Support int8 convrot embedding lookup (CORE-371) by kijai · Pull Request #15035 · Comfy-Org/ComfyUI · GitHub
Skip to content

Support int8 convrot embedding lookup (CORE-371) - #15035

Merged
comfyanonymous merged 2 commits into
Comfy-Org:masterfrom
kijai:int8_embedding
Jul 28, 2026
Merged

comfyanonymous merged 2 commits into
Comfy-Org:masterfrom
kijai:int8_embedding

Conversation

@kijai

@kijai kijai commented Jul 22, 2026

Copy link
Copy Markdown
Collaborator

Adds support for int8 embedding lookup, useful for text generation models with lm_head or embed_tokens layers such as used in Gemma4.

Compared to the fp8 scaled (the only previously supported quant format for embed lookup), int8 with convrot reaches much lower relative error at ~0.8% vs ~2.6% of fp8_scaled.

Since the embedding lookup is only ~0.10% of the token time, speed is not an issue and I only included the eager backend.

Requires comfy-kitchen PR: Comfy-Org/comfy-kitchen#83

@coderabbitai

coderabbitai Bot commented Jul 22, 2026 •

Copy link
Copy Markdown

@kijai kijai changed the title Support int8 convrot embedding lookup Support int8 convrot embedding lookup (CORE-371) Jul 24, 2026
@comfyanonymous
comfyanonymous merged commit 3d41e3e into Comfy-Org:master Jul 28, 2026
15 checks passed
@github-actions github-actions Bot locked and limited conversation to collaborators Jul 28, 2026
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants