Instructions to use RealSafe/RealSafe-R1-32B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use RealSafe/RealSafe-R1-32B with Transformers:

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

pipe = pipeline("text-generation", model="RealSafe/RealSafe-R1-32B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("RealSafe/RealSafe-R1-32B")
model = AutoModelForCausalLM.from_pretrained("RealSafe/RealSafe-R1-32B")
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

vLLM

How to use RealSafe/RealSafe-R1-32B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "RealSafe/RealSafe-R1-32B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "RealSafe/RealSafe-R1-32B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/RealSafe/RealSafe-R1-32B

SGLang

How to use RealSafe/RealSafe-R1-32B 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 "RealSafe/RealSafe-R1-32B" \
    --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": "RealSafe/RealSafe-R1-32B",
		"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 "RealSafe/RealSafe-R1-32B" \
        --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": "RealSafe/RealSafe-R1-32B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use RealSafe/RealSafe-R1-32B with Docker Model Runner:
```
docker model run hf.co/RealSafe/RealSafe-R1-32B
```

RealSafe-R1-32B

File size: 1,870 Bytes

549d19c

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9846153846153847,
  "eval_steps": 500,
  "global_step": 56,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.17582417582417584,
      "grad_norm": 2.1388941533505665,
      "learning_rate": 4.921457902821578e-06,
      "loss": 1.154,
      "step": 10
    },
    {
      "epoch": 0.3516483516483517,
      "grad_norm": 1.1259907080099065,
      "learning_rate": 4.093559974371725e-06,
      "loss": 0.8713,
      "step": 20
    },
    {
      "epoch": 0.5274725274725275,
      "grad_norm": 0.6382059432052274,
      "learning_rate": 2.6569762988232838e-06,
      "loss": 0.762,
      "step": 30
    },
    {
      "epoch": 0.7032967032967034,
      "grad_norm": 0.5375312211343909,
      "learning_rate": 1.160433012552508e-06,
      "loss": 0.7202,
      "step": 40
    },
    {
      "epoch": 0.8791208791208791,
      "grad_norm": 0.5032147276883348,
      "learning_rate": 1.7555878527937164e-07,
      "loss": 0.6998,
      "step": 50
    },
    {
      "epoch": 0.9846153846153847,
      "step": 56,
      "total_flos": 55816833990656.0,
      "train_loss": 0.8266126939228603,
      "train_runtime": 16924.1422,
      "train_samples_per_second": 0.86,
      "train_steps_per_second": 0.003
    }
  ],
  "logging_steps": 10,
  "max_steps": 56,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 55816833990656.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}