Unverified50% confidenceFactExact time
v0.34.4 fixes an intermittent "model not found" error that occasionally occurred when a large number of models were stored locally.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/4/2026
First Seen
Valid until: 1/2/2027
Sources
Related Entities
Related Claims
UnverifiedThe model Ollama downloads by default is a quantized version; Q4 means each parameter is stored with 4-bit precision, which loses a small amount of precision but greatly reduces the file size.57% similarUnverifiedSetting a lower num_ctx parameter (such as 2048) in Ollama's Modelfile can limit the context window, reduce KV Cache memory usage, and avoid the sharp performance degradation caused by insufficient memory.57% similarUnverifiedWhen deploying large models locally, VRAM usage consists of two parts: model weights and runtime memory (primarily the KV cache)56% similarUnverifiedWhen the context is extended, approximately 20% of the 20.7B dense model's parameters overflow to memory, causing speed to plummet to 5.7 tokens per second, while Dwarf Star is almost unaffected.56% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/971727API
curl https://kongchang.com/api/v1/knowledge/claims/971727MCP
get_claim(id=971727)