RedHatAI/Qwen3-4B-FP8-dynamic

Quantization vs Qwen/Qwen3-4B. RedHatAI/Qwen3-4B-FP8-dynamic is a quantized version of the other (F8_E4M3/BF16).

1 other repos hold byte-for-byte identical weight files.

On the Hub · Downloads on Model Pulse