Instructions to use openbmb/MiniCPM5-2B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use openbmb/MiniCPM5-2B with Transformers:

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

pipe = pipeline("text-generation", model="openbmb/MiniCPM5-2B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("openbmb/MiniCPM5-2B")
model = AutoModelForCausalLM.from_pretrained("openbmb/MiniCPM5-2B", 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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use openbmb/MiniCPM5-2B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "openbmb/MiniCPM5-2B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "openbmb/MiniCPM5-2B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/openbmb/MiniCPM5-2B

SGLang

How to use openbmb/MiniCPM5-2B 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 "openbmb/MiniCPM5-2B" \
    --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": "openbmb/MiniCPM5-2B",
		"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 "openbmb/MiniCPM5-2B" \
        --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": "openbmb/MiniCPM5-2B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use openbmb/MiniCPM5-2B with Docker Model Runner:
```
docker model run hf.co/openbmb/MiniCPM5-2B
```

Remove the dead `<tool_sep>` branch from chat_template.jinja (fixes OpenBMB/MiniCPM#379)

#19

by lhz7891444 - opened 12 days ago

base: refs/heads/main

←

from: refs/pr/19

Discussion Files changed

-66

lhz7891444

12 days ago

Remove the dead <tool_sep> branch from chat_template.jinja (fixes OpenBMB/MiniCPM#379)

What

Two changes, nothing else:

Delete the <tool_sep> handling block inside the assistant branch (former lines 54–117). It split content on <tool_sep> and tried to interleave the tool-call XML between the text segments by accumulating processed_content inside for loops.
Change the trailing tool-call block guard from {%- if message.tool_calls and not has_tool_sep %} to {%- if message.tool_calls %}.

Diff: 66 lines removed, 1 line changed. Parameter rendering, reasoning handling, tool-response handling and the generation prompt are untouched.

Why

The branch never worked. Jinja2 does not propagate {% set %} assignments made inside a for loop to the enclosing scope, so processed_content kept its initial value content_parts[0] — everything after the first <tool_sep> was silently dropped (see #379 for the minimal reproducer). has_tool_sep was never defined anywhere, so the trailing guard was always true and the tool-call XML was emitted by the fallback block regardless.

The branch is unreachable in normal serving. <tool_sep> is not a special token in the tokenizer, the SGLang minicpm5_detector neither produces nor consumes it, and the model does not emit it (0/10 samples across prompts designed to elicit "text, then tool call, then text"). Only a hand-crafted assistant message containing the literal string <tool_sep> reaches this code, and in that case it loses data.

Removing it is therefore a pure cleanup with one observable effect: an assistant content that happens to contain <tool_sep> is now preserved verbatim instead of being truncated.

Remove the dead `<tool_sep>` branch from chat_template.jinja (fixes OpenBMB/MiniCPM#379)930f6388

xcjthu changed pull request status to merged 4 days ago

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment