voxtral_tts: enable CUDA backend with 4w quantization (Ampere + Blackwell pre-exported artifacts)#19093
Open
seyeong-han wants to merge 10 commits intopytorch:mainfrom
Open
voxtral_tts: enable CUDA backend with 4w quantization (Ampere + Blackwell pre-exported artifacts)#19093seyeong-han wants to merge 10 commits intopytorch:mainfrom
seyeong-han wants to merge 10 commits intopytorch:mainfrom
Commits
Commits on Apr 16, 2026
Commits on Apr 17, 2026
Commits on Apr 21, 2026
- committed
Young Han
Commits on Apr 23, 2026
- committedYoung Han
- committedYoung Han
- committedYoung Han
- committedYoung Han
Commits on Apr 24, 2026
- committed
- committed
- committed