update model card README.md
Browse files
README.md
CHANGED
|
@@ -4,9 +4,24 @@ tags:
|
|
| 4 |
- generated_from_trainer
|
| 5 |
datasets:
|
| 6 |
- common_voice_12_0
|
|
|
|
|
|
|
| 7 |
model-index:
|
| 8 |
- name: wav2vec2-large-xls-r-1b-frisian
|
| 9 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
---
|
| 11 |
|
| 12 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -16,13 +31,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-1b](https://huggingface.co/facebook/wav2vec2-xls-r-1b) on the common_voice_12_0 dataset.
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
-
-
|
| 20 |
-
-
|
| 21 |
-
- eval_runtime: 211.4996
|
| 22 |
-
- eval_samples_per_second: 14.605
|
| 23 |
-
- eval_steps_per_second: 1.83
|
| 24 |
-
- epoch: 14.71
|
| 25 |
-
- step: 1750
|
| 26 |
|
| 27 |
## Model description
|
| 28 |
|
|
@@ -53,6 +63,35 @@ The following hyperparameters were used during training:
|
|
| 53 |
- num_epochs: 50
|
| 54 |
- mixed_precision_training: Native AMP
|
| 55 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 56 |
### Framework versions
|
| 57 |
|
| 58 |
- Transformers 4.27.3
|
|
|
|
| 4 |
- generated_from_trainer
|
| 5 |
datasets:
|
| 6 |
- common_voice_12_0
|
| 7 |
+
metrics:
|
| 8 |
+
- wer
|
| 9 |
model-index:
|
| 10 |
- name: wav2vec2-large-xls-r-1b-frisian
|
| 11 |
+
results:
|
| 12 |
+
- task:
|
| 13 |
+
name: Automatic Speech Recognition
|
| 14 |
+
type: automatic-speech-recognition
|
| 15 |
+
dataset:
|
| 16 |
+
name: common_voice_12_0
|
| 17 |
+
type: common_voice_12_0
|
| 18 |
+
config: fy-NL
|
| 19 |
+
split: validation
|
| 20 |
+
args: fy-NL
|
| 21 |
+
metrics:
|
| 22 |
+
- name: Wer
|
| 23 |
+
type: wer
|
| 24 |
+
value: 0.1685917915949865
|
| 25 |
---
|
| 26 |
|
| 27 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 31 |
|
| 32 |
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-1b](https://huggingface.co/facebook/wav2vec2-xls-r-1b) on the common_voice_12_0 dataset.
|
| 33 |
It achieves the following results on the evaluation set:
|
| 34 |
+
- Loss: 0.2748
|
| 35 |
+
- Wer: 0.1686
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 36 |
|
| 37 |
## Model description
|
| 38 |
|
|
|
|
| 63 |
- num_epochs: 50
|
| 64 |
- mixed_precision_training: Native AMP
|
| 65 |
|
| 66 |
+
### Training results
|
| 67 |
+
|
| 68 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 69 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
| 70 |
+
| 4.6809 | 2.1 | 250 | 2.0948 | 0.9948 |
|
| 71 |
+
| 1.2928 | 4.2 | 500 | 0.4505 | 0.4003 |
|
| 72 |
+
| 0.7887 | 6.3 | 750 | 0.3410 | 0.3287 |
|
| 73 |
+
| 0.7422 | 8.4 | 1000 | 0.3017 | 0.2756 |
|
| 74 |
+
| 0.7277 | 10.5 | 1250 | 0.3014 | 0.2624 |
|
| 75 |
+
| 0.6339 | 12.61 | 1500 | 0.2833 | 0.2398 |
|
| 76 |
+
| 0.5284 | 14.71 | 1750 | 0.2970 | 0.2404 |
|
| 77 |
+
| 0.5186 | 16.81 | 2000 | 0.2886 | 0.2400 |
|
| 78 |
+
| 0.515 | 18.91 | 2250 | 0.2891 | 0.2335 |
|
| 79 |
+
| 0.5199 | 21.01 | 2500 | 0.2985 | 0.2261 |
|
| 80 |
+
| 0.5228 | 23.11 | 2750 | 0.3026 | 0.2187 |
|
| 81 |
+
| 0.5102 | 25.21 | 3000 | 0.2829 | 0.1994 |
|
| 82 |
+
| 0.463 | 27.31 | 3250 | 0.2885 | 0.2012 |
|
| 83 |
+
| 0.5072 | 29.41 | 3500 | 0.2936 | 0.1971 |
|
| 84 |
+
| 0.4581 | 31.51 | 3750 | 0.2979 | 0.1912 |
|
| 85 |
+
| 0.4103 | 33.61 | 4000 | 0.2935 | 0.1875 |
|
| 86 |
+
| 0.3414 | 35.71 | 4250 | 0.2999 | 0.1860 |
|
| 87 |
+
| 0.4484 | 37.82 | 4500 | 0.2917 | 0.1810 |
|
| 88 |
+
| 0.3523 | 39.92 | 4750 | 0.2875 | 0.1759 |
|
| 89 |
+
| 0.3763 | 42.02 | 5000 | 0.2901 | 0.1758 |
|
| 90 |
+
| 0.2416 | 44.12 | 5250 | 0.2707 | 0.1740 |
|
| 91 |
+
| 0.1878 | 46.22 | 5500 | 0.2707 | 0.1717 |
|
| 92 |
+
| 0.1623 | 48.32 | 5750 | 0.2748 | 0.1686 |
|
| 93 |
+
|
| 94 |
+
|
| 95 |
### Framework versions
|
| 96 |
|
| 97 |
- Transformers 4.27.3
|