shokrydev/anoner-medgemma

shokrydev/ anoner-medgemma

6-bit (Q6_K) GGUF port of MedGemma-1.5-4B-it, a 4B multimodal instruction-tuned model based on the Gemma 3 architecture, offering improved medical reasoning and expanded imaging/document understanding.

ollama run shokrydev/anoner-medgemma

curl http://localhost:11434/api/chat \
  -d '{
    "model": "shokrydev/anoner-medgemma",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='shokrydev/anoner-medgemma',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'shokrydev/anoner-medgemma',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

No models have been pushed.

Readme

This model is an import of unsloth/medgemma-1.5-4b-it-Q6_K.gguf, available on Hugging Face. It is a 6-bit quantized (Q6_K) port of MedGemma-1.5-4B-IT, a 4B multimodal instruction-tuned model built on the Gemma 3 decoder-only transformer architecture and the first open-weights reasoning model published by Google.

MedGemma 1.5 4B is an updated version of MedGemma 1 4B that retains the same architecture while delivering improved performance on medical text reasoning and expanded support for medical imaging and document-understanding tasks, as described in the benchmarking results.

At 3.19 GB, it is compact enough to be loaded locally onto a GPU alongside other models, as in the case of Anoner.

6-bit (Q6_K) GGUF port of MedGemma-1.5-4B-it, a 4B multimodal instruction-tuned model based on the Gemma 3 architecture, offering improved medical reasoning and expanded imaging/document understanding.

Readme

References