Adding `safetensors` variant of this model
#3
by
SFconvertbot
- opened
- README.md +4 -8
- language_model/5gram.bin +0 -3
- language_model/unigrams.txt +0 -0
- model.safetensors +0 -3
- preprocessor_config.json +10 -10
- pytorch_model.bin +0 -3
README.md
CHANGED
|
@@ -10,7 +10,6 @@ datasets:
|
|
| 10 |
- mozilla-foundation/common_voice_8_0
|
| 11 |
metrics:
|
| 12 |
- wer
|
| 13 |
-
- cer
|
| 14 |
base_model: facebook/wav2vec2-xls-r-300m
|
| 15 |
model-index:
|
| 16 |
- name: wav2vec2-large-xls-r-300m-Urdu
|
|
@@ -30,7 +29,6 @@ model-index:
|
|
| 30 |
value: 16.7
|
| 31 |
name: Test CER
|
| 32 |
new_version: kingabzpro/whisper-large-v3-turbo-urdu
|
| 33 |
-
pipeline_tag: automatic-speech-recognition
|
| 34 |
---
|
| 35 |
|
| 36 |
|
|
@@ -56,8 +54,6 @@ python eval.py --model_id kingabzpro/wav2vec2-large-xls-r-300m-Urdu --dataset mo
|
|
| 56 |
### Inference With LM
|
| 57 |
|
| 58 |
```python
|
| 59 |
-
#pip install pyctcdecode kenlm
|
| 60 |
-
|
| 61 |
from datasets import load_dataset, Audio
|
| 62 |
from transformers import pipeline
|
| 63 |
model = "kingabzpro/wav2vec2-large-xls-r-300m-Urdu"
|
|
@@ -65,18 +61,18 @@ data = load_dataset("mozilla-foundation/common_voice_8_0",
|
|
| 65 |
"ur",
|
| 66 |
split="test",
|
| 67 |
streaming=True,
|
| 68 |
-
|
| 69 |
|
| 70 |
-
sample_iter = iter(data.cast_column("
|
| 71 |
Audio(sampling_rate=16_000)))
|
| 72 |
sample = next(sample_iter)
|
| 73 |
|
| 74 |
asr = pipeline("automatic-speech-recognition", model=model)
|
| 75 |
-
prediction = asr(sample["
|
| 76 |
chunk_length_s=5,
|
| 77 |
stride_length_s=1)
|
| 78 |
prediction
|
| 79 |
-
# => {'text': '
|
| 80 |
```
|
| 81 |
|
| 82 |
|
|
|
|
| 10 |
- mozilla-foundation/common_voice_8_0
|
| 11 |
metrics:
|
| 12 |
- wer
|
|
|
|
| 13 |
base_model: facebook/wav2vec2-xls-r-300m
|
| 14 |
model-index:
|
| 15 |
- name: wav2vec2-large-xls-r-300m-Urdu
|
|
|
|
| 29 |
value: 16.7
|
| 30 |
name: Test CER
|
| 31 |
new_version: kingabzpro/whisper-large-v3-turbo-urdu
|
|
|
|
| 32 |
---
|
| 33 |
|
| 34 |
|
|
|
|
| 54 |
### Inference With LM
|
| 55 |
|
| 56 |
```python
|
|
|
|
|
|
|
| 57 |
from datasets import load_dataset, Audio
|
| 58 |
from transformers import pipeline
|
| 59 |
model = "kingabzpro/wav2vec2-large-xls-r-300m-Urdu"
|
|
|
|
| 61 |
"ur",
|
| 62 |
split="test",
|
| 63 |
streaming=True,
|
| 64 |
+
use_auth_token=True)
|
| 65 |
|
| 66 |
+
sample_iter = iter(data.cast_column("path",
|
| 67 |
Audio(sampling_rate=16_000)))
|
| 68 |
sample = next(sample_iter)
|
| 69 |
|
| 70 |
asr = pipeline("automatic-speech-recognition", model=model)
|
| 71 |
+
prediction = asr(sample["path"]["array"],
|
| 72 |
chunk_length_s=5,
|
| 73 |
stride_length_s=1)
|
| 74 |
prediction
|
| 75 |
+
# => {'text': 'اب یہ ونگین لمحاتانکھار دلمیں میںفوث کریلیا اجائ'}
|
| 76 |
```
|
| 77 |
|
| 78 |
|
language_model/5gram.bin
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:a51741105f0d3b54ac8ede33500c8f80fcd6b03900080a45a5661ec8e58b776f
|
| 3 |
-
size 165145481
|
|
|
|
|
|
|
|
|
|
|
|
language_model/unigrams.txt
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
model.safetensors
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:9281f8afcaaff6f9774595749c225cf30ab5a7377644eed35079e46c6d770cfd
|
| 3 |
-
size 1262041132
|
|
|
|
|
|
|
|
|
|
|
|
preprocessor_config.json
CHANGED
|
@@ -1,10 +1,10 @@
|
|
| 1 |
-
{
|
| 2 |
-
"do_normalize": true,
|
| 3 |
-
"feature_extractor_type": "Wav2Vec2FeatureExtractor",
|
| 4 |
-
"feature_size": 1,
|
| 5 |
-
"padding_side": "right",
|
| 6 |
-
"padding_value": 0.0,
|
| 7 |
-
"processor_class": "Wav2Vec2ProcessorWithLM",
|
| 8 |
-
"return_attention_mask": true,
|
| 9 |
-
"sampling_rate": 16000
|
| 10 |
-
}
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_normalize": true,
|
| 3 |
+
"feature_extractor_type": "Wav2Vec2FeatureExtractor",
|
| 4 |
+
"feature_size": 1,
|
| 5 |
+
"padding_side": "right",
|
| 6 |
+
"padding_value": 0.0,
|
| 7 |
+
"processor_class": "Wav2Vec2ProcessorWithLM",
|
| 8 |
+
"return_attention_mask": true,
|
| 9 |
+
"sampling_rate": 16000
|
| 10 |
+
}
|
pytorch_model.bin
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:dc727bffccba0121b0f32f9c8f5cb9e2f6fed08733925ad5fae6f9ce5c0de681
|
| 3 |
-
size 1262157361
|
|
|
|
|
|
|
|
|
|
|
|