Andrew Moore
Moore2877
AI & ML interests
None yet
Organizations
None yet
kwarg for no template-driven skill triggering
👍 1
2
#7 opened 3 months ago
by
NikiKrutan
merge with froggeric
1
#6 opened 3 months ago
by
l0rdraiden
Breaks Qwen 3.8 CoT
1
#5 opened 3 months ago
by
R3tr000
plans for v17? now that qwen 3.8 is out?
👍 1
6
#4 opened 3 months ago
by
Shad3
Solves llama.cpp reasoning display
🤝 1
1
#2 opened 4 months ago
by
FilterJoe
good model ,have Q5 ?~~
🤗 1
6
#1 opened 5 months ago
by
mslxd
chat template v21.3
🔥 7
13
#51 opened 5 months ago
by
froggeric
Runtime-adaptive tool-call format: XML on vLLM, Hermes JSON on llama.cpp — no config needed
3
#52 opened 5 months ago
by
Moore2877
Runtime-adaptive tool-call format: XML on vLLM, Hermes JSON on llama.cpp — no config needed
1
#53 opened 5 months ago
by
Moore2877
Fix: preserve_thinking default regressed to false in v20 (contradicts v19 changelog & README)
3
#46 opened 5 months ago
by
Moore2877
Enhancement: support Anthropic-style `message.thinking` alongside `reasoning_content`
👍 1
1
#50 opened 5 months ago
by
Moore2877
Fix: <|think_on|>/<|think_off|> honored from tool responses (prompt-injection surface)
1
#49 opened 5 months ago
by
Moore2877
Fix: inline-quoted </think> in assistant history swallows half the message as reasoning
1
#48 opened 5 months ago
by
Moore2877
Fix: false-positive tool-error detection (marker anywhere in response) + add npm/git markers
1
#47 opened 5 months ago
by
Moore2877
Fix: tool calls silently lost on llama.cpp/ik_llama — switch XML format to Hermes JSON
1
#45 opened 5 months ago
by
Moore2877