recommend(): 0.5 GB tolerance so nameplate min_vram tiers are reachable - #59
Merged
Conversation
Fixes #56. min_vram thresholds are nameplate card sizes but vram_gb is free VRAM; an 8 GB card reports ~7.5-7.9 free, so the flux-family 8 GB tiers (and every other nameplate-valued threshold) silently dropped to the CPU tier on exactly the cards they target.
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Fixes #56.
The tier table's
min_vramvalues are nameplate card sizes (8, 12, 16...), butrecommend()compares them against detected FREE VRAM, and no card reports its nameplate as free: a real 8 GB card shows ~7.5-7.9 GB. So every nameplate-valued tier (the flux-family 8 GB nf4 tiers, SDXL's 12 GB full-GPU tier, flux2's 16 GB bf16 tier) was unreachable on exactly the card it targets, silently dropping to the CPU tier.One-line fix, option 2 from the issue: a 0.5 GB tolerance in the single comparison, keeping the tier table's readable nameplate numbers. The CPU tier (min_vram = 0) is unaffected. Tests pin the boundary:
vram_gb = 7.6selects the flux2 8 GB GPU tier,7.4drops to CPU; all 92 existing recommend tests unchanged.