Datasets:

bdatm-project
/

evaluation-results-task2

model_name stringlengths 25 56	exact_match float64 0 0.64	em_extracted float64 0 0.64	same_move float64 0 0.98	legal float64 0.11 0.98	to_sq_match float64 0 0.98	from_sq_match float64 0.01 0.98	piece_match float64 0.03 0.98	multiline float64 0 1	adapter_sha stringlengths 64 64 ⌀	ambiguous float64 0 0.04
Vanilla Qwen2.5-VL (Zero-Shot)	0.0025	0.0025	0.0025	0.1075	0.0025	0.0125	0.03	0.0425	null	0.0025
Qwen + LoRA Fine-Tuning (from HF)	0.62	0.62	0.9775	0.9775	0.9775	0.9775	0.9775	0	24f0a8806dd4f9c7d37cf88d10903577a6ed6b7b22126578a68b3da3ab38a6f7	0.035
Qwen + LoRA (Zigzag - TF)	0.2575	0.2575	0.4675	0.55	0.47	0.5125	0.5475	0	24f0a8806dd4f9c7d37cf88d10903577a6ed6b7b22126578a68b3da3ab38a6f7	0.025
Qwen + LoRA (Spiral - TF)	0.0475	0.0475	0.0825	0.1475	0.0825	0.1325	0.1475	0	24f0a8806dd4f9c7d37cf88d10903577a6ed6b7b22126578a68b3da3ab38a6f7	0.015
Qwen + LoRA (File_wise - TF)	0.0425	0.0425	0.1	0.205	0.1	0.19	0.205	0	24f0a8806dd4f9c7d37cf88d10903577a6ed6b7b22126578a68b3da3ab38a6f7	0.0075
Qwen + LoRA (Zigzag - SFT)	0.5475	0.5475	0.92	0.9225	0.92	0.9225	0.9225	0	67828ae646cad642eb58253cfbaf909fa9fc55f1dcaf37b90ac3eb6a7c07c3d4	0.0375
Qwen + LoRA (Spiral - SFT)	0.5775	0.5775	0.9	0.915	0.905	0.9075	0.91	0	27ae53231a670fe4c6c441f98731e01a9027685c1849018c5369582ab00789e7	0.0325
Qwen + LoRA (File_wise - SFT)	0.64	0.64	0.9775	0.9775	0.9775	0.9775	0.9775	0	0d78c66fadebc9e2578cd027b4f6ee05dbe19c1e2b9981008646fc5440429ed7	0.0375
Qwen + LoRA (Learned Reordering)	0	0.53	0.87	0.88	0.87	0.88	0.88	1	18d0d72c23cbe24a8ef0eb4707655b1b3c617da3467c0df0c12a743299fc4820	0.0375
Vanilla Qwen3.5-0.8B (Zero-Shot) [tok256]	0	0	0	0.13	0	0.005	0.03	0.0025	null	0
Vanilla Qwen3.5-0.8B (3-Shot) [tok256]	0.0075	0.0075	0.0075	0.175	0.01	0.045	0.0575	0	null	0.035
Qwen + LoRA (Raster - SFT) [tok256]	0.6025	0.6025	0.9375	0.95	0.9375	0.95	0.95	0	76f477b50405aba63d65c41c7eef1cbe6f0fd4bfa1b1903a7326d42e5927a6d8	0.0325
Qwen + LoRA Raster, read zigzag (TF) [tok256]	0.5275	0.5275	0.825	0.8525	0.825	0.85	0.8525	0	76f477b50405aba63d65c41c7eef1cbe6f0fd4bfa1b1903a7326d42e5927a6d8	0.0325
Qwen + LoRA Raster, read spiral_center_out (TF) [tok256]	0.145	0.145	0.2025	0.3225	0.2025	0.2925	0.3175	0	76f477b50405aba63d65c41c7eef1cbe6f0fd4bfa1b1903a7326d42e5927a6d8	0.0125
Qwen + LoRA Raster, read file_wise (TF) [tok256]	0.0975	0.0975	0.1525	0.2975	0.1525	0.265	0.295	0	76f477b50405aba63d65c41c7eef1cbe6f0fd4bfa1b1903a7326d42e5927a6d8	0.0075
Qwen + LoRA Raster, read hilbert (TF) [tok256]	0.29	0.29	0.4275	0.535	0.43	0.51	0.5275	0	76f477b50405aba63d65c41c7eef1cbe6f0fd4bfa1b1903a7326d42e5927a6d8	0.0125
Qwen + LoRA (Zigzag - SFT) [tok256]	0.535	0.535	0.895	0.9075	0.895	0.9075	0.9075	0	772c529ff97062446e71d3673ba5bedceca234e0e5346adbecb486afcba10a37	0.0325
Qwen + LoRA (Spiral_center_out - SFT) [tok256]	0.54	0.54	0.825	0.8475	0.8275	0.8425	0.845	0	d2359c1707b3bc8ccdd888b12cbac1e2b796f2ad97468ce11e5946aaa9d1b600	0.0325
Qwen + LoRA (File_wise - SFT) [tok256]	0.5475	0.5475	0.885	0.905	0.885	0.9	0.905	0	8dbb823e14d8eaf39dfa1564f274d32b48565de76eb1767c707b930345633cf8	0.03
Qwen + LoRA (Hilbert - SFT) [tok256]	0.525	0.525	0.8575	0.875	0.8575	0.8725	0.875	0	edb5272751b27e847f6146021db841f7c3ba19d331b34c4840f24ee6b675fdca	0.03
Qwen + LoRA (Learned Reordering) [tok256]	0.5775	0.5775	0.8925	0.8975	0.8925	0.895	0.8975	0	df37d131a44bf72023ae143b1e8cb6f54aff69c3224c1fe1a40827e6175a7d3f	0.035
Qwen + LoRA (Learned Reordering, frozen policy) [tok256]	0.595	0.595	0.8825	0.905	0.8825	0.9025	0.905	0	7f4cfe2689e4795b2c2017ebce16743bf4c6d4caad259b62f5eb720cbcb7a150	0.035

Model-comparison table for this task: one row per evaluated model, written by push_results_table in src/eval/results.py. Per-sample predictions are in per_sample/.

Downloads last month: 164

Number of rows:

Total file size:

485 kB