Model first push
Browse files- checkpoint-17500/config.json +108 -0
- checkpoint-17500/optimizer.pt +3 -0
- checkpoint-17500/preprocessor_config.json +9 -0
- checkpoint-17500/pytorch_model.bin +3 -0
- checkpoint-17500/rng_state.pth +3 -0
- checkpoint-17500/scheduler.pt +3 -0
- checkpoint-17500/trainer_state.json +541 -0
- checkpoint-17500/training_args.bin +3 -0
- config.json +108 -0
- optimizer.pt +3 -0
- preprocessor_config.json +9 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- runs/Nov14_01-43-39_70156f070b9c/1668390306.856677/events.out.tfevents.1668390306.70156f070b9c.85.1 +3 -0
- runs/Nov14_01-43-39_70156f070b9c/events.out.tfevents.1668390306.70156f070b9c.85.0 +3 -0
- scheduler.pt +3 -0
- trainer_state.json +556 -0
- training_args.bin +3 -0
checkpoint-17500/config.json
ADDED
|
@@ -0,0 +1,108 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "yip-i/uaspeech-pretrained",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"adapter_kernel_size": 3,
|
| 5 |
+
"adapter_stride": 2,
|
| 6 |
+
"add_adapter": false,
|
| 7 |
+
"apply_spec_augment": true,
|
| 8 |
+
"architectures": [
|
| 9 |
+
"Wav2Vec2ForCTC"
|
| 10 |
+
],
|
| 11 |
+
"attention_dropout": 0.0,
|
| 12 |
+
"bos_token_id": 1,
|
| 13 |
+
"classifier_proj_size": 256,
|
| 14 |
+
"codevector_dim": 256,
|
| 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_dropout_prob": 0.0,
|
| 58 |
+
"hidden_size": 768,
|
| 59 |
+
"initializer_range": 0.02,
|
| 60 |
+
"intermediate_size": 3072,
|
| 61 |
+
"layer_norm_eps": 1e-05,
|
| 62 |
+
"layerdrop": 0.0,
|
| 63 |
+
"mask_feature_length": 10,
|
| 64 |
+
"mask_feature_min_masks": 0,
|
| 65 |
+
"mask_feature_prob": 0.0,
|
| 66 |
+
"mask_time_length": 10,
|
| 67 |
+
"mask_time_min_masks": 2,
|
| 68 |
+
"mask_time_prob": 0.65,
|
| 69 |
+
"model_type": "wav2vec2",
|
| 70 |
+
"num_adapter_layers": 3,
|
| 71 |
+
"num_attention_heads": 12,
|
| 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": 12,
|
| 78 |
+
"num_negatives": 100,
|
| 79 |
+
"output_hidden_size": 768,
|
| 80 |
+
"pad_token_id": 29,
|
| 81 |
+
"proj_codevector_dim": 256,
|
| 82 |
+
"tdnn_dilation": [
|
| 83 |
+
1,
|
| 84 |
+
2,
|
| 85 |
+
3,
|
| 86 |
+
1,
|
| 87 |
+
1
|
| 88 |
+
],
|
| 89 |
+
"tdnn_dim": [
|
| 90 |
+
512,
|
| 91 |
+
512,
|
| 92 |
+
512,
|
| 93 |
+
512,
|
| 94 |
+
1500
|
| 95 |
+
],
|
| 96 |
+
"tdnn_kernel": [
|
| 97 |
+
5,
|
| 98 |
+
3,
|
| 99 |
+
3,
|
| 100 |
+
1,
|
| 101 |
+
1
|
| 102 |
+
],
|
| 103 |
+
"torch_dtype": "float32",
|
| 104 |
+
"transformers_version": "4.23.1",
|
| 105 |
+
"use_weighted_layer_sum": false,
|
| 106 |
+
"vocab_size": 32,
|
| 107 |
+
"xvector_output_dim": 512
|
| 108 |
+
}
|
checkpoint-17500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:536b181a3a322aaf0b4d78d18751437c33b961e6655cc4c049ea3b207faaa6e9
|
| 3 |
+
size 721685265
|
checkpoint-17500/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": false,
|
| 8 |
+
"sampling_rate": 16000
|
| 9 |
+
}
|
checkpoint-17500/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d6ff44d661cae8cb3e5d07e2957fb41f2ea18f4544836f52130d34eddfd5089
|
| 3 |
+
size 377702321
|
checkpoint-17500/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8e06a5780cb4c5a5ab12f02e995c2100f2d68a7b1512c016d743690d9188cb6c
|
| 3 |
+
size 14567
|
checkpoint-17500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4f78aeb70f5e1d3045d4779dc880fe517a3726f5e89e28a06e1feee3b56dbde0
|
| 3 |
+
size 623
|
checkpoint-17500/trainer_state.json
ADDED
|
@@ -0,0 +1,541 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 28.40909090909091,
|
| 5 |
+
"global_step": 17500,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.81,
|
| 12 |
+
"learning_rate": 5e-05,
|
| 13 |
+
"loss": 8.6952,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.81,
|
| 18 |
+
"eval_loss": 3.0705761909484863,
|
| 19 |
+
"eval_runtime": 8.0191,
|
| 20 |
+
"eval_samples_per_second": 25.813,
|
| 21 |
+
"eval_steps_per_second": 3.242,
|
| 22 |
+
"eval_wer": 1.1315192743764173,
|
| 23 |
+
"step": 500
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 1.62,
|
| 27 |
+
"learning_rate": 0.0001,
|
| 28 |
+
"loss": 3.2723,
|
| 29 |
+
"step": 1000
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 1.62,
|
| 33 |
+
"eval_loss": 2.9638686180114746,
|
| 34 |
+
"eval_runtime": 8.0684,
|
| 35 |
+
"eval_samples_per_second": 25.656,
|
| 36 |
+
"eval_steps_per_second": 3.222,
|
| 37 |
+
"eval_wer": 1.0,
|
| 38 |
+
"step": 1000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 2.44,
|
| 42 |
+
"learning_rate": 9.713958810068651e-05,
|
| 43 |
+
"loss": 3.2191,
|
| 44 |
+
"step": 1500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 2.44,
|
| 48 |
+
"eval_loss": 3.0145716667175293,
|
| 49 |
+
"eval_runtime": 8.0711,
|
| 50 |
+
"eval_samples_per_second": 25.647,
|
| 51 |
+
"eval_steps_per_second": 3.221,
|
| 52 |
+
"eval_wer": 1.0,
|
| 53 |
+
"step": 1500
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 3.25,
|
| 57 |
+
"learning_rate": 9.4279176201373e-05,
|
| 58 |
+
"loss": 3.0698,
|
| 59 |
+
"step": 2000
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 3.25,
|
| 63 |
+
"eval_loss": 2.9972307682037354,
|
| 64 |
+
"eval_runtime": 8.0326,
|
| 65 |
+
"eval_samples_per_second": 25.77,
|
| 66 |
+
"eval_steps_per_second": 3.237,
|
| 67 |
+
"eval_wer": 1.3356009070294785,
|
| 68 |
+
"step": 2000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 4.06,
|
| 72 |
+
"learning_rate": 9.14187643020595e-05,
|
| 73 |
+
"loss": 3.1918,
|
| 74 |
+
"step": 2500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 4.06,
|
| 78 |
+
"eval_loss": 2.9555039405822754,
|
| 79 |
+
"eval_runtime": 8.0447,
|
| 80 |
+
"eval_samples_per_second": 25.731,
|
| 81 |
+
"eval_steps_per_second": 3.232,
|
| 82 |
+
"eval_wer": 1.0,
|
| 83 |
+
"step": 2500
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 4.87,
|
| 87 |
+
"learning_rate": 8.8558352402746e-05,
|
| 88 |
+
"loss": 3.0932,
|
| 89 |
+
"step": 3000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 4.87,
|
| 93 |
+
"eval_loss": 2.9776482582092285,
|
| 94 |
+
"eval_runtime": 8.053,
|
| 95 |
+
"eval_samples_per_second": 25.705,
|
| 96 |
+
"eval_steps_per_second": 3.229,
|
| 97 |
+
"eval_wer": 1.0,
|
| 98 |
+
"step": 3000
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 5.68,
|
| 102 |
+
"learning_rate": 8.569794050343249e-05,
|
| 103 |
+
"loss": 3.2271,
|
| 104 |
+
"step": 3500
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 5.68,
|
| 108 |
+
"eval_loss": 3.055955410003662,
|
| 109 |
+
"eval_runtime": 8.1309,
|
| 110 |
+
"eval_samples_per_second": 25.458,
|
| 111 |
+
"eval_steps_per_second": 3.198,
|
| 112 |
+
"eval_wer": 1.3378684807256236,
|
| 113 |
+
"step": 3500
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 6.49,
|
| 117 |
+
"learning_rate": 8.283752860411899e-05,
|
| 118 |
+
"loss": 3.2925,
|
| 119 |
+
"step": 4000
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 6.49,
|
| 123 |
+
"eval_loss": 2.9724416732788086,
|
| 124 |
+
"eval_runtime": 8.0677,
|
| 125 |
+
"eval_samples_per_second": 25.658,
|
| 126 |
+
"eval_steps_per_second": 3.223,
|
| 127 |
+
"eval_wer": 1.3900226757369614,
|
| 128 |
+
"step": 4000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 7.31,
|
| 132 |
+
"learning_rate": 7.99771167048055e-05,
|
| 133 |
+
"loss": 3.2195,
|
| 134 |
+
"step": 4500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 7.31,
|
| 138 |
+
"eval_loss": 3.5231425762176514,
|
| 139 |
+
"eval_runtime": 8.1433,
|
| 140 |
+
"eval_samples_per_second": 25.42,
|
| 141 |
+
"eval_steps_per_second": 3.193,
|
| 142 |
+
"eval_wer": 1.0,
|
| 143 |
+
"step": 4500
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 8.12,
|
| 147 |
+
"learning_rate": 7.711670480549199e-05,
|
| 148 |
+
"loss": 3.3582,
|
| 149 |
+
"step": 5000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 8.12,
|
| 153 |
+
"eval_loss": 3.749286413192749,
|
| 154 |
+
"eval_runtime": 8.0117,
|
| 155 |
+
"eval_samples_per_second": 25.837,
|
| 156 |
+
"eval_steps_per_second": 3.245,
|
| 157 |
+
"eval_wer": 1.0249433106575965,
|
| 158 |
+
"step": 5000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 8.93,
|
| 162 |
+
"learning_rate": 7.42562929061785e-05,
|
| 163 |
+
"loss": 3.3233,
|
| 164 |
+
"step": 5500
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 8.93,
|
| 168 |
+
"eval_loss": 3.3524105548858643,
|
| 169 |
+
"eval_runtime": 8.2055,
|
| 170 |
+
"eval_samples_per_second": 25.227,
|
| 171 |
+
"eval_steps_per_second": 3.169,
|
| 172 |
+
"eval_wer": 1.0,
|
| 173 |
+
"step": 5500
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 9.74,
|
| 177 |
+
"learning_rate": 7.1395881006865e-05,
|
| 178 |
+
"loss": 3.1679,
|
| 179 |
+
"step": 6000
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 9.74,
|
| 183 |
+
"eval_loss": 3.186889410018921,
|
| 184 |
+
"eval_runtime": 8.0182,
|
| 185 |
+
"eval_samples_per_second": 25.816,
|
| 186 |
+
"eval_steps_per_second": 3.243,
|
| 187 |
+
"eval_wer": 1.3877551020408163,
|
| 188 |
+
"step": 6000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 10.55,
|
| 192 |
+
"learning_rate": 6.853546910755149e-05,
|
| 193 |
+
"loss": 3.1393,
|
| 194 |
+
"step": 6500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 10.55,
|
| 198 |
+
"eval_loss": 3.0084877014160156,
|
| 199 |
+
"eval_runtime": 8.0014,
|
| 200 |
+
"eval_samples_per_second": 25.871,
|
| 201 |
+
"eval_steps_per_second": 3.249,
|
| 202 |
+
"eval_wer": 1.2947845804988662,
|
| 203 |
+
"step": 6500
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 11.36,
|
| 207 |
+
"learning_rate": 6.5675057208238e-05,
|
| 208 |
+
"loss": 3.1699,
|
| 209 |
+
"step": 7000
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 11.36,
|
| 213 |
+
"eval_loss": 2.997206449508667,
|
| 214 |
+
"eval_runtime": 8.1145,
|
| 215 |
+
"eval_samples_per_second": 25.51,
|
| 216 |
+
"eval_steps_per_second": 3.204,
|
| 217 |
+
"eval_wer": 1.1337868480725624,
|
| 218 |
+
"step": 7000
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 12.18,
|
| 222 |
+
"learning_rate": 6.281464530892449e-05,
|
| 223 |
+
"loss": 3.3382,
|
| 224 |
+
"step": 7500
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 12.18,
|
| 228 |
+
"eval_loss": 2.9686625003814697,
|
| 229 |
+
"eval_runtime": 8.0023,
|
| 230 |
+
"eval_samples_per_second": 25.868,
|
| 231 |
+
"eval_steps_per_second": 3.249,
|
| 232 |
+
"eval_wer": 1.3877551020408163,
|
| 233 |
+
"step": 7500
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 12.99,
|
| 237 |
+
"learning_rate": 5.9954233409610984e-05,
|
| 238 |
+
"loss": 3.0454,
|
| 239 |
+
"step": 8000
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 12.99,
|
| 243 |
+
"eval_loss": 2.968928337097168,
|
| 244 |
+
"eval_runtime": 8.2051,
|
| 245 |
+
"eval_samples_per_second": 25.228,
|
| 246 |
+
"eval_steps_per_second": 3.169,
|
| 247 |
+
"eval_wer": 1.383219954648526,
|
| 248 |
+
"step": 8000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 13.8,
|
| 252 |
+
"learning_rate": 5.709382151029748e-05,
|
| 253 |
+
"loss": 3.0609,
|
| 254 |
+
"step": 8500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 13.8,
|
| 258 |
+
"eval_loss": 2.902815580368042,
|
| 259 |
+
"eval_runtime": 8.0338,
|
| 260 |
+
"eval_samples_per_second": 25.766,
|
| 261 |
+
"eval_steps_per_second": 3.236,
|
| 262 |
+
"eval_wer": 1.3900226757369614,
|
| 263 |
+
"step": 8500
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 14.61,
|
| 267 |
+
"learning_rate": 5.423340961098399e-05,
|
| 268 |
+
"loss": 3.0224,
|
| 269 |
+
"step": 9000
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 14.61,
|
| 273 |
+
"eval_loss": 2.9064831733703613,
|
| 274 |
+
"eval_runtime": 8.2142,
|
| 275 |
+
"eval_samples_per_second": 25.2,
|
| 276 |
+
"eval_steps_per_second": 3.165,
|
| 277 |
+
"eval_wer": 1.3877551020408163,
|
| 278 |
+
"step": 9000
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 15.42,
|
| 282 |
+
"learning_rate": 5.137299771167048e-05,
|
| 283 |
+
"loss": 3.0156,
|
| 284 |
+
"step": 9500
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 15.42,
|
| 288 |
+
"eval_loss": 2.8854894638061523,
|
| 289 |
+
"eval_runtime": 7.9943,
|
| 290 |
+
"eval_samples_per_second": 25.893,
|
| 291 |
+
"eval_steps_per_second": 3.252,
|
| 292 |
+
"eval_wer": 1.3900226757369614,
|
| 293 |
+
"step": 9500
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 16.23,
|
| 297 |
+
"learning_rate": 4.851258581235698e-05,
|
| 298 |
+
"loss": 3.0317,
|
| 299 |
+
"step": 10000
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 16.23,
|
| 303 |
+
"eval_loss": 2.8956820964813232,
|
| 304 |
+
"eval_runtime": 8.1264,
|
| 305 |
+
"eval_samples_per_second": 25.473,
|
| 306 |
+
"eval_steps_per_second": 3.199,
|
| 307 |
+
"eval_wer": 1.3900226757369614,
|
| 308 |
+
"step": 10000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 17.05,
|
| 312 |
+
"learning_rate": 4.565217391304348e-05,
|
| 313 |
+
"loss": 3.0184,
|
| 314 |
+
"step": 10500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 17.05,
|
| 318 |
+
"eval_loss": 2.88946270942688,
|
| 319 |
+
"eval_runtime": 8.0387,
|
| 320 |
+
"eval_samples_per_second": 25.751,
|
| 321 |
+
"eval_steps_per_second": 3.234,
|
| 322 |
+
"eval_wer": 1.3900226757369614,
|
| 323 |
+
"step": 10500
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 17.86,
|
| 327 |
+
"learning_rate": 4.279176201372998e-05,
|
| 328 |
+
"loss": 3.0852,
|
| 329 |
+
"step": 11000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 17.86,
|
| 333 |
+
"eval_loss": 2.8936383724212646,
|
| 334 |
+
"eval_runtime": 8.1682,
|
| 335 |
+
"eval_samples_per_second": 25.342,
|
| 336 |
+
"eval_steps_per_second": 3.183,
|
| 337 |
+
"eval_wer": 1.3900226757369614,
|
| 338 |
+
"step": 11000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 18.67,
|
| 342 |
+
"learning_rate": 3.993135011441648e-05,
|
| 343 |
+
"loss": 3.0017,
|
| 344 |
+
"step": 11500
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 18.67,
|
| 348 |
+
"eval_loss": 2.8695127964019775,
|
| 349 |
+
"eval_runtime": 8.169,
|
| 350 |
+
"eval_samples_per_second": 25.34,
|
| 351 |
+
"eval_steps_per_second": 3.183,
|
| 352 |
+
"eval_wer": 1.3900226757369614,
|
| 353 |
+
"step": 11500
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 19.48,
|
| 357 |
+
"learning_rate": 3.707093821510298e-05,
|
| 358 |
+
"loss": 2.9337,
|
| 359 |
+
"step": 12000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 19.48,
|
| 363 |
+
"eval_loss": 2.8768134117126465,
|
| 364 |
+
"eval_runtime": 8.1024,
|
| 365 |
+
"eval_samples_per_second": 25.548,
|
| 366 |
+
"eval_steps_per_second": 3.209,
|
| 367 |
+
"eval_wer": 1.3900226757369614,
|
| 368 |
+
"step": 12000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 20.29,
|
| 372 |
+
"learning_rate": 3.421052631578947e-05,
|
| 373 |
+
"loss": 3.0017,
|
| 374 |
+
"step": 12500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 20.29,
|
| 378 |
+
"eval_loss": 2.8580081462860107,
|
| 379 |
+
"eval_runtime": 8.1811,
|
| 380 |
+
"eval_samples_per_second": 25.302,
|
| 381 |
+
"eval_steps_per_second": 3.178,
|
| 382 |
+
"eval_wer": 1.3900226757369614,
|
| 383 |
+
"step": 12500
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 21.1,
|
| 387 |
+
"learning_rate": 3.135011441647597e-05,
|
| 388 |
+
"loss": 2.9472,
|
| 389 |
+
"step": 13000
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 21.1,
|
| 393 |
+
"eval_loss": 2.839965343475342,
|
| 394 |
+
"eval_runtime": 8.2839,
|
| 395 |
+
"eval_samples_per_second": 24.988,
|
| 396 |
+
"eval_steps_per_second": 3.139,
|
| 397 |
+
"eval_wer": 1.3900226757369614,
|
| 398 |
+
"step": 13000
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 21.92,
|
| 402 |
+
"learning_rate": 2.8489702517162476e-05,
|
| 403 |
+
"loss": 3.0214,
|
| 404 |
+
"step": 13500
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 21.92,
|
| 408 |
+
"eval_loss": 2.856123924255371,
|
| 409 |
+
"eval_runtime": 8.1176,
|
| 410 |
+
"eval_samples_per_second": 25.5,
|
| 411 |
+
"eval_steps_per_second": 3.203,
|
| 412 |
+
"eval_wer": 1.3900226757369614,
|
| 413 |
+
"step": 13500
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 22.73,
|
| 417 |
+
"learning_rate": 2.562929061784897e-05,
|
| 418 |
+
"loss": 2.9336,
|
| 419 |
+
"step": 14000
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 22.73,
|
| 423 |
+
"eval_loss": 2.879206895828247,
|
| 424 |
+
"eval_runtime": 8.1391,
|
| 425 |
+
"eval_samples_per_second": 25.433,
|
| 426 |
+
"eval_steps_per_second": 3.194,
|
| 427 |
+
"eval_wer": 1.3900226757369614,
|
| 428 |
+
"step": 14000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 23.54,
|
| 432 |
+
"learning_rate": 2.276887871853547e-05,
|
| 433 |
+
"loss": 3.0134,
|
| 434 |
+
"step": 14500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 23.54,
|
| 438 |
+
"eval_loss": 2.8472487926483154,
|
| 439 |
+
"eval_runtime": 8.0887,
|
| 440 |
+
"eval_samples_per_second": 25.591,
|
| 441 |
+
"eval_steps_per_second": 3.214,
|
| 442 |
+
"eval_wer": 1.3900226757369614,
|
| 443 |
+
"step": 14500
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 24.35,
|
| 447 |
+
"learning_rate": 1.990846681922197e-05,
|
| 448 |
+
"loss": 2.9433,
|
| 449 |
+
"step": 15000
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 24.35,
|
| 453 |
+
"eval_loss": 2.8818936347961426,
|
| 454 |
+
"eval_runtime": 8.193,
|
| 455 |
+
"eval_samples_per_second": 25.266,
|
| 456 |
+
"eval_steps_per_second": 3.173,
|
| 457 |
+
"eval_wer": 1.3900226757369614,
|
| 458 |
+
"step": 15000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 25.16,
|
| 462 |
+
"learning_rate": 1.7048054919908468e-05,
|
| 463 |
+
"loss": 2.8536,
|
| 464 |
+
"step": 15500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 25.16,
|
| 468 |
+
"eval_loss": 2.837463140487671,
|
| 469 |
+
"eval_runtime": 8.1663,
|
| 470 |
+
"eval_samples_per_second": 25.348,
|
| 471 |
+
"eval_steps_per_second": 3.184,
|
| 472 |
+
"eval_wer": 1.3900226757369614,
|
| 473 |
+
"step": 15500
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 25.97,
|
| 477 |
+
"learning_rate": 1.4187643020594965e-05,
|
| 478 |
+
"loss": 2.8742,
|
| 479 |
+
"step": 16000
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 25.97,
|
| 483 |
+
"eval_loss": 2.857389450073242,
|
| 484 |
+
"eval_runtime": 8.194,
|
| 485 |
+
"eval_samples_per_second": 25.262,
|
| 486 |
+
"eval_steps_per_second": 3.173,
|
| 487 |
+
"eval_wer": 1.3900226757369614,
|
| 488 |
+
"step": 16000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 26.79,
|
| 492 |
+
"learning_rate": 1.1327231121281464e-05,
|
| 493 |
+
"loss": 2.8298,
|
| 494 |
+
"step": 16500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 26.79,
|
| 498 |
+
"eval_loss": 2.982081651687622,
|
| 499 |
+
"eval_runtime": 8.1935,
|
| 500 |
+
"eval_samples_per_second": 25.264,
|
| 501 |
+
"eval_steps_per_second": 3.173,
|
| 502 |
+
"eval_wer": 1.3900226757369614,
|
| 503 |
+
"step": 16500
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 27.6,
|
| 507 |
+
"learning_rate": 8.466819221967964e-06,
|
| 508 |
+
"loss": 2.7439,
|
| 509 |
+
"step": 17000
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 27.6,
|
| 513 |
+
"eval_loss": 3.2741587162017822,
|
| 514 |
+
"eval_runtime": 8.2267,
|
| 515 |
+
"eval_samples_per_second": 25.162,
|
| 516 |
+
"eval_steps_per_second": 3.16,
|
| 517 |
+
"eval_wer": 1.3900226757369614,
|
| 518 |
+
"step": 17000
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 28.41,
|
| 522 |
+
"learning_rate": 5.606407322654463e-06,
|
| 523 |
+
"loss": 2.7008,
|
| 524 |
+
"step": 17500
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 28.41,
|
| 528 |
+
"eval_loss": 3.365966320037842,
|
| 529 |
+
"eval_runtime": 8.203,
|
| 530 |
+
"eval_samples_per_second": 25.235,
|
| 531 |
+
"eval_steps_per_second": 3.17,
|
| 532 |
+
"eval_wer": 1.3900226757369614,
|
| 533 |
+
"step": 17500
|
| 534 |
+
}
|
| 535 |
+
],
|
| 536 |
+
"max_steps": 18480,
|
| 537 |
+
"num_train_epochs": 30,
|
| 538 |
+
"total_flos": 3.9780154122268984e+18,
|
| 539 |
+
"trial_name": null,
|
| 540 |
+
"trial_params": null
|
| 541 |
+
}
|
checkpoint-17500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:151803200d49964e6d9f1364dede5203d89e69881055e526220d7cecd61ffdfb
|
| 3 |
+
size 3375
|
config.json
ADDED
|
@@ -0,0 +1,108 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "yip-i/uaspeech-pretrained",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"adapter_kernel_size": 3,
|
| 5 |
+
"adapter_stride": 2,
|
| 6 |
+
"add_adapter": false,
|
| 7 |
+
"apply_spec_augment": true,
|
| 8 |
+
"architectures": [
|
| 9 |
+
"Wav2Vec2ForCTC"
|
| 10 |
+
],
|
| 11 |
+
"attention_dropout": 0.0,
|
| 12 |
+
"bos_token_id": 1,
|
| 13 |
+
"classifier_proj_size": 256,
|
| 14 |
+
"codevector_dim": 256,
|
| 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_dropout_prob": 0.0,
|
| 58 |
+
"hidden_size": 768,
|
| 59 |
+
"initializer_range": 0.02,
|
| 60 |
+
"intermediate_size": 3072,
|
| 61 |
+
"layer_norm_eps": 1e-05,
|
| 62 |
+
"layerdrop": 0.0,
|
| 63 |
+
"mask_feature_length": 10,
|
| 64 |
+
"mask_feature_min_masks": 0,
|
| 65 |
+
"mask_feature_prob": 0.0,
|
| 66 |
+
"mask_time_length": 10,
|
| 67 |
+
"mask_time_min_masks": 2,
|
| 68 |
+
"mask_time_prob": 0.65,
|
| 69 |
+
"model_type": "wav2vec2",
|
| 70 |
+
"num_adapter_layers": 3,
|
| 71 |
+
"num_attention_heads": 12,
|
| 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": 12,
|
| 78 |
+
"num_negatives": 100,
|
| 79 |
+
"output_hidden_size": 768,
|
| 80 |
+
"pad_token_id": 29,
|
| 81 |
+
"proj_codevector_dim": 256,
|
| 82 |
+
"tdnn_dilation": [
|
| 83 |
+
1,
|
| 84 |
+
2,
|
| 85 |
+
3,
|
| 86 |
+
1,
|
| 87 |
+
1
|
| 88 |
+
],
|
| 89 |
+
"tdnn_dim": [
|
| 90 |
+
512,
|
| 91 |
+
512,
|
| 92 |
+
512,
|
| 93 |
+
512,
|
| 94 |
+
1500
|
| 95 |
+
],
|
| 96 |
+
"tdnn_kernel": [
|
| 97 |
+
5,
|
| 98 |
+
3,
|
| 99 |
+
3,
|
| 100 |
+
1,
|
| 101 |
+
1
|
| 102 |
+
],
|
| 103 |
+
"torch_dtype": "float32",
|
| 104 |
+
"transformers_version": "4.23.1",
|
| 105 |
+
"use_weighted_layer_sum": false,
|
| 106 |
+
"vocab_size": 32,
|
| 107 |
+
"xvector_output_dim": 512
|
| 108 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:16c9cd721072f1285d9a09541b16d36154b0a65769ba2304ebd3cf1a47a3ccd9
|
| 3 |
+
size 721685265
|
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": false,
|
| 8 |
+
"sampling_rate": 16000
|
| 9 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c3b8ce0b861b34249acd899833cc5d17966f7c3820981c93099f3388451d2d17
|
| 3 |
+
size 377702321
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:49daf610bd0be93ec34116c5dabb671af4b737094b6d5999d22a7ff4458cd59e
|
| 3 |
+
size 14503
|
runs/Nov14_01-43-39_70156f070b9c/1668390306.856677/events.out.tfevents.1668390306.70156f070b9c.85.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0ca0942bb58cf7715bde732b0a6e3d088ea04147768ab715db2dee9de56e0a1f
|
| 3 |
+
size 5471
|
runs/Nov14_01-43-39_70156f070b9c/events.out.tfevents.1668390306.70156f070b9c.85.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:62b5c681d69c7373b8ec47ffc5e62b6cbd4ccfd518965e2f66ceae9414986bf6
|
| 3 |
+
size 22613
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b5d4a13226e6eeef330f9ea7271fdddd531065e7435f741b314c8ee6f8e1c979
|
| 3 |
+
size 623
|
trainer_state.json
ADDED
|
@@ -0,0 +1,556 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 29.22077922077922,
|
| 5 |
+
"global_step": 18000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.81,
|
| 12 |
+
"learning_rate": 5e-05,
|
| 13 |
+
"loss": 8.6952,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.81,
|
| 18 |
+
"eval_loss": 3.0705761909484863,
|
| 19 |
+
"eval_runtime": 8.0191,
|
| 20 |
+
"eval_samples_per_second": 25.813,
|
| 21 |
+
"eval_steps_per_second": 3.242,
|
| 22 |
+
"eval_wer": 1.1315192743764173,
|
| 23 |
+
"step": 500
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 1.62,
|
| 27 |
+
"learning_rate": 0.0001,
|
| 28 |
+
"loss": 3.2723,
|
| 29 |
+
"step": 1000
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 1.62,
|
| 33 |
+
"eval_loss": 2.9638686180114746,
|
| 34 |
+
"eval_runtime": 8.0684,
|
| 35 |
+
"eval_samples_per_second": 25.656,
|
| 36 |
+
"eval_steps_per_second": 3.222,
|
| 37 |
+
"eval_wer": 1.0,
|
| 38 |
+
"step": 1000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 2.44,
|
| 42 |
+
"learning_rate": 9.713958810068651e-05,
|
| 43 |
+
"loss": 3.2191,
|
| 44 |
+
"step": 1500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 2.44,
|
| 48 |
+
"eval_loss": 3.0145716667175293,
|
| 49 |
+
"eval_runtime": 8.0711,
|
| 50 |
+
"eval_samples_per_second": 25.647,
|
| 51 |
+
"eval_steps_per_second": 3.221,
|
| 52 |
+
"eval_wer": 1.0,
|
| 53 |
+
"step": 1500
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 3.25,
|
| 57 |
+
"learning_rate": 9.4279176201373e-05,
|
| 58 |
+
"loss": 3.0698,
|
| 59 |
+
"step": 2000
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 3.25,
|
| 63 |
+
"eval_loss": 2.9972307682037354,
|
| 64 |
+
"eval_runtime": 8.0326,
|
| 65 |
+
"eval_samples_per_second": 25.77,
|
| 66 |
+
"eval_steps_per_second": 3.237,
|
| 67 |
+
"eval_wer": 1.3356009070294785,
|
| 68 |
+
"step": 2000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 4.06,
|
| 72 |
+
"learning_rate": 9.14187643020595e-05,
|
| 73 |
+
"loss": 3.1918,
|
| 74 |
+
"step": 2500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 4.06,
|
| 78 |
+
"eval_loss": 2.9555039405822754,
|
| 79 |
+
"eval_runtime": 8.0447,
|
| 80 |
+
"eval_samples_per_second": 25.731,
|
| 81 |
+
"eval_steps_per_second": 3.232,
|
| 82 |
+
"eval_wer": 1.0,
|
| 83 |
+
"step": 2500
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 4.87,
|
| 87 |
+
"learning_rate": 8.8558352402746e-05,
|
| 88 |
+
"loss": 3.0932,
|
| 89 |
+
"step": 3000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 4.87,
|
| 93 |
+
"eval_loss": 2.9776482582092285,
|
| 94 |
+
"eval_runtime": 8.053,
|
| 95 |
+
"eval_samples_per_second": 25.705,
|
| 96 |
+
"eval_steps_per_second": 3.229,
|
| 97 |
+
"eval_wer": 1.0,
|
| 98 |
+
"step": 3000
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 5.68,
|
| 102 |
+
"learning_rate": 8.569794050343249e-05,
|
| 103 |
+
"loss": 3.2271,
|
| 104 |
+
"step": 3500
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 5.68,
|
| 108 |
+
"eval_loss": 3.055955410003662,
|
| 109 |
+
"eval_runtime": 8.1309,
|
| 110 |
+
"eval_samples_per_second": 25.458,
|
| 111 |
+
"eval_steps_per_second": 3.198,
|
| 112 |
+
"eval_wer": 1.3378684807256236,
|
| 113 |
+
"step": 3500
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 6.49,
|
| 117 |
+
"learning_rate": 8.283752860411899e-05,
|
| 118 |
+
"loss": 3.2925,
|
| 119 |
+
"step": 4000
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 6.49,
|
| 123 |
+
"eval_loss": 2.9724416732788086,
|
| 124 |
+
"eval_runtime": 8.0677,
|
| 125 |
+
"eval_samples_per_second": 25.658,
|
| 126 |
+
"eval_steps_per_second": 3.223,
|
| 127 |
+
"eval_wer": 1.3900226757369614,
|
| 128 |
+
"step": 4000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 7.31,
|
| 132 |
+
"learning_rate": 7.99771167048055e-05,
|
| 133 |
+
"loss": 3.2195,
|
| 134 |
+
"step": 4500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 7.31,
|
| 138 |
+
"eval_loss": 3.5231425762176514,
|
| 139 |
+
"eval_runtime": 8.1433,
|
| 140 |
+
"eval_samples_per_second": 25.42,
|
| 141 |
+
"eval_steps_per_second": 3.193,
|
| 142 |
+
"eval_wer": 1.0,
|
| 143 |
+
"step": 4500
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 8.12,
|
| 147 |
+
"learning_rate": 7.711670480549199e-05,
|
| 148 |
+
"loss": 3.3582,
|
| 149 |
+
"step": 5000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 8.12,
|
| 153 |
+
"eval_loss": 3.749286413192749,
|
| 154 |
+
"eval_runtime": 8.0117,
|
| 155 |
+
"eval_samples_per_second": 25.837,
|
| 156 |
+
"eval_steps_per_second": 3.245,
|
| 157 |
+
"eval_wer": 1.0249433106575965,
|
| 158 |
+
"step": 5000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 8.93,
|
| 162 |
+
"learning_rate": 7.42562929061785e-05,
|
| 163 |
+
"loss": 3.3233,
|
| 164 |
+
"step": 5500
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 8.93,
|
| 168 |
+
"eval_loss": 3.3524105548858643,
|
| 169 |
+
"eval_runtime": 8.2055,
|
| 170 |
+
"eval_samples_per_second": 25.227,
|
| 171 |
+
"eval_steps_per_second": 3.169,
|
| 172 |
+
"eval_wer": 1.0,
|
| 173 |
+
"step": 5500
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 9.74,
|
| 177 |
+
"learning_rate": 7.1395881006865e-05,
|
| 178 |
+
"loss": 3.1679,
|
| 179 |
+
"step": 6000
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 9.74,
|
| 183 |
+
"eval_loss": 3.186889410018921,
|
| 184 |
+
"eval_runtime": 8.0182,
|
| 185 |
+
"eval_samples_per_second": 25.816,
|
| 186 |
+
"eval_steps_per_second": 3.243,
|
| 187 |
+
"eval_wer": 1.3877551020408163,
|
| 188 |
+
"step": 6000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 10.55,
|
| 192 |
+
"learning_rate": 6.853546910755149e-05,
|
| 193 |
+
"loss": 3.1393,
|
| 194 |
+
"step": 6500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 10.55,
|
| 198 |
+
"eval_loss": 3.0084877014160156,
|
| 199 |
+
"eval_runtime": 8.0014,
|
| 200 |
+
"eval_samples_per_second": 25.871,
|
| 201 |
+
"eval_steps_per_second": 3.249,
|
| 202 |
+
"eval_wer": 1.2947845804988662,
|
| 203 |
+
"step": 6500
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 11.36,
|
| 207 |
+
"learning_rate": 6.5675057208238e-05,
|
| 208 |
+
"loss": 3.1699,
|
| 209 |
+
"step": 7000
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 11.36,
|
| 213 |
+
"eval_loss": 2.997206449508667,
|
| 214 |
+
"eval_runtime": 8.1145,
|
| 215 |
+
"eval_samples_per_second": 25.51,
|
| 216 |
+
"eval_steps_per_second": 3.204,
|
| 217 |
+
"eval_wer": 1.1337868480725624,
|
| 218 |
+
"step": 7000
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 12.18,
|
| 222 |
+
"learning_rate": 6.281464530892449e-05,
|
| 223 |
+
"loss": 3.3382,
|
| 224 |
+
"step": 7500
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 12.18,
|
| 228 |
+
"eval_loss": 2.9686625003814697,
|
| 229 |
+
"eval_runtime": 8.0023,
|
| 230 |
+
"eval_samples_per_second": 25.868,
|
| 231 |
+
"eval_steps_per_second": 3.249,
|
| 232 |
+
"eval_wer": 1.3877551020408163,
|
| 233 |
+
"step": 7500
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 12.99,
|
| 237 |
+
"learning_rate": 5.9954233409610984e-05,
|
| 238 |
+
"loss": 3.0454,
|
| 239 |
+
"step": 8000
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 12.99,
|
| 243 |
+
"eval_loss": 2.968928337097168,
|
| 244 |
+
"eval_runtime": 8.2051,
|
| 245 |
+
"eval_samples_per_second": 25.228,
|
| 246 |
+
"eval_steps_per_second": 3.169,
|
| 247 |
+
"eval_wer": 1.383219954648526,
|
| 248 |
+
"step": 8000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 13.8,
|
| 252 |
+
"learning_rate": 5.709382151029748e-05,
|
| 253 |
+
"loss": 3.0609,
|
| 254 |
+
"step": 8500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 13.8,
|
| 258 |
+
"eval_loss": 2.902815580368042,
|
| 259 |
+
"eval_runtime": 8.0338,
|
| 260 |
+
"eval_samples_per_second": 25.766,
|
| 261 |
+
"eval_steps_per_second": 3.236,
|
| 262 |
+
"eval_wer": 1.3900226757369614,
|
| 263 |
+
"step": 8500
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 14.61,
|
| 267 |
+
"learning_rate": 5.423340961098399e-05,
|
| 268 |
+
"loss": 3.0224,
|
| 269 |
+
"step": 9000
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 14.61,
|
| 273 |
+
"eval_loss": 2.9064831733703613,
|
| 274 |
+
"eval_runtime": 8.2142,
|
| 275 |
+
"eval_samples_per_second": 25.2,
|
| 276 |
+
"eval_steps_per_second": 3.165,
|
| 277 |
+
"eval_wer": 1.3877551020408163,
|
| 278 |
+
"step": 9000
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 15.42,
|
| 282 |
+
"learning_rate": 5.137299771167048e-05,
|
| 283 |
+
"loss": 3.0156,
|
| 284 |
+
"step": 9500
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 15.42,
|
| 288 |
+
"eval_loss": 2.8854894638061523,
|
| 289 |
+
"eval_runtime": 7.9943,
|
| 290 |
+
"eval_samples_per_second": 25.893,
|
| 291 |
+
"eval_steps_per_second": 3.252,
|
| 292 |
+
"eval_wer": 1.3900226757369614,
|
| 293 |
+
"step": 9500
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 16.23,
|
| 297 |
+
"learning_rate": 4.851258581235698e-05,
|
| 298 |
+
"loss": 3.0317,
|
| 299 |
+
"step": 10000
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 16.23,
|
| 303 |
+
"eval_loss": 2.8956820964813232,
|
| 304 |
+
"eval_runtime": 8.1264,
|
| 305 |
+
"eval_samples_per_second": 25.473,
|
| 306 |
+
"eval_steps_per_second": 3.199,
|
| 307 |
+
"eval_wer": 1.3900226757369614,
|
| 308 |
+
"step": 10000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 17.05,
|
| 312 |
+
"learning_rate": 4.565217391304348e-05,
|
| 313 |
+
"loss": 3.0184,
|
| 314 |
+
"step": 10500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 17.05,
|
| 318 |
+
"eval_loss": 2.88946270942688,
|
| 319 |
+
"eval_runtime": 8.0387,
|
| 320 |
+
"eval_samples_per_second": 25.751,
|
| 321 |
+
"eval_steps_per_second": 3.234,
|
| 322 |
+
"eval_wer": 1.3900226757369614,
|
| 323 |
+
"step": 10500
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 17.86,
|
| 327 |
+
"learning_rate": 4.279176201372998e-05,
|
| 328 |
+
"loss": 3.0852,
|
| 329 |
+
"step": 11000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 17.86,
|
| 333 |
+
"eval_loss": 2.8936383724212646,
|
| 334 |
+
"eval_runtime": 8.1682,
|
| 335 |
+
"eval_samples_per_second": 25.342,
|
| 336 |
+
"eval_steps_per_second": 3.183,
|
| 337 |
+
"eval_wer": 1.3900226757369614,
|
| 338 |
+
"step": 11000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 18.67,
|
| 342 |
+
"learning_rate": 3.993135011441648e-05,
|
| 343 |
+
"loss": 3.0017,
|
| 344 |
+
"step": 11500
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 18.67,
|
| 348 |
+
"eval_loss": 2.8695127964019775,
|
| 349 |
+
"eval_runtime": 8.169,
|
| 350 |
+
"eval_samples_per_second": 25.34,
|
| 351 |
+
"eval_steps_per_second": 3.183,
|
| 352 |
+
"eval_wer": 1.3900226757369614,
|
| 353 |
+
"step": 11500
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 19.48,
|
| 357 |
+
"learning_rate": 3.707093821510298e-05,
|
| 358 |
+
"loss": 2.9337,
|
| 359 |
+
"step": 12000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 19.48,
|
| 363 |
+
"eval_loss": 2.8768134117126465,
|
| 364 |
+
"eval_runtime": 8.1024,
|
| 365 |
+
"eval_samples_per_second": 25.548,
|
| 366 |
+
"eval_steps_per_second": 3.209,
|
| 367 |
+
"eval_wer": 1.3900226757369614,
|
| 368 |
+
"step": 12000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 20.29,
|
| 372 |
+
"learning_rate": 3.421052631578947e-05,
|
| 373 |
+
"loss": 3.0017,
|
| 374 |
+
"step": 12500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 20.29,
|
| 378 |
+
"eval_loss": 2.8580081462860107,
|
| 379 |
+
"eval_runtime": 8.1811,
|
| 380 |
+
"eval_samples_per_second": 25.302,
|
| 381 |
+
"eval_steps_per_second": 3.178,
|
| 382 |
+
"eval_wer": 1.3900226757369614,
|
| 383 |
+
"step": 12500
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 21.1,
|
| 387 |
+
"learning_rate": 3.135011441647597e-05,
|
| 388 |
+
"loss": 2.9472,
|
| 389 |
+
"step": 13000
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 21.1,
|
| 393 |
+
"eval_loss": 2.839965343475342,
|
| 394 |
+
"eval_runtime": 8.2839,
|
| 395 |
+
"eval_samples_per_second": 24.988,
|
| 396 |
+
"eval_steps_per_second": 3.139,
|
| 397 |
+
"eval_wer": 1.3900226757369614,
|
| 398 |
+
"step": 13000
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 21.92,
|
| 402 |
+
"learning_rate": 2.8489702517162476e-05,
|
| 403 |
+
"loss": 3.0214,
|
| 404 |
+
"step": 13500
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 21.92,
|
| 408 |
+
"eval_loss": 2.856123924255371,
|
| 409 |
+
"eval_runtime": 8.1176,
|
| 410 |
+
"eval_samples_per_second": 25.5,
|
| 411 |
+
"eval_steps_per_second": 3.203,
|
| 412 |
+
"eval_wer": 1.3900226757369614,
|
| 413 |
+
"step": 13500
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 22.73,
|
| 417 |
+
"learning_rate": 2.562929061784897e-05,
|
| 418 |
+
"loss": 2.9336,
|
| 419 |
+
"step": 14000
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 22.73,
|
| 423 |
+
"eval_loss": 2.879206895828247,
|
| 424 |
+
"eval_runtime": 8.1391,
|
| 425 |
+
"eval_samples_per_second": 25.433,
|
| 426 |
+
"eval_steps_per_second": 3.194,
|
| 427 |
+
"eval_wer": 1.3900226757369614,
|
| 428 |
+
"step": 14000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 23.54,
|
| 432 |
+
"learning_rate": 2.276887871853547e-05,
|
| 433 |
+
"loss": 3.0134,
|
| 434 |
+
"step": 14500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 23.54,
|
| 438 |
+
"eval_loss": 2.8472487926483154,
|
| 439 |
+
"eval_runtime": 8.0887,
|
| 440 |
+
"eval_samples_per_second": 25.591,
|
| 441 |
+
"eval_steps_per_second": 3.214,
|
| 442 |
+
"eval_wer": 1.3900226757369614,
|
| 443 |
+
"step": 14500
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 24.35,
|
| 447 |
+
"learning_rate": 1.990846681922197e-05,
|
| 448 |
+
"loss": 2.9433,
|
| 449 |
+
"step": 15000
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 24.35,
|
| 453 |
+
"eval_loss": 2.8818936347961426,
|
| 454 |
+
"eval_runtime": 8.193,
|
| 455 |
+
"eval_samples_per_second": 25.266,
|
| 456 |
+
"eval_steps_per_second": 3.173,
|
| 457 |
+
"eval_wer": 1.3900226757369614,
|
| 458 |
+
"step": 15000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 25.16,
|
| 462 |
+
"learning_rate": 1.7048054919908468e-05,
|
| 463 |
+
"loss": 2.8536,
|
| 464 |
+
"step": 15500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 25.16,
|
| 468 |
+
"eval_loss": 2.837463140487671,
|
| 469 |
+
"eval_runtime": 8.1663,
|
| 470 |
+
"eval_samples_per_second": 25.348,
|
| 471 |
+
"eval_steps_per_second": 3.184,
|
| 472 |
+
"eval_wer": 1.3900226757369614,
|
| 473 |
+
"step": 15500
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 25.97,
|
| 477 |
+
"learning_rate": 1.4187643020594965e-05,
|
| 478 |
+
"loss": 2.8742,
|
| 479 |
+
"step": 16000
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 25.97,
|
| 483 |
+
"eval_loss": 2.857389450073242,
|
| 484 |
+
"eval_runtime": 8.194,
|
| 485 |
+
"eval_samples_per_second": 25.262,
|
| 486 |
+
"eval_steps_per_second": 3.173,
|
| 487 |
+
"eval_wer": 1.3900226757369614,
|
| 488 |
+
"step": 16000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 26.79,
|
| 492 |
+
"learning_rate": 1.1327231121281464e-05,
|
| 493 |
+
"loss": 2.8298,
|
| 494 |
+
"step": 16500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 26.79,
|
| 498 |
+
"eval_loss": 2.982081651687622,
|
| 499 |
+
"eval_runtime": 8.1935,
|
| 500 |
+
"eval_samples_per_second": 25.264,
|
| 501 |
+
"eval_steps_per_second": 3.173,
|
| 502 |
+
"eval_wer": 1.3900226757369614,
|
| 503 |
+
"step": 16500
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 27.6,
|
| 507 |
+
"learning_rate": 8.466819221967964e-06,
|
| 508 |
+
"loss": 2.7439,
|
| 509 |
+
"step": 17000
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 27.6,
|
| 513 |
+
"eval_loss": 3.2741587162017822,
|
| 514 |
+
"eval_runtime": 8.2267,
|
| 515 |
+
"eval_samples_per_second": 25.162,
|
| 516 |
+
"eval_steps_per_second": 3.16,
|
| 517 |
+
"eval_wer": 1.3900226757369614,
|
| 518 |
+
"step": 17000
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 28.41,
|
| 522 |
+
"learning_rate": 5.606407322654463e-06,
|
| 523 |
+
"loss": 2.7008,
|
| 524 |
+
"step": 17500
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 28.41,
|
| 528 |
+
"eval_loss": 3.365966320037842,
|
| 529 |
+
"eval_runtime": 8.203,
|
| 530 |
+
"eval_samples_per_second": 25.235,
|
| 531 |
+
"eval_steps_per_second": 3.17,
|
| 532 |
+
"eval_wer": 1.3900226757369614,
|
| 533 |
+
"step": 17500
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 29.22,
|
| 537 |
+
"learning_rate": 2.745995423340961e-06,
|
| 538 |
+
"loss": 2.7087,
|
| 539 |
+
"step": 18000
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 29.22,
|
| 543 |
+
"eval_loss": 3.812533140182495,
|
| 544 |
+
"eval_runtime": 8.2382,
|
| 545 |
+
"eval_samples_per_second": 25.127,
|
| 546 |
+
"eval_steps_per_second": 3.156,
|
| 547 |
+
"eval_wer": 1.3900226757369614,
|
| 548 |
+
"step": 18000
|
| 549 |
+
}
|
| 550 |
+
],
|
| 551 |
+
"max_steps": 18480,
|
| 552 |
+
"num_train_epochs": 30,
|
| 553 |
+
"total_flos": 4.093361821503758e+18,
|
| 554 |
+
"trial_name": null,
|
| 555 |
+
"trial_params": null
|
| 556 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:151803200d49964e6d9f1364dede5203d89e69881055e526220d7cecd61ffdfb
|
| 3 |
+
size 3375
|