Verified record · from the answer index

mixtral-8x7b-instruct-v0-1

This container houses the Mixtral-8x7B-Instruct-v0.1, a sophisticated, instruction-tuned model from Mistral AI. It uses a Sparse Mixture of Experts (SMoE) architecture, dynamically selecting 2 of its 8 experts per token for greater efficien

model:ngc:nim/mistralai/mixtral-8x7b-instruct-v0-1