update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,77 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
tags:
|
| 4 |
+
- generated_from_trainer
|
| 5 |
+
datasets:
|
| 6 |
+
- common_voice_10_0
|
| 7 |
+
model-index:
|
| 8 |
+
- name: fine-tune-Wav2Vec2-XLS-R-300M-Indonesia-3
|
| 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 |
+
# fine-tune-Wav2Vec2-XLS-R-300M-Indonesia-3
|
| 16 |
+
|
| 17 |
+
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the common_voice_10_0 dataset.
|
| 18 |
+
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 0.2856
|
| 20 |
+
- Wer: 0.1931
|
| 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: 0.0001
|
| 40 |
+
- train_batch_size: 36
|
| 41 |
+
- eval_batch_size: 8
|
| 42 |
+
- seed: 42
|
| 43 |
+
- gradient_accumulation_steps: 2
|
| 44 |
+
- total_train_batch_size: 72
|
| 45 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 46 |
+
- lr_scheduler_type: linear
|
| 47 |
+
- lr_scheduler_warmup_steps: 500
|
| 48 |
+
- num_epochs: 60
|
| 49 |
+
- mixed_precision_training: Native AMP
|
| 50 |
+
|
| 51 |
+
### Training results
|
| 52 |
+
|
| 53 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 54 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
| 55 |
+
| 2.8503 | 4.0 | 460 | 1.2215 | 0.8013 |
|
| 56 |
+
| 0.2917 | 8.0 | 920 | 0.2824 | 0.3055 |
|
| 57 |
+
| 0.1487 | 12.0 | 1380 | 0.2719 | 0.2593 |
|
| 58 |
+
| 0.0904 | 16.0 | 1840 | 0.2536 | 0.2448 |
|
| 59 |
+
| 0.0741 | 20.0 | 2300 | 0.2620 | 0.2325 |
|
| 60 |
+
| 0.0635 | 24.0 | 2760 | 0.2953 | 0.2354 |
|
| 61 |
+
| 0.0541 | 28.0 | 3220 | 0.2683 | 0.2184 |
|
| 62 |
+
| 0.048 | 32.0 | 3680 | 0.2707 | 0.2097 |
|
| 63 |
+
| 0.0435 | 36.0 | 4140 | 0.2764 | 0.2111 |
|
| 64 |
+
| 0.0377 | 40.0 | 4600 | 0.2874 | 0.2048 |
|
| 65 |
+
| 0.0352 | 44.0 | 5060 | 0.2758 | 0.1999 |
|
| 66 |
+
| 0.0304 | 48.0 | 5520 | 0.2808 | 0.1969 |
|
| 67 |
+
| 0.0285 | 52.0 | 5980 | 0.2860 | 0.1943 |
|
| 68 |
+
| 0.0258 | 56.0 | 6440 | 0.2867 | 0.1943 |
|
| 69 |
+
| 0.0239 | 60.0 | 6900 | 0.2856 | 0.1931 |
|
| 70 |
+
|
| 71 |
+
|
| 72 |
+
### Framework versions
|
| 73 |
+
|
| 74 |
+
- Transformers 4.22.2
|
| 75 |
+
- Pytorch 1.10.0+cu113
|
| 76 |
+
- Datasets 2.5.2
|
| 77 |
+
- Tokenizers 0.12.1
|