Repository navigation
Upstream sync 11/N: merge 9d808e2309 fastsafetensors loader (conflict) - #1105
Merged
Merged
Conversation
…oject#40183) Signed-off-by: Git Bisector <gitbisector@gmail.com> Signed-off-by: gitbisector <gitbisector@gmail.com> Signed-off-by: git bisector <gitbisector@gmail.com> Co-authored-by: Claude <noreply@anthropic.com> Co-authored-by: Cyrus Leung <tlleungac@connect.ust.hk>
Conflict-only step of the upstream catch-up. 9d808e2 is "[Core] Use fastsafetensors ParallelLoader for weight loading (vllm-project#40183)". vllm/envs.py: both sides append declarations to the same block - the fork's AWQ switches and upstream's VLLM_FASTSAFETENSORS_QUEUE_SIZE. Purely additive, keep both. Verified each of the four still has a matching runtime entry in the environment_variables dict. envs.py is the only fork-modified file the commit touches; its other two files (weight_utils.py and a fastsafetensors test) carry no fork changes. Signed-off-by: Robert Esclapez Garcia <robert.garcia@amd.com>
mgehre-amd
approved these changes
Aug 13, 2026
1 task done
eble-amd
approved these changes
Aug 13, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Context
Eleventh step of the batched upstream catch-up. Stacked on #1103; the chain below it is #1100 → #1101 → #1102 → #1103 (#1097 and #1098 have landed on
gfx11).Conflict-only step: exactly one upstream commit.
9d808e2309"[Core] Use fastsafetensors ParallelLoader for weight loading (vllm-project#40183)"mainConflict and resolution
vllm/envs.py— same shape as #1102: both sides append declarations to the same block. The fork's AWQ switches (VLLM_USE_TRITON_AWQ_GEMV,VLLM_ALLOW_UNFUSED_AWQ_GEMM,VLLM_AWQ_USE_TN_GEMM) and upstream'sVLLM_FASTSAFETENSORS_QUEUE_SIZE. Purely additive — kept both.Post-merge checks
environment_variablesdict.envs.pyis the only fork-modified file this commit touches; the other two (weight_utils.pyand a fastsafetensors test) carry no fork changes, so there is no fork code depending on the refactored loader API.Test plan