wav2vec2-large-xls-r-300m-czech-colab-finetuned
Browse files
README.md
CHANGED
|
@@ -5,9 +5,24 @@ tags:
|
|
| 5 |
- generated_from_trainer
|
| 6 |
datasets:
|
| 7 |
- voxpopuli
|
|
|
|
|
|
|
| 8 |
model-index:
|
| 9 |
- name: wav2vec2-large-xls-r-300m-czech-colab-finetuned
|
| 10 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
---
|
| 12 |
|
| 13 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -17,12 +32,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [facebook/wav2vec2-lv-60-espeak-cv-ft](https://huggingface.co/facebook/wav2vec2-lv-60-espeak-cv-ft) on the voxpopuli dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
-
|
| 21 |
-
-
|
| 22 |
-
- eval_runtime: 119.4799
|
| 23 |
-
- eval_samples_per_second: 4.185
|
| 24 |
-
- eval_steps_per_second: 0.527
|
| 25 |
-
- step: 0
|
| 26 |
|
| 27 |
## Model description
|
| 28 |
|
|
@@ -53,6 +64,26 @@ The following hyperparameters were used during training:
|
|
| 53 |
- num_epochs: 50
|
| 54 |
- mixed_precision_training: Native AMP
|
| 55 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 56 |
### Framework versions
|
| 57 |
|
| 58 |
- Transformers 4.35.2
|
|
|
|
| 5 |
- generated_from_trainer
|
| 6 |
datasets:
|
| 7 |
- voxpopuli
|
| 8 |
+
metrics:
|
| 9 |
+
- wer
|
| 10 |
model-index:
|
| 11 |
- name: wav2vec2-large-xls-r-300m-czech-colab-finetuned
|
| 12 |
+
results:
|
| 13 |
+
- task:
|
| 14 |
+
name: Automatic Speech Recognition
|
| 15 |
+
type: automatic-speech-recognition
|
| 16 |
+
dataset:
|
| 17 |
+
name: voxpopuli
|
| 18 |
+
type: voxpopuli
|
| 19 |
+
config: cs
|
| 20 |
+
split: test
|
| 21 |
+
args: cs
|
| 22 |
+
metrics:
|
| 23 |
+
- name: Wer
|
| 24 |
+
type: wer
|
| 25 |
+
value: 0.6178421298458664
|
| 26 |
---
|
| 27 |
|
| 28 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 32 |
|
| 33 |
This model is a fine-tuned version of [facebook/wav2vec2-lv-60-espeak-cv-ft](https://huggingface.co/facebook/wav2vec2-lv-60-espeak-cv-ft) on the voxpopuli dataset.
|
| 34 |
It achieves the following results on the evaluation set:
|
| 35 |
+
- Loss: 624.5939
|
| 36 |
+
- Wer: 0.6178
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
|
| 38 |
## Model description
|
| 39 |
|
|
|
|
| 64 |
- num_epochs: 50
|
| 65 |
- mixed_precision_training: Native AMP
|
| 66 |
|
| 67 |
+
### Training results
|
| 68 |
+
|
| 69 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 70 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
| 71 |
+
| 3007.3212 | 3.51 | 100 | 1006.7374 | 0.9865 |
|
| 72 |
+
| 354.3011 | 7.02 | 200 | 563.6080 | 0.9980 |
|
| 73 |
+
| 211.5289 | 10.53 | 300 | 599.5796 | 0.9165 |
|
| 74 |
+
| 187.8653 | 14.04 | 400 | 447.1478 | 0.8099 |
|
| 75 |
+
| 163.1056 | 17.54 | 500 | 430.5204 | 0.6875 |
|
| 76 |
+
| 143.0342 | 21.05 | 600 | 413.8947 | 0.6850 |
|
| 77 |
+
| 116.0388 | 24.56 | 700 | 435.5743 | 0.6737 |
|
| 78 |
+
| 95.5554 | 28.07 | 800 | 490.6329 | 0.6339 |
|
| 79 |
+
| 80.6966 | 31.58 | 900 | 493.9658 | 0.6344 |
|
| 80 |
+
| 68.7335 | 35.09 | 1000 | 525.7507 | 0.6263 |
|
| 81 |
+
| 58.3269 | 38.6 | 1100 | 582.5747 | 0.6128 |
|
| 82 |
+
| 54.3181 | 42.11 | 1200 | 600.8087 | 0.6308 |
|
| 83 |
+
| 48.5287 | 45.61 | 1300 | 594.6959 | 0.6112 |
|
| 84 |
+
| 43.041 | 49.12 | 1400 | 624.5939 | 0.6178 |
|
| 85 |
+
|
| 86 |
+
|
| 87 |
### Framework versions
|
| 88 |
|
| 89 |
- Transformers 4.35.2
|