update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,68 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
tags:
|
| 3 |
+
- generated_from_trainer
|
| 4 |
+
model-index:
|
| 5 |
+
- name: timit-supervised
|
| 6 |
+
results: []
|
| 7 |
+
---
|
| 8 |
+
|
| 9 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 10 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 11 |
+
|
| 12 |
+
# timit-supervised
|
| 13 |
+
|
| 14 |
+
This model is a fine-tuned version of [Experiments/single_dataset/timit-supervised/checkpoint-3500](https://huggingface.co/Experiments/single_dataset/timit-supervised/checkpoint-3500) on the None dataset.
|
| 15 |
+
It achieves the following results on the evaluation set:
|
| 16 |
+
- Loss: 0.1272
|
| 17 |
+
- Wer: 0.0532
|
| 18 |
+
|
| 19 |
+
## Model description
|
| 20 |
+
|
| 21 |
+
More information needed
|
| 22 |
+
|
| 23 |
+
## Intended uses & limitations
|
| 24 |
+
|
| 25 |
+
More information needed
|
| 26 |
+
|
| 27 |
+
## Training and evaluation data
|
| 28 |
+
|
| 29 |
+
More information needed
|
| 30 |
+
|
| 31 |
+
## Training procedure
|
| 32 |
+
|
| 33 |
+
### Training hyperparameters
|
| 34 |
+
|
| 35 |
+
The following hyperparameters were used during training:
|
| 36 |
+
- learning_rate: 0.0001
|
| 37 |
+
- train_batch_size: 16
|
| 38 |
+
- eval_batch_size: 8
|
| 39 |
+
- 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: 1000
|
| 43 |
+
- num_epochs: 20
|
| 44 |
+
- mixed_precision_training: Native AMP
|
| 45 |
+
|
| 46 |
+
### Training results
|
| 47 |
+
|
| 48 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 49 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
| 50 |
+
| 0.0554 | 1.77 | 500 | 0.1310 | 0.0697 |
|
| 51 |
+
| 0.0509 | 3.53 | 1000 | 0.1497 | 0.0710 |
|
| 52 |
+
| 0.038 | 5.3 | 1500 | 0.1190 | 0.0659 |
|
| 53 |
+
| 0.0328 | 7.07 | 2000 | 0.0926 | 0.0596 |
|
| 54 |
+
| 0.0247 | 8.83 | 2500 | 0.0873 | 0.0570 |
|
| 55 |
+
| 0.0229 | 10.6 | 3000 | 0.0890 | 0.0532 |
|
| 56 |
+
| 0.0183 | 12.37 | 3500 | 0.0969 | 0.0532 |
|
| 57 |
+
| 0.0326 | 14.13 | 4000 | 0.0809 | 0.0469 |
|
| 58 |
+
| 0.03 | 15.9 | 4500 | 0.0758 | 0.0444 |
|
| 59 |
+
| 0.0264 | 17.67 | 5000 | 0.0973 | 0.0520 |
|
| 60 |
+
| 0.0244 | 19.43 | 5500 | 0.1272 | 0.0532 |
|
| 61 |
+
|
| 62 |
+
|
| 63 |
+
### Framework versions
|
| 64 |
+
|
| 65 |
+
- Transformers 4.14.1
|
| 66 |
+
- Pytorch 1.10.2
|
| 67 |
+
- Datasets 1.18.2
|
| 68 |
+
- Tokenizers 0.10.3
|