Update README.md
Browse files
README.md
CHANGED
|
@@ -32,11 +32,11 @@ tags:
|
|
| 32 |
|
| 33 |
## Summary
|
| 34 |
|
| 35 |
-
The "langtech-veu/faster-
|
| 36 |
|
| 37 |
## Model Description
|
| 38 |
|
| 39 |
-
The "langtech-veu/faster-
|
| 40 |
|
| 41 |
## Intended Uses and Limitations
|
| 42 |
|
|
@@ -70,7 +70,7 @@ To transcribe audio in Catalan using this model, you can follow this example:
|
|
| 70 |
```python
|
| 71 |
from faster_whisper import WhisperModel
|
| 72 |
|
| 73 |
-
model_size = "langtech-veu/faster-
|
| 74 |
|
| 75 |
# Run on GPU with FP16
|
| 76 |
model = WhisperModel(model_size, device="cuda", compute_type="float16")
|
|
@@ -95,8 +95,8 @@ for segment in segments:
|
|
| 95 |
This model is not a direct result of training. It is a conversion of a [Whisper](https://huggingface.co/openai/whisper-large-v3) model using [faster-whisper](https://github.com/guillaumekln/faster-whisper/tree/master). The procedure to create the model is as follows:
|
| 96 |
|
| 97 |
```bash
|
| 98 |
-
ct2-transformers-converter --model langtech-veu/
|
| 99 |
-
--output_dir
|
| 100 |
--copy_files preprocessor_config.json
|
| 101 |
--quantization float16
|
| 102 |
```
|
|
@@ -105,10 +105,10 @@ ct2-transformers-converter --model langtech-veu/Whisper_bsc_large_v3_cat
|
|
| 105 |
If this model contributes to your research, please cite the work:
|
| 106 |
```
|
| 107 |
@misc{takanori2025whisperbsclarge3cat,
|
| 108 |
-
title={Acoustic Model in Catalan:
|
| 109 |
author={Sanchez Shiromizu, Lucas Takanori; Hernandez Mena, Carlos Daniel; Messaoudi, Abir; España i Bonet, Cristina; Cortada Garcia, Marti},
|
| 110 |
organization={Barcelona Supercomputing Center},
|
| 111 |
-
url={https://huggingface.co/langtech-veu/
|
| 112 |
year={2025}
|
| 113 |
}
|
| 114 |
```
|
|
|
|
| 32 |
|
| 33 |
## Summary
|
| 34 |
|
| 35 |
+
The "langtech-veu/faster-whisper-bsc-large-v3-cat" is an acoustic model based on a [faster-whisper](https://github.com/guillaumekln/faster-whisper/tree/master) version of [langtech-veu/whisper-bsc-large-v3-cat](https://huggingface.co/langtech-veu/whisper-bsc-large-v3-cat) suitable for Automatic Speech Recognition in Catalan.
|
| 36 |
|
| 37 |
## Model Description
|
| 38 |
|
| 39 |
+
The "langtech-veu/faster-whisper-bsc-large-v3-cat" is the result of converting the [langtech-veu/whisper-bsc-large-v3-cat](https://huggingface.co/langtech-veu/whisper-bsc-large-v3-cat) into a lighter model using a Python module called [faster-whisper](https://github.com/guillaumekln/faster-whisper/tree/master).
|
| 40 |
|
| 41 |
## Intended Uses and Limitations
|
| 42 |
|
|
|
|
| 70 |
```python
|
| 71 |
from faster_whisper import WhisperModel
|
| 72 |
|
| 73 |
+
model_size = "langtech-veu/faster-whisper-bsc-large-v3-cat"
|
| 74 |
|
| 75 |
# Run on GPU with FP16
|
| 76 |
model = WhisperModel(model_size, device="cuda", compute_type="float16")
|
|
|
|
| 95 |
This model is not a direct result of training. It is a conversion of a [Whisper](https://huggingface.co/openai/whisper-large-v3) model using [faster-whisper](https://github.com/guillaumekln/faster-whisper/tree/master). The procedure to create the model is as follows:
|
| 96 |
|
| 97 |
```bash
|
| 98 |
+
ct2-transformers-converter --model langtech-veu/whisper-bsc-large-v3-cat
|
| 99 |
+
--output_dir faster-whisper-bsc-large-v3-cat
|
| 100 |
--copy_files preprocessor_config.json
|
| 101 |
--quantization float16
|
| 102 |
```
|
|
|
|
| 105 |
If this model contributes to your research, please cite the work:
|
| 106 |
```
|
| 107 |
@misc{takanori2025whisperbsclarge3cat,
|
| 108 |
+
title={Acoustic Model in Catalan: whisper-bsc-large-v3-cat.},
|
| 109 |
author={Sanchez Shiromizu, Lucas Takanori; Hernandez Mena, Carlos Daniel; Messaoudi, Abir; España i Bonet, Cristina; Cortada Garcia, Marti},
|
| 110 |
organization={Barcelona Supercomputing Center},
|
| 111 |
+
url={https://huggingface.co/langtech-veu/faster-whisper-bsc-large-v3-cat},
|
| 112 |
year={2025}
|
| 113 |
}
|
| 114 |
```
|