JM
akierum
·
AI & ML interests
None yet
Recent Activity
new activity about 14 hours ago
bartowski/nex-agi_Nex-N2-mini-GGUF:Fails with using chrome dev tools Take screenshot new activity about 15 hours ago
protoLabsAI/ThinkingCap-Qwen3.6-27B-MTP-GGUF:Janai imports model with wrong nameOrganizations
None yet
Fails with using chrome dev tools Take screenshot
#4 opened about 14 hours ago
by
akierum
Janai imports model with wrong name
3
#6 opened 1 day ago
by
akierum
Why LMstudio downbloads mmproj-F32.gguf instead of mmproj-F16.gguf for Q8 ?
#32 opened 1 day ago
by
akierum
Thanks for this
🤗 1
1
#5 opened 2 days ago
by
jcbtc
please provide vision mmproj file
🤝 1
3
#4 opened 2 days ago
by
akierum
Q8 seems best version tested so far even better than unsloth Q8_XL compared by codeneedle
2
#3 opened 2 days ago
by
akierum
ThinkingCap-Qwen3.6-27B - Without Q8_XL Quant to match Unsloth this is waste of time sorry.
1
#9 opened 2 days ago
by
akierum
codeneedle has bad results
2
#3 opened 6 days ago
by
akierum
So is this better then original Q8 version
#2 opened 6 days ago
by
akierum
included jinja template wrong default parameters why?
1
#2 opened 7 days ago
by
akierum
No vision support where is mmproj file?
4
#1 opened 7 days ago
by
akierum
Can not use 35B with claude code and codex.
2
#19 opened 12 days ago
by
beginor
Reasoning degradation and tool calling failures at 70-80K context tokens
👍 1
31
#22 opened 11 days ago
by
Sdoh
Aider Polyglot Benchmark Results (C++ & Python) — Q4 through Q8
👍 5
3
#23 opened 11 days ago
by
YukiTomita-CC
Breaks claude code
13
#41 opened 30 days ago
by
widdae
31B-Dense model missing from HuggingFace
👍👀 28
15
#2 opened 15 days ago
by
smcleod
Very impressive with openclaw! and fixed chat template by Opus
🔥 1
5
#11 opened 14 days ago
by
Andy5k
Does Ornith-1.0-35B preserve historical reasoning traces like Qwen3.6?
1
#11 opened 14 days ago
by
eflaneble
vision support?
1
#3 opened 18 days ago
by
akierum
Fails to load on LM Studio 0.4.16 (Build 2)
3
#1 opened 20 days ago
by
akierum