---
base_model: IndexTeam/Index-Translate-2B
library_name: mlx
license: apache-2.0
pipeline_tag: translation
tags:
- mlx
- translation
- bf16
---

# Index-Translate-2B-MLX-bf16

Unquantized MLX (bfloat16) export of [IndexTeam/Index-Translate-2B](https://huggingface.co/IndexTeam/Index-Translate-2B) — the same weights in
MLX's safetensors layout. Re-encoded only; no retraining, no fine-tuning.

Text translation only: the upstream checkpoint's vision tower and MTP head are not included.
Runs with [mlx-lm](https://github.com/ml-explore/mlx-lm) on Apple Silicon; not a llama.cpp / GGUF model.

## Files

| file | bytes | sha256 |
|---|---:|---|
| `model.safetensors` | 3,763,691,809 | `d58283244c9691eb68f39a91a0189b0f03469d66f1d951952a2fca4fd97eb43d` |

Plus `config.json`, `tokenizer.json`, `tokenizer_config.json`, `vocab.json`, `merges.txt`,
`generation_config.json` (copied from upstream), and `MANIFEST.json` (conversion record).

## Attribution

Weights: [IndexTeam/Index-Translate-2B](https://huggingface.co/IndexTeam/Index-Translate-2B) (apache-2.0) · [family repo](https://github.com/bilibili/Index-Translate).
This release only re-encodes the upstream weights; no new rights are claimed.