Instructions to use SpyRL/SpyRL-Qwen3-4B-Math with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use SpyRL/SpyRL-Qwen3-4B-Math with Transformers:

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

pipe = pipeline("text-generation", model="SpyRL/SpyRL-Qwen3-4B-Math")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("SpyRL/SpyRL-Qwen3-4B-Math")
model = AutoModelForCausalLM.from_pretrained("SpyRL/SpyRL-Qwen3-4B-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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use SpyRL/SpyRL-Qwen3-4B-Math with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "SpyRL/SpyRL-Qwen3-4B-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": "SpyRL/SpyRL-Qwen3-4B-Math",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/SpyRL/SpyRL-Qwen3-4B-Math

SGLang

How to use SpyRL/SpyRL-Qwen3-4B-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 "SpyRL/SpyRL-Qwen3-4B-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": "SpyRL/SpyRL-Qwen3-4B-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 "SpyRL/SpyRL-Qwen3-4B-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": "SpyRL/SpyRL-Qwen3-4B-Math",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use SpyRL/SpyRL-Qwen3-4B-Math with Docker Model Runner:
```
docker model run hf.co/SpyRL/SpyRL-Qwen3-4B-Math
```

Add link to paper

by nielsr HF Staff - opened about 21 hours ago

base: refs/heads/main

←

from: refs/pr/1

Discussion Files changed

+13

-11

Files changed (1) hide show

README.md +13 -11

README.md CHANGED Viewed

@@ -1,21 +1,23 @@
 ---
-license: apache-2.0
 base_model: Qwen/Qwen3-4B-Instruct-2507
-pipeline_tag: text-generation
-library_name: transformers
 language:
-  - en
 tags:
-  - spyrl
-  - rlsvr
-  - self-play
-  - reinforcement-learning
-  - mathematical-reasoning
-  - reasoning
 ---
 # SpyRL-Qwen3-4B-Math
 Qwen3-4B-Instruct-2507 trained with SpyRL on mathematical reasoning. In the performing stage each agent designs *and* solves a problem grounded in a math-heavy web document; the spy sees the same document with a contiguous 40% span masked out.
 Trained with **SpyRL**, the reference implementation of **RLSVR** (Reinforcement Learning with Self-Verifiable Rewards) from the COLM 2026 paper *From RLVR to RLSVR: Task Transformation Induces Self-Verifiable Rewards for Open-Ended LLM Self-Improvement*.
@@ -89,4 +91,4 @@ bash spyrl/train_math_reasoning.sh
     booktitle = {Conference on Language Modeling (COLM)},
     year      = {2026}
 }
-```

 ---
 base_model: Qwen/Qwen3-4B-Instruct-2507
 language:
+- en
+library_name: transformers
+license: apache-2.0
+pipeline_tag: text-generation
 tags:
+- spyrl
+- rlsvr
+- self-play
+- reinforcement-learning
+- mathematical-reasoning
+- reasoning
 ---
 # SpyRL-Qwen3-4B-Math
+This model was presented in the paper [From RLVR to RLSVR: Task Transformation Induces Self-Verifiable Rewards for Open-Ended LLM Self-Improvement](https://huggingface.co/papers/2607.23802).
 Qwen3-4B-Instruct-2507 trained with SpyRL on mathematical reasoning. In the performing stage each agent designs *and* solves a problem grounded in a math-heavy web document; the spy sees the same document with a contiguous 40% span masked out.
 Trained with **SpyRL**, the reference implementation of **RLSVR** (Reinforcement Learning with Self-Verifiable Rewards) from the COLM 2026 paper *From RLVR to RLSVR: Task Transformation Induces Self-Verifiable Rewards for Open-Ended LLM Self-Improvement*.
     booktitle = {Conference on Language Modeling (COLM)},
     year      = {2026}
 }
+```