Latest ContentInference Cookbook Model Library

Llama 3.1 70B

Llama 3.1 70B is a multilingual dense transformer large language model designed for advanced text generation, reasoning, and large-scale AI applications. The model features a 70B parameter architecture with 80 transformer layers, 64 attention heads, and an 8,192 hidden size, utilizing Grouped Query Attention (GQA) for scalable and efficient inference. It supports up to a 128K token context window with RoPE scaling for long-context understanding. Optimized for multilingual text and code generation, it is widely used for research, conversational AI, and enterprise-grade LLM deployments.

Type	Dense LLM
Capabilities	Text Generation, Instruction Following, Reasoning, Mathematical Reasoning+5 more
Group Release Date	July 22, 2024
Links	Blog\|HF Model Card
License	Llama3.1

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 --gpus all 
 --shm-size 128g 
 -p 8000:8000 
 -v ~/.cache/huggingface:/root/.cache/huggingface 
 -e HF_TOKEN='YOUR_HF_TOKEN' 
 --ipc=host 
 lmsysorg/sglang:v0.5.8-cu130 
 python3 -m sglang.launch_server 
 --model-path meta-llama/Llama-3.1-70B 
 --host 0.0.0.0 
 --port 8000 
 --max-prefill-tokens 65536 
 --max-running-requests 1024 
 --tp 8 
 --mem-fraction-static 0.95 
 --trust-remote-code