DeepSeek-Coder-V2-Lite model GPU/RAM requirement
HashedViking opened this issue · comments
Sergey Zhdanov commented
Hi, thank you for the amazing work! In the readme you say "DeepSeek-Coder-V2 in BF16 format for inference, 80GB*8 GPUs are required".
How much GPU/RAM is needed for inference for DeepSeek-Coder-V2-Lite model?
Daya Guo commented
40GB *1 GPUs in BF16 format
orderer0001 commented
40GB *1 GPUs in BF16 format
3*24G 4090 GPU doesn’t work?
Daya Guo commented
40GB *1 GPUs in BF16 format
3*24G 4090 GPU doesn’t work?
If you have 3*24G 4090 GPUs, you need to enable TP 2 or PP 2. Alternatively, you can use the quantized version of the model on Ollama.