---
language:
- eu
license: cc-by-4.0
pipeline_tag: automatic-speech-recognition
tags:
- automatic-speech-recognition
- speech
- asr
- basque
- euskara
- nemo
- parakeet
- fastconformer
- tdt
base_model: nvidia/parakeet-tdt-0.6b-v3
datasets:
- asierhv/composite_corpus_eu_v2.1
model-index:
- name: parakeet-tdt-0.6b-v3-basque
  results:
  - task:
      type: automatic-speech-recognition
      name: Automatic Speech Recognition
    dataset:
      type: custom
      name: Composite Basque test splits (CV/Parliament/OSLR)
    metrics:
    - type: wer
      name: test_cv WER
      value: 6.92
    - type: wer
      name: test_parl WER
      value: 4.36
    - type: wer
      name: test_oslr WER
      value: 14.52
---

# Parakeet TDT 0.6B v3 — Basque (Euskara)

Basque ASR model obtained by fine-tuning [nvidia/parakeet-tdt-0.6b-v3](https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3) on [asierhv/composite_corpus_eu_v2.1](https://huggingface.co/datasets/asierhv/composite_corpus_eu_v2.1).

## What this model is

- Base architecture: FastConformer RNNT-TDT (Parakeet TDT 0.6B v3)
- Language: Basque (`eu`)
- Target task: Automatic speech recognition (16 kHz speech)
- Framework: NVIDIA NeMo
- Artifact format: `.nemo`

## Training data provenance

Fine-tuning used the composite Basque corpus:

- [asierhv/composite_corpus_eu_v2.1](https://huggingface.co/datasets/asierhv/composite_corpus_eu_v2.1)

This composite corpus includes Basque speech/transcript data from:

- Mozilla Common Voice (Basque)
- Basque Parliament
- OpenSLR (Basque)

Reference datasets:

- Mozilla Common Voice: https://commonvoice.mozilla.org/
- Basque Parliament corpus (as included in the composite dataset above)
- OpenSLR Basque resources: https://www.openslr.org/

## Evaluation summary

WER was computed on held-out test splits created from the same composite source family.

| Split | Baseline WER (base model) | Fine-tuned WER | Absolute gain |
|---|---:|---:|---:|
| `test_cv` | 108.47% | **6.92%** | +101.55 |
| `test_parl` | 107.61% | **4.36%** | +103.25 |
| `test_oslr` | 108.52% | **14.52%** | +94.00 |

Notes:

- Baseline is the original multilingual/English-oriented base model, which performs poorly on Basque (WER > 100% is expected).
- Fine-tuned metrics above are from the final selected checkpoint/model export.

## Quick start

### 1) Load from local `.nemo`

```python
import nemo.collections.asr as nemo_asr

asr_model = nemo_asr.models.ASRModel.restore_from("parakeet-tdt-0.6b-v3-basque.nemo")
asr_model = asr_model.cuda().eval()  # optional if GPU is available

pred = asr_model.transcribe(
    audio=["/path/to/audio.wav"],
    use_lhotse=False,
    batch_size=1,
    num_workers=0,
    verbose=False,
)

# pred may be list[str] or list[Hypothesis], depending on settings/version
if len(pred) and hasattr(pred[0], "text"):
    print(pred[0].text)
else:
    print(pred[0])
```

### 2) Load from Hugging Face repo

```python
import nemo.collections.asr as nemo_asr

asr_model = nemo_asr.models.ASRModel.from_pretrained(
    model_name="xezpeleta/parakeet-tdt-0.6b-v3-basque"
)

out = asr_model.transcribe(
    audio=["/path/to/audio.wav"],
    use_lhotse=False,
    batch_size=1,
    num_workers=0,
    verbose=False,
)

print(out[0].text if hasattr(out[0], "text") else out[0])
```

### 3) Batch transcription snippet

```python
audio_paths = [
    "/path/a.wav",
    "/path/b.wav",
    "/path/c.wav",
]

outs = asr_model.transcribe(
    audio=audio_paths,
    use_lhotse=False,
    batch_size=16,
    num_workers=4,
    verbose=True,
)

texts = [o.text if hasattr(o, "text") else o for o in outs]
for path, txt in zip(audio_paths, texts):
    print(path, "=>", txt)
```

## Recommended runtime settings

- Use 16 kHz mono WAV when possible.
- For NeMo 2.7 environments, use `use_lhotse=False` in `transcribe(...)` if you observe inference/runtime issues in constrained container setups.
- Increase `batch_size` gradually to match available GPU memory.

## Fine-tuning recipe (high-level)

- Full-model fine-tuning from `nvidia/parakeet-tdt-0.6b-v3`
- Optimizer: AdamW
- LR: `1e-4`
- Scheduler: CosineAnnealing with warmup
- Effective batch size: 64 (`batch_size=8`, gradient accumulation=8)
- Precision: BF16 mixed
- Hardware used: NVIDIA L40 (48 GB)

## Limitations

- Performance can degrade on strong code-switching, noisy far-field audio, or domains far from training data.
- This model is optimized for Basque ASR; behavior on other languages is not the target.
- Proper text normalization and punctuation/casing post-processing may still be needed for production use.

## Citation and acknowledgements

If you use this model, please cite/credit:

1. Base model: [nvidia/parakeet-tdt-0.6b-v3](https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3)
2. Training dataset: [asierhv/composite_corpus_eu_v2.1](https://huggingface.co/datasets/asierhv/composite_corpus_eu_v2.1)
3. Underlying source collections: Mozilla Common Voice, Basque Parliament corpus, OpenSLR Basque resources

## License

This derivative model follows the base model and dataset terms. Keep attribution and license obligations from all upstream assets.