This repository contains a full merged model for bidirectional Luganda-English speech-to-speech translation. It was created by merging the LoRA adapter yigagilbert/stepaudio2-mini-luganda-english-bidirectional-s2st-lora into stepfun-ai/Step-Audio-2-mini.

The separate adapter-only repository should remain available for users who prefer PEFT loading or want the smaller adapter artifact. This full repository is intended for simpler deployment and inference where loading a single model repo is preferable.

Research and development for Luganda-English speech translation in both directions:

Validate outputs with native speakers before production or high-stakes use.

The model was trained from yigagilbert/luganda-english-cleaned-v1-split by expanding each aligned speech pair into two direction-conditioned supervised examples. Each example uses a direction-specific system prompt so that one adapter can learn both translation directions.

The checkpoint used for this merge was evaluated on 200 held-out validation examples per direction. The metrics below were generated with the adapter-loaded fine-tuned model before merge; the merged model contains the same adapted weights and is expected to match these results aside from normal deterministic or runtime differences. Re-run evaluation directly on this repository before a strict release if exact reproducibility is required.

Speech metrics were computed on 20 audio-aligned samples per direction using WavLM-large SpeechBERTScore-style similarity and MFCC+DTW MCD. These are supporting audio-channel diagnostics, not standalone translation-quality measures.

The bidirectional model does not recover unrelated multilingual speech translation ability on small FLEURS probes: Spanish -> English BLEU is 0.293 and Japanese -> English BLEU is 0.163 on 20 samples each. However, it does recover target-script compliance on Luganda -> Chinese and Luganda -> Japanese prompt probes (100% target-script hit rate in both cases). The semantic quality of those composition outputs is poor, with roundtrip BLEU below 1, so these probes should not be read as successful Luganda -> Chinese or Luganda -> Japanese translation.

This model demonstrates that one direction-conditioned LoRA adaptation of Step-Audio-2-mini can learn both Luganda -> English and English -> Luganda speech translation. Compared with the earlier Luganda-to-English-only adapter, it adds English-to-Luganda capability but is weaker on Luganda-to-English. The bidirectional run was shorter (9,000 steps) than the focused one-direction run, so the Luganda-to-English gap combines multi-task interference and training duration.

If this repository includes token2wav/, those assets are provided to support waveform synthesis from generated audio tokens. Some inference clients may still use the official Step-Audio2 runtime code for token-to-waveform conversion.

The training code and adapter metadata are Apache-2.0. Because this merged repository contains base-model weights, users must also comply with the base model license and any dataset licensing constraints.