Downloads · 30 days
4
1% of all-time downloads
clarityIPA-wav2vec2-base-v1.1
SC1R24/clarityIPA-wav2vec2-base-v1.1
clarityIPA-wav2vec2-base-v1.1 is a automatic speech recognition model from SC1R24. Use it when you need speech turned into text. It is set up for transformers.
Finetuned wav2vec2-base for recognizing International Phonetic Alphabet (IPA) phonemes
All-time downloads
310
Public
Parameters
94.4M
378 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors378 MB · 100%
At a glance
- Task
- Automatic Speech Recognition
- Library
- transformers
- Model type
- wav2vec2
- Access
- Public
- Created
- Jan 24, 2026
- Updated
- Jan 26, 2026
- SHA
- eba8bc74
Model card
From the Hugging Face model README
Model Card for Model ID
Finetuned wav2vec2-base for recognizing International Phonetic Alphabet (IPA) phonemes
Training & Evaluation Metrics
| Step | Training Loss | Validation Loss | CER |
|---|---|---|---|
| 1000 | 2.332100 | 1.889206 | 0.529458 |
| 2000 | 0.689400 | 0.458334 | 0.111666 |
| 3000 | 0.539600 | 0.408651 | 0.105563 |
| 4000 | 0.442600 | 0.343108 | 0.091160 |
| 5000 | 0.400800 | 0.319820 | 0.085694 |
| 6000 | 0.357900 | 0.293541 | 0.081331 |
| 7000 | 0.326500 | 0.276007 | 0.080334 |
| 8000 | 0.303200 | 0.263738 | 0.079040 |
| 9000 | 0.264200 | 0.261444 | 0.076742 |
| 10000 | 0.263600 | 0.261173 | 0.073419 |
| 11000 | 0.230200 | 0.250614 | 0.074479 |
| 12000 | 0.217700 | 0.262788 | 0.071432 |
| 13000 | 0.228900 | 0.254010 | 0.072054 ← Selected checkpoint |
| 14000 | 0.203400 | 0.263125 | 0.071623 |
- Task
- Automatic Speech Recognition
- Library
- transformers
- Type
- wav2vec2
- Created
- Jan 24, 2026
- Updated
- Jan 26, 2026