18.552,
|
| 771 |
+
"eval_steps_per_second": 2.324,
|
| 772 |
+
"eval_wer": 0.3941823606190565,
|
| 773 |
+
"step": 5100
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 41.94,
|
| 777 |
+
"learning_rate": 6.0322580645161295e-05,
|
| 778 |
+
"loss": 0.3093,
|
| 779 |
+
"step": 5200
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 41.94,
|
| 783 |
+
"eval_loss": 0.31502029299736023,
|
| 784 |
+
"eval_runtime": 97.1686,
|
| 785 |
+
"eval_samples_per_second": 18.977,
|
| 786 |
+
"eval_steps_per_second": 2.377,
|
| 787 |
+
"eval_wer": 0.3930635838150289,
|
| 788 |
+
"step": 5200
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 42.74,
|
| 792 |
+
"learning_rate": 5.763440860215054e-05,
|
| 793 |
+
"loss": 0.2949,
|
| 794 |
+
"step": 5300
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 42.74,
|
| 798 |
+
"eval_loss": 0.3235335052013397,
|
| 799 |
+
"eval_runtime": 102.899,
|
| 800 |
+
"eval_samples_per_second": 17.92,
|
| 801 |
+
"eval_steps_per_second": 2.245,
|
| 802 |
+
"eval_wer": 0.3953011374230841,
|
| 803 |
+
"step": 5300
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 43.55,
|
| 807 |
+
"learning_rate": 5.494623655913979e-05,
|
| 808 |
+
"loss": 0.3013,
|
| 809 |
+
"step": 5400
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 43.55,
|
| 813 |
+
"eval_loss": 0.3291964530944824,
|
| 814 |
+
"eval_runtime": 97.4704,
|
| 815 |
+
"eval_samples_per_second": 18.919,
|
| 816 |
+
"eval_steps_per_second": 2.37,
|
| 817 |
+
"eval_wer": 0.3934365094163714,
|
| 818 |
+
"step": 5400
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 44.35,
|
| 822 |
+
"learning_rate": 5.225806451612903e-05,
|
| 823 |
+
"loss": 0.3092,
|
| 824 |
+
"step": 5500
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 44.35,
|
| 828 |
+
"eval_loss": 0.3066520392894745,
|
| 829 |
+
"eval_runtime": 96.9166,
|
| 830 |
+
"eval_samples_per_second": 19.027,
|
| 831 |
+
"eval_steps_per_second": 2.383,
|
| 832 |
+
"eval_wer": 0.38616446019019207,
|
| 833 |
+
"step": 5500
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 45.16,
|
| 837 |
+
"learning_rate": 4.956989247311828e-05,
|
| 838 |
+
"loss": 0.2938,
|
| 839 |
+
"step": 5600
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 45.16,
|
| 843 |
+
"eval_loss": 0.32365530729293823,
|
| 844 |
+
"eval_runtime": 98.8953,
|
| 845 |
+
"eval_samples_per_second": 18.646,
|
| 846 |
+
"eval_steps_per_second": 2.336,
|
| 847 |
+
"eval_wer": 0.38756293119522656,
|
| 848 |
+
"step": 5600
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 45.97,
|
| 852 |
+
"learning_rate": 4.688172043010753e-05,
|
| 853 |
+
"loss": 0.2985,
|
| 854 |
+
"step": 5700
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 45.97,
|
| 858 |
+
"eval_loss": 0.3358381390571594,
|
| 859 |
+
"eval_runtime": 96.7245,
|
| 860 |
+
"eval_samples_per_second": 19.064,
|
| 861 |
+
"eval_steps_per_second": 2.388,
|
| 862 |
+
"eval_wer": 0.3876561625955622,
|
| 863 |
+
"step": 5700
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 46.77,
|
| 867 |
+
"learning_rate": 4.4193548387096775e-05,
|
| 868 |
+
"loss": 0.2943,
|
| 869 |
+
"step": 5800
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 46.77,
|
| 873 |
+
"eval_loss": 0.3283197283744812,
|
| 874 |
+
"eval_runtime": 98.5499,
|
| 875 |
+
"eval_samples_per_second": 18.711,
|
| 876 |
+
"eval_steps_per_second": 2.344,
|
| 877 |
+
"eval_wer": 0.3928771210143576,
|
| 878 |
+
"step": 5800
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 47.58,
|
| 882 |
+
"learning_rate": 4.1505376344086025e-05,
|
| 883 |
+
"loss": 0.2754,
|
| 884 |
+
"step": 5900
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 47.58,
|
| 888 |
+
"eval_loss": 0.33124613761901855,
|
| 889 |
+
"eval_runtime": 97.2557,
|
| 890 |
+
"eval_samples_per_second": 18.96,
|
| 891 |
+
"eval_steps_per_second": 2.375,
|
| 892 |
+
"eval_wer": 0.38952079060227485,
|
| 893 |
+
"step": 5900
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 48.39,
|
| 897 |
+
"learning_rate": 3.881720430107527e-05,
|
| 898 |
+
"loss": 0.2906,
|
| 899 |
+
"step": 6000
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 48.39,
|
| 903 |
+
"eval_loss": 0.33012691140174866,
|
| 904 |
+
"eval_runtime": 99.7418,
|
| 905 |
+
"eval_samples_per_second": 18.488,
|
| 906 |
+
"eval_steps_per_second": 2.316,
|
| 907 |
+
"eval_wer": 0.39045310460563115,
|
| 908 |
+
"step": 6000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 49.19,
|
| 912 |
+
"learning_rate": 3.612903225806452e-05,
|
| 913 |
+
"loss": 0.2751,
|
| 914 |
+
"step": 6100
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 49.19,
|
| 918 |
+
"eval_loss": 0.3261188566684723,
|
| 919 |
+
"eval_runtime": 98.7276,
|
| 920 |
+
"eval_samples_per_second": 18.678,
|
| 921 |
+
"eval_steps_per_second": 2.34,
|
| 922 |
+
"eval_wer": 0.3873764683945553,
|
| 923 |
+
"step": 6100
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 50.0,
|
| 927 |
+
"learning_rate": 3.344086021505377e-05,
|
| 928 |
+
"loss": 0.2825,
|
| 929 |
+
"step": 6200
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 50.0,
|
| 933 |
+
"eval_loss": 0.3322080671787262,
|
| 934 |
+
"eval_runtime": 97.0188,
|
| 935 |
+
"eval_samples_per_second": 19.007,
|
| 936 |
+
"eval_steps_per_second": 2.381,
|
| 937 |
+
"eval_wer": 0.38700354279321275,
|
| 938 |
+
"step": 6200
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 50.81,
|
| 942 |
+
"learning_rate": 3.075268817204301e-05,
|
| 943 |
+
"loss": 0.2727,
|
| 944 |
+
"step": 6300
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 50.81,
|
| 948 |
+
"eval_loss": 0.3448057770729065,
|
| 949 |
+
"eval_runtime": 98.9754,
|
| 950 |
+
"eval_samples_per_second": 18.631,
|
| 951 |
+
"eval_steps_per_second": 2.334,
|
| 952 |
+
"eval_wer": 0.3928771210143576,
|
| 953 |
+
"step": 6300
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 51.61,
|
| 957 |
+
"learning_rate": 2.806451612903226e-05,
|
| 958 |
+
"loss": 0.2896,
|
| 959 |
+
"step": 6400
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 51.61,
|
| 963 |
+
"eval_loss": 0.33276790380477905,
|
| 964 |
+
"eval_runtime": 96.9108,
|
| 965 |
+
"eval_samples_per_second": 19.028,
|
| 966 |
+
"eval_steps_per_second": 2.384,
|
| 967 |
+
"eval_wer": 0.38858847659891854,
|
| 968 |
+
"step": 6400
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 52.42,
|
| 972 |
+
"learning_rate": 2.537634408602151e-05,
|
| 973 |
+
"loss": 0.2753,
|
| 974 |
+
"step": 6500
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 52.42,
|
| 978 |
+
"eval_loss": 0.32856670022010803,
|
| 979 |
+
"eval_runtime": 105.68,
|
| 980 |
+
"eval_samples_per_second": 17.449,
|
| 981 |
+
"eval_steps_per_second": 2.186,
|
| 982 |
+
"eval_wer": 0.3876561625955622,
|
| 983 |
+
"step": 6500
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 53.23,
|
| 987 |
+
"learning_rate": 2.268817204301075e-05,
|
| 988 |
+
"loss": 0.271,
|
| 989 |
+
"step": 6600
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 53.23,
|
| 993 |
+
"eval_loss": 0.3287724256515503,
|
| 994 |
+
"eval_runtime": 100.7427,
|
| 995 |
+
"eval_samples_per_second": 18.304,
|
| 996 |
+
"eval_steps_per_second": 2.293,
|
| 997 |
+
"eval_wer": 0.3876561625955622,
|
| 998 |
+
"step": 6600
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 54.03,
|
| 1002 |
+
"learning_rate": 2e-05,
|
| 1003 |
+
"loss": 0.2664,
|
| 1004 |
+
"step": 6700
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 54.03,
|
| 1008 |
+
"eval_loss": 0.3290824890136719,
|
| 1009 |
+
"eval_runtime": 98.5577,
|
| 1010 |
+
"eval_samples_per_second": 18.71,
|
| 1011 |
+
"eval_steps_per_second": 2.344,
|
| 1012 |
+
"eval_wer": 0.3893343278016036,
|
| 1013 |
+
"step": 6700
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 54.84,
|
| 1017 |
+
"learning_rate": 1.7311827956989248e-05,
|
| 1018 |
+
"loss": 0.2491,
|
| 1019 |
+
"step": 6800
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 54.84,
|
| 1023 |
+
"eval_loss": 0.33568552136421204,
|
| 1024 |
+
"eval_runtime": 103.2565,
|
| 1025 |
+
"eval_samples_per_second": 17.858,
|
| 1026 |
+
"eval_steps_per_second": 2.237,
|
| 1027 |
+
"eval_wer": 0.388215550997576,
|
| 1028 |
+
"step": 6800
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 55.65,
|
| 1032 |
+
"learning_rate": 1.4623655913978495e-05,
|
| 1033 |
+
"loss": 0.2569,
|
| 1034 |
+
"step": 6900
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 55.65,
|
| 1038 |
+
"eval_loss": 0.3253258466720581,
|
| 1039 |
+
"eval_runtime": 104.0214,
|
| 1040 |
+
"eval_samples_per_second": 17.727,
|
| 1041 |
+
"eval_steps_per_second": 2.221,
|
| 1042 |
+
"eval_wer": 0.3876561625955622,
|
| 1043 |
+
"step": 6900
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 56.45,
|
| 1047 |
+
"learning_rate": 1.1935483870967743e-05,
|
| 1048 |
+
"loss": 0.2568,
|
| 1049 |
+
"step": 7000
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 56.45,
|
| 1053 |
+
"eval_loss": 0.3199854791164398,
|
| 1054 |
+
"eval_runtime": 103.944,
|
| 1055 |
+
"eval_samples_per_second": 17.74,
|
| 1056 |
+
"eval_steps_per_second": 2.222,
|
| 1057 |
+
"eval_wer": 0.3868170799925415,
|
| 1058 |
+
"step": 7000
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 57.26,
|
| 1062 |
+
"learning_rate": 9.24731182795699e-06,
|
| 1063 |
+
"loss": 0.2668,
|
| 1064 |
+
"step": 7100
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 57.26,
|
| 1068 |
+
"eval_loss": 0.32368558645248413,
|
| 1069 |
+
"eval_runtime": 103.5633,
|
| 1070 |
+
"eval_samples_per_second": 17.806,
|
| 1071 |
+
"eval_steps_per_second": 2.231,
|
| 1072 |
+
"eval_wer": 0.3862576915905277,
|
| 1073 |
+
"step": 7100
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 58.06,
|
| 1077 |
+
"learning_rate": 6.586021505376344e-06,
|
| 1078 |
+
"loss": 0.253,
|
| 1079 |
+
"step": 7200
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 58.06,
|
| 1083 |
+
"eval_loss": 0.3248105049133301,
|
| 1084 |
+
"eval_runtime": 103.7822,
|
| 1085 |
+
"eval_samples_per_second": 17.768,
|
| 1086 |
+
"eval_steps_per_second": 2.226,
|
| 1087 |
+
"eval_wer": 0.38504568338616446,
|
| 1088 |
+
"step": 7200
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 58.87,
|
| 1092 |
+
"learning_rate": 3.8978494623655915e-06,
|
| 1093 |
+
"loss": 0.2621,
|
| 1094 |
+
"step": 7300
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 58.87,
|
| 1098 |
+
"eval_loss": 0.3238443434238434,
|
| 1099 |
+
"eval_runtime": 97.9464,
|
| 1100 |
+
"eval_samples_per_second": 18.827,
|
| 1101 |
+
"eval_steps_per_second": 2.358,
|
| 1102 |
+
"eval_wer": 0.38448629498415066,
|
| 1103 |
+
"step": 7300
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 59.68,
|
| 1107 |
+
"learning_rate": 1.2096774193548388e-06,
|
| 1108 |
+
"loss": 0.2571,
|
| 1109 |
+
"step": 7400
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 59.68,
|
| 1113 |
+
"eval_loss": 0.3273279666900635,
|
| 1114 |
+
"eval_runtime": 101.5169,
|
| 1115 |
+
"eval_samples_per_second": 18.164,
|
| 1116 |
+
"eval_steps_per_second": 2.275,
|
| 1117 |
+
"eval_wer": 0.3852321461868357,
|
| 1118 |
+
"step": 7400
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 60.0,
|
| 1122 |
+
"step": 7440,
|
| 1123 |
+
"total_flos": 2.7388066347434852e+19,
|
| 1124 |
+
"train_loss": 0.8272683016715511,
|
| 1125 |
+
"train_runtime": 19767.0967,
|
| 1126 |
+
"train_samples_per_second": 12.035,
|
| 1127 |
+
"train_steps_per_second": 0.376
|
| 1128 |
+
}
|
| 1129 |
+
],
|
| 1130 |
+
"max_steps": 7440,
|
| 1131 |
+
"num_train_epochs": 60,
|
| 1132 |
+
"total_flos": 2.7388066347434852e+19,
|
| 1133 |
+
"trial_name": null,
|
| 1134 |
+
"trial_params": null
|
| 1135 |
+
}
|
vocab.json
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"
|
|
|
|
| 1 |
+
{"y": 0, "h": 1, "c": 2, "d": 3, "r": 4, "t": 5, "é": 6, "n": 8, "！": 9, "x": 10, "v": 11, "p": 12, "z": 13, "l": 14, "j": 15, "m": 16, "b": 17, "i": 18, "e": 19, "g": 20, "a": 21, "，": 22, "u": 23, "f": 24, "s": 25, "k": 26, "o": 27, "w": 28, "|": 7, "[UNK]": 29, "[PAD]": 30}
|