bvassie/nemotron-3.5-lightning:30b-mtp

bvassie/ nemotron-3.5-lightning:30b-mtp

8 Downloads Updated yesterday

NVIDIA Nemotron 3.5 Lightning is an open 30B mixture-of-experts (MoE) model with 3B active parameters built for always-on agents.

ollama run bvassie/nemotron-3.5-lightning:30b-mtp

curl http://localhost:11434/api/chat \
  -d '{
    "model": "bvassie/nemotron-3.5-lightning:30b-mtp",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='bvassie/nemotron-3.5-lightning:30b-mtp',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'bvassie/nemotron-3.5-lightning:30b-mtp',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Details

Updated yesterday

yesterday

792f1bcbecda · 20GB ·

model

archnemotron_h_moe

parameters31.6B

quantizationQ4_0

19GB

draft

GGUF��>��general.architecture��nemotron_h_moe��general.type�

1.2GB

params

{ "draft_num_predict": 2, "temperature": 1, "top_p": 0.95 }

53B

template

13B

Readme

Created using:

https://huggingface.co/ggml-org/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-GGUF

Modelfile:

FROM ./NVIDIA-Nemotron-3.5-Lightning-30B-A3B-Q4_0.gguf
DRAFT ./mtp-NVIDIA-Nemotron-3.5-Lightning-30B-A3B-Q4_0.gguf
TEMPLATE {{ .Prompt }}
RENDERER nemotron-3.5-nano
PARSER nemotron-3.5-nano
PARAMETER draft_num_predict 2
PARAMETER temperature 1
PARAMETER top_p 0.95