End of training
Browse files- README.md +81 -0
- config.json +76 -0
- model.safetensors +3 -0
- preprocessor_config.json +10 -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/hubert-large-ls960-ft
|
| 4 |
+
tags:
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
metrics:
|
| 7 |
+
- wer
|
| 8 |
+
model-index:
|
| 9 |
+
- name: hubert-large-ls960-ft-V2-50
|
| 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 |
+
# hubert-large-ls960-ft-V2-50
|
| 17 |
+
|
| 18 |
+
This model is a fine-tuned version of [facebook/hubert-large-ls960-ft](https://huggingface.co/facebook/hubert-large-ls960-ft) on an unknown dataset.
|
| 19 |
+
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 0.9222
|
| 21 |
+
- Wer: 0.0732
|
| 22 |
+
- Per: 0.0540
|
| 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 |
+
| 12.8849 | 1.0 | 818 | 4.5925 | 0.9551 | 0.9649 |
|
| 55 |
+
| 2.7511 | 2.0 | 1636 | 1.7073 | 0.4693 | 0.4560 |
|
| 56 |
+
| 1.1653 | 3.0 | 2454 | 1.1204 | 0.1534 | 0.1317 |
|
| 57 |
+
| 0.7529 | 4.0 | 3272 | 1.0336 | 0.1055 | 0.0841 |
|
| 58 |
+
| 0.6309 | 5.0 | 4090 | 1.0015 | 0.1023 | 0.0817 |
|
| 59 |
+
| 0.5354 | 6.0 | 4908 | 1.0387 | 0.0992 | 0.0777 |
|
| 60 |
+
| 0.4907 | 7.0 | 5726 | 0.9957 | 0.1087 | 0.0893 |
|
| 61 |
+
| 0.4326 | 8.0 | 6544 | 0.8882 | 0.1091 | 0.0844 |
|
| 62 |
+
| 0.4148 | 9.0 | 7362 | 0.9542 | 0.0830 | 0.0638 |
|
| 63 |
+
| 0.3779 | 10.0 | 8180 | 0.9479 | 0.0690 | 0.0501 |
|
| 64 |
+
| 0.3502 | 11.0 | 8998 | 0.9840 | 0.0689 | 0.0491 |
|
| 65 |
+
| 0.3294 | 12.0 | 9816 | 1.0877 | 0.0694 | 0.0491 |
|
| 66 |
+
| 0.3239 | 13.0 | 10634 | 0.8955 | 0.0731 | 0.0534 |
|
| 67 |
+
| 0.3069 | 14.0 | 11452 | 0.8547 | 0.0776 | 0.0580 |
|
| 68 |
+
| 0.2689 | 15.0 | 12270 | 0.9683 | 0.0720 | 0.0525 |
|
| 69 |
+
| 0.2486 | 16.0 | 13088 | 0.9282 | 0.0704 | 0.0519 |
|
| 70 |
+
| 0.2291 | 17.0 | 13906 | 0.9004 | 0.0671 | 0.0481 |
|
| 71 |
+
| 0.2294 | 18.0 | 14724 | 0.9242 | 0.0747 | 0.0547 |
|
| 72 |
+
| 0.2151 | 19.0 | 15542 | 0.9400 | 0.0747 | 0.0554 |
|
| 73 |
+
| 0.2109 | 20.0 | 16360 | 0.9222 | 0.0732 | 0.0540 |
|
| 74 |
+
|
| 75 |
+
|
| 76 |
+
### Framework versions
|
| 77 |
+
|
| 78 |
+
- Transformers 4.35.2
|
| 79 |
+
- Pytorch 2.1.0+cu121
|
| 80 |
+
- Datasets 2.16.1
|
| 81 |
+
- Tokenizers 0.15.0
|
config.json
ADDED
|
@@ -0,0 +1,76 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "facebook/hubert-large-ls960-ft",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"apply_spec_augment": true,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"HubertForCTC"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 1,
|
| 10 |
+
"classifier_proj_size": 256,
|
| 11 |
+
"conv_bias": true,
|
| 12 |
+
"conv_dim": [
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512
|
| 20 |
+
],
|
| 21 |
+
"conv_kernel": [
|
| 22 |
+
10,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
3,
|
| 27 |
+
2,
|
| 28 |
+
2
|
| 29 |
+
],
|
| 30 |
+
"conv_stride": [
|
| 31 |
+
5,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2,
|
| 37 |
+
2
|
| 38 |
+
],
|
| 39 |
+
"ctc_loss_reduction": "mean",
|
| 40 |
+
"ctc_zero_infinity": false,
|
| 41 |
+
"diversity_loss_weight": 0.1,
|
| 42 |
+
"do_stable_layer_norm": true,
|
| 43 |
+
"eos_token_id": 2,
|
| 44 |
+
"feat_extract_activation": "gelu",
|
| 45 |
+
"feat_extract_dropout": 0.0,
|
| 46 |
+
"feat_extract_norm": "layer",
|
| 47 |
+
"feat_proj_dropout": 0.1,
|
| 48 |
+
"feat_proj_layer_norm": true,
|
| 49 |
+
"final_dropout": 0.1,
|
| 50 |
+
"gradient_checkpointing": false,
|
| 51 |
+
"hidden_act": "gelu",
|
| 52 |
+
"hidden_dropout": 0.1,
|
| 53 |
+
"hidden_dropout_prob": 0.1,
|
| 54 |
+
"hidden_size": 1024,
|
| 55 |
+
"initializer_range": 0.02,
|
| 56 |
+
"intermediate_size": 4096,
|
| 57 |
+
"layer_norm_eps": 1e-05,
|
| 58 |
+
"layerdrop": 0.1,
|
| 59 |
+
"mask_feature_length": 10,
|
| 60 |
+
"mask_feature_min_masks": 0,
|
| 61 |
+
"mask_feature_prob": 0.0,
|
| 62 |
+
"mask_time_length": 10,
|
| 63 |
+
"mask_time_min_masks": 2,
|
| 64 |
+
"mask_time_prob": 0.05,
|
| 65 |
+
"model_type": "hubert",
|
| 66 |
+
"num_attention_heads": 16,
|
| 67 |
+
"num_conv_pos_embedding_groups": 16,
|
| 68 |
+
"num_conv_pos_embeddings": 128,
|
| 69 |
+
"num_feat_extract_layers": 7,
|
| 70 |
+
"num_hidden_layers": 24,
|
| 71 |
+
"pad_token_id": 37,
|
| 72 |
+
"torch_dtype": "float32",
|
| 73 |
+
"transformers_version": "4.35.2",
|
| 74 |
+
"use_weighted_layer_sum": false,
|
| 75 |
+
"vocab_size": 40
|
| 76 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:31ac49565ea02b3ed0b0f9560d188d0531abac9a05ecca4462eca9664026db41
|
| 3 |
+
size 1261970648
|
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 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,348 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.854691743850708,
|
| 3 |
+
"best_model_checkpoint": "nrshoudi/hubert-large-ls960-ft-V2-50/checkpoint-11452",
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 16360,
|
| 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": 12.8849,
|
| 15 |
+
"step": 818
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 1.0,
|
| 19 |
+
"eval_loss": 4.592483043670654,
|
| 20 |
+
"eval_per": 0.9649250918684632,
|
| 21 |
+
"eval_runtime": 118.1886,
|
| 22 |
+
"eval_samples_per_second": 6.879,
|
| 23 |
+
"eval_steps_per_second": 3.444,
|
| 24 |
+
"eval_wer": 0.9551117540012132,
|
| 25 |
+
"step": 818
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 2.0,
|
| 29 |
+
"learning_rate": 0.0001,
|
| 30 |
+
"loss": 2.7511,
|
| 31 |
+
"step": 1636
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 2.0,
|
| 35 |
+
"eval_loss": 1.7072813510894775,
|
| 36 |
+
"eval_per": 0.4559738057099783,
|
| 37 |
+
"eval_runtime": 117.681,
|
| 38 |
+
"eval_samples_per_second": 6.909,
|
| 39 |
+
"eval_steps_per_second": 3.459,
|
| 40 |
+
"eval_wer": 0.4692734823386683,
|
| 41 |
+
"step": 1636
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 3.0,
|
| 45 |
+
"learning_rate": 9.444444444444444e-05,
|
| 46 |
+
"loss": 1.1653,
|
| 47 |
+
"step": 2454
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 3.0,
|
| 51 |
+
"eval_loss": 1.1204495429992676,
|
| 52 |
+
"eval_per": 0.1317016866107604,
|
| 53 |
+
"eval_runtime": 118.0673,
|
| 54 |
+
"eval_samples_per_second": 6.886,
|
| 55 |
+
"eval_steps_per_second": 3.447,
|
| 56 |
+
"eval_wer": 0.15342261210396155,
|
| 57 |
+
"step": 2454
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 4.0,
|
| 61 |
+
"learning_rate": 8.888888888888889e-05,
|
| 62 |
+
"loss": 0.7529,
|
| 63 |
+
"step": 3272
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 4.0,
|
| 67 |
+
"eval_loss": 1.033559799194336,
|
| 68 |
+
"eval_per": 0.08411853387355131,
|
| 69 |
+
"eval_runtime": 118.2209,
|
| 70 |
+
"eval_samples_per_second": 6.877,
|
| 71 |
+
"eval_steps_per_second": 3.443,
|
| 72 |
+
"eval_wer": 0.10554803788903924,
|
| 73 |
+
"step": 3272
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 5.0,
|
| 77 |
+
"learning_rate": 8.333333333333334e-05,
|
| 78 |
+
"loss": 0.6309,
|
| 79 |
+
"step": 4090
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 5.0,
|
| 83 |
+
"eval_loss": 1.0015147924423218,
|
| 84 |
+
"eval_per": 0.08166870818807123,
|
| 85 |
+
"eval_runtime": 118.3158,
|
| 86 |
+
"eval_samples_per_second": 6.871,
|
| 87 |
+
"eval_steps_per_second": 3.44,
|
| 88 |
+
"eval_wer": 0.10228174140264103,
|
| 89 |
+
"step": 4090
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 6.0,
|
| 93 |
+
"learning_rate": 7.777777777777778e-05,
|
| 94 |
+
"loss": 0.5354,
|
| 95 |
+
"step": 4908
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 6.0,
|
| 99 |
+
"eval_loss": 1.0387498140335083,
|
| 100 |
+
"eval_per": 0.07768774144916611,
|
| 101 |
+
"eval_runtime": 118.2614,
|
| 102 |
+
"eval_samples_per_second": 6.875,
|
| 103 |
+
"eval_steps_per_second": 3.442,
|
| 104 |
+
"eval_wer": 0.09924875180812841,
|
| 105 |
+
"step": 4908
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 7.0,
|
| 109 |
+
"learning_rate": 7.222222222222222e-05,
|
| 110 |
+
"loss": 0.4907,
|
| 111 |
+
"step": 5726
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 7.0,
|
| 115 |
+
"eval_loss": 0.9956713318824768,
|
| 116 |
+
"eval_per": 0.08925374540657684,
|
| 117 |
+
"eval_runtime": 117.9222,
|
| 118 |
+
"eval_samples_per_second": 6.894,
|
| 119 |
+
"eval_steps_per_second": 3.451,
|
| 120 |
+
"eval_wer": 0.10872101161868322,
|
| 121 |
+
"step": 5726
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 8.0,
|
| 125 |
+
"learning_rate": 6.666666666666667e-05,
|
| 126 |
+
"loss": 0.4326,
|
| 127 |
+
"step": 6544
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 8.0,
|
| 131 |
+
"eval_loss": 0.8882040977478027,
|
| 132 |
+
"eval_per": 0.08440120606802977,
|
| 133 |
+
"eval_runtime": 118.2986,
|
| 134 |
+
"eval_samples_per_second": 6.872,
|
| 135 |
+
"eval_steps_per_second": 3.44,
|
| 136 |
+
"eval_wer": 0.10914096402407727,
|
| 137 |
+
"step": 6544
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 9.0,
|
| 141 |
+
"learning_rate": 6.111111111111112e-05,
|
| 142 |
+
"loss": 0.4148,
|
| 143 |
+
"step": 7362
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 9.0,
|
| 147 |
+
"eval_loss": 0.9541579484939575,
|
| 148 |
+
"eval_per": 0.06376613587110148,
|
| 149 |
+
"eval_runtime": 118.3333,
|
| 150 |
+
"eval_samples_per_second": 6.87,
|
| 151 |
+
"eval_steps_per_second": 3.439,
|
| 152 |
+
"eval_wer": 0.08301059213289161,
|
| 153 |
+
"step": 7362
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 10.0,
|
| 157 |
+
"learning_rate": 5.555555555555556e-05,
|
| 158 |
+
"loss": 0.3779,
|
| 159 |
+
"step": 8180
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 10.0,
|
| 163 |
+
"eval_loss": 0.9479207396507263,
|
| 164 |
+
"eval_per": 0.05010364647130877,
|
| 165 |
+
"eval_runtime": 118.3528,
|
| 166 |
+
"eval_samples_per_second": 6.869,
|
| 167 |
+
"eval_steps_per_second": 3.439,
|
| 168 |
+
"eval_wer": 0.06901217861975643,
|
| 169 |
+
"step": 8180
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 11.0,
|
| 173 |
+
"learning_rate": 5e-05,
|
| 174 |
+
"loss": 0.3502,
|
| 175 |
+
"step": 8998
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 11.0,
|
| 179 |
+
"eval_loss": 0.9840469360351562,
|
| 180 |
+
"eval_per": 0.04906718175822105,
|
| 181 |
+
"eval_runtime": 118.5969,
|
| 182 |
+
"eval_samples_per_second": 6.855,
|
| 183 |
+
"eval_steps_per_second": 3.432,
|
| 184 |
+
"eval_wer": 0.06887219448462507,
|
| 185 |
+
"step": 8998
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 12.0,
|
| 189 |
+
"learning_rate": 4.4444444444444447e-05,
|
| 190 |
+
"loss": 0.3294,
|
| 191 |
+
"step": 9816
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 12.0,
|
| 195 |
+
"eval_loss": 1.087670922279358,
|
| 196 |
+
"eval_per": 0.04906718175822105,
|
| 197 |
+
"eval_runtime": 118.227,
|
| 198 |
+
"eval_samples_per_second": 6.877,
|
| 199 |
+
"eval_steps_per_second": 3.443,
|
| 200 |
+
"eval_wer": 0.06943213102515049,
|
| 201 |
+
"step": 9816
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 13.0,
|
| 205 |
+
"learning_rate": 3.888888888888889e-05,
|
| 206 |
+
"loss": 0.3239,
|
| 207 |
+
"step": 10634
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 13.0,
|
| 211 |
+
"eval_loss": 0.8955483436584473,
|
| 212 |
+
"eval_per": 0.05335437670781117,
|
| 213 |
+
"eval_runtime": 118.657,
|
| 214 |
+
"eval_samples_per_second": 6.852,
|
| 215 |
+
"eval_steps_per_second": 3.43,
|
| 216 |
+
"eval_wer": 0.07311837991694274,
|
| 217 |
+
"step": 10634
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 14.0,
|
| 221 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 222 |
+
"loss": 0.3069,
|
| 223 |
+
"step": 11452
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 14.0,
|
| 227 |
+
"eval_loss": 0.854691743850708,
|
| 228 |
+
"eval_per": 0.05804202393291247,
|
| 229 |
+
"eval_runtime": 119.0248,
|
| 230 |
+
"eval_samples_per_second": 6.831,
|
| 231 |
+
"eval_steps_per_second": 3.419,
|
| 232 |
+
"eval_wer": 0.077597872241146,
|
| 233 |
+
"step": 11452
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 15.0,
|
| 237 |
+
"learning_rate": 2.777777777777778e-05,
|
| 238 |
+
"loss": 0.2689,
|
| 239 |
+
"step": 12270
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 15.0,
|
| 243 |
+
"eval_loss": 0.968323290348053,
|
| 244 |
+
"eval_per": 0.05245924809196269,
|
| 245 |
+
"eval_runtime": 118.5754,
|
| 246 |
+
"eval_samples_per_second": 6.856,
|
| 247 |
+
"eval_steps_per_second": 3.432,
|
| 248 |
+
"eval_wer": 0.07199850683589193,
|
| 249 |
+
"step": 12270
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 16.0,
|
| 253 |
+
"learning_rate": 2.2222222222222223e-05,
|
| 254 |
+
"loss": 0.2486,
|
| 255 |
+
"step": 13088
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 16.0,
|
| 259 |
+
"eval_loss": 0.9282185435295105,
|
| 260 |
+
"eval_per": 0.05189390370300575,
|
| 261 |
+
"eval_runtime": 118.5748,
|
| 262 |
+
"eval_samples_per_second": 6.856,
|
| 263 |
+
"eval_steps_per_second": 3.432,
|
| 264 |
+
"eval_wer": 0.07041201997106994,
|
| 265 |
+
"step": 13088
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 17.0,
|
| 269 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 270 |
+
"loss": 0.2291,
|
| 271 |
+
"step": 13906
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 17.0,
|
| 275 |
+
"eval_loss": 0.900354266166687,
|
| 276 |
+
"eval_per": 0.048101385093752946,
|
| 277 |
+
"eval_runtime": 118.0532,
|
| 278 |
+
"eval_samples_per_second": 6.887,
|
| 279 |
+
"eval_steps_per_second": 3.448,
|
| 280 |
+
"eval_wer": 0.06714572348467174,
|
| 281 |
+
"step": 13906
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 18.0,
|
| 285 |
+
"learning_rate": 1.1111111111111112e-05,
|
| 286 |
+
"loss": 0.2294,
|
| 287 |
+
"step": 14724
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 18.0,
|
| 291 |
+
"eval_loss": 0.9242432117462158,
|
| 292 |
+
"eval_per": 0.05472062564779045,
|
| 293 |
+
"eval_runtime": 118.8448,
|
| 294 |
+
"eval_samples_per_second": 6.841,
|
| 295 |
+
"eval_steps_per_second": 3.425,
|
| 296 |
+
"eval_wer": 0.07470486678176473,
|
| 297 |
+
"step": 14724
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 19.0,
|
| 301 |
+
"learning_rate": 5.555555555555556e-06,
|
| 302 |
+
"loss": 0.2151,
|
| 303 |
+
"step": 15542
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 19.0,
|
| 307 |
+
"eval_loss": 0.9400033354759216,
|
| 308 |
+
"eval_per": 0.05540375011778008,
|
| 309 |
+
"eval_runtime": 118.591,
|
| 310 |
+
"eval_samples_per_second": 6.855,
|
| 311 |
+
"eval_steps_per_second": 3.432,
|
| 312 |
+
"eval_wer": 0.07470486678176473,
|
| 313 |
+
"step": 15542
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 20.0,
|
| 317 |
+
"learning_rate": 0.0,
|
| 318 |
+
"loss": 0.2109,
|
| 319 |
+
"step": 16360
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 20.0,
|
| 323 |
+
"eval_loss": 0.9222491979598999,
|
| 324 |
+
"eval_per": 0.05396683312918119,
|
| 325 |
+
"eval_runtime": 119.3623,
|
| 326 |
+
"eval_samples_per_second": 6.811,
|
| 327 |
+
"eval_steps_per_second": 3.41,
|
| 328 |
+
"eval_wer": 0.07321170267369698,
|
| 329 |
+
"step": 16360
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 20.0,
|
| 333 |
+
"step": 16360,
|
| 334 |
+
"total_flos": 5.249198170404504e+18,
|
| 335 |
+
"train_loss": 1.15744112849527,
|
| 336 |
+
"train_runtime": 11016.1558,
|
| 337 |
+
"train_samples_per_second": 2.97,
|
| 338 |
+
"train_steps_per_second": 1.485
|
| 339 |
+
}
|
| 340 |
+
],
|
| 341 |
+
"logging_steps": 500,
|
| 342 |
+
"max_steps": 16360,
|
| 343 |
+
"num_train_epochs": 20,
|
| 344 |
+
"save_steps": 500,
|
| 345 |
+
"total_flos": 5.249198170404504e+18,
|
| 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 sha256:702228b134a59e0f61047ef176b39e03ee953452fa04c1dab1c68c95827243a6
|
| 3 |
+
size 4600
|