Llama 3.1 Nemotron 70B Instruct HF
nvidia/Llama-3.1-Nemotron-70B-Instruct-HF
Open Source · chat · open-weights
Open
Alert me on changes
Intelligence
Context
131.1K
Max output
1K
Weights
Open
API $/1M
$1.20 / $1.20
Modalities
text
Released
12 Oct 2024
Intelligence Index via Artificial Analysis · 0–100, higher is better
License: llama3.1 · nvidia/Llama-3.1-Nemotron-70B-Instruct-HF
AI summary
● machine-written
Llama 3.1 Nemotron 70B Instruct released by NVIDIA
NVIDIA's Llama 3.1 Nemotron 70B Instruct is an open-weights language model designed for generating precise, helpful responses across diverse domains. Built on Llama 3.1 70B architecture with Reinforcement Learning from Human Feedback (RLHF), it supports a 131K token context window and targets applications requiring high accuracy in response generation.
What's new
- Open-weights model released October 2024 with RLHF fine-tuning for alignment
- 131,072 token context window for extended context handling
- Scores 8 on Artificial Analysis Intelligence Index, above average for comparable models
- Outputs 278.9 tokens per second in independent benchmarks
Best for
High-accuracy response generation across multiple domainsApplications requiring helpful, aligned outputsLong-context tasks with up to 131K token windowProduction deployments via open-source or API providers
Source: https://huggingface.co/nvidia/Llama-3.1-Nemotron-70B-Instruct-HF