RedHatAI/DeepSeek-R1-Distill-Qwen-32B-FP8-dynamic

Quantization vs deepseek-ai/DeepSeek-R1-Distill-Qwen-32B. RedHatAI/DeepSeek-R1-Distill-Qwen-32B-FP8-dynamic is a quantized version of the other (F8_E4M3/BF16).

On the Hub · Downloads on Model Pulse