Instructions to use djdeniro/MiniMax-M2.7-MXFP416 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use djdeniro/MiniMax-M2.7-MXFP416 with Transformers:

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

pipe = pipeline("text-generation", model="djdeniro/MiniMax-M2.7-MXFP416", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("djdeniro/MiniMax-M2.7-MXFP416", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("djdeniro/MiniMax-M2.7-MXFP416", trust_remote_code=True, device_map="auto")
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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use djdeniro/MiniMax-M2.7-MXFP416 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "djdeniro/MiniMax-M2.7-MXFP416"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "djdeniro/MiniMax-M2.7-MXFP416",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/djdeniro/MiniMax-M2.7-MXFP416

SGLang

How to use djdeniro/MiniMax-M2.7-MXFP416 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 "djdeniro/MiniMax-M2.7-MXFP416" \
    --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": "djdeniro/MiniMax-M2.7-MXFP416",
		"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 "djdeniro/MiniMax-M2.7-MXFP416" \
        --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": "djdeniro/MiniMax-M2.7-MXFP416",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use djdeniro/MiniMax-M2.7-MXFP416 with Docker Model Runner:
```
docker model run hf.co/djdeniro/MiniMax-M2.7-MXFP416
```

djdeniro commited on Jun 9

Commit

1a7f400

verified ·

1 Parent(s): 7bfe6c8

Fix MTP info: not available in quantized variant; update performance

Browse files

Files changed (1) hide show

README.md +2 -3

README.md CHANGED Viewed

@@ -39,7 +39,7 @@ The result fits in ~17.5 GiB per GPU (TP8) while retaining near-BF16 quality.
 - **456B total params** (sparse), **~30B activated** per token
 - **256 experts** per MoE layer, top-8 routing, 62 transformer layers
-- **3 MTP layers** for speculative decoding
 - **200k context window**
 - Native tool-calling support
@@ -81,10 +81,9 @@ docker run --name minimax-mxfp416 \
   -e ROCR_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 \
   -e TRUST_REMOTE_CODE=1 \
   -v /path/to/models:/app/models:ro \
-  -v /path/to/patches:/patches:ro \
   -p 8000:8000 \
   tcclaviger/vllm22:latest \
-  bash -c "cp /patches/vllm22_minimax_m2.py /app/vllm/vllm/model_executor/models/minimax_m2.py && \
     pip install -q sentencepiece && \
     exec vllm serve /app/models/MiniMax-M2.7-MXFP416 \
       --served-model-name minimax-m2.7-mxfp416 \

 - **456B total params** (sparse), **~30B activated** per token
 - **256 experts** per MoE layer, top-8 routing, 62 transformer layers
+- Config declares `use_mtp: True` (1 layer, 3 modules), but MTP weights were stripped during quantization — not available in this variant
 - **200k context window**
 - Native tool-calling support
   -e ROCR_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 \
   -e TRUST_REMOTE_CODE=1 \
   -v /path/to/models:/app/models:ro \
   -p 8000:8000 \
   tcclaviger/vllm22:latest \
+  bash -c "cp /app/models/vllm22_minimax_m2.py /app/vllm/vllm/model_executor/models/minimax_m2.py && \
     pip install -q sentencepiece && \
     exec vllm serve /app/models/MiniMax-M2.7-MXFP416 \
       --served-model-name minimax-m2.7-mxfp416 \