from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model = AutoModelForCausalLM.from_pretrained(
    "ttlynne/qwen3vl-4b-gspo-merged",
    trust_remote_code=True,
    torch_dtype=torch.bfloat16,
    device_map="auto"
)

tokenizer = AutoTokenizer.from_pretrained(
    "ttlynne/qwen3vl-4b-gspo-merged",
    trust_remote_code=True
)

# Example inference
messages = [
    {"role": "user", "content": "Solve: 2x + 5 = 13"}
]

text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)

inputs = tokenizer([text], return_tensors="pt").to(model.device)

outputs = model.generate(
    **inputs,
    max_new_tokens=512,
    temperature=0.7
)

print(tokenizer.decode(outputs[0], skip_special_tokens=True))

📈 Performance

(Add your evaluation results here)

🙏 Acknowledgements

Base model: Qwen Team
Training framework: veRL