Commit ·
288c286
1
Parent(s): d20b3ce
Add wav2vec fine-tuned model files
Browse files- added_tokens.json +4 -0
- checkpoint-350/added_tokens.json +4 -0
- checkpoint-350/config.json +108 -0
- checkpoint-350/model.safetensors +3 -0
- checkpoint-350/optimizer.pt +3 -0
- checkpoint-350/preprocessor_config.json +11 -0
- checkpoint-350/rng_state.pth +3 -0
- checkpoint-350/scaler.pt +3 -0
- checkpoint-350/scheduler.pt +3 -0
- checkpoint-350/special_tokens_map.json +28 -0
- checkpoint-350/tokenizer_config.json +53 -0
- checkpoint-350/trainer_state.json +279 -0
- checkpoint-350/training_args.bin +3 -0
- checkpoint-350/vocab.json +31 -0
- checkpoint-375/added_tokens.json +4 -0
- checkpoint-375/config.json +108 -0
- checkpoint-375/model.safetensors +3 -0
- checkpoint-375/optimizer.pt +3 -0
- checkpoint-375/preprocessor_config.json +11 -0
- checkpoint-375/rng_state.pth +3 -0
- checkpoint-375/scaler.pt +3 -0
- checkpoint-375/scheduler.pt +3 -0
- checkpoint-375/special_tokens_map.json +28 -0
- checkpoint-375/tokenizer_config.json +53 -0
- checkpoint-375/trainer_state.json +293 -0
- checkpoint-375/training_args.bin +3 -0
- checkpoint-375/vocab.json +31 -0
- config.json +108 -0
- evaluation_results.csv +57 -0
- model.safetensors +3 -0
- preprocessor_config.json +11 -0
- special_tokens_map.json +28 -0
- tokenizer_config.json +53 -0
- training_log.tsv +26 -0
- vocab.json +31 -0
added_tokens.json
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"</s>": 30,
|
| 3 |
+
"<s>": 29
|
| 4 |
+
}
|
checkpoint-350/added_tokens.json
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"</s>": 30,
|
| 3 |
+
"<s>": 29
|
| 4 |
+
}
|
checkpoint-350/config.json
ADDED
|
@@ -0,0 +1,108 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"adapter_attn_dim": null,
|
| 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": 0.0,
|
| 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": false,
|
| 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": 0.0,
|
| 53 |
+
"feat_quantizer_dropout": 0.0,
|
| 54 |
+
"final_dropout": 0.0,
|
| 55 |
+
"gradient_checkpointing": false,
|
| 56 |
+
"hidden_act": "gelu",
|
| 57 |
+
"hidden_dropout": 0.0,
|
| 58 |
+
"hidden_size": 1024,
|
| 59 |
+
"initializer_range": 0.02,
|
| 60 |
+
"intermediate_size": 4096,
|
| 61 |
+
"layer_norm_eps": 1e-05,
|
| 62 |
+
"layerdrop": 0.0,
|
| 63 |
+
"mask_feature_length": 10,
|
| 64 |
+
"mask_feature_min_masks": 0,
|
| 65 |
+
"mask_feature_prob": 0.0,
|
| 66 |
+
"mask_time_length": 10,
|
| 67 |
+
"mask_time_min_masks": 2,
|
| 68 |
+
"mask_time_prob": 0.0,
|
| 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": 28,
|
| 81 |
+
"proj_codevector_dim": 768,
|
| 82 |
+
"tdnn_dilation": [
|
| 83 |
+
1,
|
| 84 |
+
2,
|
| 85 |
+
3,
|
| 86 |
+
1,
|
| 87 |
+
1
|
| 88 |
+
],
|
| 89 |
+
"tdnn_dim": [
|
| 90 |
+
512,
|
| 91 |
+
512,
|
| 92 |
+
512,
|
| 93 |
+
512,
|
| 94 |
+
1500
|
| 95 |
+
],
|
| 96 |
+
"tdnn_kernel": [
|
| 97 |
+
5,
|
| 98 |
+
3,
|
| 99 |
+
3,
|
| 100 |
+
1,
|
| 101 |
+
1
|
| 102 |
+
],
|
| 103 |
+
"torch_dtype": "float32",
|
| 104 |
+
"transformers_version": "4.55.4",
|
| 105 |
+
"use_weighted_layer_sum": false,
|
| 106 |
+
"vocab_size": 31,
|
| 107 |
+
"xvector_output_dim": 512
|
| 108 |
+
}
|
checkpoint-350/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ed5f427487a4f75146b4e9dc480494f8cfaa16c8507284e695d3b23b353eb34c
|
| 3 |
+
size 1261934580
|
checkpoint-350/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a8f716e20a540ad57a89a97cf60f2059a70665e1d6504d27ca6d95bdde5765f5
|
| 3 |
+
size 806521803
|
checkpoint-350/preprocessor_config.json
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
+
"return_tensors": "np",
|
| 10 |
+
"sampling_rate": 16000
|
| 11 |
+
}
|
checkpoint-350/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:123848c4e5af199b0a3023db0dc50b82d8d7f7a2d2e95b352af9861d5d70da79
|
| 3 |
+
size 14645
|
checkpoint-350/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eca071cf113770e6722dad4a3d7b2eab7ca077ae7e6cfa66af1ff8bc547d5284
|
| 3 |
+
size 1383
|
checkpoint-350/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e07e2030005258fea5b5de0d579feff99fe4606f17cd41417eb090d35fca8f99
|
| 3 |
+
size 1465
|
checkpoint-350/special_tokens_map.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
{
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": true,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false
|
| 9 |
+
},
|
| 10 |
+
{
|
| 11 |
+
"content": "</s>",
|
| 12 |
+
"lstrip": false,
|
| 13 |
+
"normalized": true,
|
| 14 |
+
"rstrip": false,
|
| 15 |
+
"single_word": false
|
| 16 |
+
}
|
| 17 |
+
],
|
| 18 |
+
"bos_token": "<s>",
|
| 19 |
+
"eos_token": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false
|
| 25 |
+
},
|
| 26 |
+
"pad_token": "[PAD]",
|
| 27 |
+
"unk_token": "[UNK]"
|
| 28 |
+
}
|
checkpoint-350/tokenizer_config.json
ADDED
|
@@ -0,0 +1,53 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"27": {
|
| 4 |
+
"content": "[UNK]",
|
| 5 |
+
"lstrip": true,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": true,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": false
|
| 10 |
+
},
|
| 11 |
+
"28": {
|
| 12 |
+
"content": "[PAD]",
|
| 13 |
+
"lstrip": true,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": true,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": false
|
| 18 |
+
},
|
| 19 |
+
"29": {
|
| 20 |
+
"content": "<s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"30": {
|
| 28 |
+
"content": "</s>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
"additional_special_tokens": [
|
| 37 |
+
"<s>",
|
| 38 |
+
"</s>"
|
| 39 |
+
],
|
| 40 |
+
"bos_token": "<s>",
|
| 41 |
+
"clean_up_tokenization_spaces": false,
|
| 42 |
+
"do_lower_case": false,
|
| 43 |
+
"eos_token": "</s>",
|
| 44 |
+
"extra_special_tokens": {},
|
| 45 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 46 |
+
"pad_token": "[PAD]",
|
| 47 |
+
"processor_class": "Wav2Vec2Processor",
|
| 48 |
+
"replace_word_delimiter_char": " ",
|
| 49 |
+
"target_lang": null,
|
| 50 |
+
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
| 51 |
+
"unk_token": "[UNK]",
|
| 52 |
+
"word_delimiter_token": "|"
|
| 53 |
+
}
|
checkpoint-350/trainer_state.json
ADDED
|
@@ -0,0 +1,279 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 14.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 350,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.4032258064516129,
|
| 14 |
+
"grad_norm": 15.945989608764648,
|
| 15 |
+
"learning_rate": 5.4e-07,
|
| 16 |
+
"loss": 2.942,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.8064516129032258,
|
| 21 |
+
"grad_norm": 15.050374984741211,
|
| 22 |
+
"learning_rate": 1.14e-06,
|
| 23 |
+
"loss": 3.345,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.2016129032258065,
|
| 28 |
+
"grad_norm": 9.487156867980957,
|
| 29 |
+
"learning_rate": 1.74e-06,
|
| 30 |
+
"loss": 2.6443,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 1.6048387096774195,
|
| 35 |
+
"grad_norm": 14.825271606445312,
|
| 36 |
+
"learning_rate": 2.34e-06,
|
| 37 |
+
"loss": 3.0528,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 2.0,
|
| 42 |
+
"grad_norm": 15.298517227172852,
|
| 43 |
+
"learning_rate": 2.9400000000000002e-06,
|
| 44 |
+
"loss": 2.2951,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 2.403225806451613,
|
| 49 |
+
"grad_norm": 13.529952049255371,
|
| 50 |
+
"learning_rate": 3.54e-06,
|
| 51 |
+
"loss": 2.3061,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 2.806451612903226,
|
| 56 |
+
"grad_norm": 12.780255317687988,
|
| 57 |
+
"learning_rate": 4.14e-06,
|
| 58 |
+
"loss": 1.828,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 3.2016129032258065,
|
| 63 |
+
"grad_norm": 4.392550468444824,
|
| 64 |
+
"learning_rate": 4.74e-06,
|
| 65 |
+
"loss": 1.7883,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 3.6048387096774195,
|
| 70 |
+
"grad_norm": 6.868764400482178,
|
| 71 |
+
"learning_rate": 5.34e-06,
|
| 72 |
+
"loss": 1.3883,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 4.0,
|
| 77 |
+
"grad_norm": 11.529573440551758,
|
| 78 |
+
"learning_rate": 5.940000000000001e-06,
|
| 79 |
+
"loss": 1.2804,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 4.403225806451613,
|
| 84 |
+
"grad_norm": 3.888104200363159,
|
| 85 |
+
"learning_rate": 6.54e-06,
|
| 86 |
+
"loss": 1.1184,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 4.806451612903226,
|
| 91 |
+
"grad_norm": 5.543831825256348,
|
| 92 |
+
"learning_rate": 7.14e-06,
|
| 93 |
+
"loss": 0.8838,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 5.201612903225806,
|
| 98 |
+
"grad_norm": 3.6435844898223877,
|
| 99 |
+
"learning_rate": 7.74e-06,
|
| 100 |
+
"loss": 0.7359,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 5.604838709677419,
|
| 105 |
+
"grad_norm": 3.873931407928467,
|
| 106 |
+
"learning_rate": 8.340000000000001e-06,
|
| 107 |
+
"loss": 0.6677,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 6.0,
|
| 112 |
+
"grad_norm": 6.526176929473877,
|
| 113 |
+
"learning_rate": 8.939999999999999e-06,
|
| 114 |
+
"loss": 0.5343,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 6.403225806451613,
|
| 119 |
+
"grad_norm": 2.3237783908843994,
|
| 120 |
+
"learning_rate": 9.54e-06,
|
| 121 |
+
"loss": 0.4596,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 6.806451612903226,
|
| 126 |
+
"grad_norm": 3.2708520889282227,
|
| 127 |
+
"learning_rate": 1.0140000000000001e-05,
|
| 128 |
+
"loss": 0.371,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 7.201612903225806,
|
| 133 |
+
"grad_norm": 1.420652985572815,
|
| 134 |
+
"learning_rate": 1.074e-05,
|
| 135 |
+
"loss": 0.397,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 7.604838709677419,
|
| 140 |
+
"grad_norm": 2.8168821334838867,
|
| 141 |
+
"learning_rate": 1.134e-05,
|
| 142 |
+
"loss": 0.2335,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 8.0,
|
| 147 |
+
"grad_norm": 0.872466504573822,
|
| 148 |
+
"learning_rate": 1.1940000000000001e-05,
|
| 149 |
+
"loss": 0.2535,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 8.403225806451612,
|
| 154 |
+
"grad_norm": 1.6869771480560303,
|
| 155 |
+
"learning_rate": 1.254e-05,
|
| 156 |
+
"loss": 0.1642,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 8.806451612903226,
|
| 161 |
+
"grad_norm": 8.301424026489258,
|
| 162 |
+
"learning_rate": 1.314e-05,
|
| 163 |
+
"loss": 0.112,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 9.201612903225806,
|
| 168 |
+
"grad_norm": 0.7436397671699524,
|
| 169 |
+
"learning_rate": 1.374e-05,
|
| 170 |
+
"loss": 0.1257,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 9.60483870967742,
|
| 175 |
+
"grad_norm": 1.891860842704773,
|
| 176 |
+
"learning_rate": 1.434e-05,
|
| 177 |
+
"loss": 0.0623,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 10.0,
|
| 182 |
+
"grad_norm": 0.5202131867408752,
|
| 183 |
+
"learning_rate": 1.4940000000000001e-05,
|
| 184 |
+
"loss": 0.0572,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 10.403225806451612,
|
| 189 |
+
"grad_norm": 0.0677497610449791,
|
| 190 |
+
"learning_rate": 1.554e-05,
|
| 191 |
+
"loss": 0.0212,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 10.806451612903226,
|
| 196 |
+
"grad_norm": 0.07801195234060287,
|
| 197 |
+
"learning_rate": 1.614e-05,
|
| 198 |
+
"loss": 0.0243,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 11.201612903225806,
|
| 203 |
+
"grad_norm": 0.11515898257493973,
|
| 204 |
+
"learning_rate": 1.6740000000000002e-05,
|
| 205 |
+
"loss": 0.0321,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 11.60483870967742,
|
| 210 |
+
"grad_norm": 0.11826858669519424,
|
| 211 |
+
"learning_rate": 1.734e-05,
|
| 212 |
+
"loss": 0.0254,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 12.0,
|
| 217 |
+
"grad_norm": 0.13558819890022278,
|
| 218 |
+
"learning_rate": 1.794e-05,
|
| 219 |
+
"loss": 0.014,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 12.403225806451612,
|
| 224 |
+
"grad_norm": 0.03777763620018959,
|
| 225 |
+
"learning_rate": 1.854e-05,
|
| 226 |
+
"loss": 0.0301,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 12.806451612903226,
|
| 231 |
+
"grad_norm": 0.0405518114566803,
|
| 232 |
+
"learning_rate": 1.914e-05,
|
| 233 |
+
"loss": 0.0027,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 13.201612903225806,
|
| 238 |
+
"grad_norm": 0.03762541711330414,
|
| 239 |
+
"learning_rate": 1.974e-05,
|
| 240 |
+
"loss": 0.0026,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 13.60483870967742,
|
| 245 |
+
"grad_norm": 0.06563286483287811,
|
| 246 |
+
"learning_rate": 2.0340000000000002e-05,
|
| 247 |
+
"loss": 0.0279,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 14.0,
|
| 252 |
+
"grad_norm": 0.12380703538656235,
|
| 253 |
+
"learning_rate": 2.094e-05,
|
| 254 |
+
"loss": 0.0063,
|
| 255 |
+
"step": 350
|
| 256 |
+
}
|
| 257 |
+
],
|
| 258 |
+
"logging_steps": 10,
|
| 259 |
+
"max_steps": 375,
|
| 260 |
+
"num_input_tokens_seen": 0,
|
| 261 |
+
"num_train_epochs": 15,
|
| 262 |
+
"save_steps": 50,
|
| 263 |
+
"stateful_callbacks": {
|
| 264 |
+
"TrainerControl": {
|
| 265 |
+
"args": {
|
| 266 |
+
"should_epoch_stop": false,
|
| 267 |
+
"should_evaluate": false,
|
| 268 |
+
"should_log": false,
|
| 269 |
+
"should_save": true,
|
| 270 |
+
"should_training_stop": false
|
| 271 |
+
},
|
| 272 |
+
"attributes": {}
|
| 273 |
+
}
|
| 274 |
+
},
|
| 275 |
+
"total_flos": 1.651986705536446e+17,
|
| 276 |
+
"train_batch_size": 4,
|
| 277 |
+
"trial_name": null,
|
| 278 |
+
"trial_params": null
|
| 279 |
+
}
|
checkpoint-350/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:65214c45db9799ff8e38f62a8a12f53de739d750564cf70464decc61c84f64ab
|
| 3 |
+
size 5777
|
checkpoint-350/vocab.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"[PAD]": 28,
|
| 3 |
+
"[UNK]": 27,
|
| 4 |
+
"a": 1,
|
| 5 |
+
"b": 2,
|
| 6 |
+
"c": 3,
|
| 7 |
+
"d": 4,
|
| 8 |
+
"e": 5,
|
| 9 |
+
"f": 6,
|
| 10 |
+
"g": 7,
|
| 11 |
+
"h": 8,
|
| 12 |
+
"i": 9,
|
| 13 |
+
"j": 10,
|
| 14 |
+
"k": 11,
|
| 15 |
+
"l": 12,
|
| 16 |
+
"m": 13,
|
| 17 |
+
"n": 14,
|
| 18 |
+
"o": 15,
|
| 19 |
+
"p": 16,
|
| 20 |
+
"q": 17,
|
| 21 |
+
"r": 18,
|
| 22 |
+
"s": 19,
|
| 23 |
+
"t": 20,
|
| 24 |
+
"u": 21,
|
| 25 |
+
"v": 22,
|
| 26 |
+
"w": 23,
|
| 27 |
+
"x": 24,
|
| 28 |
+
"y": 25,
|
| 29 |
+
"z": 26,
|
| 30 |
+
"|": 0
|
| 31 |
+
}
|
checkpoint-375/added_tokens.json
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"</s>": 30,
|
| 3 |
+
"<s>": 29
|
| 4 |
+
}
|
checkpoint-375/config.json
ADDED
|
@@ -0,0 +1,108 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"adapter_attn_dim": null,
|
| 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": 0.0,
|
| 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": false,
|
| 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": 0.0,
|
| 53 |
+
"feat_quantizer_dropout": 0.0,
|
| 54 |
+
"final_dropout": 0.0,
|
| 55 |
+
"gradient_checkpointing": false,
|
| 56 |
+
"hidden_act": "gelu",
|
| 57 |
+
"hidden_dropout": 0.0,
|
| 58 |
+
"hidden_size": 1024,
|
| 59 |
+
"initializer_range": 0.02,
|
| 60 |
+
"intermediate_size": 4096,
|
| 61 |
+
"layer_norm_eps": 1e-05,
|
| 62 |
+
"layerdrop": 0.0,
|
| 63 |
+
"mask_feature_length": 10,
|
| 64 |
+
"mask_feature_min_masks": 0,
|
| 65 |
+
"mask_feature_prob": 0.0,
|
| 66 |
+
"mask_time_length": 10,
|
| 67 |
+
"mask_time_min_masks": 2,
|
| 68 |
+
"mask_time_prob": 0.0,
|
| 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": 28,
|
| 81 |
+
"proj_codevector_dim": 768,
|
| 82 |
+
"tdnn_dilation": [
|
| 83 |
+
1,
|
| 84 |
+
2,
|
| 85 |
+
3,
|
| 86 |
+
1,
|
| 87 |
+
1
|
| 88 |
+
],
|
| 89 |
+
"tdnn_dim": [
|
| 90 |
+
512,
|
| 91 |
+
512,
|
| 92 |
+
512,
|
| 93 |
+
512,
|
| 94 |
+
1500
|
| 95 |
+
],
|
| 96 |
+
"tdnn_kernel": [
|
| 97 |
+
5,
|
| 98 |
+
3,
|
| 99 |
+
3,
|
| 100 |
+
1,
|
| 101 |
+
1
|
| 102 |
+
],
|
| 103 |
+
"torch_dtype": "float32",
|
| 104 |
+
"transformers_version": "4.55.4",
|
| 105 |
+
"use_weighted_layer_sum": false,
|
| 106 |
+
"vocab_size": 31,
|
| 107 |
+
"xvector_output_dim": 512
|
| 108 |
+
}
|
checkpoint-375/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4a78453f5c1780262440dd4fd6465ec500764cf4eccb2e7bad9a7e7f5ffcc2e5
|
| 3 |
+
size 1261934580
|
checkpoint-375/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d46c229febe3e84451b6b8ab287e9ba97b56be3c714e4eddacaf6087128d1b55
|
| 3 |
+
size 806521803
|
checkpoint-375/preprocessor_config.json
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
+
"return_tensors": "np",
|
| 10 |
+
"sampling_rate": 16000
|
| 11 |
+
}
|
checkpoint-375/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:214119e423d32a6bd19a61d9c6c73a5bb02729dfeacd793faf03518656d409c4
|
| 3 |
+
size 14645
|
checkpoint-375/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ffa100fb89452a105fbe870f5976239d27ca580780da5887426b11aec30bf36e
|
| 3 |
+
size 1383
|
checkpoint-375/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f0d44bba9ff495f2f261cd3e092af3e4a799899ba7c44ee5b3bc6534485b82ea
|
| 3 |
+
size 1465
|
checkpoint-375/special_tokens_map.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
{
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": true,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false
|
| 9 |
+
},
|
| 10 |
+
{
|
| 11 |
+
"content": "</s>",
|
| 12 |
+
"lstrip": false,
|
| 13 |
+
"normalized": true,
|
| 14 |
+
"rstrip": false,
|
| 15 |
+
"single_word": false
|
| 16 |
+
}
|
| 17 |
+
],
|
| 18 |
+
"bos_token": "<s>",
|
| 19 |
+
"eos_token": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false
|
| 25 |
+
},
|
| 26 |
+
"pad_token": "[PAD]",
|
| 27 |
+
"unk_token": "[UNK]"
|
| 28 |
+
}
|
checkpoint-375/tokenizer_config.json
ADDED
|
@@ -0,0 +1,53 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"27": {
|
| 4 |
+
"content": "[UNK]",
|
| 5 |
+
"lstrip": true,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": true,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": false
|
| 10 |
+
},
|
| 11 |
+
"28": {
|
| 12 |
+
"content": "[PAD]",
|
| 13 |
+
"lstrip": true,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": true,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": false
|
| 18 |
+
},
|
| 19 |
+
"29": {
|
| 20 |
+
"content": "<s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"30": {
|
| 28 |
+
"content": "</s>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
"additional_special_tokens": [
|
| 37 |
+
"<s>",
|
| 38 |
+
"</s>"
|
| 39 |
+
],
|
| 40 |
+
"bos_token": "<s>",
|
| 41 |
+
"clean_up_tokenization_spaces": false,
|
| 42 |
+
"do_lower_case": false,
|
| 43 |
+
"eos_token": "</s>",
|
| 44 |
+
"extra_special_tokens": {},
|
| 45 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 46 |
+
"pad_token": "[PAD]",
|
| 47 |
+
"processor_class": "Wav2Vec2Processor",
|
| 48 |
+
"replace_word_delimiter_char": " ",
|
| 49 |
+
"target_lang": null,
|
| 50 |
+
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
| 51 |
+
"unk_token": "[UNK]",
|
| 52 |
+
"word_delimiter_token": "|"
|
| 53 |
+
}
|
checkpoint-375/trainer_state.json
ADDED
|
@@ -0,0 +1,293 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 15.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 375,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.4032258064516129,
|
| 14 |
+
"grad_norm": 15.945989608764648,
|
| 15 |
+
"learning_rate": 5.4e-07,
|
| 16 |
+
"loss": 2.942,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.8064516129032258,
|
| 21 |
+
"grad_norm": 15.050374984741211,
|
| 22 |
+
"learning_rate": 1.14e-06,
|
| 23 |
+
"loss": 3.345,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.2016129032258065,
|
| 28 |
+
"grad_norm": 9.487156867980957,
|
| 29 |
+
"learning_rate": 1.74e-06,
|
| 30 |
+
"loss": 2.6443,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 1.6048387096774195,
|
| 35 |
+
"grad_norm": 14.825271606445312,
|
| 36 |
+
"learning_rate": 2.34e-06,
|
| 37 |
+
"loss": 3.0528,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 2.0,
|
| 42 |
+
"grad_norm": 15.298517227172852,
|
| 43 |
+
"learning_rate": 2.9400000000000002e-06,
|
| 44 |
+
"loss": 2.2951,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 2.403225806451613,
|
| 49 |
+
"grad_norm": 13.529952049255371,
|
| 50 |
+
"learning_rate": 3.54e-06,
|
| 51 |
+
"loss": 2.3061,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 2.806451612903226,
|
| 56 |
+
"grad_norm": 12.780255317687988,
|
| 57 |
+
"learning_rate": 4.14e-06,
|
| 58 |
+
"loss": 1.828,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 3.2016129032258065,
|
| 63 |
+
"grad_norm": 4.392550468444824,
|
| 64 |
+
"learning_rate": 4.74e-06,
|
| 65 |
+
"loss": 1.7883,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 3.6048387096774195,
|
| 70 |
+
"grad_norm": 6.868764400482178,
|
| 71 |
+
"learning_rate": 5.34e-06,
|
| 72 |
+
"loss": 1.3883,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 4.0,
|
| 77 |
+
"grad_norm": 11.529573440551758,
|
| 78 |
+
"learning_rate": 5.940000000000001e-06,
|
| 79 |
+
"loss": 1.2804,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 4.403225806451613,
|
| 84 |
+
"grad_norm": 3.888104200363159,
|
| 85 |
+
"learning_rate": 6.54e-06,
|
| 86 |
+
"loss": 1.1184,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 4.806451612903226,
|
| 91 |
+
"grad_norm": 5.543831825256348,
|
| 92 |
+
"learning_rate": 7.14e-06,
|
| 93 |
+
"loss": 0.8838,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 5.201612903225806,
|
| 98 |
+
"grad_norm": 3.6435844898223877,
|
| 99 |
+
"learning_rate": 7.74e-06,
|
| 100 |
+
"loss": 0.7359,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 5.604838709677419,
|
| 105 |
+
"grad_norm": 3.873931407928467,
|
| 106 |
+
"learning_rate": 8.340000000000001e-06,
|
| 107 |
+
"loss": 0.6677,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 6.0,
|
| 112 |
+
"grad_norm": 6.526176929473877,
|
| 113 |
+
"learning_rate": 8.939999999999999e-06,
|
| 114 |
+
"loss": 0.5343,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 6.403225806451613,
|
| 119 |
+
"grad_norm": 2.3237783908843994,
|
| 120 |
+
"learning_rate": 9.54e-06,
|
| 121 |
+
"loss": 0.4596,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 6.806451612903226,
|
| 126 |
+
"grad_norm": 3.2708520889282227,
|
| 127 |
+
"learning_rate": 1.0140000000000001e-05,
|
| 128 |
+
"loss": 0.371,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 7.201612903225806,
|
| 133 |
+
"grad_norm": 1.420652985572815,
|
| 134 |
+
"learning_rate": 1.074e-05,
|
| 135 |
+
"loss": 0.397,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 7.604838709677419,
|
| 140 |
+
"grad_norm": 2.8168821334838867,
|
| 141 |
+
"learning_rate": 1.134e-05,
|
| 142 |
+
"loss": 0.2335,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 8.0,
|
| 147 |
+
"grad_norm": 0.872466504573822,
|
| 148 |
+
"learning_rate": 1.1940000000000001e-05,
|
| 149 |
+
"loss": 0.2535,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 8.403225806451612,
|
| 154 |
+
"grad_norm": 1.6869771480560303,
|
| 155 |
+
"learning_rate": 1.254e-05,
|
| 156 |
+
"loss": 0.1642,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 8.806451612903226,
|
| 161 |
+
"grad_norm": 8.301424026489258,
|
| 162 |
+
"learning_rate": 1.314e-05,
|
| 163 |
+
"loss": 0.112,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 9.201612903225806,
|
| 168 |
+
"grad_norm": 0.7436397671699524,
|
| 169 |
+
"learning_rate": 1.374e-05,
|
| 170 |
+
"loss": 0.1257,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 9.60483870967742,
|
| 175 |
+
"grad_norm": 1.891860842704773,
|
| 176 |
+
"learning_rate": 1.434e-05,
|
| 177 |
+
"loss": 0.0623,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 10.0,
|
| 182 |
+
"grad_norm": 0.5202131867408752,
|
| 183 |
+
"learning_rate": 1.4940000000000001e-05,
|
| 184 |
+
"loss": 0.0572,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 10.403225806451612,
|
| 189 |
+
"grad_norm": 0.0677497610449791,
|
| 190 |
+
"learning_rate": 1.554e-05,
|
| 191 |
+
"loss": 0.0212,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 10.806451612903226,
|
| 196 |
+
"grad_norm": 0.07801195234060287,
|
| 197 |
+
"learning_rate": 1.614e-05,
|
| 198 |
+
"loss": 0.0243,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 11.201612903225806,
|
| 203 |
+
"grad_norm": 0.11515898257493973,
|
| 204 |
+
"learning_rate": 1.6740000000000002e-05,
|
| 205 |
+
"loss": 0.0321,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 11.60483870967742,
|
| 210 |
+
"grad_norm": 0.11826858669519424,
|
| 211 |
+
"learning_rate": 1.734e-05,
|
| 212 |
+
"loss": 0.0254,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 12.0,
|
| 217 |
+
"grad_norm": 0.13558819890022278,
|
| 218 |
+
"learning_rate": 1.794e-05,
|
| 219 |
+
"loss": 0.014,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 12.403225806451612,
|
| 224 |
+
"grad_norm": 0.03777763620018959,
|
| 225 |
+
"learning_rate": 1.854e-05,
|
| 226 |
+
"loss": 0.0301,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 12.806451612903226,
|
| 231 |
+
"grad_norm": 0.0405518114566803,
|
| 232 |
+
"learning_rate": 1.914e-05,
|
| 233 |
+
"loss": 0.0027,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 13.201612903225806,
|
| 238 |
+
"grad_norm": 0.03762541711330414,
|
| 239 |
+
"learning_rate": 1.974e-05,
|
| 240 |
+
"loss": 0.0026,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 13.60483870967742,
|
| 245 |
+
"grad_norm": 0.06563286483287811,
|
| 246 |
+
"learning_rate": 2.0340000000000002e-05,
|
| 247 |
+
"loss": 0.0279,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 14.0,
|
| 252 |
+
"grad_norm": 0.12380703538656235,
|
| 253 |
+
"learning_rate": 2.094e-05,
|
| 254 |
+
"loss": 0.0063,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 14.403225806451612,
|
| 259 |
+
"grad_norm": 0.015481448732316494,
|
| 260 |
+
"learning_rate": 2.154e-05,
|
| 261 |
+
"loss": 0.0021,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 14.806451612903226,
|
| 266 |
+
"grad_norm": 0.036212582141160965,
|
| 267 |
+
"learning_rate": 2.214e-05,
|
| 268 |
+
"loss": 0.0019,
|
| 269 |
+
"step": 370
|
| 270 |
+
}
|
| 271 |
+
],
|
| 272 |
+
"logging_steps": 10,
|
| 273 |
+
"max_steps": 375,
|
| 274 |
+
"num_input_tokens_seen": 0,
|
| 275 |
+
"num_train_epochs": 15,
|
| 276 |
+
"save_steps": 50,
|
| 277 |
+
"stateful_callbacks": {
|
| 278 |
+
"TrainerControl": {
|
| 279 |
+
"args": {
|
| 280 |
+
"should_epoch_stop": false,
|
| 281 |
+
"should_evaluate": false,
|
| 282 |
+
"should_log": false,
|
| 283 |
+
"should_save": true,
|
| 284 |
+
"should_training_stop": true
|
| 285 |
+
},
|
| 286 |
+
"attributes": {}
|
| 287 |
+
}
|
| 288 |
+
},
|
| 289 |
+
"total_flos": 1.770072662194541e+17,
|
| 290 |
+
"train_batch_size": 4,
|
| 291 |
+
"trial_name": null,
|
| 292 |
+
"trial_params": null
|
| 293 |
+
}
|
checkpoint-375/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:65214c45db9799ff8e38f62a8a12f53de739d750564cf70464decc61c84f64ab
|
| 3 |
+
size 5777
|
checkpoint-375/vocab.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"[PAD]": 28,
|
| 3 |
+
"[UNK]": 27,
|
| 4 |
+
"a": 1,
|
| 5 |
+
"b": 2,
|
| 6 |
+
"c": 3,
|
| 7 |
+
"d": 4,
|
| 8 |
+
"e": 5,
|
| 9 |
+
"f": 6,
|
| 10 |
+
"g": 7,
|
| 11 |
+
"h": 8,
|
| 12 |
+
"i": 9,
|
| 13 |
+
"j": 10,
|
| 14 |
+
"k": 11,
|
| 15 |
+
"l": 12,
|
| 16 |
+
"m": 13,
|
| 17 |
+
"n": 14,
|
| 18 |
+
"o": 15,
|
| 19 |
+
"p": 16,
|
| 20 |
+
"q": 17,
|
| 21 |
+
"r": 18,
|
| 22 |
+
"s": 19,
|
| 23 |
+
"t": 20,
|
| 24 |
+
"u": 21,
|
| 25 |
+
"v": 22,
|
| 26 |
+
"w": 23,
|
| 27 |
+
"x": 24,
|
| 28 |
+
"y": 25,
|
| 29 |
+
"z": 26,
|
| 30 |
+
"|": 0
|
| 31 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,108 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"adapter_attn_dim": null,
|
| 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": 0.0,
|
| 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": false,
|
| 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": 0.0,
|
| 53 |
+
"feat_quantizer_dropout": 0.0,
|
| 54 |
+
"final_dropout": 0.0,
|
| 55 |
+
"gradient_checkpointing": false,
|
| 56 |
+
"hidden_act": "gelu",
|
| 57 |
+
"hidden_dropout": 0.0,
|
| 58 |
+
"hidden_size": 1024,
|
| 59 |
+
"initializer_range": 0.02,
|
| 60 |
+
"intermediate_size": 4096,
|
| 61 |
+
"layer_norm_eps": 1e-05,
|
| 62 |
+
"layerdrop": 0.0,
|
| 63 |
+
"mask_feature_length": 10,
|
| 64 |
+
"mask_feature_min_masks": 0,
|
| 65 |
+
"mask_feature_prob": 0.0,
|
| 66 |
+
"mask_time_length": 10,
|
| 67 |
+
"mask_time_min_masks": 2,
|
| 68 |
+
"mask_time_prob": 0.0,
|
| 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": 28,
|
| 81 |
+
"proj_codevector_dim": 768,
|
| 82 |
+
"tdnn_dilation": [
|
| 83 |
+
1,
|
| 84 |
+
2,
|
| 85 |
+
3,
|
| 86 |
+
1,
|
| 87 |
+
1
|
| 88 |
+
],
|
| 89 |
+
"tdnn_dim": [
|
| 90 |
+
512,
|
| 91 |
+
512,
|
| 92 |
+
512,
|
| 93 |
+
512,
|
| 94 |
+
1500
|
| 95 |
+
],
|
| 96 |
+
"tdnn_kernel": [
|
| 97 |
+
5,
|
| 98 |
+
3,
|
| 99 |
+
3,
|
| 100 |
+
1,
|
| 101 |
+
1
|
| 102 |
+
],
|
| 103 |
+
"torch_dtype": "float32",
|
| 104 |
+
"transformers_version": "4.55.4",
|
| 105 |
+
"use_weighted_layer_sum": false,
|
| 106 |
+
"vocab_size": 31,
|
| 107 |
+
"xvector_output_dim": 512
|
| 108 |
+
}
|
evaluation_results.csv
ADDED
|
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
Reference,Prediction,WER,CER
|
| 2 |
+
wi,,1.0,1.0
|
| 3 |
+
wi,,1.0,1.0
|
| 4 |
+
wi,,1.0,1.0
|
| 5 |
+
wi,,1.0,1.0
|
| 6 |
+
wi,,1.0,1.0
|
| 7 |
+
wi,<unk>,1.0,2.5
|
| 8 |
+
wi,,1.0,1.0
|
| 9 |
+
wi,,1.0,1.0
|
| 10 |
+
wi,,1.0,1.0
|
| 11 |
+
wo,,1.0,1.0
|
| 12 |
+
wo,,1.0,1.0
|
| 13 |
+
wo,,1.0,1.0
|
| 14 |
+
wo,,1.0,1.0
|
| 15 |
+
wo,,1.0,1.0
|
| 16 |
+
wo,,1.0,1.0
|
| 17 |
+
wo,,1.0,1.0
|
| 18 |
+
wo,,1.0,1.0
|
| 19 |
+
wo,,1.0,1.0
|
| 20 |
+
wo,,1.0,1.0
|
| 21 |
+
wo,,1.0,1.0
|
| 22 |
+
wu,,1.0,1.0
|
| 23 |
+
wu,,1.0,1.0
|
| 24 |
+
wu,,1.0,1.0
|
| 25 |
+
wu,,1.0,1.0
|
| 26 |
+
wu,,1.0,1.0
|
| 27 |
+
wu,,1.0,1.0
|
| 28 |
+
wu,,1.0,1.0
|
| 29 |
+
wu,,1.0,1.0
|
| 30 |
+
wu,,1.0,1.0
|
| 31 |
+
ya,,1.0,1.0
|
| 32 |
+
ya,,1.0,1.0
|
| 33 |
+
ya,,1.0,1.0
|
| 34 |
+
ya,,1.0,1.0
|
| 35 |
+
ya,,1.0,1.0
|
| 36 |
+
ya,,1.0,1.0
|
| 37 |
+
ya,,1.0,1.0
|
| 38 |
+
ye,,1.0,1.0
|
| 39 |
+
ye,,1.0,1.0
|
| 40 |
+
ye,,1.0,1.0
|
| 41 |
+
ye,,1.0,1.0
|
| 42 |
+
ye,,1.0,1.0
|
| 43 |
+
yi,,1.0,1.0
|
| 44 |
+
yi,,1.0,1.0
|
| 45 |
+
yi,,1.0,1.0
|
| 46 |
+
yi,,1.0,1.0
|
| 47 |
+
yi,,1.0,1.0
|
| 48 |
+
yo,,1.0,1.0
|
| 49 |
+
yo,,1.0,1.0
|
| 50 |
+
yo,,1.0,1.0
|
| 51 |
+
yo,</s>,1.0,2.0
|
| 52 |
+
yo,,1.0,1.0
|
| 53 |
+
yu,,1.0,1.0
|
| 54 |
+
yu,,1.0,1.0
|
| 55 |
+
yu,,1.0,1.0
|
| 56 |
+
yu,,1.0,1.0
|
| 57 |
+
yu,,1.0,1.0
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4a78453f5c1780262440dd4fd6465ec500764cf4eccb2e7bad9a7e7f5ffcc2e5
|
| 3 |
+
size 1261934580
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
+
"return_tensors": "np",
|
| 10 |
+
"sampling_rate": 16000
|
| 11 |
+
}
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
{
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": true,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false
|
| 9 |
+
},
|
| 10 |
+
{
|
| 11 |
+
"content": "</s>",
|
| 12 |
+
"lstrip": false,
|
| 13 |
+
"normalized": true,
|
| 14 |
+
"rstrip": false,
|
| 15 |
+
"single_word": false
|
| 16 |
+
}
|
| 17 |
+
],
|
| 18 |
+
"bos_token": "<s>",
|
| 19 |
+
"eos_token": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false
|
| 25 |
+
},
|
| 26 |
+
"pad_token": "[PAD]",
|
| 27 |
+
"unk_token": "[UNK]"
|
| 28 |
+
}
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,53 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"27": {
|
| 4 |
+
"content": "[UNK]",
|
| 5 |
+
"lstrip": true,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": true,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": false
|
| 10 |
+
},
|
| 11 |
+
"28": {
|
| 12 |
+
"content": "[PAD]",
|
| 13 |
+
"lstrip": true,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": true,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": false
|
| 18 |
+
},
|
| 19 |
+
"29": {
|
| 20 |
+
"content": "<s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"30": {
|
| 28 |
+
"content": "</s>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
"additional_special_tokens": [
|
| 37 |
+
"<s>",
|
| 38 |
+
"</s>"
|
| 39 |
+
],
|
| 40 |
+
"bos_token": "<s>",
|
| 41 |
+
"clean_up_tokenization_spaces": false,
|
| 42 |
+
"do_lower_case": false,
|
| 43 |
+
"eos_token": "</s>",
|
| 44 |
+
"extra_special_tokens": {},
|
| 45 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 46 |
+
"pad_token": "[PAD]",
|
| 47 |
+
"processor_class": "Wav2Vec2Processor",
|
| 48 |
+
"replace_word_delimiter_char": " ",
|
| 49 |
+
"target_lang": null,
|
| 50 |
+
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
| 51 |
+
"unk_token": "[UNK]",
|
| 52 |
+
"word_delimiter_token": "|"
|
| 53 |
+
}
|
training_log.tsv
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
timestamp step epoch train_loss grad_norm learning_rate
|
| 2 |
+
2025-09-21T21:41:21.703893 10 0.32 903.355000 614.498596 1.00e-07
|
| 3 |
+
2025-09-21T21:41:27.856246 20 0.65 902.448600 617.974670 5.00e-07
|
| 4 |
+
2025-09-21T21:41:33.781063 30 0.97 886.335200 578.719543 1.00e-06
|
| 5 |
+
2025-09-21T21:41:39.650969 40 1.29 941.081400 763.962463 1.50e-06
|
| 6 |
+
2025-09-21T21:41:45.448703 50 1.61 878.512000 745.441284 2.00e-06
|
| 7 |
+
2025-09-21T21:41:46.303569 50 1.61 NA NA NA
|
| 8 |
+
2025-09-21T21:41:53.272236 60 1.94 870.239400 805.721558 2.50e-06
|
| 9 |
+
2025-09-21T21:41:58.923912 70 2.26 883.066600 810.874512 3.00e-06
|
| 10 |
+
2025-09-21T21:42:04.647459 80 2.58 858.037100 876.961365 3.50e-06
|
| 11 |
+
2025-09-21T21:42:10.439181 90 2.90 879.551700 879.198303 4.00e-06
|
| 12 |
+
2025-09-21T21:42:16.088489 100 3.23 838.324500 1011.899841 4.50e-06
|
| 13 |
+
2025-09-21T21:42:16.862886 100 3.23 NA NA NA
|
| 14 |
+
2025-09-21T21:42:24.087341 110 3.55 789.837500 1295.429565 5.00e-06
|
| 15 |
+
2025-09-21T21:42:29.817688 120 3.87 745.176600 1823.199463 5.50e-06
|
| 16 |
+
2025-09-21T21:42:35.473868 130 4.19 701.951900 1844.962524 6.00e-06
|
| 17 |
+
2025-09-21T21:42:41.363224 140 4.52 661.328600 1836.961670 6.50e-06
|
| 18 |
+
2025-09-21T21:42:47.037840 150 4.84 558.101000 1857.315308 7.00e-06
|
| 19 |
+
2025-09-21T21:42:47.769415 150 4.84 NA NA NA
|
| 20 |
+
2025-09-21T21:42:54.652980 160 5.16 499.938000 2131.156982 7.50e-06
|
| 21 |
+
2025-09-21T21:43:00.440291 170 5.48 450.506500 1810.863647 8.00e-06
|
| 22 |
+
2025-09-21T21:43:06.233803 180 5.81 390.488400 1685.968994 8.50e-06
|
| 23 |
+
2025-09-21T21:43:12.164436 190 6.13 324.997900 1826.587402 9.00e-06
|
| 24 |
+
2025-09-21T21:43:17.924886 200 6.45 280.306800 1618.513672 9.50e-06
|
| 25 |
+
2025-09-21T21:43:18.676823 200 6.45 NA NA NA
|
| 26 |
+
2025-09-21T21:43:20.259029 200 6.45 NA NA NA
|
vocab.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"[PAD]": 28,
|
| 3 |
+
"[UNK]": 27,
|
| 4 |
+
"a": 1,
|
| 5 |
+
"b": 2,
|
| 6 |
+
"c": 3,
|
| 7 |
+
"d": 4,
|
| 8 |
+
"e": 5,
|
| 9 |
+
"f": 6,
|
| 10 |
+
"g": 7,
|
| 11 |
+
"h": 8,
|
| 12 |
+
"i": 9,
|
| 13 |
+
"j": 10,
|
| 14 |
+
"k": 11,
|
| 15 |
+
"l": 12,
|
| 16 |
+
"m": 13,
|
| 17 |
+
"n": 14,
|
| 18 |
+
"o": 15,
|
| 19 |
+
"p": 16,
|
| 20 |
+
"q": 17,
|
| 21 |
+
"r": 18,
|
| 22 |
+
"s": 19,
|
| 23 |
+
"t": 20,
|
| 24 |
+
"u": 21,
|
| 25 |
+
"v": 22,
|
| 26 |
+
"w": 23,
|
| 27 |
+
"x": 24,
|
| 28 |
+
"y": 25,
|
| 29 |
+
"z": 26,
|
| 30 |
+
"|": 0
|
| 31 |
+
}
|