---
language:
- tr
- hi
license: cc-by-nc-4.0
base_model: CohereLabs/tiny-aya-base
base_model_relation: adapter
tags:
- superseded
- checkpoint-artifact
- tinyaya
pipeline_tag: audio-to-audio
pretty_name: TinyAya Stage 2 (superseded run artifact)
---

# TinyAya Stage 2 — superseded run artifact

> ### ⚠️ Not a release. Use v0.3 instead.
> This repo holds a **single intermediate checkpoint** pushed automatically by
> the trainer during an early Stage-2 run. It has no evaluation, no checkpoint
> suite, and no guarantee about which step it represents.
>
> The released model is
> **[`tr-hi-s2st-v0.3`](https://huggingface.co/tiny-aya-translate/tr-hi-s2st-v0.3)**,
> which ships ~89 checkpoints, a full evaluation, and audio samples.

Contents are the standard composite layout — `peft_adapter/`, `projection.pt`,
`depth_decoder.pt`, `audio_heads.pt`, `text_embed.pt`, `metadata.json`. Kept
public for provenance rather than deleted.

Weights are LoRA derivatives of `CohereLabs/tiny-aya-base` (CC-BY-NC-4.0) and
inherit its **non-commercial** terms.

## Code

| repo | what it does |
|---|---|
| [`model`](https://github.com/tiny-aya-simultaneous-translation/model) | Stage-2 training, evaluation harness and TPU launch tooling |

## Project

**TinyAya Stage 2** — Turkish⇄Hindi speech-to-speech translation with a text
inner-monologue: a LoRA-adapted Cohere2 backbone driving a **frozen** Moshi depth
decoder over Mimi codes.

The v0.3 run covered **76,250 steps / 2.07 epochs** on a Cloud TPU v6e-16
(best val composite **2.8199** @ step 76,000). Read honestly: the text
inner-monologue **learns to translate** (free-run chrF++ ~25.7 / 25.1), while
**intelligible audio synthesis remains the frontier** (ASR-chrF++ 3.7 / 9.6
against a 92.1 / 86.6 ground-truth-audio ceiling) — bounded by the frozen depth
decoder, not by translation understanding.

- **Results:** [v0.3 evaluation report](https://github.com/tiny-aya-simultaneous-translation/model/blob/main/docs/v0.3-eval-report.md)
- **Training run:** [W&B `xzcb60bl`](https://wandb.ai/cataluna84/tinyaya-stage2-tpu/runs/xzcb60bl) · [emergence report](https://wandb.ai/cataluna84/tinyaya-stage2-tpu/reports/TinyAya-v0.3-Emergence-and-Data-Efficiency--VmlldzoxNzU1OTU1NQ==)
- **Blog:** [Adapting Moshi for Low-Resource Speech Translation](https://labscommunity.cohere.com/blog/2026/adapting-moshi-low-resource-speech-translation/)

Compute for the v0.3 run was provided by **Google's TPU Research Cloud (TRC)**.