Instructions to use MarinaraSpaghetti/Nemomix-v1.0-12B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use MarinaraSpaghetti/Nemomix-v1.0-12B with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="MarinaraSpaghetti/Nemomix-v1.0-12B")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("MarinaraSpaghetti/Nemomix-v1.0-12B")
model = AutoModelForCausalLM.from_pretrained("MarinaraSpaghetti/Nemomix-v1.0-12B", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use MarinaraSpaghetti/Nemomix-v1.0-12B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "MarinaraSpaghetti/Nemomix-v1.0-12B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MarinaraSpaghetti/Nemomix-v1.0-12B",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/MarinaraSpaghetti/Nemomix-v1.0-12B

SGLang

How to use MarinaraSpaghetti/Nemomix-v1.0-12B with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "MarinaraSpaghetti/Nemomix-v1.0-12B" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MarinaraSpaghetti/Nemomix-v1.0-12B",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "MarinaraSpaghetti/Nemomix-v1.0-12B" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MarinaraSpaghetti/Nemomix-v1.0-12B",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use MarinaraSpaghetti/Nemomix-v1.0-12B with Docker Model Runner:
```
docker model run hf.co/MarinaraSpaghetti/Nemomix-v1.0-12B
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

V4.0 is the best one, use that one.

Information

Description

My main goal with this one was to merge the smartness of the base Instruct Nemo with the better prose from the different roleplaying fine-tunes. This is version v0.1, still to be tested. Weights shamelessly stolen from ParasiticRogue (thank you, friend). All credits and thanks go to Intervitens, Mistralai, NeverSleep and ShuttleAI for providing amazing models used in the merge.

Instruct

Both Mistral Instruct and ChatML should work.

<s>[INST] {system} [/INST]{assistant}</s>[INST] {user} [/INST]

Or...

<|im_start|>system
{system}<|im_end|>
<|im_start|>user
{user}<|im_end|>
<|im_start|>assistant
{assistant}<|im_end|>

Settings

Lower Temperature of 0.35 recommended, although I had luck with Temperatures above one (1.0-1.2) if you crank up the Min P (0.01-0.1). Run with base DRY of 0.8/1.75/2/0 and you're good to go.

GGUF

https://huggingface.co/MarinaraSpaghetti/Nemomix-v0.1-12B-GGUF

Other Versions

V1: https://huggingface.co/MarinaraSpaghetti/Nemomix-v1.0-12B

V2: https://huggingface.co/MarinaraSpaghetti/Nemomix-v2.0-12B

V3: https://huggingface.co/MarinaraSpaghetti/Nemomix-v3.0-12B

V4: https://huggingface.co/MarinaraSpaghetti/Nemomix-v4.0-12B

Nemomix-v0.1-12B

This is a merge of pre-trained language models created using mergekit.

Merge Details

Merge Method

This model was merged using the DARE TIES merge method using F:\mergekit\mistralaiMistral-Nemo-Base-2407 as a base.

Models Merged

The following models were included in the merge:

F:\mergekit\intervitens_mini-magnum-12b-v1.1
F:\mergekit\mistralaiMistral-Nemo-Instruct-2407
F:\mergekit\NeverSleep_Lumimaid-v0.2-12B
F:\mergekit\shuttleai_shuttle-2.5-mini

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: F:\mergekit\shuttleai_shuttle-2.5-mini
    parameters:
      weight: 0.16
      density: 0.42
  - model: F:\mergekit\NeverSleep_Lumimaid-v0.2-12B
    parameters:
      weight: 0.22
      density: 0.54
  - model: F:\mergekit\intervitens_mini-magnum-12b-v1.1
    parameters:
      weight: 0.28
      density: 0.66
  - model: F:\mergekit\mistralaiMistral-Nemo-Instruct-2407
    parameters:
      weight: 0.34
      density: 0.78
merge_method: dare_ties
base_model: F:\mergekit\mistralaiMistral-Nemo-Base-2407
parameters:
  int8_mask: true
dtype: bfloat16