End of training
Browse files- README.md +81 -0
- config.json +109 -0
- preprocessor_config.json +10 -0
- pytorch_model.bin +3 -0
- trainer_state.json +348 -0
- training_args.bin +3 -0
README.md
ADDED
|
@@ -0,0 +1,81 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: facebook/wav2vec2-xls-r-300m
|
| 4 |
+
tags:
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
metrics:
|
| 7 |
+
- wer
|
| 8 |
+
model-index:
|
| 9 |
+
- name: wav2vec_arabic_mdd
|
| 10 |
+
results: []
|
| 11 |
+
---
|
| 12 |
+
|
| 13 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 14 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 15 |
+
|
| 16 |
+
# wav2vec_arabic_mdd
|
| 17 |
+
|
| 18 |
+
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on an unknown dataset.
|
| 19 |
+
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 0.3264
|
| 21 |
+
- Wer: 0.0417
|
| 22 |
+
- Per: 0.0321
|
| 23 |
+
|
| 24 |
+
## Model description
|
| 25 |
+
|
| 26 |
+
More information needed
|
| 27 |
+
|
| 28 |
+
## Intended uses & limitations
|
| 29 |
+
|
| 30 |
+
More information needed
|
| 31 |
+
|
| 32 |
+
## Training and evaluation data
|
| 33 |
+
|
| 34 |
+
More information needed
|
| 35 |
+
|
| 36 |
+
## Training procedure
|
| 37 |
+
|
| 38 |
+
### Training hyperparameters
|
| 39 |
+
|
| 40 |
+
The following hyperparameters were used during training:
|
| 41 |
+
- learning_rate: 0.0001
|
| 42 |
+
- train_batch_size: 2
|
| 43 |
+
- eval_batch_size: 2
|
| 44 |
+
- seed: 42
|
| 45 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 46 |
+
- lr_scheduler_type: linear
|
| 47 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 48 |
+
- num_epochs: 20
|
| 49 |
+
|
| 50 |
+
### Training results
|
| 51 |
+
|
| 52 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer | Per |
|
| 53 |
+
|:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|
|
| 54 |
+
| 8.2138 | 1.0 | 1617 | 3.1149 | 1.0 | 1.0 |
|
| 55 |
+
| 1.1507 | 2.0 | 3234 | 0.3763 | 0.0733 | 0.0570 |
|
| 56 |
+
| 0.2 | 3.0 | 4851 | 0.3199 | 0.0634 | 0.0493 |
|
| 57 |
+
| 0.1328 | 4.0 | 6468 | 0.3447 | 0.0530 | 0.0399 |
|
| 58 |
+
| 0.1045 | 5.0 | 8085 | 0.3005 | 0.0514 | 0.0401 |
|
| 59 |
+
| 0.088 | 6.0 | 9702 | 0.3276 | 0.0517 | 0.0391 |
|
| 60 |
+
| 0.0756 | 7.0 | 11319 | 0.3367 | 0.0512 | 0.0388 |
|
| 61 |
+
| 0.062 | 8.0 | 12936 | 0.3933 | 0.0572 | 0.0422 |
|
| 62 |
+
| 0.0603 | 9.0 | 14553 | 0.3331 | 0.0475 | 0.0359 |
|
| 63 |
+
| 0.0517 | 10.0 | 16170 | 0.3014 | 0.0500 | 0.0372 |
|
| 64 |
+
| 0.046 | 11.0 | 17787 | 0.3067 | 0.0477 | 0.0360 |
|
| 65 |
+
| 0.0398 | 12.0 | 19404 | 0.2990 | 0.0456 | 0.0348 |
|
| 66 |
+
| 0.0329 | 13.0 | 21021 | 0.3064 | 0.0450 | 0.0342 |
|
| 67 |
+
| 0.0319 | 14.0 | 22638 | 0.3336 | 0.0499 | 0.0383 |
|
| 68 |
+
| 0.0288 | 15.0 | 24255 | 0.3222 | 0.0451 | 0.0358 |
|
| 69 |
+
| 0.0258 | 16.0 | 25872 | 0.3302 | 0.0428 | 0.0324 |
|
| 70 |
+
| 0.0181 | 17.0 | 27489 | 0.3408 | 0.0433 | 0.0334 |
|
| 71 |
+
| 0.0172 | 18.0 | 29106 | 0.3216 | 0.0430 | 0.0331 |
|
| 72 |
+
| 0.0145 | 19.0 | 30723 | 0.3254 | 0.0430 | 0.0329 |
|
| 73 |
+
| 0.0127 | 20.0 | 32340 | 0.3264 | 0.0417 | 0.0321 |
|
| 74 |
+
|
| 75 |
+
|
| 76 |
+
### Framework versions
|
| 77 |
+
|
| 78 |
+
- Transformers 4.34.1
|
| 79 |
+
- Pytorch 2.1.0+cu118
|
| 80 |
+
- Datasets 2.14.6
|
| 81 |
+
- Tokenizers 0.14.1
|
config.json
ADDED
|
@@ -0,0 +1,109 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "facebook/wav2vec2-xls-r-300m",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"adapter_attn_dim": null,
|
| 5 |
+
"adapter_kernel_size": 3,
|
| 6 |
+
"adapter_stride": 2,
|
| 7 |
+
"add_adapter": false,
|
| 8 |
+
"apply_spec_augment": true,
|
| 9 |
+
"architectures": [
|
| 10 |
+
"Wav2Vec2ForCTC"
|
| 11 |
+
],
|
| 12 |
+
"attention_dropout": 0.1,
|
| 13 |
+
"bos_token_id": 1,
|
| 14 |
+
"classifier_proj_size": 256,
|
| 15 |
+
"codevector_dim": 768,
|
| 16 |
+
"contrastive_logits_temperature": 0.1,
|
| 17 |
+
"conv_bias": true,
|
| 18 |
+
"conv_dim": [
|
| 19 |
+
512,
|
| 20 |
+
512,
|
| 21 |
+
512,
|
| 22 |
+
512,
|
| 23 |
+
512,
|
| 24 |
+
512,
|
| 25 |
+
512
|
| 26 |
+
],
|
| 27 |
+
"conv_kernel": [
|
| 28 |
+
10,
|
| 29 |
+
3,
|
| 30 |
+
3,
|
| 31 |
+
3,
|
| 32 |
+
3,
|
| 33 |
+
2,
|
| 34 |
+
2
|
| 35 |
+
],
|
| 36 |
+
"conv_stride": [
|
| 37 |
+
5,
|
| 38 |
+
2,
|
| 39 |
+
2,
|
| 40 |
+
2,
|
| 41 |
+
2,
|
| 42 |
+
2,
|
| 43 |
+
2
|
| 44 |
+
],
|
| 45 |
+
"ctc_loss_reduction": "mean",
|
| 46 |
+
"ctc_zero_infinity": false,
|
| 47 |
+
"diversity_loss_weight": 0.1,
|
| 48 |
+
"do_stable_layer_norm": true,
|
| 49 |
+
"eos_token_id": 2,
|
| 50 |
+
"feat_extract_activation": "gelu",
|
| 51 |
+
"feat_extract_dropout": 0.0,
|
| 52 |
+
"feat_extract_norm": "layer",
|
| 53 |
+
"feat_proj_dropout": 0.1,
|
| 54 |
+
"feat_quantizer_dropout": 0.0,
|
| 55 |
+
"final_dropout": 0.0,
|
| 56 |
+
"gradient_checkpointing": false,
|
| 57 |
+
"hidden_act": "gelu",
|
| 58 |
+
"hidden_dropout": 0.1,
|
| 59 |
+
"hidden_size": 1024,
|
| 60 |
+
"initializer_range": 0.02,
|
| 61 |
+
"intermediate_size": 4096,
|
| 62 |
+
"layer_norm_eps": 1e-05,
|
| 63 |
+
"layerdrop": 0.1,
|
| 64 |
+
"mask_feature_length": 10,
|
| 65 |
+
"mask_feature_min_masks": 0,
|
| 66 |
+
"mask_feature_prob": 0.0,
|
| 67 |
+
"mask_time_length": 10,
|
| 68 |
+
"mask_time_min_masks": 2,
|
| 69 |
+
"mask_time_prob": 0.075,
|
| 70 |
+
"model_type": "wav2vec2",
|
| 71 |
+
"num_adapter_layers": 3,
|
| 72 |
+
"num_attention_heads": 16,
|
| 73 |
+
"num_codevector_groups": 2,
|
| 74 |
+
"num_codevectors_per_group": 320,
|
| 75 |
+
"num_conv_pos_embedding_groups": 16,
|
| 76 |
+
"num_conv_pos_embeddings": 128,
|
| 77 |
+
"num_feat_extract_layers": 7,
|
| 78 |
+
"num_hidden_layers": 24,
|
| 79 |
+
"num_negatives": 100,
|
| 80 |
+
"output_hidden_size": 1024,
|
| 81 |
+
"pad_token_id": 37,
|
| 82 |
+
"proj_codevector_dim": 768,
|
| 83 |
+
"tdnn_dilation": [
|
| 84 |
+
1,
|
| 85 |
+
2,
|
| 86 |
+
3,
|
| 87 |
+
1,
|
| 88 |
+
1
|
| 89 |
+
],
|
| 90 |
+
"tdnn_dim": [
|
| 91 |
+
512,
|
| 92 |
+
512,
|
| 93 |
+
512,
|
| 94 |
+
512,
|
| 95 |
+
1500
|
| 96 |
+
],
|
| 97 |
+
"tdnn_kernel": [
|
| 98 |
+
5,
|
| 99 |
+
3,
|
| 100 |
+
3,
|
| 101 |
+
1,
|
| 102 |
+
1
|
| 103 |
+
],
|
| 104 |
+
"torch_dtype": "float32",
|
| 105 |
+
"transformers_version": "4.34.1",
|
| 106 |
+
"use_weighted_layer_sum": false,
|
| 107 |
+
"vocab_size": 40,
|
| 108 |
+
"xvector_output_dim": 512
|
| 109 |
+
}
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_normalize": true,
|
| 3 |
+
"feature_extractor_type": "Wav2Vec2FeatureExtractor",
|
| 4 |
+
"feature_size": 1,
|
| 5 |
+
"padding_side": "right",
|
| 6 |
+
"padding_value": 0,
|
| 7 |
+
"processor_class": "Wav2Vec2Processor",
|
| 8 |
+
"return_attention_mask": true,
|
| 9 |
+
"sampling_rate": 16000
|
| 10 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:29fe62faf88173276e1072187f5f576f7c13533622f9219344bfbaa4171491a9
|
| 3 |
+
size 1262066346
|
trainer_state.json
ADDED
|
@@ -0,0 +1,348 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.29899266362190247,
|
| 3 |
+
"best_model_checkpoint": "nrshoudi/wav2vec_arabic_mdd/checkpoint-19404",
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 32340,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 1.0,
|
| 13 |
+
"learning_rate": 5e-05,
|
| 14 |
+
"loss": 8.2138,
|
| 15 |
+
"step": 1617
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 1.0,
|
| 19 |
+
"eval_loss": 3.1148788928985596,
|
| 20 |
+
"eval_per": 1.0,
|
| 21 |
+
"eval_runtime": 115.2576,
|
| 22 |
+
"eval_samples_per_second": 7.054,
|
| 23 |
+
"eval_steps_per_second": 3.531,
|
| 24 |
+
"eval_wer": 1.0,
|
| 25 |
+
"step": 1617
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 2.0,
|
| 29 |
+
"learning_rate": 0.0001,
|
| 30 |
+
"loss": 1.1507,
|
| 31 |
+
"step": 3234
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 2.0,
|
| 35 |
+
"eval_loss": 0.37626245617866516,
|
| 36 |
+
"eval_per": 0.0569820032036182,
|
| 37 |
+
"eval_runtime": 116.2275,
|
| 38 |
+
"eval_samples_per_second": 6.995,
|
| 39 |
+
"eval_steps_per_second": 3.502,
|
| 40 |
+
"eval_wer": 0.0732583640520741,
|
| 41 |
+
"step": 3234
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 3.0,
|
| 45 |
+
"learning_rate": 9.444444444444444e-05,
|
| 46 |
+
"loss": 0.2,
|
| 47 |
+
"step": 4851
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 3.0,
|
| 51 |
+
"eval_loss": 0.319857656955719,
|
| 52 |
+
"eval_per": 0.04925562988787336,
|
| 53 |
+
"eval_runtime": 115.4027,
|
| 54 |
+
"eval_samples_per_second": 7.045,
|
| 55 |
+
"eval_steps_per_second": 3.527,
|
| 56 |
+
"eval_wer": 0.06341281321450236,
|
| 57 |
+
"step": 4851
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 4.0,
|
| 61 |
+
"learning_rate": 8.888888888888889e-05,
|
| 62 |
+
"loss": 0.1328,
|
| 63 |
+
"step": 6468
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 4.0,
|
| 67 |
+
"eval_loss": 0.34467777609825134,
|
| 68 |
+
"eval_per": 0.03985677942146424,
|
| 69 |
+
"eval_runtime": 115.3838,
|
| 70 |
+
"eval_samples_per_second": 7.046,
|
| 71 |
+
"eval_steps_per_second": 3.527,
|
| 72 |
+
"eval_wer": 0.05296066445802809,
|
| 73 |
+
"step": 6468
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 5.0,
|
| 77 |
+
"learning_rate": 8.333333333333334e-05,
|
| 78 |
+
"loss": 0.1045,
|
| 79 |
+
"step": 8085
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 5.0,
|
| 83 |
+
"eval_loss": 0.30046939849853516,
|
| 84 |
+
"eval_per": 0.04013945161594271,
|
| 85 |
+
"eval_runtime": 116.1796,
|
| 86 |
+
"eval_samples_per_second": 6.998,
|
| 87 |
+
"eval_steps_per_second": 3.503,
|
| 88 |
+
"eval_wer": 0.05142083897158322,
|
| 89 |
+
"step": 8085
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 6.0,
|
| 93 |
+
"learning_rate": 7.777777777777778e-05,
|
| 94 |
+
"loss": 0.088,
|
| 95 |
+
"step": 9702
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 6.0,
|
| 99 |
+
"eval_loss": 0.32756975293159485,
|
| 100 |
+
"eval_per": 0.03907943088664845,
|
| 101 |
+
"eval_runtime": 116.0861,
|
| 102 |
+
"eval_samples_per_second": 7.003,
|
| 103 |
+
"eval_steps_per_second": 3.506,
|
| 104 |
+
"eval_wer": 0.05165414586346881,
|
| 105 |
+
"step": 9702
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 7.0,
|
| 109 |
+
"learning_rate": 7.222222222222222e-05,
|
| 110 |
+
"loss": 0.0756,
|
| 111 |
+
"step": 11319
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 7.0,
|
| 115 |
+
"eval_loss": 0.33674493432044983,
|
| 116 |
+
"eval_per": 0.03879675869216998,
|
| 117 |
+
"eval_runtime": 115.7009,
|
| 118 |
+
"eval_samples_per_second": 7.027,
|
| 119 |
+
"eval_steps_per_second": 3.518,
|
| 120 |
+
"eval_wer": 0.05118753207969763,
|
| 121 |
+
"step": 11319
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 8.0,
|
| 125 |
+
"learning_rate": 6.666666666666667e-05,
|
| 126 |
+
"loss": 0.062,
|
| 127 |
+
"step": 12936
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 8.0,
|
| 131 |
+
"eval_loss": 0.3933280110359192,
|
| 132 |
+
"eval_per": 0.042165269009705075,
|
| 133 |
+
"eval_runtime": 115.0602,
|
| 134 |
+
"eval_samples_per_second": 7.066,
|
| 135 |
+
"eval_steps_per_second": 3.537,
|
| 136 |
+
"eval_wer": 0.05720684989034576,
|
| 137 |
+
"step": 12936
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 9.0,
|
| 141 |
+
"learning_rate": 6.111111111111112e-05,
|
| 142 |
+
"loss": 0.0603,
|
| 143 |
+
"step": 14553
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 9.0,
|
| 147 |
+
"eval_loss": 0.3331439197063446,
|
| 148 |
+
"eval_per": 0.03587581268255913,
|
| 149 |
+
"eval_runtime": 115.3081,
|
| 150 |
+
"eval_samples_per_second": 7.051,
|
| 151 |
+
"eval_steps_per_second": 3.53,
|
| 152 |
+
"eval_wer": 0.04754794456628249,
|
| 153 |
+
"step": 14553
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 10.0,
|
| 157 |
+
"learning_rate": 5.555555555555556e-05,
|
| 158 |
+
"loss": 0.0517,
|
| 159 |
+
"step": 16170
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 10.0,
|
| 163 |
+
"eval_loss": 0.301434725522995,
|
| 164 |
+
"eval_per": 0.03717139357391878,
|
| 165 |
+
"eval_runtime": 115.3305,
|
| 166 |
+
"eval_samples_per_second": 7.049,
|
| 167 |
+
"eval_steps_per_second": 3.529,
|
| 168 |
+
"eval_wer": 0.049974336241892583,
|
| 169 |
+
"step": 16170
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 11.0,
|
| 173 |
+
"learning_rate": 5e-05,
|
| 174 |
+
"loss": 0.046,
|
| 175 |
+
"step": 17787
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 11.0,
|
| 179 |
+
"eval_loss": 0.30668768286705017,
|
| 180 |
+
"eval_per": 0.036017148779798364,
|
| 181 |
+
"eval_runtime": 115.2426,
|
| 182 |
+
"eval_samples_per_second": 7.055,
|
| 183 |
+
"eval_steps_per_second": 3.532,
|
| 184 |
+
"eval_wer": 0.04768792870141384,
|
| 185 |
+
"step": 17787
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 12.0,
|
| 189 |
+
"learning_rate": 4.4444444444444447e-05,
|
| 190 |
+
"loss": 0.0398,
|
| 191 |
+
"step": 19404
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 12.0,
|
| 195 |
+
"eval_loss": 0.29899266362190247,
|
| 196 |
+
"eval_per": 0.03481579195326486,
|
| 197 |
+
"eval_runtime": 115.6042,
|
| 198 |
+
"eval_samples_per_second": 7.033,
|
| 199 |
+
"eval_steps_per_second": 3.521,
|
| 200 |
+
"eval_wer": 0.04558816667444356,
|
| 201 |
+
"step": 19404
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 13.0,
|
| 205 |
+
"learning_rate": 3.888888888888889e-05,
|
| 206 |
+
"loss": 0.0329,
|
| 207 |
+
"step": 21021
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 13.0,
|
| 211 |
+
"eval_loss": 0.3064207434654236,
|
| 212 |
+
"eval_per": 0.034226891548101386,
|
| 213 |
+
"eval_runtime": 116.0675,
|
| 214 |
+
"eval_samples_per_second": 7.005,
|
| 215 |
+
"eval_steps_per_second": 3.507,
|
| 216 |
+
"eval_wer": 0.04498156875554104,
|
| 217 |
+
"step": 21021
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 14.0,
|
| 221 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 222 |
+
"loss": 0.0319,
|
| 223 |
+
"step": 22638
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 14.0,
|
| 227 |
+
"eval_loss": 0.3336479663848877,
|
| 228 |
+
"eval_per": 0.03825497031941958,
|
| 229 |
+
"eval_runtime": 115.1932,
|
| 230 |
+
"eval_samples_per_second": 7.058,
|
| 231 |
+
"eval_steps_per_second": 3.533,
|
| 232 |
+
"eval_wer": 0.04988101348513835,
|
| 233 |
+
"step": 22638
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 15.0,
|
| 237 |
+
"learning_rate": 2.777777777777778e-05,
|
| 238 |
+
"loss": 0.0288,
|
| 239 |
+
"step": 24255
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 15.0,
|
| 243 |
+
"eval_loss": 0.32215777039527893,
|
| 244 |
+
"eval_per": 0.03575803260152643,
|
| 245 |
+
"eval_runtime": 115.0522,
|
| 246 |
+
"eval_samples_per_second": 7.066,
|
| 247 |
+
"eval_steps_per_second": 3.538,
|
| 248 |
+
"eval_wer": 0.04512155289067239,
|
| 249 |
+
"step": 24255
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 16.0,
|
| 253 |
+
"learning_rate": 2.2222222222222223e-05,
|
| 254 |
+
"loss": 0.0258,
|
| 255 |
+
"step": 25872
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 16.0,
|
| 259 |
+
"eval_loss": 0.33019357919692993,
|
| 260 |
+
"eval_per": 0.03241307830019787,
|
| 261 |
+
"eval_runtime": 115.355,
|
| 262 |
+
"eval_samples_per_second": 7.048,
|
| 263 |
+
"eval_steps_per_second": 3.528,
|
| 264 |
+
"eval_wer": 0.042788483971816525,
|
| 265 |
+
"step": 25872
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 17.0,
|
| 269 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 270 |
+
"loss": 0.0181,
|
| 271 |
+
"step": 27489
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 17.0,
|
| 275 |
+
"eval_loss": 0.3408127427101135,
|
| 276 |
+
"eval_per": 0.03335531894845944,
|
| 277 |
+
"eval_runtime": 115.3097,
|
| 278 |
+
"eval_samples_per_second": 7.051,
|
| 279 |
+
"eval_steps_per_second": 3.53,
|
| 280 |
+
"eval_wer": 0.0432550977555877,
|
| 281 |
+
"step": 27489
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 18.0,
|
| 285 |
+
"learning_rate": 1.1111111111111112e-05,
|
| 286 |
+
"loss": 0.0172,
|
| 287 |
+
"step": 29106
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 18.0,
|
| 291 |
+
"eval_loss": 0.3215967118740082,
|
| 292 |
+
"eval_per": 0.033119758786394046,
|
| 293 |
+
"eval_runtime": 115.4137,
|
| 294 |
+
"eval_samples_per_second": 7.044,
|
| 295 |
+
"eval_steps_per_second": 3.526,
|
| 296 |
+
"eval_wer": 0.043021790863702114,
|
| 297 |
+
"step": 29106
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 19.0,
|
| 301 |
+
"learning_rate": 5.555555555555556e-06,
|
| 302 |
+
"loss": 0.0145,
|
| 303 |
+
"step": 30723
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 19.0,
|
| 307 |
+
"eval_loss": 0.3254193663597107,
|
| 308 |
+
"eval_per": 0.03293131065674173,
|
| 309 |
+
"eval_runtime": 115.2743,
|
| 310 |
+
"eval_samples_per_second": 7.053,
|
| 311 |
+
"eval_steps_per_second": 3.531,
|
| 312 |
+
"eval_wer": 0.042975129485324996,
|
| 313 |
+
"step": 30723
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 20.0,
|
| 317 |
+
"learning_rate": 0.0,
|
| 318 |
+
"loss": 0.0127,
|
| 319 |
+
"step": 32340
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 20.0,
|
| 323 |
+
"eval_loss": 0.32637345790863037,
|
| 324 |
+
"eval_per": 0.0321304061057194,
|
| 325 |
+
"eval_runtime": 115.2538,
|
| 326 |
+
"eval_samples_per_second": 7.054,
|
| 327 |
+
"eval_steps_per_second": 3.531,
|
| 328 |
+
"eval_wer": 0.04166861089076571,
|
| 329 |
+
"step": 32340
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 20.0,
|
| 333 |
+
"step": 32340,
|
| 334 |
+
"total_flos": 1.0355998175044045e+19,
|
| 335 |
+
"train_loss": 0.520349072083029,
|
| 336 |
+
"train_runtime": 19017.9936,
|
| 337 |
+
"train_samples_per_second": 3.401,
|
| 338 |
+
"train_steps_per_second": 1.7
|
| 339 |
+
}
|
| 340 |
+
],
|
| 341 |
+
"logging_steps": 500,
|
| 342 |
+
"max_steps": 32340,
|
| 343 |
+
"num_train_epochs": 20,
|
| 344 |
+
"save_steps": 500,
|
| 345 |
+
"total_flos": 1.0355998175044045e+19,
|
| 346 |
+
"trial_name": null,
|
| 347 |
+
"trial_params": null
|
| 348 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:84b4a7ac0e1c6b093474175c4fd5e8c724dc3907c9fb9bd69ef08bd309794b0d
|
| 3 |
+
size 4536
|