update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,89 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
tags:
|
| 4 |
+
- generated_from_trainer
|
| 5 |
+
metrics:
|
| 6 |
+
- wer
|
| 7 |
+
model-index:
|
| 8 |
+
- name: hubert_model
|
| 9 |
+
results: []
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 13 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 14 |
+
|
| 15 |
+
# hubert_model
|
| 16 |
+
|
| 17 |
+
This model is a fine-tuned version of [facebook/hubert-base-ls960](https://huggingface.co/facebook/hubert-base-ls960) on the None dataset.
|
| 18 |
+
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 4.7173
|
| 20 |
+
- Wer: 1.0
|
| 21 |
+
|
| 22 |
+
## Model description
|
| 23 |
+
|
| 24 |
+
More information needed
|
| 25 |
+
|
| 26 |
+
## Intended uses & limitations
|
| 27 |
+
|
| 28 |
+
More information needed
|
| 29 |
+
|
| 30 |
+
## Training and evaluation data
|
| 31 |
+
|
| 32 |
+
More information needed
|
| 33 |
+
|
| 34 |
+
## Training procedure
|
| 35 |
+
|
| 36 |
+
### Training hyperparameters
|
| 37 |
+
|
| 38 |
+
The following hyperparameters were used during training:
|
| 39 |
+
- learning_rate: 1e-05
|
| 40 |
+
- train_batch_size: 64
|
| 41 |
+
- eval_batch_size: 16
|
| 42 |
+
- seed: 42
|
| 43 |
+
- gradient_accumulation_steps: 2
|
| 44 |
+
- total_train_batch_size: 128
|
| 45 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 46 |
+
- lr_scheduler_type: linear
|
| 47 |
+
- lr_scheduler_warmup_steps: 125
|
| 48 |
+
- num_epochs: 3
|
| 49 |
+
- mixed_precision_training: Native AMP
|
| 50 |
+
|
| 51 |
+
### Training results
|
| 52 |
+
|
| 53 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 54 |
+
|:-------------:|:-----:|:----:|:---------------:|:---:|
|
| 55 |
+
| 55.5107 | 0.11 | 100 | 93.6947 | 1.0 |
|
| 56 |
+
| 29.8329 | 0.22 | 200 | 53.0718 | 1.0 |
|
| 57 |
+
| 22.4958 | 0.32 | 300 | 42.6961 | 1.0 |
|
| 58 |
+
| 19.1734 | 0.43 | 400 | 34.1686 | 1.0 |
|
| 59 |
+
| 15.9615 | 0.54 | 500 | 27.1054 | 1.0 |
|
| 60 |
+
| 13.1077 | 0.65 | 600 | 21.2901 | 1.0 |
|
| 61 |
+
| 11.0162 | 0.76 | 700 | 16.6558 | 1.0 |
|
| 62 |
+
| 9.3359 | 0.87 | 800 | 13.1283 | 1.0 |
|
| 63 |
+
| 8.2754 | 0.97 | 900 | 10.6005 | 1.0 |
|
| 64 |
+
| 7.1321 | 1.08 | 1000 | 8.7120 | 1.0 |
|
| 65 |
+
| 6.2621 | 1.19 | 1100 | 7.4866 | 1.0 |
|
| 66 |
+
| 5.8109 | 1.3 | 1200 | 6.6416 | 1.0 |
|
| 67 |
+
| 5.386 | 1.41 | 1300 | 6.1307 | 1.0 |
|
| 68 |
+
| 5.1782 | 1.51 | 1400 | 5.8103 | 1.0 |
|
| 69 |
+
| 4.9481 | 1.62 | 1500 | 5.6119 | 1.0 |
|
| 70 |
+
| 4.8722 | 1.73 | 1600 | 5.4872 | 1.0 |
|
| 71 |
+
| 4.7617 | 1.84 | 1700 | 5.3270 | 1.0 |
|
| 72 |
+
| 4.717 | 1.95 | 1800 | 5.2877 | 1.0 |
|
| 73 |
+
| 4.6256 | 2.06 | 1900 | 5.6727 | 1.0 |
|
| 74 |
+
| 4.6255 | 2.16 | 2000 | 5.4983 | 1.0 |
|
| 75 |
+
| 4.5977 | 2.27 | 2100 | 5.2167 | 1.0 |
|
| 76 |
+
| 4.5797 | 2.38 | 2200 | 4.9743 | 1.0 |
|
| 77 |
+
| 4.5616 | 2.49 | 2300 | 4.8446 | 1.0 |
|
| 78 |
+
| 4.5476 | 2.6 | 2400 | 4.7885 | 1.0 |
|
| 79 |
+
| 4.5516 | 2.71 | 2500 | 4.7597 | 1.0 |
|
| 80 |
+
| 4.5343 | 2.81 | 2600 | 4.7309 | 1.0 |
|
| 81 |
+
| 4.586 | 2.92 | 2700 | 4.7173 | 1.0 |
|
| 82 |
+
|
| 83 |
+
|
| 84 |
+
### Framework versions
|
| 85 |
+
|
| 86 |
+
- Transformers 4.25.1
|
| 87 |
+
- Pytorch 1.12.0
|
| 88 |
+
- Datasets 2.7.1
|
| 89 |
+
- Tokenizers 0.13.2
|