Commit Graph

2 Commits

Author SHA1 Message Date
eea73f1ce5 fix: reorder QUANT_LEVELS so q4_0 (4.0b) comes before turbo2 (1.5b)
All checks were successful
Smoke Test / smoke (pull_request) Successful in 33s
q4_0 has higher bits_per_channel (4.0 vs 1.5), meaning higher quality.
The selection logic iterates QUANT_LEVELS in order, picking the first
level that fits available memory. q4_0 should be tried before turbo2
because it produces better quality output when memory allows.

Closes #139
2026-04-21 11:19:54 +00:00
f7ea01cb65 feat: auto-select quantization based on available VRAM (#81) 2026-04-15 15:03:04 +00:00