Commit
·
aa5ed2c
1
Parent(s):
517f43e
update model card README.md
Browse files
README.md
CHANGED
|
@@ -8,56 +8,9 @@ tags:
|
|
| 8 |
- collectivat/tv3_parla
|
| 9 |
- projecte-aina/parlament_parla
|
| 10 |
- generated_from_trainer
|
| 11 |
-
- robust-speech-event
|
| 12 |
-
datasets:
|
| 13 |
-
- mozilla-foundation/common_voice_8_0
|
| 14 |
-
- collectivat/tv3_parla
|
| 15 |
-
- projecte-aina/parlament_parla
|
| 16 |
model-index:
|
| 17 |
- name: wav2vec2-xls-r-300m-ca
|
| 18 |
-
results:
|
| 19 |
-
- task:
|
| 20 |
-
name: Speech Recognition
|
| 21 |
-
type: automatic-speech-recognition
|
| 22 |
-
dataset:
|
| 23 |
-
name: mozilla-foundation/common_voice_8_0 ca
|
| 24 |
-
type: mozilla-foundation/common_voice_8_0
|
| 25 |
-
args: ca
|
| 26 |
-
metrics:
|
| 27 |
-
- name: Test WER
|
| 28 |
-
type: wer
|
| 29 |
-
value: 0.15636874077301
|
| 30 |
-
- name: Test CER
|
| 31 |
-
type: cer
|
| 32 |
-
value: 0.04086725403909639
|
| 33 |
-
- task:
|
| 34 |
-
name: Speech Recognition
|
| 35 |
-
type: automatic-speech-recognition
|
| 36 |
-
dataset:
|
| 37 |
-
name: projecte-aina/parlament_parla ca
|
| 38 |
-
type: projecte-aina/parlament_parla
|
| 39 |
-
args: clean
|
| 40 |
-
metrics:
|
| 41 |
-
- name: Test WER
|
| 42 |
-
type: wer
|
| 43 |
-
value: 0.09940385143350199
|
| 44 |
-
- name: Test CER
|
| 45 |
-
type: cer
|
| 46 |
-
value: 0.026906712890009454
|
| 47 |
-
- task:
|
| 48 |
-
name: Speech Recognition
|
| 49 |
-
type: automatic-speech-recognition
|
| 50 |
-
dataset:
|
| 51 |
-
name: collectivat/tv3_parla ca
|
| 52 |
-
type: collectivat/tv3_parla
|
| 53 |
-
args: ca
|
| 54 |
-
metrics:
|
| 55 |
-
- name: Test WER
|
| 56 |
-
type: wer
|
| 57 |
-
value: 0.27349193517342263
|
| 58 |
-
- name: Test CER
|
| 59 |
-
type: cer
|
| 60 |
-
value: 0.11571091827304163
|
| 61 |
---
|
| 62 |
|
| 63 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -67,8 +20,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 67 |
|
| 68 |
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the MOZILLA-FOUNDATION/COMMON_VOICE_8_0 - CA dataset.
|
| 69 |
It achieves the following results on the evaluation set:
|
| 70 |
-
- Loss: 0.
|
| 71 |
-
- Wer: 0.
|
| 72 |
|
| 73 |
## Model description
|
| 74 |
|
|
@@ -96,7 +49,7 @@ The following hyperparameters were used during training:
|
|
| 96 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 97 |
- lr_scheduler_type: linear
|
| 98 |
- lr_scheduler_warmup_steps: 2000
|
| 99 |
-
- num_epochs:
|
| 100 |
- mixed_precision_training: Native AMP
|
| 101 |
|
| 102 |
### Training results
|
|
@@ -125,6 +78,29 @@ The following hyperparameters were used during training:
|
|
| 125 |
| 1.1407 | 5.33 | 10000 | 0.2768 | 0.1798 |
|
| 126 |
| 1.1401 | 5.59 | 10500 | 0.2780 | 0.1816 |
|
| 127 |
| 1.1333 | 5.86 | 11000 | 0.2748 | 0.1750 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 128 |
|
| 129 |
|
| 130 |
### Framework versions
|
|
|
|
| 8 |
- collectivat/tv3_parla
|
| 9 |
- projecte-aina/parlament_parla
|
| 10 |
- generated_from_trainer
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
model-index:
|
| 12 |
- name: wav2vec2-xls-r-300m-ca
|
| 13 |
+
results: []
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 14 |
---
|
| 15 |
|
| 16 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 20 |
|
| 21 |
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the MOZILLA-FOUNDATION/COMMON_VOICE_8_0 - CA dataset.
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 0.2549
|
| 24 |
+
- Wer: 0.1573
|
| 25 |
|
| 26 |
## Model description
|
| 27 |
|
|
|
|
| 49 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 50 |
- lr_scheduler_type: linear
|
| 51 |
- lr_scheduler_warmup_steps: 2000
|
| 52 |
+
- num_epochs: 12.0
|
| 53 |
- mixed_precision_training: Native AMP
|
| 54 |
|
| 55 |
### Training results
|
|
|
|
| 78 |
| 1.1407 | 5.33 | 10000 | 0.2768 | 0.1798 |
|
| 79 |
| 1.1401 | 5.59 | 10500 | 0.2780 | 0.1816 |
|
| 80 |
| 1.1333 | 5.86 | 11000 | 0.2748 | 0.1750 |
|
| 81 |
+
| 1.1571 | 6.13 | 11500 | 0.2808 | 0.1708 |
|
| 82 |
+
| 1.1505 | 6.39 | 12000 | 0.2726 | 0.1692 |
|
| 83 |
+
| 1.1519 | 6.66 | 12500 | 0.2749 | 0.1654 |
|
| 84 |
+
| 1.136 | 6.93 | 13000 | 0.2765 | 0.1643 |
|
| 85 |
+
| 1.1326 | 7.19 | 13500 | 0.2706 | 0.1668 |
|
| 86 |
+
| 1.1342 | 7.46 | 14000 | 0.2665 | 0.1638 |
|
| 87 |
+
| 1.1286 | 7.72 | 14500 | 0.2669 | 0.1636 |
|
| 88 |
+
| 1.1243 | 7.99 | 15000 | 0.2619 | 0.1623 |
|
| 89 |
+
| 1.1173 | 8.26 | 15500 | 0.2652 | 0.1604 |
|
| 90 |
+
| 1.1129 | 8.52 | 16000 | 0.2610 | 0.1598 |
|
| 91 |
+
| 1.1091 | 8.79 | 16500 | 0.2608 | 0.1584 |
|
| 92 |
+
| 1.1053 | 9.06 | 17000 | 0.2633 | 0.1664 |
|
| 93 |
+
| 1.1004 | 9.32 | 17500 | 0.2594 | 0.1662 |
|
| 94 |
+
| 1.0995 | 9.59 | 18000 | 0.2623 | 0.1569 |
|
| 95 |
+
| 1.0964 | 9.86 | 18500 | 0.2624 | 0.1597 |
|
| 96 |
+
| 1.09 | 10.12 | 19000 | 0.2577 | 0.1578 |
|
| 97 |
+
| 1.089 | 10.39 | 19500 | 0.2574 | 0.1531 |
|
| 98 |
+
| 1.0864 | 10.66 | 20000 | 0.2556 | 0.1546 |
|
| 99 |
+
| 1.0806 | 10.92 | 20500 | 0.2548 | 0.1583 |
|
| 100 |
+
| 1.0842 | 11.19 | 21000 | 0.2550 | 0.1542 |
|
| 101 |
+
| 1.0805 | 11.45 | 21500 | 0.2561 | 0.1524 |
|
| 102 |
+
| 1.0722 | 11.72 | 22000 | 0.2540 | 0.1566 |
|
| 103 |
+
| 1.0763 | 11.99 | 22500 | 0.2549 | 0.1572 |
|
| 104 |
|
| 105 |
|
| 106 |
### Framework versions
|