lukasmwerner/mark-tagger:1b

lukasmwerner/ mark-tagger:1b

22 Downloads Updated 1 year ago

JSON Website Tagging Model for mark bookmark manager

ollama run lukasmwerner/mark-tagger:1b

curl http://localhost:11434/api/chat \
  -d '{
    "model": "lukasmwerner/mark-tagger:1b",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='lukasmwerner/mark-tagger:1b',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

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

Details

Updated 1 year ago

1 year ago

960b50875af7 · 1.3GB

model

archllama

parameters1.24B

quantizationQ8_0

1.3GB

Readme

1B model: lukasmwerner/mark-tagger:1b is based on the llama3.2:1b instruct model and is fine tuned on a personal dataset of bookmark behavior

Recommended temperature in mark: 0.2

Mark Prompt Template

Generate 3-5 relevant tags for this content. Avoid redundancy or similar sounding tags. Avoid putting the website title in the tags as well. Return only the tags as a JSON array of strings. No additional explanation needed.

Title: {{title}}
URL: {{url}}
Description: {{description}}