Commit
·
e1c59af
1
Parent(s):
ad231eb
update model card README.md
Browse files
README.md
CHANGED
|
@@ -2,8 +2,6 @@
|
|
| 2 |
license: apache-2.0
|
| 3 |
tags:
|
| 4 |
- generated_from_trainer
|
| 5 |
-
metrics:
|
| 6 |
-
- wer
|
| 7 |
model-index:
|
| 8 |
- name: wav2vec2-large-asr-th
|
| 9 |
results: []
|
|
@@ -14,11 +12,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 14 |
|
| 15 |
# wav2vec2-large-asr-th
|
| 16 |
|
| 17 |
-
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on
|
| 18 |
-
It achieves the following results on the evaluation set:
|
| 19 |
-
- Loss: 0.8027
|
| 20 |
-
- Wer: 0.5915
|
| 21 |
-
- Cer: 0.9337
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
@@ -37,7 +31,7 @@ More information needed
|
|
| 37 |
### Training hyperparameters
|
| 38 |
|
| 39 |
The following hyperparameters were used during training:
|
| 40 |
-
- learning_rate: 0.
|
| 41 |
- train_batch_size: 16
|
| 42 |
- eval_batch_size: 8
|
| 43 |
- seed: 42
|
|
@@ -46,40 +40,9 @@ The following hyperparameters were used during training:
|
|
| 46 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 47 |
- lr_scheduler_type: linear
|
| 48 |
- lr_scheduler_warmup_steps: 200
|
| 49 |
-
- training_steps:
|
| 50 |
- mixed_precision_training: Native AMP
|
| 51 |
|
| 52 |
-
### Training results
|
| 53 |
-
|
| 54 |
-
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|
| 55 |
-
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|
|
| 56 |
-
| 3.7792 | 2.2 | 200 | 3.8202 | 1.0 | 1.0000 |
|
| 57 |
-
| 3.5786 | 4.39 | 400 | 3.6976 | 1.0 | 1.0000 |
|
| 58 |
-
| 3.5284 | 6.59 | 600 | 3.5717 | 1.0 | 1.0000 |
|
| 59 |
-
| 2.8338 | 8.79 | 800 | 2.5502 | 1.0 | 0.9847 |
|
| 60 |
-
| 1.52 | 10.98 | 1000 | 1.3082 | 0.8934 | 0.9465 |
|
| 61 |
-
| 1.1756 | 13.19 | 1200 | 1.0115 | 0.7787 | 0.9418 |
|
| 62 |
-
| 0.8762 | 15.38 | 1400 | 0.8637 | 0.7175 | 0.9389 |
|
| 63 |
-
| 0.798 | 17.58 | 1600 | 0.8104 | 0.6865 | 0.9380 |
|
| 64 |
-
| 0.6566 | 19.78 | 1800 | 0.7661 | 0.6651 | 0.9366 |
|
| 65 |
-
| 0.6085 | 21.97 | 2000 | 0.7702 | 0.6563 | 0.9365 |
|
| 66 |
-
| 0.5786 | 24.17 | 2200 | 0.7600 | 0.6570 | 0.9359 |
|
| 67 |
-
| 0.4955 | 26.37 | 2400 | 0.7820 | 0.6476 | 0.9354 |
|
| 68 |
-
| 0.4624 | 28.57 | 2600 | 0.7417 | 0.6227 | 0.9349 |
|
| 69 |
-
| 0.4263 | 30.77 | 2800 | 0.7748 | 0.6288 | 0.9350 |
|
| 70 |
-
| 0.4259 | 32.96 | 3000 | 0.7448 | 0.6148 | 0.9344 |
|
| 71 |
-
| 0.3838 | 35.16 | 3200 | 0.7742 | 0.6199 | 0.9345 |
|
| 72 |
-
| 0.3691 | 37.36 | 3400 | 0.7552 | 0.6013 | 0.9342 |
|
| 73 |
-
| 0.3818 | 39.56 | 3600 | 0.7659 | 0.6005 | 0.9341 |
|
| 74 |
-
| 0.3288 | 41.75 | 3800 | 0.7760 | 0.6006 | 0.9340 |
|
| 75 |
-
| 0.3189 | 43.95 | 4000 | 0.7773 | 0.6020 | 0.9340 |
|
| 76 |
-
| 0.3276 | 46.15 | 4200 | 0.7943 | 0.5958 | 0.9339 |
|
| 77 |
-
| 0.3216 | 48.35 | 4400 | 0.7853 | 0.5934 | 0.9338 |
|
| 78 |
-
| 0.3203 | 50.55 | 4600 | 0.8036 | 0.5962 | 0.9339 |
|
| 79 |
-
| 0.3232 | 52.74 | 4800 | 0.8027 | 0.5915 | 0.9337 |
|
| 80 |
-
| 0.3225 | 54.94 | 5000 | 0.7960 | 0.5924 | 0.9337 |
|
| 81 |
-
|
| 82 |
-
|
| 83 |
### Framework versions
|
| 84 |
|
| 85 |
- Transformers 4.26.1
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
tags:
|
| 4 |
- generated_from_trainer
|
|
|
|
|
|
|
| 5 |
model-index:
|
| 6 |
- name: wav2vec2-large-asr-th
|
| 7 |
results: []
|
|
|
|
| 12 |
|
| 13 |
# wav2vec2-large-asr-th
|
| 14 |
|
| 15 |
+
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on an unknown dataset.
|
|
|
|
|
|
|
|
|
|
|
|
|
| 16 |
|
| 17 |
## Model description
|
| 18 |
|
|
|
|
| 31 |
### Training hyperparameters
|
| 32 |
|
| 33 |
The following hyperparameters were used during training:
|
| 34 |
+
- learning_rate: 0.0002
|
| 35 |
- train_batch_size: 16
|
| 36 |
- eval_batch_size: 8
|
| 37 |
- seed: 42
|
|
|
|
| 40 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 41 |
- lr_scheduler_type: linear
|
| 42 |
- lr_scheduler_warmup_steps: 200
|
| 43 |
+
- training_steps: 3200
|
| 44 |
- mixed_precision_training: Native AMP
|
| 45 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 46 |
### Framework versions
|
| 47 |
|
| 48 |
- Transformers 4.26.1
|