update model card README.md
Browse files
README.md
ADDED
@@ -0,0 +1,76 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
tags:
|
4 |
+
- generated_from_trainer
|
5 |
+
datasets:
|
6 |
+
- filipino_voice
|
7 |
+
model-index:
|
8 |
+
- name: english-filipino-wav2vec2-l-xls-r-test-06
|
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 |
+
# english-filipino-wav2vec2-l-xls-r-test-06
|
16 |
+
|
17 |
+
This model is a fine-tuned version of [jonatasgrosman/wav2vec2-large-xlsr-53-english](https://huggingface.co/jonatasgrosman/wav2vec2-large-xlsr-53-english) on the filipino_voice dataset.
|
18 |
+
It achieves the following results on the evaluation set:
|
19 |
+
- Loss: 1.5806
|
20 |
+
- Wer: 0.6568
|
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: 0.002
|
40 |
+
- train_batch_size: 8
|
41 |
+
- eval_batch_size: 8
|
42 |
+
- seed: 42
|
43 |
+
- gradient_accumulation_steps: 2
|
44 |
+
- total_train_batch_size: 16
|
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: 30
|
49 |
+
- mixed_precision_training: Native AMP
|
50 |
+
|
51 |
+
### Training results
|
52 |
+
|
53 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
54 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
55 |
+
| 3.0031 | 2.09 | 400 | 1.2366 | 0.8780 |
|
56 |
+
| 0.9084 | 4.19 | 800 | 1.0653 | 0.8081 |
|
57 |
+
| 0.6484 | 6.28 | 1200 | 1.1648 | 0.8258 |
|
58 |
+
| 0.5335 | 8.38 | 1600 | 1.0903 | 0.7542 |
|
59 |
+
| 0.4359 | 10.47 | 2000 | 0.9466 | 0.7058 |
|
60 |
+
| 0.3629 | 12.57 | 2400 | 0.9266 | 0.7048 |
|
61 |
+
| 0.3057 | 14.66 | 2800 | 1.0879 | 0.7018 |
|
62 |
+
| 0.2477 | 16.75 | 3200 | 1.1113 | 0.7022 |
|
63 |
+
| 0.208 | 18.85 | 3600 | 1.1345 | 0.6742 |
|
64 |
+
| 0.1781 | 20.94 | 4000 | 1.3117 | 0.6974 |
|
65 |
+
| 0.1465 | 23.04 | 4400 | 1.3248 | 0.6916 |
|
66 |
+
| 0.1288 | 25.13 | 4800 | 1.4306 | 0.6523 |
|
67 |
+
| 0.1108 | 27.23 | 5200 | 1.5155 | 0.6685 |
|
68 |
+
| 0.099 | 29.32 | 5600 | 1.5806 | 0.6568 |
|
69 |
+
|
70 |
+
|
71 |
+
### Framework versions
|
72 |
+
|
73 |
+
- Transformers 4.11.3
|
74 |
+
- Pytorch 1.10.0+cu113
|
75 |
+
- Datasets 1.18.3
|
76 |
+
- Tokenizers 0.10.3
|