Timmy_Foundation/turboquant
15
0
Fork 0
Code Issues 38 Pull Requests 32 Actions Packages Projects Releases Wiki Activity

fix: quant selector quality-order assertion (#139)
All checks were successful
Smoke Test / smoke (pull_request) Successful in 14s
Details

Browse Source
This commit is contained in:
Alexander Whitestone
2026-04-21 11:25:16 +00:00
parent 492c1cdcfd
commit fc67bd7085
1 changed files with 10 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
evolution/quant_selector.py
View File

@@ -62,16 +62,6 @@ QUANT_LEVELS = [
min_memory_headroom_gb=3.0,
description="3-bit TurboQuant. High quality, ~6x KV compression."
),
QuantLevel(
name="turbo2",
bits_per_channel=1.5,
compression_ratio=10.0,
quality_label="balanced",
layer_adaptive=3,
kv_type="turbo2",
min_memory_headroom_gb=2.0,
description="2-bit TurboQuant. Balanced, ~10x KV compression."
),
QuantLevel(
name="q4_0",
bits_per_channel=4.0,
@@ -82,6 +72,16 @@ QUANT_LEVELS = [
min_memory_headroom_gb=1.5,
description="Standard 4-bit quant. Fast fallback, no TurboQuant."
),
QuantLevel(
name="turbo2",
bits_per_channel=1.5,
compression_ratio=10.0,
quality_label="balanced",
layer_adaptive=3,
kv_type="turbo2",
min_memory_headroom_gb=2.0,
description="2-bit TurboQuant. Balanced, ~10x KV compression."
),
]