Instructions to use Nickyang/FastCuRL-1.5B-V3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Nickyang/FastCuRL-1.5B-V3 with Transformers:

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

pipe = pipeline("text-generation", model="Nickyang/FastCuRL-1.5B-V3")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Nickyang/FastCuRL-1.5B-V3")
model = AutoModelForCausalLM.from_pretrained("Nickyang/FastCuRL-1.5B-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]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Nickyang/FastCuRL-1.5B-V3 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Nickyang/FastCuRL-1.5B-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": "Nickyang/FastCuRL-1.5B-V3",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Nickyang/FastCuRL-1.5B-V3

SGLang

How to use Nickyang/FastCuRL-1.5B-V3 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 "Nickyang/FastCuRL-1.5B-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": "Nickyang/FastCuRL-1.5B-V3",
		"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 "Nickyang/FastCuRL-1.5B-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": "Nickyang/FastCuRL-1.5B-V3",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Nickyang/FastCuRL-1.5B-V3 with Docker Model Runner:
```
docker model run hf.co/Nickyang/FastCuRL-1.5B-V3
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

FastCuRL-1.5B-V3

FastCuRL Overview

2025-05-23

We release FastCuRL-1.5B-V3 and FastCuRL-1.5B-V2.

2025-03-17

We release FastCuRL-1.5B-Preview, a slow-thinking reasoning model that outperforms the previous SoTA DeepScaleR-1.5B-Preview with 50% training steps! We adapt a novel curriculum-guided iterative lengthening reinforcement learning to the DeepSeek-R1-Distill-Qwen-1.5B and observe continuous performance improvement as training steps increase. To better reproduce our work and advance research progress, we open-source our code, model, and data.

Code: https://github.com/nick7nlp/FastCuRL

2025-03-21

Paper: https://arxiv.org/abs/2503.17287

Key Results

We report Pass@1 accuracy averaged over 16 samples for each problem.

Model	AIME 2024	MATH 500	AMC 2023	Minerva Math	OlympiadBench	Avg.
Qwen2.5-Math-7B-Instruct	13.3	79.8	50.6	34.6	40.7	43.8
rStar-Math-7B	26.7	78.4	47.5	-	47.1	-
Eurus-2-7B-PRIME	26.7	79.2	57.8	38.6	42.1	48.9
Qwen2.5-7B-SimpleRL	26.7	82.4	62.5	39.7	43.3	50.9
DeepSeek-R1-Distill-Qwen-1.5B	28.8	82.8	62.9	26.5	43.3	48.9
Still-1.5B	32.5	84.4	66.7	29.0	45.4	51.6
DeepScaleR-1.5B-Preview	43.1	87.8	73.6	30.2	50.0	57.0
FastCuRL-1.5B-Preview	43.1	88.0	74.2	31.6	50.4	57.5
FastCuRL-1.5B-V2	47.5	89.3	77.0	32.8	53.3	60.0
FastCuRL-1.5B-V3	49.6	90.5	78.5	34.7	54.5	61.6

Training Data

Following DeepScaleR, our training dataset consists of 40,315 unique problem-answer pairs compiled from:

AIME problems (1984-2023)
AMC problems (before 2023)
Omni-MATH dataset
Still dataset

Acknowledgements

Our training experiments are powered by our heavily modified fork of verl and deepscaler.
Our model is trained on top of DeepSeek-R1-Distill-Qwen-1.5B.