Citing @adsilva264, the account says llama.cpp's fit option treats context as a floor, pushing a 64K setting to the full 256K and leaving half the VRAM idle at 20–30 tokens per second. Turning fit off, pinning layers and cutting context to 128K raised...
VQV Signal
OPEN SOURCE
WATCH
TECHNICAL
Community tip: llama.cpp's fit setting was halving Qwen3.5 speed on a 16GB card
Citing @adsilva264, the account says llama.cpp's fit option treats context as a floor, pushing a 64K setting to the full 256K and leaving half the VRAM idle at 20–30 tokens per second. Turning fit off, pinning layers and cutting context to 128K raised...
Citing @adsilva264, the account says llama.cpp's fit option treats context as a floor, pushing a 64K setting to the full 256K and leaving half the VRAM idle at 20–30 tokens per second. Turning fit off, pinning layers and cutting context to 128K raised decoding to 74 tokens per second on an RTX 5080 (the author's...
VQV organizes public signals from inspectable sources. It does not independently verify the underlying report.
Signal Strength 88%
Technical label WATCH
Public Interest 45
Category OPEN SOURCE
Reader Depth TECHNICAL
Signal Strength reflects source quality, relevance, freshness and evidence. Public Interest helps organize discovery; it is not proof of truth.
Public Interest components
Recognizable Entity Score 82
Practical Impact Score 8
Novelty Interest Score 70
Consequence Score 18
Curiosity Score 0
Shareability Score 60
Models
Topics
VQV surfaced this signal because it is recent, relevant to Open Source LLMs, connected to Scobleizer News Pulse (X).
No login, cookies, social SDKs, or automatic posting.