Commit
•
a00787a
1
Parent(s):
33b71f6
Update README.md
Browse files
README.md
CHANGED
|
@@ -4,64 +4,25 @@ tags:
|
|
| 4 |
metrics:
|
| 5 |
- wer
|
| 6 |
model-index:
|
| 7 |
-
- name: seq-xls-r-fleurs_zu-run3-asr_xh-
|
| 8 |
results: []
|
|
|
|
|
|
|
| 9 |
---
|
| 10 |
|
| 11 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 12 |
should probably proofread and complete it, then remove this comment. -->
|
| 13 |
|
| 14 |
-
# seq-xls-r-fleurs_zu-run3-asr_xh-
|
| 15 |
-
|
| 16 |
-
This model is a fine-tuned version of [lucas-meyer/xls-r-fleurs_zu-run3](https://huggingface.co/lucas-meyer/xls-r-fleurs_zu-run3) on an unknown dataset.
|
| 17 |
-
It achieves the following results on the evaluation set:
|
| 18 |
-
- Loss: 0.4363
|
| 19 |
-
- Wer: 0.4998
|
| 20 |
-
|
| 21 |
-
## Model description
|
| 22 |
-
|
| 23 |
-
More information needed
|
| 24 |
-
|
| 25 |
-
## Intended uses & limitations
|
| 26 |
-
|
| 27 |
-
More information needed
|
| 28 |
-
|
| 29 |
-
## Training and evaluation data
|
| 30 |
-
|
| 31 |
-
More information needed
|
| 32 |
-
|
| 33 |
-
## Training procedure
|
| 34 |
-
|
| 35 |
-
### Training hyperparameters
|
| 36 |
-
|
| 37 |
-
The following hyperparameters were used during training:
|
| 38 |
-
- learning_rate: 0.0003
|
| 39 |
-
- train_batch_size: 8
|
| 40 |
-
- eval_batch_size: 8
|
| 41 |
-
- seed: 42
|
| 42 |
-
- gradient_accumulation_steps: 2
|
| 43 |
-
- total_train_batch_size: 16
|
| 44 |
-
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
-
- lr_scheduler_type: linear
|
| 46 |
-
- lr_scheduler_warmup_steps: 500
|
| 47 |
-
- num_epochs: 30
|
| 48 |
-
- mixed_precision_training: Native AMP
|
| 49 |
-
|
| 50 |
-
### Training results
|
| 51 |
-
|
| 52 |
-
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 53 |
-
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
| 54 |
-
| 5.4903 | 1.59 | 250 | 0.9489 | 0.7525 |
|
| 55 |
-
| 0.6278 | 3.18 | 500 | 0.4877 | 0.6284 |
|
| 56 |
-
| 0.3831 | 4.78 | 750 | 0.4181 | 0.5640 |
|
| 57 |
-
| 0.2639 | 6.37 | 1000 | 0.3858 | 0.5018 |
|
| 58 |
-
| 0.1961 | 7.96 | 1250 | 0.4228 | 0.4995 |
|
| 59 |
-
| 0.1579 | 9.55 | 1500 | 0.4363 | 0.4998 |
|
| 60 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 61 |
|
| 62 |
### Framework versions
|
| 63 |
|
| 64 |
- Transformers 4.28.0
|
| 65 |
- Pytorch 2.0.1+cu117
|
| 66 |
- Datasets 2.14.4
|
| 67 |
-
- Tokenizers 0.13.3
|
|
|
|
| 4 |
metrics:
|
| 5 |
- wer
|
| 6 |
model-index:
|
| 7 |
+
- name: seq-xls-r-fleurs_zu-run3-asr_xh-run7-LM
|
| 8 |
results: []
|
| 9 |
+
datasets:
|
| 10 |
+
- lucas-meyer/asr_xh
|
| 11 |
---
|
| 12 |
|
| 13 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 14 |
should probably proofread and complete it, then remove this comment. -->
|
| 15 |
|
| 16 |
+
# seq-xls-r-fleurs_zu-run3-asr_xh-run7-LM
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
|
| 18 |
+
This model is a fine-tuned version of [lucas-meyer/xls-r-fleurs_zu-run3](https://huggingface.co/lucas-meyer/xls-r-fleurs_zu-run3) on asr_af with an added LM (5-gram Kneser-Ney).
|
| 19 |
+
It achieves the following results:
|
| 20 |
+
- Loss: 0.3923
|
| 21 |
+
- Wer: 0.3993
|
| 22 |
|
| 23 |
### Framework versions
|
| 24 |
|
| 25 |
- Transformers 4.28.0
|
| 26 |
- Pytorch 2.0.1+cu117
|
| 27 |
- Datasets 2.14.4
|
| 28 |
+
- Tokenizers 0.13.3
|