Llama-3-8B-Instruct-ANIMTOOLS-Encoder

Built with Meta Llama 3

Unmodified base weights of meta-llama/Meta-Llama-3-8B-Instruct, redistributed under the terms of the Meta Llama 3 Community License as the text-encoder base for ANIMTOOLS (an AI motion-generation toolset for Houdini by Old Machines Digital). The ANIMTOOLS installer downloads this repository so end users do not need a HuggingFace account for the gated upstream repo.

This repository contains exactly the files the ANIMTOOLS pipeline loads: the four safetensors shards, config.json, and model.safetensors.index.json. Tokenizer files are loaded from the MIT-licensed McGill-NLP LLM2Vec adapter repositories, which the encoder applies on top of these weights.

License

Use of these weights is governed by the Meta Llama 3 Community License and the accompanying Acceptable Use Policy. See NOTICE. By downloading this repository you agree to those terms.

This is an independent redistribution: no endorsement by Meta is implied. Llama 3 and Meta are trademarks of Meta Platforms, Inc.

Downloads last month
4
Safetensors
Model size
8B params
Tensor type
BF16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for oldmachines/Llama-3-8B-Instruct-ANIMTOOLS-Encoder

Finetuned
(1159)
this model