Infrastructure
Models & Tools
OLAB provides open-source and internal model inference, containerized GPU compute, and industrial-scale HPC resources for research at NYU Langone.
Models
llama3-3-70b-chatAvailableMeta Llama 3.3 70B — general-purpose chat and instruction following.
llama3-3-70B-DSR1AvailableMeta Llama 3.3 70B with DeepSeek R1 distillation — enhanced reasoning.
NYUTronComing soonNYU Langone clinical language model — trained on clinical notes.
Lang-OneComing soonInternal clinical language model for healthcare applications.
Access available models through the inference platform. Requires NYU Langone network access.
Tools & Resources
LLM Inference Service
OpenAI-compatible, high-performance inference on H100 GPUs. Includes GUI and API access — free for internal NYU Langone research.
View docs →Enroot Container Guide
Run containerized GPU workloads on HPC clusters with NVIDIA Enroot and Slurm integration.
View guide →Compute Resources
Industrial-scale GPU infrastructure — A100, H100, and GH nodes through our NVIDIA and HPE partnerships.
View specs →