Automatic Speech Recognition
NeMo
ONNX
GGUF
English
parakeet
tdt
speaker-adaptation
english
tech-vocabulary
irish-english
dictation
fritter
Instructions to use Fritter-ASR/Fritter with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- NeMo
How to use Fritter-ASR/Fritter with NeMo:
import nemo.collections.asr as nemo_asr asr_model = nemo_asr.models.ASRModel.from_pretrained("Fritter-ASR/Fritter") transcriptions = asr_model.transcribe(["file.wav"]) - Notebooks
- Google Colab
- Kaggle
Card: Fritter v1.4 (r39-general)
Browse files
README.md
CHANGED
|
@@ -20,7 +20,7 @@ datasets:
|
|
| 20 |
- openslr/librispeech_asr
|
| 21 |
---
|
| 22 |
|
| 23 |
-
# Fritter v1.
|
| 24 |
|
| 25 |
> Formerly [`danielrosehill/Fritter`](https://huggingface.co/danielrosehill/Fritter), where Fritter was a proof of concept. It moved to the Fritter-ASR organisation in November 2026, and new releases are published here.
|
| 26 |
|
|
@@ -42,7 +42,7 @@ A full fine-tune of [`nvidia/parakeet-tdt-0.6b-v2`](https://huggingface.co/nvidi
|
|
| 42 |
speaker's read-aloud phone recordings (an Irish English speaker living in Israel), run `r11-general`, trained
|
| 43 |
2026-10-27. Output keeps v2's casing and punctuation; numbers are written in digits.
|
| 44 |
|
| 45 |
-
**Current version: v1.
|
| 46 |
|
| 47 |
**Try it in your browser:** [Fritter-ASR/Fritter-Demo](https://huggingface.co/spaces/Fritter-ASR/Fritter-Demo) runs the int8 ONNX build on your own device.
|
| 48 |
|
|
|
|
| 20 |
- openslr/librispeech_asr
|
| 21 |
---
|
| 22 |
|
| 23 |
+
# Fritter v1.4: Parakeet TDT 0.6B v2 for English dictation, no code-switching
|
| 24 |
|
| 25 |
> Formerly [`danielrosehill/Fritter`](https://huggingface.co/danielrosehill/Fritter), where Fritter was a proof of concept. It moved to the Fritter-ASR organisation in November 2026, and new releases are published here.
|
| 26 |
|
|
|
|
| 42 |
speaker's read-aloud phone recordings (an Irish English speaker living in Israel), run `r11-general`, trained
|
| 43 |
2026-10-27. Output keeps v2's casing and punctuation; numbers are written in digits.
|
| 44 |
|
| 45 |
+
**Current version: v1.4** (run `r39-general`, trained 2026-11-10): v1.3 continued on 359.3 min of new own takes plus a replay sample (takes up to 60 s, cost-capped batches). The weights and every build in this repo are v1.4. v1.1 to v1.4 have not been benchmarked, so the results below are v1.0's (run `r11-general`).
|
| 46 |
|
| 47 |
**Try it in your browser:** [Fritter-ASR/Fritter-Demo](https://huggingface.co/spaces/Fritter-ASR/Fritter-Demo) runs the int8 ONNX build on your own device.
|
| 48 |
|