Sample Page Title

Tencent Hunyuan researchers have launched HY-MT1.5, a multilingual machine translation household that targets each cellular units and cloud methods with the identical coaching recipe and metrics. HY-MT1.5 consists of two translation fashions, HY-MT1.5-1.8B and HY-MT1.5-7B, helps mutual translation throughout 33 languages with 5 ethnic and dialect variations, and is obtainable on GitHub and Hugging Face beneath open weights.

Mannequin household and deployment targets

HY-MT1.5-7B is an upgraded model of the WMT25 championship system Hunyuan-MT-7B. It’s optimized for explanatory translation and combined language eventualities, and provides native help for terminology intervention, contextual translation and formatted translation.

HY-MT1.5-1.8B is the compact variant. It has lower than one third the parameters of HY-MT1.5-7B however delivers comparable translation efficiency within the reported benchmarks. After quantization, the 1.8B mannequin can run on edge units and help actual time translation.

The quantized HY-MT1.5-1.8B operates on units with about 1 GB of reminiscence and reaches a mean response time of about 0.18 seconds for Chinese language inputs of round 50 tokens, whereas surpassing mainstream industrial translation APIs in high quality. HY-MT1.5-7B targets server and excessive finish edge deployment, the place latency round 0.45 seconds is appropriate in alternate for increased high quality.

Holistic coaching framework

The analysis staff defines HY-MT1.5 as a translation particular language mannequin skilled with a multi stage pipeline.

The pipeline has 5 major parts:

1. Normal pre coaching: The bottom mannequin is first pre-trained on massive scale multilingual textual content with a language modeling goal. This builds shared representations throughout languages.
2. MT oriented pre coaching: The mannequin is then uncovered to parallel corpora and translation oriented aims. This step aligns the technology distribution with actual translation duties reasonably than open ended textual content technology.
3. Supervised high-quality tuning: Top quality sentence and doc stage parallel information is used to high-quality tune the mannequin with supervised loss. This stage sharpens literal correctness, area protection and path particular habits, akin to ZH to EN versus EN to ZH.
4. On coverage distillation from 7B to 1.8B: HY-MT1.5-7B is used as a instructor for HY-MT1.5-1.8B. The analysis staff collects about 1 million monolingual prompts throughout the 33 languages, runs them by the instructor and makes use of reverse Kullback Leibler divergence on the coed rollouts to match the instructor distribution. This yields a 1.8B pupil that inherits many of the 7B mannequin’s translation habits with a lot decrease price.
5. Reinforcement studying with rubrics based mostly analysis: Within the remaining stage, each fashions are optimized with a bunch relative coverage optimization type algorithm and a rubrics based mostly reward mannequin. Human reviewers rating translations on a number of axes akin to accuracy, fluency, idiomaticity and cultural appropriateness. The reward mannequin distills these scores and guides the coverage replace.

This pipeline is restricted to machine translation. It differs from chat oriented LLM coaching by combining translation centric supervised information, on coverage distillation inside the translation area and RL tuned with high-quality grained translation rubrics.

Benchmark outcomes in opposition to open and industrial methods

HY-MT1.5 is evaluated on Flores 200, WMT25 and a Mandarin to minority language benchmark utilizing XCOMET-XXL and CometKiwi.

Key outcomes from the above Desk within the report:

• On Flores 200, HY-MT1.5-7B reaches XCOMET-XXL scores of 0.8690 for ZH to XX, 0.9093 for EN to XX and 0.8098 for XX to XX. It outperforms translation specialised fashions akin to iFLYTEK Translator and Doubao Translator and matches or exceeds medium sized normal fashions like Qwen3-235B-A22B.
• On WMT25, HY-MT1.5-7B reaches XCOMET-XXL 0.6159. That is about 0.065 increased than Gemini 3.0 Professional and considerably above translation oriented fashions akin to Seed-X-PPO-7B and Tower-Plus-72B. HY-MT1.5-1.8B scores 0.5308, which nonetheless exceeds many medium sized normal fashions and translation methods.
• On Mandarin to minority language pairs, HY-MT1.5-7B achieves 0.6174 in XCOMET-XXL, increased than all baselines together with Gemini 3.0 Professional. The 1.8B variant reaches 0.5806 and nonetheless surpasses a number of very massive fashions like DeepSeek-V3.2.

In human analysis on a 0 to 4 scale for Chinese language to English and English to Chinese language, HY-MT1.5-1.8B achieves a mean rating of two.74, which is increased than Baidu, iFLYTEK, Doubao, Microsoft and Google translator methods beneath the identical protocol.

Sensible options for product use

The fashions expose three immediate pushed capabilities that matter in manufacturing methods:

1. Terminology intervention: A immediate template helps you to inject time period mappings akin to “混元珠 → Chaos Pearl”. With out the mapping, the mannequin outputs an ambiguous transliteration. With the mapping, it enforces a constant area particular time period. That is essential for authorized, medical or model constrained content material.
2. Context conscious translation: A second template accepts a context block plus the sentence to translate. The report exhibits the phrase “pilot” misinterpreted as an individual when context is absent. When a paragraph about TV collection is added, the mannequin accurately interprets “pilot” as an episode.
3. Format preserving translation: A 3rd template wraps the supply in <supply> tags and marks spans with <sn> tags. The instruction forces the mannequin to maintain tags and output inside <goal> tags. This permits HTML or XML like textual content to outlive translation with construction preserved.

These are applied as immediate codecs, so they’re obtainable even if you name the general public weights by commonplace LLM stacks.

Quantization and edge deployment

HY-MT1.5-1.8B is evaluated with FP8 and Int4 put up coaching quantization utilizing GPTQ.

The above Desk 4 exhibits:

• FP8 retains XCOMET-XXL scores very near the complete precision mannequin, for instance 0.8379 versus 0.8361 for ZH to XX.
• Int4 reduces measurement additional however introduces clear high quality drops on Flores 200.

On Hugging Face, Tencent publishes each FP8 and GPTQ Int4 variants for HY-MT1.5-1.8B and HY-MT1.5-7B, together with GGUF variations for native inference stacks. Quantization is the mechanism that allows the reported 1 GB reminiscence deployment and low latency on shopper {hardware}.

Key Takeaways

• HY-MT1.5 is a 2 mannequin translation household, HY-MT1.5-1.8B and HY-MT1.5-7B, supporting mutual translation throughout 33 languages plus 5 dialect or variant types, launched with open weights on GitHub and Hugging Face.
• HY-MT1.5-1.8B is a distillation based mostly edge mannequin that runs on about 1 GB reminiscence with round 0.18 seconds latency for 50 token Chinese language inputs, whereas attaining trade main efficiency amongst fashions of comparable measurement and surpassing most industrial translation APIs.
• HY-MT1.5-7B is an upgraded WMT25 champion system that reaches roughly 95 p.c of Gemini 3.0 Professional on Flores 200 and surpasses it on WMT25 and Mandarin minority benchmarks, competing with a lot bigger open and closed fashions.
• Each fashions are skilled with a holistic translation particular pipeline that mixes normal and MT oriented pre coaching, supervised high-quality tuning, on coverage distillation and reinforcement studying guided by rubric based mostly human analysis, which is essential to their high quality and effectivity commerce off.
• HY-MT1.5 exposes manufacturing oriented options by prompts, together with terminology intervention, context conscious translation and format preserving translation, and ships FP8, Int4 and GGUF variants so groups can deploy on units or servers with commonplace LLM stacks.

Try the Paper, Mannequin Weights on HF and GitHub Repo. Additionally, be happy to comply with us on Twitter and don’t overlook to hitch our 100k+ ML SubReddit and Subscribe to our Publication. Wait! are you on telegram? now you’ll be able to be a part of us on telegram as effectively.

Michal Sutter is a knowledge science skilled with a Grasp of Science in Knowledge Science from the College of Padova. With a strong basis in statistical evaluation, machine studying, and information engineering, Michal excels at reworking complicated datasets into actionable insights.