China, Shanghai
LLM Reinforcement Learning Framework Intern - 2027
What the posting asks for
- Names
- PyTorch, Python, veRL
- Doctorate
- Not mentioned
Read out of the employer's own description. Absence means the posting does not say, not that the answer is no.
The role
We are now looking for exceptionally dedicated individuals to join our team as LLM Reinforcement Learning Framework Intern!
At NVIDIA, we are renowned for our continual progress in the fields of graphics and parallel computing. By joining our Architecture group, you will be part of a dynamic team that is spearheading the AI revolution. This remarkable opportunity will allow you to work on innovative projects that extend the frontiers of GPU technology and AI.
What you'll be doing:
- Help NVIDIA Nemo RL to be efficiently working under large scale post train workload.
- Adopt SOTA RL algorithms to Nemo RL.
- Improve RL rollout performance.
What we need to see:
- Having experience with VeRL, Miles, Slime, or Nemo RL is highly beneficial.
- Strong proficiency in programming languages such as Python, C++, or CUDA.
- Ambitious approach with a proven track record of taking initiative and delivering results.
- Experience with PyTorch, RLHF (Reinforcement Learning with Human Feedback), and LLM training frameworks like FSDP/Megatron LLM is a plus.
- Familiarity with agentic RL and inference engine development is preferred.
- Pursuing A degree in Computer Science, Engineering, or a related field
Published by NVIDIA on their own careers page and reproduced here unedited. Read it at NVIDIA.
Apply at NVIDIA → Applications go directly to NVIDIA. This board does not sit in between, take a fee from you, or see your application.
What this listing does not tell you
Listed 7 days, which is recent for this board. Of the 85 post-training and RL roles this board has watched from listing to removal, 4% were gone from their employer's careers page by day 7, and the median came down after 104 days. That is a description of other listings that have already ended, not a prediction about this one: this board records when a listing disappears, never why, and a posting still up is not on a clock it can see.
Pay is not confirmed for this role. Check the employer’s posting for current compensation. Explore published bands from other roles →
NVIDIA has 33 roles open on this board, 11 of them in post-training and RL. 1 of those does publish a band, which you can compare; pay can differ by role and location.
Free. One email on Thursdays.
More roles like this
Matched by discipline, title, listed location and work arrangement.
Same discipline: Post-Training & RL · Both have internship titles · Shared listed location
Sunnyvale$65/hr
Same discipline: Post-Training & RL · Both have internship titles
Sunnyvale$65/hr
Same discipline: Post-Training & RL · Both have internship titles
US, CA, Santa Clara
Same discipline: Post-Training & RL · Both have internship titles
Zurich, Switzerland
Same discipline: Post-Training & RL · Both have internship titles