---
language:
- ny
tags:
- automatic-speech-recognition
- whisper
- chichewa
- nyanja
- fleurs
license: apache-2.0
pipeline_tag: automatic-speech-recognition
base_model: openai/whisper-small
---

# Fleurs Synthetic Whisper Small Fleurs Synthetic Hours0P75 Checkpoint

## Summary

This repository contains a Whisper checkpoint for Chichewa/Nyanja
automatic speech recognition, fine-tuned from `openai/whisper-small`.

- Experiment type: `fleurs-synthetic`
- Base model: `openai/whisper-small`
- Training condition: `fleurs_synthetic_hours0p75`
- Release artifact: full fine-tuned checkpoint selected from the best training checkpoint

## Intended use

This checkpoint is intended for research and evaluation on Chichewa/Nyanja ASR.
It is not a production-ready speech system and should be validated carefully
before downstream use.

## How to use

This repository contains a full fine-tuned checkpoint. It can be loaded
directly with Transformers.

```python
from transformers import AutoModelForSpeechSeq2Seq, AutoProcessor

model_id = "ai4good-labyrinth/fleurs-synthetic-hours0p75-whisper-small-no-language"
processor = AutoProcessor.from_pretrained(model_id)
model = AutoModelForSpeechSeq2Seq.from_pretrained(model_id)
```

## Training data

- Training source: `FLEURS train + Prepared dataset`
- Evaluation source during training: `FLEURS dev`
- Train examples before duration filtering: `8243`
- Train examples after duration filtering: `8202`
- Dev examples before duration filtering: `311`
- Dev examples after duration filtering: `305`
- Duration filter used during training: `min_duration_seconds=0.0`, `max_duration_seconds=30.0`

## Training procedure

- Fine-tuning script: `experiments/whisper_finetune/finetune_whisper.py`
- Base model: `openai/whisper-small`
- Task: `transcribe`
- Language hint during training/evaluation: none, corresponding to `--language auto` in standalone evaluation
- Mixed precision: `no`
- Gradient checkpointing: `False`
- Selected checkpoint step: `6600`
- Selected checkpoint epoch: `6.43`

## Training-time dev selection

The best checkpoint was selected using trainer-side dev evaluation on the
duration-filtered FLEURS dev split.

- Dev WER: `0.5012`
- Dev CER: `0.1509`
- Dev loss: `0.7995`

These values come from the training pipeline and may differ slightly from
standalone post-hoc evaluation because the decoding path is not perfectly
identical.

## Evaluation protocol

Standalone evaluation is recommended for the final release. Filtered and
unfiltered results should be reported separately.

- Filtered evaluation: `min_duration_seconds=0`, `max_duration_seconds=30`
- Unfiltered evaluation: no duration constraint
- Decoding task: `transcribe`
- Language hint: `auto`

## Evaluation summary

| Dataset | Split | Setting | Num examples | WER | CER | Notes |
|---|---|---|---:|---:|---:|---|
| FLEURS | dev | filtered | 305 | 0.5063 | 0.1414 | Filtered to 30 seconds |
| FLEURS | dev | unfiltered | TBD | TBD | TBD | Standalone eval pending |
| FLEURS | test | filtered | 745 | 0.4965 | 0.1460 | Filtered to 30 seconds |
| FLEURS | test | unfiltered | TBD | TBD | TBD | Standalone eval pending |
| Zambezi | dev | filtered | TBD | TBD | TBD | Standalone eval pending |
| Zambezi | dev | unfiltered | TBD | TBD | TBD | Standalone eval pending |
| Zambezi | test | filtered | 427 | 0.6231 | 0.1545 | Filtered to 30 seconds |
| Zambezi | test | unfiltered | 428 | 0.6203 | 0.1536 | No duration filter |
## Files in this repository

- Model weights and config: repository root
- Processor/tokenizer files: repository root
- Evaluation JSON files: `eval/...`

## Known limitations

- Whisper does not provide an official Nyanja/Chichewa language token.
- Users must also comply with the upstream dataset licenses and any upstream model license obligations.
- Standalone evaluation and trainer-side evaluation can differ slightly even on the same split and duration filter.
- Cross-dataset results should be interpreted carefully because transcription conventions may differ across corpora.

## Citation

If you use this checkpoint, please cite:

- the Whisper paper
- the FLEURS dataset
- this repository

```bibtex
@misc{fleurs_synthetic_hours0p75_whisper_small_no_language_2026,
  title        = {Fleurs Synthetic Whisper Small Fleurs Synthetic Hours0P75 Checkpoint},
  author       = {AI4Good Labyrinth Team},
  year         = {2026},
  howpublished = {\url{https://huggingface.co/ai4good-labyrinth/fleurs-synthetic-hours0p75-whisper-small-no-language}},
  note         = {Whisper fine-tuning for Chichewa/Nyanja ASR}
}
```