🤝 Open to Collab, Open to Learn\Work
DaveJski.
psychofanPLAYS
AI & ML interests
AI fanatic, my ADHD fell in love with it. Looking to expand into quanting, fine-tuning and training my own models. If you know what you’re doing, and looking for someone to team up with, shoot me a message!
Recent Activity
liked a model about 17 hours ago
Lightricks/LTX-2.5 updated a collection 17 days ago
Best_Picks liked a model 17 days ago
ukisai/Swift-Qwen3.8-27bOrganizations
None yet
5090 Qwen3.8-27B NVFP4 Best setup for real world use (vLLM wins over llama)
17
#132 opened about 2 months ago
by 
marco9899
Can we expect Ornith based on q3.8 ??
#13 opened about 1 month ago
by psychofanPLAYS
Horrible model .... Overthink eats all context tokens
👍😔 6
23
#136 opened about 2 months ago
by 
kremerneil
Why Qwen3.8-27B overthinks? Here the reason and partial fix confirmed by benchmarks.
❤️👀 25
29
#76 opened about 2 months ago
by 
LuffyTheFox
FP8 KV Cache Calibration
2
#109 opened about 2 months ago
by 
peterdab
why model return stop casually
3
#111 opened about 2 months ago
by 
xuguowong
notice how they all do good at around 2t
11
#36 opened about 2 months ago
by 
wesdoffical
New activity in DavidAU/Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-NEO-MAX-MTP-GGUF 2 months ago
Abliteration not working
👍 3
64
#13 opened 3 months ago
by 
Dhrhciebcy
New activity in DavidAU/Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-NEO-MAX-MTP-GGUF 2 months ago
Tested >200k context on 2 RTX 3090's (48GB VRAM)
10
#9 opened 3 months ago
by 
Mick277
can u do this 711 method to uncensored heretic qwen 3.6 35b moe version
👍 3
5
#15 opened 3 months ago
by 
allenhuggingface
Single Coder Setting?
2
#37 opened 3 months ago
by 
exlr8in