nm-testing/DeepSeek-Coder-V2-Lite-Instruct-FP8

Quantization vs deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct. nm-testing/DeepSeek-Coder-V2-Lite-Instruct-FP8 is a quantized version of the other (F8_E4M3/BF16).

On the Hub · Downloads on Model Pulse