HCAE HCAE (Hybrid Convolutional-Attention Encoder) is a next-generation family of lightweight text embedding models designed for extreme efficiency wit HeavensHackDev/HCAE-21M-Base Feature Extraction • 23.9M • Updated Apr 8 HeavensHackDev/HCAE-21M-Instruct Feature Extraction • 23.9M • Updated Apr 8 • 2 HeavensHackDev/HCAE-21M-v1.1-Instruct Feature Extraction • 24.4M • Updated Apr 9 • 39 HeavensHackDev/HCAE-21M-v1.1-Base Feature Extraction • 24.4M • Updated Apr 9 • 43
Gear The Gear family of LLMs HeavensHackDev/Gear-1-160m Text Generation • 0.2B • Updated Sep 16, 2025 • 13 HeavensHackDev/Gear-2-500m Text Generation • 0.5B • Updated Dec 7, 2025 • 11
HCAE HCAE (Hybrid Convolutional-Attention Encoder) is a next-generation family of lightweight text embedding models designed for extreme efficiency wit HeavensHackDev/HCAE-21M-Base Feature Extraction • 23.9M • Updated Apr 8 HeavensHackDev/HCAE-21M-Instruct Feature Extraction • 23.9M • Updated Apr 8 • 2 HeavensHackDev/HCAE-21M-v1.1-Instruct Feature Extraction • 24.4M • Updated Apr 9 • 39 HeavensHackDev/HCAE-21M-v1.1-Base Feature Extraction • 24.4M • Updated Apr 9 • 43
Gear The Gear family of LLMs HeavensHackDev/Gear-1-160m Text Generation • 0.2B • Updated Sep 16, 2025 • 13 HeavensHackDev/Gear-2-500m Text Generation • 0.5B • Updated Dec 7, 2025 • 11
