Quark-150M-128K-Zero

Try Try

Quark-150M-128K-Zero has 165M parameters (165M active per token) and takes 661 MB on disk. See where its parameters live, layer by layer, and what hardware it needs.

At a glance

empero-ai/Quark-150M-128K-Zero
Repo createdNov 15, 2026updated Nov 15, 2026
Model typeDenseFull attention (GQA)
InputsText
Total parameters165M
Active per token165M100% of the model
ExpertsNone (dense)
Max context (from config)128K tokens
Layers16
On disk661 MB1 file
PrecisionFP32 (100%)
QuantizationNoneoriginal precision (FP32)
Fewest GPUs1× RTX 4060a single 8 GB card · weights only
FamilyFine-tune of empero-ai/Quark-150M-128K-Base
Licenseapache-2.0
GitHubNot linked