cmsmanhattan/ jirack-ultra-1b-q4:latest

Updated 9 hours ago

ollama run cmsmanhattan/jirack-ultra-1b-q4

curl http://localhost:11434/api/chat \
  -d '{
    "model": "cmsmanhattan/jirack-ultra-1b-q4",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='cmsmanhattan/jirack-ultra-1b-q4',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'cmsmanhattan/jirack-ultra-1b-q4',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Details

Updated 9 hours ago

9 hours ago

5c4068bd70c9 · 1.1GB ·

model

archqwen2

parameters1.78B

quantizationQ4_K_M

1.1GB

template

<｜User｜>{{ .Prompt }}<｜Assistant｜>

42B

system

You are JiRack, a helpful and honest AI assistant.

50B

params

{ "num_ctx": 8192, "repeat_penalty": 1.1, "stop": [ "<｜end▁of▁sentence｜>

142B

Readme

JiRack Ultra 1B Q4_K_M

Fast ~1.5B CPU model from CMS Manhattan. BitNet-style ternary path, extended tokenizer with Routing, Tool-call and Robotics tags. Built on a redesigned DeepSeek R1 / Qwen2 stack. This tag is Q4_K_M.

HF: https://huggingface.co/CMSManhattan/JiRackUltra_1b

Run

ollama run cmsmanhattan/jirack-ultra-1b-q4

What it is

Cloud-ready small expert for RAG and local inference
Tokenizer tags for routing, tool calls, robotics
Quant from FP16; ternary architecture can go smaller with QAT
Java / Spring AI and Gorilla-style tool-call examples on the HF card

Hardware

Q4_K_M: about 1.1 GB on disk, 2–4 GB RAM recommended.

Contact

grabko@cmsmanhattan.com
+1 (516) 777-0945
New York, USA

License

MIT on weights. Commercial terms for UI / enterprise on the HF page.