llama.cpp
llama: Wire up Qwen3.5/Qwen3.5MoE tensors for NVFP4 support
#20506

Merged

llama: Wire up Qwen3.5/Qwen3.5MoE tensors for NVFP4 support #20506

CISC merged 2 commits into ggml-org:master from michaelw9999:nvfp4-qwen35-nvfp4-loader-fix

llama : fix Qwen3.5 NVFP4 scale loading

478822c3

michaelw9999 requested a review from

CISC 119 days ago

michaelw9999 requested a review from

copilot-pull-request-reviewer 119 days ago

github-actions added model

copilot-pull-request-reviewer commented on 2026-03-13

michaelw9999 changed the title ~~ggml: Wire up Qwen3.5/Qwen3.5MoE tensors for NVFP4 support~~ llama: Wire up Qwen3.5/Qwen3.5MoE tensors for NVFP4 support 119 days ago

Merge branch 'ggml-org:master' into nvfp4-qwen35-nvfp4-loader-fix

a89e95fc

CISC approved these changes on 2026-03-14

CISC merged d23355af into master 117 days ago

michaelw9999 deleted the nvfp4-qwen35-nvfp4-loader-fix branch 117 days ago

Reviewers

CISC

copilot-pull-request-reviewer

Assignees

No one assigned

Labels

model

Milestone

No milestone

llama.cpp llama: Wire up Qwen3.5/Qwen3.5MoE tensors for NVFP4 support #20506 Merged

llama: Wire up Qwen3.5/Qwen3.5MoE tensors for NVFP4 support #20506

llama.cpp
llama: Wire up Qwen3.5/Qwen3.5MoE tensors for NVFP4 support
#20506

Merged