Instructions to use KaraKaraWarehouse/Matsutei-Qwen2.5-72b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use KaraKaraWarehouse/Matsutei-Qwen2.5-72b with Transformers:

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

pipe = pipeline("text-generation", model="KaraKaraWarehouse/Matsutei-Qwen2.5-72b")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("KaraKaraWarehouse/Matsutei-Qwen2.5-72b")
model = AutoModelForMultimodalLM.from_pretrained("KaraKaraWarehouse/Matsutei-Qwen2.5-72b")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use KaraKaraWarehouse/Matsutei-Qwen2.5-72b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "KaraKaraWarehouse/Matsutei-Qwen2.5-72b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KaraKaraWarehouse/Matsutei-Qwen2.5-72b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/KaraKaraWarehouse/Matsutei-Qwen2.5-72b

SGLang

How to use KaraKaraWarehouse/Matsutei-Qwen2.5-72b 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 "KaraKaraWarehouse/Matsutei-Qwen2.5-72b" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KaraKaraWarehouse/Matsutei-Qwen2.5-72b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "KaraKaraWarehouse/Matsutei-Qwen2.5-72b" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KaraKaraWarehouse/Matsutei-Qwen2.5-72b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use KaraKaraWarehouse/Matsutei-Qwen2.5-72b with Docker Model Runner:
```
docker model run hf.co/KaraKaraWarehouse/Matsutei-Qwen2.5-72b
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Matsutei-Qwen2.5-72b

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

SteyrCannon had a weird vibe issue when it comes to world book lore info (When starting from afresh, it might get confused when world info is injected due to double chat messages). So I've been falling back to my EurobeatVARemix merge. This merge should address that inital issue but I think there's other quirks to this one.

Quants & Hosts

Merge Details

Merge Method

This model was merged using the TIES merge method using EVA-UNIT-01/EVA-Qwen2.5-72B-v0.1 as a base.

Models Merged

The following models were included in the merge:

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: KaraKaraWitch/SteyrCannon-Qwen2.5-72b
    parameters:
      density: 0.25
      weight: 0.5
  - model: EVA-UNIT-01/EVA-Qwen2.5-72B-v0.1
    parameters:
      density: 0.5
      weight: 0.75
  - model: ZeusLabs/Chronos-Platinum-72B
    parameters:
      density: 0.5
      weight: 0.5
  - model: m8than/banana-2-b-72b
    parameters:
      density: 0.65
      weight: 0.40
  

merge_method: ties
base_model: EVA-UNIT-01/EVA-Qwen2.5-72B-v0.1
parameters:
  normalize: true
dtype: bfloat16