Role positioning:

As a large model training engineer, you will be at the forefront of Web3 and AI technology integration, focusing on the training and optimization of large language models (LLM). You will work with cross-functional teams to drive innovative applications of models in the Web3 field. If you are passionate about artificial intelligence, good at solving complex problems, and eager to grow in a creative and technically leadership environment, this will be your ideal position!

-Job Responsibilities:

• Model training and optimization: Design and implement large language model training strategies, including supervised fine-tuning (SFT), reinforcement learning (such as GRPO, PPO) and other methods, to enhance the model's intelligence in the Web3 field.

• Data processing and generation: Build a high-quality training dataset, perform data distillation and long-short chain of thought (Long&Short Chain of Thought, CoT) data generation, and ensure that the model has strong inference abilities.

• Model architecture and evaluation: Explore and apply advanced model architectures such as expert mixing (MoE), develop model evaluation frameworks and metrics, and continuously optimize model performance.

• Distributed training and deployment: Develop and maintain distributed training schemes for models to ensure efficient training and stable deployment.

• Technological frontier exploration: track the research dynamics in the AI field, such as OpenAiGPT-4.5, DeepSeek-R1, etc., and promote technical innovation and application in actual business.

-Position requirements:

1. Educational background: Bachelor's degree, master's degree or doctorate in Computer Science, Artificial Intelligence, Machine Learning or related fields.

2. Technical skills:

• Proficient in Transformer architecture, familiar with TransformerReinforcement Learning (TRL), PyTorch or TensorFlow deep learning-based learning frameworks, etc.

• Have large language model fine-tuning experience, familiar with reasoning-oriented reinforcement learning (Reasoning-Oriented Reinforcement Learning, RORL) technology,

• Familiar with distributed training frameworks, with practical experience in model parallelism, Flash Attention, LoRA, etc.

3. Engineering capability:

• Proficient in Python, Go, etc. programming language, with good coding style and software engineering practical experience.

• Familiar with model serving technologies such as Triton, vLLM, TGI, etc., those with inference optimization experience are preferred.

4. Research ability:

• Able to read and implement cutting-edge papers, write technical reports or blogs.

• Priority will be given to those with papers published at top conferences (such as NeurIPS, ICLR, ICML, ACL) or open-source project contributors.

5. Soft skills:

• Have excellent team collaboration and communication skills, and be able to efficiently collaborate with cross-functional teams.

• In-depth understanding of open-source AI communities, with preference for relevant project contributors.