Instructions to use meituan-longcat/LongCat-Flash-Prover with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use meituan-longcat/LongCat-Flash-Prover with Transformers:

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

pipe = pipeline("text-generation", model="meituan-longcat/LongCat-Flash-Prover", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import LongcatCausalLM
model = LongcatCausalLM.from_pretrained("meituan-longcat/LongCat-Flash-Prover", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use meituan-longcat/LongCat-Flash-Prover with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/meituan-longcat/LongCat-Flash-Prover

SGLang

How to use meituan-longcat/LongCat-Flash-Prover 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 "meituan-longcat/LongCat-Flash-Prover" \
    --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": "meituan-longcat/LongCat-Flash-Prover",
		"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 "meituan-longcat/LongCat-Flash-Prover" \
        --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": "meituan-longcat/LongCat-Flash-Prover",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use meituan-longcat/LongCat-Flash-Prover with Docker Model Runner:
```
docker model run hf.co/meituan-longcat/LongCat-Flash-Prover
```

Update README.md

by maternion - opened Mar 24

base: refs/heads/main

←

from: refs/pr/2

Discussion Files changed

-1

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -52,7 +52,7 @@ tags:
 ## Introduction
-We introduce **LongCat-Flash-Prover**, a flagship $560$-billion-parameter open-source Mixture-of-Experts (MoE) model that advances Native Formal Reasoning in Lean4 through agentic tool-integrated reasoning (TIR).
 We decompose the native formal reasoning task into three independent formal capabilities, i.e., auto-formalization, sketching, and proving.
 To facilitate these capabilities, we propose a Hybrid-Experts Iteration Framework to expand high-quality task trajectories, including generating a formal statement based on a given informal problem, producing a whole-proof directly from the statement, or a lemma-style sketch.
 During agentic RL, we present a Hierarchical Importance Sampling Policy Optimization (HisPO) algorithm, which aims to stabilize the MoE model training on such long-horizon tasks. It employs a gradient masking strategy that accounts for the policy staleness and the inherent train-inference engine discrepancies at both sequence and token levels.

 ## Introduction
+We introduce **LongCat-Flash-Prover**, a flagship 560-billion-parameter open-source Mixture-of-Experts (MoE) model that advances Native Formal Reasoning in Lean4 through agentic tool-integrated reasoning (TIR).
 We decompose the native formal reasoning task into three independent formal capabilities, i.e., auto-formalization, sketching, and proving.
 To facilitate these capabilities, we propose a Hybrid-Experts Iteration Framework to expand high-quality task trajectories, including generating a formal statement based on a given informal problem, producing a whole-proof directly from the statement, or a lemma-style sketch.
 During agentic RL, we present a Hierarchical Importance Sampling Policy Optimization (HisPO) algorithm, which aims to stabilize the MoE model training on such long-horizon tasks. It employs a gradient masking strategy that accounts for the policy staleness and the inherent train-inference engine discrepancies at both sequence and token levels.