Instructions to use michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-GRPO with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-GRPO with Transformers:

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

pipe = pipeline("text-generation", model="michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-GRPO")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-GRPO")
model = AutoModelForCausalLM.from_pretrained("michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-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 michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-GRPO with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-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": "michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-GRPO",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-GRPO

SGLang

How to use michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-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 "michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-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": "michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-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 "michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-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": "michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-GRPO",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-GRPO with Docker Model Runner:
```
docker model run hf.co/michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-GRPO
```

used GRPO on top of Qwen/Qwen2.5-Math-1.5B model on the gsm8k training set

used REINFORCE loss with baseline
used per-sample length normalization for loss aggregation

wandb training run metrics

hyperparameters:

lr = 3e-5
n_grpo_steps = 100
advantage_eps = 1e-6
rollout_batch_size = 256
group_size = 8
gradient_accumulation_steps = 128
epochs_per_rollout_batch = 1
train_batch_size = 256 # on policy since same as rollout_batch_size
use_std_normalization = True

optim = torch.optim.AdamW(
    hf_model.parameters(), lr=lr, weight_decay=0.0, betas=(0.9, 0.95)
)
scheduler = LinearLR(optim, start_factor=1.0, end_factor=0.1, total_iters=n_grpo_steps)

prompt template:

A conversation between User and Assistant. The User asks a question, and the Assistant solves it. The Assistant first thinks about the reasoning process in the mind and then provides the User with the answer. The reasoning process is enclosed within <think> </think> and answer is enclosed within <answer> </answer> tags, respectively, i.e., <think> reasoning process here </think> <answer> answer here </answer>.
User: {question}
Assistant: <think>

reward format checks for proper use of <think></think> and <answer></answer> tags

performance on GSM8K test set

correct format: 1172/1319
correct reward: 966/1319

Downloads last month: 14

Safetensors

Model size

2B params

Tensor type

BF16

Model tree for michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-GRPO

Base model

Qwen/Qwen2.5-1.5B

Finetuned

Qwen/Qwen2.5-Math-1.5B

Finetuned

(208)

this model

michaelbzhu
/

Qwen2.5-Math-1.5B-GSM8K-GRPO

Model tree for michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-GRPO

Dataset used to train michaelbzhu/Qwen2.5-Math-1.5B-GSM8K-GRPO