Upload folder using huggingface_hub
Browse files- added_tokens.json +4 -0
- checkpoint-9000/config.json +107 -0
- checkpoint-9000/model.safetensors +3 -0
- checkpoint-9000/optimizer.pt +3 -0
- checkpoint-9000/preprocessor_config.json +9 -0
- checkpoint-9000/rng_state.pth +3 -0
- checkpoint-9000/scaler.pt +3 -0
- checkpoint-9000/scheduler.pt +3 -0
- checkpoint-9000/trainer_state.json +322 -0
- checkpoint-9000/training_args.bin +3 -0
- checkpoint-9380/config.json +107 -0
- checkpoint-9380/model.safetensors +3 -0
- checkpoint-9380/optimizer.pt +3 -0
- checkpoint-9380/preprocessor_config.json +9 -0
- checkpoint-9380/rng_state.pth +3 -0
- checkpoint-9380/scaler.pt +3 -0
- checkpoint-9380/scheduler.pt +3 -0
- checkpoint-9380/trainer_state.json +322 -0
- checkpoint-9380/training_args.bin +3 -0
- config.json +107 -0
- model.safetensors +3 -0
- preprocessor_config.json +10 -0
- special_tokens_map.json +6 -0
- tokenizer_config.json +49 -0
- vocab.json +31 -0
added_tokens.json
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"</s>": 30,
|
| 3 |
+
"<s>": 29
|
| 4 |
+
}
|
checkpoint-9000/config.json
ADDED
|
@@ -0,0 +1,107 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": 0.1,
|
| 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 |
+
"dtype": "float32",
|
| 49 |
+
"eos_token_id": 2,
|
| 50 |
+
"feat_extract_activation": "gelu",
|
| 51 |
+
"feat_extract_dropout": 0.0,
|
| 52 |
+
"feat_extract_norm": "layer",
|
| 53 |
+
"feat_proj_dropout": 0.0,
|
| 54 |
+
"feat_quantizer_dropout": 0.0,
|
| 55 |
+
"final_dropout": 0.0,
|
| 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": 0.0,
|
| 66 |
+
"mask_time_length": 10,
|
| 67 |
+
"mask_time_min_masks": 2,
|
| 68 |
+
"mask_time_prob": 0.05,
|
| 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 |
+
"transformers_version": "4.57.1",
|
| 104 |
+
"use_weighted_layer_sum": false,
|
| 105 |
+
"vocab_size": 31,
|
| 106 |
+
"xvector_output_dim": 512
|
| 107 |
+
}
|
checkpoint-9000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8db5711c31e5f7ad116f5c3d180ae61193ce81bb987dd2345da0b1acfed838f7
|
| 3 |
+
size 1261934580
|
checkpoint-9000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b41d6b1b35b8587b835e43e091aacac26d0288ea0ce7d3928afbc9b43f700d69
|
| 3 |
+
size 2524125339
|
checkpoint-9000/preprocessor_config.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_normalize": true,
|
| 3 |
+
"feature_extractor_type": "Wav2Vec2FeatureExtractor",
|
| 4 |
+
"feature_size": 1,
|
| 5 |
+
"padding_side": "right",
|
| 6 |
+
"padding_value": 0.0,
|
| 7 |
+
"return_attention_mask": true,
|
| 8 |
+
"sampling_rate": 16000
|
| 9 |
+
}
|
checkpoint-9000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ab423d30467d2b97c43f53301849392e580c540cc194fb0f77c00cc58ae26e38
|
| 3 |
+
size 14709
|
checkpoint-9000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:29c4b3a2db27ec7adc685046699ff42abf833a0a0e3afa0ab1f8ae149ccb5902
|
| 3 |
+
size 1383
|
checkpoint-9000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0e81ea76d57fb75982d9efccd8becd9b66781bedc91ada6e391446ff001eae96
|
| 3 |
+
size 1465
|
checkpoint-9000/trainer_state.json
ADDED
|
@@ -0,0 +1,322 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 9000,
|
| 3 |
+
"best_metric": 0.3495483628152051,
|
| 4 |
+
"best_model_checkpoint": "/content/drive/MyDrive/Model-Finetuning/wav2vec2/model/zu_wav2vec2/checkpoint-9000",
|
| 5 |
+
"epoch": 9.5952,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 9000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.5333333333333333,
|
| 14 |
+
"grad_norm": 2.2490434646606445,
|
| 15 |
+
"learning_rate": 0.00029939999999999996,
|
| 16 |
+
"loss": 5.3513,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.5333333333333333,
|
| 21 |
+
"eval_loss": 3.067410707473755,
|
| 22 |
+
"eval_runtime": 142.738,
|
| 23 |
+
"eval_samples_per_second": 21.018,
|
| 24 |
+
"eval_steps_per_second": 2.627,
|
| 25 |
+
"eval_wer": 1.0,
|
| 26 |
+
"step": 500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 1.0661333333333334,
|
| 30 |
+
"grad_norm": 1.0280228853225708,
|
| 31 |
+
"learning_rate": 0.0002831418918918919,
|
| 32 |
+
"loss": 2.4793,
|
| 33 |
+
"step": 1000
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 1.0661333333333334,
|
| 37 |
+
"eval_loss": 1.1224186420440674,
|
| 38 |
+
"eval_runtime": 144.1294,
|
| 39 |
+
"eval_samples_per_second": 20.815,
|
| 40 |
+
"eval_steps_per_second": 2.602,
|
| 41 |
+
"eval_wer": 0.992002258185924,
|
| 42 |
+
"step": 1000
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 1.5994666666666668,
|
| 46 |
+
"grad_norm": 1.6041176319122314,
|
| 47 |
+
"learning_rate": 0.00026624999999999994,
|
| 48 |
+
"loss": 0.9637,
|
| 49 |
+
"step": 1500
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 1.5994666666666668,
|
| 53 |
+
"eval_loss": 0.6741820573806763,
|
| 54 |
+
"eval_runtime": 143.9596,
|
| 55 |
+
"eval_samples_per_second": 20.839,
|
| 56 |
+
"eval_steps_per_second": 2.605,
|
| 57 |
+
"eval_wer": 0.7161272111403839,
|
| 58 |
+
"step": 1500
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 2.1322666666666668,
|
| 62 |
+
"grad_norm": 1.4150700569152832,
|
| 63 |
+
"learning_rate": 0.00024935810810810806,
|
| 64 |
+
"loss": 0.717,
|
| 65 |
+
"step": 2000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 2.1322666666666668,
|
| 69 |
+
"eval_loss": 0.5053547620773315,
|
| 70 |
+
"eval_runtime": 145.0264,
|
| 71 |
+
"eval_samples_per_second": 20.686,
|
| 72 |
+
"eval_steps_per_second": 2.586,
|
| 73 |
+
"eval_wer": 0.620530673692134,
|
| 74 |
+
"step": 2000
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 2.6656,
|
| 78 |
+
"grad_norm": 1.3749128580093384,
|
| 79 |
+
"learning_rate": 0.0002324662162162162,
|
| 80 |
+
"loss": 0.5979,
|
| 81 |
+
"step": 2500
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 2.6656,
|
| 85 |
+
"eval_loss": 0.4564489424228668,
|
| 86 |
+
"eval_runtime": 145.9072,
|
| 87 |
+
"eval_samples_per_second": 20.561,
|
| 88 |
+
"eval_steps_per_second": 2.57,
|
| 89 |
+
"eval_wer": 0.5550432818968761,
|
| 90 |
+
"step": 2500
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 3.1984,
|
| 94 |
+
"grad_norm": 1.1167936325073242,
|
| 95 |
+
"learning_rate": 0.00021557432432432432,
|
| 96 |
+
"loss": 0.5388,
|
| 97 |
+
"step": 3000
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 3.1984,
|
| 101 |
+
"eval_loss": 0.4222851097583771,
|
| 102 |
+
"eval_runtime": 146.8407,
|
| 103 |
+
"eval_samples_per_second": 20.43,
|
| 104 |
+
"eval_steps_per_second": 2.554,
|
| 105 |
+
"eval_wer": 0.5333082423786225,
|
| 106 |
+
"step": 3000
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 3.7317333333333336,
|
| 110 |
+
"grad_norm": 1.713294267654419,
|
| 111 |
+
"learning_rate": 0.00019868243243243244,
|
| 112 |
+
"loss": 0.4762,
|
| 113 |
+
"step": 3500
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 3.7317333333333336,
|
| 117 |
+
"eval_loss": 0.3905419409275055,
|
| 118 |
+
"eval_runtime": 146.5994,
|
| 119 |
+
"eval_samples_per_second": 20.464,
|
| 120 |
+
"eval_steps_per_second": 2.558,
|
| 121 |
+
"eval_wer": 0.4974595408355288,
|
| 122 |
+
"step": 3500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 4.2645333333333335,
|
| 126 |
+
"grad_norm": 2.082676887512207,
|
| 127 |
+
"learning_rate": 0.0001817905405405405,
|
| 128 |
+
"loss": 0.4398,
|
| 129 |
+
"step": 4000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 4.2645333333333335,
|
| 133 |
+
"eval_loss": 0.3778238892555237,
|
| 134 |
+
"eval_runtime": 145.863,
|
| 135 |
+
"eval_samples_per_second": 20.567,
|
| 136 |
+
"eval_steps_per_second": 2.571,
|
| 137 |
+
"eval_wer": 0.48692133985698155,
|
| 138 |
+
"step": 4000
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 4.797866666666667,
|
| 142 |
+
"grad_norm": 2.2089028358459473,
|
| 143 |
+
"learning_rate": 0.00016489864864864862,
|
| 144 |
+
"loss": 0.4098,
|
| 145 |
+
"step": 4500
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 4.797866666666667,
|
| 149 |
+
"eval_loss": 0.35478344559669495,
|
| 150 |
+
"eval_runtime": 146.045,
|
| 151 |
+
"eval_samples_per_second": 20.542,
|
| 152 |
+
"eval_steps_per_second": 2.568,
|
| 153 |
+
"eval_wer": 0.4574708317651487,
|
| 154 |
+
"step": 4500
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 5.330666666666667,
|
| 158 |
+
"grad_norm": 1.058657169342041,
|
| 159 |
+
"learning_rate": 0.00014800675675675674,
|
| 160 |
+
"loss": 0.3823,
|
| 161 |
+
"step": 5000
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 5.330666666666667,
|
| 165 |
+
"eval_loss": 0.3355245292186737,
|
| 166 |
+
"eval_runtime": 145.8678,
|
| 167 |
+
"eval_samples_per_second": 20.567,
|
| 168 |
+
"eval_steps_per_second": 2.571,
|
| 169 |
+
"eval_wer": 0.45079036507339104,
|
| 170 |
+
"step": 5000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 5.864,
|
| 174 |
+
"grad_norm": 1.4570820331573486,
|
| 175 |
+
"learning_rate": 0.00013111486486486486,
|
| 176 |
+
"loss": 0.3513,
|
| 177 |
+
"step": 5500
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 5.864,
|
| 181 |
+
"eval_loss": 0.32364729046821594,
|
| 182 |
+
"eval_runtime": 147.2305,
|
| 183 |
+
"eval_samples_per_second": 20.376,
|
| 184 |
+
"eval_steps_per_second": 2.547,
|
| 185 |
+
"eval_wer": 0.4314076025592774,
|
| 186 |
+
"step": 5500
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 6.3968,
|
| 190 |
+
"grad_norm": 2.4523627758026123,
|
| 191 |
+
"learning_rate": 0.00011422297297297297,
|
| 192 |
+
"loss": 0.3111,
|
| 193 |
+
"step": 6000
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 6.3968,
|
| 197 |
+
"eval_loss": 0.3129761517047882,
|
| 198 |
+
"eval_runtime": 146.9224,
|
| 199 |
+
"eval_samples_per_second": 20.419,
|
| 200 |
+
"eval_steps_per_second": 2.552,
|
| 201 |
+
"eval_wer": 0.4113662024840045,
|
| 202 |
+
"step": 6000
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 6.930133333333333,
|
| 206 |
+
"grad_norm": 1.3307918310165405,
|
| 207 |
+
"learning_rate": 9.733108108108107e-05,
|
| 208 |
+
"loss": 0.3032,
|
| 209 |
+
"step": 6500
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 6.930133333333333,
|
| 213 |
+
"eval_loss": 0.301826536655426,
|
| 214 |
+
"eval_runtime": 147.6721,
|
| 215 |
+
"eval_samples_per_second": 20.315,
|
| 216 |
+
"eval_steps_per_second": 2.539,
|
| 217 |
+
"eval_wer": 0.4039330071509221,
|
| 218 |
+
"step": 6500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 7.462933333333333,
|
| 222 |
+
"grad_norm": 1.828587532043457,
|
| 223 |
+
"learning_rate": 8.043918918918918e-05,
|
| 224 |
+
"loss": 0.2698,
|
| 225 |
+
"step": 7000
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 7.462933333333333,
|
| 229 |
+
"eval_loss": 0.29749611020088196,
|
| 230 |
+
"eval_runtime": 146.8529,
|
| 231 |
+
"eval_samples_per_second": 20.429,
|
| 232 |
+
"eval_steps_per_second": 2.554,
|
| 233 |
+
"eval_wer": 0.3873729770417764,
|
| 234 |
+
"step": 7000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 7.996266666666667,
|
| 238 |
+
"grad_norm": 1.7805229425430298,
|
| 239 |
+
"learning_rate": 6.354729729729729e-05,
|
| 240 |
+
"loss": 0.2661,
|
| 241 |
+
"step": 7500
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 7.996266666666667,
|
| 245 |
+
"eval_loss": 0.2899162769317627,
|
| 246 |
+
"eval_runtime": 147.9735,
|
| 247 |
+
"eval_samples_per_second": 20.274,
|
| 248 |
+
"eval_steps_per_second": 2.534,
|
| 249 |
+
"eval_wer": 0.3744824990590892,
|
| 250 |
+
"step": 7500
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 8.529066666666667,
|
| 254 |
+
"grad_norm": 1.059174656867981,
|
| 255 |
+
"learning_rate": 4.66554054054054e-05,
|
| 256 |
+
"loss": 0.233,
|
| 257 |
+
"step": 8000
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 8.529066666666667,
|
| 261 |
+
"eval_loss": 0.2895633280277252,
|
| 262 |
+
"eval_runtime": 147.1645,
|
| 263 |
+
"eval_samples_per_second": 20.385,
|
| 264 |
+
"eval_steps_per_second": 2.548,
|
| 265 |
+
"eval_wer": 0.36356793375987956,
|
| 266 |
+
"step": 8000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 9.061866666666667,
|
| 270 |
+
"grad_norm": 0.938797116279602,
|
| 271 |
+
"learning_rate": 2.976351351351351e-05,
|
| 272 |
+
"loss": 0.2336,
|
| 273 |
+
"step": 8500
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 9.061866666666667,
|
| 277 |
+
"eval_loss": 0.2831208407878876,
|
| 278 |
+
"eval_runtime": 147.624,
|
| 279 |
+
"eval_samples_per_second": 20.322,
|
| 280 |
+
"eval_steps_per_second": 2.54,
|
| 281 |
+
"eval_wer": 0.35641701166729395,
|
| 282 |
+
"step": 8500
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 9.5952,
|
| 286 |
+
"grad_norm": 6.92275333404541,
|
| 287 |
+
"learning_rate": 1.287162162162162e-05,
|
| 288 |
+
"loss": 0.2137,
|
| 289 |
+
"step": 9000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 9.5952,
|
| 293 |
+
"eval_loss": 0.27855604887008667,
|
| 294 |
+
"eval_runtime": 146.2994,
|
| 295 |
+
"eval_samples_per_second": 20.506,
|
| 296 |
+
"eval_steps_per_second": 2.563,
|
| 297 |
+
"eval_wer": 0.3495483628152051,
|
| 298 |
+
"step": 9000
|
| 299 |
+
}
|
| 300 |
+
],
|
| 301 |
+
"logging_steps": 500,
|
| 302 |
+
"max_steps": 9380,
|
| 303 |
+
"num_input_tokens_seen": 0,
|
| 304 |
+
"num_train_epochs": 10,
|
| 305 |
+
"save_steps": 500,
|
| 306 |
+
"stateful_callbacks": {
|
| 307 |
+
"TrainerControl": {
|
| 308 |
+
"args": {
|
| 309 |
+
"should_epoch_stop": false,
|
| 310 |
+
"should_evaluate": false,
|
| 311 |
+
"should_log": false,
|
| 312 |
+
"should_save": true,
|
| 313 |
+
"should_training_stop": false
|
| 314 |
+
},
|
| 315 |
+
"attributes": {}
|
| 316 |
+
}
|
| 317 |
+
},
|
| 318 |
+
"total_flos": 2.0824055747211284e+19,
|
| 319 |
+
"train_batch_size": 8,
|
| 320 |
+
"trial_name": null,
|
| 321 |
+
"trial_params": null
|
| 322 |
+
}
|
checkpoint-9000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c12d2489260fb1dce96735e9306c209e6bf050b06014d9d2a71fdd38ea6d3b18
|
| 3 |
+
size 5905
|
checkpoint-9380/config.json
ADDED
|
@@ -0,0 +1,107 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": 0.1,
|
| 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 |
+
"dtype": "float32",
|
| 49 |
+
"eos_token_id": 2,
|
| 50 |
+
"feat_extract_activation": "gelu",
|
| 51 |
+
"feat_extract_dropout": 0.0,
|
| 52 |
+
"feat_extract_norm": "layer",
|
| 53 |
+
"feat_proj_dropout": 0.0,
|
| 54 |
+
"feat_quantizer_dropout": 0.0,
|
| 55 |
+
"final_dropout": 0.0,
|
| 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": 0.0,
|
| 66 |
+
"mask_time_length": 10,
|
| 67 |
+
"mask_time_min_masks": 2,
|
| 68 |
+
"mask_time_prob": 0.05,
|
| 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 |
+
"transformers_version": "4.57.1",
|
| 104 |
+
"use_weighted_layer_sum": false,
|
| 105 |
+
"vocab_size": 31,
|
| 106 |
+
"xvector_output_dim": 512
|
| 107 |
+
}
|
checkpoint-9380/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aac87e9d106ee918dca48798d06dac102725c911256e9acb514d741cf8b67a14
|
| 3 |
+
size 1261934580
|
checkpoint-9380/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e19e8cc902ae522249c74f84a0f30b48bb134885433cec7036183aa88dc9a243
|
| 3 |
+
size 2524125339
|
checkpoint-9380/preprocessor_config.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_normalize": true,
|
| 3 |
+
"feature_extractor_type": "Wav2Vec2FeatureExtractor",
|
| 4 |
+
"feature_size": 1,
|
| 5 |
+
"padding_side": "right",
|
| 6 |
+
"padding_value": 0.0,
|
| 7 |
+
"return_attention_mask": true,
|
| 8 |
+
"sampling_rate": 16000
|
| 9 |
+
}
|
checkpoint-9380/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2afd1ee849497350918429fff5988fbb283e8f5721c15d50bd3b2a09f7d49989
|
| 3 |
+
size 14709
|
checkpoint-9380/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4c800a258cb6ea8d19d26cdcad02e64d16a436f29343297becbaa069b3f0418c
|
| 3 |
+
size 1383
|
checkpoint-9380/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ce187e3f555b7fd7243a8f31396ec053a90fed362680018d8b2aa39fb391bb01
|
| 3 |
+
size 1465
|
checkpoint-9380/trainer_state.json
ADDED
|
@@ -0,0 +1,322 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 9000,
|
| 3 |
+
"best_metric": 0.3495483628152051,
|
| 4 |
+
"best_model_checkpoint": "/content/drive/MyDrive/Model-Finetuning/wav2vec2/model/zu_wav2vec2/checkpoint-9000",
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 9380,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.5333333333333333,
|
| 14 |
+
"grad_norm": 2.2490434646606445,
|
| 15 |
+
"learning_rate": 0.00029939999999999996,
|
| 16 |
+
"loss": 5.3513,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.5333333333333333,
|
| 21 |
+
"eval_loss": 3.067410707473755,
|
| 22 |
+
"eval_runtime": 142.738,
|
| 23 |
+
"eval_samples_per_second": 21.018,
|
| 24 |
+
"eval_steps_per_second": 2.627,
|
| 25 |
+
"eval_wer": 1.0,
|
| 26 |
+
"step": 500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 1.0661333333333334,
|
| 30 |
+
"grad_norm": 1.0280228853225708,
|
| 31 |
+
"learning_rate": 0.0002831418918918919,
|
| 32 |
+
"loss": 2.4793,
|
| 33 |
+
"step": 1000
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 1.0661333333333334,
|
| 37 |
+
"eval_loss": 1.1224186420440674,
|
| 38 |
+
"eval_runtime": 144.1294,
|
| 39 |
+
"eval_samples_per_second": 20.815,
|
| 40 |
+
"eval_steps_per_second": 2.602,
|
| 41 |
+
"eval_wer": 0.992002258185924,
|
| 42 |
+
"step": 1000
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 1.5994666666666668,
|
| 46 |
+
"grad_norm": 1.6041176319122314,
|
| 47 |
+
"learning_rate": 0.00026624999999999994,
|
| 48 |
+
"loss": 0.9637,
|
| 49 |
+
"step": 1500
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 1.5994666666666668,
|
| 53 |
+
"eval_loss": 0.6741820573806763,
|
| 54 |
+
"eval_runtime": 143.9596,
|
| 55 |
+
"eval_samples_per_second": 20.839,
|
| 56 |
+
"eval_steps_per_second": 2.605,
|
| 57 |
+
"eval_wer": 0.7161272111403839,
|
| 58 |
+
"step": 1500
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 2.1322666666666668,
|
| 62 |
+
"grad_norm": 1.4150700569152832,
|
| 63 |
+
"learning_rate": 0.00024935810810810806,
|
| 64 |
+
"loss": 0.717,
|
| 65 |
+
"step": 2000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 2.1322666666666668,
|
| 69 |
+
"eval_loss": 0.5053547620773315,
|
| 70 |
+
"eval_runtime": 145.0264,
|
| 71 |
+
"eval_samples_per_second": 20.686,
|
| 72 |
+
"eval_steps_per_second": 2.586,
|
| 73 |
+
"eval_wer": 0.620530673692134,
|
| 74 |
+
"step": 2000
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 2.6656,
|
| 78 |
+
"grad_norm": 1.3749128580093384,
|
| 79 |
+
"learning_rate": 0.0002324662162162162,
|
| 80 |
+
"loss": 0.5979,
|
| 81 |
+
"step": 2500
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 2.6656,
|
| 85 |
+
"eval_loss": 0.4564489424228668,
|
| 86 |
+
"eval_runtime": 145.9072,
|
| 87 |
+
"eval_samples_per_second": 20.561,
|
| 88 |
+
"eval_steps_per_second": 2.57,
|
| 89 |
+
"eval_wer": 0.5550432818968761,
|
| 90 |
+
"step": 2500
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 3.1984,
|
| 94 |
+
"grad_norm": 1.1167936325073242,
|
| 95 |
+
"learning_rate": 0.00021557432432432432,
|
| 96 |
+
"loss": 0.5388,
|
| 97 |
+
"step": 3000
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 3.1984,
|
| 101 |
+
"eval_loss": 0.4222851097583771,
|
| 102 |
+
"eval_runtime": 146.8407,
|
| 103 |
+
"eval_samples_per_second": 20.43,
|
| 104 |
+
"eval_steps_per_second": 2.554,
|
| 105 |
+
"eval_wer": 0.5333082423786225,
|
| 106 |
+
"step": 3000
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 3.7317333333333336,
|
| 110 |
+
"grad_norm": 1.713294267654419,
|
| 111 |
+
"learning_rate": 0.00019868243243243244,
|
| 112 |
+
"loss": 0.4762,
|
| 113 |
+
"step": 3500
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 3.7317333333333336,
|
| 117 |
+
"eval_loss": 0.3905419409275055,
|
| 118 |
+
"eval_runtime": 146.5994,
|
| 119 |
+
"eval_samples_per_second": 20.464,
|
| 120 |
+
"eval_steps_per_second": 2.558,
|
| 121 |
+
"eval_wer": 0.4974595408355288,
|
| 122 |
+
"step": 3500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 4.2645333333333335,
|
| 126 |
+
"grad_norm": 2.082676887512207,
|
| 127 |
+
"learning_rate": 0.0001817905405405405,
|
| 128 |
+
"loss": 0.4398,
|
| 129 |
+
"step": 4000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 4.2645333333333335,
|
| 133 |
+
"eval_loss": 0.3778238892555237,
|
| 134 |
+
"eval_runtime": 145.863,
|
| 135 |
+
"eval_samples_per_second": 20.567,
|
| 136 |
+
"eval_steps_per_second": 2.571,
|
| 137 |
+
"eval_wer": 0.48692133985698155,
|
| 138 |
+
"step": 4000
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 4.797866666666667,
|
| 142 |
+
"grad_norm": 2.2089028358459473,
|
| 143 |
+
"learning_rate": 0.00016489864864864862,
|
| 144 |
+
"loss": 0.4098,
|
| 145 |
+
"step": 4500
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 4.797866666666667,
|
| 149 |
+
"eval_loss": 0.35478344559669495,
|
| 150 |
+
"eval_runtime": 146.045,
|
| 151 |
+
"eval_samples_per_second": 20.542,
|
| 152 |
+
"eval_steps_per_second": 2.568,
|
| 153 |
+
"eval_wer": 0.4574708317651487,
|
| 154 |
+
"step": 4500
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 5.330666666666667,
|
| 158 |
+
"grad_norm": 1.058657169342041,
|
| 159 |
+
"learning_rate": 0.00014800675675675674,
|
| 160 |
+
"loss": 0.3823,
|
| 161 |
+
"step": 5000
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 5.330666666666667,
|
| 165 |
+
"eval_loss": 0.3355245292186737,
|
| 166 |
+
"eval_runtime": 145.8678,
|
| 167 |
+
"eval_samples_per_second": 20.567,
|
| 168 |
+
"eval_steps_per_second": 2.571,
|
| 169 |
+
"eval_wer": 0.45079036507339104,
|
| 170 |
+
"step": 5000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 5.864,
|
| 174 |
+
"grad_norm": 1.4570820331573486,
|
| 175 |
+
"learning_rate": 0.00013111486486486486,
|
| 176 |
+
"loss": 0.3513,
|
| 177 |
+
"step": 5500
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 5.864,
|
| 181 |
+
"eval_loss": 0.32364729046821594,
|
| 182 |
+
"eval_runtime": 147.2305,
|
| 183 |
+
"eval_samples_per_second": 20.376,
|
| 184 |
+
"eval_steps_per_second": 2.547,
|
| 185 |
+
"eval_wer": 0.4314076025592774,
|
| 186 |
+
"step": 5500
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 6.3968,
|
| 190 |
+
"grad_norm": 2.4523627758026123,
|
| 191 |
+
"learning_rate": 0.00011422297297297297,
|
| 192 |
+
"loss": 0.3111,
|
| 193 |
+
"step": 6000
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 6.3968,
|
| 197 |
+
"eval_loss": 0.3129761517047882,
|
| 198 |
+
"eval_runtime": 146.9224,
|
| 199 |
+
"eval_samples_per_second": 20.419,
|
| 200 |
+
"eval_steps_per_second": 2.552,
|
| 201 |
+
"eval_wer": 0.4113662024840045,
|
| 202 |
+
"step": 6000
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 6.930133333333333,
|
| 206 |
+
"grad_norm": 1.3307918310165405,
|
| 207 |
+
"learning_rate": 9.733108108108107e-05,
|
| 208 |
+
"loss": 0.3032,
|
| 209 |
+
"step": 6500
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 6.930133333333333,
|
| 213 |
+
"eval_loss": 0.301826536655426,
|
| 214 |
+
"eval_runtime": 147.6721,
|
| 215 |
+
"eval_samples_per_second": 20.315,
|
| 216 |
+
"eval_steps_per_second": 2.539,
|
| 217 |
+
"eval_wer": 0.4039330071509221,
|
| 218 |
+
"step": 6500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 7.462933333333333,
|
| 222 |
+
"grad_norm": 1.828587532043457,
|
| 223 |
+
"learning_rate": 8.043918918918918e-05,
|
| 224 |
+
"loss": 0.2698,
|
| 225 |
+
"step": 7000
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 7.462933333333333,
|
| 229 |
+
"eval_loss": 0.29749611020088196,
|
| 230 |
+
"eval_runtime": 146.8529,
|
| 231 |
+
"eval_samples_per_second": 20.429,
|
| 232 |
+
"eval_steps_per_second": 2.554,
|
| 233 |
+
"eval_wer": 0.3873729770417764,
|
| 234 |
+
"step": 7000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 7.996266666666667,
|
| 238 |
+
"grad_norm": 1.7805229425430298,
|
| 239 |
+
"learning_rate": 6.354729729729729e-05,
|
| 240 |
+
"loss": 0.2661,
|
| 241 |
+
"step": 7500
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 7.996266666666667,
|
| 245 |
+
"eval_loss": 0.2899162769317627,
|
| 246 |
+
"eval_runtime": 147.9735,
|
| 247 |
+
"eval_samples_per_second": 20.274,
|
| 248 |
+
"eval_steps_per_second": 2.534,
|
| 249 |
+
"eval_wer": 0.3744824990590892,
|
| 250 |
+
"step": 7500
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 8.529066666666667,
|
| 254 |
+
"grad_norm": 1.059174656867981,
|
| 255 |
+
"learning_rate": 4.66554054054054e-05,
|
| 256 |
+
"loss": 0.233,
|
| 257 |
+
"step": 8000
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 8.529066666666667,
|
| 261 |
+
"eval_loss": 0.2895633280277252,
|
| 262 |
+
"eval_runtime": 147.1645,
|
| 263 |
+
"eval_samples_per_second": 20.385,
|
| 264 |
+
"eval_steps_per_second": 2.548,
|
| 265 |
+
"eval_wer": 0.36356793375987956,
|
| 266 |
+
"step": 8000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 9.061866666666667,
|
| 270 |
+
"grad_norm": 0.938797116279602,
|
| 271 |
+
"learning_rate": 2.976351351351351e-05,
|
| 272 |
+
"loss": 0.2336,
|
| 273 |
+
"step": 8500
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 9.061866666666667,
|
| 277 |
+
"eval_loss": 0.2831208407878876,
|
| 278 |
+
"eval_runtime": 147.624,
|
| 279 |
+
"eval_samples_per_second": 20.322,
|
| 280 |
+
"eval_steps_per_second": 2.54,
|
| 281 |
+
"eval_wer": 0.35641701166729395,
|
| 282 |
+
"step": 8500
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 9.5952,
|
| 286 |
+
"grad_norm": 6.92275333404541,
|
| 287 |
+
"learning_rate": 1.287162162162162e-05,
|
| 288 |
+
"loss": 0.2137,
|
| 289 |
+
"step": 9000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 9.5952,
|
| 293 |
+
"eval_loss": 0.27855604887008667,
|
| 294 |
+
"eval_runtime": 146.2994,
|
| 295 |
+
"eval_samples_per_second": 20.506,
|
| 296 |
+
"eval_steps_per_second": 2.563,
|
| 297 |
+
"eval_wer": 0.3495483628152051,
|
| 298 |
+
"step": 9000
|
| 299 |
+
}
|
| 300 |
+
],
|
| 301 |
+
"logging_steps": 500,
|
| 302 |
+
"max_steps": 9380,
|
| 303 |
+
"num_input_tokens_seen": 0,
|
| 304 |
+
"num_train_epochs": 10,
|
| 305 |
+
"save_steps": 500,
|
| 306 |
+
"stateful_callbacks": {
|
| 307 |
+
"TrainerControl": {
|
| 308 |
+
"args": {
|
| 309 |
+
"should_epoch_stop": false,
|
| 310 |
+
"should_evaluate": false,
|
| 311 |
+
"should_log": false,
|
| 312 |
+
"should_save": true,
|
| 313 |
+
"should_training_stop": true
|
| 314 |
+
},
|
| 315 |
+
"attributes": {}
|
| 316 |
+
}
|
| 317 |
+
},
|
| 318 |
+
"total_flos": 2.167851661799376e+19,
|
| 319 |
+
"train_batch_size": 8,
|
| 320 |
+
"trial_name": null,
|
| 321 |
+
"trial_params": null
|
| 322 |
+
}
|
checkpoint-9380/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c12d2489260fb1dce96735e9306c209e6bf050b06014d9d2a71fdd38ea6d3b18
|
| 3 |
+
size 5905
|
config.json
ADDED
|
@@ -0,0 +1,107 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": 0.1,
|
| 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 |
+
"dtype": "float32",
|
| 49 |
+
"eos_token_id": 2,
|
| 50 |
+
"feat_extract_activation": "gelu",
|
| 51 |
+
"feat_extract_dropout": 0.0,
|
| 52 |
+
"feat_extract_norm": "layer",
|
| 53 |
+
"feat_proj_dropout": 0.0,
|
| 54 |
+
"feat_quantizer_dropout": 0.0,
|
| 55 |
+
"final_dropout": 0.0,
|
| 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": 0.0,
|
| 66 |
+
"mask_time_length": 10,
|
| 67 |
+
"mask_time_min_masks": 2,
|
| 68 |
+
"mask_time_prob": 0.05,
|
| 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 |
+
"transformers_version": "4.57.1",
|
| 104 |
+
"use_weighted_layer_sum": false,
|
| 105 |
+
"vocab_size": 31,
|
| 106 |
+
"xvector_output_dim": 512
|
| 107 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8db5711c31e5f7ad116f5c3d180ae61193ce81bb987dd2345da0b1acfed838f7
|
| 3 |
+
size 1261934580
|
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 |
+
}
|
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 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
+
"bos_token": "<s>",
|
| 37 |
+
"clean_up_tokenization_spaces": false,
|
| 38 |
+
"do_lower_case": false,
|
| 39 |
+
"eos_token": "</s>",
|
| 40 |
+
"extra_special_tokens": {},
|
| 41 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 42 |
+
"pad_token": "[PAD]",
|
| 43 |
+
"processor_class": "Wav2Vec2Processor",
|
| 44 |
+
"replace_word_delimiter_char": " ",
|
| 45 |
+
"target_lang": null,
|
| 46 |
+
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
| 47 |
+
"unk_token": "[UNK]",
|
| 48 |
+
"word_delimiter_token": "|"
|
| 49 |
+
}
|
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 |
+
}
|