cmsmanhattan/ jirack-ultra-32b-q2:latest

1 Download Updated 1 month ago

ollama run cmsmanhattan/jirack-ultra-32b-q2

curl http://localhost:11434/api/chat \
  -d '{
    "model": "cmsmanhattan/jirack-ultra-32b-q2",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='cmsmanhattan/jirack-ultra-32b-q2',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'cmsmanhattan/jirack-ultra-32b-q2',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Details

Updated 1 month ago

1 month ago

a34010c3ed3e · 12GB

model

archqwen2

parameters32.8B

quantizationQ2_K

12GB

system

You are JiRack, a helpful and honest AI assistant.

50B

params

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

142B

template

<｜User｜>{{ .Prompt }}<｜Assistant｜>

42B

Readme

JiRack Ultra 32B Q2_K

Large ~32B model from CMS Manhattan. BitNet-style ternary path, extended tokenizer with Routing, Tool-call and Robotics tags. Production Q2_K build for maximum compression on constrained hardware.

HF: https://huggingface.co/CMSManhattan/JiRackUltra_32b

Run

ollama run cmsmanhattan/jirack-ultra-32b-q2

What it is

Flagship mid/large size in the Ultra line — stronger reasoning, coding, and tool use than 14B
Same JiRack tokenizer tags: routing, tools, robotics
Q2_K quant for the smallest practical disk and RAM footprint at 32B
Fits serious RAG, multi-agent, and Spring / Java tool-call stacks on tighter boxes
American ternary stack from CMS Manhattan (New York)

Hardware

Q2_K: roughly 11–13 GB on disk.
Recommend 16–20 GB VRAM (GPU) or 24–32 GB system RAM (CPU).

Contact

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

License

MIT on weights unless stated otherwise on the HF card. Commercial terms for UI / enterprise on request.