Instructions to use Nimbz/knifeayumu_Cydonia-v4.1-MS3.2-Magnum-Diamond-24B_6.66bpw_H8_EXL3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Nimbz/knifeayumu_Cydonia-v4.1-MS3.2-Magnum-Diamond-24B_6.66bpw_H8_EXL3 with Transformers:

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

pipe = pipeline("text-generation", model="Nimbz/knifeayumu_Cydonia-v4.1-MS3.2-Magnum-Diamond-24B_6.66bpw_H8_EXL3")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Nimbz/knifeayumu_Cydonia-v4.1-MS3.2-Magnum-Diamond-24B_6.66bpw_H8_EXL3")
model = AutoModelForCausalLM.from_pretrained("Nimbz/knifeayumu_Cydonia-v4.1-MS3.2-Magnum-Diamond-24B_6.66bpw_H8_EXL3", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Nimbz/knifeayumu_Cydonia-v4.1-MS3.2-Magnum-Diamond-24B_6.66bpw_H8_EXL3 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Nimbz/knifeayumu_Cydonia-v4.1-MS3.2-Magnum-Diamond-24B_6.66bpw_H8_EXL3"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Nimbz/knifeayumu_Cydonia-v4.1-MS3.2-Magnum-Diamond-24B_6.66bpw_H8_EXL3",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/Nimbz/knifeayumu_Cydonia-v4.1-MS3.2-Magnum-Diamond-24B_6.66bpw_H8_EXL3

SGLang

How to use Nimbz/knifeayumu_Cydonia-v4.1-MS3.2-Magnum-Diamond-24B_6.66bpw_H8_EXL3 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 "Nimbz/knifeayumu_Cydonia-v4.1-MS3.2-Magnum-Diamond-24B_6.66bpw_H8_EXL3" \
    --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": "Nimbz/knifeayumu_Cydonia-v4.1-MS3.2-Magnum-Diamond-24B_6.66bpw_H8_EXL3",
		"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 "Nimbz/knifeayumu_Cydonia-v4.1-MS3.2-Magnum-Diamond-24B_6.66bpw_H8_EXL3" \
        --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": "Nimbz/knifeayumu_Cydonia-v4.1-MS3.2-Magnum-Diamond-24B_6.66bpw_H8_EXL3",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Nimbz/knifeayumu_Cydonia-v4.1-MS3.2-Magnum-Diamond-24B_6.66bpw_H8_EXL3 with Docker Model Runner:
```
docker model run hf.co/Nimbz/knifeayumu_Cydonia-v4.1-MS3.2-Magnum-Diamond-24B_6.66bpw_H8_EXL3
```

Configuration Parsing Warning:In config.json: "quantization_config.bits" must be an integer

EXL3 6.66bpw H8 quant, tested using it with 24k q8 context within 24GB VRAM.

Original: https://huggingface.co/knifeayumu/Cydonia-v4.1-MS3.2-Magnum-Diamond-24B

Cydonia-v4.1-MS3.2-Magnum-Diamond-24B

Recipe based on knifeayumu/Cydonia-v1.2-Magnum-v4-22B. Just an update to those who are interested.

Image to Video Generation Info

Wan-AI/Wan2.2-I2V-A14B was used to turn this image from the previous merge (slightly cropped) to an animation utilising lightx2v/Wan2.2-Lightning for faster generation and pollockjj/ComfyUI-MultiGPU nodes.

Merge Details

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

Merge Method

This model was merged using the SLERP merge method.

Models Merged

The following models were included in the merge:

TheDrummer/Cydonia-24B-v4.1
Doctor-Shotgun/MS3.2-24B-Magnum-Diamond

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: TheDrummer/Cydonia-24B-v4.1
  - model: Doctor-Shotgun/MS3.2-24B-Magnum-Diamond
merge_method: slerp
base_model: TheDrummer/Cydonia-24B-v4.1
parameters:
  t: [0.1, 0.3, 0.6, 0.3, 0.1]
dtype: bfloat16

Downloads last month: 6

Safetensors

Model size

10B params

Tensor type

F16

I16

Model tree for Nimbz/knifeayumu_Cydonia-v4.1-MS3.2-Magnum-Diamond-24B_6.66bpw_H8_EXL3

Base model

knifeayumu/Cydonia-v4.1-MS3.2-Magnum-Diamond-24B

Quantized

(8)

this model