Upload 12 files
Browse files- config.json +109 -0
- optimizer.pt +3 -0
- preprocessor_config.json +10 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +6 -0
- tokenizer_config.json +11 -0
- trainer_state.json +946 -0
- training_args.bin +3 -0
- vocab.json +50 -0
config.json
ADDED
|
@@ -0,0 +1,109 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "/content/drive/MyDrive/model_ft/checkpoint-40000",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"adapter_kernel_size": 3,
|
| 5 |
+
"adapter_stride": 2,
|
| 6 |
+
"add_adapter": false,
|
| 7 |
+
"apply_spec_augment": true,
|
| 8 |
+
"architectures": [
|
| 9 |
+
"Wav2Vec2ForCTC"
|
| 10 |
+
],
|
| 11 |
+
"attention_dropout": 0.1,
|
| 12 |
+
"bos_token_id": 1,
|
| 13 |
+
"classifier_proj_size": 256,
|
| 14 |
+
"codevector_dim": 768,
|
| 15 |
+
"contrastive_logits_temperature": 0.1,
|
| 16 |
+
"conv_bias": true,
|
| 17 |
+
"conv_dim": [
|
| 18 |
+
512,
|
| 19 |
+
512,
|
| 20 |
+
512,
|
| 21 |
+
512,
|
| 22 |
+
512,
|
| 23 |
+
512,
|
| 24 |
+
512
|
| 25 |
+
],
|
| 26 |
+
"conv_kernel": [
|
| 27 |
+
10,
|
| 28 |
+
3,
|
| 29 |
+
3,
|
| 30 |
+
3,
|
| 31 |
+
3,
|
| 32 |
+
2,
|
| 33 |
+
2
|
| 34 |
+
],
|
| 35 |
+
"conv_stride": [
|
| 36 |
+
5,
|
| 37 |
+
2,
|
| 38 |
+
2,
|
| 39 |
+
2,
|
| 40 |
+
2,
|
| 41 |
+
2,
|
| 42 |
+
2
|
| 43 |
+
],
|
| 44 |
+
"ctc_loss_reduction": "mean",
|
| 45 |
+
"ctc_zero_infinity": true,
|
| 46 |
+
"diversity_loss_weight": 0.1,
|
| 47 |
+
"do_stable_layer_norm": true,
|
| 48 |
+
"eos_token_id": 2,
|
| 49 |
+
"feat_extract_activation": "gelu",
|
| 50 |
+
"feat_extract_dropout": 0.0,
|
| 51 |
+
"feat_extract_norm": "layer",
|
| 52 |
+
"feat_proj_dropout": 0.1,
|
| 53 |
+
"feat_quantizer_dropout": 0.0,
|
| 54 |
+
"final_dropout": 0.0,
|
| 55 |
+
"gradient_checkpointing": false,
|
| 56 |
+
"hidden_act": "gelu",
|
| 57 |
+
"hidden_dropout": 0.1,
|
| 58 |
+
"hidden_size": 1024,
|
| 59 |
+
"initializer_range": 0.02,
|
| 60 |
+
"intermediate_size": 4096,
|
| 61 |
+
"layer_norm_eps": 1e-05,
|
| 62 |
+
"layerdrop": 0.1,
|
| 63 |
+
"mask_feature_length": 10,
|
| 64 |
+
"mask_feature_min_masks": 0,
|
| 65 |
+
"mask_feature_prob": 0.004,
|
| 66 |
+
"mask_time_length": 10,
|
| 67 |
+
"mask_time_min_masks": 2,
|
| 68 |
+
"mask_time_prob": 0.075,
|
| 69 |
+
"model_type": "wav2vec2",
|
| 70 |
+
"num_adapter_layers": 3,
|
| 71 |
+
"num_attention_heads": 16,
|
| 72 |
+
"num_codevector_groups": 2,
|
| 73 |
+
"num_codevectors_per_group": 320,
|
| 74 |
+
"num_conv_pos_embedding_groups": 16,
|
| 75 |
+
"num_conv_pos_embeddings": 128,
|
| 76 |
+
"num_feat_extract_layers": 7,
|
| 77 |
+
"num_hidden_layers": 24,
|
| 78 |
+
"num_negatives": 100,
|
| 79 |
+
"output_hidden_size": 1024,
|
| 80 |
+
"pad_token_id": 0,
|
| 81 |
+
"pretrained_model_name_or_path": "/content/drive/MyDrive/model_ft/checkpoint-40000",
|
| 82 |
+
"proj_codevector_dim": 768,
|
| 83 |
+
"tdnn_dilation": [
|
| 84 |
+
1,
|
| 85 |
+
2,
|
| 86 |
+
3,
|
| 87 |
+
1,
|
| 88 |
+
1
|
| 89 |
+
],
|
| 90 |
+
"tdnn_dim": [
|
| 91 |
+
512,
|
| 92 |
+
512,
|
| 93 |
+
512,
|
| 94 |
+
512,
|
| 95 |
+
1500
|
| 96 |
+
],
|
| 97 |
+
"tdnn_kernel": [
|
| 98 |
+
5,
|
| 99 |
+
3,
|
| 100 |
+
3,
|
| 101 |
+
1,
|
| 102 |
+
1
|
| 103 |
+
],
|
| 104 |
+
"torch_dtype": "float32",
|
| 105 |
+
"transformers_version": "4.24.0",
|
| 106 |
+
"use_weighted_layer_sum": false,
|
| 107 |
+
"vocab_size": 48,
|
| 108 |
+
"xvector_output_dim": 512
|
| 109 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:951b81d481a1e2cec25044e85395503a4aa9d0d43fc349ed96978d3c8653dd7c
|
| 3 |
+
size 2490552961
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_normalize": true,
|
| 3 |
+
"feature_extractor_type": "Wav2Vec2FeatureExtractor",
|
| 4 |
+
"feature_size": 1,
|
| 5 |
+
"padding_side": "right",
|
| 6 |
+
"padding_value": 0.0,
|
| 7 |
+
"processor_class": "Wav2Vec2Processor",
|
| 8 |
+
"return_attention_mask": true,
|
| 9 |
+
"sampling_rate": 16000
|
| 10 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:87b5541b7d80379c0ae0d5cb53c56d5ed9c67debc4f75c4892a88a034b90e9c1
|
| 3 |
+
size 1262098669
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f1d1b12c7f9fe21094a4717a8f9ad70f2f46c7a3117636a199509032dad90adb
|
| 3 |
+
size 14639
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b5eef804c8b9f91f0fc0c9e9b8aa1ae0eeac472a98143189d47bfb352f4aa8f
|
| 3 |
+
size 557
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:58a99bf1e9c0f65595ce423b2dc75af3991c15e55775c10969849c46314e2902
|
| 3 |
+
size 627
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"eos_token": "</s>",
|
| 4 |
+
"pad_token": "<pad>",
|
| 5 |
+
"unk_token": "<unk>"
|
| 6 |
+
}
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"do_lower_case": false,
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"pad_token": "<pad>",
|
| 6 |
+
"processor_class": "Wav2Vec2Processor",
|
| 7 |
+
"replace_word_delimiter_char": " ",
|
| 8 |
+
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
| 9 |
+
"unk_token": "<unk>",
|
| 10 |
+
"word_delimiter_token": "|"
|
| 11 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,946 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.24188034188034188,
|
| 3 |
+
"best_model_checkpoint": "/content/drive/MyDrive/model_ft_2/checkpoint-15000",
|
| 4 |
+
"epoch": 17.647058823529413,
|
| 5 |
+
"global_step": 15000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.12,
|
| 12 |
+
"learning_rate": 5.000000000000001e-07,
|
| 13 |
+
"loss": 0.3163,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.24,
|
| 18 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 19 |
+
"loss": 0.2851,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.35,
|
| 24 |
+
"learning_rate": 1.5e-06,
|
| 25 |
+
"loss": 0.2783,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.47,
|
| 30 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 31 |
+
"loss": 0.3097,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.59,
|
| 36 |
+
"learning_rate": 2.5e-06,
|
| 37 |
+
"loss": 0.286,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.71,
|
| 42 |
+
"learning_rate": 3e-06,
|
| 43 |
+
"loss": 0.3008,
|
| 44 |
+
"step": 600
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.82,
|
| 48 |
+
"learning_rate": 3.5e-06,
|
| 49 |
+
"loss": 0.2617,
|
| 50 |
+
"step": 700
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.94,
|
| 54 |
+
"learning_rate": 4.000000000000001e-06,
|
| 55 |
+
"loss": 0.2949,
|
| 56 |
+
"step": 800
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.06,
|
| 60 |
+
"learning_rate": 4.5e-06,
|
| 61 |
+
"loss": 0.2744,
|
| 62 |
+
"step": 900
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.18,
|
| 66 |
+
"learning_rate": 5e-06,
|
| 67 |
+
"loss": 0.2932,
|
| 68 |
+
"step": 1000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.29,
|
| 72 |
+
"learning_rate": 5.500000000000001e-06,
|
| 73 |
+
"loss": 0.289,
|
| 74 |
+
"step": 1100
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.41,
|
| 78 |
+
"learning_rate": 6e-06,
|
| 79 |
+
"loss": 0.2984,
|
| 80 |
+
"step": 1200
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 1.53,
|
| 84 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 85 |
+
"loss": 0.3028,
|
| 86 |
+
"step": 1300
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.65,
|
| 90 |
+
"learning_rate": 7e-06,
|
| 91 |
+
"loss": 0.2607,
|
| 92 |
+
"step": 1400
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 1.76,
|
| 96 |
+
"learning_rate": 7.500000000000001e-06,
|
| 97 |
+
"loss": 0.2906,
|
| 98 |
+
"step": 1500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.88,
|
| 102 |
+
"learning_rate": 8.000000000000001e-06,
|
| 103 |
+
"loss": 0.2755,
|
| 104 |
+
"step": 1600
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 2.0,
|
| 108 |
+
"learning_rate": 8.5e-06,
|
| 109 |
+
"loss": 0.2894,
|
| 110 |
+
"step": 1700
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 2.12,
|
| 114 |
+
"learning_rate": 9e-06,
|
| 115 |
+
"loss": 0.282,
|
| 116 |
+
"step": 1800
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.24,
|
| 120 |
+
"learning_rate": 9.5e-06,
|
| 121 |
+
"loss": 0.2715,
|
| 122 |
+
"step": 1900
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 2.35,
|
| 126 |
+
"learning_rate": 1e-05,
|
| 127 |
+
"loss": 0.2845,
|
| 128 |
+
"step": 2000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.47,
|
| 132 |
+
"learning_rate": 1e-05,
|
| 133 |
+
"loss": 0.2698,
|
| 134 |
+
"step": 2100
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 2.59,
|
| 138 |
+
"learning_rate": 1e-05,
|
| 139 |
+
"loss": 0.2803,
|
| 140 |
+
"step": 2200
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 2.71,
|
| 144 |
+
"learning_rate": 1e-05,
|
| 145 |
+
"loss": 0.2812,
|
| 146 |
+
"step": 2300
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 2.82,
|
| 150 |
+
"learning_rate": 1e-05,
|
| 151 |
+
"loss": 0.2947,
|
| 152 |
+
"step": 2400
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 2.94,
|
| 156 |
+
"learning_rate": 1e-05,
|
| 157 |
+
"loss": 0.2529,
|
| 158 |
+
"step": 2500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 3.06,
|
| 162 |
+
"learning_rate": 1e-05,
|
| 163 |
+
"loss": 0.298,
|
| 164 |
+
"step": 2600
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 3.18,
|
| 168 |
+
"learning_rate": 1e-05,
|
| 169 |
+
"loss": 0.2671,
|
| 170 |
+
"step": 2700
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 3.29,
|
| 174 |
+
"learning_rate": 1e-05,
|
| 175 |
+
"loss": 0.2619,
|
| 176 |
+
"step": 2800
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.41,
|
| 180 |
+
"learning_rate": 1e-05,
|
| 181 |
+
"loss": 0.2546,
|
| 182 |
+
"step": 2900
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 3.53,
|
| 186 |
+
"learning_rate": 1e-05,
|
| 187 |
+
"loss": 0.2776,
|
| 188 |
+
"step": 3000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 3.65,
|
| 192 |
+
"learning_rate": 1e-05,
|
| 193 |
+
"loss": 0.2869,
|
| 194 |
+
"step": 3100
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 3.76,
|
| 198 |
+
"learning_rate": 1e-05,
|
| 199 |
+
"loss": 0.2664,
|
| 200 |
+
"step": 3200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 3.88,
|
| 204 |
+
"learning_rate": 1e-05,
|
| 205 |
+
"loss": 0.2893,
|
| 206 |
+
"step": 3300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 4.0,
|
| 210 |
+
"learning_rate": 1e-05,
|
| 211 |
+
"loss": 0.2691,
|
| 212 |
+
"step": 3400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 4.12,
|
| 216 |
+
"learning_rate": 1e-05,
|
| 217 |
+
"loss": 0.2729,
|
| 218 |
+
"step": 3500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 4.24,
|
| 222 |
+
"learning_rate": 1e-05,
|
| 223 |
+
"loss": 0.2606,
|
| 224 |
+
"step": 3600
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 4.35,
|
| 228 |
+
"learning_rate": 1e-05,
|
| 229 |
+
"loss": 0.2731,
|
| 230 |
+
"step": 3700
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 4.47,
|
| 234 |
+
"learning_rate": 1e-05,
|
| 235 |
+
"loss": 0.245,
|
| 236 |
+
"step": 3800
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 4.59,
|
| 240 |
+
"learning_rate": 1e-05,
|
| 241 |
+
"loss": 0.2644,
|
| 242 |
+
"step": 3900
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 4.71,
|
| 246 |
+
"learning_rate": 1e-05,
|
| 247 |
+
"loss": 0.2651,
|
| 248 |
+
"step": 4000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 4.82,
|
| 252 |
+
"learning_rate": 1e-05,
|
| 253 |
+
"loss": 0.2437,
|
| 254 |
+
"step": 4100
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 4.94,
|
| 258 |
+
"learning_rate": 1e-05,
|
| 259 |
+
"loss": 0.2566,
|
| 260 |
+
"step": 4200
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 5.06,
|
| 264 |
+
"learning_rate": 1e-05,
|
| 265 |
+
"loss": 0.2582,
|
| 266 |
+
"step": 4300
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 5.18,
|
| 270 |
+
"learning_rate": 1e-05,
|
| 271 |
+
"loss": 0.2261,
|
| 272 |
+
"step": 4400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 5.29,
|
| 276 |
+
"learning_rate": 1e-05,
|
| 277 |
+
"loss": 0.2685,
|
| 278 |
+
"step": 4500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 5.41,
|
| 282 |
+
"learning_rate": 1e-05,
|
| 283 |
+
"loss": 0.2459,
|
| 284 |
+
"step": 4600
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 5.53,
|
| 288 |
+
"learning_rate": 1e-05,
|
| 289 |
+
"loss": 0.2536,
|
| 290 |
+
"step": 4700
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 5.65,
|
| 294 |
+
"learning_rate": 1e-05,
|
| 295 |
+
"loss": 0.2592,
|
| 296 |
+
"step": 4800
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 5.76,
|
| 300 |
+
"learning_rate": 1e-05,
|
| 301 |
+
"loss": 0.2491,
|
| 302 |
+
"step": 4900
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 5.88,
|
| 306 |
+
"learning_rate": 1e-05,
|
| 307 |
+
"loss": 0.2434,
|
| 308 |
+
"step": 5000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 5.88,
|
| 312 |
+
"eval_cer": 0.0821832842301132,
|
| 313 |
+
"eval_loss": 0.575729250907898,
|
| 314 |
+
"eval_runtime": 8.2591,
|
| 315 |
+
"eval_samples_per_second": 36.324,
|
| 316 |
+
"eval_steps_per_second": 18.162,
|
| 317 |
+
"eval_wer": 0.25256410256410255,
|
| 318 |
+
"step": 5000
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 6.0,
|
| 322 |
+
"learning_rate": 1e-05,
|
| 323 |
+
"loss": 0.2547,
|
| 324 |
+
"step": 5100
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 6.12,
|
| 328 |
+
"learning_rate": 1e-05,
|
| 329 |
+
"loss": 0.2459,
|
| 330 |
+
"step": 5200
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 6.24,
|
| 334 |
+
"learning_rate": 1e-05,
|
| 335 |
+
"loss": 0.2566,
|
| 336 |
+
"step": 5300
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 6.35,
|
| 340 |
+
"learning_rate": 1e-05,
|
| 341 |
+
"loss": 0.2498,
|
| 342 |
+
"step": 5400
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 6.47,
|
| 346 |
+
"learning_rate": 1e-05,
|
| 347 |
+
"loss": 0.2492,
|
| 348 |
+
"step": 5500
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 6.59,
|
| 352 |
+
"learning_rate": 1e-05,
|
| 353 |
+
"loss": 0.2491,
|
| 354 |
+
"step": 5600
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 6.71,
|
| 358 |
+
"learning_rate": 1e-05,
|
| 359 |
+
"loss": 0.2519,
|
| 360 |
+
"step": 5700
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 6.82,
|
| 364 |
+
"learning_rate": 1e-05,
|
| 365 |
+
"loss": 0.2426,
|
| 366 |
+
"step": 5800
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 6.94,
|
| 370 |
+
"learning_rate": 1e-05,
|
| 371 |
+
"loss": 0.2175,
|
| 372 |
+
"step": 5900
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 7.06,
|
| 376 |
+
"learning_rate": 1e-05,
|
| 377 |
+
"loss": 0.2506,
|
| 378 |
+
"step": 6000
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 7.18,
|
| 382 |
+
"learning_rate": 1e-05,
|
| 383 |
+
"loss": 0.2342,
|
| 384 |
+
"step": 6100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 7.29,
|
| 388 |
+
"learning_rate": 1e-05,
|
| 389 |
+
"loss": 0.2418,
|
| 390 |
+
"step": 6200
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 7.41,
|
| 394 |
+
"learning_rate": 1e-05,
|
| 395 |
+
"loss": 0.2345,
|
| 396 |
+
"step": 6300
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 7.53,
|
| 400 |
+
"learning_rate": 1e-05,
|
| 401 |
+
"loss": 0.2358,
|
| 402 |
+
"step": 6400
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 7.65,
|
| 406 |
+
"learning_rate": 1e-05,
|
| 407 |
+
"loss": 0.2367,
|
| 408 |
+
"step": 6500
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 7.76,
|
| 412 |
+
"learning_rate": 1e-05,
|
| 413 |
+
"loss": 0.2402,
|
| 414 |
+
"step": 6600
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 7.88,
|
| 418 |
+
"learning_rate": 1e-05,
|
| 419 |
+
"loss": 0.2448,
|
| 420 |
+
"step": 6700
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 8.0,
|
| 424 |
+
"learning_rate": 1e-05,
|
| 425 |
+
"loss": 0.2458,
|
| 426 |
+
"step": 6800
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 8.12,
|
| 430 |
+
"learning_rate": 1e-05,
|
| 431 |
+
"loss": 0.2408,
|
| 432 |
+
"step": 6900
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 8.24,
|
| 436 |
+
"learning_rate": 1e-05,
|
| 437 |
+
"loss": 0.2425,
|
| 438 |
+
"step": 7000
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 8.35,
|
| 442 |
+
"learning_rate": 1e-05,
|
| 443 |
+
"loss": 0.2302,
|
| 444 |
+
"step": 7100
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 8.47,
|
| 448 |
+
"learning_rate": 1e-05,
|
| 449 |
+
"loss": 0.2426,
|
| 450 |
+
"step": 7200
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 8.59,
|
| 454 |
+
"learning_rate": 1e-05,
|
| 455 |
+
"loss": 0.2101,
|
| 456 |
+
"step": 7300
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 8.71,
|
| 460 |
+
"learning_rate": 1e-05,
|
| 461 |
+
"loss": 0.2328,
|
| 462 |
+
"step": 7400
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 8.82,
|
| 466 |
+
"learning_rate": 1e-05,
|
| 467 |
+
"loss": 0.2457,
|
| 468 |
+
"step": 7500
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 8.94,
|
| 472 |
+
"learning_rate": 1e-05,
|
| 473 |
+
"loss": 0.2351,
|
| 474 |
+
"step": 7600
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 9.06,
|
| 478 |
+
"learning_rate": 1e-05,
|
| 479 |
+
"loss": 0.2483,
|
| 480 |
+
"step": 7700
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 9.18,
|
| 484 |
+
"learning_rate": 1e-05,
|
| 485 |
+
"loss": 0.2135,
|
| 486 |
+
"step": 7800
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 9.29,
|
| 490 |
+
"learning_rate": 1e-05,
|
| 491 |
+
"loss": 0.2227,
|
| 492 |
+
"step": 7900
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 9.41,
|
| 496 |
+
"learning_rate": 1e-05,
|
| 497 |
+
"loss": 0.2342,
|
| 498 |
+
"step": 8000
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 9.53,
|
| 502 |
+
"learning_rate": 1e-05,
|
| 503 |
+
"loss": 0.2534,
|
| 504 |
+
"step": 8100
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 9.65,
|
| 508 |
+
"learning_rate": 1e-05,
|
| 509 |
+
"loss": 0.2368,
|
| 510 |
+
"step": 8200
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 9.76,
|
| 514 |
+
"learning_rate": 1e-05,
|
| 515 |
+
"loss": 0.2531,
|
| 516 |
+
"step": 8300
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 9.88,
|
| 520 |
+
"learning_rate": 1e-05,
|
| 521 |
+
"loss": 0.2474,
|
| 522 |
+
"step": 8400
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 10.0,
|
| 526 |
+
"learning_rate": 1e-05,
|
| 527 |
+
"loss": 0.2049,
|
| 528 |
+
"step": 8500
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 10.12,
|
| 532 |
+
"learning_rate": 1e-05,
|
| 533 |
+
"loss": 0.231,
|
| 534 |
+
"step": 8600
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 10.24,
|
| 538 |
+
"learning_rate": 1e-05,
|
| 539 |
+
"loss": 0.2384,
|
| 540 |
+
"step": 8700
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 10.35,
|
| 544 |
+
"learning_rate": 1e-05,
|
| 545 |
+
"loss": 0.2363,
|
| 546 |
+
"step": 8800
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 10.47,
|
| 550 |
+
"learning_rate": 1e-05,
|
| 551 |
+
"loss": 0.2125,
|
| 552 |
+
"step": 8900
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 10.59,
|
| 556 |
+
"learning_rate": 1e-05,
|
| 557 |
+
"loss": 0.224,
|
| 558 |
+
"step": 9000
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 10.71,
|
| 562 |
+
"learning_rate": 1e-05,
|
| 563 |
+
"loss": 0.2301,
|
| 564 |
+
"step": 9100
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 10.82,
|
| 568 |
+
"learning_rate": 1e-05,
|
| 569 |
+
"loss": 0.2598,
|
| 570 |
+
"step": 9200
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 10.94,
|
| 574 |
+
"learning_rate": 1e-05,
|
| 575 |
+
"loss": 0.2121,
|
| 576 |
+
"step": 9300
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 11.06,
|
| 580 |
+
"learning_rate": 1e-05,
|
| 581 |
+
"loss": 0.2444,
|
| 582 |
+
"step": 9400
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 11.18,
|
| 586 |
+
"learning_rate": 1e-05,
|
| 587 |
+
"loss": 0.2409,
|
| 588 |
+
"step": 9500
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 11.29,
|
| 592 |
+
"learning_rate": 1e-05,
|
| 593 |
+
"loss": 0.2274,
|
| 594 |
+
"step": 9600
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 11.41,
|
| 598 |
+
"learning_rate": 1e-05,
|
| 599 |
+
"loss": 0.2246,
|
| 600 |
+
"step": 9700
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 11.53,
|
| 604 |
+
"learning_rate": 1e-05,
|
| 605 |
+
"loss": 0.2256,
|
| 606 |
+
"step": 9800
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 11.65,
|
| 610 |
+
"learning_rate": 1e-05,
|
| 611 |
+
"loss": 0.2244,
|
| 612 |
+
"step": 9900
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 11.76,
|
| 616 |
+
"learning_rate": 1e-05,
|
| 617 |
+
"loss": 0.2246,
|
| 618 |
+
"step": 10000
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 11.76,
|
| 622 |
+
"eval_cer": 0.08288106683206699,
|
| 623 |
+
"eval_loss": 0.6029604077339172,
|
| 624 |
+
"eval_runtime": 8.1986,
|
| 625 |
+
"eval_samples_per_second": 36.592,
|
| 626 |
+
"eval_steps_per_second": 18.296,
|
| 627 |
+
"eval_wer": 0.2551282051282051,
|
| 628 |
+
"step": 10000
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 11.88,
|
| 632 |
+
"learning_rate": 9.903e-06,
|
| 633 |
+
"loss": 0.2299,
|
| 634 |
+
"step": 10100
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 12.0,
|
| 638 |
+
"learning_rate": 9.803e-06,
|
| 639 |
+
"loss": 0.2131,
|
| 640 |
+
"step": 10200
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 12.12,
|
| 644 |
+
"learning_rate": 9.703000000000002e-06,
|
| 645 |
+
"loss": 0.2134,
|
| 646 |
+
"step": 10300
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 12.24,
|
| 650 |
+
"learning_rate": 9.603000000000001e-06,
|
| 651 |
+
"loss": 0.2388,
|
| 652 |
+
"step": 10400
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 12.35,
|
| 656 |
+
"learning_rate": 9.503e-06,
|
| 657 |
+
"loss": 0.2209,
|
| 658 |
+
"step": 10500
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 12.47,
|
| 662 |
+
"learning_rate": 9.403000000000001e-06,
|
| 663 |
+
"loss": 0.2347,
|
| 664 |
+
"step": 10600
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 12.59,
|
| 668 |
+
"learning_rate": 9.303e-06,
|
| 669 |
+
"loss": 0.2302,
|
| 670 |
+
"step": 10700
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 12.71,
|
| 674 |
+
"learning_rate": 9.203000000000002e-06,
|
| 675 |
+
"loss": 0.2174,
|
| 676 |
+
"step": 10800
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 12.82,
|
| 680 |
+
"learning_rate": 9.103e-06,
|
| 681 |
+
"loss": 0.2224,
|
| 682 |
+
"step": 10900
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 12.94,
|
| 686 |
+
"learning_rate": 9.003e-06,
|
| 687 |
+
"loss": 0.2269,
|
| 688 |
+
"step": 11000
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 13.06,
|
| 692 |
+
"learning_rate": 8.903000000000001e-06,
|
| 693 |
+
"loss": 0.2426,
|
| 694 |
+
"step": 11100
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 13.18,
|
| 698 |
+
"learning_rate": 8.803e-06,
|
| 699 |
+
"loss": 0.209,
|
| 700 |
+
"step": 11200
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 13.29,
|
| 704 |
+
"learning_rate": 8.703e-06,
|
| 705 |
+
"loss": 0.2103,
|
| 706 |
+
"step": 11300
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 13.41,
|
| 710 |
+
"learning_rate": 8.603e-06,
|
| 711 |
+
"loss": 0.228,
|
| 712 |
+
"step": 11400
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 13.53,
|
| 716 |
+
"learning_rate": 8.503e-06,
|
| 717 |
+
"loss": 0.2256,
|
| 718 |
+
"step": 11500
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 13.65,
|
| 722 |
+
"learning_rate": 8.404000000000001e-06,
|
| 723 |
+
"loss": 0.211,
|
| 724 |
+
"step": 11600
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 13.76,
|
| 728 |
+
"learning_rate": 8.304e-06,
|
| 729 |
+
"loss": 0.1967,
|
| 730 |
+
"step": 11700
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 13.88,
|
| 734 |
+
"learning_rate": 8.204000000000001e-06,
|
| 735 |
+
"loss": 0.2417,
|
| 736 |
+
"step": 11800
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 14.0,
|
| 740 |
+
"learning_rate": 8.104e-06,
|
| 741 |
+
"loss": 0.2268,
|
| 742 |
+
"step": 11900
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 14.12,
|
| 746 |
+
"learning_rate": 8.004e-06,
|
| 747 |
+
"loss": 0.227,
|
| 748 |
+
"step": 12000
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 14.24,
|
| 752 |
+
"learning_rate": 7.904000000000001e-06,
|
| 753 |
+
"loss": 0.2186,
|
| 754 |
+
"step": 12100
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 14.35,
|
| 758 |
+
"learning_rate": 7.804e-06,
|
| 759 |
+
"loss": 0.2125,
|
| 760 |
+
"step": 12200
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 14.47,
|
| 764 |
+
"learning_rate": 7.704000000000001e-06,
|
| 765 |
+
"loss": 0.2235,
|
| 766 |
+
"step": 12300
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 14.59,
|
| 770 |
+
"learning_rate": 7.604e-06,
|
| 771 |
+
"loss": 0.2089,
|
| 772 |
+
"step": 12400
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 14.71,
|
| 776 |
+
"learning_rate": 7.5040000000000005e-06,
|
| 777 |
+
"loss": 0.1967,
|
| 778 |
+
"step": 12500
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 14.82,
|
| 782 |
+
"learning_rate": 7.404e-06,
|
| 783 |
+
"loss": 0.2175,
|
| 784 |
+
"step": 12600
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 14.94,
|
| 788 |
+
"learning_rate": 7.304000000000001e-06,
|
| 789 |
+
"loss": 0.2157,
|
| 790 |
+
"step": 12700
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 15.06,
|
| 794 |
+
"learning_rate": 7.204000000000001e-06,
|
| 795 |
+
"loss": 0.1907,
|
| 796 |
+
"step": 12800
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 15.18,
|
| 800 |
+
"learning_rate": 7.104000000000001e-06,
|
| 801 |
+
"loss": 0.2072,
|
| 802 |
+
"step": 12900
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 15.29,
|
| 806 |
+
"learning_rate": 7.004000000000001e-06,
|
| 807 |
+
"loss": 0.2137,
|
| 808 |
+
"step": 13000
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 15.41,
|
| 812 |
+
"learning_rate": 6.904e-06,
|
| 813 |
+
"loss": 0.2137,
|
| 814 |
+
"step": 13100
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 15.53,
|
| 818 |
+
"learning_rate": 6.804e-06,
|
| 819 |
+
"loss": 0.2097,
|
| 820 |
+
"step": 13200
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 15.65,
|
| 824 |
+
"learning_rate": 6.7040000000000005e-06,
|
| 825 |
+
"loss": 0.2007,
|
| 826 |
+
"step": 13300
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 15.76,
|
| 830 |
+
"learning_rate": 6.604000000000001e-06,
|
| 831 |
+
"loss": 0.2183,
|
| 832 |
+
"step": 13400
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 15.88,
|
| 836 |
+
"learning_rate": 6.504e-06,
|
| 837 |
+
"loss": 0.1844,
|
| 838 |
+
"step": 13500
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 16.0,
|
| 842 |
+
"learning_rate": 6.404e-06,
|
| 843 |
+
"loss": 0.225,
|
| 844 |
+
"step": 13600
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 16.12,
|
| 848 |
+
"learning_rate": 6.304e-06,
|
| 849 |
+
"loss": 0.2141,
|
| 850 |
+
"step": 13700
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 16.24,
|
| 854 |
+
"learning_rate": 6.204e-06,
|
| 855 |
+
"loss": 0.2047,
|
| 856 |
+
"step": 13800
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 16.35,
|
| 860 |
+
"learning_rate": 6.104000000000001e-06,
|
| 861 |
+
"loss": 0.2019,
|
| 862 |
+
"step": 13900
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 16.47,
|
| 866 |
+
"learning_rate": 6.004000000000001e-06,
|
| 867 |
+
"loss": 0.1923,
|
| 868 |
+
"step": 14000
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 16.59,
|
| 872 |
+
"learning_rate": 5.9040000000000006e-06,
|
| 873 |
+
"loss": 0.2213,
|
| 874 |
+
"step": 14100
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 16.71,
|
| 878 |
+
"learning_rate": 5.804000000000001e-06,
|
| 879 |
+
"loss": 0.2035,
|
| 880 |
+
"step": 14200
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 16.82,
|
| 884 |
+
"learning_rate": 5.704000000000001e-06,
|
| 885 |
+
"loss": 0.2044,
|
| 886 |
+
"step": 14300
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 16.94,
|
| 890 |
+
"learning_rate": 5.604000000000001e-06,
|
| 891 |
+
"loss": 0.2055,
|
| 892 |
+
"step": 14400
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 17.06,
|
| 896 |
+
"learning_rate": 5.505000000000001e-06,
|
| 897 |
+
"loss": 0.2051,
|
| 898 |
+
"step": 14500
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 17.18,
|
| 902 |
+
"learning_rate": 5.405e-06,
|
| 903 |
+
"loss": 0.2114,
|
| 904 |
+
"step": 14600
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 17.29,
|
| 908 |
+
"learning_rate": 5.305e-06,
|
| 909 |
+
"loss": 0.2135,
|
| 910 |
+
"step": 14700
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 17.41,
|
| 914 |
+
"learning_rate": 5.205e-06,
|
| 915 |
+
"loss": 0.2079,
|
| 916 |
+
"step": 14800
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 17.53,
|
| 920 |
+
"learning_rate": 5.105e-06,
|
| 921 |
+
"loss": 0.183,
|
| 922 |
+
"step": 14900
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 17.65,
|
| 926 |
+
"learning_rate": 5.0049999999999995e-06,
|
| 927 |
+
"loss": 0.2098,
|
| 928 |
+
"step": 15000
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 17.65,
|
| 932 |
+
"eval_cer": 0.07985734222360055,
|
| 933 |
+
"eval_loss": 0.6368189454078674,
|
| 934 |
+
"eval_runtime": 7.9866,
|
| 935 |
+
"eval_samples_per_second": 37.563,
|
| 936 |
+
"eval_steps_per_second": 18.781,
|
| 937 |
+
"eval_wer": 0.24188034188034188,
|
| 938 |
+
"step": 15000
|
| 939 |
+
}
|
| 940 |
+
],
|
| 941 |
+
"max_steps": 20000,
|
| 942 |
+
"num_train_epochs": 24,
|
| 943 |
+
"total_flos": 1.2884780636980976e+19,
|
| 944 |
+
"trial_name": null,
|
| 945 |
+
"trial_params": null
|
| 946 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:79e1ada8254cb35f8a0b1e7aa66f544c0524985be7ec8d3107a77116a4f7d861
|
| 3 |
+
size 3451
|
vocab.json
ADDED
|
@@ -0,0 +1,50 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"</s>": 2,
|
| 3 |
+
"<pad>": 0,
|
| 4 |
+
"<s>": 1,
|
| 5 |
+
"<unk>": 3,
|
| 6 |
+
"a": 5,
|
| 7 |
+
"b": 7,
|
| 8 |
+
"c": 8,
|
| 9 |
+
"d": 10,
|
| 10 |
+
"e": 14,
|
| 11 |
+
"f": 17,
|
| 12 |
+
"g": 18,
|
| 13 |
+
"h": 20,
|
| 14 |
+
"i": 22,
|
| 15 |
+
"k": 24,
|
| 16 |
+
"l": 25,
|
| 17 |
+
"m": 26,
|
| 18 |
+
"n": 27,
|
| 19 |
+
"o": 28,
|
| 20 |
+
"p": 30,
|
| 21 |
+
"r": 31,
|
| 22 |
+
"s": 32,
|
| 23 |
+
"t": 35,
|
| 24 |
+
"u": 38,
|
| 25 |
+
"w": 40,
|
| 26 |
+
"x": 41,
|
| 27 |
+
"y": 42,
|
| 28 |
+
"z": 43,
|
| 29 |
+
"|": 4,
|
| 30 |
+
"ā": 6,
|
| 31 |
+
"č": 9,
|
| 32 |
+
"ē": 15,
|
| 33 |
+
"ġ": 19,
|
| 34 |
+
"ī": 23,
|
| 35 |
+
"ō": 29,
|
| 36 |
+
"š": 33,
|
| 37 |
+
"ū": 39,
|
| 38 |
+
"ž": 44,
|
| 39 |
+
"ə": 16,
|
| 40 |
+
"ʔ": 47,
|
| 41 |
+
"ʕ": 46,
|
| 42 |
+
"ḍ": 12,
|
| 43 |
+
"ḏ": 11,
|
| 44 |
+
"ḓ": 13,
|
| 45 |
+
"ḥ": 21,
|
| 46 |
+
"ṣ": 34,
|
| 47 |
+
"ṭ": 37,
|
| 48 |
+
"ṯ": 36,
|
| 49 |
+
"ẓ": 45
|
| 50 |
+
}
|