daily chooser 2026-08-18: chose Qwen3.6-35B-A3B-8bit; correct upgrade→Qwen3-Next-80B-A3B-4bit (44.8GB fits), drop gpt-oss-120b (too big)

a52578d368382bba59be361ba58b4271e8c221dc · 2026-08-18 04:51:37 -0700 · Steve Abrams

Files touched

A data/chosen-model.txt

Diff

commit a52578d368382bba59be361ba58b4271e8c221dc
Author: Steve Abrams <steve@designerwallcoverings.com>
Date:   Tue Aug 18 04:51:37 2026 -0700

    daily chooser 2026-08-18: chose Qwen3.6-35B-A3B-8bit; correct upgrade→Qwen3-Next-80B-A3B-4bit (44.8GB fits), drop gpt-oss-120b (too big)
---
 data/chosen-model.txt | 1 +
 1 file changed, 1 insertion(+)

diff --git a/data/chosen-model.txt b/data/chosen-model.txt
new file mode 100644
index 0000000..e88704a
--- /dev/null
+++ b/data/chosen-model.txt
@@ -0,0 +1 @@
+mlx-community/Qwen3.6-35B-A3B-8bit

← 0747d9f purge remaining 90GB refs -> 55GB weights ceiling everywhere · back to Local Model Leaderboard Watch · battle: eval harness + 35B baseline (11/12, 55.2 tok/s, long 6963c73 →