Contact person: Chinese-English AI Translation Training

Phone:Show

Email:Show

Location: Wan Chai, Lao People's Democratic Republic

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"I’m building an AI system that turns Chinese business emails, proposals and internal memos into clear, semi-formal English. The base model is in place; what it needs now is carefully prepared data and targeted evaluation so it learns the exact tone and terminology of day-to-day corporate communication.

Your core mission is to assemble and refine a parallel corpus that captures this setting. I’ll share a short tone brief; you’ll expand it into practical annotation guidelines, collect or create high-quality sentence-aligned pairs, and make sure every entry is deduplicated, UTF-8 clean and ready for fine-tuning.

Once the data is ready, run a small evaluation pass—BLEU with SacreBLEU (or a comparable metric) plus a concise human review—to measure progress and surface consistent error patterns. Wrap up with a brief report that flags weaknesses, recommends the next training tweak and documents the results.

Deliverables
• 10 000+ Chinese → English business-communication pairs in CSV or TMX
• Expanded guideline highlighting semi-formal tone, idioms and company-specific terms
• Evaluation script (Python) and results sheet
• Two-page findings & recommendation report

Acceptance criteria: 95 % or better human-verified accuracy on a random sample and a ≥3-point BLEU gain over the current baseline.

If you’ve worked with Hugging Face datasets, OpenAI fine-tuning, or corporate translation memories, mention it so I can see how your background fits." (client-provided description)