Instructions to use Ziyao1010/Qwen2.5-1.5B-Open-R1-GRPO-Math with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Ziyao1010/Qwen2.5-1.5B-Open-R1-GRPO-Math with Transformers:

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

pipe = pipeline("text-generation", model="Ziyao1010/Qwen2.5-1.5B-Open-R1-GRPO-Math")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Ziyao1010/Qwen2.5-1.5B-Open-R1-GRPO-Math")
model = AutoModelForCausalLM.from_pretrained("Ziyao1010/Qwen2.5-1.5B-Open-R1-GRPO-Math", 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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Ziyao1010/Qwen2.5-1.5B-Open-R1-GRPO-Math with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Ziyao1010/Qwen2.5-1.5B-Open-R1-GRPO-Math"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Ziyao1010/Qwen2.5-1.5B-Open-R1-GRPO-Math",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Ziyao1010/Qwen2.5-1.5B-Open-R1-GRPO-Math

SGLang

How to use Ziyao1010/Qwen2.5-1.5B-Open-R1-GRPO-Math 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 "Ziyao1010/Qwen2.5-1.5B-Open-R1-GRPO-Math" \
    --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": "Ziyao1010/Qwen2.5-1.5B-Open-R1-GRPO-Math",
		"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 "Ziyao1010/Qwen2.5-1.5B-Open-R1-GRPO-Math" \
        --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": "Ziyao1010/Qwen2.5-1.5B-Open-R1-GRPO-Math",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Ziyao1010/Qwen2.5-1.5B-Open-R1-GRPO-Math with Docker Model Runner:
```
docker model run hf.co/Ziyao1010/Qwen2.5-1.5B-Open-R1-GRPO-Math
```

Qwen2.5-1.5B-Open-R1-GRPO-Math / trainer_state.json

Ziyao1010

Model save

79bddb9 verified about 1 year ago

Raw History Blame Contribute Delete

3.2 kB

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9230769230769231,
  "eval_steps": 500,
  "global_step": 3,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": -5.921875,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1020.0,
      "completions/mean_length": 674.978515625,
      "completions/mean_terminated_length": 620.6162719726562,
      "completions/min_length": 44.0,
      "completions/min_terminated_length": 44.0,
      "epoch": 0.3076923076923077,
      "frac_reward_zero_std": 0.28125,
      "grad_norm": 0.24190807765558528,
      "kl": 0.00023603439331054688,
      "learning_rate": 0.0,
      "loss": 0.0155,
      "num_tokens": 417029.0,
      "reward": 0.06982421875,
      "reward_std": 0.17549802362918854,
      "rewards/accuracy_reward/mean": 0.05078125,
      "rewards/accuracy_reward/std": 0.21976542472839355,
      "rewards/format_reward/mean": 0.001953125,
      "rewards/format_reward/std": 0.04419417306780815,
      "rewards/tag_count_reward/mean": 0.01708984375,
      "rewards/tag_count_reward/std": 0.10527165234088898,
      "step": 1
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": -6.3671875,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1023.0,
      "completions/mean_length": 652.9970703125,
      "completions/mean_terminated_length": 621.1537475585938,
      "completions/min_length": 82.5,
      "completions/min_terminated_length": 82.5,
      "epoch": 0.9230769230769231,
      "frac_reward_zero_std": 0.171875,
      "kl": 0.006235480308532715,
      "num_tokens": 1242274.0,
      "reward": 0.33154296875,
      "reward_std": 0.2335953712463379,
      "rewards/accuracy_reward/mean": 0.0263671875,
      "rewards/accuracy_reward/std": 0.1589299514889717,
      "rewards/format_reward/mean": 0.009765625,
      "rewards/format_reward/std": 0.09631745889782906,
      "rewards/tag_count_reward/mean": 0.29541015625,
      "rewards/tag_count_reward/std": 0.2025153785943985,
      "step": 3,
      "total_flos": 0.0,
      "train_loss": 0.011118112752834955,
      "train_runtime": 226.5086,
      "train_samples_per_second": 0.441,
      "train_steps_per_second": 0.018
    }
  ],
  "logging_steps": 10,
  "max_steps": 4,
  "num_input_tokens_seen": 1242274,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}

	{
	"best_global_step": null,
	"best_metric": null,
	"best_model_checkpoint": null,
	"epoch": 0.9230769230769231,
	"eval_steps": 500,
	"global_step": 3,
	"is_hyper_param_search": false,
	"is_local_process_zero": true,
	"is_world_process_zero": true,
	"log_history": [
	{
	"clip_ratio/high_max": 0.0,
	"clip_ratio/high_mean": 0.0,
	"clip_ratio/low_mean": 0.0,
	"clip_ratio/low_min": 0.0,
	"clip_ratio/region_mean": 0.0,
	"completions/clipped_ratio": -5.921875,
	"completions/max_length": 1024.0,
	"completions/max_terminated_length": 1020.0,
	"completions/mean_length": 674.978515625,
	"completions/mean_terminated_length": 620.6162719726562,
	"completions/min_length": 44.0,
	"completions/min_terminated_length": 44.0,
	"epoch": 0.3076923076923077,
	"frac_reward_zero_std": 0.28125,
	"grad_norm": 0.24190807765558528,
	"kl": 0.00023603439331054688,
	"learning_rate": 0.0,
	"loss": 0.0155,
	"num_tokens": 417029.0,
	"reward": 0.06982421875,
	"reward_std": 0.17549802362918854,
	"rewards/accuracy_reward/mean": 0.05078125,
	"rewards/accuracy_reward/std": 0.21976542472839355,
	"rewards/format_reward/mean": 0.001953125,
	"rewards/format_reward/std": 0.04419417306780815,
	"rewards/tag_count_reward/mean": 0.01708984375,
	"rewards/tag_count_reward/std": 0.10527165234088898,
	"step": 1
	},
	{
	"clip_ratio/high_max": 0.0,
	"clip_ratio/high_mean": 0.0,
	"clip_ratio/low_mean": 0.0,
	"clip_ratio/low_min": 0.0,
	"clip_ratio/region_mean": 0.0,
	"completions/clipped_ratio": -6.3671875,
	"completions/max_length": 1024.0,
	"completions/max_terminated_length": 1023.0,
	"completions/mean_length": 652.9970703125,
	"completions/mean_terminated_length": 621.1537475585938,
	"completions/min_length": 82.5,
	"completions/min_terminated_length": 82.5,
	"epoch": 0.9230769230769231,
	"frac_reward_zero_std": 0.171875,
	"kl": 0.006235480308532715,
	"num_tokens": 1242274.0,
	"reward": 0.33154296875,
	"reward_std": 0.2335953712463379,
	"rewards/accuracy_reward/mean": 0.0263671875,
	"rewards/accuracy_reward/std": 0.1589299514889717,
	"rewards/format_reward/mean": 0.009765625,
	"rewards/format_reward/std": 0.09631745889782906,
	"rewards/tag_count_reward/mean": 0.29541015625,
	"rewards/tag_count_reward/std": 0.2025153785943985,
	"step": 3,
	"total_flos": 0.0,
	"train_loss": 0.011118112752834955,
	"train_runtime": 226.5086,
	"train_samples_per_second": 0.441,
	"train_steps_per_second": 0.018
	}
	],
	"logging_steps": 10,
	"max_steps": 4,
	"num_input_tokens_seen": 1242274,
	"num_train_epochs": 1,
	"save_steps": 500,
	"stateful_callbacks": {
	"TrainerControl": {
	"args": {
	"should_epoch_stop": false,
	"should_evaluate": false,
	"should_log": false,
	"should_save": false,
	"should_training_stop": false
	},
	"attributes": {}
	}
	},
	"total_flos": 0.0,
	"train_batch_size": 16,
	"trial_name": null,
	"trial_params": null
	}