Hugging Face's logo

avemio
/
German-RAG-WHISPER-LARGE-v3-TURBO-HESSIAN-AI

Automatic Speech Recognition
Transformers
Safetensors
German
whisper
Generated from Trainer
Inference Endpoints
Model card Files Community
avemio-digital commited on
Commit
e497c89
·
verified ·
1 Parent(s): d35dc49

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +7 -7
README.md CHANGED
@@ -31,19 +31,19 @@ model-index:
31
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
32
  should probably proofread and complete it, then remove this comment. -->
33
 
34
- # Whisper Large -v3 Turbo German - GRAG
35
 
36
  This model is a fine-tuned version of [openai/whisper-large-v3-turbo](https://huggingface.co/openai/whisper-large-v3-turbo) on a carefully curated dataset.
37
 
38
 
39
  ## Evaluations - Word error rate
40
 
41
- | Dataset | openai-whisper-large-v3-turbo | openai-whisper-large-v3 | primeline-whisper-large-v3-german | nyrahealth-CrisperWhisper (large)| primeline-whisper-large-v3-turbo-german |
42
- |-------------------------------------|-------------------------------|-------------------------|-----------------------------------|---------------------------|-----------------------------------------|
43
- | Tuda-De | 8.300 | 7.884 | 7.711 | **5.148** | 6.441 |
44
- | common_voice_19_0 | 3.849 | 3.484 | 3.215 | **1.927** | 3.200 |
45
- | multilingual librispeech | 3.203 | 2.832 | 2.129 | 2.815 | **2.070** |
46
- | All | 3.649 | 3.279 | 2.734 | 2.662 | **2.628** |
47
 
48
  The data and code for evaluations are available [here](https://huggingface.co/datasets/flozi00/asr-german-mixed-evals)
49
 
 
31
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
32
  should probably proofread and complete it, then remove this comment. -->
33
 
34
+ # GRAG-WHISPER-LARGE-v3-TURBO
35
 
36
  This model is a fine-tuned version of [openai/whisper-large-v3-turbo](https://huggingface.co/openai/whisper-large-v3-turbo) on a carefully curated dataset.
37
 
38
 
39
  ## Evaluations - Word error rate
40
 
41
+ | Dataset | openai-whisper-large-v3-turbo | GRAG-WHISPER-LARGE-v3-TURBO | primeline-whisper-large-v3-turbo-german |
42
+ |-------------------------------------|-------------------------------|-------------------------|-----------------------------------|
43
+ | Tuda-De | 8.195 | 6.360 | 6.441 |
44
+ | common_voice_19_0 | 3.839 | 3.249 | 3.217 |
45
+ | multilingual librispeech | 3.202 | 2.071 | 2.067 |
46
+ | All | 3.641 | 2.633 | 2.630 |
47
 
48
  The data and code for evaluations are available [here](https://huggingface.co/datasets/flozi00/asr-german-mixed-evals)
49