amd/gpt-oss-20b-WFP8-AFP8-KVFP8

Quantization vs openai/gpt-oss-20b. amd/gpt-oss-20b-WFP8-AFP8-KVFP8 is a quantized version of the other (F8_E4M3/BF16).

On the Hub · Downloads on Model Pulse