---
title: Index-Translate-omni-35B-FP8
canonical_url: "https://www.modelscope.cn/models/IndexTeam/Index-Translate-omni-35B-FP8"
md_url: "https://www.modelscope.cn/models/IndexTeam/Index-Translate-omni-35B-FP8.md"
repository: IndexTeam/Index-Translate-omni-35B-FP8
last_updated: 2026-11-06
license: apache-2.0
model_type:
  - qwen3_5_moe
architectures:
  - Qwen3_5MoeForConditionalGeneration
parameters: 35.1B
tensor_type:
  - BF16
  - F8_E4M3
  - F32
library_name:
  - safetensors
language:
  - zh
  - en
downloads: 28
stars: 0
tags:
  - index-translate
  - multimodal
  - translation
  - image-to-text
  - automatic-speech-recognition
  - fp8
---

# Index-Translate-omni-35B-FP8

> Index-Translate-omni-35B-FP8 - IndexTeam 在 ModelScope 开源的模型。Index-Translate-omni-35B-FP8

IndexTeam/Index-Translate-omni-35B-FP8 是 ModelScope 魔搭社区上的 35.1B 参数机器学习模型，采用 apache-2.0 许可。

- **Repository**: IndexTeam/Index-Translate-omni-35B-FP8
- **License**: apache-2.0
- **Parameters**: 35.1B
- **Tags**: index-translate, multimodal, translation, image-to-text, automatic-speech-recognition, fp8
- **Downloads**: 28
- **Stars**: 0
- **Last updated**: 2026-11-06

Source: https://www.modelscope.cn/models/IndexTeam/Index-Translate-omni-35B-FP8

---

# Index-Translate-omni-35B-FP8

Quantization source: [IndexTeam/Index-Translate-omni-35B on ModelScope](https://modelscope.cn/models/IndexTeam/Index-Translate-omni-35B). This export uses the complete public source snapshot, with 74 files pinned by SHA-256, including 693 language tensors, 333 visual tensors and 529 shared-native audio tensors. Source manifest SHA-256: `d3c6a02b3eb99d843f1aff5b3066d926d8c99a9fb359dd5cdd81945309c87f05`.

Built on the Index-Translate family, with multilingual text translation, terminology and formatting constraints, image translation, speech recognition, and bidirectional Chinese–English speech translation.

## Original Model Reference

Source: [Index-Translate-Omni model card](https://modelscope.cn/models/IndexTeam/Index-Translate-omni-35B). The tables below reproduce the upstream results for the original models. They are not new benchmark results for this quantized export; see the format-specific verification below for the checks actually performed here.

## Text Translation

| Model | FLORES COMET22 ↑ | WMT24++ COMET22 ↑ | WMT26 Judge ↑ | MEME ↑ |
|---|---:|---:|---:|---:|
| Index-Translate 2B | 0.8655 | 0.8489 | 60.26 | 0.6443 |
| Index-Translate-Omni 2B | 0.8655 | 0.8489 | 60.26 | 0.6443 |
| Index-Translate 9B | 0.8789 | 0.8601 | 75.35 | 0.7387 |
| Index-Translate-Omni 9B | 0.8789 | 0.8601 | 75.35 | 0.7387 |
| Index-Translate 35BA3B (preview) | 0.8794 | 0.8586 | 76.76 | 0.7405 |
| Index-Translate-Omni 35BA3B | 0.8795 | 0.8593 | 78.01 | 0.7405 |

### Instruction Following

| Model | instTrans Quality / IFscore ↑ | instTrans_minor Quality / IFscore ↑ / Off-target ↓ | IFMTBench XCOMET-XXL / IFscore ↑ |
|---|---:|---:|---:|---:|
| Index-Translate 2B | 0.5391 / 0.7569 | 0.3050 / 0.6586 / 3.97% | 0.7712 / 0.7584 |
| Index-Translate-Omni 2B | 0.5391 / 0.7569 | 0.3050 / 0.6586 / 3.97% | 0.7712 / 0.7584 |
| Index-Translate 9B | 0.6771 / 0.8209 | 0.5222 / 0.7725 / 3.47% | 0.7957 / 0.8760 |
| Index-Translate-Omni 9B | 0.6771 / 0.8209 | 0.5222 / 0.7725 / 3.47% | 0.7957 / 0.8760 |
| Index-Translate 35BA3B (preview) | 0.6901 / 0.8336 | 0.5151 / 0.7715 / 4.05% | 0.7926 / 0.8991 |
| Index-Translate-Omni 35BA3B | 0.6867 / 0.8371 | 0.5163 / 0.7843 / 3.62% | 0.7934 / 0.9070 |

Original Index-Translate figures come from the [published benchmarks](https://huggingface.co/IndexTeam/Index-Translate-9B#evaluation). The original 35BA3B reference is the published preview checkpoint.

## Speech

CER is a percentage (lower is better). BLEU and chrF2 are higher-is-better. Translation cells show BLEU / chrF2 / target CER.

| Model | Chinese ASR CER ↓ | English ASR CER ↓ | English → Chinese | Chinese → English |
|---|---:|---:|---:|---:|
| Index-Translate-Omni 2B | 13.8113 | 34.2560 | 36.982 / 30.357 / 43.3452 | 30.747 / 51.152 / 48.0670 |
| Index-Translate-Omni 9B | 17.8917 | 14.4097 | 43.507 / 36.312 / 41.3767 | 34.656 / 55.376 / 46.3720 |
| Index-Translate-Omni 35BA3B | 13.0053 | 15.4950 | 40.074 / 32.862 / 44.8097 | 32.411 / 53.452 / 46.1700 |

35BA3B uses one shared audio encoder and connector for all four tasks, with repetition penalty 1.1 for English ASR and the original decoding for the other tasks. Results use the same 30 fixed samples per task. One English → Chinese case reached the 256-token limit with no target translation and remains in the score. Sample counts differ across model sizes.

## Image Translation

COMET22 × 100 (higher is better). Each Omni model is compared with its Qwen3.5 baseline on paired evaluated subsets.

| Model | MMTIT ↑ | FLORES-rendered ↑ | Vistra ↑ |
|---|---:|---:|---:|
| Qwen3.5-2B | 52.164 | 50.187 | 55.326 |
| Index-Translate-Omni 2B | 65.809 | 53.612 | 65.596 |
| Qwen3.5-9B | 73.814 | 67.804 | 68.963 |
| Index-Translate-Omni 9B | 77.604 | 68.438 | 71.815 |
| Qwen3.5-35B-A3B | 75.602 | 70.005 | 71.067 |
| Index-Translate-Omni 35BA3B | 77.011 | 69.891 | 71.482 |

Image scores are retained from the evaluated image configurations and cover the evaluated subsets.


## FP8 Export

Quantized omni translation model with language and vision weights, original BF16 audio modules, and a portable runtime.

- **Language:** FP8 weights with per-channel scales and dynamic per-token FP8 activations (`FP8_DYNAMIC`).
- **Vision:** original FP32 tensors, byte-identical to the assembled source.
- **Audio:** original BF16 audio encoder and connector, with a portable runtime.
- `lm_head` and token embeddings are not quantized. Linear-attention projections remain in the quantization recipe.

The included `JointModel` explicitly requests `dequantize=True` for compressed-tensors. This compatibility path uses BF16 compute after loading; it is not a claim of FP8 kernel acceleration.

The public 35B shared-native audio encoder and decoder profile are preserved. English ASR uses repetition penalty 1.1; Chinese ASR uses the public request-local generation guard with `max_new_tokens=256`. Audio generation is limited to 256 new tokens.



## Loading

Download the entire repository, including `audio/` and `audio-tokenizer/`:

```bash
pip install -r requirements-joint.txt
python - <<'PY'
from joint_runtime import JointModel

model = JointModel('.', device_map='auto')
print(model.text_generate([{'role': 'user', 'content': 'Translate into English: 人工智能正在改变内容创作。'}]))
print(model.audio_generate('<audio>\nTranscribe the speech and translate it into English.', ['/path/to/audio.wav']))
PY
```

`JointModel` also supports `image_generate(messages)` with Transformers image-text messages. The runtime extracts only the required helpers from the supplied bound audio code; it does not require the original training repository. Keep the visual weights in FP32 and the audio weights in BF16.

## Verification

Saved-checkpoint reload and text generation passed. The portable runtime also passed text and image-description calls, plus real WAV speech-to-English, English ASR and Chinese ASR calls on an H100. Text perplexity is `38.0772` over `250` held-out tokens using the reference loading mode above. These are small smoke tests, not a comprehensive translation, vision or speech-quality benchmark.

The tensor audit checked 31070 quantized linear weights, 333 byte-identical visual tensors, 529 byte-identical audio tensors, and 263 byte-identical preserved language tensors. Exact outputs and limitations are in `validation.json`; file hashes are in `SHA256SUMS.txt`.

This repository does not include GGUF files; GGUF is a separate language/vision export without the end-to-end audio pipeline.

## Mirrors

- Hugging Face: https://huggingface.co/IndexTeam/Index-Translate-omni-35B-FP8
- ModelScope: https://modelscope.cn/models/IndexTeam/Index-Translate-omni-35B-FP8
