Responsibilities:
1. Convert the current drug discovery data into data that can be used for prospective model testing (benchmarks) and and model training
2. Finetune and post-train open source models such as Qwen, LLaMA, DeepSeek and others for AI for science tasks
3. Integrate expert post-trained models into the current drug discovery workflows, interfaces and tools
Qualifications:
1. Over 2-3 years experience in LLM post-training experience.
2. Master degree or above in computer science, AI related major
3. Demonstrated experience in setting up the environment, sequence, and data for post-training of open source models; also experience in benchmarking the performance of the post-trained models
4. Increasing the efficiency of training to minimize training cost
5. Experience in post training in major LLM vendors is preferred
