HelloKKMe/grounding-r1-7B

Introduction

Reinforcement learning (RL) (e.g., GRPO) helps with grounding because of its inherent objective alignment—rewarding successful clicks—rather than encouraging long textual Chain-of-Thought (CoT) reasoning. Unlike approaches that rely heavily on verbose CoT reasoning, GRPO directly incentivizes actionable and grounded responses. Based on findings from our blog, we share state-of-the-art GUI grounding models trained using GRPO.

Performance

We follow the standard evaluation protocol and benchmark our model on three challenging datasets. Our method consistently achieves the best results among all open-source model families. Below are the comparative results:

Model	Size	Open Source	ScreenSpot-V2	ScreenSpotPro	OSWORLD-G
OpenAI CUA	—	❌	87.9	23.4	—
Claude 3.7	—	❌	87.6	27.7	—
JEDI-7B	7B	✅	91.7	39.5	54.1
SE-GUI	7B	✅	90.3	47.0	—
UI-TARS	7B	✅	91.6	35.7	47.5
UI-TARS-1.5*	7B	✅	89.7*	42.0*	64.2*
UGround-v1-7B	7B	✅	—	31.1	36.4
Qwen2.5-VL-32B-Instruct	32B	✅	91.9*	48.0	59.6*
UGround-v1-72B	72B	✅	—	34.5	—
Qwen2.5-VL-72B-Instruct	72B	✅	94.00*	53.3	62.2*
UI-TARS	72B	✅	90.3	38.1	—
Grounding-R1 (Ours)	7B	✅	92.4 (∆ +2.7)	50.1(∆ +8.1)	67.7 (∆ +3.5)
Grounding-R1 (Ours)	32B	✅	93.2 (∆ +1.3)	53.6 (∆ +5.6)	61.9(∆ +2.3)
Grounding-R1 (Ours)	72B	✅	94.8(∆ +0.8)	58.4 (∆ +5.1)	66.7(∆ +4.5)

Note:

• Model size is indicated in billions (B) of parameters.
• A dash (—) denotes results that are currently unavailable.
• A superscript asterisk (﹡) denotes our evaluated result.
• UI-TARS-1.5 7B, Qwen2.5-VL-32B-Instruct, and Qwen2.5-VL-72B-Instruct are applied as our baseline models.
• ∆ indicates the performance improvement (∆) of our model compared to its baseline.

Inference

Below is a code snippet demonstrating how to run inference using a trained model.

from PIL import Image
from qwen_vl_utils import process_vision_info, smart_resize
from transformers import Qwen2_5_VLForConditionalGeneration, AutoProcessor
import torch
import re

SYSTEM_PROMPT = '''
You are an expert UI element locator. Given a GUI image and a user's element description, provide the coordinates of the specified element as a single (x,y) point. The image resolution is height {height} and width {width}. For elements with area, return the center point.

Output the coordinate pair exactly:
(x,y)
'''
SYSTEM_PROMPT=SYSTEM_PROMPT.strip()

# Function to extract coordinates from model output
def extract_coordinates(raw_string):
    try:
        matches = re.findall(r"\((-?\d*\.?\d+),\s*(-?\d*\.?\d+)\)", raw_string)
        return [tuple(map(int, match)) for match in matches][0]
    except:
        return 0,0

# Load model and processor
model_path = "HelloKKMe/grounding-r1-7B"
max_new_tokens = 32

model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
    model_path,
    torch_dtype=torch.bfloat16,
    attn_implementation="flash_attention_2",
    device_map="auto"
)
processor = AutoProcessor.from_pretrained(
    model_path,
    min_pixels=3136,
    max_pixels= 4096 * 2160
)

# Load and resize image
image = Image.open("file path")
instruction = "description"  # Instruction for grounding
width, height = image.width, image.height

resized_height, resized_width = smart_resize(
    image.height,
    image.width,
    factor=processor.image_processor.patch_size * processor.image_processor.merge_size,
    min_pixels=processor.image_processor.min_pixels,
    max_pixels=processor.image_processor.max_pixels,
)
resized_image = image.resize((resized_width, resized_height))
scale_x, scale_y = width / resized_width, height / resized_height

# Prepare system and user messages
system_message = {
   "role": "system",
   "content": SYSTEM_PROMPT.format(height=resized_height,width=resized_width)
}

user_message = {
    "role": "user",
    "content": [
        {"type": "image", "image": resized_image},
        {"type": "text", "text": instruction}
    ]
}

# Tokenize and prepare inputs
image_inputs, video_inputs = process_vision_info([system_message, user_message])
text = processor.apply_chat_template([system_message, user_message], tokenize=False, add_generation_prompt=True)
inputs = processor(text=[text], images=image_inputs, videos=video_inputs, padding=True, return_tensors="pt")
inputs = inputs.to(model.device)

# Generate prediction
output_ids = model.generate(**inputs, max_new_tokens=max_new_tokens, do_sample=False, temperature=1.0, use_cache=True)
generated_ids = [output_ids[len(input_ids):] for input_ids, output_ids in zip(inputs.input_ids, output_ids)]
output_text = processor.batch_decode(generated_ids, skip_special_tokens=True, clean_up_tokenization_spaces=True)[0]

# Extract and rescale coordinates
pred_x, pred_y  = extract_coordinates(output_text) 
pred_x*=scale_x
pred_y*=scale_y 
print(pred_x,pred_y)

Refer to our code for more details.