Hugging Face's logo

TheBloke
/
dolphin-2.6-mixtral-8x7b-GPTQ

Text Generation
Transformers
Safetensors
English
mixtral
conversational
text-generation-inference
4-bit precision
gptq
Model card Files Community
1
New discussion

KeyError: 'model.layers.0.block_sparse_moe.experts.0.w1.g_idx' when running with tensor parallelism on vllm

1
#1 opened 12 months ago by KronusCon