Latest Content
DeepSeek R1 Distill Llama 8B icon

DeepSeek R1 Distill Llama 8B

NVIDIA
DeepSeek-R1-Distill-Llama-8B is a compact dense reasoning-focused language model developed by DeepSeek AI through distillation from the larger DeepSeek-R1 model. It features an 8B parameter transformer architecture with 32 layers, 32 attention heads, and a 4,096 hidden size, built on the Llama architecture and fine-tuned using reasoning traces generated by DeepSeek-R1. The model transfers advanced reasoning behaviors into a smaller efficient model optimized for mathematics, coding, and logical analysis, supporting a 128K token context window using extended Llama-3 style RoPE scaling.
TypeDense LLM
CapabilitiesText Generation, Instruction Following, Reasoning, Mathematical Reasoning+4 more
Paper/Blog
Paper|Blog|HF Model Card
LicenseMIT

Inference Instructions

Deploy and run this model on NVIDIA B200 GPUs using the command below. Copy the command to get started with inference.

CONSOLE
docker run -it --rm 
 --runtime=nvidia 
 --gpus all 
 --ipc=host 
 --shm-size=128g 
 -p 8000:8000 
 -v ~/.cache/huggingface:/root/.cache/huggingface 
 -e HF_TOKEN='YOUR_HF_TOKEN' 
 -e LD_LIBRARY_PATH='/usr/local/nvidia/lib64:/usr/local/nvidia/lib:/usr/lib/x86_64-linux-gnu' 
 vllm/vllm-openai:v0.15.0-cu130 
 deepseek-ai/DeepSeek-R1-Distill-Llama-8B 
  --tensor-parallel-size 8 
 --max-num-batched-tokens 65536 
 --max-model-len auto 
  --gpu-memory-utilization 0.95 
 --max-num-seqs 1024 
 --disable-log-requests 
 --trust-remote-code

Model Benchmarks

Each model was tested with a fixed input size and total token volume while increasing concurrency to measure serving performance under load.

ITL vs Concurrency

Time to First Token

Throughput Scaling

Total Tokens/sec vs Avg TTFT

Vultr Cloud GPU

NVIDIA HGX B200

Deploy NVIDIA B200 on Vultr Cloud GPU

Deploy NVIDIA B200 instancesContact Sales