AI & ML interests
We do Ternary Models
Recent Activity
View all activity
Articles
kgrabko
updated 9
models 1 day ago
CMSManhattan/JiRackCoderReasoning_8b
Updated • 1
CMSManhattan/JiRackCoderReasoning_14b
Updated • 1
CMSManhattan/JiRackCoderReasoning_32b
Updated • 1
CMSManhattan/JiRackTernaryPro_1b
Text Generation • 1B • Updated • 403 • 1
CMSManhattan/JiRackUltra_1b
Text Generation • 2B • Updated • 6.3k
CMSManhattan/JiRackUltra_7b
Text Generation • 8B • Updated • 10.3k • 2
CMSManhattan/JiRackUltra_32b
Text Generation • 33B • Updated • 9.14k
CMSManhattan/JiRackUltra_14b
Text Generation • 15B • Updated • 83.8k • 1
CMSManhattan/JiRackDeltaNet_27b
Text Generation • 27B • Updated • 2.71k
Article
JiRack Ultra blew up the Hugging Face charts
CMSManhattan
• Article
Almost Opus 4.6 Max quality with JiRack DeltaNet 27B — but it runs on your PC
CMSManhattan
• Post
61
Ternary Transformers & Micro-Agent Architecture
CMSManhattan : Center Business Solutions Inc.
JiRack — Ternary Transformers & Micro-Agent Architecture
We build highly efficient large language models using 1.58-bit ternary weights {-1, 0, 1} for extreme compression and fast CPU/GPU inference.
Core focus:
JiRack Ternary Transformer Architecture — fresh Qwen base, trained on DeepSeek-style datasets, optimized for fast CPU inference (MIT License)
JiRack Micro-Agent Deployment — specialized small models + smart router for low-cost agentic systems
Production-ready ONNX Runtime & Docker inference stacks
Public Models
ModelSizeStatusJiRackUltra series (1B / 7B / 14B / 32B)—Released
CMSManhattan/JiRackUltra_1b
CMSManhattan/JiRackUltra_7b
CMSManhattan/JiRackUltra_14b
CMSManhattan/JiRackUltra_32b
JiRackTernary series1B → 10B+ReleasedJiRackPrecisionTokenizer—Released
Mission
Democratize frontier-scale language models through extreme efficiency. Train and run powerful models on accessible hardware without sacrificing quality.
Solved issues
Benefits of JiRack Micro-Agent Architecture:
Solves catastrophic forgetting during training by using small, specialized models for each domain, managed by a smart router Enables extremely cheap inference using ternary models Significantly reduces cloud inference costs while maintaining high performance In classical architecture, an expensive model has to search for MCP-agents every time, while JiRack uses a very small model and cheap router for agent tasks, saving big money right from the start Considered one of the best approaches for enterprise AI deployments
Hugging Face: CMSManhattan
Ollama : https://ollama.com/cmsmanhattan
Docker Hub: cmsmanhattan Contact: grabko@cmsmanhattan.com
CMSManhattan : Center Business Solutions Inc.
JiRack — Ternary Transformers & Micro-Agent Architecture
We build highly efficient large language models using 1.58-bit ternary weights {-1, 0, 1} for extreme compression and fast CPU/GPU inference.
Core focus:
JiRack Ternary Transformer Architecture — fresh Qwen base, trained on DeepSeek-style datasets, optimized for fast CPU inference (MIT License)
JiRack Micro-Agent Deployment — specialized small models + smart router for low-cost agentic systems
Production-ready ONNX Runtime & Docker inference stacks
Public Models
ModelSizeStatusJiRackUltra series (1B / 7B / 14B / 32B)—Released
CMSManhattan/JiRackUltra_1b
CMSManhattan/JiRackUltra_7b
CMSManhattan/JiRackUltra_14b
CMSManhattan/JiRackUltra_32b
JiRackTernary series1B → 10B+ReleasedJiRackPrecisionTokenizer—Released
Mission
Democratize frontier-scale language models through extreme efficiency. Train and run powerful models on accessible hardware without sacrificing quality.
Solved issues
Benefits of JiRack Micro-Agent Architecture:
Solves catastrophic forgetting during training by using small, specialized models for each domain, managed by a smart router Enables extremely cheap inference using ternary models Significantly reduces cloud inference costs while maintaining high performance In classical architecture, an expensive model has to search for MCP-agents every time, while JiRack uses a very small model and cheap router for agent tasks, saving big money right from the start Considered one of the best approaches for enterprise AI deployments
Hugging Face:
Ollama : https://ollama.com/cmsmanhattan
Docker Hub: cmsmanhattan Contact: grabko@cmsmanhattan.com
- 1 reply
kgrabko
published a
model 10 days ago
kgrabko
published a
model 11 days ago
kgrabko
updated a
model 11 days ago
kgrabko
updated a
model 16 days ago
kgrabko
updated a
model 23 days ago
kgrabko
published a
model about 1 month ago
kgrabko
updated a
model about 1 month ago