Juk Armstrong
jukofyork
AI & ML interests
None yet
Recent Activity
liked
a model
1 day ago
patched-codes/Llama-3.2-1B-FastApply
liked
a model
7 days ago
AnnaWegmann/Style-Embedding
new activity
15 days ago
jukofyork/command-a-03-2025-DRAFT-0.8B-v3.0-GGUF:Draft/Speculative decoding
Organizations
None yet
Draft/Speculative decoding
3
#1 opened 17 days ago
by
yano2mch
Draft model with ik_llama.cpp
👍
1
2
#1 opened 5 months ago
by
Chico70
Create Boy in love
#17 opened 26 days ago
by
Imma9907
“The doom lies in yourself, not in your name.”
👀
6
248
#15 opened 4 months ago
by
jukofyork
Amazing It worked, Thanks.
1
#1 opened about 1 month ago
by
pankajmathur
The "Softmax Bottleneck" might be interesting to look at
2
#1 opened about 1 month ago
by
jukofyork
Was this really trained during QAT using a symmetric 4bit quant with only 15/16 values used?
👀
9
3
#26 opened 2 months ago
by
jukofyork
These don't really work.
5
#1 opened 2 months ago
by
Lockout
The data seems to have lost the `task_language` field?
2
#2 opened 3 months ago
by
jukofyork
Test 2: Opening chapter of a 'Grimdark' trilogy
#2 opened 3 months ago
by
jukofyork
Test 1: Battlestar Galactica in the prose of Haruki Murakami
#1 opened 3 months ago
by
jukofyork
Where are the control vectors and what do people use now?
4
#16 opened 3 months ago
by
rookaw
“The doom lies in yourself, not in your name.”
👀
6
248
#15 opened 4 months ago
by
jukofyork
Where are the control vectors and what do people use now?
4
#16 opened 3 months ago
by
rookaw
“The doom lies in yourself, not in your name.”
👀
6
248
#15 opened 4 months ago
by
jukofyork
Eval request: grok-2
🚀
🔥
2
2
#376 opened 3 months ago
by
ChuckMcSneed