Meta-Llama-3-70B-Instruct is the instruction-tuned version of Meta's 70 billion parameter Llama 3 model. It excels at dialogue, reasoning, and following complex instructions. The model weights are publicly available on Hugging Face and can be used with the Transformers library or optimized inference runtimes. It is intended for developers and researchers building advanced AI applications.
Meta Llama 3 70B Instruct is a Foundation models & chat product. It focuses on running a high-performance 70B open-weight LLM locally for instruction-tuned text generation. It is built as an open-source project for developers and researchers. Meta Llama 3 70B Instruct is open source under the Open Source license. It runs on the web, the command line, and API.
Behind Meta Llama 3 70B Instruct is Meta, based in the United States, and the product first shipped in 2024. The GitHub repository has been archived. It operates in a well-populated space: PulseGate tracks 11 similar tools. Key capabilities include Chat Template, Instruction Following, and 70B Parameters.
Latest indexed changes and source events
meta-llama/Meta-Llama-3-70B-Instruct verified by the PulseGate indexer
⚠ Archived
