Instructions to use jorbix/Qwen2.5-1.5B-Open-R1-GRPO with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use jorbix/Qwen2.5-1.5B-Open-R1-GRPO with Transformers:

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

pipe = pipeline("text-generation", model="jorbix/Qwen2.5-1.5B-Open-R1-GRPO")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("jorbix/Qwen2.5-1.5B-Open-R1-GRPO")
model = AutoModelForCausalLM.from_pretrained("jorbix/Qwen2.5-1.5B-Open-R1-GRPO", 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 jorbix/Qwen2.5-1.5B-Open-R1-GRPO with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "jorbix/Qwen2.5-1.5B-Open-R1-GRPO"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "jorbix/Qwen2.5-1.5B-Open-R1-GRPO",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/jorbix/Qwen2.5-1.5B-Open-R1-GRPO

SGLang

How to use jorbix/Qwen2.5-1.5B-Open-R1-GRPO 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 "jorbix/Qwen2.5-1.5B-Open-R1-GRPO" \
    --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": "jorbix/Qwen2.5-1.5B-Open-R1-GRPO",
		"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 "jorbix/Qwen2.5-1.5B-Open-R1-GRPO" \
        --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": "jorbix/Qwen2.5-1.5B-Open-R1-GRPO",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use jorbix/Qwen2.5-1.5B-Open-R1-GRPO with Docker Model Runner:
```
docker model run hf.co/jorbix/Qwen2.5-1.5B-Open-R1-GRPO
```

Qwen2.5-1.5B-Open-R1-GRPO

File size: 3,985 Bytes

6f31dab

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8888888888888888,
  "eval_steps": 500,
  "global_step": 6,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio": 0.0,
      "completion_length": 49.59999942779541,
      "epoch": 0.14814814814814814,
      "grad_norm": 0.796875,
      "kl": 0.0,
      "learning_rate": 2e-05,
      "loss": 0.0275,
      "reward": 0.008593750128056854,
      "reward_std": 0.034375000512227416,
      "rewards/format_reward": 0.0,
      "rewards/reward_asserty_verify_with_z3_v2": 0.007812500116415322,
      "rewards/tag_count_reward": 0.0007812500116415322,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 47.14687633514404,
      "epoch": 0.2962962962962963,
      "grad_norm": 2.390625,
      "kl": 0.0,
      "learning_rate": 1.8090169943749477e-05,
      "loss": 0.0268,
      "reward": 0.03593750135041773,
      "reward_std": 0.11172563303261995,
      "rewards/format_reward": 0.0,
      "rewards/reward_asserty_verify_with_z3_v2": 0.032812500139698386,
      "rewards/tag_count_reward": 0.0031250000465661287,
      "step": 2
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 43.42812538146973,
      "epoch": 0.4444444444444444,
      "grad_norm": 112.5,
      "kl": 5.2669981345534325,
      "learning_rate": 1.3090169943749475e-05,
      "loss": 0.1044,
      "reward": 0.2042931616306305,
      "reward_std": 0.15884361974895,
      "rewards/format_reward": 0.0,
      "rewards/reward_asserty_verify_with_z3_v2": 0.1925744116306305,
      "rewards/tag_count_reward": 0.011718750291038305,
      "step": 3
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 248.70000076293945,
      "epoch": 0.5925925925925926,
      "grad_norm": 54.75,
      "kl": 1.006903812289238,
      "learning_rate": 6.909830056250527e-06,
      "loss": -0.018,
      "reward": 0.17103422805666924,
      "reward_std": 0.2909920923411846,
      "rewards/format_reward": 0.015625000232830644,
      "rewards/reward_asserty_verify_with_z3_v2": 0.08119047619402409,
      "rewards/tag_count_reward": 0.07421874813735485,
      "step": 4
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 372.0031318664551,
      "epoch": 0.7407407407407407,
      "grad_norm": 7.4375,
      "kl": 0.1897900253534317,
      "learning_rate": 1.9098300562505266e-06,
      "loss": -0.1511,
      "reward": 0.18147081695497036,
      "reward_std": 0.2571421265602112,
      "rewards/format_reward": 0.009375000139698386,
      "rewards/reward_asserty_verify_with_z3_v2": 0.1197520662099123,
      "rewards/tag_count_reward": 0.052343751303851604,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 331.5047149658203,
      "epoch": 0.8888888888888888,
      "grad_norm": 2.109375,
      "kl": 0.14362791739404202,
      "learning_rate": 0.0,
      "loss": -0.0532,
      "reward": 0.1194010404869914,
      "reward_std": 0.2147055435925722,
      "rewards/format_reward": 0.0031250000465661287,
      "rewards/reward_asserty_verify_with_z3_v2": 0.049869792885147035,
      "rewards/tag_count_reward": 0.06640625,
      "step": 6
    },
    {
      "epoch": 0.8888888888888888,
      "step": 6,
      "total_flos": 0.0,
      "train_loss": -0.010596248631676039,
      "train_runtime": 239.2366,
      "train_samples_per_second": 0.556,
      "train_steps_per_second": 0.025
    }
  ],
  "logging_steps": 1,
  "max_steps": 6,
  "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": 0.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}