harald/nixcfg
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Revert "fix(halo): 27 only"

Browse source 
This reverts commit 72e7bf613f.
This commit is contained in:
Harald Hoyer 2026-05-20 07:05:27 +02:00
parent 72e7bf613f
commit 31e491e314
1 changed files with 9 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

10
systems/x86_64-linux/halo/models.ini
View file

@ -1,6 +1,6 @@
version = 1 version = 1
[Qwen3.6-27B] [*]
flash-attn = on flash-attn = on
jinja = true jinja = true
n-gpu-layers = 99 n-gpu-layers = 99
@ -17,6 +17,14 @@ no-context-shift = true
chat-template-kwargs = {"preserve_thinking": true} chat-template-kwargs = {"preserve_thinking": true}
fit = on fit = on
spec-type = draft-mtp spec-type = draft-mtp
[Qwen3.6-35B-A3B]
hf = unsloth/Qwen3.6-35B-A3B-MTP-GGUF:UD-Q8_K_XL
spec-draft-n-max = 3
parallel = 1
c = 262144
[Qwen3.6-27B]
hf = unsloth/Qwen3.6-27B-MTP-GGUF:Q6_K hf = unsloth/Qwen3.6-27B-MTP-GGUF:Q6_K
spec-draft-n-max = 6 spec-draft-n-max = 6
parallel = 2 parallel = 2