FP-8 version please 🥺
#7
by
nikhilfande
- opened
Could you release fp8 version of this model succ that we could fit into 2XH100 gpus for production ready deployment
Hi, the developers have released several quantized versions. The FP8 variant is available in the ‘Qwen/Qwen3-Coder-Next-FP8’ directory.
Oh great, thanks for confirming
nikhilfande
changed discussion status to
closed