Release ClickContrast-trained SelectGround-8B

README.md

@@ -0,0 +1,52 @@
+---
+license: apache-2.0
+library_name: peft
+base_model: Qwen/Qwen3-VL-8B-Instruct
+base_model_relation: adapter
+datasets:
+- ruotian/ClickContrast
+pipeline_tag: image-text-to-text
+tags:
+- gui-grounding
+- computer-use
+- qwen3-vl
+- lora
+- selectground
+---
+
+# SelectGround-8B
+
+SelectGround maps a screenshot and instruction to one click.
+It learns from target–distractor pairs using coordinate supervision and an auxiliary attention-based selection loss.
+Latent Competitor Revisit (LCR) reuses the learned selector to revisit competing regions and compare independently decoded coordinates.
+
+| Inference | ScreenSpot-Pro | MMBench-GUI L2 | OSWorld-G |
+|---|---:|---:|---:|
+| Direct | 66.034 | 86.283 | 70.196 |
+| LCR | 73.182 | 88.008 | 71.961 |
+
+Results use all 1,581 ScreenSpot-Pro and 3,594 MMBench-GUI L2 examples, and the 510 target-bearing OSWorld-G examples.
+The LCR comparison weights are fixed per benchmark and documented in the code repository.
+
+## Contents
+
+This release contains the LoRA adapter and the auxiliary selection head.
+The inference code loads them on **Qwen/Qwen3-VL-8B-Instruct** at revision **0c351dd01ed87e9c1b53cbc748cba10e6187ff3b**.
+It uses the backbone's frozen visual encoder and merger.
+This is one directly trained checkpoint, not an average of checkpoints.
+
+## Use and reproduce
+
+See [SelectGround](https://github.com/zhangruotian/SelectGround) for installation, the exact training recipe, direct inference, LCR, and evaluation of all three benchmarks.
+
+~~~bash
+python infer.py --model ruotian/SelectGround-8B --image screenshot.png --instruction "Click the Save button"
+python infer.py --model ruotian/SelectGround-8B --image screenshot.png --instruction "Click the Save button" --lcr
+~~~
+
+Output **point** is in original-image pixels. **normalized_point** uses 0–1000 coordinates.
+Generation is greedy, uses at most 32 tokens, and preserves the image aspect ratio under an 8,847,360-pixel budget.
+
+Training starts from the pinned backbone with seed 20260625 and uses [ClickContrast](https://huggingface.co/datasets/ruotian/ClickContrast).
+The selected recipe uses 110 initial-stage updates followed by 10 refinement updates.
+The code release fixes the full schedules, sampling, loss, and optimizer-state boundaries.

adapter_config.json

@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen3-VL-8B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 128,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 64,
+  "rank_pattern": {},
+  "revision": "0c351dd01ed87e9c1b53cbc748cba10e6187ff3b",
+  "target_modules": [
+    "o_proj",
+    "k_proj",
+    "up_proj",
+    "gate_proj",
+    "q_proj",
+    "v_proj",
+    "down_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37dbcf5e4db0b20d9c90dc42170f33143a5f5c3c4c413ca7878774737bb79cb9
+size 349251816

selection_head.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:31d5b513154c74d88d8af50ddf35a4876ac55b7e7295f2508b68067fd68a10b7
+size 2458