How to use wjldw/ToolPRM-Base-v3 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="wjldw/ToolPRM-Base-v3") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)
# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("wjldw/ToolPRM-Base-v3") model = AutoModelForCausalLM.from_pretrained("wjldw/ToolPRM-Base-v3") 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]:]))
How to use wjldw/ToolPRM-Base-v3 with vLLM:
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "wjldw/ToolPRM-Base-v3" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "wjldw/ToolPRM-Base-v3", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'
docker model run hf.co/wjldw/ToolPRM-Base-v3
How to use wjldw/ToolPRM-Base-v3 with SGLang:
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "wjldw/ToolPRM-Base-v3" \ --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": "wjldw/ToolPRM-Base-v3", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'
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 "wjldw/ToolPRM-Base-v3" \ --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": "wjldw/ToolPRM-Base-v3", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'
How to use wjldw/ToolPRM-Base-v3 with Docker Model Runner:
