Commit
•
dcd30ec
1
Parent(s):
501cb1d
up
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- all_results.json +14 -0
- checkpoint-3000/config.json +107 -0
- checkpoint-3000/optimizer.pt +3 -0
- checkpoint-3000/preprocessor_config.json +9 -0
- checkpoint-3000/pytorch_model.bin +3 -0
- checkpoint-3000/rng_state_0.pth +3 -0
- checkpoint-3000/rng_state_1.pth +3 -0
- checkpoint-3000/rng_state_2.pth +3 -0
- checkpoint-3000/rng_state_3.pth +3 -0
- checkpoint-3000/rng_state_4.pth +3 -0
- checkpoint-3000/rng_state_5.pth +3 -0
- checkpoint-3000/rng_state_6.pth +3 -0
- checkpoint-3000/rng_state_7.pth +3 -0
- checkpoint-3000/scaler.pt +3 -0
- checkpoint-3000/scheduler.pt +3 -0
- checkpoint-3000/trainer_state.json +250 -0
- checkpoint-3000/training_args.bin +3 -0
- checkpoint-3500/config.json +107 -0
- checkpoint-3500/optimizer.pt +3 -0
- checkpoint-3500/preprocessor_config.json +9 -0
- checkpoint-3500/pytorch_model.bin +3 -0
- checkpoint-3500/rng_state_0.pth +3 -0
- checkpoint-3500/rng_state_1.pth +3 -0
- checkpoint-3500/rng_state_2.pth +3 -0
- checkpoint-3500/rng_state_3.pth +3 -0
- checkpoint-3500/rng_state_4.pth +3 -0
- checkpoint-3500/rng_state_5.pth +3 -0
- checkpoint-3500/rng_state_6.pth +3 -0
- checkpoint-3500/rng_state_7.pth +3 -0
- checkpoint-3500/scaler.pt +3 -0
- checkpoint-3500/scheduler.pt +3 -0
- checkpoint-3500/trainer_state.json +289 -0
- checkpoint-3500/training_args.bin +3 -0
- checkpoint-4000/config.json +107 -0
- checkpoint-4000/optimizer.pt +3 -0
- checkpoint-4000/preprocessor_config.json +9 -0
- checkpoint-4000/pytorch_model.bin +3 -0
- checkpoint-4000/rng_state_0.pth +3 -0
- checkpoint-4000/rng_state_1.pth +3 -0
- checkpoint-4000/rng_state_2.pth +3 -0
- checkpoint-4000/rng_state_3.pth +3 -0
- checkpoint-4000/rng_state_4.pth +3 -0
- checkpoint-4000/rng_state_5.pth +3 -0
- checkpoint-4000/rng_state_6.pth +3 -0
- checkpoint-4000/rng_state_7.pth +3 -0
- checkpoint-4000/scaler.pt +3 -0
- checkpoint-4000/scheduler.pt +3 -0
- checkpoint-4000/trainer_state.json +328 -0
- checkpoint-4000/training_args.bin +3 -0
- config.json +11 -12
all_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 150.0,
|
| 3 |
+
"eval_loss": 0.4378179907798767,
|
| 4 |
+
"eval_runtime": 1.9751,
|
| 5 |
+
"eval_samples": 270,
|
| 6 |
+
"eval_samples_per_second": 136.702,
|
| 7 |
+
"eval_steps_per_second": 4.557,
|
| 8 |
+
"eval_wer": 0.16863370547581075,
|
| 9 |
+
"train_loss": 1.9741196327647943,
|
| 10 |
+
"train_runtime": 1808.1454,
|
| 11 |
+
"train_samples": 905,
|
| 12 |
+
"train_samples_per_second": 75.077,
|
| 13 |
+
"train_steps_per_second": 2.406
|
| 14 |
+
}
|
checkpoint-3000/config.json
ADDED
|
@@ -0,0 +1,107 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "./",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 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 |
+
"hidden_act": "gelu",
|
| 56 |
+
"hidden_dropout": 0.0,
|
| 57 |
+
"hidden_size": 1024,
|
| 58 |
+
"initializer_range": 0.02,
|
| 59 |
+
"intermediate_size": 4096,
|
| 60 |
+
"layer_norm_eps": 1e-05,
|
| 61 |
+
"layerdrop": 0.0,
|
| 62 |
+
"mask_feature_length": 64,
|
| 63 |
+
"mask_feature_min_masks": 0,
|
| 64 |
+
"mask_feature_prob": 0.25,
|
| 65 |
+
"mask_time_length": 10,
|
| 66 |
+
"mask_time_min_masks": 2,
|
| 67 |
+
"mask_time_prob": 0.75,
|
| 68 |
+
"model_type": "wav2vec2",
|
| 69 |
+
"num_adapter_layers": 3,
|
| 70 |
+
"num_attention_heads": 16,
|
| 71 |
+
"num_codevector_groups": 2,
|
| 72 |
+
"num_codevectors_per_group": 320,
|
| 73 |
+
"num_conv_pos_embedding_groups": 16,
|
| 74 |
+
"num_conv_pos_embeddings": 128,
|
| 75 |
+
"num_feat_extract_layers": 7,
|
| 76 |
+
"num_hidden_layers": 24,
|
| 77 |
+
"num_negatives": 100,
|
| 78 |
+
"output_hidden_size": 1024,
|
| 79 |
+
"pad_token_id": 48,
|
| 80 |
+
"proj_codevector_dim": 768,
|
| 81 |
+
"tdnn_dilation": [
|
| 82 |
+
1,
|
| 83 |
+
2,
|
| 84 |
+
3,
|
| 85 |
+
1,
|
| 86 |
+
1
|
| 87 |
+
],
|
| 88 |
+
"tdnn_dim": [
|
| 89 |
+
512,
|
| 90 |
+
512,
|
| 91 |
+
512,
|
| 92 |
+
512,
|
| 93 |
+
1500
|
| 94 |
+
],
|
| 95 |
+
"tdnn_kernel": [
|
| 96 |
+
5,
|
| 97 |
+
3,
|
| 98 |
+
3,
|
| 99 |
+
1,
|
| 100 |
+
1
|
| 101 |
+
],
|
| 102 |
+
"torch_dtype": "float32",
|
| 103 |
+
"transformers_version": "4.15.0.dev0",
|
| 104 |
+
"use_weighted_layer_sum": false,
|
| 105 |
+
"vocab_size": 52,
|
| 106 |
+
"xvector_output_dim": 512
|
| 107 |
+
}
|
checkpoint-3000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fe99480dea945a1a380d12be09b6856fda798f10a34f11f503aab627a2fd9e54
|
| 3 |
+
size 2490485393
|
checkpoint-3000/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,
|
| 7 |
+
"return_attention_mask": true,
|
| 8 |
+
"sampling_rate": 16000
|
| 9 |
+
}
|
checkpoint-3000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:07d839729b1094bfc1dad9cf44e7a3baf0db79456bcca4a1c0e4c452c21bebc3
|
| 3 |
+
size 1262136881
|
checkpoint-3000/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e48eaa59b156324d3c752bac6c979d12f1a4b38fc7d4e4f56ae9b260e174b2b5
|
| 3 |
+
size 14567
|
checkpoint-3000/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4f1ae93e2dd3cdba783c24773712aae4ddc605a264e2fa159af37ccaf70c21e2
|
| 3 |
+
size 14503
|
checkpoint-3000/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:44cb9a419c137a795c59ada5bf76b4ae022572edd9068af0610737d45ab2090f
|
| 3 |
+
size 14567
|
checkpoint-3000/rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:05a9d8080a2008b1dfce194265d322150b21b366293a3020c93e0b227cb74efe
|
| 3 |
+
size 14503
|
checkpoint-3000/rng_state_4.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2eb127c2e2bd207cdfe9c6d2ee0ac5ae0e8fc5abb58b9c0391004b938a5aaafd
|
| 3 |
+
size 14567
|
checkpoint-3000/rng_state_5.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5b5a4fa54612b3e0e99132c4bbf6e5800eb2bc1e83300840263ff1aad4a17795
|
| 3 |
+
size 14567
|
checkpoint-3000/rng_state_6.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ccba11918942a57e0c0cbc717f7f6cadbfd906edc19bc89af674b132ce7de846
|
| 3 |
+
size 14567
|
checkpoint-3000/rng_state_7.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:76e0a7ce570523f52846def5a4e8fd2a312bdcc9d099b503d3eabc27e670477b
|
| 3 |
+
size 14631
|
checkpoint-3000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3e282386c88b92c8801eadb3ac3e2e1c684c2653af5b93ca015f68731ca894c0
|
| 3 |
+
size 559
|
checkpoint-3000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9662679f467de6dc69408f77a06738ea4f16c091cbe75449af8e7685b8b8ab08
|
| 3 |
+
size 623
|
checkpoint-3000/trainer_state.json
ADDED
|
@@ -0,0 +1,250 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 103.44827586206897,
|
| 5 |
+
"global_step": 3000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 3.45,
|
| 12 |
+
"learning_rate": 7.2e-06,
|
| 13 |
+
"loss": 16.2031,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 6.9,
|
| 18 |
+
"learning_rate": 1.47e-05,
|
| 19 |
+
"loss": 6.6774,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 10.34,
|
| 24 |
+
"learning_rate": 2.2199999999999998e-05,
|
| 25 |
+
"loss": 4.7603,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 13.79,
|
| 30 |
+
"learning_rate": 2.97e-05,
|
| 31 |
+
"loss": 3.9198,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 17.24,
|
| 36 |
+
"learning_rate": 3.7199999999999996e-05,
|
| 37 |
+
"loss": 3.5558,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 17.24,
|
| 42 |
+
"eval_loss": 3.5216922760009766,
|
| 43 |
+
"eval_runtime": 2.4797,
|
| 44 |
+
"eval_samples_per_second": 108.882,
|
| 45 |
+
"eval_steps_per_second": 3.629,
|
| 46 |
+
"eval_wer": 0.9812865497076023,
|
| 47 |
+
"step": 500
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 20.69,
|
| 51 |
+
"learning_rate": 4.4699999999999996e-05,
|
| 52 |
+
"loss": 3.4799,
|
| 53 |
+
"step": 600
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 24.14,
|
| 57 |
+
"learning_rate": 5.2199999999999995e-05,
|
| 58 |
+
"loss": 3.4271,
|
| 59 |
+
"step": 700
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 27.59,
|
| 63 |
+
"learning_rate": 5.97e-05,
|
| 64 |
+
"loss": 3.2779,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 31.03,
|
| 69 |
+
"learning_rate": 6.72e-05,
|
| 70 |
+
"loss": 2.9317,
|
| 71 |
+
"step": 900
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"epoch": 34.48,
|
| 75 |
+
"learning_rate": 7.47e-05,
|
| 76 |
+
"loss": 2.2202,
|
| 77 |
+
"step": 1000
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 34.48,
|
| 81 |
+
"eval_loss": 0.9991578459739685,
|
| 82 |
+
"eval_runtime": 1.9665,
|
| 83 |
+
"eval_samples_per_second": 137.299,
|
| 84 |
+
"eval_steps_per_second": 4.577,
|
| 85 |
+
"eval_wer": 0.2637958532695375,
|
| 86 |
+
"step": 1000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 37.93,
|
| 90 |
+
"learning_rate": 7.285074626865672e-05,
|
| 91 |
+
"loss": 1.835,
|
| 92 |
+
"step": 1100
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 41.38,
|
| 96 |
+
"learning_rate": 7.061194029850745e-05,
|
| 97 |
+
"loss": 1.653,
|
| 98 |
+
"step": 1200
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 44.83,
|
| 102 |
+
"learning_rate": 6.83731343283582e-05,
|
| 103 |
+
"loss": 1.5243,
|
| 104 |
+
"step": 1300
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 48.28,
|
| 108 |
+
"learning_rate": 6.613432835820895e-05,
|
| 109 |
+
"loss": 1.4343,
|
| 110 |
+
"step": 1400
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 51.72,
|
| 114 |
+
"learning_rate": 6.38955223880597e-05,
|
| 115 |
+
"loss": 1.376,
|
| 116 |
+
"step": 1500
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 51.72,
|
| 120 |
+
"eval_loss": 0.5396047830581665,
|
| 121 |
+
"eval_runtime": 1.8956,
|
| 122 |
+
"eval_samples_per_second": 142.435,
|
| 123 |
+
"eval_steps_per_second": 4.748,
|
| 124 |
+
"eval_wer": 0.1946836788942052,
|
| 125 |
+
"step": 1500
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 55.17,
|
| 129 |
+
"learning_rate": 6.165671641791044e-05,
|
| 130 |
+
"loss": 1.3168,
|
| 131 |
+
"step": 1600
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 58.62,
|
| 135 |
+
"learning_rate": 5.941791044776119e-05,
|
| 136 |
+
"loss": 1.2773,
|
| 137 |
+
"step": 1700
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 62.07,
|
| 141 |
+
"learning_rate": 5.717910447761193e-05,
|
| 142 |
+
"loss": 1.2455,
|
| 143 |
+
"step": 1800
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 65.52,
|
| 147 |
+
"learning_rate": 5.494029850746268e-05,
|
| 148 |
+
"loss": 1.1881,
|
| 149 |
+
"step": 1900
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 68.97,
|
| 153 |
+
"learning_rate": 5.272388059701492e-05,
|
| 154 |
+
"loss": 1.1565,
|
| 155 |
+
"step": 2000
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 68.97,
|
| 159 |
+
"eval_loss": 0.4707144498825073,
|
| 160 |
+
"eval_runtime": 1.8963,
|
| 161 |
+
"eval_samples_per_second": 142.381,
|
| 162 |
+
"eval_steps_per_second": 4.746,
|
| 163 |
+
"eval_wer": 0.1773524720893142,
|
| 164 |
+
"step": 2000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 72.41,
|
| 168 |
+
"learning_rate": 5.048507462686567e-05,
|
| 169 |
+
"loss": 1.1295,
|
| 170 |
+
"step": 2100
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 75.86,
|
| 174 |
+
"learning_rate": 4.8268656716417906e-05,
|
| 175 |
+
"loss": 1.0983,
|
| 176 |
+
"step": 2200
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 79.31,
|
| 180 |
+
"learning_rate": 4.6029850746268655e-05,
|
| 181 |
+
"loss": 1.0716,
|
| 182 |
+
"step": 2300
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 82.76,
|
| 186 |
+
"learning_rate": 4.37910447761194e-05,
|
| 187 |
+
"loss": 1.0582,
|
| 188 |
+
"step": 2400
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 86.21,
|
| 192 |
+
"learning_rate": 4.155223880597015e-05,
|
| 193 |
+
"loss": 1.0112,
|
| 194 |
+
"step": 2500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 86.21,
|
| 198 |
+
"eval_loss": 0.46029242873191833,
|
| 199 |
+
"eval_runtime": 1.9007,
|
| 200 |
+
"eval_samples_per_second": 142.05,
|
| 201 |
+
"eval_steps_per_second": 4.735,
|
| 202 |
+
"eval_wer": 0.17533227006911217,
|
| 203 |
+
"step": 2500
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 89.66,
|
| 207 |
+
"learning_rate": 3.931343283582089e-05,
|
| 208 |
+
"loss": 0.9962,
|
| 209 |
+
"step": 2600
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 93.1,
|
| 213 |
+
"learning_rate": 3.707462686567164e-05,
|
| 214 |
+
"loss": 0.9746,
|
| 215 |
+
"step": 2700
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 96.55,
|
| 219 |
+
"learning_rate": 3.483582089552238e-05,
|
| 220 |
+
"loss": 0.9592,
|
| 221 |
+
"step": 2800
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 100.0,
|
| 225 |
+
"learning_rate": 3.259701492537313e-05,
|
| 226 |
+
"loss": 0.9364,
|
| 227 |
+
"step": 2900
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 103.45,
|
| 231 |
+
"learning_rate": 3.0358208955223878e-05,
|
| 232 |
+
"loss": 0.9387,
|
| 233 |
+
"step": 3000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 103.45,
|
| 237 |
+
"eval_loss": 0.4466392397880554,
|
| 238 |
+
"eval_runtime": 1.9017,
|
| 239 |
+
"eval_samples_per_second": 141.977,
|
| 240 |
+
"eval_steps_per_second": 4.733,
|
| 241 |
+
"eval_wer": 0.1721424774056353,
|
| 242 |
+
"step": 3000
|
| 243 |
+
}
|
| 244 |
+
],
|
| 245 |
+
"max_steps": 4350,
|
| 246 |
+
"num_train_epochs": 150,
|
| 247 |
+
"total_flos": 1.2910510097926455e+19,
|
| 248 |
+
"trial_name": null,
|
| 249 |
+
"trial_params": null
|
| 250 |
+
}
|
checkpoint-3000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f6f38b03969c3e29b358a2037b4f61da5251f02d2d9436641d5a5d94788db60e
|
| 3 |
+
size 2927
|
checkpoint-3500/config.json
ADDED
|
@@ -0,0 +1,107 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "./",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 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 |
+
"hidden_act": "gelu",
|
| 56 |
+
"hidden_dropout": 0.0,
|
| 57 |
+
"hidden_size": 1024,
|
| 58 |
+
"initializer_range": 0.02,
|
| 59 |
+
"intermediate_size": 4096,
|
| 60 |
+
"layer_norm_eps": 1e-05,
|
| 61 |
+
"layerdrop": 0.0,
|
| 62 |
+
"mask_feature_length": 64,
|
| 63 |
+
"mask_feature_min_masks": 0,
|
| 64 |
+
"mask_feature_prob": 0.25,
|
| 65 |
+
"mask_time_length": 10,
|
| 66 |
+
"mask_time_min_masks": 2,
|
| 67 |
+
"mask_time_prob": 0.75,
|
| 68 |
+
"model_type": "wav2vec2",
|
| 69 |
+
"num_adapter_layers": 3,
|
| 70 |
+
"num_attention_heads": 16,
|
| 71 |
+
"num_codevector_groups": 2,
|
| 72 |
+
"num_codevectors_per_group": 320,
|
| 73 |
+
"num_conv_pos_embedding_groups": 16,
|
| 74 |
+
"num_conv_pos_embeddings": 128,
|
| 75 |
+
"num_feat_extract_layers": 7,
|
| 76 |
+
"num_hidden_layers": 24,
|
| 77 |
+
"num_negatives": 100,
|
| 78 |
+
"output_hidden_size": 1024,
|
| 79 |
+
"pad_token_id": 48,
|
| 80 |
+
"proj_codevector_dim": 768,
|
| 81 |
+
"tdnn_dilation": [
|
| 82 |
+
1,
|
| 83 |
+
2,
|
| 84 |
+
3,
|
| 85 |
+
1,
|
| 86 |
+
1
|
| 87 |
+
],
|
| 88 |
+
"tdnn_dim": [
|
| 89 |
+
512,
|
| 90 |
+
512,
|
| 91 |
+
512,
|
| 92 |
+
512,
|
| 93 |
+
1500
|
| 94 |
+
],
|
| 95 |
+
"tdnn_kernel": [
|
| 96 |
+
5,
|
| 97 |
+
3,
|
| 98 |
+
3,
|
| 99 |
+
1,
|
| 100 |
+
1
|
| 101 |
+
],
|
| 102 |
+
"torch_dtype": "float32",
|
| 103 |
+
"transformers_version": "4.15.0.dev0",
|
| 104 |
+
"use_weighted_layer_sum": false,
|
| 105 |
+
"vocab_size": 52,
|
| 106 |
+
"xvector_output_dim": 512
|
| 107 |
+
}
|
checkpoint-3500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3f421d2b1e3f06b4cff391206ab29dcb32cf2aac5324f7d41ebf8aca6fada942
|
| 3 |
+
size 2490485393
|
checkpoint-3500/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,
|
| 7 |
+
"return_attention_mask": true,
|
| 8 |
+
"sampling_rate": 16000
|
| 9 |
+
}
|
checkpoint-3500/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8507f1fd4951af8ae7c4f333000a1b1ca0b8429480c2955ef2548d28c5793ab
|
| 3 |
+
size 1262136881
|
checkpoint-3500/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8788c98275b28b5b9ec4bfcd1f218ccda9c88b355673c8d00abef3cc369028cc
|
| 3 |
+
size 14567
|
checkpoint-3500/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c3caaf21470c55a26a83a6db1f9e0b5e1adc0a4b02d7d570660a3bc828b3d8d6
|
| 3 |
+
size 14503
|
checkpoint-3500/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:77d386527eecef36648e701aca8cbe60644c0913930ddef107ad583bcd96ed69
|
| 3 |
+
size 14567
|
checkpoint-3500/rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0244895a4a72f32d72d09ba5f40370b298c8da4f5c7cd488b7c7aea9f0568ba0
|
| 3 |
+
size 14567
|
checkpoint-3500/rng_state_4.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:86af25e4eb41d0743b2678df4a77f634ab5b517e8354374f65f13d45d22cdbbf
|
| 3 |
+
size 14567
|
checkpoint-3500/rng_state_5.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5457892e73154387ee22e28a873ddc0c92ba3c1ca39c4efef6d0bf475deaeeb7
|
| 3 |
+
size 14567
|
checkpoint-3500/rng_state_6.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea5e60f39a6c296b52983d9a64b3924d0c3cf6d0930834aeba8d037f54e78a8e
|
| 3 |
+
size 14567
|
checkpoint-3500/rng_state_7.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5062184afe4371e809d6e392f3ce930b22620b6864cc1c51d79270e651eb17bb
|
| 3 |
+
size 14567
|
checkpoint-3500/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d09341b70ef57760e07267e67bbaee462df7adf8d23d92a20cc576715bd2b094
|
| 3 |
+
size 559
|
checkpoint-3500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3fbe3309fc6bfffbdf44f870e02817da89e5efe88906cbd19563e50da8624cfc
|
| 3 |
+
size 623
|
checkpoint-3500/trainer_state.json
ADDED
|
@@ -0,0 +1,289 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 120.6896551724138,
|
| 5 |
+
"global_step": 3500,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 3.45,
|
| 12 |
+
"learning_rate": 7.2e-06,
|
| 13 |
+
"loss": 16.2031,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 6.9,
|
| 18 |
+
"learning_rate": 1.47e-05,
|
| 19 |
+
"loss": 6.6774,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 10.34,
|
| 24 |
+
"learning_rate": 2.2199999999999998e-05,
|
| 25 |
+
"loss": 4.7603,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 13.79,
|
| 30 |
+
"learning_rate": 2.97e-05,
|
| 31 |
+
"loss": 3.9198,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 17.24,
|
| 36 |
+
"learning_rate": 3.7199999999999996e-05,
|
| 37 |
+
"loss": 3.5558,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 17.24,
|
| 42 |
+
"eval_loss": 3.5216922760009766,
|
| 43 |
+
"eval_runtime": 2.4797,
|
| 44 |
+
"eval_samples_per_second": 108.882,
|
| 45 |
+
"eval_steps_per_second": 3.629,
|
| 46 |
+
"eval_wer": 0.9812865497076023,
|
| 47 |
+
"step": 500
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 20.69,
|
| 51 |
+
"learning_rate": 4.4699999999999996e-05,
|
| 52 |
+
"loss": 3.4799,
|
| 53 |
+
"step": 600
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 24.14,
|
| 57 |
+
"learning_rate": 5.2199999999999995e-05,
|
| 58 |
+
"loss": 3.4271,
|
| 59 |
+
"step": 700
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 27.59,
|
| 63 |
+
"learning_rate": 5.97e-05,
|
| 64 |
+
"loss": 3.2779,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 31.03,
|
| 69 |
+
"learning_rate": 6.72e-05,
|
| 70 |
+
"loss": 2.9317,
|
| 71 |
+
"step": 900
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"epoch": 34.48,
|
| 75 |
+
"learning_rate": 7.47e-05,
|
| 76 |
+
"loss": 2.2202,
|
| 77 |
+
"step": 1000
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 34.48,
|
| 81 |
+
"eval_loss": 0.9991578459739685,
|
| 82 |
+
"eval_runtime": 1.9665,
|
| 83 |
+
"eval_samples_per_second": 137.299,
|
| 84 |
+
"eval_steps_per_second": 4.577,
|
| 85 |
+
"eval_wer": 0.2637958532695375,
|
| 86 |
+
"step": 1000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 37.93,
|
| 90 |
+
"learning_rate": 7.285074626865672e-05,
|
| 91 |
+
"loss": 1.835,
|
| 92 |
+
"step": 1100
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 41.38,
|
| 96 |
+
"learning_rate": 7.061194029850745e-05,
|
| 97 |
+
"loss": 1.653,
|
| 98 |
+
"step": 1200
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 44.83,
|
| 102 |
+
"learning_rate": 6.83731343283582e-05,
|
| 103 |
+
"loss": 1.5243,
|
| 104 |
+
"step": 1300
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 48.28,
|
| 108 |
+
"learning_rate": 6.613432835820895e-05,
|
| 109 |
+
"loss": 1.4343,
|
| 110 |
+
"step": 1400
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 51.72,
|
| 114 |
+
"learning_rate": 6.38955223880597e-05,
|
| 115 |
+
"loss": 1.376,
|
| 116 |
+
"step": 1500
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 51.72,
|
| 120 |
+
"eval_loss": 0.5396047830581665,
|
| 121 |
+
"eval_runtime": 1.8956,
|
| 122 |
+
"eval_samples_per_second": 142.435,
|
| 123 |
+
"eval_steps_per_second": 4.748,
|
| 124 |
+
"eval_wer": 0.1946836788942052,
|
| 125 |
+
"step": 1500
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 55.17,
|
| 129 |
+
"learning_rate": 6.165671641791044e-05,
|
| 130 |
+
"loss": 1.3168,
|
| 131 |
+
"step": 1600
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 58.62,
|
| 135 |
+
"learning_rate": 5.941791044776119e-05,
|
| 136 |
+
"loss": 1.2773,
|
| 137 |
+
"step": 1700
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 62.07,
|
| 141 |
+
"learning_rate": 5.717910447761193e-05,
|
| 142 |
+
"loss": 1.2455,
|
| 143 |
+
"step": 1800
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 65.52,
|
| 147 |
+
"learning_rate": 5.494029850746268e-05,
|
| 148 |
+
"loss": 1.1881,
|
| 149 |
+
"step": 1900
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 68.97,
|
| 153 |
+
"learning_rate": 5.272388059701492e-05,
|
| 154 |
+
"loss": 1.1565,
|
| 155 |
+
"step": 2000
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 68.97,
|
| 159 |
+
"eval_loss": 0.4707144498825073,
|
| 160 |
+
"eval_runtime": 1.8963,
|
| 161 |
+
"eval_samples_per_second": 142.381,
|
| 162 |
+
"eval_steps_per_second": 4.746,
|
| 163 |
+
"eval_wer": 0.1773524720893142,
|
| 164 |
+
"step": 2000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 72.41,
|
| 168 |
+
"learning_rate": 5.048507462686567e-05,
|
| 169 |
+
"loss": 1.1295,
|
| 170 |
+
"step": 2100
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 75.86,
|
| 174 |
+
"learning_rate": 4.8268656716417906e-05,
|
| 175 |
+
"loss": 1.0983,
|
| 176 |
+
"step": 2200
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 79.31,
|
| 180 |
+
"learning_rate": 4.6029850746268655e-05,
|
| 181 |
+
"loss": 1.0716,
|
| 182 |
+
"step": 2300
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 82.76,
|
| 186 |
+
"learning_rate": 4.37910447761194e-05,
|
| 187 |
+
"loss": 1.0582,
|
| 188 |
+
"step": 2400
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 86.21,
|
| 192 |
+
"learning_rate": 4.155223880597015e-05,
|
| 193 |
+
"loss": 1.0112,
|
| 194 |
+
"step": 2500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 86.21,
|
| 198 |
+
"eval_loss": 0.46029242873191833,
|
| 199 |
+
"eval_runtime": 1.9007,
|
| 200 |
+
"eval_samples_per_second": 142.05,
|
| 201 |
+
"eval_steps_per_second": 4.735,
|
| 202 |
+
"eval_wer": 0.17533227006911217,
|
| 203 |
+
"step": 2500
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 89.66,
|
| 207 |
+
"learning_rate": 3.931343283582089e-05,
|
| 208 |
+
"loss": 0.9962,
|
| 209 |
+
"step": 2600
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 93.1,
|
| 213 |
+
"learning_rate": 3.707462686567164e-05,
|
| 214 |
+
"loss": 0.9746,
|
| 215 |
+
"step": 2700
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 96.55,
|
| 219 |
+
"learning_rate": 3.483582089552238e-05,
|
| 220 |
+
"loss": 0.9592,
|
| 221 |
+
"step": 2800
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 100.0,
|
| 225 |
+
"learning_rate": 3.259701492537313e-05,
|
| 226 |
+
"loss": 0.9364,
|
| 227 |
+
"step": 2900
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 103.45,
|
| 231 |
+
"learning_rate": 3.0358208955223878e-05,
|
| 232 |
+
"loss": 0.9387,
|
| 233 |
+
"step": 3000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 103.45,
|
| 237 |
+
"eval_loss": 0.4466392397880554,
|
| 238 |
+
"eval_runtime": 1.9017,
|
| 239 |
+
"eval_samples_per_second": 141.977,
|
| 240 |
+
"eval_steps_per_second": 4.733,
|
| 241 |
+
"eval_wer": 0.1721424774056353,
|
| 242 |
+
"step": 3000
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 106.9,
|
| 246 |
+
"learning_rate": 2.811940298507462e-05,
|
| 247 |
+
"loss": 0.9223,
|
| 248 |
+
"step": 3100
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 110.34,
|
| 252 |
+
"learning_rate": 2.588059701492537e-05,
|
| 253 |
+
"loss": 0.8839,
|
| 254 |
+
"step": 3200
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 113.79,
|
| 258 |
+
"learning_rate": 2.364179104477612e-05,
|
| 259 |
+
"loss": 0.8845,
|
| 260 |
+
"step": 3300
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 117.24,
|
| 264 |
+
"learning_rate": 2.1402985074626863e-05,
|
| 265 |
+
"loss": 0.8604,
|
| 266 |
+
"step": 3400
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 120.69,
|
| 270 |
+
"learning_rate": 1.9164179104477612e-05,
|
| 271 |
+
"loss": 0.8526,
|
| 272 |
+
"step": 3500
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 120.69,
|
| 276 |
+
"eval_loss": 0.43604278564453125,
|
| 277 |
+
"eval_runtime": 1.8975,
|
| 278 |
+
"eval_samples_per_second": 142.29,
|
| 279 |
+
"eval_steps_per_second": 4.743,
|
| 280 |
+
"eval_wer": 0.17044125465178098,
|
| 281 |
+
"step": 3500
|
| 282 |
+
}
|
| 283 |
+
],
|
| 284 |
+
"max_steps": 4350,
|
| 285 |
+
"num_train_epochs": 150,
|
| 286 |
+
"total_flos": 1.5065166616188682e+19,
|
| 287 |
+
"trial_name": null,
|
| 288 |
+
"trial_params": null
|
| 289 |
+
}
|
checkpoint-3500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f6f38b03969c3e29b358a2037b4f61da5251f02d2d9436641d5a5d94788db60e
|
| 3 |
+
size 2927
|
checkpoint-4000/config.json
ADDED
|
@@ -0,0 +1,107 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "./",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 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 |
+
"hidden_act": "gelu",
|
| 56 |
+
"hidden_dropout": 0.0,
|
| 57 |
+
"hidden_size": 1024,
|
| 58 |
+
"initializer_range": 0.02,
|
| 59 |
+
"intermediate_size": 4096,
|
| 60 |
+
"layer_norm_eps": 1e-05,
|
| 61 |
+
"layerdrop": 0.0,
|
| 62 |
+
"mask_feature_length": 64,
|
| 63 |
+
"mask_feature_min_masks": 0,
|
| 64 |
+
"mask_feature_prob": 0.25,
|
| 65 |
+
"mask_time_length": 10,
|
| 66 |
+
"mask_time_min_masks": 2,
|
| 67 |
+
"mask_time_prob": 0.75,
|
| 68 |
+
"model_type": "wav2vec2",
|
| 69 |
+
"num_adapter_layers": 3,
|
| 70 |
+
"num_attention_heads": 16,
|
| 71 |
+
"num_codevector_groups": 2,
|
| 72 |
+
"num_codevectors_per_group": 320,
|
| 73 |
+
"num_conv_pos_embedding_groups": 16,
|
| 74 |
+
"num_conv_pos_embeddings": 128,
|
| 75 |
+
"num_feat_extract_layers": 7,
|
| 76 |
+
"num_hidden_layers": 24,
|
| 77 |
+
"num_negatives": 100,
|
| 78 |
+
"output_hidden_size": 1024,
|
| 79 |
+
"pad_token_id": 48,
|
| 80 |
+
"proj_codevector_dim": 768,
|
| 81 |
+
"tdnn_dilation": [
|
| 82 |
+
1,
|
| 83 |
+
2,
|
| 84 |
+
3,
|
| 85 |
+
1,
|
| 86 |
+
1
|
| 87 |
+
],
|
| 88 |
+
"tdnn_dim": [
|
| 89 |
+
512,
|
| 90 |
+
512,
|
| 91 |
+
512,
|
| 92 |
+
512,
|
| 93 |
+
1500
|
| 94 |
+
],
|
| 95 |
+
"tdnn_kernel": [
|
| 96 |
+
5,
|
| 97 |
+
3,
|
| 98 |
+
3,
|
| 99 |
+
1,
|
| 100 |
+
1
|
| 101 |
+
],
|
| 102 |
+
"torch_dtype": "float32",
|
| 103 |
+
"transformers_version": "4.15.0.dev0",
|
| 104 |
+
"use_weighted_layer_sum": false,
|
| 105 |
+
"vocab_size": 52,
|
| 106 |
+
"xvector_output_dim": 512
|
| 107 |
+
}
|
checkpoint-4000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e9b7156890d92f7e3e7d73a964ed92fda2116a2a252494fbe66415631ae94962
|
| 3 |
+
size 2490485393
|
checkpoint-4000/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,
|
| 7 |
+
"return_attention_mask": true,
|
| 8 |
+
"sampling_rate": 16000
|
| 9 |
+
}
|
checkpoint-4000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b00fd64e7e95f566a2fae32794ebbe237a19914930bcdccec19c516a3fca89e7
|
| 3 |
+
size 1262136881
|
checkpoint-4000/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c27c8e6d4c33919229c7d4240a66b6e288e3a01b097eba84264ed3974d264a01
|
| 3 |
+
size 14567
|
checkpoint-4000/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72ec11b564dc40fad0622c5eec645266b0d2893c3bfeae1edb8d94e3d7be93e4
|
| 3 |
+
size 14567
|
checkpoint-4000/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0ab56d59383d5509dfa905bff398502522b42d99e2170d36e951a07fea75222d
|
| 3 |
+
size 14567
|
checkpoint-4000/rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:640572036962b8c35ede1eb981f6ccf5c6e4c045f16a43fdbac641b47b9a7b3c
|
| 3 |
+
size 14567
|
checkpoint-4000/rng_state_4.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dfdc4dd455cb23066cb93b1612945ab0a244a95ac4350fb7a8ff9d70e0222ed1
|
| 3 |
+
size 14567
|
checkpoint-4000/rng_state_5.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f6bc63bed88616cba15942352fff38898fc793302db4c0bc3055c5d6f63b1c01
|
| 3 |
+
size 14567
|
checkpoint-4000/rng_state_6.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b2abd59205bcb4e28532b2fda4cb473371a64026fe2a28d95a5f767ad422f5a1
|
| 3 |
+
size 14503
|
checkpoint-4000/rng_state_7.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1f9f72a8b6e105d83979c9e66f262d8265bf1b4ae2763d5a63dbde9b12fa6a07
|
| 3 |
+
size 14567
|
checkpoint-4000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:75dc1f3df92aaeaa0e2e26fd93ca618cb37b77b3395d56a08ef59bb2bb1f2047
|
| 3 |
+
size 559
|
checkpoint-4000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e415907330c3a0a40649be82e9bcced404f2cb43e4bba013232cd32680765684
|
| 3 |
+
size 623
|
checkpoint-4000/trainer_state.json
ADDED
|
@@ -0,0 +1,328 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 137.93103448275863,
|
| 5 |
+
"global_step": 4000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 3.45,
|
| 12 |
+
"learning_rate": 7.2e-06,
|
| 13 |
+
"loss": 16.2031,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 6.9,
|
| 18 |
+
"learning_rate": 1.47e-05,
|
| 19 |
+
"loss": 6.6774,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 10.34,
|
| 24 |
+
"learning_rate": 2.2199999999999998e-05,
|
| 25 |
+
"loss": 4.7603,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 13.79,
|
| 30 |
+
"learning_rate": 2.97e-05,
|
| 31 |
+
"loss": 3.9198,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 17.24,
|
| 36 |
+
"learning_rate": 3.7199999999999996e-05,
|
| 37 |
+
"loss": 3.5558,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 17.24,
|
| 42 |
+
"eval_loss": 3.5216922760009766,
|
| 43 |
+
"eval_runtime": 2.4797,
|
| 44 |
+
"eval_samples_per_second": 108.882,
|
| 45 |
+
"eval_steps_per_second": 3.629,
|
| 46 |
+
"eval_wer": 0.9812865497076023,
|
| 47 |
+
"step": 500
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 20.69,
|
| 51 |
+
"learning_rate": 4.4699999999999996e-05,
|
| 52 |
+
"loss": 3.4799,
|
| 53 |
+
"step": 600
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 24.14,
|
| 57 |
+
"learning_rate": 5.2199999999999995e-05,
|
| 58 |
+
"loss": 3.4271,
|
| 59 |
+
"step": 700
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 27.59,
|
| 63 |
+
"learning_rate": 5.97e-05,
|
| 64 |
+
"loss": 3.2779,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 31.03,
|
| 69 |
+
"learning_rate": 6.72e-05,
|
| 70 |
+
"loss": 2.9317,
|
| 71 |
+
"step": 900
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"epoch": 34.48,
|
| 75 |
+
"learning_rate": 7.47e-05,
|
| 76 |
+
"loss": 2.2202,
|
| 77 |
+
"step": 1000
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 34.48,
|
| 81 |
+
"eval_loss": 0.9991578459739685,
|
| 82 |
+
"eval_runtime": 1.9665,
|
| 83 |
+
"eval_samples_per_second": 137.299,
|
| 84 |
+
"eval_steps_per_second": 4.577,
|
| 85 |
+
"eval_wer": 0.2637958532695375,
|
| 86 |
+
"step": 1000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 37.93,
|
| 90 |
+
"learning_rate": 7.285074626865672e-05,
|
| 91 |
+
"loss": 1.835,
|
| 92 |
+
"step": 1100
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 41.38,
|
| 96 |
+
"learning_rate": 7.061194029850745e-05,
|
| 97 |
+
"loss": 1.653,
|
| 98 |
+
"step": 1200
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 44.83,
|
| 102 |
+
"learning_rate": 6.83731343283582e-05,
|
| 103 |
+
"loss": 1.5243,
|
| 104 |
+
"step": 1300
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 48.28,
|
| 108 |
+
"learning_rate": 6.613432835820895e-05,
|
| 109 |
+
"loss": 1.4343,
|
| 110 |
+
"step": 1400
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 51.72,
|
| 114 |
+
"learning_rate": 6.38955223880597e-05,
|
| 115 |
+
"loss": 1.376,
|
| 116 |
+
"step": 1500
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 51.72,
|
| 120 |
+
"eval_loss": 0.5396047830581665,
|
| 121 |
+
"eval_runtime": 1.8956,
|
| 122 |
+
"eval_samples_per_second": 142.435,
|
| 123 |
+
"eval_steps_per_second": 4.748,
|
| 124 |
+
"eval_wer": 0.1946836788942052,
|
| 125 |
+
"step": 1500
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 55.17,
|
| 129 |
+
"learning_rate": 6.165671641791044e-05,
|
| 130 |
+
"loss": 1.3168,
|
| 131 |
+
"step": 1600
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 58.62,
|
| 135 |
+
"learning_rate": 5.941791044776119e-05,
|
| 136 |
+
"loss": 1.2773,
|
| 137 |
+
"step": 1700
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 62.07,
|
| 141 |
+
"learning_rate": 5.717910447761193e-05,
|
| 142 |
+
"loss": 1.2455,
|
| 143 |
+
"step": 1800
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 65.52,
|
| 147 |
+
"learning_rate": 5.494029850746268e-05,
|
| 148 |
+
"loss": 1.1881,
|
| 149 |
+
"step": 1900
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 68.97,
|
| 153 |
+
"learning_rate": 5.272388059701492e-05,
|
| 154 |
+
"loss": 1.1565,
|
| 155 |
+
"step": 2000
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 68.97,
|
| 159 |
+
"eval_loss": 0.4707144498825073,
|
| 160 |
+
"eval_runtime": 1.8963,
|
| 161 |
+
"eval_samples_per_second": 142.381,
|
| 162 |
+
"eval_steps_per_second": 4.746,
|
| 163 |
+
"eval_wer": 0.1773524720893142,
|
| 164 |
+
"step": 2000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 72.41,
|
| 168 |
+
"learning_rate": 5.048507462686567e-05,
|
| 169 |
+
"loss": 1.1295,
|
| 170 |
+
"step": 2100
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 75.86,
|
| 174 |
+
"learning_rate": 4.8268656716417906e-05,
|
| 175 |
+
"loss": 1.0983,
|
| 176 |
+
"step": 2200
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 79.31,
|
| 180 |
+
"learning_rate": 4.6029850746268655e-05,
|
| 181 |
+
"loss": 1.0716,
|
| 182 |
+
"step": 2300
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 82.76,
|
| 186 |
+
"learning_rate": 4.37910447761194e-05,
|
| 187 |
+
"loss": 1.0582,
|
| 188 |
+
"step": 2400
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 86.21,
|
| 192 |
+
"learning_rate": 4.155223880597015e-05,
|
| 193 |
+
"loss": 1.0112,
|
| 194 |
+
"step": 2500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 86.21,
|
| 198 |
+
"eval_loss": 0.46029242873191833,
|
| 199 |
+
"eval_runtime": 1.9007,
|
| 200 |
+
"eval_samples_per_second": 142.05,
|
| 201 |
+
"eval_steps_per_second": 4.735,
|
| 202 |
+
"eval_wer": 0.17533227006911217,
|
| 203 |
+
"step": 2500
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 89.66,
|
| 207 |
+
"learning_rate": 3.931343283582089e-05,
|
| 208 |
+
"loss": 0.9962,
|
| 209 |
+
"step": 2600
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 93.1,
|
| 213 |
+
"learning_rate": 3.707462686567164e-05,
|
| 214 |
+
"loss": 0.9746,
|
| 215 |
+
"step": 2700
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 96.55,
|
| 219 |
+
"learning_rate": 3.483582089552238e-05,
|
| 220 |
+
"loss": 0.9592,
|
| 221 |
+
"step": 2800
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 100.0,
|
| 225 |
+
"learning_rate": 3.259701492537313e-05,
|
| 226 |
+
"loss": 0.9364,
|
| 227 |
+
"step": 2900
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 103.45,
|
| 231 |
+
"learning_rate": 3.0358208955223878e-05,
|
| 232 |
+
"loss": 0.9387,
|
| 233 |
+
"step": 3000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 103.45,
|
| 237 |
+
"eval_loss": 0.4466392397880554,
|
| 238 |
+
"eval_runtime": 1.9017,
|
| 239 |
+
"eval_samples_per_second": 141.977,
|
| 240 |
+
"eval_steps_per_second": 4.733,
|
| 241 |
+
"eval_wer": 0.1721424774056353,
|
| 242 |
+
"step": 3000
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 106.9,
|
| 246 |
+
"learning_rate": 2.811940298507462e-05,
|
| 247 |
+
"loss": 0.9223,
|
| 248 |
+
"step": 3100
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 110.34,
|
| 252 |
+
"learning_rate": 2.588059701492537e-05,
|
| 253 |
+
"loss": 0.8839,
|
| 254 |
+
"step": 3200
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 113.79,
|
| 258 |
+
"learning_rate": 2.364179104477612e-05,
|
| 259 |
+
"loss": 0.8845,
|
| 260 |
+
"step": 3300
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 117.24,
|
| 264 |
+
"learning_rate": 2.1402985074626863e-05,
|
| 265 |
+
"loss": 0.8604,
|
| 266 |
+
"step": 3400
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 120.69,
|
| 270 |
+
"learning_rate": 1.9164179104477612e-05,
|
| 271 |
+
"loss": 0.8526,
|
| 272 |
+
"step": 3500
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 120.69,
|
| 276 |
+
"eval_loss": 0.43604278564453125,
|
| 277 |
+
"eval_runtime": 1.8975,
|
| 278 |
+
"eval_samples_per_second": 142.29,
|
| 279 |
+
"eval_steps_per_second": 4.743,
|
| 280 |
+
"eval_wer": 0.17044125465178098,
|
| 281 |
+
"step": 3500
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 124.14,
|
| 285 |
+
"learning_rate": 1.6925373134328355e-05,
|
| 286 |
+
"loss": 0.8311,
|
| 287 |
+
"step": 3600
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 127.59,
|
| 291 |
+
"learning_rate": 1.4686567164179104e-05,
|
| 292 |
+
"loss": 0.832,
|
| 293 |
+
"step": 3700
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 131.03,
|
| 297 |
+
"learning_rate": 1.244776119402985e-05,
|
| 298 |
+
"loss": 0.812,
|
| 299 |
+
"step": 3800
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 134.48,
|
| 303 |
+
"learning_rate": 1.0208955223880597e-05,
|
| 304 |
+
"loss": 0.8016,
|
| 305 |
+
"step": 3900
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 137.93,
|
| 309 |
+
"learning_rate": 7.970149253731343e-06,
|
| 310 |
+
"loss": 0.8018,
|
| 311 |
+
"step": 4000
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 137.93,
|
| 315 |
+
"eval_loss": 0.437362939119339,
|
| 316 |
+
"eval_runtime": 1.9635,
|
| 317 |
+
"eval_samples_per_second": 137.506,
|
| 318 |
+
"eval_steps_per_second": 4.584,
|
| 319 |
+
"eval_wer": 0.1720361509835194,
|
| 320 |
+
"step": 4000
|
| 321 |
+
}
|
| 322 |
+
],
|
| 323 |
+
"max_steps": 4350,
|
| 324 |
+
"num_train_epochs": 150,
|
| 325 |
+
"total_flos": 1.7217444822080553e+19,
|
| 326 |
+
"trial_name": null,
|
| 327 |
+
"trial_params": null
|
| 328 |
+
}
|
checkpoint-4000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f6f38b03969c3e29b358a2037b4f61da5251f02d2d9436641d5a5d94788db60e
|
| 3 |
+
size 2927
|
config.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "
|
| 3 |
-
"activation_dropout": 0.
|
| 4 |
"adapter_kernel_size": 3,
|
| 5 |
"adapter_stride": 2,
|
| 6 |
"add_adapter": false,
|
|
@@ -8,7 +8,7 @@
|
|
| 8 |
"architectures": [
|
| 9 |
"Wav2Vec2ForCTC"
|
| 10 |
],
|
| 11 |
-
"attention_dropout": 0.
|
| 12 |
"bos_token_id": 1,
|
| 13 |
"classifier_proj_size": 256,
|
| 14 |
"codevector_dim": 768,
|
|
@@ -41,7 +41,7 @@
|
|
| 41 |
2,
|
| 42 |
2
|
| 43 |
],
|
| 44 |
-
"ctc_loss_reduction": "
|
| 45 |
"ctc_zero_infinity": false,
|
| 46 |
"diversity_loss_weight": 0.1,
|
| 47 |
"do_stable_layer_norm": true,
|
|
@@ -49,23 +49,22 @@
|
|
| 49 |
"feat_extract_activation": "gelu",
|
| 50 |
"feat_extract_dropout": 0.0,
|
| 51 |
"feat_extract_norm": "layer",
|
| 52 |
-
"feat_proj_dropout": 0.
|
| 53 |
"feat_quantizer_dropout": 0.0,
|
| 54 |
"final_dropout": 0.0,
|
| 55 |
-
"gradient_checkpointing": false,
|
| 56 |
"hidden_act": "gelu",
|
| 57 |
-
"hidden_dropout": 0.
|
| 58 |
"hidden_size": 1024,
|
| 59 |
"initializer_range": 0.02,
|
| 60 |
"intermediate_size": 4096,
|
| 61 |
"layer_norm_eps": 1e-05,
|
| 62 |
-
"layerdrop": 0.
|
| 63 |
-
"mask_feature_length":
|
| 64 |
"mask_feature_min_masks": 0,
|
| 65 |
-
"mask_feature_prob": 0.
|
| 66 |
"mask_time_length": 10,
|
| 67 |
"mask_time_min_masks": 2,
|
| 68 |
-
"mask_time_prob": 0.
|
| 69 |
"model_type": "wav2vec2",
|
| 70 |
"num_adapter_layers": 3,
|
| 71 |
"num_attention_heads": 16,
|
|
@@ -77,7 +76,7 @@
|
|
| 77 |
"num_hidden_layers": 24,
|
| 78 |
"num_negatives": 100,
|
| 79 |
"output_hidden_size": 1024,
|
| 80 |
-
"pad_token_id":
|
| 81 |
"proj_codevector_dim": 768,
|
| 82 |
"tdnn_dilation": [
|
| 83 |
1,
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "./",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
"adapter_kernel_size": 3,
|
| 5 |
"adapter_stride": 2,
|
| 6 |
"add_adapter": false,
|
|
|
|
| 8 |
"architectures": [
|
| 9 |
"Wav2Vec2ForCTC"
|
| 10 |
],
|
| 11 |
+
"attention_dropout": 0.0,
|
| 12 |
"bos_token_id": 1,
|
| 13 |
"classifier_proj_size": 256,
|
| 14 |
"codevector_dim": 768,
|
|
|
|
| 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,
|
|
|
|
| 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 |
"hidden_act": "gelu",
|
| 56 |
+
"hidden_dropout": 0.0,
|
| 57 |
"hidden_size": 1024,
|
| 58 |
"initializer_range": 0.02,
|
| 59 |
"intermediate_size": 4096,
|
| 60 |
"layer_norm_eps": 1e-05,
|
| 61 |
+
"layerdrop": 0.0,
|
| 62 |
+
"mask_feature_length": 64,
|
| 63 |
"mask_feature_min_masks": 0,
|
| 64 |
+
"mask_feature_prob": 0.25,
|
| 65 |
"mask_time_length": 10,
|
| 66 |
"mask_time_min_masks": 2,
|
| 67 |
+
"mask_time_prob": 0.75,
|
| 68 |
"model_type": "wav2vec2",
|
| 69 |
"num_adapter_layers": 3,
|
| 70 |
"num_attention_heads": 16,
|
|
|
|
| 76 |
"num_hidden_layers": 24,
|
| 77 |
"num_negatives": 100,
|
| 78 |
"output_hidden_size": 1024,
|
| 79 |
+
"pad_token_id": 48,
|
| 80 |
"proj_codevector_dim": 768,
|
| 81 |
"tdnn_dilation": [
|
| 82 |
1,
|