Add files using upload-large-folder tool

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Camera_Motion.log

@@ -0,0 +1,335 @@
+[2026-06-08 06:08:02 UTC] COMMAND /root/ckpts/micromamba/envs/vbench2/bin/python -u evaluate.py --videos_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Camera_Motion --dimension Camera_Motion --output_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Camera_Motion --full_json_dir /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json --mode vbench_standard
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Camera_Motion', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Camera_Motion', dimension=['Camera_Motion'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Camera_Motion/results_2026-06-08-06:08:05_full_info.json
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Camera_Motion/results_2026-06-08-06:08:05_full_info.json
+Downloading: "https://github.com/facebookresearch/co-tracker/zipball/main" to /root/ckpts/torch/hub/main.zip
+Downloading: "https://huggingface.co/facebook/cotracker/resolve/main/cotracker2.pth" to /root/ckpts/torch/hub/checkpoints/cotracker2.pth
+
  0%|          | 0.00/195M [00:00<?, ?B/s]
  0%|          | 128k/195M [00:00<03:23, 1.00MB/s]
  6%|▌         | 11.1M/195M [00:00<00:03, 60.6MB/s]
 18%|█▊        | 34.5M/195M [00:00<00:01, 139MB/s] 
 35%|███▌      | 68.8M/195M [00:00<00:00, 222MB/s]
 48%|████▊     | 94.5M/195M [00:00<00:00, 239MB/s]
 62%|██████▏   | 120M/195M [00:00<00:00, 249MB/s] 
 75%|███████▌  | 146M/195M [00:00<00:00, 257MB/s]
 88%|████████▊ | 171M/195M [00:00<00:00, 257MB/s]
+
  0%|          | 0/108 [00:00<?, ?it/s]2026-06-08 06:08:12,169 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:08:16,728 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:08:20,841 - pyscenedetect - INFO - Detecting scenes...
+
  1%|          | 1/108 [00:12<22:55, 12.85s/it]2026-06-08 06:08:25,022 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:08:29,015 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:08:32,990 - pyscenedetect - INFO - Detecting scenes...
+
  2%|▏         | 2/108 [00:24<21:54, 12.40s/it]2026-06-08 06:08:37,115 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:08:41,380 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:08:45,454 - pyscenedetect - INFO - Detecting scenes...
+
  3%|▎         | 3/108 [00:37<21:42, 12.41s/it]2026-06-08 06:08:49,531 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:08:53,694 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:08:57,799 - pyscenedetect - INFO - Detecting scenes...
+
  4%|▎         | 4/108 [00:49<21:25, 12.36s/it]2026-06-08 06:09:01,819 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:09:05,973 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:09:09,969 - pyscenedetect - INFO - Detecting scenes...
+
  5%|▍         | 5/108 [01:01<21:04, 12.27s/it]2026-06-08 06:09:13,934 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:09:18,077 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:09:22,084 - pyscenedetect - INFO - Detecting scenes...
+
  6%|▌         | 6/108 [01:14<21:18, 12.53s/it]2026-06-08 06:09:26,962 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:09:32,024 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:09:35,985 - pyscenedetect - INFO - Detecting scenes...
+
  6%|▋         | 7/108 [01:27<21:21, 12.69s/it]2026-06-08 06:09:39,975 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:09:44,153 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:09:48,121 - pyscenedetect - INFO - Detecting scenes...
+
  7%|▋         | 8/108 [01:39<20:50, 12.51s/it]2026-06-08 06:09:52,084 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:09:56,222 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:10:00,243 - pyscenedetect - INFO - Detecting scenes...
+
  8%|▊         | 9/108 [01:52<20:26, 12.39s/it]2026-06-08 06:10:04,220 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:10:08,216 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:10:12,208 - pyscenedetect - INFO - Detecting scenes...
+
  9%|▉         | 10/108 [02:04<20:03, 12.28s/it]2026-06-08 06:10:16,258 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:10:20,220 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:10:24,190 - pyscenedetect - INFO - Detecting scenes...
+
 10%|█         | 11/108 [02:16<19:40, 12.17s/it]2026-06-08 06:10:28,169 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:10:32,162 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:10:36,257 - pyscenedetect - INFO - Detecting scenes...
+
 11%|█         | 12/108 [02:28<19:26, 12.15s/it]2026-06-08 06:10:40,272 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:10:44,310 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:10:48,443 - pyscenedetect - INFO - Detecting scenes...
+
 12%|█▏        | 13/108 [02:40<19:15, 12.16s/it]2026-06-08 06:10:52,461 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:10:56,581 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:11:00,626 - pyscenedetect - INFO - Detecting scenes...
+
 13%|█▎        | 14/108 [02:52<19:04, 12.17s/it]2026-06-08 06:11:04,673 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:11:08,609 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:11:12,620 - pyscenedetect - INFO - Detecting scenes...
+
 14%|█▍        | 15/108 [03:04<18:44, 12.10s/it]2026-06-08 06:11:16,579 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:11:20,517 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:11:24,672 - pyscenedetect - INFO - Detecting scenes...
+
 15%|█▍        | 16/108 [03:16<18:32, 12.10s/it]2026-06-08 06:11:28,682 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:11:32,649 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:11:36,678 - pyscenedetect - INFO - Detecting scenes...
+
 16%|█▌        | 17/108 [03:28<18:20, 12.09s/it]2026-06-08 06:11:40,769 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:11:44,744 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:11:48,953 - pyscenedetect - INFO - Detecting scenes...
+
 17%|█▋        | 18/108 [03:40<18:09, 12.10s/it]2026-06-08 06:11:52,890 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:11:56,850 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:12:00,782 - pyscenedetect - INFO - Detecting scenes...
+
 18%|█▊        | 19/108 [03:52<17:50, 12.03s/it]2026-06-08 06:12:04,757 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:12:08,647 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:12:12,544 - pyscenedetect - INFO - Detecting scenes...
+
 19%|█▊        | 20/108 [04:04<17:33, 11.98s/it]2026-06-08 06:12:16,603 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:12:20,515 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:12:24,490 - pyscenedetect - INFO - Detecting scenes...
+
 19%|█▉        | 21/108 [04:17<17:37, 12.16s/it]2026-06-08 06:12:29,192 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:12:34,172 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:12:38,939 - pyscenedetect - INFO - Detecting scenes...
+
 20%|██        | 22/108 [04:31<18:19, 12.79s/it]2026-06-08 06:12:43,453 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:12:49,060 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:12:53,676 - pyscenedetect - INFO - Detecting scenes...
+
 21%|██▏       | 23/108 [04:46<19:04, 13.46s/it]2026-06-08 06:12:58,472 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:13:03,433 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:13:08,141 - pyscenedetect - INFO - Detecting scenes...
+
 22%|██▏       | 24/108 [05:01<19:25, 13.87s/it]2026-06-08 06:13:13,310 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:13:18,379 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:13:23,152 - pyscenedetect - INFO - Detecting scenes...
+
 23%|██▎       | 25/108 [05:15<19:29, 14.09s/it]2026-06-08 06:13:27,914 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:13:32,612 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:13:37,297 - pyscenedetect - INFO - Detecting scenes...
+
 24%|██▍       | 26/108 [05:29<19:14, 14.08s/it]2026-06-08 06:13:41,979 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:13:47,090 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:13:52,655 - pyscenedetect - INFO - Detecting scenes...
+
 25%|██▌       | 27/108 [05:45<19:33, 14.49s/it]2026-06-08 06:13:57,409 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:14:01,793 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:14:06,368 - pyscenedetect - INFO - Detecting scenes...
+
 26%|██▌       | 28/108 [05:58<18:59, 14.25s/it]2026-06-08 06:14:11,096 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:14:16,094 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:14:21,190 - pyscenedetect - INFO - Detecting scenes...
+
 27%|██▋       | 29/108 [06:14<19:11, 14.58s/it]2026-06-08 06:14:26,436 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:14:31,213 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:14:36,137 - pyscenedetect - INFO - Detecting scenes...
+
 28%|██▊       | 30/108 [06:29<19:14, 14.80s/it]2026-06-08 06:14:41,752 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:14:46,881 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:14:51,575 - pyscenedetect - INFO - Detecting scenes...
+
 29%|██▊       | 31/108 [06:43<18:49, 14.67s/it]2026-06-08 06:14:56,119 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:15:01,049 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:15:05,721 - pyscenedetect - INFO - Detecting scenes...
+
 30%|██▉       | 32/108 [06:58<18:28, 14.59s/it]2026-06-08 06:15:10,533 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:15:15,309 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:15:20,031 - pyscenedetect - INFO - Detecting scenes...
+
 31%|███       | 33/108 [07:13<18:15, 14.61s/it]2026-06-08 06:15:25,168 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:15:30,452 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:15:34,957 - pyscenedetect - INFO - Detecting scenes...
+
 31%|███▏      | 34/108 [07:27<17:55, 14.54s/it]2026-06-08 06:15:39,539 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:15:44,731 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:15:49,637 - pyscenedetect - INFO - Detecting scenes...
+
 32%|███▏      | 35/108 [07:42<17:59, 14.79s/it]2026-06-08 06:15:54,938 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:15:59,846 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:16:04,766 - pyscenedetect - INFO - Detecting scenes...
+
 33%|███▎      | 36/108 [07:57<17:47, 14.83s/it]2026-06-08 06:16:09,837 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:16:15,582 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:16:21,126 - pyscenedetect - INFO - Detecting scenes...
+
 34%|███▍      | 37/108 [08:14<18:19, 15.48s/it]2026-06-08 06:16:26,845 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:16:31,741 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:16:36,791 - pyscenedetect - INFO - Detecting scenes...
+
 35%|███▌      | 38/108 [08:29<17:45, 15.23s/it]2026-06-08 06:16:41,473 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:16:46,320 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:16:51,107 - pyscenedetect - INFO - Detecting scenes...
+
 36%|███▌      | 39/108 [08:43<17:15, 15.00s/it]2026-06-08 06:16:55,954 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:17:00,799 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:17:05,885 - pyscenedetect - INFO - Detecting scenes...
+
 37%|███▋      | 40/108 [08:58<16:55, 14.93s/it]2026-06-08 06:17:10,712 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:17:15,444 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:17:20,620 - pyscenedetect - INFO - Detecting scenes...
+
 38%|███▊      | 41/108 [09:12<16:30, 14.79s/it]2026-06-08 06:17:25,169 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:17:30,588 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:17:35,224 - pyscenedetect - INFO - Detecting scenes...
+
 39%|███▉      | 42/108 [09:27<16:14, 14.77s/it]2026-06-08 06:17:39,897 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:17:44,574 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:17:50,171 - pyscenedetect - INFO - Detecting scenes...
+
 40%|███▉      | 43/108 [09:42<16:00, 14.78s/it]2026-06-08 06:17:54,693 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:17:59,600 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:18:04,281 - pyscenedetect - INFO - Detecting scenes...
+
 41%|████      | 44/108 [09:57<15:41, 14.71s/it]2026-06-08 06:18:09,263 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:18:14,815 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:18:19,388 - pyscenedetect - INFO - Detecting scenes...
+
 42%|████▏     | 45/108 [10:11<15:29, 14.75s/it]2026-06-08 06:18:24,105 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:18:28,181 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:18:32,377 - pyscenedetect - INFO - Detecting scenes...
+
 43%|████▎     | 46/108 [10:24<14:32, 14.07s/it]2026-06-08 06:18:36,607 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:18:40,591 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:18:44,580 - pyscenedetect - INFO - Detecting scenes...
+
 44%|████▎     | 47/108 [10:36<13:41, 13.47s/it]2026-06-08 06:18:48,653 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:18:52,730 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:18:56,816 - pyscenedetect - INFO - Detecting scenes...
+
 44%|████▍     | 48/108 [10:48<13:08, 13.14s/it]2026-06-08 06:19:01,028 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:19:05,241 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:19:09,279 - pyscenedetect - INFO - Detecting scenes...
+
 45%|████▌     | 49/108 [11:01<12:39, 12.88s/it]2026-06-08 06:19:13,287 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:19:17,256 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:19:21,150 - pyscenedetect - INFO - Detecting scenes...
+
 46%|████▋     | 50/108 [11:13<12:09, 12.58s/it]2026-06-08 06:19:25,171 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:19:29,276 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:19:33,245 - pyscenedetect - INFO - Detecting scenes...
+
 47%|████▋     | 51/108 [11:25<11:50, 12.46s/it]2026-06-08 06:19:37,361 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:19:41,473 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:19:45,494 - pyscenedetect - INFO - Detecting scenes...
+
 48%|████▊     | 52/108 [11:37<11:32, 12.36s/it]2026-06-08 06:19:49,473 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:19:53,536 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:19:57,597 - pyscenedetect - INFO - Detecting scenes...
+
 49%|████▉     | 53/108 [11:49<11:15, 12.28s/it]2026-06-08 06:20:01,579 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:20:05,513 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:20:09,522 - pyscenedetect - INFO - Detecting scenes...
+
 50%|█████     | 54/108 [12:01<10:58, 12.19s/it]2026-06-08 06:20:13,560 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:20:17,596 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:20:21,674 - pyscenedetect - INFO - Detecting scenes...
+
 51%|█████     | 55/108 [12:13<10:45, 12.18s/it]2026-06-08 06:20:25,718 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:20:29,734 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:20:33,846 - pyscenedetect - INFO - Detecting scenes...
+
 52%|█████▏    | 56/108 [12:25<10:33, 12.19s/it]2026-06-08 06:20:37,919 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:20:41,951 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:20:46,018 - pyscenedetect - INFO - Detecting scenes...
+
 53%|█████▎    | 57/108 [12:37<10:20, 12.17s/it]2026-06-08 06:20:50,047 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:20:54,017 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:20:58,043 - pyscenedetect - INFO - Detecting scenes...
+
 54%|█████▎    | 58/108 [12:49<10:06, 12.14s/it]2026-06-08 06:21:02,100 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:21:06,192 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:21:10,224 - pyscenedetect - INFO - Detecting scenes...
+
 55%|█████▍    | 59/108 [13:02<09:58, 12.22s/it]2026-06-08 06:21:14,521 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:21:18,515 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:21:22,508 - pyscenedetect - INFO - Detecting scenes...
+
 56%|█████▌    | 60/108 [13:14<09:45, 12.21s/it]2026-06-08 06:21:26,696 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:21:30,624 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:21:34,634 - pyscenedetect - INFO - Detecting scenes...
+
 56%|█████▋    | 61/108 [13:26<09:30, 12.14s/it]2026-06-08 06:21:38,692 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:21:42,807 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:21:46,801 - pyscenedetect - INFO - Detecting scenes...
+
 57%|█████▋    | 62/108 [13:38<09:19, 12.16s/it]2026-06-08 06:21:50,881 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:21:55,692 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:22:00,503 - pyscenedetect - INFO - Detecting scenes...
+
 58%|█████▊    | 63/108 [13:52<09:28, 12.64s/it]2026-06-08 06:22:04,660 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:22:08,686 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:22:12,681 - pyscenedetect - INFO - Detecting scenes...
+
 59%|█████▉    | 64/108 [14:04<09:10, 12.50s/it]2026-06-08 06:22:16,838 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:22:20,914 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:22:24,894 - pyscenedetect - INFO - Detecting scenes...
+
 60%|██████    | 65/108 [14:16<08:53, 12.40s/it]2026-06-08 06:22:28,983 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:22:33,025 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:22:37,062 - pyscenedetect - INFO - Detecting scenes...
+
 61%|██████    | 66/108 [14:28<08:37, 12.31s/it]2026-06-08 06:22:41,098 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:22:45,088 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:22:49,157 - pyscenedetect - INFO - Detecting scenes...
+
 62%|██████▏   | 67/108 [14:41<08:25, 12.32s/it]2026-06-08 06:22:53,434 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:22:58,305 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:23:03,120 - pyscenedetect - INFO - Detecting scenes...
+
 63%|██████▎   | 68/108 [14:55<08:40, 13.02s/it]2026-06-08 06:23:08,092 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:23:13,085 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:23:18,279 - pyscenedetect - INFO - Detecting scenes...
+
 64%|██████▍   | 69/108 [15:10<08:49, 13.57s/it]2026-06-08 06:23:22,945 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:23:27,548 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:23:32,072 - pyscenedetect - INFO - Detecting scenes...
+
 65%|██████▍   | 70/108 [15:24<08:41, 13.73s/it]2026-06-08 06:23:37,051 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:23:41,768 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:23:46,512 - pyscenedetect - INFO - Detecting scenes...
+
 66%|██████▌   | 71/108 [15:39<08:32, 13.86s/it]2026-06-08 06:23:51,201 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:23:56,025 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:24:00,890 - pyscenedetect - INFO - Detecting scenes...
+
 67%|██████▋   | 72/108 [15:53<08:24, 14.01s/it]2026-06-08 06:24:05,565 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:24:10,354 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:24:15,595 - pyscenedetect - INFO - Detecting scenes...
+
 68%|██████▊   | 73/108 [16:08<08:21, 14.33s/it]2026-06-08 06:24:20,628 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:24:25,557 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:24:30,516 - pyscenedetect - INFO - Detecting scenes...
+
 69%|██████▊   | 74/108 [16:23<08:16, 14.62s/it]2026-06-08 06:24:35,927 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:24:42,131 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:24:47,750 - pyscenedetect - INFO - Detecting scenes...
+
 69%|██████▉   | 75/108 [16:40<08:26, 15.35s/it]2026-06-08 06:24:52,980 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:24:57,774 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:25:02,793 - pyscenedetect - INFO - Detecting scenes...
+
 70%|███████   | 76/108 [16:55<08:08, 15.26s/it]2026-06-08 06:25:08,021 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:25:12,696 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:25:17,424 - pyscenedetect - INFO - Detecting scenes...
+
 71%|███████▏  | 77/108 [17:09<07:40, 14.87s/it]2026-06-08 06:25:21,991 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:25:27,298 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:25:32,091 - pyscenedetect - INFO - Detecting scenes...
+
 72%|███████▏  | 78/108 [17:24<07:26, 14.88s/it]2026-06-08 06:25:36,885 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:25:42,139 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:25:47,128 - pyscenedetect - INFO - Detecting scenes...
+
 73%|███████▎  | 79/108 [17:39<07:10, 14.84s/it]2026-06-08 06:25:51,629 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:25:56,424 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:26:01,073 - pyscenedetect - INFO - Detecting scenes...
+
 74%|███████▍  | 80/108 [17:53<06:52, 14.74s/it]2026-06-08 06:26:06,152 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:26:10,998 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:26:15,769 - pyscenedetect - INFO - Detecting scenes...
+
 75%|███████▌  | 81/108 [18:08<06:35, 14.65s/it]2026-06-08 06:26:20,590 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:26:25,625 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:26:30,620 - pyscenedetect - INFO - Detecting scenes...
+
 76%|███████▌  | 82/108 [18:23<06:25, 14.84s/it]2026-06-08 06:26:35,870 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:26:40,905 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:26:45,485 - pyscenedetect - INFO - Detecting scenes...
+
 77%|███████▋  | 83/108 [18:37<06:06, 14.67s/it]2026-06-08 06:26:50,144 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:26:55,101 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:27:00,058 - pyscenedetect - INFO - Detecting scenes...
+
 78%|███████▊  | 84/108 [18:53<05:55, 14.82s/it]2026-06-08 06:27:05,318 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:27:09,977 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:27:14,716 - pyscenedetect - INFO - Detecting scenes...
+
 79%|███████▊  | 85/108 [19:07<05:38, 14.71s/it]2026-06-08 06:27:19,761 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:27:25,253 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:27:29,991 - pyscenedetect - INFO - Detecting scenes...
+
 80%|███████▉  | 86/108 [19:22<05:25, 14.81s/it]2026-06-08 06:27:34,800 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:27:39,662 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:27:44,908 - pyscenedetect - INFO - Detecting scenes...
+
 81%|████████  | 87/108 [19:37<05:12, 14.89s/it]2026-06-08 06:27:49,891 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:27:54,689 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:27:59,228 - pyscenedetect - INFO - Detecting scenes...
+
 81%|████████▏ | 88/108 [19:51<04:51, 14.59s/it]2026-06-08 06:28:03,780 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:28:08,512 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:28:13,567 - pyscenedetect - INFO - Detecting scenes...
+
 82%|████████▏ | 89/108 [20:05<04:36, 14.53s/it]2026-06-08 06:28:18,158 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:28:22,908 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:28:28,003 - pyscenedetect - INFO - Detecting scenes...
+
 83%|████████▎ | 90/108 [20:20<04:24, 14.67s/it]2026-06-08 06:28:33,190 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:28:38,096 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:28:43,132 - pyscenedetect - INFO - Detecting scenes...
+
 84%|████████▍ | 91/108 [20:35<04:08, 14.64s/it]2026-06-08 06:28:47,747 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:28:52,175 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:28:56,517 - pyscenedetect - INFO - Detecting scenes...
+
 85%|████████▌ | 92/108 [20:48<03:46, 14.19s/it]2026-06-08 06:29:00,859 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:29:04,962 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:29:09,004 - pyscenedetect - INFO - Detecting scenes...
+
 86%|████████▌ | 93/108 [21:00<03:23, 13.58s/it]2026-06-08 06:29:13,015 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:29:17,165 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:29:21,212 - pyscenedetect - INFO - Detecting scenes...
+
 87%|████████▋ | 94/108 [21:13<03:04, 13.16s/it]2026-06-08 06:29:25,203 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:29:29,229 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:29:33,345 - pyscenedetect - INFO - Detecting scenes...
+
 88%|████████▊ | 95/108 [21:25<02:47, 12.88s/it]2026-06-08 06:29:37,430 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:29:41,537 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:29:45,658 - pyscenedetect - INFO - Detecting scenes...
+
 89%|████████▉ | 96/108 [21:37<02:32, 12.69s/it]2026-06-08 06:29:49,664 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:29:53,704 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:29:57,875 - pyscenedetect - INFO - Detecting scenes...
+
 90%|████████▉ | 97/108 [21:49<02:18, 12.56s/it]2026-06-08 06:30:01,936 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:30:06,030 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:30:10,010 - pyscenedetect - INFO - Detecting scenes...
+
 91%|█████████ | 98/108 [22:01<02:04, 12.46s/it]2026-06-08 06:30:14,145 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:30:18,240 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:30:22,265 - pyscenedetect - INFO - Detecting scenes...
+
 92%|█████████▏| 99/108 [22:14<01:51, 12.39s/it]2026-06-08 06:30:26,392 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:30:30,436 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:30:34,555 - pyscenedetect - INFO - Detecting scenes...
+
 93%|█████████▎| 100/108 [22:26<01:38, 12.32s/it]2026-06-08 06:30:38,536 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:30:42,525 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:30:46,601 - pyscenedetect - INFO - Detecting scenes...
+
 94%|█████████▎| 101/108 [22:38<01:25, 12.26s/it]2026-06-08 06:30:50,653 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:30:54,718 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:30:58,766 - pyscenedetect - INFO - Detecting scenes...
+
 94%|█████████▍| 102/108 [22:50<01:13, 12.20s/it]2026-06-08 06:31:02,715 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:31:06,606 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:31:10,697 - pyscenedetect - INFO - Detecting scenes...
+
 95%|█████████▌| 103/108 [23:02<01:00, 12.12s/it]2026-06-08 06:31:14,656 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:31:18,647 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:31:22,716 - pyscenedetect - INFO - Detecting scenes...
+
 96%|█████████▋| 104/108 [23:14<00:48, 12.14s/it]2026-06-08 06:31:26,829 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:31:30,915 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:31:35,022 - pyscenedetect - INFO - Detecting scenes...
+
 97%|█████████▋| 105/108 [23:26<00:36, 12.18s/it]2026-06-08 06:31:39,090 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:31:43,083 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:31:47,124 - pyscenedetect - INFO - Detecting scenes...
+
 98%|█████████▊| 106/108 [23:39<00:24, 12.16s/it]2026-06-08 06:31:51,219 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:31:55,222 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:31:59,309 - pyscenedetect - INFO - Detecting scenes...
+
 99%|█████████▉| 107/108 [23:51<00:12, 12.16s/it]2026-06-08 06:32:03,370 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:32:07,319 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 06:32:11,318 - pyscenedetect - INFO - Detecting scenes...
+
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Camera_Motion/results_2026-06-08-06:08:05_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Complex_Landscape.log

@@ -0,0 +1,24 @@
+[2026-06-08 06:37:04 UTC] COMMAND /root/ckpts/micromamba/envs/vbench2/bin/python -u evaluate.py --videos_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Complex_Landscape --dimension Complex_Landscape --output_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Complex_Landscape --full_json_dir /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json --mode vbench_standard
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Complex_Landscape', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Complex_Landscape', dimension=['Complex_Landscape'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+File exists: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+File exists: /root/ckpts/vbench2/Qwen/Qwen2.5-7B-Instruct
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Complex_Landscape/results_2026-06-08-06:37:08_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/LLaVA_NeXT/llava/model/llava_arch.py:215: SyntaxWarning: "is not" with a literal. Did you mean "!="?
+  if slower_img_feat is not 0:
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Complex_Landscape/results_2026-06-08-06:37:08_full_info.json
+Loaded LLaVA model: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+You are using a model of type llava to instantiate a model of type llava_qwen. This is not supported for all configurations of models and can yield errors.
+Sliding Window Attention is enabled but not implemented for `sdpa`; unexpected results may be encountered.
+Loading vision tower: google/siglip-so400m-patch14-384
+2026-06-08 06:37:14,718 - accelerate.utils.modeling - INFO - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk).
+
+Model Class: LlavaQwenForCausalLM
+2026-06-08 06:37:20,156 - accelerate.utils.modeling - INFO - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk).
+
+
  0%|          | 0/30 [00:00<?, ?it/s]The attention mask is not set and cannot be inferred from input because pad token is same as eos token. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
+
  3%|▎         | 1/30 [00:53<26:05, 53.97s/it]
  7%|▋         | 2/30 [01:48<25:26, 54.51s/it]
 10%|█         | 3/30 [02:38<23:33, 52.33s/it]
 13%|█▎        | 4/30 [03:13<19:44, 45.56s/it]
 17%|█▋        | 5/30 [04:05<19:54, 47.79s/it]
 20%|██        | 6/30 [05:04<20:40, 51.70s/it]
 23%|██▎       | 7/30 [05:52<19:18, 50.37s/it]
 27%|██▋       | 8/30 [06:34<17:31, 47.78s/it]
 30%|███       | 9/30 [07:31<17:44, 50.68s/it]
 33%|███▎      | 10/30 [08:34<18:08, 54.40s/it]
 37%|█��█▋      | 11/30 [09:38<18:09, 57.36s/it]
 40%|████      | 12/30 [10:32<16:55, 56.44s/it]
 43%|████▎     | 13/30 [11:52<17:59, 63.47s/it]
 47%|████▋     | 14/30 [12:44<15:58, 59.88s/it]
 50%|█████     | 15/30 [13:25<13:35, 54.37s/it]
 53%|█████▎    | 16/30 [14:23<12:57, 55.53s/it]
 57%|█████▋    | 17/30 [15:15<11:48, 54.48s/it]
 60%|██████    | 18/30 [16:05<10:36, 53.05s/it]
 63%|██████▎   | 19/30 [16:55<09:32, 52.00s/it]
 67%|██████▋   | 20/30 [17:59<09:17, 55.70s/it]
 70%|███████   | 21/30 [18:50<08:09, 54.38s/it]
 73%|███████▎  | 22/30 [19:47<07:20, 55.04s/it]
 77%|███████▋  | 23/30 [20:48<06:38, 56.94s/it]
 80%|████████  | 24/30 [21:40<05:33, 55.51s/it]
 83%|████████▎ | 25/30 [22:29<04:26, 53.27s/it]
 87%|████████▋ | 26/30 [23:23<03:34, 53.74s/it]
 90%|█████████ | 27/30 [24:09<02:34, 51.45s/it]
 93%|█████████▎| 28/30 [24:53<01:38, 49.08s/it]
 97%|█████████▋| 29/30 [25:40<00:48, 48.32s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Complex_Landscape/results_2026-06-08-06:37:08_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Complex_Plot.log

@@ -0,0 +1,24 @@
+[2026-06-08 06:08:02 UTC] COMMAND /root/ckpts/micromamba/envs/vbench2/bin/python -u evaluate.py --videos_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Complex_Plot --dimension Complex_Plot --output_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Complex_Plot --full_json_dir /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json --mode vbench_standard
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Complex_Plot', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Complex_Plot', dimension=['Complex_Plot'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+File exists: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+File exists: /root/ckpts/vbench2/Qwen/Qwen2.5-7B-Instruct
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Complex_Plot/results_2026-06-08-06:08:05_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/LLaVA_NeXT/llava/model/llava_arch.py:215: SyntaxWarning: "is not" with a literal. Did you mean "!="?
+  if slower_img_feat is not 0:
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Complex_Plot/results_2026-06-08-06:08:05_full_info.json
+Loaded LLaVA model: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+You are using a model of type llava to instantiate a model of type llava_qwen. This is not supported for all configurations of models and can yield errors.
+Sliding Window Attention is enabled but not implemented for `sdpa`; unexpected results may be encountered.
+Loading vision tower: google/siglip-so400m-patch14-384
+2026-06-08 06:08:16,942 - accelerate.utils.modeling - INFO - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk).
+
+Model Class: LlavaQwenForCausalLM
+2026-06-08 06:08:23,048 - accelerate.utils.modeling - INFO - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk).
+
+
  0%|          | 0/60 [00:00<?, ?it/s]The attention mask is not set and cannot be inferred from input because pad token is same as eos token. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
+
  2%|▏         | 1/60 [00:35<35:01, 35.63s/it]
  3%|▎         | 2/60 [01:10<33:52, 35.04s/it]
  5%|▌         | 3/60 [01:56<38:19, 40.34s/it]
  7%|▋         | 4/60 [02:42<39:43, 42.56s/it]
  8%|▊         | 5/60 [03:11<34:19, 37.44s/it]
 10%|█         | 6/60 [03:53<35:14, 39.15s/it]
 12%|█▏        | 7/60 [04:35<35:13, 39.89s/it]
 13%|█▎        | 8/60 [05:06<32:09, 37.11s/it]
 15%|█▌        | 9/60 [05:41<31:05, 36.58s/it]
 17%|█▋        | 10/60 [06:25<32:27, 38.94s/it]
 18%|█▊        | 11/60 [07:07<32:20, 39.61s/it]
 20%|██        | 12/60 [07:50<32:36, 40.77s/it]
 22%|██▏       | 13/60 [08:26<30:48, 39.34s/it]
 23%|██▎       | 14/60 [08:59<28:39, 37.38s/it]
 25%|██▌       | 15/60 [09:43<29:34, 39.43s/it]
 27%|██▋       | 16/60 [10:20<28:18, 38.59s/it]
 28%|██▊       | 17/60 [10:52<26:20, 36.76s/it]
 30%|███       | 18/60 [11:30<25:51, 36.93s/it]
 32%|███▏      | 19/60 [11:55<22:53, 33.49s/it]
 33%|███▎      | 20/60 [12:24<21:31, 32.29s/it]
 35%|███▌      | 21/60 [12:59<21:23, 32.90s/it]
 37%|███▋      | 22/60 [13:27<19:59, 31.58s/it]
 38%|███▊      | 23/60 [14:02<19:59, 32.41s/it]
 40%|████      | 24/60 [14:34<19:25, 32.36s/it]
 42%|████▏     | 25/60 [15:15<20:23, 34.95s/it]
 43%|████▎     | 26/60 [15:48<19:25, 34.28s/it]
 45%|████▌     | 27/60 [16:28<19:53, 36.18s/it]
 47%|████▋     | 28/60 [17:13<20:36, 38.63s/it]
 48%|████▊     | 29/60 [17:44<18:54, 36.59s/it]
 50%|█████     | 30/60 [18:20<18:04, 36.15s/it]
 52%|█████▏    | 31/60 [19:03<18:27, 38.21s/it]
 53%|█████▎    | 32/60 [19:40<17:43, 37.98s/it]
 55%|█████▌    | 33/60 [20:14<16:36, 36.89s/it]
 57%|█████▋    | 34/60 [20:49<15:42, 36.25s/it]
 58%|█████▊    | 35/60 [21:24<14:54, 35.79s/it]
 60%|██████    | 36/60 [21:56<13:49, 34.58s/it]
 62%|██████▏   | 37/60 [22:32<13:27, 35.13s/it]
 63%|██████▎   | 38/60 [23:18<14:02, 38.32s/it]
 65%|██████▌   | 39/60 [23:54<13:13, 37.80s/it]
 67%|██████▋   | 40/60 [24:38<13:11, 39.57s/it]
 68%|██████▊   | 41/60 [25:14<12:12, 38.54s/it]
 70%|███████   | 42/60 [25:48<11:05, 37.00s/it]
 72%|███████▏  | 43/60 [26:22<10:14, 36.17s/it]
 73%|███████▎  | 44/60 [27:06<10:19, 38.73s/it]
 75%|███████▌  | 45/60 [27:50<10:01, 40.09s/it]
 77%|███████▋  | 46/60 [28:33<09:35, 41.13s/it]
 78%|███████▊  | 47/60 [29:15<08:56, 41.27s/it]
 80%|████████  | 48/60 [30:04<08:43, 43.59s/it]
 82%|████████▏ | 49/60 [30:53<08:16, 45.17s/it]
 83%|████████▎ | 50/60 [31:23<06:48, 40.83s/it]
 85%|████████▌ | 51/60 [31:57<05:47, 38.63s/it]
 87%|████████▋ | 52/60 [32:24<04:41, 35.17s/it]
 88%|████████▊ | 53/60 [33:00<04:07, 35.34s/it]
 90%|█████████ | 54/60 [33:33<03:28, 34.70s/it]
 92%|█████████▏| 55/60 [34:05<02:49, 33.88s/it]
 93%|█████████▎| 56/60 [34:39<02:16, 34.04s/it]
 95%|█████████▌| 57/60 [35:12<01:41, 33.75s/it]
 97%|█████████▋| 58/60 [35:48<01:08, 34.39s/it]
 98%|█████████▊| 59/60 [36:27<00:35, 35.79s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Complex_Plot/results_2026-06-08-06:08:05_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Composition.log

@@ -0,0 +1,21 @@
+[2026-06-08 06:08:02 UTC] COMMAND /root/ckpts/micromamba/envs/vbench2/bin/python -u evaluate.py --videos_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Composition --dimension Composition --output_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Composition --full_json_dir /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json --mode vbench_standard
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Composition', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Composition', dimension=['Composition'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+File exists: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Composition/results_2026-06-08-06:08:05_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/LLaVA_NeXT/llava/model/llava_arch.py:215: SyntaxWarning: "is not" with a literal. Did you mean "!="?
+  if slower_img_feat is not 0:
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Composition/results_2026-06-08-06:08:05_full_info.json
+Loaded LLaVA model: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+You are using a model of type llava to instantiate a model of type llava_qwen. This is not supported for all configurations of models and can yield errors.
+Sliding Window Attention is enabled but not implemented for `sdpa`; unexpected results may be encountered.
+Loading vision tower: google/siglip-so400m-patch14-384
+2026-06-08 06:08:17,208 - accelerate.utils.modeling - INFO - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk).
+
+Model Class: LlavaQwenForCausalLM
+
  0%|          | 0/53 [00:00<?, ?it/s]The attention mask is not set and cannot be inferred from input because pad token is same as eos token. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
+
  2%|▏         | 1/53 [00:22<19:48, 22.87s/it]
  4%|▍         | 2/53 [00:40<17:00, 20.00s/it]
  6%|▌         | 3/53 [01:03<17:31, 21.02s/it]
  8%|▊         | 4/53 [01:25<17:42, 21.69s/it]
  9%|▉         | 5/53 [01:43<16:18, 20.38s/it]
 11%|█▏        | 6/53 [02:01<15:17, 19.53s/it]
 13%|█▎        | 7/53 [02:11<12:25, 16.21s/it]
 15%|█▌        | 8/53 [02:29<12:37, 16.84s/it]
 17%|█▋        | 9/53 [02:51<13:34, 18.50s/it]
 19%|█▉        | 10/53 [03:13<14:07, 19.70s/it]
 21%|██        | 11/53 [03:36<14:19, 20.46s/it]
 23%|██▎       | 12/53 [03:58<14:20, 21.00s/it]
 25%|██▍       | 13/53 [04:21<14:21, 21.54s/it]
 26%|██▋       | 14/53 [04:35<12:34, 19.35s/it]
 28%|██▊       | 15/53 [04:53<12:05, 19.09s/it]
 30%|███       | 16/53 [05:03<10:05, 16.37s/it]
 32%|███▏      | 17/53 [05:31<11:48, 19.69s/it]
 34%|███▍      | 18/53 [05:41<09:51, 16.89s/it]
 36%|███▌      | 19/53 [06:13<12:04, 21.30s/it]
 38%|███▊      | 20/53 [06:31<11:15, 20.47s/it]
 40%|███▉      | 21/53 [06:42<09:17, 17.42s/it]
 42%|████▏     | 22/53 [07:25<12:59, 25.14s/it]
 43%|████▎     | 23/53 [07:46<11:55, 23.85s/it]
 45%|████▌     | 24/53 [08:17<12:39, 26.20s/it]
 47%|████▋     | 25/53 [08:27<09:56, 21.31s/it]
 49%|████▉     | 26/53 [08:37<08:03, 17.90s/it]
 51%|█████     | 27/53 [08:47<06:41, 15.46s/it]
 53%|█████▎    | 28/53 [08:57<05:45, 13.82s/it]
 55%|█████▍    | 29/53 [09:20<06:39, 16.64s/it]
 57%|█████▋    | 30/53 [09:30<05:36, 14.62s/it]
 58%|█████▊    | 31/53 [09:40<04:49, 13.14s/it]
 60%|██████    | 32/53 [09:56<04:55, 14.06s/it]
 62%|██████▏   | 33/53 [10:06<04:14, 12.75s/it]
 64%|██████▍   | 34/53 [10:15<03:43, 11.76s/it]
 66%|██████▌   | 35/53 [10:24<03:17, 10.98s/it]
 68%|██████▊   | 36/53 [10:33<02:57, 10.43s/it]
 70%|██████▉   | 37/53 [10:43<02:41, 10.07s/it]
 72%|███████▏  | 38/53 [10:52<02:26,  9.79s/it]
 74%|███████▎  | 39/53 [11:01<02:14,  9.61s/it]
 75%|███████▌  | 40/53 [11:17<02:29, 11.48s/it]
 77%|███████▋  | 41/53 [11:26<02:09, 10.82s/it]
 79%|███████▉  | 42/53 [11:35<01:53, 10.31s/it]
 81%|████████  | 43/53 [11:44<01:39,  9.97s/it]
 83%|████████▎ | 44/53 [12:26<02:56, 19.62s/it]
 85%|████████▍ | 45/53 [12:55<02:58, 22.37s/it]
 87%|████████▋ | 46/53 [13:24<02:50, 24.33s/it]
 89%|████████▊ | 47/53 [13:53<02:34, 25.69s/it]
 91%|█████████ | 48/53 [14:22<02:13, 26.65s/it]
 92%|█████████▏| 49/53 [14:51<01:49, 27.41s/it]
 94%|█████████▍| 50/53 [15:21<01:24, 28.06s/it]
 96%|█████████▌| 51/53 [15:50<00:57, 28.51s/it]
 98%|█████████▊| 52/53 [16:27<00:30, 30.86s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Composition/results_2026-06-08-06:08:05_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Diversity.log

@@ -0,0 +1,14 @@
+[2026-06-08 06:08:02 UTC] COMMAND /root/ckpts/micromamba/envs/vbench2/bin/python -u evaluate.py --videos_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Diversity --dimension Diversity --output_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Diversity --full_json_dir /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json --mode vbench_standard
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Diversity', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Diversity', dimension=['Diversity'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Diversity/results_2026-06-08-06:08:05_full_info.json
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Diversity/results_2026-06-08-06:08:05_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
+  warnings.warn(
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG19_Weights.IMAGENET1K_V1`. You can also use `weights=VGG19_Weights.DEFAULT` to get the most up-to-date weights.
+  warnings.warn(msg)
+Downloading: "https://download.pytorch.org/models/vgg19-dcbb9e9d.pth" to /root/ckpts/torch/hub/checkpoints/vgg19-dcbb9e9d.pth
+
  0%|          | 0.00/548M [00:00<?, ?B/s]
  0%|          | 2.38M/548M [00:00<00:23, 24.6MB/s]
  2%|▏         | 11.6M/548M [00:00<00:08, 66.7MB/s]
  5%|▌         | 27.5M/548M [00:00<00:04, 112MB/s] 
 10%|▉         | 53.0M/548M [00:00<00:03, 173MB/s]
 15%|█▍        | 79.9M/548M [00:00<00:02, 212MB/s]
 20%|█▉        | 107M/548M [00:00<00:01, 237MB/s] 
 25%|██▍       | 134M/548M [00:00<00:01, 253MB/s]
 30%|██▉       | 162M/548M [00:00<00:01, 264MB/s]
 34%|███▍      | 189M/548M [00:00<00:01, 270MB/s]
 39%|███▉      | 216M/548M [00:01<00:01, 275MB/s]
 44%|████▍     | 244M/548M [00:01<00:01, 279MB/s]
 50%|████▉     | 272M/548M [00:01<00:01, 282MB/s]
 54%|█████▍    | 299M/548M [00:01<00:00, 282MB/s]
 59%|█████▉    | 326M/548M [00:01<00:00, 284MB/s]
 65%|██████▍   | 354M/548M [00:01<00:00, 285MB/s]
 69%|██████▉   | 381M/548M [00:01<00:00, 285MB/s]
 74%|███████▍  | 408M/548M [00:01<00:00, 272MB/s]
 79%|███████▉  | 434M/548M [00:01<00:00, 262MB/s]
 84%|████████▍ | 462M/548M [00:01<00:00, 269MB/s]
 89%|████████▉ | 489M/548M [00:02<00:00, 275MB/s]
 94%|█████████▍| 517M/548M [00:02<00:00, 278MB/s]
 99%|█████████▉| 543M/548M [00:02<00:00, 278MB/s]
+
  0%|          | 0/10 [00:00<?, ?it/s]
 10%|█         | 1/10 [10:13<1:32:01, 613.55s/it]
 20%|██        | 2/10 [20:51<1:23:44, 628.04s/it]
 30%|███       | 3/10 [31:23<1:13:28, 629.80s/it]
 40%|████      | 4/10 [42:08<1:03:34, 635.76s/it]
 50%|█████     | 5/10 [52:53<53:15, 639.12s/it]  
 60%|██████    | 6/10 [1:03:33<42:36, 639.25s/it]
 70%|███████   | 7/10 [1:14:03<31:48, 636.27s/it]
 80%|████████  | 8/10 [1:24:29<21:05, 632.99s/it]
 90%|█████████ | 9/10 [1:35:19<10:38, 638.31s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Diversity/results_2026-06-08-06:08:05_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Dynamic_Attribute.log

@@ -0,0 +1,21 @@
+[2026-06-08 06:17:03 UTC] COMMAND /root/ckpts/micromamba/envs/vbench2/bin/python -u evaluate.py --videos_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Dynamic_Attribute --dimension Dynamic_Attribute --output_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Dynamic_Attribute --full_json_dir /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json --mode vbench_standard
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Dynamic_Attribute', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Dynamic_Attribute', dimension=['Dynamic_Attribute'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+File exists: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Dynamic_Attribute/results_2026-06-08-06:17:06_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/LLaVA_NeXT/llava/model/llava_arch.py:215: SyntaxWarning: "is not" with a literal. Did you mean "!="?
+  if slower_img_feat is not 0:
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Dynamic_Attribute/results_2026-06-08-06:17:06_full_info.json
+Loaded LLaVA model: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+You are using a model of type llava to instantiate a model of type llava_qwen. This is not supported for all configurations of models and can yield errors.
+Sliding Window Attention is enabled but not implemented for `sdpa`; unexpected results may be encountered.
+Loading vision tower: google/siglip-so400m-patch14-384
+2026-06-08 06:17:13,618 - accelerate.utils.modeling - INFO - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk).
+
+Model Class: LlavaQwenForCausalLM
+
  0%|          | 0/91 [00:00<?, ?it/s]The attention mask is not set and cannot be inferred from input because pad token is same as eos token. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
+
  1%|          | 1/91 [00:23<35:00, 23.34s/it]
  2%|▏         | 2/91 [00:46<34:38, 23.35s/it]
  3%|▎         | 3/91 [01:09<34:06, 23.26s/it]
  4%|▍         | 4/91 [01:32<33:10, 22.88s/it]
  5%|▌         | 5/91 [01:54<32:25, 22.62s/it]
  7%|▋         | 6/91 [02:16<31:53, 22.51s/it]
  8%|▊         | 7/91 [02:38<31:24, 22.43s/it]
  9%|▉         | 8/91 [03:00<30:51, 22.31s/it]
 10%|▉         | 9/91 [03:22<30:22, 22.23s/it]
 11%|█         | 10/91 [03:45<29:57, 22.19s/it]
 12%|█▏        | 11/91 [04:07<29:37, 22.22s/it]
 13%|█▎        | 12/91 [04:29<29:14, 22.21s/it]
 14%|█▍        | 13/91 [04:51<28:57, 22.27s/it]
 15%|█▌        | 14/91 [05:14<28:33, 22.26s/it]
 16%|█▋        | 15/91 [05:36<28:07, 22.21s/it]
 18%|█▊        | 16/91 [05:58<27:53, 22.32s/it]
 19%|█▊        | 17/91 [06:21<27:40, 22.44s/it]
 20%|█▉        | 18/91 [06:44<27:17, 22.44s/it]
 21%|██        | 19/91 [07:06<27:04, 22.56s/it]
 22%|██▏       | 20/91 [07:29<26:52, 22.72s/it]
 23%|██▎       | 21/91 [07:53<26:44, 22.93s/it]
 24%|██▍       | 22/91 [08:16<26:17, 22.86s/it]
 25%|██▌       | 23/91 [08:38<25:49, 22.79s/it]
 26%|██▋       | 24/91 [09:01<25:33, 22.89s/it]
 27%|██▋       | 25/91 [09:24<25:08, 22.85s/it]
 29%|██▊       | 26/91 [09:47<24:42, 22.81s/it]
 30%|██▉       | 27/91 [10:09<24:14, 22.73s/it]
 31%|███       | 28/91 [10:33<24:02, 22.89s/it]
 32%|███▏      | 29/91 [10:56<23:40, 22.91s/it]
 33%|███▎      | 30/91 [11:19<23:18, 22.92s/it]
 34%|███▍      | 31/91 [11:41<22:47, 22.79s/it]
 35%|███▌      | 32/91 [11:50<18:19, 18.64s/it]
 36%|███▋      | 33/91 [11:59<15:14, 15.76s/it]
 37%|███▋      | 34/91 [12:08<12:58, 13.66s/it]
 38%|███▊      | 35/91 [12:17<11:26, 12.25s/it]
 40%|███▉      | 36/91 [12:26<10:17, 11.24s/it]
 41%|████      | 37/91 [12:34<09:28, 10.53s/it]
 42%|████▏     | 38/91 [12:43<08:52, 10.05s/it]
 43%|████▎     | 39/91 [12:52<08:25,  9.72s/it]
 44%|████▍     | 40/91 [13:01<08:02,  9.47s/it]
 45%|████▌     | 41/91 [13:10<07:45,  9.30s/it]
 46%|████▌     | 42/91 [13:19<07:29,  9.18s/it]
 47%|████▋     | 43/91 [13:28<07:14,  9.06s/it]
 48%|████▊     | 44/91 [13:37<07:05,  9.05s/it]
 49%|████▉     | 45/91 [13:46<06:56,  9.05s/it]
 51%|█████     | 46/91 [13:55<06:48,  9.07s/it]
 52%|█████▏    | 47/91 [14:04<06:39,  9.07s/it]
 53%|█████▎    | 48/91 [14:13<06:29,  9.05s/it]
 54%|█████▍    | 49/91 [14:22<06:19,  9.04s/it]
 55%|█████▍    | 50/91 [14:31<06:10,  9.05s/it]
 56%|█████▌    | 51/91 [14:40<06:01,  9.04s/it]
 57%|█████▋    | 52/91 [14:49<05:52,  9.03s/it]
 58%|█████▊    | 53/91 [14:58<05:43,  9.03s/it]
 59%|█████▉    | 54/91 [15:07<05:35,  9.08s/it]
 60%|██████    | 55/91 [15:16<05:26,  9.08s/it]
 62%|██████▏   | 56/91 [15:25<05:15,  9.02s/it]
 63%|██████▎   | 57/91 [15:34<05:07,  9.05s/it]
 64%|██████▎   | 58/91 [15:44<04:59,  9.08s/it]
 65%|██████▍   | 59/91 [15:53<04:49,  9.05s/it]
 66%|██████▌   | 60/91 [16:02<04:43,  9.14s/it]
 67%|██████▋   | 61/91 [16:11<04:36,  9.21s/it]
 68%|██████▊   | 62/91 [16:21<04:29,  9.29s/it]
 69%|██████▉   | 63/91 [16:31<04:26,  9.53s/it]
 70%|███████   | 64/91 [16:40<04:16,  9.49s/it]
 71%|███████▏  | 65/91 [16:50<04:08,  9.56s/it]
 73%|███████▎  | 66/91 [17:00<04:00,  9.60s/it]
 74%|███████▎  | 67/91 [17:10<03:52,  9.69s/it]
 75%|███████▍  | 68/91 [17:33<05:17, 13.80s/it]
 76%|███████▌  | 69/91 [17:56<06:02, 16.47s/it]
 77%|███████▋  | 70/91 [18:18<06:25, 18.35s/it]
 78%|███████▊  | 71/91 [18:41<06:33, 19.66s/it]
 79%|███████▉  | 72/91 [19:04<06:30, 20.57s/it]
 80%|████████  | 73/91 [19:27<06:22, 21.22s/it]
 81%|████████▏ | 74/91 [19:50<06:10, 21.80s/it]
 82%|████████▏ | 75/91 [20:12<05:51, 21.97s/it]
 84%|████████▎ | 76/91 [20:35<05:32, 22.15s/it]
 85%|████████▍ | 77/91 [20:58<05:13, 22.41s/it]
 86%|████████▌ | 78/91 [21:20<04:52, 22.52s/it]
 87%|████████▋ | 79/91 [21:43<04:30, 22.51s/it]
 88%|████████▊ | 80/91 [22:05<04:07, 22.47s/it]
 89%|████████▉ | 81/91 [22:28<03:44, 22.44s/it]
 90%|█████████ | 82/91 [22:50<03:20, 22.30s/it]
 91%|█████████ | 83/91 [23:12<02:57, 22.21s/it]
 92%|█████████▏| 84/91 [23:34<02:35, 22.15s/it]
 93%|█████████▎| 85/91 [23:56<02:12, 22.06s/it]
 95%|█████████▍| 86/91 [24:18<01:50, 22.03s/it]
 96%|█████████▌| 87/91 [24:40<01:28, 22.04s/it]
 97%|█████████▋| 88/91 [25:02<01:06, 22.03s/it]
 98%|█████████▊| 89/91 [25:24<00:44, 22.03s/it]
 99%|█████████▉| 90/91 [25:46<00:21, 21.99s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Dynamic_Attribute/results_2026-06-08-06:17:06_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Dynamic_Spatial_Relationship.log

@@ -0,0 +1,21 @@
+[2026-06-08 06:08:02 UTC] COMMAND /root/ckpts/micromamba/envs/vbench2/bin/python -u evaluate.py --videos_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Dynamic_Spatial_Relationship --dimension Dynamic_Spatial_Relationship --output_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Dynamic_Spatial_Relationship --full_json_dir /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json --mode vbench_standard
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Dynamic_Spatial_Relationship', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Dynamic_Spatial_Relationship', dimension=['Dynamic_Spatial_Relationship'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+File exists: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Dynamic_Spatial_Relationship/results_2026-06-08-06:08:05_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/LLaVA_NeXT/llava/model/llava_arch.py:215: SyntaxWarning: "is not" with a literal. Did you mean "!="?
+  if slower_img_feat is not 0:
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Dynamic_Spatial_Relationship/results_2026-06-08-06:08:05_full_info.json
+Loaded LLaVA model: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+You are using a model of type llava to instantiate a model of type llava_qwen. This is not supported for all configurations of models and can yield errors.
+Sliding Window Attention is enabled but not implemented for `sdpa`; unexpected results may be encountered.
+Loading vision tower: google/siglip-so400m-patch14-384
+2026-06-08 06:08:17,167 - accelerate.utils.modeling - INFO - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk).
+
+Model Class: LlavaQwenForCausalLM
+
  0%|          | 0/69 [00:00<?, ?it/s]The attention mask is not set and cannot be inferred from input because pad token is same as eos token. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
+
  1%|▏         | 1/69 [00:07<08:04,  7.13s/it]
  3%|▎         | 2/69 [00:13<07:30,  6.72s/it]
  4%|▍         | 3/69 [00:20<07:16,  6.62s/it]
  6%|▌         | 4/69 [00:26<07:05,  6.54s/it]
  7%|▋         | 5/69 [00:33<06:59,  6.55s/it]
  9%|▊         | 6/69 [00:39<06:47,  6.47s/it]
 10%|█         | 7/69 [00:45<06:41,  6.47s/it]
 12%|█▏        | 8/69 [00:52<06:32,  6.44s/it]
 13%|█▎        | 9/69 [00:58<06:23,  6.40s/it]
 14%|█▍        | 10/69 [01:04<06:18,  6.41s/it]
 16%|█▌        | 11/69 [01:11<06:13,  6.44s/it]
 17%|█▋        | 12/69 [01:17<06:05,  6.42s/it]
 19%|█▉        | 13/69 [01:24<06:01,  6.46s/it]
 20%|██        | 14/69 [01:30<05:52,  6.41s/it]
 22%|██▏       | 15/69 [01:37<05:46,  6.42s/it]
 23%|██▎       | 16/69 [01:43<05:38,  6.38s/it]
 25%|██▍       | 17/69 [01:49<05:31,  6.37s/it]
 26%|██▌       | 18/69 [01:56<05:24,  6.35s/it]
 28%|██▊       | 19/69 [02:02<05:17,  6.34s/it]
 29%|██▉       | 20/69 [02:08<05:10,  6.33s/it]
 30%|███       | 21/69 [02:15<05:06,  6.38s/it]
 32%|███▏      | 22/69 [02:21<05:01,  6.41s/it]
 33%|███▎      | 23/69 [02:28<04:56,  6.44s/it]
 35%|███▍      | 24/69 [02:34<04:48,  6.41s/it]
 36%|███▌      | 25/69 [02:40<04:41,  6.40s/it]
 38%|███▊      | 26/69 [02:47<04:34,  6.39s/it]
 39%|███▉      | 27/69 [02:53<04:28,  6.40s/it]
 41%|████      | 28/69 [03:00<04:23,  6.42s/it]
 42%|████▏     | 29/69 [03:06<04:16,  6.41s/it]
 43%|████▎     | 30/69 [03:13<04:11,  6.45s/it]
 45%|████▍     | 31/69 [03:19<04:04,  6.43s/it]
 46%|████▋     | 32/69 [03:25<03:58,  6.43s/it]
 48%|████▊     | 33/69 [03:32<03:51,  6.44s/it]
 49%|████▉     | 34/69 [03:38<03:44,  6.41s/it]
 51%|█████     | 35/69 [03:45<03:40,  6.47s/it]
 52%|█████▏    | 36/69 [03:51<03:33,  6.46s/it]
 54%|█████▎    | 37/69 [03:58<03:25,  6.41s/it]
 55%|█████▌    | 38/69 [04:04<03:19,  6.43s/it]
 57%|█████▋    | 39/69 [04:11<03:20,  6.67s/it]
 58%|█████▊    | 40/69 [04:19<03:19,  6.89s/it]
 59%|█████▉    | 41/69 [04:26<03:16,  7.03s/it]
 61%|██████    | 42/69 [04:33<03:08,  6.98s/it]
 62%|██████▏   | 43/69 [04:40<02:58,  6.88s/it]
 64%|██████▍   | 44/69 [04:47<02:55,  7.02s/it]
 65%|██████▌   | 45/69 [04:55<02:53,  7.21s/it]
 67%|██████▋   | 46/69 [05:01<02:43,  7.12s/it]
 68%|██████▊   | 47/69 [05:09<02:37,  7.17s/it]
 70%|██████▉   | 48/69 [05:16<02:34,  7.34s/it]
 71%|███████   | 49/69 [05:23<02:24,  7.20s/it]
 72%|███████▏  | 50/69 [05:30<02:15,  7.11s/it]
 74%|███████▍  | 51/69 [05:37<02:05,  6.98s/it]
 75%|███████▌  | 52/69 [05:44<01:57,  6.90s/it]
 77%|███████▋  | 53/69 [05:51<01:51,  6.98s/it]
 78%|███████▊  | 54/69 [05:59<01:51,  7.44s/it]
 80%|███████▉  | 55/69 [06:07<01:44,  7.48s/it]
 81%|████████  | 56/69 [06:15<01:38,  7.54s/it]
 83%|████████▎ | 57/69 [06:22<01:29,  7.50s/it]
 84%|████████▍ | 58/69 [06:29<01:21,  7.45s/it]
 86%|████████▌ | 59/69 [06:36<01:13,  7.36s/it]
 87%|████████▋ | 60/69 [06:43<01:05,  7.26s/it]
 88%|████████▊ | 61/69 [06:51<00:58,  7.25s/it]
 90%|████████▉ | 62/69 [06:58<00:50,  7.19s/it]
 91%|█████████▏| 63/69 [07:05<00:43,  7.27s/it]
 93%|█████████▎| 64/69 [07:13<00:36,  7.34s/it]
 94%|█████████▍| 65/69 [07:21<00:30,  7.61s/it]
 96%|█████████▌| 66/69 [07:28<00:22,  7.50s/it]
 97%|█████████▋| 67/69 [07:36<00:15,  7.54s/it]
 99%|█████████▊| 68/69 [07:44<00:07,  7.66s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Dynamic_Spatial_Relationship/results_2026-06-08-06:08:05_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Human_Anatomy.log

@@ -0,0 +1,23 @@
+[2026-06-08 06:08:02 UTC] COMMAND /root/ckpts/micromamba/envs/vbench2/bin/python -u evaluate.py --videos_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Human_Anatomy --dimension Human_Anatomy --output_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Anatomy --full_json_dir /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json --mode vbench_standard
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Anatomy', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Human_Anatomy', dimension=['Human_Anatomy'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Anatomy/results_2026-06-08-06:08:05_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Anatomy/results_2026-06-08-06:08:05_full_info.json
+Loads checkpoint by local backend from path: /root/ckpts/vbench2/YOLO-World/yolo_world_v2_xl_obj365v1_goldg_cc3mlite_pretrain-5daf1395.pth
+2026-06-08 06:08:15,379 - lvis.lvis - INFO - Loading annotations.
+2026-06-08 06:08:16,825 - lvis.lvis - INFO - Creating index.
+2026-06-08 06:08:16,874 - lvis.lvis - INFO - Index created.
+Loads checkpoint by local backend from path: /root/ckpts/vbench2/YOLO-World/yolo_world_v2_xl_obj365v1_goldg_cc3mlite_pretrain-5daf1395.pth
+2026-06-08 06:08:21,169 - lvis.lvis - INFO - Loading annotations.
+2026-06-08 06:08:22,077 - lvis.lvis - INFO - Creating index.
+2026-06-08 06:08:22,130 - lvis.lvis - INFO - Index created.
+=> merge config from vbench2/third_party/ViTDetector/simmim_finetune__vit_base__img224__800ep.yaml
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/torch/functional.py:505: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at /pytorch/aten/src/ATen/native/TensorShape.cpp:4317.)
+  return _VF.meshgrid(tensors, **kwargs)  # type: ignore[attr-defined]
+=> merge config from vbench2/third_party/ViTDetector/simmim_finetune__vit_base__img224__800ep.yaml
+=> merge config from vbench2/third_party/ViTDetector/simmim_finetune__vit_base__img224__800ep.yaml
+
  0%|          | 0/360 [00:00<?, ?it/s]
  0%|          | 1/360 [00:11<1:11:43, 11.99s/it]
  1%|          | 2/360 [00:24<1:12:59, 12.23s/it]
  1%|          | 3/360 [00:37<1:14:51, 12.58s/it]
  1%|          | 4/360 [00:50<1:16:26, 12.88s/it]
  1%|▏         | 5/360 [01:03<1:15:28, 12.76s/it]
  2%|▏         | 6/360 [01:16<1:15:42, 12.83s/it]
  2%|▏         | 7/360 [01:29<1:16:43, 13.04s/it]
  2%|▏         | 8/360 [01:47<1:25:37, 14.60s/it]
  2%|▎         | 9/360 [02:06<1:33:23, 15.96s/it]
  3%|▎         | 10/360 [02:18<1:25:39, 14.69s/it]
  3%|▎         | 11/360 [02:31<1:21:59, 14.10s/it]
  3%|▎         | 12/360 [02:43<1:18:58, 13.62s/it]
  4%|▎         | 13/360 [02:56<1:17:28, 13.40s/it]
  4%|▍         | 14/360 [03:09<1:15:33, 13.10s/it]
  4%|▍         | 15/360 [03:21<1:14:05, 12.88s/it]
  4%|▍         | 16/360 [03:39<1:22:25, 14.38s/it]
  5%|▍         | 17/360 [04:08<1:48:03, 18.90s/it]
  5%|▌         | 18/360 [04:48<2:23:10, 25.12s/it]
  5%|▌         | 19/360 [05:05<2:09:38, 22.81s/it]
  6%|▌         | 20/360 [05:22<1:58:48, 20.97s/it]
  6%|▌         | 21/360 [05:38<1:49:51, 19.44s/it]
  6%|▌         | 22/360 [05:54<1:44:00, 18.46s/it]
  6%|▋         | 23/360 [06:11<1:40:42, 17.93s/it]
  7%|▋         | 24/360 [06:27<1:37:10, 17.35s/it]
  7%|▋         | 25/360 [06:59<2:02:08, 21.88s/it]
  7%|▋         | 26/360 [07:31<2:18:38, 24.91s/it]
  8%|▊         | 27/360 [07:54<2:15:07, 24.35s/it]
  8%|▊         | 28/360 [08:11<2:02:43, 22.18s/it]
  8%|▊         | 29/360 [08:28<1:54:06, 20.68s/it]
  8%|▊         | 30/360 [08:47<1:49:45, 19.96s/it]
  9%|▊         | 31/360 [09:05<1:46:43, 19.46s/it]
  9%|▉         | 32/360 [09:22<1:42:53, 18.82s/it]
  9%|▉         | 33/360 [09:40<1:40:24, 18.42s/it]
  9%|▉         | 34/360 [10:11<2:00:30, 22.18s/it]
 10%|▉         | 35/360 [10:36<2:05:50, 23.23s/it]
 10%|█         | 36/360 [11:01<2:07:59, 23.70s/it]
 10%|█         | 37/360 [11:14<1:49:38, 20.37s/it]
 11%|█         | 38/360 [11:27<1:37:39, 18.20s/it]
 11%|█         | 39/360 [11:39<1:27:33, 16.37s/it]
 11%|█         | 40/360 [11:52<1:22:16, 15.43s/it]
 11%|█▏        | 41/360 [12:05<1:17:37, 14.60s/it]
 12%|█▏        | 42/360 [12:18<1:15:00, 14.15s/it]
 12%|█▏        | 43/360 [12:32<1:15:12, 14.23s/it]
 12%|█▏        | 44/360 [12:52<1:22:46, 15.72s/it]
 12%|█▎        | 45/360 [12:59<1:09:48, 13.30s/it]
 13%|█▎        | 46/360 [13:12<1:08:24, 13.07s/it]
 13%|█▎        | 47/360 [13:24<1:07:18, 12.90s/it]
 13%|█▎        | 48/360 [13:40<1:10:43, 13.60s/it]
 14%|█▎        | 49/360 [13:52<1:08:47, 13.27s/it]
 14%|█▍        | 50/360 [14:08<1:12:14, 13.98s/it]
 14%|█▍        | 51/360 [14:22<1:11:49, 13.95s/it]
 14%|█▍        | 52/360 [14:55<1:41:16, 19.73s/it]
 15%|█▍        | 53/360 [15:14<1:40:45, 19.69s/it]
 15%|█▌        | 54/360 [15:46<1:59:22, 23.41s/it]
 15%|█▌        | 55/360 [16:04<1:50:38, 21.76s/it]
 16%|█▌        | 56/360 [16:24<1:47:41, 21.26s/it]
 16%|█▌        | 57/360 [16:43<1:43:02, 20.40s/it]
 16%|█▌        | 58/360 [17:00<1:37:50, 19.44s/it]
 16%|█▋        | 59/360 [17:17<1:33:09, 18.57s/it]
 17%|█▋        | 60/360 [17:32<1:28:15, 17.65s/it]
 17%|█▋        | 61/360 [17:53<1:32:21, 18.53s/it]
 17%|█▋        | 62/360 [18:11<1:32:18, 18.59s/it]
 18%|█▊        | 63/360 [18:29<1:31:08, 18.41s/it]
 18%|█▊        | 64/360 [18:47<1:29:38, 18.17s/it]
 18%|█▊        | 65/360 [19:05<1:28:35, 18.02s/it]
 18%|█▊        | 66/360 [19:24<1:29:36, 18.29s/it]
 19%|█▊        | 67/360 [19:50<1:40:30, 20.58s/it]
 19%|█▉        | 68/360 [20:17<1:49:45, 22.55s/it]
 19%|█▉        | 69/360 [20:37<1:45:56, 21.84s/it]
 19%|█▉        | 70/360 [20:53<1:36:55, 20.05s/it]
 20%|█▉        | 71/360 [21:07<1:28:44, 18.42s/it]
 20%|██        | 72/360 [21:19<1:18:53, 16.44s/it]
 20%|██        | 73/360 [21:32<1:12:52, 15.24s/it]
 21%|██        | 74/360 [21:49<1:16:10, 15.98s/it]
 21%|██        | 75/360 [22:07<1:18:42, 16.57s/it]
 21%|██        | 76/360 [22:27<1:22:27, 17.42s/it]
 21%|██▏       | 77/360 [22:44<1:21:20, 17.25s/it]
 22%|██▏       | 78/360 [23:01<1:21:45, 17.40s/it]
 22%|██▏       | 79/360 [23:12<1:12:19, 15.44s/it]
 22%|██▏       | 80/360 [23:22<1:04:02, 13.72s/it]
 22%|██▎       | 81/360 [23:34<1:01:52, 13.31s/it]
 23%|██▎       | 82/360 [23:46<58:53, 12.71s/it]  
 23%|██▎       | 83/360 [23:58<58:36, 12.70s/it]
 23%|██▎       | 84/360 [24:08<54:04, 11.75s/it]
 24%|██▎       | 85/360 [24:22<56:47, 12.39s/it]
 24%|██▍       | 86/360 [24:40<1:05:06, 14.26s/it]
 24%|██▍       | 87/360 [24:55<1:05:42, 14.44s/it]
 24%|██▍       | 88/360 [25:15<1:12:16, 15.94s/it]
 25%|██▍       | 89/360 [25:36<1:19:27, 17.59s/it]
 25%|██▌       | 90/360 [26:02<1:30:36, 20.13s/it]
 25%|██▌       | 91/360 [26:20<1:26:51, 19.37s/it]
 26%|██▌       | 92/360 [26:44<1:33:44, 20.99s/it]
 26%|██▌       | 93/360 [27:08<1:37:23, 21.89s/it]
 26%|██▌       | 94/360 [27:47<1:59:06, 26.87s/it]
 26%|██▋       | 95/360 [28:17<2:02:40, 27.78s/it]
 27%|██▋       | 96/360 [28:41<1:57:47, 26.77s/it]
 27%|██▋       | 97/360 [28:56<1:41:36, 23.18s/it]
 27%|██▋       | 98/360 [29:11<1:30:27, 20.72s/it]
 28%|██▊       | 99/360 [29:28<1:25:24, 19.64s/it]
 28%|██▊       | 100/360 [29:43<1:18:27, 18.10s/it]
 28%|██▊       | 101/360 [30:00<1:16:51, 17.80s/it]
 28%|██▊       | 102/360 [30:17<1:15:38, 17.59s/it]
 29%|██▊       | 103/360 [30:34<1:14:52, 17.48s/it]
 29%|██▉       | 104/360 [30:47<1:09:21, 16.25s/it]
 29%|██▉       | 105/360 [31:05<1:11:04, 16.72s/it]
 29%|██▉       | 106/360 [31:16<1:03:49, 15.08s/it]
 30%|██▉       | 107/360 [31:28<58:46, 13.94s/it]  
 30%|███       | 108/360 [31:38<54:23, 12.95s/it]
 30%|███       | 109/360 [31:50<52:26, 12.54s/it]
 31%|███       | 110/360 [32:00<49:07, 11.79s/it]
 31%|███       | 111/360 [32:11<48:08, 11.60s/it]
 31%|███       | 112/360 [32:20<43:52, 10.61s/it]
 31%|███▏      | 113/360 [32:28<41:35, 10.10s/it]
 32%|███▏      | 114/360 [32:40<43:18, 10.56s/it]
 32%|███▏      | 115/360 [32:52<44:29, 10.90s/it]
 32%|███▏      | 116/360 [33:00<41:42, 10.25s/it]
 32%|███▎      | 117/360 [33:13<43:49, 10.82s/it]
 33%|███▎      | 118/360 [33:25<45:00, 11.16s/it]
 33%|███▎      | 119/360 [33:39<48:41, 12.12s/it]
 33%|███▎      | 120/360 [33:51<47:59, 12.00s/it]
 34%|███▎      | 121/360 [34:02<47:28, 11.92s/it]
 34%|███▍      | 122/360 [34:15<47:51, 12.07s/it]
 34%|███▍      | 123/360 [34:30<50:56, 12.90s/it]
 34%|███▍      | 124/360 [34:44<52:03, 13.23s/it]
 35%|███▍      | 125/360 [34:55<50:06, 12.79s/it]
 35%|███▌      | 126/360 [35:07<48:40, 12.48s/it]
 35%|███▌      | 127/360 [35:18<46:19, 11.93s/it]
 36%|███▌      | 128/360 [35:29<45:22, 11.73s/it]
 36%|███▌      | 129/360 [35:46<50:43, 13.18s/it]
 36%|███▌      | 130/360 [36:04<56:47, 14.81s/it]
 36%|███▋      | 131/360 [36:29<1:07:26, 17.67s/it]
 37%|███▋      | 132/360 [36:48<1:08:49, 18.11s/it]
 37%|███▋      | 133/360 [37:03<1:05:32, 17.33s/it]
 37%|███▋      | 134/360 [37:20<1:04:28, 17.12s/it]
 38%|███▊      | 135/360 [37:34<1:00:31, 16.14s/it]
 38%|███▊      | 136/360 [37:52<1:02:12, 16.66s/it]
 38%|███▊      | 137/360 [38:08<1:01:05, 16.44s/it]
 38%|███▊      | 138/360 [38:26<1:02:53, 17.00s/it]
 39%|███▊      | 139/360 [38:44<1:03:23, 17.21s/it]
 39%|███▉      | 140/360 [39:00<1:02:37, 17.08s/it]
 39%|███▉      | 141/360 [39:17<1:01:46, 16.93s/it]
 39%|███▉      | 142/360 [39:32<59:28, 16.37s/it]  
 40%|███▉      | 143/360 [39:46<57:02, 15.77s/it]
 40%|████      | 144/360 [40:01<55:39, 15.46s/it]
 40%|████      | 145/360 [40:22<1:01:40, 17.21s/it]
 41%|████      | 146/360 [40:45<1:07:26, 18.91s/it]
 41%|████      | 147/360 [41:05<1:07:51, 19.11s/it]
 41%|████      | 148/360 [41:24<1:07:26, 19.09s/it]
 41%|████▏     | 149/360 [41:38<1:02:22, 17.74s/it]
 42%|████▏     | 150/360 [41:53<59:11, 16.91s/it]  
 42%|████▏     | 151/360 [42:07<55:47, 16.02s/it]
 42%|████▏     | 152/360 [42:20<52:09, 15.04s/it]
 42%|████▎     | 153/360 [42:32<48:57, 14.19s/it]
 43%|████▎     | 154/360 [42:43<44:55, 13.08s/it]
 43%|████▎     | 155/360 [42:53<42:11, 12.35s/it]
 43%|████▎     | 156/360 [43:04<39:58, 11.76s/it]
 44%|████▎     | 157/360 [43:15<38:41, 11.43s/it]
 44%|████▍     | 158/360 [43:26<38:12, 11.35s/it]
 44%|████▍     | 159/360 [43:36<36:33, 10.92s/it]
 44%|████▍     | 160/360 [43:48<38:10, 11.45s/it]
 45%|████▍     | 161/360 [44:01<39:28, 11.90s/it]
 45%|████▌     | 162/360 [44:13<38:53, 11.78s/it]
 45%|████▌     | 163/360 [44:25<38:57, 11.87s/it]
 46%|████▌     | 164/360 [44:37<39:19, 12.04s/it]
 46%|████▌     | 165/360 [44:49<38:44, 11.92s/it]
 46%|████▌     | 166/360 [45:00<37:46, 11.68s/it]
 46%|████▋     | 167/360 [45:11<36:30, 11.35s/it]
 47%|████▋     | 168/360 [45:21<35:32, 11.10s/it]
 47%|████▋     | 169/360 [45:32<35:28, 11.14s/it]
 47%|████▋     | 170/360 [45:38<30:31,  9.64s/it]
 48%|████▊     | 171/360 [45:52<34:07, 10.83s/it]
 48%|████▊     | 172/360 [46:04<35:07, 11.21s/it]
 48%|████▊     | 173/360 [46:18<36:59, 11.87s/it]
 48%|████▊     | 174/360 [46:34<41:25, 13.36s/it]
 49%|████▊     | 175/360 [46:51<44:32, 14.44s/it]
 49%|████▉     | 176/360 [47:09<47:34, 15.52s/it]
 49%|████▉     | 177/360 [47:27<49:07, 16.10s/it]
 49%|████▉     | 178/360 [47:45<50:38, 16.69s/it]
 50%|████▉     | 179/360 [48:04<52:11, 17.30s/it]
 50%|█████     | 180/360 [48:24<54:16, 18.09s/it]
 50%|█████     | 181/360 [48:45<56:31, 18.95s/it]
 51%|█████     | 182/360 [49:05<57:10, 19.27s/it]
 51%|█████     | 183/360 [49:24<56:34, 19.18s/it]
 51%|█████     | 184/360 [49:39<52:41, 17.96s/it]
 51%|█████▏    | 185/360 [49:54<49:40, 17.03s/it]
 52%|█████▏    | 186/360 [50:08<47:32, 16.39s/it]
 52%|█████▏    | 187/360 [50:26<47:59, 16.64s/it]
 52%|█████▏    | 188/360 [50:42<47:15, 16.48s/it]
 52%|█████▎    | 189/360 [51:00<48:26, 17.00s/it]
 53%|█████▎    | 190/360 [51:16<47:42, 16.84s/it]
 53%|█████▎    | 191/360 [51:33<47:15, 16.78s/it]
 53%|█████▎    | 192/360 [51:48<45:30, 16.25s/it]
 54%|█████▎    | 193/360 [52:09<49:17, 17.71s/it]
 54%|█████▍    | 194/360 [52:24<46:35, 16.84s/it]
 54%|█████▍    | 195/360 [52:39<44:52, 16.32s/it]
 54%|█████▍    | 196/360 [52:50<39:58, 14.62s/it]
 55%|█████▍    | 197/360 [53:00<36:26, 13.41s/it]
 55%|█████▌    | 198/360 [53:12<34:39, 12.83s/it]
 55%|█████▌    | 199/360 [53:24<34:03, 12.69s/it]
 56%|█████▌    | 200/360 [53:37<34:10, 12.82s/it]
 56%|█████▌    | 201/360 [53:52<35:38, 13.45s/it]
 56%|█████▌    | 202/360 [54:04<34:23, 13.06s/it]
 56%|█████▋    | 203/360 [54:17<34:00, 12.99s/it]
 57%|█████▋    | 204/360 [54:28<32:18, 12.43s/it]
 57%|█████▋    | 205/360 [54:40<31:25, 12.16s/it]
 57%|█████▋    | 206/360 [54:51<30:28, 11.88s/it]
 57%|█████▊    | 207/360 [55:04<30:49, 12.09s/it]
 58%|█████▊    | 208/360 [55:16<30:28, 12.03s/it]
 58%|█████▊    | 209/360 [55:27<30:07, 11.97s/it]
 58%|█████▊    | 210/360 [55:36<27:03, 10.82s/it]
 59%|█████▊    | 211/360 [55:48<28:13, 11.37s/it]
 59%|█████▉    | 212/360 [56:02<29:32, 11.97s/it]
 59%|█████▉    | 213/360 [56:14<29:44, 12.14s/it]
 59%|█████▉    | 214/360 [56:25<28:51, 11.86s/it]
 60%|█████▉    | 215/360 [56:37<28:44, 11.89s/it]
 60%|██████    | 216/360 [56:50<29:00, 12.09s/it]
 60%|██████    | 217/360 [57:03<29:51, 12.53s/it]
 61%|██████    | 218/360 [57:19<31:51, 13.46s/it]
 61%|██████    | 219/360 [57:33<32:20, 13.76s/it]
 61%|██████    | 220/360 [57:51<35:05, 15.04s/it]
 61%|██████▏   | 221/360 [58:09<36:49, 15.90s/it]
 62%|██████▏   | 222/360 [58:24<35:30, 15.44s/it]
 62%|██████▏   | 223/360 [58:36<33:00, 14.46s/it]
 62%|██████▏   | 224/360 [58:48<31:07, 13.73s/it]
 62%|██████▎   | 225/360 [59:01<30:44, 13.66s/it]
 63%|██████▎   | 226/360 [59:21<34:23, 15.40s/it]
 63%|██████▎   | 227/360 [59:39<36:13, 16.35s/it]
 63%|██████▎   | 228/360 [59:56<35:47, 16.27s/it]
 64%|██████▎   | 229/360 [1:00:10<34:35, 15.84s/it]
 64%|██████▍   | 230/360 [1:00:30<36:34, 16.88s/it]
 64%|██████▍   | 231/360 [1:00:45<35:26, 16.49s/it]
 64%|██████▍   | 232/360 [1:01:00<34:15, 16.06s/it]
 65%|██████▍   | 233/360 [1:01:15<33:13, 15.70s/it]
 65%|██████▌   | 234/360 [1:01:31<32:58, 15.70s/it]
 65%|██████▌   | 235/360 [1:01:47<32:46, 15.73s/it]
 66%|██████▌   | 236/360 [1:02:06<34:55, 16.90s/it]
 66%|██████▌   | 237/360 [1:02:25<35:27, 17.30s/it]
 66%|██████▌   | 238/360 [1:02:40<33:55, 16.69s/it]
 66%|██████▋   | 239/360 [1:02:57<33:41, 16.70s/it]
 67%|██████▋   | 240/360 [1:03:12<32:39, 16.33s/it]
 67%|██████▋   | 241/360 [1:03:23<29:01, 14.64s/it]
 67%|██████▋   | 242/360 [1:03:34<26:41, 13.57s/it]
 68%|██████▊   | 243/360 [1:03:43<24:00, 12.31s/it]
 68%|██████▊   | 244/360 [1:03:54<22:48, 11.80s/it]
 68%|██████▊   | 245/360 [1:04:05<22:23, 11.68s/it]
 68%|██████▊   | 246/360 [1:04:16<21:56, 11.55s/it]
 69%|██████▊   | 247/360 [1:04:29<22:06, 11.74s/it]
 69%|██████▉   | 248/360 [1:04:40<21:45, 11.66s/it]
 69%|██████▉   | 249/360 [1:04:54<22:45, 12.30s/it]
 69%|██████▉   | 250/360 [1:05:04<21:33, 11.76s/it]
 70%|██████▉   | 251/360 [1:05:16<21:15, 11.71s/it]
 70%|███████   | 252/360 [1:05:26<20:09, 11.20s/it]
 70%|███████   | 253/360 [1:05:37<20:00, 11.22s/it]
 71%|███████   | 254/360 [1:05:49<19:52, 11.25s/it]
 71%|███████   | 255/360 [1:06:00<19:53, 11.37s/it]
 71%|███████   | 256/360 [1:06:13<20:30, 11.83s/it]
 71%|███████▏  | 257/360 [1:06:24<19:47, 11.53s/it]
 72%|███████▏  | 258/360 [1:06:40<22:06, 13.00s/it]
 72%|███████▏  | 259/360 [1:06:56<23:27, 13.93s/it]
 72%|███████▏  | 260/360 [1:07:09<22:34, 13.55s/it]
 72%|███████▎  | 261/360 [1:07:26<24:03, 14.58s/it]
 73%|███████▎  | 262/360 [1:07:36<21:41, 13.28s/it]
 73%|███████▎  | 263/360 [1:08:02<27:21, 16.93s/it]
 73%|███████▎  | 264/360 [1:08:22<28:38, 17.90s/it]
 74%|███████▎  | 265/360 [1:08:50<32:59, 20.84s/it]
 74%|███████▍  | 266/360 [1:09:09<32:05, 20.48s/it]
 74%|███████▍  | 267/360 [1:09:29<31:11, 20.12s/it]
 74%|███████▍  | 268/360 [1:09:52<32:20, 21.09s/it]
 75%|███████▍  | 269/360 [1:10:11<31:08, 20.53s/it]
 75%|███████▌  | 270/360 [1:10:32<30:52, 20.58s/it]
 75%|███████▌  | 271/360 [1:10:52<30:07, 20.31s/it]
 76%|███████▌  | 272/360 [1:11:14<30:32, 20.82s/it]
 76%|███████▌  | 273/360 [1:11:37<31:26, 21.68s/it]
 76%|███████▌  | 274/360 [1:11:55<29:26, 20.54s/it]
 76%|███████▋  | 275/360 [1:12:15<28:38, 20.22s/it]
 77%|███████▋  | 276/360 [1:12:31<26:38, 19.03s/it]
 77%|███████▋  | 277/360 [1:12:41<22:33, 16.31s/it]
 77%|███████▋  | 278/360 [1:12:57<22:18, 16.33s/it]
 78%|███████▊  | 279/360 [1:13:12<21:18, 15.78s/it]
 78%|███████▊  | 280/360 [1:13:30<21:55, 16.44s/it]
 78%|███████▊  | 281/360 [1:13:52<23:53, 18.14s/it]
 78%|███████▊  | 282/360 [1:14:05<21:45, 16.73s/it]
 79%|███████▊  | 283/360 [1:14:20<20:52, 16.26s/it]
 79%|███████▉  | 284/360 [1:14:34<19:37, 15.49s/it]
 79%|███████▉  | 285/360 [1:14:50<19:21, 15.49s/it]
 79%|███████▉  | 286/360 [1:15:09<20:42, 16.79s/it]
 80%|███████▉  | 287/360 [1:15:27<20:39, 16.97s/it]
 80%|████████  | 288/360 [1:15:45<20:38, 17.20s/it]
 80%|████████  | 289/360 [1:15:59<19:22, 16.37s/it]
 81%|████████  | 290/360 [1:16:14<18:45, 16.08s/it]
 81%|████████  | 291/360 [1:16:26<17:06, 14.87s/it]
 81%|████████  | 292/360 [1:16:42<16:56, 14.95s/it]
 81%|████████▏ | 293/360 [1:16:57<16:50, 15.09s/it]
 82%|████████▏ | 294/360 [1:17:14<17:17, 15.72s/it]
 82%|████████▏ | 295/360 [1:17:34<18:16, 16.87s/it]
 82%|████████▏ | 296/360 [1:17:51<18:09, 17.02s/it]
 82%|████████▎ | 297/360 [1:18:08<17:51, 17.00s/it]
 83%|████████▎ | 298/360 [1:18:28<18:21, 17.76s/it]
 83%|████████▎ | 299/360 [1:18:45<17:50, 17.55s/it]
 83%|████████▎ | 300/360 [1:19:05<18:24, 18.41s/it]
 84%|████████▎ | 301/360 [1:19:31<20:27, 20.80s/it]
 84%|████████▍ | 302/360 [1:19:57<21:31, 22.26s/it]
 84%|████████▍ | 303/360 [1:20:21<21:45, 22.91s/it]
 84%|████████▍ | 304/360 [1:20:41<20:29, 21.95s/it]
 85%|████████▍ | 305/360 [1:20:57<18:28, 20.16s/it]
 85%|████████▌ | 306/360 [1:21:14<17:12, 19.12s/it]
 85%|████████▌ | 307/360 [1:21:33<16:46, 18.99s/it]
 86%|████████▌ | 308/360 [1:21:52<16:36, 19.17s/it]
 86%|████████▌ | 309/360 [1:22:10<15:59, 18.82s/it]
 86%|████████▌ | 310/360 [1:22:35<17:17, 20.75s/it]
 86%|████████▋ | 311/360 [1:22:57<17:12, 21.06s/it]
 87%|████████▋ | 312/360 [1:23:20<17:14, 21.55s/it]
 87%|████████▋ | 313/360 [1:23:37<15:52, 20.27s/it]
 87%|████████▋ | 314/360 [1:23:56<15:15, 19.91s/it]
 88%|████████▊ | 315/360 [1:24:18<15:20, 20.46s/it]
 88%|████████▊ | 316/360 [1:24:33<13:54, 18.96s/it]
 88%|████████▊ | 317/360 [1:24:50<13:06, 18.29s/it]
 88%|████████▊ | 318/360 [1:25:08<12:39, 18.09s/it]
 89%|████████▊ | 319/360 [1:25:22<11:36, 16.99s/it]
 89%|████████▉ | 320/360 [1:25:37<10:47, 16.18s/it]
 89%|████████▉ | 321/360 [1:25:51<10:06, 15.56s/it]
 89%|████████▉ | 322/360 [1:26:09<10:27, 16.52s/it]
 90%|████████▉ | 323/360 [1:26:28<10:34, 17.16s/it]
 90%|█████████ | 324/360 [1:26:47<10:39, 17.75s/it]
 90%|█████████ | 325/360 [1:27:00<09:30, 16.29s/it]
 91%|█████████ | 326/360 [1:27:14<08:46, 15.48s/it]
 91%|█████████ | 327/360 [1:27:27<08:11, 14.90s/it]
 91%|█████████ | 328/360 [1:27:40<07:40, 14.40s/it]
 91%|█████████▏| 329/360 [1:27:58<07:59, 15.48s/it]
 92%|█████████▏| 330/360 [1:28:21<08:44, 17.47s/it]
 92%|█████████▏| 331/360 [1:28:33<07:44, 16.01s/it]
 92%|█████████▏| 332/360 [1:28:58<08:43, 18.71s/it]
 92%|█████████▎| 333/360 [1:29:20<08:47, 19.53s/it]
 93%|█████████▎| 334/360 [1:29:45<09:14, 21.34s/it]
 93%|█████████▎| 335/360 [1:30:21<10:40, 25.63s/it]
 93%|█████████▎| 336/360 [1:30:51<10:47, 26.97s/it]
 94%|█████████▎| 337/360 [1:31:12<09:42, 25.35s/it]
 94%|█████████▍| 338/360 [1:31:35<08:58, 24.49s/it]
 94%|█████████▍| 339/360 [1:31:59<08:28, 24.24s/it]
 94%|█████████▍| 340/360 [1:32:27<08:27, 25.38s/it]
 95%|█████████▍| 341/360 [1:32:54<08:10, 25.83s/it]
 95%|█████████▌| 342/360 [1:33:22<07:59, 26.64s/it]
 95%|█████████▌| 343/360 [1:33:41<06:55, 24.47s/it]
 96%|█████████▌| 344/360 [1:34:01<06:10, 23.13s/it]
 96%|█████████▌| 345/360 [1:34:24<05:44, 22.95s/it]
 96%|█████████▌| 346/360 [1:34:45<05:14, 22.49s/it]
 96%|█████████▋| 347/360 [1:35:09<04:55, 22.75s/it]
 97%|█████████▋| 348/360 [1:35:27<04:14, 21.25s/it]
 97%|█████████▋| 349/360 [1:35:36<03:14, 17.71s/it]
 97%|█████████▋| 350/360 [1:35:48<02:40, 16.01s/it]
 98%|█████████▊| 351/360 [1:36:01<02:15, 15.09s/it]
 98%|█████████▊| 352/360 [1:36:19<02:08, 16.08s/it]
 98%|█████████▊| 353/360 [1:36:41<02:04, 17.75s/it]
 98%|█████████▊| 354/360 [1:36:59<01:47, 17.91s/it]
 99%|█████████▊| 355/360 [1:37:20<01:33, 18.71s/it]
 99%|█████████▉| 356/360 [1:37:37<01:13, 18.27s/it]
 99%|█████████▉| 357/360 [1:37:54<00:53, 17.79s/it]
 99%|█████████▉| 358/360 [1:38:08<00:33, 16.72s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Anatomy/results_2026-06-08-06:08:05_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Human_Clothes.log

@@ -0,0 +1,23 @@
+[2026-06-08 06:08:02 UTC] COMMAND /root/ckpts/micromamba/envs/vbench2/bin/python -u evaluate.py --videos_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Human_Clothes --dimension Human_Clothes --output_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Clothes --full_json_dir /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json --mode vbench_standard
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Clothes', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Human_Clothes', dimension=['Human_Clothes'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+File exists: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Clothes/results_2026-06-08-06:08:05_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/LLaVA_NeXT/llava/model/llava_arch.py:215: SyntaxWarning: "is not" with a literal. Did you mean "!="?
+  if slower_img_feat is not 0:
+/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/LLaVA_NeXT/llava/model/llava_arch.py:215: SyntaxWarning: "is not" with a literal. Did you mean "!="?
+  if slower_img_feat is not 0:
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Clothes/results_2026-06-08-06:08:05_full_info.json
+Loaded LLaVA model: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+You are using a model of type llava to instantiate a model of type llava_qwen. This is not supported for all configurations of models and can yield errors.
+Sliding Window Attention is enabled but not implemented for `sdpa`; unexpected results may be encountered.
+Loading vision tower: google/siglip-so400m-patch14-384
+2026-06-08 06:08:16,058 - accelerate.utils.modeling - INFO - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk).
+
+Model Class: LlavaQwenForCausalLM
+
  0%|          | 0/75 [00:00<?, ?it/s]The attention mask is not set and cannot be inferred from input because pad token is same as eos token. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
+
  1%|▏         | 1/75 [00:24<30:17, 24.56s/it]
  3%|▎         | 2/75 [00:47<28:37, 23.53s/it]
  4%|▍         | 3/75 [01:10<27:51, 23.22s/it]
  5%|▌         | 4/75 [01:33<27:18, 23.08s/it]
  7%|▋         | 5/75 [01:55<26:50, 23.00s/it]
  8%|▊         | 6/75 [02:18<26:22, 22.93s/it]
  9%|▉         | 7/75 [02:41<25:55, 22.88s/it]
 11%|█         | 8/75 [03:04<25:34, 22.90s/it]
 12%|█▏        | 9/75 [03:27<25:10, 22.89s/it]
 13%|█▎        | 10/75 [03:50<24:48, 22.90s/it]
 15%|█▍        | 11/75 [04:04<21:29, 20.15s/it]
 16%|█▌        | 12/75 [04:23<20:53, 19.90s/it]
 17%|█▋        | 13/75 [04:46<21:39, 20.97s/it]
 19%|█▊        | 14/75 [05:10<22:09, 21.79s/it]
 20%|██        | 15/75 [05:34<22:21, 22.35s/it]
 21%|██▏       | 16/75 [05:57<22:18, 22.68s/it]
 23%|██▎       | 17/75 [06:12<19:37, 20.31s/it]
 24%|██▍       | 18/75 [06:27<17:40, 18.61s/it]
 25%|██▌       | 19/75 [06:46<17:26, 18.70s/it]
 27%|██▋       | 20/75 [07:00<15:57, 17.41s/it]
 28%|██▊       | 21/75 [07:24<17:20, 19.26s/it]
 29%|██▉       | 22/75 [07:47<18:12, 20.61s/it]
 31%|███       | 23/75 [08:11<18:39, 21.53s/it]
 32%|███▏      | 24/75 [08:35<18:53, 22.23s/it]
 33%|███▎      | 25/75 [08:59<18:53, 22.66s/it]
 35%|███▍      | 26/75 [09:22<18:38, 22.82s/it]
 36%|███▌      | 27/75 [09:45<18:26, 23.06s/it]
 37%|███▋      | 28/75 [10:05<17:16, 22.05s/it]
 39%|███▊      | 29/75 [10:28<17:04, 22.28s/it]
 40%|████      | 30/75 [10:51<16:49, 22.43s/it]
 41%|████▏     | 31/75 [11:13<16:32, 22.55s/it]
 43%|████▎     | 32/75 [11:36<16:12, 22.61s/it]
 44%|████▍     | 33/75 [11:59<15:53, 22.70s/it]
 45%|████▌     | 34/75 [12:17<14:36, 21.39s/it]
 47%|████▋     | 35/75 [12:40<14:30, 21.77s/it]
 48%|████▊     | 36/75 [13:03<14:21, 22.10s/it]
 49%|████▉     | 37/75 [13:26<14:07, 22.32s/it]
 51%|█████     | 38/75 [13:49<13:52, 22.50s/it]
 52%|█████▏    | 39/75 [14:11<13:32, 22.58s/it]
 53%|█████▎    | 40/75 [14:34<13:12, 22.66s/it]
 55%|█████▍    | 41/75 [14:58<13:01, 22.99s/it]
 56%|█████▌    | 42/75 [15:22<12:51, 23.37s/it]
 57%|█████▋    | 43/75 [15:46<12:28, 23.40s/it]
 59%|█████▊    | 44/75 [16:10<12:08, 23.51s/it]
 60%|██████    | 45/75 [16:33<11:48, 23.63s/it]
 61%|██████▏   | 46/75 [16:53<10:49, 22.39s/it]
 63%|██████▎   | 47/75 [17:17<10:41, 22.91s/it]
 64%|██████▍   | 48/75 [17:41<10:23, 23.10s/it]
 65%|██████▌   | 49/75 [18:04<10:03, 23.21s/it]
 67%|██████▋   | 50/75 [18:28<09:45, 23.42s/it]
 68%|██████▊   | 51/75 [18:52<09:23, 23.48s/it]
 69%|██████▉   | 52/75 [19:15<08:57, 23.38s/it]
 71%|███████   | 53/75 [19:34<08:07, 22.17s/it]
 72%|███████▏  | 54/75 [19:53<07:25, 21.23s/it]
 73%|███████▎  | 55/75 [20:17<07:18, 21.91s/it]
 75%|███████▍  | 56/75 [20:40<07:05, 22.42s/it]
 76%|███████▌  | 57/75 [21:03<06:46, 22.57s/it]
 77%|███████▋  | 58/75 [21:22<06:02, 21.32s/it]
 79%|███████▊  | 59/75 [21:44<05:47, 21.73s/it]
 80%|████████  | 60/75 [22:07<05:30, 22.04s/it]
 81%|████████▏ | 61/75 [22:30<05:11, 22.23s/it]
 83%|████████▎ | 62/75 [22:52<04:51, 22.39s/it]
 84%|████████▍ | 63/75 [23:11<04:14, 21.21s/it]
 85%|████████▌ | 64/75 [23:34<03:58, 21.70s/it]
 87%|████████▋ | 65/75 [23:57<03:40, 22.03s/it]
 88%|████████▊ | 66/75 [24:15<03:08, 20.90s/it]
 89%|████████▉ | 67/75 [24:38<02:51, 21.49s/it]
 91%|█████████ | 68/75 [25:01<02:33, 21.88s/it]
 92%|█████████▏| 69/75 [25:24<02:14, 22.44s/it]
 93%|█████████▎| 70/75 [25:48<01:53, 22.75s/it]
 95%|█████████▍| 71/75 [26:12<01:32, 23.13s/it]
 96%|█████████▌| 72/75 [26:35<01:09, 23.29s/it]
 97%|█████████▋| 73/75 [26:59<00:46, 23.29s/it]
 99%|█████████▊| 74/75 [27:22<00:23, 23.38s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Clothes/results_2026-06-08-06:08:05_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Human_Identity.log

@@ -0,0 +1,16 @@
+[2026-06-08 06:08:02 UTC] COMMAND /root/ckpts/micromamba/envs/vbench2/bin/python -u evaluate.py --videos_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Human_Identity --dimension Human_Identity --output_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Identity --full_json_dir /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json --mode vbench_standard
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Identity', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Human_Identity', dimension=['Human_Identity'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Identity/results_2026-06-08-06:08:05_full_info.json
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Identity/results_2026-06-08-06:08:05_full_info.json
+Downloading: "https://github.com/ternaus/retinaface/releases/download/0.01/retinaface_resnet50_2020-07-20-f168fae3c.zip" to /root/ckpts/torch/hub/checkpoints/retinaface_resnet50_2020-07-20-f168fae3c.zip
+
  0%|          | 0.00/96.9M [00:00<?, ?B/s]
 10%|█         | 10.1M/96.9M [00:00<00:01, 82.9MB/s]
 19%|█▊        | 18.1M/96.9M [00:00<00:01, 66.7MB/s]
 25%|██▌       | 24.6M/96.9M [00:00<00:01, 57.8MB/s]
 33%|███▎      | 32.4M/96.9M [00:00<00:01, 65.2MB/s]
 44%|████▍     | 42.6M/96.9M [00:00<00:00, 78.2MB/s]
 59%|█████▉    | 57.2M/96.9M [00:00<00:00, 101MB/s] 
 70%|██████▉   | 67.4M/96.9M [00:00<00:00, 97.6MB/s]
 79%|███████▉  | 77.0M/96.9M [00:01<00:00, 76.4MB/s]
 88%|████████▊ | 85.1M/96.9M [00:01<00:00, 71.8MB/s]
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/torch/hub.py:879: FutureWarning: Falling back to the old format < 1.6. This support will be deprecated in favor of default zipfile format introduced in 1.6. Please redo torch.save() to save it in the new zipfile format.
+  return _legacy_zip_load(cached_file, model_dir, map_location, weights_only)
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
+  warnings.warn(
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=None`.
+  warnings.warn(msg)
+
  0%|          | 0/46 [00:00<?, ?it/s]
  2%|▏         | 1/46 [01:14<55:46, 74.37s/it]
  4%|▍         | 2/46 [02:27<53:51, 73.45s/it]
  7%|▋         | 3/46 [03:39<52:25, 73.14s/it]
  9%|▊         | 4/46 [04:57<52:23, 74.84s/it]
 11%|█         | 5/46 [05:53<46:28, 68.01s/it]
 13%|█▎        | 6/46 [07:15<48:32, 72.81s/it]
 15%|█▌        | 7/46 [07:46<38:24, 59.10s/it]
 17%|█▋        | 8/46 [09:10<42:25, 66.97s/it]
 20%|█▉        | 9/46 [09:39<33:59, 55.11s/it]
 22%|██▏       | 10/46 [10:33<32:52, 54.80s/it]
 24%|██▍       | 11/46 [11:48<35:33, 60.97s/it]
 26%|██▌       | 12/46 [13:02<36:52, 65.09s/it]
 28%|██▊       | 13/46 [14:17<37:25, 68.05s/it]
 30%|███       | 14/46 [14:45<29:45, 55.79s/it]
 33%|███▎      | 15/46 [14:48<20:40, 40.03s/it]
 35%|███▍      | 16/46 [14:51<14:29, 28.98s/it]
 37%|███▋      | 17/46 [15:23<14:20, 29.68s/it]
 39%|███▉      | 18/46 [16:20<17:47, 38.13s/it]
 41%|████▏     | 19/46 [16:51<16:06, 35.81s/it]
 43%|████▎     | 20/46 [16:55<11:21, 26.21s/it]
 46%|████▌     | 21/46 [17:25<11:24, 27.38s/it]
 48%|████▊     | 22/46 [17:56<11:27, 28.64s/it]
 50%|█████     | 23/46 [18:00<08:03, 21.03s/it]
 52%|█████▏    | 24/46 [18:58<11:48, 32.19s/it]
 54%|█████▍    | 25/46 [19:55<13:55, 39.79s/it]
 57%|█████▋    | 26/46 [19:59<09:38, 28.90s/it]
 59%|█████▊    | 27/46 [20:28<09:10, 28.98s/it]
 61%|██████    | 28/46 [20:32<06:26, 21.46s/it]
 63%|██████▎   | 29/46 [21:01<06:42, 23.70s/it]
 65%|██████▌   | 30/46 [21:28<06:35, 24.75s/it]
 67%|██████▋   | 31/46 [21:54<06:17, 25.16s/it]
 70%|██████▉   | 32/46 [22:45<07:40, 32.89s/it]
 72%|███████▏  | 33/46 [23:35<08:15, 38.09s/it]
 74%|███████▍  | 34/46 [23:38<05:30, 27.56s/it]
 76%|███████▌  | 35/46 [23:41<03:42, 20.19s/it]
 78%|███████▊  | 36/46 [24:32<04:53, 29.37s/it]
 80%|████████  | 37/46 [24:35<03:13, 21.49s/it]
 83%|████████▎ | 38/46 [24:38<02:07, 15.97s/it]
 85%|████████▍ | 39/46 [25:05<02:14, 19.26s/it]
 87%|████████▋ | 40/46 [25:08<01:26, 14.39s/it]
 89%|████████▉ | 41/46 [25:38<01:35, 19.03s/it]
 91%|█████████▏| 42/46 [26:09<01:29, 22.48s/it]
 93%|█████████▎| 43/46 [27:07<01:39, 33.27s/it]
 96%|█████████▌| 44/46 [28:07<01:22, 41.20s/it]
 98%|█████████▊| 45/46 [28:37<00:37, 37.86s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Identity/results_2026-06-08-06:08:05_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Human_Interaction.log

@@ -0,0 +1,23 @@
+[2026-06-08 06:33:04 UTC] COMMAND /root/ckpts/micromamba/envs/vbench2/bin/python -u evaluate.py --videos_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Human_Interaction --dimension Human_Interaction --output_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Interaction --full_json_dir /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json --mode vbench_standard
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Interaction', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Human_Interaction', dimension=['Human_Interaction'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+File exists: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+File exists: /root/ckpts/vbench2/Qwen/Qwen2.5-7B-Instruct
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Interaction/results_2026-06-08-06:33:06_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/LLaVA_NeXT/llava/model/llava_arch.py:215: SyntaxWarning: "is not" with a literal. Did you mean "!="?
+  if slower_img_feat is not 0:
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Interaction/results_2026-06-08-06:33:06_full_info.json
+Loaded LLaVA model: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+You are using a model of type llava to instantiate a model of type llava_qwen. This is not supported for all configurations of models and can yield errors.
+Sliding Window Attention is enabled but not implemented for `sdpa`; unexpected results may be encountered.
+Loading vision tower: google/siglip-so400m-patch14-384
+
+Model Class: LlavaQwenForCausalLM
+2026-06-08 06:33:17,343 - accelerate.utils.modeling - INFO - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk).
+
+
  0%|          | 0/100 [00:00<?, ?it/s]The attention mask is not set and cannot be inferred from input because pad token is same as eos token. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
+
  1%|          | 1/100 [00:51<1:25:36, 51.88s/it]
  2%|▏         | 2/100 [01:38<1:19:51, 48.89s/it]
  3%|▎         | 3/100 [02:16<1:10:42, 43.74s/it]
  4%|▍         | 4/100 [02:56<1:07:49, 42.39s/it]
  5%|▌         | 5/100 [03:45<1:11:00, 44.85s/it]
  6%|▌         | 6/100 [04:38<1:14:39, 47.65s/it]
  7%|▋         | 7/100 [05:23<1:12:29, 46.77s/it]
  8%|▊         | 8/100 [06:15<1:14:02, 48.29s/it]
  9%|▉         | 9/100 [06:52<1:08:03, 44.88s/it]
 10%|█         | 10/100 [07:33<1:05:11, 43.46s/it]
 11%|█         | 11/100 [08:12<1:02:50, 42.36s/it]
 12%|█▏        | 12/100 [08:56<1:02:32, 42.64s/it]
 13%|█▎        | 13/100 [09:32<59:07, 40.78s/it]  
 14%|█▍        | 14/100 [10:12<57:59, 40.45s/it]
 15%|█▌        | 15/100 [10:54<58:01, 40.96s/it]
 16%|█▌        | 16/100 [11:45<1:01:23, 43.85s/it]
 17%|█▋        | 17/100 [12:27<1:00:12, 43.52s/it]
 18%|█▊        | 18/100 [13:06<57:36, 42.16s/it]  
 19%|█▉        | 19/100 [13:57<1:00:10, 44.58s/it]
 20%|██        | 20/100 [14:44<1:00:25, 45.32s/it]
 21%|██        | 21/100 [15:31<1:00:33, 46.00s/it]
 22%|██▏       | 22/100 [16:15<59:00, 45.39s/it]  
 23%|██▎       | 23/100 [17:03<59:06, 46.06s/it]
 24%|██▍       | 24/100 [17:38<54:10, 42.78s/it]
 25%|██▌       | 25/100 [18:17<52:16, 41.82s/it]
 26%|██▌       | 26/100 [18:58<50:57, 41.32s/it]
 27%|██▋       | 27/100 [19:43<51:38, 42.45s/it]
 28%|██▊       | 28/100 [20:22<49:50, 41.54s/it]
 29%|██▉       | 29/100 [21:06<50:08, 42.38s/it]
 30%|███       | 30/100 [21:53<50:58, 43.69s/it]
 31%|███       | 31/100 [22:39<51:00, 44.36s/it]
 32%|███▏      | 32/100 [23:23<50:14, 44.33s/it]
 33%|███▎      | 33/100 [24:10<50:15, 45.00s/it]
 34%|███▍      | 34/100 [24:50<47:49, 43.48s/it]
 35%|███▌      | 35/100 [25:33<47:06, 43.48s/it]
 36%|███▌      | 36/100 [26:18<46:37, 43.71s/it]
 37%|███▋      | 37/100 [27:15<50:13, 47.84s/it]
 38%|███▊      | 38/100 [27:58<47:59, 46.45s/it]
 39%|███▉      | 39/100 [28:42<46:22, 45.62s/it]
 40%|████      | 40/100 [29:23<44:20, 44.35s/it]
 41%|████      | 41/100 [30:02<41:48, 42.51s/it]
 42%|████▏     | 42/100 [30:37<38:58, 40.32s/it]
 43%|████▎     | 43/100 [31:23<39:54, 42.01s/it]
 44%|████▍     | 44/100 [31:59<37:27, 40.14s/it]
 45%|████▌     | 45/100 [32:41<37:19, 40.73s/it]
 46%|████▌     | 46/100 [33:26<37:57, 42.17s/it]
 47%|████▋     | 47/100 [34:10<37:46, 42.76s/it]
 48%|████▊     | 48/100 [34:53<36:56, 42.62s/it]
 49%|████▉     | 49/100 [35:39<37:18, 43.89s/it]
 50%|█████     | 50/100 [36:20<35:37, 42.75s/it]
 51%|█████     | 51/100 [37:17<38:32, 47.20s/it]
 52%|█████▏    | 52/100 [37:57<36:01, 45.03s/it]
 53%|█████▎    | 53/100 [38:34<33:16, 42.47s/it]
 54%|█████▍    | 54/100 [39:14<31:58, 41.71s/it]
 55%|█████▌    | 55/100 [39:56<31:32, 42.06s/it]
 56%|█████▌    | 56/100 [40:39<30:55, 42.17s/it]
 57%|█████▋    | 57/100 [41:22<30:29, 42.56s/it]
 58%|█████▊    | 58/100 [42:04<29:33, 42.24s/it]
 59%|█████▉    | 59/100 [42:44<28:24, 41.57s/it]
 60%|██████    | 60/100 [43:30<28:44, 43.11s/it]
 61%|██████    | 61/100 [44:15<28:21, 43.62s/it]
 62%|██████▏   | 62/100 [44:56<27:09, 42.89s/it]
 63%|██████▎   | 63/100 [45:40<26:35, 43.12s/it]
 64%|██████▍   | 64/100 [46:23<25:52, 43.12s/it]
 65%|██████▌   | 65/100 [47:13<26:21, 45.17s/it]
 66%|██████▌   | 66/100 [47:54<24:50, 43.84s/it]
 67%|██████▋   | 67/100 [48:55<27:01, 49.14s/it]
 68%|██████▊   | 68/100 [49:40<25:26, 47.70s/it]
 69%|██████▉   | 69/100 [50:16<22:51, 44.24s/it]
 70%|███████   | 70/100 [50:53<21:05, 42.18s/it]
 71%|███████   | 71/100 [51:32<19:51, 41.08s/it]
 72%|███████▏  | 72/100 [52:12<19:03, 40.83s/it]
 73%|███████▎  | 73/100 [52:49<17:51, 39.70s/it]
 74%|███████▍  | 74/100 [53:39<18:32, 42.80s/it]
 75%|███████▌  | 75/100 [54:32<19:06, 45.87s/it]
 76%|███████▌  | 76/100 [55:14<17:52, 44.69s/it]
 77%|███████▋  | 77/100 [56:11<18:28, 48.20s/it]
 78%|███████▊  | 78/100 [56:53<16:59, 46.33s/it]
 79%|███████▉  | 79/100 [57:40<16:20, 46.70s/it]
 80%|████████  | 80/100 [58:21<14:57, 44.87s/it]
 81%|████████  | 81/100 [59:08<14:25, 45.53s/it]
 82%|████████▏ | 82/100 [59:42<12:36, 42.01s/it]
 83%|████████▎ | 83/100 [1:00:28<12:16, 43.34s/it]
 84%|████████▍ | 84/100 [1:01:12<11:38, 43.67s/it]
 85%|████████▌ | 85/100 [1:01:49<10:25, 41.67s/it]
 86%|████████▌ | 86/100 [1:02:27<09:26, 40.47s/it]
 87%|████████▋ | 87/100 [1:03:13<09:06, 42.04s/it]
 88%|████████▊ | 88/100 [1:03:52<08:14, 41.17s/it]
 89%|████████▉ | 89/100 [1:04:54<08:41, 47.38s/it]
 90%|█████████ | 90/100 [1:05:49<08:18, 49.80s/it]
 91%|█████████ | 91/100 [1:06:34<07:15, 48.37s/it]
 92%|█████████▏| 92/100 [1:07:19<06:19, 47.39s/it]
 93%|█████████▎| 93/100 [1:08:07<05:31, 47.39s/it]
 94%|█████████▍| 94/100 [1:08:53<04:42, 47.03s/it]
 95%|█████████▌| 95/100 [1:09:42<03:58, 47.60s/it]
 96%|█████████▌| 96/100 [1:10:26<03:06, 46.62s/it]
 97%|█████████▋| 97/100 [1:11:06<02:13, 44.52s/it]
 98%|█████████▊| 98/100 [1:11:48<01:27, 43.95s/it]
 99%|█████████▉| 99/100 [1:12:33<00:44, 44.25s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Interaction/results_2026-06-08-06:33:06_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Instance_Preservation_rerun.log

@@ -0,0 +1,59 @@
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Instance_Preservation', dimension=['Instance_Preservation'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation/results_2026-06-08-07:26:41_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+[INFO:swift] Successfully registered `/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/Instance_detector/swift/llm/dataset/data/dataset_info.json`.
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation/results_2026-06-08-07:26:41_full_info.json
+[INFO:swift] Successfully loaded /root/ckpts/vbench2/instance_anomaly_detector/model/args.json.
+[INFO:swift] Loading the model using model_dir: /root/ckpts/huggingface/Qwen/Qwen2.5-VL-3B-Instruct
+[INFO:swift] Setting torch_dtype: torch.bfloat16
+Using a slow image processor as `use_fast` is unset and a slow processor was saved with this model. `use_fast=True` will be the default behavior in v4.52, even if the model was saved with a slow processor. This will result in minor differences in outputs. You'll still be able to use a slow processor with `use_fast=False`.
+[INFO:swift] model_kwargs: {'device_map': 'cuda:0'}
+
+[INFO:swift] Setting image_factor: 28. You can adjust this hyperparameter through the environment variable: `IMAGE_FACTOR`.
+[INFO:swift] Setting min_pixels: 3136. You can adjust this hyperparameter through the environment variable: `MIN_PIXELS`.
+[INFO:swift] Using environment variable `MAX_PIXELS`, Setting max_pixels: 1003520.
+[INFO:swift] Setting max_ratio: 200. You can adjust this hyperparameter through the environment variable: `MAX_RATIO`.
+[INFO:swift] Using environment variable `VIDEO_MIN_PIXELS`, Setting video_min_pixels: 100352.
+[INFO:swift] Using environment variable `VIDEO_MAX_PIXELS`, Setting video_max_pixels: 602112.
+[INFO:swift] Using environment variable `VIDEO_TOTAL_PIXELS`, Setting video_total_pixels: 90316800.
+[INFO:swift] Setting frame_factor: 2. You can adjust this hyperparameter through the environment variable: `FRAME_FACTOR`.
+[INFO:swift] Setting fps: 2.0. You can adjust this hyperparameter through the environment variable: `FPS`.
+[INFO:swift] Using environment variable `FPS_MIN_FRAMES`, Setting fps_min_frames: 32.
+[INFO:swift] Using environment variable `FPS_MAX_FRAMES`, Setting fps_max_frames: 50.
+[INFO:swift] default_system: 'You are a helpful and harmless assistant.'
+[INFO:swift] max_length: 128000
+[INFO:swift] response_prefix: ''
+[INFO:swift] agent_template: hermes
+[INFO:swift] norm_bbox: none
+[INFO:swift] default_system: 'You are a helpful and harmless assistant.'
+[INFO:swift] max_length: 128000
+[INFO:swift] response_prefix: ''
+[INFO:swift] agent_template: hermes
+[INFO:swift] norm_bbox: none
+
  0%|          | 0/57 [00:00<?, ?it/s]
  0%|          | 0/57 [00:00<?, ?it/s]
+Traceback (most recent call last):
+  File "/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/evaluate.py", line 153, in <module>
+    main()
+  File "/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/evaluate.py", line 139, in main
+    my_VBench.evaluate(
+  File "/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/__init__.py", line 112, in evaluate
+    results = evaluate_func(cur_full_info_path, self.device, submodules_list, **kwargs)
+  File "/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/instance_preservation.py", line 19, in compute_instance_preservation
+    all_results, video_results = compute_anomaly(prompt_dict_ls, device, submodules_dict)
+  File "/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/Instance_detector/test.py", line 47, in compute_anomaly
+    split(os.path.dirname(video_paths[0]), model_name)
+  File "/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/Instance_detector/split.py", line 52, in split
+    split_video(os.path.join(input_folder, file), output_folder)
+  File "/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/Instance_detector/split.py", line 27, in split_video
+    duration, fps = get_video_info(filepath)
+  File "/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/Instance_detector/split.py", line 13, in get_video_info
+    result = subprocess.run(cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True)
+  File "/root/ckpts/micromamba/envs/vbench2/lib/python3.10/subprocess.py", line 503, in run
+    with Popen(*popenargs, **kwargs) as process:
+  File "/root/ckpts/micromamba/envs/vbench2/lib/python3.10/subprocess.py", line 971, in __init__
+    self._execute_child(args, executable, preexec_fn, close_fds,
+  File "/root/ckpts/micromamba/envs/vbench2/lib/python3.10/subprocess.py", line 1863, in _execute_child
+    raise child_exception_type(errno_num, err_msg, err_filename)
+FileNotFoundError: [Errno 2] No such file or directory: 'ffprobe'

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Instance_Preservation_rerun2.log

@@ -0,0 +1,36 @@
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Instance_Preservation', dimension=['Instance_Preservation'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation/results_2026-06-08-07:27:32_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+[INFO:swift] Successfully registered `/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/Instance_detector/swift/llm/dataset/data/dataset_info.json`.
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation/results_2026-06-08-07:27:32_full_info.json
+[INFO:swift] Successfully loaded /root/ckpts/vbench2/instance_anomaly_detector/model/args.json.
+[INFO:swift] Loading the model using model_dir: /root/ckpts/huggingface/Qwen/Qwen2.5-VL-3B-Instruct
+[INFO:swift] Setting torch_dtype: torch.bfloat16
+Using a slow image processor as `use_fast` is unset and a slow processor was saved with this model. `use_fast=True` will be the default behavior in v4.52, even if the model was saved with a slow processor. This will result in minor differences in outputs. You'll still be able to use a slow processor with `use_fast=False`.
+[INFO:swift] model_kwargs: {'device_map': 'cuda:0'}
+
+[INFO:swift] Setting image_factor: 28. You can adjust this hyperparameter through the environment variable: `IMAGE_FACTOR`.
+[INFO:swift] Setting min_pixels: 3136. You can adjust this hyperparameter through the environment variable: `MIN_PIXELS`.
+[INFO:swift] Using environment variable `MAX_PIXELS`, Setting max_pixels: 1003520.
+[INFO:swift] Setting max_ratio: 200. You can adjust this hyperparameter through the environment variable: `MAX_RATIO`.
+[INFO:swift] Using environment variable `VIDEO_MIN_PIXELS`, Setting video_min_pixels: 100352.
+[INFO:swift] Using environment variable `VIDEO_MAX_PIXELS`, Setting video_max_pixels: 602112.
+[INFO:swift] Using environment variable `VIDEO_TOTAL_PIXELS`, Setting video_total_pixels: 90316800.
+[INFO:swift] Setting frame_factor: 2. You can adjust this hyperparameter through the environment variable: `FRAME_FACTOR`.
+[INFO:swift] Setting fps: 2.0. You can adjust this hyperparameter through the environment variable: `FPS`.
+[INFO:swift] Using environment variable `FPS_MIN_FRAMES`, Setting fps_min_frames: 32.
+[INFO:swift] Using environment variable `FPS_MAX_FRAMES`, Setting fps_max_frames: 50.
+[INFO:swift] default_system: 'You are a helpful and harmless assistant.'
+[INFO:swift] max_length: 128000
+[INFO:swift] response_prefix: ''
+[INFO:swift] agent_template: hermes
+[INFO:swift] norm_bbox: none
+[INFO:swift] default_system: 'You are a helpful and harmless assistant.'
+[INFO:swift] max_length: 128000
+[INFO:swift] response_prefix: ''
+[INFO:swift] agent_template: hermes
+[INFO:swift] norm_bbox: none
+
  0%|          | 0/57 [00:00<?, ?it/s]qwen-vl-utils using decord to read video.
+
  2%|▏         | 1/57 [02:13<2:04:17, 133.17s/it]

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Instance_Preservation_shard_00.log

@@ -0,0 +1,38 @@
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation_shards/shard_00', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_instance_shards/selected_info_instance_shard_00.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Instance_Preservation', dimension=['Instance_Preservation'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation_shards/shard_00/results_2026-06-08-07:32:45_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+[INFO:swift] Successfully registered `/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/Instance_detector/swift/llm/dataset/data/dataset_info.json`.
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation_shards/shard_00/results_2026-06-08-07:32:45_full_info.json
+[INFO:swift] Successfully loaded /root/ckpts/vbench2/instance_anomaly_detector/model/args.json.
+[INFO:swift] Loading the model using model_dir: /root/ckpts/huggingface/Qwen/Qwen2.5-VL-3B-Instruct
+[INFO:swift] Setting torch_dtype: torch.bfloat16
+Using a slow image processor as `use_fast` is unset and a slow processor was saved with this model. `use_fast=True` will be the default behavior in v4.52, even if the model was saved with a slow processor. This will result in minor differences in outputs. You'll still be able to use a slow processor with `use_fast=False`.
+[INFO:swift] model_kwargs: {'device_map': 'cuda:0'}
+
+[INFO:swift] Setting image_factor: 28. You can adjust this hyperparameter through the environment variable: `IMAGE_FACTOR`.
+[INFO:swift] Setting min_pixels: 3136. You can adjust this hyperparameter through the environment variable: `MIN_PIXELS`.
+[INFO:swift] Using environment variable `MAX_PIXELS`, Setting max_pixels: 1003520.
+[INFO:swift] Setting max_ratio: 200. You can adjust this hyperparameter through the environment variable: `MAX_RATIO`.
+[INFO:swift] Using environment variable `VIDEO_MIN_PIXELS`, Setting video_min_pixels: 100352.
+[INFO:swift] Using environment variable `VIDEO_MAX_PIXELS`, Setting video_max_pixels: 602112.
+[INFO:swift] Using environment variable `VIDEO_TOTAL_PIXELS`, Setting video_total_pixels: 90316800.
+[INFO:swift] Setting frame_factor: 2. You can adjust this hyperparameter through the environment variable: `FRAME_FACTOR`.
+[INFO:swift] Setting fps: 2.0. You can adjust this hyperparameter through the environment variable: `FPS`.
+[INFO:swift] Using environment variable `FPS_MIN_FRAMES`, Setting fps_min_frames: 32.
+[INFO:swift] Using environment variable `FPS_MAX_FRAMES`, Setting fps_max_frames: 50.
+[INFO:swift] default_system: 'You are a helpful and harmless assistant.'
+[INFO:swift] max_length: 128000
+[INFO:swift] response_prefix: ''
+[INFO:swift] agent_template: hermes
+[INFO:swift] norm_bbox: none
+[INFO:swift] default_system: 'You are a helpful and harmless assistant.'
+[INFO:swift] max_length: 128000
+[INFO:swift] response_prefix: ''
+[INFO:swift] agent_template: hermes
+[INFO:swift] norm_bbox: none
+
  0%|          | 0/11 [00:00<?, ?it/s]qwen-vl-utils using decord to read video.
+
  9%|▉         | 1/11 [02:02<20:24, 122.40s/it]
 18%|█▊        | 2/11 [04:03<18:14, 121.63s/it]
 27%|██▋       | 3/11 [06:13<16:44, 125.52s/it]
 36%|███▋      | 4/11 [08:27<15:01, 128.78s/it]
 45%|████▌     | 5/11 [10:38<12:58, 129.76s/it]
 55%|█████▍    | 6/11 [12:38<10:32, 126.42s/it]
 64%|██████▎   | 7/11 [14:39<08:18, 124.53s/it]
 73%|███████▎  | 8/11 [16:41<06:10, 123.66s/it]
 82%|████████▏ | 9/11 [18:31<03:58, 119.40s/it]
 91%|█████████ | 10/11 [20:38<02:01, 121.67s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation_shards/shard_00/results_2026-06-08-07:32:45_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Instance_Preservation_shard_01.log

@@ -0,0 +1,38 @@
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation_shards/shard_01', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_instance_shards/selected_info_instance_shard_01.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Instance_Preservation', dimension=['Instance_Preservation'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation_shards/shard_01/results_2026-06-08-07:32:45_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+[INFO:swift] Successfully registered `/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/Instance_detector/swift/llm/dataset/data/dataset_info.json`.
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation_shards/shard_01/results_2026-06-08-07:32:45_full_info.json
+[INFO:swift] Successfully loaded /root/ckpts/vbench2/instance_anomaly_detector/model/args.json.
+[INFO:swift] Loading the model using model_dir: /root/ckpts/huggingface/Qwen/Qwen2.5-VL-3B-Instruct
+[INFO:swift] Setting torch_dtype: torch.bfloat16
+Using a slow image processor as `use_fast` is unset and a slow processor was saved with this model. `use_fast=True` will be the default behavior in v4.52, even if the model was saved with a slow processor. This will result in minor differences in outputs. You'll still be able to use a slow processor with `use_fast=False`.
+[INFO:swift] model_kwargs: {'device_map': 'cuda:0'}
+
+[INFO:swift] Setting image_factor: 28. You can adjust this hyperparameter through the environment variable: `IMAGE_FACTOR`.
+[INFO:swift] Setting min_pixels: 3136. You can adjust this hyperparameter through the environment variable: `MIN_PIXELS`.
+[INFO:swift] Using environment variable `MAX_PIXELS`, Setting max_pixels: 1003520.
+[INFO:swift] Setting max_ratio: 200. You can adjust this hyperparameter through the environment variable: `MAX_RATIO`.
+[INFO:swift] Using environment variable `VIDEO_MIN_PIXELS`, Setting video_min_pixels: 100352.
+[INFO:swift] Using environment variable `VIDEO_MAX_PIXELS`, Setting video_max_pixels: 602112.
+[INFO:swift] Using environment variable `VIDEO_TOTAL_PIXELS`, Setting video_total_pixels: 90316800.
+[INFO:swift] Setting frame_factor: 2. You can adjust this hyperparameter through the environment variable: `FRAME_FACTOR`.
+[INFO:swift] Setting fps: 2.0. You can adjust this hyperparameter through the environment variable: `FPS`.
+[INFO:swift] Using environment variable `FPS_MIN_FRAMES`, Setting fps_min_frames: 32.
+[INFO:swift] Using environment variable `FPS_MAX_FRAMES`, Setting fps_max_frames: 50.
+[INFO:swift] default_system: 'You are a helpful and harmless assistant.'
+[INFO:swift] max_length: 128000
+[INFO:swift] response_prefix: ''
+[INFO:swift] agent_template: hermes
+[INFO:swift] norm_bbox: none
+[INFO:swift] default_system: 'You are a helpful and harmless assistant.'
+[INFO:swift] max_length: 128000
+[INFO:swift] response_prefix: ''
+[INFO:swift] agent_template: hermes
+[INFO:swift] norm_bbox: none
+
  0%|          | 0/11 [00:00<?, ?it/s]qwen-vl-utils using decord to read video.
+
  9%|▉         | 1/11 [02:03<20:32, 123.30s/it]
 18%|█▊        | 2/11 [04:04<18:17, 121.97s/it]
 27%|██▋       | 3/11 [06:16<16:53, 126.68s/it]
 36%|███▋      | 4/11 [08:29<15:04, 129.16s/it]
 45%|████▌     | 5/11 [10:40<12:58, 129.83s/it]
 55%|█████▍    | 6/11 [12:41<10:33, 126.69s/it]
 64%|██████▎   | 7/11 [14:41<08:18, 124.50s/it]
 73%|███████▎  | 8/11 [16:42<06:10, 123.54s/it]
 82%|████████▏ | 9/11 [18:49<04:09, 124.70s/it]
 91%|█████████ | 10/11 [20:55<02:04, 124.97s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation_shards/shard_01/results_2026-06-08-07:32:45_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Instance_Preservation_shard_02.log

@@ -0,0 +1,38 @@
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation_shards/shard_02', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_instance_shards/selected_info_instance_shard_02.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Instance_Preservation', dimension=['Instance_Preservation'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation_shards/shard_02/results_2026-06-08-07:32:45_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+[INFO:swift] Successfully registered `/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/Instance_detector/swift/llm/dataset/data/dataset_info.json`.
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation_shards/shard_02/results_2026-06-08-07:32:45_full_info.json
+[INFO:swift] Successfully loaded /root/ckpts/vbench2/instance_anomaly_detector/model/args.json.
+[INFO:swift] Loading the model using model_dir: /root/ckpts/huggingface/Qwen/Qwen2.5-VL-3B-Instruct
+[INFO:swift] Setting torch_dtype: torch.bfloat16
+Using a slow image processor as `use_fast` is unset and a slow processor was saved with this model. `use_fast=True` will be the default behavior in v4.52, even if the model was saved with a slow processor. This will result in minor differences in outputs. You'll still be able to use a slow processor with `use_fast=False`.
+[INFO:swift] model_kwargs: {'device_map': 'cuda:0'}
+
+[INFO:swift] Setting image_factor: 28. You can adjust this hyperparameter through the environment variable: `IMAGE_FACTOR`.
+[INFO:swift] Setting min_pixels: 3136. You can adjust this hyperparameter through the environment variable: `MIN_PIXELS`.
+[INFO:swift] Using environment variable `MAX_PIXELS`, Setting max_pixels: 1003520.
+[INFO:swift] Setting max_ratio: 200. You can adjust this hyperparameter through the environment variable: `MAX_RATIO`.
+[INFO:swift] Using environment variable `VIDEO_MIN_PIXELS`, Setting video_min_pixels: 100352.
+[INFO:swift] Using environment variable `VIDEO_MAX_PIXELS`, Setting video_max_pixels: 602112.
+[INFO:swift] Using environment variable `VIDEO_TOTAL_PIXELS`, Setting video_total_pixels: 90316800.
+[INFO:swift] Setting frame_factor: 2. You can adjust this hyperparameter through the environment variable: `FRAME_FACTOR`.
+[INFO:swift] Setting fps: 2.0. You can adjust this hyperparameter through the environment variable: `FPS`.
+[INFO:swift] Using environment variable `FPS_MIN_FRAMES`, Setting fps_min_frames: 32.
+[INFO:swift] Using environment variable `FPS_MAX_FRAMES`, Setting fps_max_frames: 50.
+[INFO:swift] default_system: 'You are a helpful and harmless assistant.'
+[INFO:swift] max_length: 128000
+[INFO:swift] response_prefix: ''
+[INFO:swift] agent_template: hermes
+[INFO:swift] norm_bbox: none
+[INFO:swift] default_system: 'You are a helpful and harmless assistant.'
+[INFO:swift] max_length: 128000
+[INFO:swift] response_prefix: ''
+[INFO:swift] agent_template: hermes
+[INFO:swift] norm_bbox: none
+
  0%|          | 0/11 [00:00<?, ?it/s]qwen-vl-utils using decord to read video.
+
  9%|▉         | 1/11 [01:10<11:43, 70.34s/it]
 18%|█▊        | 2/11 [02:37<12:02, 80.31s/it]
 27%|██▋       | 3/11 [03:19<08:21, 62.67s/it]
 36%|███▋      | 4/11 [05:06<09:20, 80.09s/it]
 45%|████▌     | 5/11 [07:18<09:52, 98.81s/it]
 55%|█████▍    | 6/11 [09:31<09:12, 110.50s/it]
 64%|██████▎   | 7/11 [11:38<07:43, 115.89s/it]
 73%|███████▎  | 8/11 [13:37<05:51, 117.04s/it]
 82%|████████▏ | 9/11 [15:38<03:56, 118.20s/it]
 91%|█████████ | 10/11 [17:45<02:00, 120.98s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation_shards/shard_02/results_2026-06-08-07:32:45_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Instance_Preservation_shard_04.log

@@ -0,0 +1,38 @@
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation_shards/shard_04', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_instance_shards/selected_info_instance_shard_04.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Instance_Preservation', dimension=['Instance_Preservation'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation_shards/shard_04/results_2026-06-08-07:32:45_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+[INFO:swift] Successfully registered `/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/Instance_detector/swift/llm/dataset/data/dataset_info.json`.
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation_shards/shard_04/results_2026-06-08-07:32:45_full_info.json
+[INFO:swift] Successfully loaded /root/ckpts/vbench2/instance_anomaly_detector/model/args.json.
+[INFO:swift] Loading the model using model_dir: /root/ckpts/huggingface/Qwen/Qwen2.5-VL-3B-Instruct
+[INFO:swift] Setting torch_dtype: torch.bfloat16
+Using a slow image processor as `use_fast` is unset and a slow processor was saved with this model. `use_fast=True` will be the default behavior in v4.52, even if the model was saved with a slow processor. This will result in minor differences in outputs. You'll still be able to use a slow processor with `use_fast=False`.
+[INFO:swift] model_kwargs: {'device_map': 'cuda:0'}
+
+[INFO:swift] Setting image_factor: 28. You can adjust this hyperparameter through the environment variable: `IMAGE_FACTOR`.
+[INFO:swift] Setting min_pixels: 3136. You can adjust this hyperparameter through the environment variable: `MIN_PIXELS`.
+[INFO:swift] Using environment variable `MAX_PIXELS`, Setting max_pixels: 1003520.
+[INFO:swift] Setting max_ratio: 200. You can adjust this hyperparameter through the environment variable: `MAX_RATIO`.
+[INFO:swift] Using environment variable `VIDEO_MIN_PIXELS`, Setting video_min_pixels: 100352.
+[INFO:swift] Using environment variable `VIDEO_MAX_PIXELS`, Setting video_max_pixels: 602112.
+[INFO:swift] Using environment variable `VIDEO_TOTAL_PIXELS`, Setting video_total_pixels: 90316800.
+[INFO:swift] Setting frame_factor: 2. You can adjust this hyperparameter through the environment variable: `FRAME_FACTOR`.
+[INFO:swift] Setting fps: 2.0. You can adjust this hyperparameter through the environment variable: `FPS`.
+[INFO:swift] Using environment variable `FPS_MIN_FRAMES`, Setting fps_min_frames: 32.
+[INFO:swift] Using environment variable `FPS_MAX_FRAMES`, Setting fps_max_frames: 50.
+[INFO:swift] default_system: 'You are a helpful and harmless assistant.'
+[INFO:swift] max_length: 128000
+[INFO:swift] response_prefix: ''
+[INFO:swift] agent_template: hermes
+[INFO:swift] norm_bbox: none
+[INFO:swift] default_system: 'You are a helpful and harmless assistant.'
+[INFO:swift] max_length: 128000
+[INFO:swift] response_prefix: ''
+[INFO:swift] agent_template: hermes
+[INFO:swift] norm_bbox: none
+
  0%|          | 0/12 [00:00<?, ?it/s]qwen-vl-utils using decord to read video.
+
  8%|▊         | 1/12 [02:03<22:33, 123.04s/it]
 17%|█▋        | 2/12 [04:03<20:15, 121.55s/it]
 25%|██▌       | 3/12 [06:12<18:44, 124.96s/it]
 33%|███▎      | 4/12 [08:25<17:04, 128.10s/it]
 42%|████▏     | 5/12 [10:38<15:09, 129.94s/it]
 50%|█████     | 6/12 [12:39<12:41, 126.93s/it]
 58%|█████▊    | 7/12 [14:39<10:22, 124.52s/it]
 67%|██████▋   | 8/12 [16:41<08:15, 123.80s/it]
 75%|███████▌  | 9/12 [18:49<06:14, 124.95s/it]
 83%|████████▎ | 10/12 [20:55<04:10, 125.30s/it]
 92%|█████████▏| 11/12 [22:52<02:02, 122.77s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation_shards/shard_04/results_2026-06-08-07:32:45_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Material.log

@@ -0,0 +1,21 @@
+[2026-06-08 07:01:05 UTC] COMMAND /root/ckpts/micromamba/envs/vbench2/bin/python -u evaluate.py --videos_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Material --dimension Material --output_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Material --full_json_dir /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json --mode vbench_standard
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Material', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Material', dimension=['Material'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+File exists: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Material/results_2026-06-08-07:01:08_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/LLaVA_NeXT/llava/model/llava_arch.py:215: SyntaxWarning: "is not" with a literal. Did you mean "!="?
+  if slower_img_feat is not 0:
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Material/results_2026-06-08-07:01:08_full_info.json
+Loaded LLaVA model: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+You are using a model of type llava to instantiate a model of type llava_qwen. This is not supported for all configurations of models and can yield errors.
+Sliding Window Attention is enabled but not implemented for `sdpa`; unexpected results may be encountered.
+Loading vision tower: google/siglip-so400m-patch14-384
+2026-06-08 07:01:13,648 - accelerate.utils.modeling - INFO - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk).
+
+Model Class: LlavaQwenForCausalLM
+
  0%|          | 0/50 [00:00<?, ?it/s]The attention mask is not set and cannot be inferred from input because pad token is same as eos token. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
+
  2%|▏         | 1/50 [00:16<13:09, 16.11s/it]
  4%|▍         | 2/50 [00:31<12:32, 15.68s/it]
  6%|▌         | 3/50 [00:46<12:07, 15.47s/it]
  8%|▊         | 4/50 [01:02<11:51, 15.47s/it]
 10%|█         | 5/50 [01:17<11:39, 15.55s/it]
 12%|█▏        | 6/50 [01:33<11:23, 15.54s/it]
 14%|█▍        | 7/50 [01:48<11:05, 15.48s/it]
 16%|█▌        | 8/50 [02:04<10:48, 15.45s/it]
 18%|█▊        | 9/50 [02:19<10:32, 15.43s/it]
 20%|██        | 10/50 [02:34<10:16, 15.42s/it]
 22%|██▏       | 11/50 [02:50<10:01, 15.42s/it]
 24%|██▍       | 12/50 [03:05<09:45, 15.41s/it]
 26%|██▌       | 13/50 [03:21<09:30, 15.41s/it]
 28%|██▊       | 14/50 [03:36<09:13, 15.37s/it]
 30%|███       | 15/50 [03:51<08:56, 15.33s/it]
 32%|███▏      | 16/50 [04:07<08:42, 15.37s/it]
 34%|███▍      | 17/50 [04:23<08:34, 15.58s/it]
 36%|███▌      | 18/50 [04:39<08:24, 15.76s/it]
 38%|███▊      | 19/50 [04:55<08:10, 15.82s/it]
 40%|████      | 20/50 [05:10<07:52, 15.76s/it]
 42%|████▏     | 21/50 [05:26<07:39, 15.84s/it]
 44%|████▍     | 22/50 [05:42<07:20, 15.74s/it]
 46%|████▌     | 23/50 [05:58<07:05, 15.78s/it]
 48%|████▊     | 24/50 [06:14<06:52, 15.88s/it]
 50%|█████     | 25/50 [06:30<06:36, 15.86s/it]
 52%|█████▏    | 26/50 [06:46<06:21, 15.88s/it]
 54%|█████▍    | 27/50 [07:02<06:06, 15.95s/it]
 56%|█████▌    | 28/50 [07:17<05:48, 15.85s/it]
 58%|█████▊    | 29/50 [07:33<05:33, 15.90s/it]
 60%|██████    | 30/50 [07:49<05:17, 15.86s/it]
 62%|██████▏   | 31/50 [08:06<05:04, 16.01s/it]
 64%|██████▍   | 32/50 [08:22<04:50, 16.13s/it]
 66%|██████▌   | 33/50 [08:38<04:33, 16.07s/it]
 68%|██████▊   | 34/50 [08:54<04:17, 16.11s/it]
 70%|███████   | 35/50 [09:10<04:01, 16.10s/it]
 72%|███████▏  | 36/50 [09:26<03:45, 16.10s/it]
 74%|███████▍  | 37/50 [09:42<03:28, 16.01s/it]
 76%|███████▌  | 38/50 [09:58<03:11, 15.98s/it]
 78%|███████▊  | 39/50 [10:14<02:55, 15.97s/it]
 80%|████████  | 40/50 [10:29<02:38, 15.83s/it]
 82%|████████▏ | 41/50 [10:45<02:20, 15.66s/it]
 84%|████████▍ | 42/50 [11:00<02:04, 15.57s/it]
 86%|████████▌ | 43/50 [11:15<01:48, 15.45s/it]
 88%|████████▊ | 44/50 [11:30<01:32, 15.35s/it]
 90%|█████████ | 45/50 [11:39<01:07, 13.46s/it]
 92%|█████████▏| 46/50 [11:48<00:48, 12.14s/it]
 94%|█████████▍| 47/50 [11:57<00:33, 11.18s/it]
 96%|█████████▌| 48/50 [12:06<00:20, 10.49s/it]
 98%|█████████▊| 49/50 [12:15<00:09,  9.98s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Material/results_2026-06-08-07:01:08_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Mechanics.log

@@ -0,0 +1,21 @@
+[2026-06-08 06:46:35 UTC] COMMAND /root/ckpts/micromamba/envs/vbench2/bin/python -u evaluate.py --videos_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Mechanics --dimension Mechanics --output_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Mechanics --full_json_dir /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json --mode vbench_standard
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Mechanics', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Mechanics', dimension=['Mechanics'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+File exists: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Mechanics/results_2026-06-08-06:46:38_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/LLaVA_NeXT/llava/model/llava_arch.py:215: SyntaxWarning: "is not" with a literal. Did you mean "!="?
+  if slower_img_feat is not 0:
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Mechanics/results_2026-06-08-06:46:38_full_info.json
+Loaded LLaVA model: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+You are using a model of type llava to instantiate a model of type llava_qwen. This is not supported for all configurations of models and can yield errors.
+Sliding Window Attention is enabled but not implemented for `sdpa`; unexpected results may be encountered.
+Loading vision tower: google/siglip-so400m-patch14-384
+2026-06-08 06:46:44,838 - accelerate.utils.modeling - INFO - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk).
+
+Model Class: LlavaQwenForCausalLM
+
  0%|          | 0/57 [00:00<?, ?it/s]The attention mask is not set and cannot be inferred from input because pad token is same as eos token. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
+
  2%|▏         | 1/57 [00:17<16:12, 17.37s/it]
  4%|▎         | 2/57 [00:33<15:19, 16.71s/it]
  5%|▌         | 3/57 [00:50<15:04, 16.75s/it]
  7%|▋         | 4/57 [01:06<14:43, 16.67s/it]
  9%|▉         | 5/57 [01:24<14:34, 16.81s/it]
 11%|█         | 6/57 [01:40<14:18, 16.84s/it]
 12%|█▏        | 7/57 [01:57<14:03, 16.86s/it]
 14%|█▍        | 8/57 [02:14<13:39, 16.72s/it]
 16%|█▌        | 9/57 [02:24<11:37, 14.54s/it]
 18%|█▊        | 10/57 [02:34<10:24, 13.29s/it]
 19%|█▉        | 11/57 [02:44<09:23, 12.25s/it]
 21%|██        | 12/57 [02:54<08:44, 11.65s/it]
 23%|██▎       | 13/57 [03:11<09:35, 13.09s/it]
 25%|██▍       | 14/57 [03:21<08:44, 12.19s/it]
 26%|██▋       | 15/57 [03:31<08:03, 11.51s/it]
 28%|██▊       | 16/57 [03:47<08:46, 12.83s/it]
 30%|██▉       | 17/57 [04:02<09:08, 13.71s/it]
 32%|███▏      | 18/57 [04:18<09:17, 14.29s/it]
 33%|███▎      | 19/57 [04:34<09:21, 14.77s/it]
 35%|███▌      | 20/57 [04:50<09:17, 15.06s/it]
 37%|███▋      | 21/57 [05:05<09:09, 15.25s/it]
 39%|███▊      | 22/57 [05:15<07:51, 13.47s/it]
 40%|████      | 23/57 [05:24<06:54, 12.19s/it]
 42%|████▏     | 24/57 [05:33<06:13, 11.33s/it]
 44%|████▍     | 25/57 [05:42<05:43, 10.72s/it]
 46%|████▌     | 26/57 [05:52<05:17, 10.25s/it]
 47%|████���     | 27/57 [06:01<04:59,  9.97s/it]
 49%|████▉     | 28/57 [06:10<04:42,  9.74s/it]
 51%|█████     | 29/57 [06:20<04:30,  9.66s/it]
 53%|█████▎    | 30/57 [06:35<05:09, 11.47s/it]
 54%|█████▍    | 31/57 [06:51<05:31, 12.76s/it]
 56%|█████▌    | 32/57 [07:07<05:41, 13.65s/it]
 58%|█████▊    | 33/57 [07:22<05:42, 14.27s/it]
 60%|█████▉    | 34/57 [07:38<05:37, 14.68s/it]
 61%|██████▏   | 35/57 [07:54<05:30, 15.00s/it]
 63%|██████▎   | 36/57 [08:10<05:22, 15.37s/it]
 65%|██████▍   | 37/57 [08:27<05:14, 15.70s/it]
 67%|██████▋   | 38/57 [08:43<05:02, 15.94s/it]
 68%|██████▊   | 39/57 [09:00<04:50, 16.16s/it]
 70%|███████   | 40/57 [09:16<04:35, 16.23s/it]
 72%|███████▏  | 41/57 [09:33<04:21, 16.37s/it]
 74%|███████▎  | 42/57 [09:49<04:05, 16.39s/it]
 75%|███████▌  | 43/57 [10:06<03:49, 16.37s/it]
 77%|███████▋  | 44/57 [10:22<03:32, 16.35s/it]
 79%|███████▉  | 45/57 [10:38<03:17, 16.43s/it]
 81%|████████  | 46/57 [10:55<03:01, 16.53s/it]
 82%|████████▏ | 47/57 [11:12<02:46, 16.66s/it]
 84%|████████▍ | 48/57 [11:29<02:29, 16.64s/it]
 86%|████████▌ | 49/57 [11:45<02:12, 16.54s/it]
 88%|████████▊ | 50/57 [12:02<01:55, 16.52s/it]
 89%|████████▉ | 51/57 [12:18<01:38, 16.43s/it]
 91%|█████████ | 52/57 [12:34<01:21, 16.40s/it]
 93%|█████████▎| 53/57 [12:50<01:05, 16.34s/it]
 95%|█████████▍| 54/57 [13:07<00:48, 16.33s/it]
 96%|█████████▋| 55/57 [13:23<00:32, 16.30s/it]
 98%|█████████▊| 56/57 [13:39<00:16, 16.28s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Mechanics/results_2026-06-08-06:46:38_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Motion_Order_Understanding.log

@@ -0,0 +1,24 @@
+[2026-06-08 06:26:03 UTC] COMMAND /root/ckpts/micromamba/envs/vbench2/bin/python -u evaluate.py --videos_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Motion_Order_Understanding --dimension Motion_Order_Understanding --output_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Motion_Order_Understanding --full_json_dir /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json --mode vbench_standard
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Motion_Order_Understanding', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Motion_Order_Understanding', dimension=['Motion_Order_Understanding'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+File exists: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+File exists: /root/ckpts/vbench2/Qwen/Qwen2.5-7B-Instruct
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Motion_Order_Understanding/results_2026-06-08-06:26:07_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/LLaVA_NeXT/llava/model/llava_arch.py:215: SyntaxWarning: "is not" with a literal. Did you mean "!="?
+  if slower_img_feat is not 0:
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Motion_Order_Understanding/results_2026-06-08-06:26:07_full_info.json
+Loaded LLaVA model: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+You are using a model of type llava to instantiate a model of type llava_qwen. This is not supported for all configurations of models and can yield errors.
+Sliding Window Attention is enabled but not implemented for `sdpa`; unexpected results may be encountered.
+Loading vision tower: google/siglip-so400m-patch14-384
+2026-06-08 06:26:13,478 - accelerate.utils.modeling - INFO - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk).
+
+Model Class: LlavaQwenForCausalLM
+2026-06-08 06:26:19,877 - accelerate.utils.modeling - INFO - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk).
+
+
  0%|          | 0/99 [00:00<?, ?it/s]The attention mask is not set and cannot be inferred from input because pad token is same as eos token. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
+
  1%|          | 1/99 [00:23<38:36, 23.64s/it]
  2%|▏         | 2/99 [00:45<36:54, 22.83s/it]
  3%|▎         | 3/99 [01:08<36:08, 22.58s/it]
  4%|▍         | 4/99 [01:28<34:24, 21.73s/it]
  5%|▌         | 5/99 [01:49<33:22, 21.30s/it]
  6%|▌         | 6/99 [02:12<34:02, 21.96s/it]
  7%|▋         | 7/99 [02:32<32:44, 21.35s/it]
  8%|▊         | 8/99 [02:51<31:10, 20.55s/it]
  9%|▉         | 9/99 [03:09<29:31, 19.69s/it]
 10%|█         | 10/99 [03:29<29:20, 19.78s/it]
 11%|█         | 11/99 [03:52<30:30, 20.80s/it]
 12%|█▏        | 12/99 [04:11<29:21, 20.24s/it]
 13%|█▎        | 13/99 [04:29<28:19, 19.76s/it]
 14%|█▍        | 14/99 [04:48<27:22, 19.32s/it]
 15%|█▌        | 15/99 [05:05<26:22, 18.83s/it]
 16%|█▌        | 16/99 [05:26<26:56, 19.47s/it]
 17%|█▋        | 17/99 [05:45<26:16, 19.22s/it]
 18%|█▊        | 18/99 [06:05<26:05, 19.32s/it]
 19%|█▉        | 19/99 [06:23<25:25, 19.07s/it]
 20%|██        | 20/99 [06:40<24:29, 18.60s/it]
 21%|██        | 21/99 [06:58<23:53, 18.38s/it]
 22%|██▏       | 22/99 [07:21<25:21, 19.77s/it]
 23%|██▎       | 23/99 [07:43<25:51, 20.41s/it]
 24%|██▍       | 24/99 [08:05<26:09, 20.93s/it]
 25%|██▌       | 25/99 [08:27<26:10, 21.22s/it]
 26%|██▋       | 26/99 [08:50<26:14, 21.57s/it]
 27%|██▋       | 27/99 [09:10<25:34, 21.31s/it]
 28%|██▊       | 28/99 [09:29<24:17, 20.53s/it]
 29%|██▉       | 29/99 [09:53<25:05, 21.51s/it]
 30%|███       | 30/99 [10:13<24:09, 21.00s/it]
 31%|███▏      | 31/99 [10:38<25:06, 22.15s/it]
 32%|███▏      | 32/99 [11:01<25:10, 22.55s/it]
 33%|███▎      | 33/99 [11:21<23:53, 21.71s/it]
 34%|███▍      | 34/99 [11:45<24:12, 22.35s/it]
 35%|███▌      | 35/99 [12:06<23:27, 21.99s/it]
 36%|███▋      | 36/99 [12:28<23:15, 22.15s/it]
 37%|███▋      | 37/99 [12:49<22:27, 21.73s/it]
 38%|███▊      | 38/99 [13:10<21:46, 21.41s/it]
 39%|███▉      | 39/99 [13:29<20:40, 20.67s/it]
 40%|████      | 40/99 [13:46<19:25, 19.75s/it]
 41%|████▏     | 41/99 [14:04<18:33, 19.20s/it]
 42%|████▏     | 42/99 [14:24<18:16, 19.24s/it]
 43%|████▎     | 43/99 [14:41<17:32, 18.79s/it]
 44%|████▍     | 44/99 [14:59<17:02, 18.59s/it]
 45%|████▌     | 45/99 [15:18<16:48, 18.67s/it]
 46%|████▋     | 46/99 [15:36<16:14, 18.38s/it]
 47%|████▋     | 47/99 [15:54<15:55, 18.38s/it]
 48%|████▊     | 48/99 [16:12<15:26, 18.17s/it]
 49%|████▉     | 49/99 [16:31<15:23, 18.47s/it]
 51%|█████     | 50/99 [16:49<15:00, 18.37s/it]
 52%|█████▏    | 51/99 [17:09<15:02, 18.80s/it]
 53%|█████▎    | 52/99 [17:29<15:00, 19.16s/it]
 54%|█████▎    | 53/99 [17:50<15:11, 19.81s/it]
 55%|█████▍    | 54/99 [18:11<15:03, 20.08s/it]
 56%|█████▌    | 55/99 [18:33<15:04, 20.55s/it]
 57%|█████▋    | 56/99 [18:55<15:03, 21.01s/it]
 58%|█████▊    | 57/99 [19:16<14:37, 20.89s/it]
 59%|█████▊    | 58/99 [19:39<14:47, 21.64s/it]
 60%|█████▉    | 59/99 [20:00<14:18, 21.46s/it]
 61%|██████    | 60/99 [20:23<14:12, 21.85s/it]
 62%|██████▏   | 61/99 [20:45<13:55, 21.97s/it]
 63%|██████▎   | 62/99 [21:09<13:53, 22.52s/it]
 64%|██████▎   | 63/99 [21:29<13:07, 21.88s/it]
 65%|██████▍   | 64/99 [21:49<12:23, 21.23s/it]
 66%|██████▌   | 65/99 [22:09<11:51, 20.94s/it]
 67%|██████▋   | 66/99 [22:31<11:39, 21.21s/it]
 68%|██████▊   | 67/99 [22:53<11:28, 21.52s/it]
 69%|██████▊   | 68/99 [23:16<11:18, 21.88s/it]
 70%|██████▉   | 69/99 [23:38<10:55, 21.86s/it]
 71%|███████   | 70/99 [23:57<10:13, 21.14s/it]
 72%|███████▏  | 71/99 [24:19<09:54, 21.23s/it]
 73%|███████▎  | 72/99 [24:39<09:23, 20.87s/it]
 74%|███████▎  | 73/99 [24:58<08:47, 20.28s/it]
 75%|███████▍  | 74/99 [25:18<08:27, 20.30s/it]
 76%|███████▌  | 75/99 [25:36<07:51, 19.63s/it]
 77%|███████▋  | 76/99 [25:54<07:20, 19.13s/it]
 78%|███████▊  | 77/99 [26:13<06:59, 19.06s/it]
 79%|███████▉  | 78/99 [26:33<06:44, 19.29s/it]
 80%|███████▉  | 79/99 [26:51<06:21, 19.06s/it]
 81%|████████  | 80/99 [27:09<05:55, 18.69s/it]
 82%|████████▏ | 81/99 [27:30<05:48, 19.34s/it]
 83%|████████▎ | 82/99 [27:49<05:27, 19.28s/it]
 84%|████████▍ | 83/99 [28:08<05:06, 19.19s/it]
 85%|████████▍ | 84/99 [28:29<04:54, 19.64s/it]
 86%|████████▌ | 85/99 [28:47<04:27, 19.14s/it]
 87%|████████▋ | 86/99 [29:08<04:16, 19.73s/it]
 88%|████████▊ | 87/99 [29:31<04:08, 20.68s/it]
 89%|████████▉ | 88/99 [29:52<03:51, 21.00s/it]
 90%|████████▉ | 89/99 [30:13<03:27, 20.73s/it]
 91%|█████████ | 90/99 [30:34<03:07, 20.85s/it]
 92%|█████████▏| 91/99 [30:56<02:49, 21.15s/it]
 93%|█████████▎| 92/99 [31:16<02:25, 20.83s/it]
 94%|█████████▍| 93/99 [31:39<02:09, 21.58s/it]
 95%|█████████▍| 94/99 [32:00<01:46, 21.33s/it]
 96%|█████████▌| 95/99 [32:21<01:25, 21.29s/it]
 97%|███████��█▋| 96/99 [32:42<01:03, 21.20s/it]
 98%|█████████▊| 97/99 [33:03<00:42, 21.21s/it]
 99%|█████████▉| 98/99 [33:25<00:21, 21.43s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Motion_Order_Understanding/results_2026-06-08-06:26:07_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Motion_Rationality.log

@@ -0,0 +1,21 @@
+[2026-06-08 06:38:04 UTC] COMMAND /root/ckpts/micromamba/envs/vbench2/bin/python -u evaluate.py --videos_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Motion_Rationality --dimension Motion_Rationality --output_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Motion_Rationality --full_json_dir /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json --mode vbench_standard
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Motion_Rationality', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Motion_Rationality', dimension=['Motion_Rationality'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+File exists: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Motion_Rationality/results_2026-06-08-06:38:07_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/LLaVA_NeXT/llava/model/llava_arch.py:215: SyntaxWarning: "is not" with a literal. Did you mean "!="?
+  if slower_img_feat is not 0:
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Motion_Rationality/results_2026-06-08-06:38:07_full_info.json
+Loaded LLaVA model: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+You are using a model of type llava to instantiate a model of type llava_qwen. This is not supported for all configurations of models and can yield errors.
+Sliding Window Attention is enabled but not implemented for `sdpa`; unexpected results may be encountered.
+Loading vision tower: google/siglip-so400m-patch14-384
+2026-06-08 06:38:14,658 - accelerate.utils.modeling - INFO - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk).
+
+Model Class: LlavaQwenForCausalLM
+
  0%|          | 0/58 [00:00<?, ?it/s]The attention mask is not set and cannot be inferred from input because pad token is same as eos token. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
+
  2%|▏         | 1/58 [00:30<29:05, 30.62s/it]
  3%|▎         | 2/58 [01:00<28:10, 30.19s/it]
  5%|▌         | 3/58 [01:30<27:25, 29.92s/it]
  7%|▋         | 4/58 [01:59<26:34, 29.52s/it]
  9%|▊         | 5/58 [02:28<25:54, 29.34s/it]
 10%|█         | 6/58 [02:56<25:18, 29.21s/it]
 12%|█▏        | 7/58 [03:26<24:49, 29.20s/it]
 14%|█▍        | 8/58 [03:55<24:18, 29.18s/it]
 16%|█▌        | 9/58 [04:24<23:49, 29.17s/it]
 17%|█▋        | 10/58 [04:53<23:22, 29.22s/it]
 19%|█▉        | 11/58 [05:22<22:52, 29.20s/it]
 21%|██        | 12/58 [05:52<22:26, 29.26s/it]
 22%|██▏       | 13/58 [06:22<22:12, 29.62s/it]
 24%|██▍       | 14/58 [06:52<21:47, 29.73s/it]
 26%|██▌       | 15/58 [07:22<21:21, 29.79s/it]
 28%|██▊       | 16/58 [07:52<20:51, 29.79s/it]
 29%|██▉       | 17/58 [08:22<20:23, 29.84s/it]
 31%|███       | 18/58 [08:52<19:52, 29.81s/it]
 33%|███▎      | 19/58 [09:22<19:24, 29.86s/it]
 34%|███▍      | 20/58 [09:52<18:57, 29.94s/it]
 36%|███▌      | 21/58 [10:22<18:31, 30.04s/it]
 38%|███▊      | 22/58 [10:52<17:57, 29.92s/it]
 40%|███▉      | 23/58 [11:22<17:29, 29.98s/it]
 41%|████▏     | 24/58 [11:52<16:57, 29.92s/it]
 43%|████▎     | 25/58 [12:21<16:23, 29.81s/it]
 45%|���███▍     | 26/58 [12:50<15:45, 29.56s/it]
 47%|████▋     | 27/58 [13:19<15:12, 29.45s/it]
 48%|████▊     | 28/58 [13:48<14:38, 29.27s/it]
 50%|█████     | 29/58 [14:17<14:06, 29.20s/it]
 52%|█████▏    | 30/58 [14:46<13:35, 29.14s/it]
 53%|█████▎    | 31/58 [15:15<13:06, 29.13s/it]
 55%|█████▌    | 32/58 [15:45<12:38, 29.18s/it]
 57%|█████▋    | 33/58 [16:14<12:10, 29.21s/it]
 59%|█████▊    | 34/58 [16:44<11:46, 29.42s/it]
 60%|██████    | 35/58 [17:14<11:20, 29.57s/it]
 62%|██████▏   | 36/58 [17:44<10:53, 29.73s/it]
 64%|██████▍   | 37/58 [18:13<10:22, 29.63s/it]
 66%|██████▌   | 38/58 [18:43<09:52, 29.64s/it]
 67%|██████▋   | 39/58 [19:13<09:25, 29.78s/it]
 69%|██████▉   | 40/58 [19:43<08:59, 29.99s/it]
 71%|███████   | 41/58 [20:14<08:31, 30.06s/it]
 72%|███████▏  | 42/58 [20:43<07:58, 29.89s/it]
 74%|███████▍  | 43/58 [21:13<07:27, 29.84s/it]
 76%|███████▌  | 44/58 [21:43<06:56, 29.78s/it]
 78%|███████▊  | 45/58 [22:12<06:27, 29.79s/it]
 79%|███████▉  | 46/58 [22:42<05:57, 29.80s/it]
 81%|████████  | 47/58 [23:11<05:25, 29.60s/it]
 83%|████████▎ | 48/58 [23:40<04:54, 29.40s/it]
 84%|████████▍ | 49/58 [24:09<04:23, 29.30s/it]
 86%|████████▌ | 50/58 [24:38<03:53, 29.20s/it]
 88%|████████▊ | 51/58 [25:07<03:23, 29.10s/it]
 90%|████████▉ | 52/58 [25:36<02:54, 29.09s/it]
 91%|█████████▏| 53/58 [26:05<02:25, 29.08s/it]
 93%|█████████▎| 54/58 [26:34<01:56, 29.07s/it]
 95%|█████████▍| 55/58 [27:03<01:27, 29.04s/it]
 97%|█████████▋| 56/58 [27:33<00:58, 29.27s/it]
 98%|█████████▊| 57/58 [28:03<00:29, 29.36s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Motion_Rationality/results_2026-06-08-06:38:07_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Multi-View_Consistency.log

@@ -0,0 +1,302 @@
+[2026-06-08 07:01:35 UTC] COMMAND /root/ckpts/micromamba/envs/vbench2/bin/python -u evaluate.py --videos_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Multi-View_Consistency --dimension Multi-View_Consistency --output_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Multi-View_Consistency --full_json_dir /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json --mode vbench_standard
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Multi-View_Consistency', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Multi-View_Consistency', dimension=['Multi-View_Consistency'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Multi-View_Consistency/results_2026-06-08-07:01:38_full_info.json
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Multi-View_Consistency/results_2026-06-08-07:01:38_full_info.json
+Using cache found in /root/ckpts/torch/hub/facebookresearch_co-tracker_main
+
  0%|          | 0/97 [00:00<?, ?it/s]2026-06-08 07:01:45,054 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:01:49,283 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:01:53,224 - pyscenedetect - INFO - Detecting scenes...
+
  1%|          | 1/97 [00:12<19:34, 12.23s/it]2026-06-08 07:01:57,278 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:02:01,260 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:02:05,213 - pyscenedetect - INFO - Detecting scenes...
+
  2%|▏         | 2/97 [00:24<19:10, 12.11s/it]2026-06-08 07:02:09,301 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:02:13,243 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:02:17,227 - pyscenedetect - INFO - Detecting scenes...
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/torch/functional.py:505: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at /pytorch/aten/src/ATen/native/TensorShape.cpp:4317.)
+  return _VF.meshgrid(tensors, **kwargs)  # type: ignore[attr-defined]
+
  3%|▎         | 3/97 [00:40<21:56, 14.01s/it]2026-06-08 07:02:25,566 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:02:32,167 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:02:36,111 - pyscenedetect - INFO - Detecting scenes...
+
  4%|▍         | 4/97 [00:54<21:58, 14.18s/it]2026-06-08 07:02:40,015 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:02:43,991 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:02:48,222 - pyscenedetect - INFO - Detecting scenes...
+
  5%|▌         | 5/97 [01:07<20:37, 13.45s/it]2026-06-08 07:02:52,178 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:02:56,075 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:02:59,974 - pyscenedetect - INFO - Detecting scenes...
+
  6%|▌         | 6/97 [01:18<19:28, 12.84s/it]2026-06-08 07:03:03,833 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:03:07,781 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:03:11,736 - pyscenedetect - INFO - Detecting scenes...
+
  7%|▋         | 7/97 [01:30<18:45, 12.51s/it]2026-06-08 07:03:15,650 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:03:19,512 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:03:27,672 - pyscenedetect - INFO - Detecting scenes...
+
  8%|▊         | 8/97 [01:51<22:36, 15.24s/it]2026-06-08 07:03:36,768 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:03:46,463 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:03:50,440 - pyscenedetect - INFO - Detecting scenes...
+
  9%|▉         | 9/97 [02:09<23:26, 15.98s/it]2026-06-08 07:03:54,356 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:03:58,348 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:04:02,351 - pyscenedetect - INFO - Detecting scenes...
+
 10%|█         | 10/97 [02:21<21:21, 14.73s/it]2026-06-08 07:04:06,280 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:04:10,229 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:04:14,211 - pyscenedetect - INFO - Detecting scenes...
+
 11%|█▏        | 11/97 [02:33<19:50, 13.84s/it]2026-06-08 07:04:18,119 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:04:22,096 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:04:26,031 - pyscenedetect - INFO - Detecting scenes...
+
 12%|█▏        | 12/97 [02:44<18:44, 13.23s/it]2026-06-08 07:04:29,943 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:04:37,289 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:04:41,205 - pyscenedetect - INFO - Detecting scenes...
+
 13%|█▎        | 13/97 [03:00<19:20, 13.81s/it]2026-06-08 07:04:45,099 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:04:48,977 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:04:53,079 - pyscenedetect - INFO - Detecting scenes...
+
 14%|█▍        | 14/97 [03:16<20:01, 14.47s/it]2026-06-08 07:05:01,124 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:05:05,155 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:05:09,032 - pyscenedetect - INFO - Detecting scenes...
+
 15%|█▌        | 15/97 [03:27<18:41, 13.68s/it]2026-06-08 07:05:12,929 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:05:16,889 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:05:20,726 - pyscenedetect - INFO - Detecting scenes...
+
 16%|█▋        | 16/97 [03:39<17:38, 13.07s/it]2026-06-08 07:05:24,595 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:05:28,685 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:05:33,111 - pyscenedetect - INFO - Detecting scenes...
+
 18%|█▊        | 17/97 [03:52<17:23, 13.05s/it]2026-06-08 07:05:37,591 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:05:42,189 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:05:53,214 - pyscenedetect - INFO - Detecting scenes...
+
 19%|█▊        | 18/97 [04:12<20:05, 15.26s/it]2026-06-08 07:05:57,996 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:06:02,687 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:06:12,912 - pyscenedetect - INFO - Detecting scenes...
+
 20%|█▉        | 19/97 [04:32<21:34, 16.60s/it]2026-06-08 07:06:17,714 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:06:22,331 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:06:33,111 - pyscenedetect - INFO - Detecting scenes...
+
 21%|██        | 20/97 [04:52<22:39, 17.66s/it]2026-06-08 07:06:37,842 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:06:42,390 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:06:47,040 - pyscenedetect - INFO - Detecting scenes...
+
 22%|██▏       | 21/97 [05:06<20:55, 16.52s/it]2026-06-08 07:06:51,692 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:06:56,081 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:07:00,618 - pyscenedetect - INFO - Detecting scenes...
+
 23%|██▎       | 22/97 [05:20<19:40, 15.74s/it]2026-06-08 07:07:05,623 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:07:10,043 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:07:14,426 - pyscenedetect - INFO - Detecting scenes...
+
 24%|██▎       | 23/97 [05:33<18:32, 15.03s/it]2026-06-08 07:07:18,992 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:07:23,452 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:07:28,024 - pyscenedetect - INFO - Detecting scenes...
+
 25%|██▍       | 24/97 [05:47<17:45, 14.60s/it]2026-06-08 07:07:32,600 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:07:42,109 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:07:47,182 - pyscenedetect - INFO - Detecting scenes...
+
 26%|██▌       | 25/97 [06:12<21:05, 17.57s/it]2026-06-08 07:07:57,146 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:08:01,655 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:08:06,225 - pyscenedetect - INFO - Detecting scenes...
+
 27%|██▋       | 26/97 [06:25<19:30, 16.48s/it]2026-06-08 07:08:11,037 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:08:15,716 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:08:20,289 - pyscenedetect - INFO - Detecting scenes...
+
 28%|██▊       | 27/97 [06:39<18:14, 15.64s/it]2026-06-08 07:08:24,714 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:08:30,120 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:08:34,548 - pyscenedetect - INFO - Detecting scenes...
+
 29%|██▉       | 28/97 [06:54<17:33, 15.26s/it]2026-06-08 07:08:39,093 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:08:43,660 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:08:48,759 - pyscenedetect - INFO - Detecting scenes...
+
 30%|██▉       | 29/97 [07:08<17:10, 15.15s/it]2026-06-08 07:08:53,990 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:08:58,587 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:09:03,242 - pyscenedetect - INFO - Detecting scenes...
+
 31%|███       | 30/97 [07:28<18:18, 16.39s/it]2026-06-08 07:09:13,266 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:09:17,924 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:09:22,447 - pyscenedetect - INFO - Detecting scenes...
+
 32%|███▏      | 31/97 [07:47<18:55, 17.20s/it]2026-06-08 07:09:32,369 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:09:40,521 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:09:50,065 - pyscenedetect - INFO - Detecting scenes...
+
 33%|███▎      | 32/97 [08:09<20:17, 18.73s/it]2026-06-08 07:09:54,665 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:09:59,869 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:10:04,956 - pyscenedetect - INFO - Detecting scenes...
+
 34%|███▍      | 33/97 [08:24<18:43, 17.55s/it]2026-06-08 07:10:09,453 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:10:13,953 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:10:18,718 - pyscenedetect - INFO - Detecting scenes...
+
 35%|███▌      | 34/97 [08:43<19:02, 18.13s/it]2026-06-08 07:10:28,937 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:10:33,625 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:10:38,363 - pyscenedetect - INFO - Detecting scenes...
+
 36%|███▌      | 35/97 [08:57<17:27, 16.90s/it]2026-06-08 07:10:42,960 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:10:47,730 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:10:52,357 - pyscenedetect - INFO - Detecting scenes...
+
 37%|███▋      | 36/97 [09:16<17:49, 17.53s/it]2026-06-08 07:11:02,011 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:11:12,148 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:11:16,943 - pyscenedetect - INFO - Detecting scenes...
+
 38%|███▊      | 37/97 [09:36<18:09, 18.17s/it]2026-06-08 07:11:21,626 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:11:26,391 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:11:31,007 - pyscenedetect - INFO - Detecting scenes...
+
 39%|███▉      | 38/97 [09:50<16:43, 17.01s/it]2026-06-08 07:11:35,923 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:11:40,503 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:11:44,879 - pyscenedetect - INFO - Detecting scenes...
+
 40%|████      | 39/97 [10:03<15:14, 15.77s/it]2026-06-08 07:11:48,820 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:11:52,684 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:11:56,613 - pyscenedetect - INFO - Detecting scenes...
+
 41%|████      | 40/97 [10:15<13:49, 14.54s/it]2026-06-08 07:12:00,495 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:12:04,414 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:12:13,093 - pyscenedetect - INFO - Detecting scenes...
+
 42%|████▏     | 41/97 [10:31<14:07, 15.13s/it]2026-06-08 07:12:16,976 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:12:20,813 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:12:24,811 - pyscenedetect - INFO - Detecting scenes...
+
 43%|████▎     | 42/97 [10:43<12:55, 14.09s/it]2026-06-08 07:12:28,662 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:12:32,580 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:12:36,500 - pyscenedetect - INFO - Detecting scenes...
+
 44%|████▍     | 43/97 [10:55<12:02, 13.38s/it]2026-06-08 07:12:40,367 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:12:44,233 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:12:48,167 - pyscenedetect - INFO - Detecting scenes...
+
 45%|████▌     | 44/97 [11:07<11:22, 12.87s/it]2026-06-08 07:12:52,066 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:12:56,001 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:12:59,885 - pyscenedetect - INFO - Detecting scenes...
+
 46%|████▋     | 45/97 [11:20<11:15, 13.00s/it]2026-06-08 07:13:05,362 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:13:10,762 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:13:14,605 - pyscenedetect - INFO - Detecting scenes...
+
 47%|████▋     | 46/97 [11:33<11:04, 13.04s/it]2026-06-08 07:13:18,482 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:13:22,326 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:13:26,253 - pyscenedetect - INFO - Detecting scenes...
+
 48%|████▊     | 47/97 [11:49<11:30, 13.80s/it]2026-06-08 07:13:34,089 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:13:38,066 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:13:41,979 - pyscenedetect - INFO - Detecting scenes...
+
 49%|████▉     | 48/97 [12:00<10:46, 13.20s/it]2026-06-08 07:13:45,847 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:13:49,817 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:13:53,769 - pyscenedetect - INFO - Detecting scenes...
+
 51%|█████     | 49/97 [12:12<10:14, 12.80s/it]2026-06-08 07:13:57,726 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:14:01,683 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:14:05,730 - pyscenedetect - INFO - Detecting scenes...
+
 52%|█████▏    | 50/97 [12:24<09:48, 12.52s/it]2026-06-08 07:14:09,610 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:14:13,576 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:14:17,642 - pyscenedetect - INFO - Detecting scenes...
+
 53%|█████▎    | 51/97 [12:36<09:28, 12.35s/it]2026-06-08 07:14:21,550 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:14:25,468 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:14:29,447 - pyscenedetect - INFO - Detecting scenes...
+
 54%|█████▎    | 52/97 [12:48<09:08, 12.18s/it]2026-06-08 07:14:33,334 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:14:37,229 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:14:41,156 - pyscenedetect - INFO - Detecting scenes...
+
 55%|█████▍    | 53/97 [12:59<08:48, 12.02s/it]2026-06-08 07:14:44,985 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:14:48,895 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:14:52,868 - pyscenedetect - INFO - Detecting scenes...
+
 56%|█████▌    | 54/97 [13:11<08:33, 11.95s/it]2026-06-08 07:14:56,754 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:15:00,569 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:15:04,521 - pyscenedetect - INFO - Detecting scenes...
+
 57%|█████▋    | 55/97 [13:23<08:18, 11.87s/it]2026-06-08 07:15:08,457 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:15:12,269 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:15:16,138 - pyscenedetect - INFO - Detecting scenes...
+
 58%|█████▊    | 56/97 [13:34<08:02, 11.77s/it]2026-06-08 07:15:19,976 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:15:27,867 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:15:31,771 - pyscenedetect - INFO - Detecting scenes...
+
 59%|█████▉    | 57/97 [13:50<08:37, 12.94s/it]2026-06-08 07:15:35,672 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:15:39,546 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:15:43,463 - pyscenedetect - INFO - Detecting scenes...
+
 60%|█████▉    | 58/97 [14:02<08:08, 12.54s/it]2026-06-08 07:15:47,255 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:15:51,136 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:15:55,036 - pyscenedetect - INFO - Detecting scenes...
+
 61%|██████    | 59/97 [14:13<07:46, 12.27s/it]2026-06-08 07:15:58,895 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:16:04,623 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:16:08,976 - pyscenedetect - INFO - Detecting scenes...
+
 62%|██████▏   | 60/97 [14:28<08:00, 12.98s/it]2026-06-08 07:16:13,555 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:16:18,177 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:16:22,960 - pyscenedetect - INFO - Detecting scenes...
+
 63%|██████▎   | 61/97 [14:42<08:00, 13.34s/it]2026-06-08 07:16:27,715 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:16:37,017 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:16:47,050 - pyscenedetect - INFO - Detecting scenes...
+
 64%|██████▍   | 62/97 [15:06<09:39, 16.57s/it]2026-06-08 07:16:51,827 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:16:59,335 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:17:04,074 - pyscenedetect - INFO - Detecting scenes...
+
 65%|██████▍   | 63/97 [15:23<09:25, 16.63s/it]2026-06-08 07:17:08,587 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:17:20,134 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:17:25,337 - pyscenedetect - INFO - Detecting scenes...
+
 66%|██████▌   | 64/97 [15:45<09:56, 18.09s/it]2026-06-08 07:17:30,075 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:17:34,820 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:17:39,565 - pyscenedetect - INFO - Detecting scenes...
+
 67%|██████▋   | 65/97 [15:59<09:00, 16.89s/it]2026-06-08 07:17:44,185 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:17:49,729 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:17:56,957 - pyscenedetect - INFO - Detecting scenes...
+
 68%|██████▊   | 66/97 [16:16<08:46, 17.00s/it]2026-06-08 07:18:01,428 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:18:06,481 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:18:11,231 - pyscenedetect - INFO - Detecting scenes...
+
 69%|██████▉   | 67/97 [16:31<08:10, 16.34s/it]2026-06-08 07:18:16,226 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:18:21,011 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:18:30,809 - pyscenedetect - INFO - Detecting scenes...
+
 70%|███████   | 68/97 [16:50<08:18, 17.18s/it]2026-06-08 07:18:35,353 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:18:39,807 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:18:44,356 - pyscenedetect - INFO - Detecting scenes...
+
 71%|███████   | 69/97 [17:08<08:13, 17.63s/it]2026-06-08 07:18:54,107 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:18:59,172 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:19:04,202 - pyscenedetect - INFO - Detecting scenes...
+
 72%|███████▏  | 70/97 [17:23<07:34, 16.82s/it]2026-06-08 07:19:08,980 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:19:13,495 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:19:18,145 - pyscenedetect - INFO - Detecting scenes...
+
 73%|███████▎  | 71/97 [17:37<06:53, 15.89s/it]2026-06-08 07:19:22,689 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:19:27,299 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:19:31,948 - pyscenedetect - INFO - Detecting scenes...
+
 74%|███████▍  | 72/97 [17:51<06:22, 15.29s/it]2026-06-08 07:19:36,574 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:19:42,151 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:19:46,764 - pyscenedetect - INFO - Detecting scenes...
+
 75%|███████▌  | 73/97 [18:06<06:03, 15.14s/it]2026-06-08 07:19:51,371 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:19:56,061 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:20:01,284 - pyscenedetect - INFO - Detecting scenes...
+
 76%|███████▋  | 74/97 [18:20<05:44, 14.98s/it]2026-06-08 07:20:05,980 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:20:11,125 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:20:15,865 - pyscenedetect - INFO - Detecting scenes...
+
 77%|███████▋  | 75/97 [18:36<05:31, 15.07s/it]2026-06-08 07:20:21,271 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:20:26,123 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:20:30,914 - pyscenedetect - INFO - Detecting scenes...
+
 78%|███████▊  | 76/97 [18:50<05:10, 14.80s/it]2026-06-08 07:20:35,428 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:20:40,095 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:20:44,927 - pyscenedetect - INFO - Detecting scenes...
+
 79%|███████▉  | 77/97 [19:09<05:22, 16.14s/it]2026-06-08 07:20:54,747 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:21:02,276 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:21:07,032 - pyscenedetect - INFO - Detecting scenes...
+
 80%|████████  | 78/97 [19:27<05:17, 16.70s/it]2026-06-08 07:21:12,706 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:21:17,333 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:21:26,957 - pyscenedetect - INFO - Detecting scenes...
+
 81%|████████▏ | 79/97 [19:46<05:13, 17.40s/it]2026-06-08 07:21:31,754 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:21:36,306 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:21:40,902 - pyscenedetect - INFO - Detecting scenes...
+
 82%|████████▏ | 80/97 [20:00<04:37, 16.35s/it]2026-06-08 07:21:45,657 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:21:50,247 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:21:55,210 - pyscenedetect - INFO - Detecting scenes...
+
 84%|████████▎ | 81/97 [20:15<04:13, 15.83s/it]2026-06-08 07:22:00,253 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:22:04,904 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:22:09,534 - pyscenedetect - INFO - Detecting scenes...
+
 85%|████████▍ | 82/97 [20:29<03:48, 15.25s/it]2026-06-08 07:22:14,166 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:22:18,038 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:22:25,884 - pyscenedetect - INFO - Detecting scenes...
+
 86%|████████▌ | 83/97 [20:44<03:34, 15.36s/it]2026-06-08 07:22:29,758 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:22:33,701 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:22:37,583 - pyscenedetect - INFO - Detecting scenes...
+
 87%|████████▋ | 84/97 [20:56<03:05, 14.27s/it]2026-06-08 07:22:41,505 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:22:45,512 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:22:49,659 - pyscenedetect - INFO - Detecting scenes...
+
 88%|████████▊ | 85/97 [21:09<02:46, 13.85s/it]2026-06-08 07:22:54,359 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:22:59,155 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:23:03,071 - pyscenedetect - INFO - Detecting scenes...
+
 89%|████████▊ | 86/97 [21:21<02:28, 13.49s/it]2026-06-08 07:23:07,015 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:23:10,968 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:23:19,103 - pyscenedetect - INFO - Detecting scenes...
+
 90%|████████▉ | 87/97 [21:37<02:22, 14.22s/it]2026-06-08 07:23:22,927 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:23:26,771 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:23:30,613 - pyscenedetect - INFO - Detecting scenes...
+
 91%|█████████ | 88/97 [21:49<02:00, 13.41s/it]2026-06-08 07:23:34,469 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:23:38,409 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:23:42,337 - pyscenedetect - INFO - Detecting scenes...
+
 92%|█████████▏| 89/97 [22:01<01:43, 12.91s/it]2026-06-08 07:23:46,203 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:23:51,205 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:23:55,923 - pyscenedetect - INFO - Detecting scenes...
+
 93%|█████████▎| 90/97 [22:14<01:31, 13.08s/it]2026-06-08 07:23:59,683 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:24:03,524 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:24:07,510 - pyscenedetect - INFO - Detecting scenes...
+
 94%|█████████▍| 91/97 [22:26<01:15, 12.65s/it]2026-06-08 07:24:11,342 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:24:15,151 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:24:19,062 - pyscenedetect - INFO - Detecting scenes...
+
 95%|█████████▍| 92/97 [22:37<01:01, 12.33s/it]2026-06-08 07:24:22,899 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:24:26,789 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:24:30,542 - pyscenedetect - INFO - Detecting scenes...
+
 96%|█████████▌| 93/97 [22:49<00:48, 12.08s/it]2026-06-08 07:24:34,394 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:24:38,219 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:24:41,996 - pyscenedetect - INFO - Detecting scenes...
+
 97%|█████████▋| 94/97 [23:00<00:35, 11.90s/it]2026-06-08 07:24:45,877 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:24:49,796 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:24:58,041 - pyscenedetect - INFO - Detecting scenes...
+
 98%|█████████▊| 95/97 [23:16<00:26, 13.11s/it]2026-06-08 07:25:01,826 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:25:05,750 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:25:13,691 - pyscenedetect - INFO - Detecting scenes...
+
 99%|█████████▉| 96/97 [23:32<00:13, 13.89s/it]2026-06-08 07:25:17,541 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:25:21,451 - pyscenedetect - INFO - Detecting scenes...
+2026-06-08 07:25:25,417 - pyscenedetect - INFO - Detecting scenes...
+
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Multi-View_Consistency/results_2026-06-08-07:01:38_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/Thermotics.log

@@ -0,0 +1,21 @@
+[2026-06-08 06:52:05 UTC] COMMAND /root/ckpts/micromamba/envs/vbench2/bin/python -u evaluate.py --videos_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Thermotics --dimension Thermotics --output_path /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Thermotics --full_json_dir /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json --mode vbench_standard
+args: Namespace(output_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Thermotics', full_json_dir='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_selected_info.json', videos_path='/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/videos/Thermotics', dimension=['Thermotics'], load_ckpt_from_local=None, read_frame=None, mode='vbench_standard', prompt='None', prompt_file=None, category=None, imaging_quality_preprocessing_mode='longer')
+start evaluation
+File exists: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+Evaluation meta data saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Thermotics/results_2026-06-08-06:52:08_full_info.json
+/root/ckpts/micromamba/envs/vbench2/lib/python3.10/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/vbench2/third_party/LLaVA_NeXT/llava/model/llava_arch.py:215: SyntaxWarning: "is not" with a literal. Did you mean "!="?
+  if slower_img_feat is not 0:
+cur_full_info_path: /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Thermotics/results_2026-06-08-06:52:08_full_info.json
+Loaded LLaVA model: /root/ckpts/vbench2/lmms-lab/LLaVA-Video-7B-Qwen2
+You are using a model of type llava to instantiate a model of type llava_qwen. This is not supported for all configurations of models and can yield errors.
+Sliding Window Attention is enabled but not implemented for `sdpa`; unexpected results may be encountered.
+Loading vision tower: google/siglip-so400m-patch14-384
+2026-06-08 06:52:13,507 - accelerate.utils.modeling - INFO - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk).
+
+Model Class: LlavaQwenForCausalLM
+
  0%|          | 0/50 [00:00<?, ?it/s]The attention mask is not set and cannot be inferred from input because pad token is same as eos token. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
+
  2%|▏         | 1/50 [00:16<13:19, 16.32s/it]
  4%|▍         | 2/50 [00:32<12:49, 16.03s/it]
  6%|▌         | 3/50 [00:47<12:25, 15.87s/it]
  8%|▊         | 4/50 [01:03<12:05, 15.77s/it]
 10%|█         | 5/50 [01:19<11:48, 15.74s/it]
 12%|█▏        | 6/50 [01:34<11:30, 15.70s/it]
 14%|█▍        | 7/50 [01:50<11:15, 15.70s/it]
 16%|█▌        | 8/50 [02:06<10:58, 15.68s/it]
 18%|█▊        | 9/50 [02:21<10:43, 15.68s/it]
 20%|██        | 10/50 [02:37<10:33, 15.83s/it]
 22%|██▏       | 11/50 [02:47<09:07, 14.04s/it]
 24%|██▍       | 12/50 [03:04<09:23, 14.83s/it]
 26%|██▌       | 13/50 [03:21<09:34, 15.52s/it]
 28%|██▊       | 14/50 [03:37<09:26, 15.74s/it]
 30%|███       | 15/50 [03:54<09:16, 15.91s/it]
 32%|███▏      | 16/50 [04:10<09:07, 16.11s/it]
 34%|███▍      | 17/50 [04:27<08:54, 16.21s/it]
 36%|███▌      | 18/50 [04:43<08:40, 16.25s/it]
 38%|███▊      | 19/50 [05:00<08:26, 16.34s/it]
 40%|████      | 20/50 [05:10<07:12, 14.43s/it]
 42%|████▏     | 21/50 [05:26<07:13, 14.96s/it]
 44%|████▍     | 22/50 [05:42<07:10, 15.37s/it]
 46%|████▌     | 23/50 [05:52<06:09, 13.69s/it]
 48%|████▊     | 24/50 [06:08<06:16, 14.49s/it]
 50%|█████     | 25/50 [06:25<06:18, 15.15s/it]
 52%|█████▏    | 26/50 [06:42<06:13, 15.57s/it]
 54%|█████▍    | 27/50 [06:58<06:04, 15.83s/it]
 56%|█████▌    | 28/50 [07:15<05:53, 16.08s/it]
 58%|█████▊    | 29/50 [07:31<05:38, 16.11s/it]
 60%|██████    | 30/50 [07:48<05:26, 16.32s/it]
 62%|██████▏   | 31/50 [08:04<05:13, 16.48s/it]
 64%|██████▍   | 32/50 [08:21<04:55, 16.41s/it]
 66%|██████▌   | 33/50 [08:37<04:40, 16.50s/it]
 68%|██████▊   | 34/50 [08:53<04:21, 16.35s/it]
 70%|███████   | 35/50 [09:09<04:03, 16.24s/it]
 72%|███████▏  | 36/50 [09:25<03:45, 16.11s/it]
 74%|███████▍  | 37/50 [09:41<03:28, 16.02s/it]
 76%|███████▌  | 38/50 [10:03<03:34, 17.89s/it]
 78%|███████▊  | 39/50 [10:19<03:09, 17.23s/it]
 80%|████████  | 40/50 [10:35<02:47, 16.79s/it]
 82%|████████▏ | 41/50 [10:50<02:28, 16.48s/it]
 84%|████████▍ | 42/50 [11:06<02:10, 16.28s/it]
 86%|████████▌ | 43/50 [11:16<01:39, 14.18s/it]
 88%|████████▊ | 44/50 [11:25<01:16, 12.75s/it]
 90%|█████████ | 45/50 [11:41<01:08, 13.67s/it]
 92%|█████████▏| 46/50 [11:57<00:57, 14.30s/it]
 94%|█████████▍| 47/50 [12:12<00:44, 14.78s/it]
 96%|█████████▌| 48/50 [12:28<00:30, 15.10s/it]
 98%|█████████▊| 49/50 [12:44<00:15, 15.33s/it]
+Evaluation results saved to /nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Thermotics/results_2026-06-08-06:52:08_eval_results.json
+done

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_parallel/parallel_status.json

@@ -0,0 +1,27 @@
+{
+  "generated_at_utc": "2026-06-08T07:59:03.299108+00:00",
+  "run_id": "yw008-1-sage-videoscore2-token-credit-video-adv",
+  "step": "100",
+  "output_root": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results",
+  "completed": {
+    "Camera_Motion": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Camera_Motion/results_2026-06-08-06:08:05_eval_results.json",
+    "Complex_Plot": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Complex_Plot/results_2026-06-08-06:08:05_eval_results.json",
+    "Human_Anatomy": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Anatomy/results_2026-06-08-06:08:05_eval_results.json",
+    "Human_Identity": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Identity/results_2026-06-08-06:08:05_eval_results.json",
+    "Human_Clothes": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Clothes/results_2026-06-08-06:08:05_eval_results.json",
+    "Diversity": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Diversity/results_2026-06-08-06:08:05_eval_results.json",
+    "Composition": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Composition/results_2026-06-08-06:08:05_eval_results.json",
+    "Dynamic_Spatial_Relationship": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Dynamic_Spatial_Relationship/results_2026-06-08-06:08:05_eval_results.json",
+    "Dynamic_Attribute": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Dynamic_Attribute/results_2026-06-08-06:17:06_eval_results.json",
+    "Motion_Order_Understanding": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Motion_Order_Understanding/results_2026-06-08-06:26:07_eval_results.json",
+    "Human_Interaction": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Interaction/results_2026-06-08-06:33:06_eval_results.json",
+    "Complex_Landscape": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Complex_Landscape/results_2026-06-08-06:37:08_eval_results.json",
+    "Motion_Rationality": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Motion_Rationality/results_2026-06-08-06:38:07_eval_results.json",
+    "Instance_Preservation": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation/results_2026-06-08-07:58:10_sharded_merge_eval_results.json",
+    "Mechanics": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Mechanics/results_2026-06-08-06:46:38_eval_results.json",
+    "Thermotics": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Thermotics/results_2026-06-08-06:52:08_eval_results.json",
+    "Material": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Material/results_2026-06-08-07:01:08_eval_results.json",
+    "Multi-View_Consistency": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Multi-View_Consistency/results_2026-06-08-07:01:38_eval_results.json"
+  },
+  "failed": {}
+}

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/logs/vbench2_requirement_py312.txt

@@ -0,0 +1,197 @@
+accelerate==1.2.1
+addict==2.4.0
+aiohappyeyeballs==2.4.4
+aiohttp==3.11.11
+aiosignal==1.3.2
+albucore==0.0.19
+albumentations==1.4.19
+annotated-types==0.7.0
+antlr4-python3-runtime==4.9.3
+anyio==4.8.0
+args==0.1.0
+async-timeout==5.0.1
+attrs==24.3.0
+av==14.0.1
+beautifulsoup4==4.12.3
+# skipped for eval-only py312 setup: bitsandbytes==0.42.0
+blis==1.2.0
+boto3==1.26.90
+botocore
+braceexpand==0.1.7
+catalogue==2.0.10
+certifi==2024.12.14
+charset-normalizer==3.4.1
+click==8.1.8
+clint==0.5.1
+cloudpickle==3.1.1
+confection==0.1.5
+contourpy==1.3.1
+cycler==0.12.1
+cymem==2.0.11
+datasets==3.2.0
+decorator==5.1.1
+decord==0.6.0
+deep-sort-realtime==1.3.2
+# skipped for eval-only py312 setup: deepspeed==0.14.4
+defusedxml==0.7.1
+diffusers==0.32.1
+dill==0.3.8
+distro==1.9.0
+docker-pycreds==0.4.0
+easydict==1.13
+einops==0.8.0
+einops-exts==0.0.4
+exceptiongroup==1.2.2
+fairscale==0.4.13
+fastapi==0.115.6
+filelock==3.13.1
+filterpy==1.4.5
+fire==0.4.0
+flowlib==0.6.6.8
+fonttools==4.55.3
+frozenlist==1.5.0
+fsspec==2024.2.0
+ftfy==6.3.1
+fvcore==0.1.5.post20221221
+gdown==5.2.0
+gitdb==4.0.12
+GitPython==3.1.44
+h11==0.14.0
+h5py==3.12.1
+hjson==3.1.0
+hpsv2==1.2.0
+httpcore==1.0.7
+httpx==0.28.1
+huggingface-hub==0.30.0
+idna==3.10
+imageio==2.37.0
+imageio-ffmpeg==0.6.0
+importlib_metadata==8.5.0
+iniconfig==2.0.0
+iopath==0.1.10
+ipdb==0.13.13
+ipython==8.32.0
+Jinja2==3.1.3
+jiter==0.8.2
+jmespath==1.0.1
+joblib==1.4.2
+kiwisolver==1.4.8
+kornia==0.5.6
+kornia_rs==0.1.8
+langcodes==3.5.0
+lazy_loader==0.4
+-e git+https://github.com/LLaVA-VL/LLaVA-NeXT@79ef45a6d8b89b92d7a8525f077c3a3a9894a87d#egg=llava
+logger==1.4
+lvis==0.5.3
+MarkupSafe==2.1.5
+matplotlib==3.10.0
+mmdet==3.0.0
+mmengine==0.10.5
+mpmath==1.3.0
+msgpack==1.1.0
+multidict==6.1.0
+multiprocess==0.70.16
+murmurhash==1.0.12
+networkx==3.2.1
+nltk==3.9.1
+numpy==1.26.3
+omegaconf==2.3.0
+onnx==1.17.0
+onnxruntime>=1.17.0,<2.0.0  # patched for Python 3.12
+open_clip_torch==2.30.0
+openai==1.59.5
+opencv-python==4.10.0.84
+opencv-python-headless==4.10.0.84
+openmim==0.3.9
+packaging==24.2
+pandas==2.2.3
+peft==0.14.0
+pillow==10.2.0
+platformdirs==4.3.6
+pluggy==1.5.0
+portalocker==3.1.1
+preshed==3.0.9
+prettytable==3.14.0
+propcache==0.2.1
+protobuf==3.20.3
+py-cpuinfo==9.0.0
+pyarrow==18.1.0
+pycocoevalcap==1.2
+pycocotools==2.0.8
+pydantic==2.10.4
+pydantic_core==2.27.2
+pyecharts==2.0.8
+pynvml==12.0.0
+pyparsing==3.2.1
+PySocks==1.7.1
+pytest==7.2.0
+pytest-split==0.8.0
+python-dateutil==2.9.0.post0
+pytz==2024.2
+PyYAML==6.0.2
+qudida==0.0.4
+regex==2024.11.6
+requests==2.32.3
+rich==13.9.4
+safetensors==0.5.2
+scenedetect==0.6.5.2
+scikit-image==0.25.0
+scikit-learn==1.6.0
+scipy==1.15.0
+seaborn==0.13.2
+selenium==4.29.0
+sentencepiece==0.2.0
+sentry-sdk==2.19.2
+setproctitle==1.3.4
+shapely==2.0.7
+simplejson==3.20.1
+six==1.17.0
+smmap==5.0.2
+snapshot-selenium==0.0.2
+sniffio==1.3.1
+soupsieve==2.6
+spacy==3.8.4
+srsly==2.5.1
+stack-data==0.6.3
+starlette==0.41.3
+sympy==1.13.1
+tabulate==0.9.0
+termcolor==2.5.0
+terminaltables==3.1.10
+thinc
+thop==0.1.1.post2209072238
+threadpoolctl==3.5.0
+tifffile==2025.1.10
+tiktoken==0.8.0
+timesformer-pytorch==0.4.1
+timm==0.4.12
+tokenizers==0.21.1
+tomli==2.2.1
+tqdm==4.67.1
+traitlets==5.14.3
+transformers==4.51.0
+# skipped for eval-only py312 setup: triton==3.1.0
+typing_extensions==4.12.2
+tzdata==2024.2
+ultralytics==8.3.68
+unicorn==2.1.1
+urllib3==1.26.15
+wandb==0.19.2
+wasabi==1.1.3
+wcwidth==0.2.13
+webdataset==0.2.100
+websocket==0.2.1
+websocket-client==1.8.0
+wordcloud==1.9.4
+xxhash==3.5.0
+yacs==0.1.8
+yapf==0.43.0
+yarl==1.18.3
+yolov5
+zipp==3.21.0
+mediapipe
+retina-face
+retinaface-pytorch
+supervision
+tf-keras
+qwen-vl-utils

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_instance_shards/selected_info_instance_shard_00.json

@@ -0,0 +1,79 @@
+[
+  {
+    "prompt_en": "An orange dog is running.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.orange dog"
+  },
+  {
+    "prompt_en": "Two orange dogs are running.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "2.orange dog"
+  },
+  {
+    "prompt_en": "A brown dog is on the left of an apple, then the dog moves to the right of the apple.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.brown dog"
+  },
+  {
+    "prompt_en": "An orange cat is running.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.orange cat"
+  },
+  {
+    "prompt_en": "An orange cat and a white cat are running.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.orange cat and 1.white cat"
+  },
+  {
+    "prompt_en": "An orange cat and a white cat are fighting.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.orange cat and 1.white cat"
+  },
+  {
+    "prompt_en": "An orange cat is on the left of an apple, then the cat moves to the right of the apple.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.orange cat"
+  },
+  {
+    "prompt_en": "A beige tiger is running.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.beige tiger"
+  },
+  {
+    "prompt_en": "Two sandy kangaroos are jumping.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "2.sandy kangaroo"
+  },
+  {
+    "prompt_en": "Two sandy kangaroos are fighting.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "2.sandy kangaroo"
+  },
+  {
+    "prompt_en": "A giraffe is walking.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.giraffe"
+  }
+]

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_instance_shards/selected_info_instance_shard_01.json

@@ -0,0 +1,79 @@
+[
+  {
+    "prompt_en": "Two giraffes are walking.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "2.giraffe"
+  },
+  {
+    "prompt_en": "A brown bear is running.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.brown bear"
+  },
+  {
+    "prompt_en": "Two brown bears are running.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "2.brown bear"
+  },
+  {
+    "prompt_en": "A orange fox is running.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.orange fox"
+  },
+  {
+    "prompt_en": "A white sheep is walking.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.white sheep"
+  },
+  {
+    "prompt_en": "Two white sheep are walking.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "2.white sheep"
+  },
+  {
+    "prompt_en": "Two white sheep are eating grass.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "2.white sheep"
+  },
+  {
+    "prompt_en": "Two blue birds flies through the sky, coming close to each other and forming a complex formation.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "2.blue bird"
+  },
+  {
+    "prompt_en": "A ball rolls off the table.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.ball"
+  },
+  {
+    "prompt_en": "A blue ball rolls off the table.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.blue ball"
+  },
+  {
+    "prompt_en": "A red ball bouncing down stairs",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.red ball"
+  }
+]

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_instance_shards/selected_info_instance_shard_02.json

@@ -0,0 +1,79 @@
+[
+  {
+    "prompt_en": "A red ball and a blue ball bouncing down stairs",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.red ball and 1.blue ball"
+  },
+  {
+    "prompt_en": "A red ball falls from a height, hits the ground, and bounces back up.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.red ball"
+  },
+  {
+    "prompt_en": "A red ball and a blue ball fall from a height, hits the ground, and bounces back up.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.red ball and 1.blue ball"
+  },
+  {
+    "prompt_en": "Aerial view, aerial view. Aerial view, aerial view. One blue balls and one red balls are on the wooden table and collide horizontally, bird's-eye view.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.blue ball and 1.red ball"
+  },
+  {
+    "prompt_en": "Two basketballs spin rapidly on the floor.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "2.basketball"
+  },
+  {
+    "prompt_en": "Two footballs spin rapidly on the floor.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "2.football"
+  },
+  {
+    "prompt_en": "A red car is moving along the road.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.red car"
+  },
+  {
+    "prompt_en": "A car speeds along a winding mountain road, with trees rapidly passing by.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.car"
+  },
+  {
+    "prompt_en": "A man and a woman is doing yoga.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.man and 1.woman"
+  },
+  {
+    "prompt_en": "Two people are doing yoga.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "2.person"
+  },
+  {
+    "prompt_en": "A person is running.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.person"
+  }
+]

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_instance_shards/selected_info_instance_shard_03.json

@@ -0,0 +1,86 @@
+[
+  {
+    "prompt_en": "Two people are running.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "2.person"
+  },
+  {
+    "prompt_en": "A man is playing basketball.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.man"
+  },
+  {
+    "prompt_en": "Two men are playing basketball.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "2.man"
+  },
+  {
+    "prompt_en": "Two red balloons are floating in the air.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "2.red balloon"
+  },
+  {
+    "prompt_en": "A top is spinning.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.top"
+  },
+  {
+    "prompt_en": "A red top and a blue top are spinning.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.red top and 1.blue top"
+  },
+  {
+    "prompt_en": "Two red tops are spinning.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "2.red top"
+  },
+  {
+    "prompt_en": "A drone is floating in the air.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.drone"
+  },
+  {
+    "prompt_en": "A black drone is floating in the air.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.black drone"
+  },
+  {
+    "prompt_en": "Two black drones are floating in the air.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "2.black drone"
+  },
+  {
+    "prompt_en": "A black drone and a blue drone are floating in the air.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.black drone and 1.blue drone"
+  },
+  {
+    "prompt_en": "A droplet of water is falling.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.droplet of water"
+  }
+]

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_instance_shards/selected_info_instance_shard_04.json

@@ -0,0 +1,86 @@
+[
+  {
+    "prompt_en": "A panda changes from small to big.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.panda"
+  },
+  {
+    "prompt_en": "A panda changes from big to small.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.panda"
+  },
+  {
+    "prompt_en": "An orange dog changes from small to big.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.orange dog"
+  },
+  {
+    "prompt_en": "An orange dog changes from big to small.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.orange dog"
+  },
+  {
+    "prompt_en": "An orange cat changes from big to small.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.orange cat"
+  },
+  {
+    "prompt_en": "A fish changes from small to big.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.fish"
+  },
+  {
+    "prompt_en": "A fish changes from big to small.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.fish"
+  },
+  {
+    "prompt_en": "A beige lion changes from big to small.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.beige lion"
+  },
+  {
+    "prompt_en": "A beige tiger changes from small to big.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.beige tiger"
+  },
+  {
+    "prompt_en": "An orange cat is in front of a tree, then the cat moves to the left of the tree.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.orange cat"
+  },
+  {
+    "prompt_en": "A beige tiger is in front of a tree, then the tiger moves to the left of the tree.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.beige tiger"
+  },
+  {
+    "prompt_en": "A white sheep is in front of a tree, then the sheep moves to the left of the tree.",
+    "dimension": [
+      "Instance_Preservation"
+    ],
+    "auxiliary_info": "1.white sheep"
+  }
+]

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_score_details.csv

@@ -0,0 +1,25 @@
+model_name,dimension,score,score_percent,result_file
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Camera Motion,0.5555555555555556,55.55555555555556,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Camera_Motion/results_2026-06-08-06:08:05_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Complex Landscape,0.19111111111111098,19.111111111111097,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Complex_Landscape/results_2026-06-08-06:37:08_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Complex Plot,0.1318518518518517,13.18518518518517,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Complex_Plot/results_2026-06-08-06:08:05_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Composition,0.4329140461215935,43.291404612159354,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Composition/results_2026-06-08-06:08:05_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Diversity,0.368597237765789,36.8597237765789,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Diversity/results_2026-06-08-06:08:05_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Dynamic Attribute,0.2564102564102564,25.64102564102564,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Dynamic_Attribute/results_2026-06-08-06:17:06_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Dynamic Spatial Relationship,0.34299516908212563,34.29951690821256,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Dynamic_Spatial_Relationship/results_2026-06-08-06:08:05_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Human Anatomy,0.8927218399424536,89.27218399424535,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Anatomy/results_2026-06-08-06:08:05_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Human Clothes,0.9420289855072463,94.20289855072464,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Clothes/results_2026-06-08-06:08:05_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Human Identity,0.8589602187029363,85.89602187029362,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Identity/results_2026-06-08-06:08:05_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Human Interaction,0.5333333333333333,53.333333333333336,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Interaction/results_2026-06-08-06:33:06_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Instance Preservation,0.935672514619883,93.56725146198829,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation/results_2026-06-08-07:58:10_sharded_merge_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Material,0.43564356435643564,43.56435643564357,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Material/results_2026-06-08-07:01:08_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Mechanics,0.5934959349593496,59.34959349593496,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Mechanics/results_2026-06-08-06:46:38_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Motion Order Understanding,0.18181818181818182,18.181818181818183,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Motion_Order_Understanding/results_2026-06-08-06:26:07_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Motion Rationality,0.3563218390804598,35.63218390804598,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Motion_Rationality/results_2026-06-08-06:38:07_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Multi-View Consistency,0.49019640768588135,49.01964076858813,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Multi-View_Consistency/results_2026-06-08-07:01:38_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Thermotics,0.4460431654676259,44.60431654676259,/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Thermotics/results_2026-06-08-06:52:08_eval_results.json
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Creativity,0.4007556419436913,40.07556419436913,
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Commonsense,0.6459971768501713,64.59971768501714,
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Controllability,0.31329649416605937,31.329649416605935,
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Human Fidelity,0.8979036813842121,89.7903681384212,
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Physics,0.49134476811732314,49.13447681173231,
+yw008-1-sage-videoscore2-token-credit-video-adv_step_100,Overall,0.5498595524922915,54.98595524922914,

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_score_summary.json

@@ -0,0 +1,82 @@
+{
+  "model_name": "yw008-1-sage-videoscore2-token-credit-video-adv_step_100",
+  "generated_at_utc": "2026-06-08T07:58:19.863413+00:00",
+  "results_root": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results",
+  "score_formula": "mean of five VBench2 groups; each group is the arithmetic mean of its listed dimensions",
+  "dimension_scores": {
+    "Camera Motion": 0.5555555555555556,
+    "Complex Landscape": 0.19111111111111098,
+    "Complex Plot": 0.1318518518518517,
+    "Composition": 0.4329140461215935,
+    "Diversity": 0.368597237765789,
+    "Dynamic Attribute": 0.2564102564102564,
+    "Dynamic Spatial Relationship": 0.34299516908212563,
+    "Human Anatomy": 0.8927218399424536,
+    "Human Clothes": 0.9420289855072463,
+    "Human Identity": 0.8589602187029363,
+    "Human Interaction": 0.5333333333333333,
+    "Instance Preservation": 0.935672514619883,
+    "Material": 0.43564356435643564,
+    "Mechanics": 0.5934959349593496,
+    "Motion Order Understanding": 0.18181818181818182,
+    "Motion Rationality": 0.3563218390804598,
+    "Multi-View Consistency": 0.49019640768588135,
+    "Thermotics": 0.4460431654676259
+  },
+  "group_scores": {
+    "Creativity": 0.4007556419436913,
+    "Commonsense": 0.6459971768501713,
+    "Controllability": 0.31329649416605937,
+    "Human Fidelity": 0.8979036813842121,
+    "Physics": 0.49134476811732314
+  },
+  "overall": 0.5498595524922915,
+  "dimension_scores_percent": {
+    "Camera Motion": 55.55555555555556,
+    "Complex Landscape": 19.111111111111097,
+    "Complex Plot": 13.18518518518517,
+    "Composition": 43.291404612159354,
+    "Diversity": 36.8597237765789,
+    "Dynamic Attribute": 25.64102564102564,
+    "Dynamic Spatial Relationship": 34.29951690821256,
+    "Human Anatomy": 89.27218399424535,
+    "Human Clothes": 94.20289855072464,
+    "Human Identity": 85.89602187029362,
+    "Human Interaction": 53.333333333333336,
+    "Instance Preservation": 93.56725146198829,
+    "Material": 43.56435643564357,
+    "Mechanics": 59.34959349593496,
+    "Motion Order Understanding": 18.181818181818183,
+    "Motion Rationality": 35.63218390804598,
+    "Multi-View Consistency": 49.01964076858813,
+    "Thermotics": 44.60431654676259
+  },
+  "group_scores_percent": {
+    "Creativity": 40.07556419436913,
+    "Commonsense": 64.59971768501714,
+    "Controllability": 31.329649416605935,
+    "Human Fidelity": 89.7903681384212,
+    "Physics": 49.13447681173231
+  },
+  "overall_percent": 54.98595524922914,
+  "result_files": {
+    "Camera Motion": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Camera_Motion/results_2026-06-08-06:08:05_eval_results.json",
+    "Complex Landscape": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Complex_Landscape/results_2026-06-08-06:37:08_eval_results.json",
+    "Complex Plot": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Complex_Plot/results_2026-06-08-06:08:05_eval_results.json",
+    "Composition": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Composition/results_2026-06-08-06:08:05_eval_results.json",
+    "Diversity": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Diversity/results_2026-06-08-06:08:05_eval_results.json",
+    "Dynamic Attribute": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Dynamic_Attribute/results_2026-06-08-06:17:06_eval_results.json",
+    "Dynamic Spatial Relationship": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Dynamic_Spatial_Relationship/results_2026-06-08-06:08:05_eval_results.json",
+    "Human Anatomy": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Anatomy/results_2026-06-08-06:08:05_eval_results.json",
+    "Human Clothes": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Clothes/results_2026-06-08-06:08:05_eval_results.json",
+    "Human Identity": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Identity/results_2026-06-08-06:08:05_eval_results.json",
+    "Human Interaction": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Human_Interaction/results_2026-06-08-06:33:06_eval_results.json",
+    "Instance Preservation": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Instance_Preservation/results_2026-06-08-07:58:10_sharded_merge_eval_results.json",
+    "Material": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Material/results_2026-06-08-07:01:08_eval_results.json",
+    "Mechanics": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Mechanics/results_2026-06-08-06:46:38_eval_results.json",
+    "Motion Order Understanding": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Motion_Order_Understanding/results_2026-06-08-06:26:07_eval_results.json",
+    "Motion Rationality": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Motion_Rationality/results_2026-06-08-06:38:07_eval_results.json",
+    "Multi-View Consistency": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Multi-View_Consistency/results_2026-06-08-07:01:38_eval_results.json",
+    "Thermotics": "/nfs/ywang29/Reward_finetuning/SAGE-GRPO/outputs/yw008-1-sage-videoscore2-token-credit-video-adv/vbench2_eval/step_100/vbench2_results/Thermotics/results_2026-06-08-06:52:08_eval_results.json"
+  }
+}

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_003_seed_789_A couple strolls hand in hand down a bustling city_config.json

@@ -0,0 +1,49 @@
+{
+  "timestamp": "2026-06-13T02:54:09.689242",
+  "task": "t2v",
+  "transformer_version": "480p_t2v",
+  "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_003_seed_789_A couple strolls hand in hand down a bustling city.mp4",
+  "arguments": {
+    "prompt": "A couple strolls hand in hand down a bustling city street, their fingers intertwined, embodying a sense of unity and affection. The man, wearing a navy coat and jeans, walks confidently beside the woman, who is dressed in a stylish red trench coat and black boots. The street is alive with the sounds of distant traffic and the chatter of passersby, while the couple remains in their own world, exchanging smiles and glances. The static camera captures their every step, as they navigate the urban landscape, framed by towering buildings and vibrant storefronts, under a sky painted with the warm hues of a setting sun.",
+    "valid_video_csv": [
+      "assets/video_gen_eval_top100.csv"
+    ],
+    "negative_prompt": "",
+    "resolution": "480p",
+    "model_path": "./ckpts",
+    "aspect_ratio": "16:9",
+    "fixed_size": "480x864",
+    "num_inference_steps": 40,
+    "video_length": 121,
+    "sr": false,
+    "save_pre_sr_video": false,
+    "rewrite": false,
+    "cfg_distilled": false,
+    "enable_step_distill": false,
+    "sparse_attn": false,
+    "offloading": false,
+    "group_offloading": null,
+    "overlap_group_offloading": false,
+    "dtype": "bf16",
+    "seed": 789,
+    "image_path": null,
+    "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_003_seed_789_A couple strolls hand in hand down a bustling city.mp4",
+    "original_noise_dir": null,
+    "use_sageattn": false,
+    "sage_blocks_range": "0-53",
+    "enable_torch_compile": false,
+    "enable_cache": false,
+    "cache_type": "deepcache",
+    "no_cache_block_id": "53",
+    "cache_start_step": 11,
+    "cache_end_step": 45,
+    "total_steps": 50,
+    "cache_step_interval": 4,
+    "save_generation_config": true,
+    "checkpoint_path": "/root/yw008-1-sage-videoscore2-token-credit-video-adv/checkpoint-100/transformer",
+    "lora_path": null,
+    "use_fp8_gemm": false,
+    "quant_type": "fp8-per-token-sgl",
+    "include_patterns": "double_blocks"
+  }
+}

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_008_seed_789_From a high vantage point the camera captures a bu_config.json

@@ -0,0 +1,49 @@
+{
+  "timestamp": "2026-06-13T03:04:30.161436",
+  "task": "t2v",
+  "transformer_version": "480p_t2v",
+  "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_008_seed_789_From a high vantage point the camera captures a bu.mp4",
+  "arguments": {
+    "prompt": "From a high vantage point, the camera captures a bustling city street below, where pedestrians in colorful attire navigate the crosswalks, their movements synchronized like a choreographed dance. The soft, golden lighting of the late afternoon bathes the scene, casting long shadows and highlighting the vibrant hues of the cityscape. Cars glide smoothly along the road, their headlights twinkling like stars in the gentle light. The camera pans slowly, revealing a tapestry of urban life, where the rhythm of footsteps and the hum of traffic create a harmonious symphony, all under the warm embrace of the setting sun.",
+    "valid_video_csv": [
+      "assets/video_gen_eval_top100.csv"
+    ],
+    "negative_prompt": "",
+    "resolution": "480p",
+    "model_path": "./ckpts",
+    "aspect_ratio": "16:9",
+    "fixed_size": "480x864",
+    "num_inference_steps": 40,
+    "video_length": 121,
+    "sr": false,
+    "save_pre_sr_video": false,
+    "rewrite": false,
+    "cfg_distilled": false,
+    "enable_step_distill": false,
+    "sparse_attn": false,
+    "offloading": false,
+    "group_offloading": null,
+    "overlap_group_offloading": false,
+    "dtype": "bf16",
+    "seed": 789,
+    "image_path": null,
+    "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_008_seed_789_From a high vantage point the camera captures a bu.mp4",
+    "original_noise_dir": null,
+    "use_sageattn": false,
+    "sage_blocks_range": "0-53",
+    "enable_torch_compile": false,
+    "enable_cache": false,
+    "cache_type": "deepcache",
+    "no_cache_block_id": "53",
+    "cache_start_step": 11,
+    "cache_end_step": 45,
+    "total_steps": 50,
+    "cache_step_interval": 4,
+    "save_generation_config": true,
+    "checkpoint_path": "/root/yw008-1-sage-videoscore2-token-credit-video-adv/checkpoint-100/transformer",
+    "lora_path": null,
+    "use_fp8_gemm": false,
+    "quant_type": "fp8-per-token-sgl",
+    "include_patterns": "double_blocks"
+  }
+}

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_015_seed_789_A man in a sleek black suit and polished shoes sta_config.json

@@ -0,0 +1,49 @@
+{
+  "timestamp": "2026-06-13T03:18:55.048469",
+  "task": "t2v",
+  "transformer_version": "480p_t2v",
+  "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_015_seed_789_A man in a sleek black suit and polished shoes sta.mp4",
+  "arguments": {
+    "prompt": "A man in a sleek black suit and polished shoes stands confidently in a minimalist, modern room with soft lighting. The camera captures his entire form, highlighting his poised stance and the subtle sheen of his attire. As he turns to his left, the fabric of his suit moves fluidly, emphasizing his graceful motion. He begins to walk with purpose across the polished wooden floor, his footsteps echoing softly in the serene space. The background features a large window with sheer curtains, allowing gentle natural light to filter through, casting a warm glow on his path.",
+    "valid_video_csv": [
+      "assets/video_gen_eval_top100.csv"
+    ],
+    "negative_prompt": "",
+    "resolution": "480p",
+    "model_path": "./ckpts",
+    "aspect_ratio": "16:9",
+    "fixed_size": "480x864",
+    "num_inference_steps": 40,
+    "video_length": 121,
+    "sr": false,
+    "save_pre_sr_video": false,
+    "rewrite": false,
+    "cfg_distilled": false,
+    "enable_step_distill": false,
+    "sparse_attn": false,
+    "offloading": false,
+    "group_offloading": null,
+    "overlap_group_offloading": false,
+    "dtype": "bf16",
+    "seed": 789,
+    "image_path": null,
+    "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_015_seed_789_A man in a sleek black suit and polished shoes sta.mp4",
+    "original_noise_dir": null,
+    "use_sageattn": false,
+    "sage_blocks_range": "0-53",
+    "enable_torch_compile": false,
+    "enable_cache": false,
+    "cache_type": "deepcache",
+    "no_cache_block_id": "53",
+    "cache_start_step": 11,
+    "cache_end_step": 45,
+    "total_steps": 50,
+    "cache_step_interval": 4,
+    "save_generation_config": true,
+    "checkpoint_path": "/root/yw008-1-sage-videoscore2-token-credit-video-adv/checkpoint-100/transformer",
+    "lora_path": null,
+    "use_fp8_gemm": false,
+    "quant_type": "fp8-per-token-sgl",
+    "include_patterns": "double_blocks"
+  }
+}

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_017_seed_789_A charming little girl wearing a bright yellow rai_config.json

@@ -0,0 +1,49 @@
+{
+  "timestamp": "2026-06-13T03:23:03.878825",
+  "task": "t2v",
+  "transformer_version": "480p_t2v",
+  "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_017_seed_789_A charming little girl wearing a bright yellow rai.mp4",
+  "arguments": {
+    "prompt": "A charming little girl, wearing a bright yellow raincoat and red boots, walks confidently down a bustling city street, holding a vibrant rainbow umbrella. Beside her, a small, fluffy white dog trots happily, its leash held firmly in her left hand. The street is lined with colorful storefronts and bustling pedestrians, creating a lively urban backdrop. The girl's joyful expression and the dog's playful demeanor add a sense of warmth and innocence to the scene. As they walk, the sun peeks through the clouds, casting a gentle glow on the wet pavement, enhancing the vibrant colors of the cityscape.",
+    "valid_video_csv": [
+      "assets/video_gen_eval_top100.csv"
+    ],
+    "negative_prompt": "",
+    "resolution": "480p",
+    "model_path": "./ckpts",
+    "aspect_ratio": "16:9",
+    "fixed_size": "480x864",
+    "num_inference_steps": 40,
+    "video_length": 121,
+    "sr": false,
+    "save_pre_sr_video": false,
+    "rewrite": false,
+    "cfg_distilled": false,
+    "enable_step_distill": false,
+    "sparse_attn": false,
+    "offloading": false,
+    "group_offloading": null,
+    "overlap_group_offloading": false,
+    "dtype": "bf16",
+    "seed": 789,
+    "image_path": null,
+    "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_017_seed_789_A charming little girl wearing a bright yellow rai.mp4",
+    "original_noise_dir": null,
+    "use_sageattn": false,
+    "sage_blocks_range": "0-53",
+    "enable_torch_compile": false,
+    "enable_cache": false,
+    "cache_type": "deepcache",
+    "no_cache_block_id": "53",
+    "cache_start_step": 11,
+    "cache_end_step": 45,
+    "total_steps": 50,
+    "cache_step_interval": 4,
+    "save_generation_config": true,
+    "checkpoint_path": "/root/yw008-1-sage-videoscore2-token-credit-video-adv/checkpoint-100/transformer",
+    "lora_path": null,
+    "use_fp8_gemm": false,
+    "quant_type": "fp8-per-token-sgl",
+    "include_patterns": "double_blocks"
+  }
+}

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_019_seed_789_A static camera captures a charming scene of a lit_config.json

@@ -0,0 +1,49 @@
+{
+  "timestamp": "2026-06-13T03:27:14.602024",
+  "task": "t2v",
+  "transformer_version": "480p_t2v",
+  "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_019_seed_789_A static camera captures a charming scene of a lit.mp4",
+  "arguments": {
+    "prompt": "A static camera captures a charming scene of a little girl, wearing a bright yellow raincoat and red boots, walking down a quaint cobblestone street. Her curly hair bounces with each step as she follows a small, fluffy white dog trotting eagerly ahead. The street is lined with colorful, flower-filled window boxes and vintage lampposts, creating a picturesque, storybook setting. The girl's laughter echoes softly as the dog occasionally glances back, ensuring she keeps up. Sunlight filters through the leaves of overhanging trees, casting playful shadows on the ground, enhancing the whimsical atmosphere of their joyful stroll.",
+    "valid_video_csv": [
+      "assets/video_gen_eval_top100.csv"
+    ],
+    "negative_prompt": "",
+    "resolution": "480p",
+    "model_path": "./ckpts",
+    "aspect_ratio": "16:9",
+    "fixed_size": "480x864",
+    "num_inference_steps": 40,
+    "video_length": 121,
+    "sr": false,
+    "save_pre_sr_video": false,
+    "rewrite": false,
+    "cfg_distilled": false,
+    "enable_step_distill": false,
+    "sparse_attn": false,
+    "offloading": false,
+    "group_offloading": null,
+    "overlap_group_offloading": false,
+    "dtype": "bf16",
+    "seed": 789,
+    "image_path": null,
+    "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_019_seed_789_A static camera captures a charming scene of a lit.mp4",
+    "original_noise_dir": null,
+    "use_sageattn": false,
+    "sage_blocks_range": "0-53",
+    "enable_torch_compile": false,
+    "enable_cache": false,
+    "cache_type": "deepcache",
+    "no_cache_block_id": "53",
+    "cache_start_step": 11,
+    "cache_end_step": 45,
+    "total_steps": 50,
+    "cache_step_interval": 4,
+    "save_generation_config": true,
+    "checkpoint_path": "/root/yw008-1-sage-videoscore2-token-credit-video-adv/checkpoint-100/transformer",
+    "lora_path": null,
+    "use_fp8_gemm": false,
+    "quant_type": "fp8-per-token-sgl",
+    "include_patterns": "double_blocks"
+  }
+}

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_032_seed_789_A charming Korean male star dressed in a casual wh_config.json

@@ -0,0 +1,49 @@
+{
+  "timestamp": "2026-06-13T03:54:11.032264",
+  "task": "t2v",
+  "transformer_version": "480p_t2v",
+  "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_032_seed_789_A charming Korean male star dressed in a casual wh.mp4",
+  "arguments": {
+    "prompt": "A charming Korean male star, dressed in a casual white t-shirt and jeans, stands in a sun-dappled forest clearing, surrounded by towering trees and vibrant greenery. He carefully prepares sushi on a rustic wooden table, his hands skillfully crafting each roll with precision and grace. The sunlight filters through the leaves, casting playful shadows on his focused expression. Nearby, a gentle stream babbles, adding a serene soundtrack to the scene. As he finishes, he presents the sushi with a warm smile, the forest's natural beauty enhancing the delightful culinary moment.",
+    "valid_video_csv": [
+      "assets/video_gen_eval_top100.csv"
+    ],
+    "negative_prompt": "",
+    "resolution": "480p",
+    "model_path": "./ckpts",
+    "aspect_ratio": "16:9",
+    "fixed_size": "480x864",
+    "num_inference_steps": 40,
+    "video_length": 121,
+    "sr": false,
+    "save_pre_sr_video": false,
+    "rewrite": false,
+    "cfg_distilled": false,
+    "enable_step_distill": false,
+    "sparse_attn": false,
+    "offloading": false,
+    "group_offloading": null,
+    "overlap_group_offloading": false,
+    "dtype": "bf16",
+    "seed": 789,
+    "image_path": null,
+    "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_032_seed_789_A charming Korean male star dressed in a casual wh.mp4",
+    "original_noise_dir": null,
+    "use_sageattn": false,
+    "sage_blocks_range": "0-53",
+    "enable_torch_compile": false,
+    "enable_cache": false,
+    "cache_type": "deepcache",
+    "no_cache_block_id": "53",
+    "cache_start_step": 11,
+    "cache_end_step": 45,
+    "total_steps": 50,
+    "cache_step_interval": 4,
+    "save_generation_config": true,
+    "checkpoint_path": "/root/yw008-1-sage-videoscore2-token-credit-video-adv/checkpoint-100/transformer",
+    "lora_path": null,
+    "use_fp8_gemm": false,
+    "quant_type": "fp8-per-token-sgl",
+    "include_patterns": "double_blocks"
+  }
+}

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_034_seed_789_A vibrant group of Caribbean dancers adorned in da_config.json

@@ -0,0 +1,49 @@
+{
+  "timestamp": "2026-06-13T03:58:22.425535",
+  "task": "t2v",
+  "transformer_version": "480p_t2v",
+  "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_034_seed_789_A vibrant group of Caribbean dancers adorned in da.mp4",
+  "arguments": {
+    "prompt": "A vibrant group of Caribbean dancers, adorned in dazzling costumes bursting with feathers and sequins, captivates the carnival crowd. The camera captures their every move, following the rhythmic sway of their hips and the intricate footwork that matches the pulsating beats of lively calypso music. Their costumes shimmer under the bright sunlight, with hues of turquoise, magenta, and gold creating a kaleidoscope of color. The dancers' expressions are joyous and infectious, their energy palpable as they twirl and leap, feathers fluttering with each movement. The scene is alive with the sounds of steel drums and cheering spectators, encapsulating the essence of Caribbean celebration.",
+    "valid_video_csv": [
+      "assets/video_gen_eval_top100.csv"
+    ],
+    "negative_prompt": "",
+    "resolution": "480p",
+    "model_path": "./ckpts",
+    "aspect_ratio": "16:9",
+    "fixed_size": "480x864",
+    "num_inference_steps": 40,
+    "video_length": 121,
+    "sr": false,
+    "save_pre_sr_video": false,
+    "rewrite": false,
+    "cfg_distilled": false,
+    "enable_step_distill": false,
+    "sparse_attn": false,
+    "offloading": false,
+    "group_offloading": null,
+    "overlap_group_offloading": false,
+    "dtype": "bf16",
+    "seed": 789,
+    "image_path": null,
+    "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_034_seed_789_A vibrant group of Caribbean dancers adorned in da.mp4",
+    "original_noise_dir": null,
+    "use_sageattn": false,
+    "sage_blocks_range": "0-53",
+    "enable_torch_compile": false,
+    "enable_cache": false,
+    "cache_type": "deepcache",
+    "no_cache_block_id": "53",
+    "cache_start_step": 11,
+    "cache_end_step": 45,
+    "total_steps": 50,
+    "cache_step_interval": 4,
+    "save_generation_config": true,
+    "checkpoint_path": "/root/yw008-1-sage-videoscore2-token-credit-video-adv/checkpoint-100/transformer",
+    "lora_path": null,
+    "use_fp8_gemm": false,
+    "quant_type": "fp8-per-token-sgl",
+    "include_patterns": "double_blocks"
+  }
+}

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_041_seed_789_A lively group of South Asian women adorned in vib_config.json

@@ -0,0 +1,49 @@
+{
+  "timestamp": "2026-06-13T04:13:00.997837",
+  "task": "t2v",
+  "transformer_version": "480p_t2v",
+  "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_041_seed_789_A lively group of South Asian women adorned in vib.mp4",
+  "arguments": {
+    "prompt": "A lively group of South Asian women, adorned in vibrant saris of red, gold, and emerald, gracefully perform a traditional dance at a bustling festival. The camera captures their synchronized movements, as their bangles jingle and anklets chime, creating a symphony of sound. The vivid colors of their outfits shimmer under the festival lights, enhancing the joyous atmosphere. Their expressions are a blend of concentration and joy, reflecting the cultural richness of the celebration. The background is filled with colorful decorations and a cheering crowd, adding to the festive ambiance and highlighting the beauty of their cultural heritage.",
+    "valid_video_csv": [
+      "assets/video_gen_eval_top100.csv"
+    ],
+    "negative_prompt": "",
+    "resolution": "480p",
+    "model_path": "./ckpts",
+    "aspect_ratio": "16:9",
+    "fixed_size": "480x864",
+    "num_inference_steps": 40,
+    "video_length": 121,
+    "sr": false,
+    "save_pre_sr_video": false,
+    "rewrite": false,
+    "cfg_distilled": false,
+    "enable_step_distill": false,
+    "sparse_attn": false,
+    "offloading": false,
+    "group_offloading": null,
+    "overlap_group_offloading": false,
+    "dtype": "bf16",
+    "seed": 789,
+    "image_path": null,
+    "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_041_seed_789_A lively group of South Asian women adorned in vib.mp4",
+    "original_noise_dir": null,
+    "use_sageattn": false,
+    "sage_blocks_range": "0-53",
+    "enable_torch_compile": false,
+    "enable_cache": false,
+    "cache_type": "deepcache",
+    "no_cache_block_id": "53",
+    "cache_start_step": 11,
+    "cache_end_step": 45,
+    "total_steps": 50,
+    "cache_step_interval": 4,
+    "save_generation_config": true,
+    "checkpoint_path": "/root/yw008-1-sage-videoscore2-token-credit-video-adv/checkpoint-100/transformer",
+    "lora_path": null,
+    "use_fp8_gemm": false,
+    "quant_type": "fp8-per-token-sgl",
+    "include_patterns": "double_blocks"
+  }
+}

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_059_seed_789_A solitary figure sits on the edge of a dimly lit_config.json

@@ -0,0 +1,49 @@
+{
+  "timestamp": "2026-06-13T04:50:08.995067",
+  "task": "t2v",
+  "transformer_version": "480p_t2v",
+  "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_059_seed_789_A solitary figure sits on the edge of a dimly lit.mp4",
+  "arguments": {
+    "prompt": "A solitary figure sits on the edge of a dimly lit room, their silhouette partially illuminated by a single flickering candle, casting shadows that dance across the walls. Their face, a canvas of raw emotion, reveals furrowed brows and tear-filled eyes, reflecting the storm within. The camera zooms in, capturing the trembling of their hands as they clutch a crumpled letter, a symbol of their inner conflict. The room's muted colors and sparse furnishings mirror their isolation, while the soft, haunting melody playing in the background underscores their emotional turmoil. As the camera pans out, the individual takes a deep, shuddering breath, their gaze distant, lost in the labyrinth of their thoughts.",
+    "valid_video_csv": [
+      "assets/video_gen_eval_top100.csv"
+    ],
+    "negative_prompt": "",
+    "resolution": "480p",
+    "model_path": "./ckpts",
+    "aspect_ratio": "16:9",
+    "fixed_size": "480x864",
+    "num_inference_steps": 40,
+    "video_length": 121,
+    "sr": false,
+    "save_pre_sr_video": false,
+    "rewrite": false,
+    "cfg_distilled": false,
+    "enable_step_distill": false,
+    "sparse_attn": false,
+    "offloading": false,
+    "group_offloading": null,
+    "overlap_group_offloading": false,
+    "dtype": "bf16",
+    "seed": 789,
+    "image_path": null,
+    "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_059_seed_789_A solitary figure sits on the edge of a dimly lit.mp4",
+    "original_noise_dir": null,
+    "use_sageattn": false,
+    "sage_blocks_range": "0-53",
+    "enable_torch_compile": false,
+    "enable_cache": false,
+    "cache_type": "deepcache",
+    "no_cache_block_id": "53",
+    "cache_start_step": 11,
+    "cache_end_step": 45,
+    "total_steps": 50,
+    "cache_step_interval": 4,
+    "save_generation_config": true,
+    "checkpoint_path": "/root/yw008-1-sage-videoscore2-token-credit-video-adv/checkpoint-100/transformer",
+    "lora_path": null,
+    "use_fp8_gemm": false,
+    "quant_type": "fp8-per-token-sgl",
+    "include_patterns": "double_blocks"
+  }
+}

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_064_seed_789_In a dimly lit smoke-filled room a diverse group o_config.json

@@ -0,0 +1,49 @@
+{
+  "timestamp": "2026-06-13T05:00:23.859828",
+  "task": "t2v",
+  "transformer_version": "480p_t2v",
+  "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_064_seed_789_In a dimly lit smoke-filled room a diverse group o.mp4",
+  "arguments": {
+    "prompt": "In a dimly lit, smoke-filled room, a diverse group of individuals huddles around a large table cluttered with blueprints, maps, and digital tablets, their faces illuminated by the soft glow of a single overhead lamp. The camera pans over their intense expressions, capturing the leader, a sharp-eyed woman in a tailored suit, outlining the heist's intricate details with precision. As they strategize, the camera zooms in on a young man, his brow furrowed, grappling with the moral implications of their plan. Tension mounts as they debate the risks and rewards, their voices a mix of excitement and apprehension. The scene shifts to a montage of their preparations, each member honing their skills, from hacking security systems to practicing stealth maneuvers. The camera lingers on their faces, revealing moments of doubt and resolve, as they confront the ethical dilemmas of their audacious scheme.",
+    "valid_video_csv": [
+      "assets/video_gen_eval_top100.csv"
+    ],
+    "negative_prompt": "",
+    "resolution": "480p",
+    "model_path": "./ckpts",
+    "aspect_ratio": "16:9",
+    "fixed_size": "480x864",
+    "num_inference_steps": 40,
+    "video_length": 121,
+    "sr": false,
+    "save_pre_sr_video": false,
+    "rewrite": false,
+    "cfg_distilled": false,
+    "enable_step_distill": false,
+    "sparse_attn": false,
+    "offloading": false,
+    "group_offloading": null,
+    "overlap_group_offloading": false,
+    "dtype": "bf16",
+    "seed": 789,
+    "image_path": null,
+    "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_064_seed_789_In a dimly lit smoke-filled room a diverse group o.mp4",
+    "original_noise_dir": null,
+    "use_sageattn": false,
+    "sage_blocks_range": "0-53",
+    "enable_torch_compile": false,
+    "enable_cache": false,
+    "cache_type": "deepcache",
+    "no_cache_block_id": "53",
+    "cache_start_step": 11,
+    "cache_end_step": 45,
+    "total_steps": 50,
+    "cache_step_interval": 4,
+    "save_generation_config": true,
+    "checkpoint_path": "/root/yw008-1-sage-videoscore2-token-credit-video-adv/checkpoint-100/transformer",
+    "lora_path": null,
+    "use_fp8_gemm": false,
+    "quant_type": "fp8-per-token-sgl",
+    "include_patterns": "double_blocks"
+  }
+}

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_075_seed_789_In the dim light of dawn a delicate spiders web gl_config.json

@@ -0,0 +1,49 @@
+{
+  "timestamp": "2026-06-13T05:22:56.575710",
+  "task": "t2v",
+  "transformer_version": "480p_t2v",
+  "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_075_seed_789_In the dim light of dawn a delicate spiders web gl.mp4",
+  "arguments": {
+    "prompt": "In the dim light of dawn, a delicate spider's web glistens with dew, each droplet a miniature universe suspended in the intricate silk. As the camera zooms in, the droplets reveal swirling galaxies, vibrant with stars and colorful nebulae, creating a cosmic spectacle. The dark, moody background enhances the ethereal beauty of this natural wonder, as if the web itself is a portal to the cosmos. The gentle sway of the web in the breeze adds a sense of motion, making the galaxies appear alive, twinkling with the mysteries of the universe.",
+    "valid_video_csv": [
+      "assets/video_gen_eval_top100.csv"
+    ],
+    "negative_prompt": "",
+    "resolution": "480p",
+    "model_path": "./ckpts",
+    "aspect_ratio": "16:9",
+    "fixed_size": "480x864",
+    "num_inference_steps": 40,
+    "video_length": 121,
+    "sr": false,
+    "save_pre_sr_video": false,
+    "rewrite": false,
+    "cfg_distilled": false,
+    "enable_step_distill": false,
+    "sparse_attn": false,
+    "offloading": false,
+    "group_offloading": null,
+    "overlap_group_offloading": false,
+    "dtype": "bf16",
+    "seed": 789,
+    "image_path": null,
+    "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_075_seed_789_In the dim light of dawn a delicate spiders web gl.mp4",
+    "original_noise_dir": null,
+    "use_sageattn": false,
+    "sage_blocks_range": "0-53",
+    "enable_torch_compile": false,
+    "enable_cache": false,
+    "cache_type": "deepcache",
+    "no_cache_block_id": "53",
+    "cache_start_step": 11,
+    "cache_end_step": 45,
+    "total_steps": 50,
+    "cache_step_interval": 4,
+    "save_generation_config": true,
+    "checkpoint_path": "/root/yw008-1-sage-videoscore2-token-credit-video-adv/checkpoint-100/transformer",
+    "lora_path": null,
+    "use_fp8_gemm": false,
+    "quant_type": "fp8-per-token-sgl",
+    "include_patterns": "double_blocks"
+  }
+}

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_076_seed_789_The camera soars above a lively medieval marketpla_config.json

@@ -0,0 +1,49 @@
+{
+  "timestamp": "2026-06-13T05:24:59.686641",
+  "task": "t2v",
+  "transformer_version": "480p_t2v",
+  "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_076_seed_789_The camera soars above a lively medieval marketpla.mp4",
+  "arguments": {
+    "prompt": "The camera soars above a lively medieval marketplace, capturing vibrant stalls brimming with colorful fabrics, fresh produce, and handcrafted goods. Merchants in traditional garb enthusiastically barter with animated townsfolk, while children weave through the crowd, laughter echoing. The scene transitions smoothly, morphing into a bustling modern cityscape. Towering skyscrapers pierce the sky, their glass facades reflecting the vibrant hues of neon lights. The streets below teem with life, cars weaving through traffic, and pedestrians navigating the urban jungle. The juxtaposition of eras highlights the evolution of commerce and community, from quaint cobblestones to sleek, illuminated avenues.",
+    "valid_video_csv": [
+      "assets/video_gen_eval_top100.csv"
+    ],
+    "negative_prompt": "",
+    "resolution": "480p",
+    "model_path": "./ckpts",
+    "aspect_ratio": "16:9",
+    "fixed_size": "480x864",
+    "num_inference_steps": 40,
+    "video_length": 121,
+    "sr": false,
+    "save_pre_sr_video": false,
+    "rewrite": false,
+    "cfg_distilled": false,
+    "enable_step_distill": false,
+    "sparse_attn": false,
+    "offloading": false,
+    "group_offloading": null,
+    "overlap_group_offloading": false,
+    "dtype": "bf16",
+    "seed": 789,
+    "image_path": null,
+    "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_076_seed_789_The camera soars above a lively medieval marketpla.mp4",
+    "original_noise_dir": null,
+    "use_sageattn": false,
+    "sage_blocks_range": "0-53",
+    "enable_torch_compile": false,
+    "enable_cache": false,
+    "cache_type": "deepcache",
+    "no_cache_block_id": "53",
+    "cache_start_step": 11,
+    "cache_end_step": 45,
+    "total_steps": 50,
+    "cache_step_interval": 4,
+    "save_generation_config": true,
+    "checkpoint_path": "/root/yw008-1-sage-videoscore2-token-credit-video-adv/checkpoint-100/transformer",
+    "lora_path": null,
+    "use_fp8_gemm": false,
+    "quant_type": "fp8-per-token-sgl",
+    "include_patterns": "double_blocks"
+  }
+}

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_085_seed_789_The camera glides through a majestic glacial canyo_config.json

@@ -0,0 +1,49 @@
+{
+  "timestamp": "2026-06-13T05:43:32.852771",
+  "task": "t2v",
+  "transformer_version": "480p_t2v",
+  "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_085_seed_789_The camera glides through a majestic glacial canyo.mp4",
+  "arguments": {
+    "prompt": "The camera glides through a majestic glacial canyon, its icy walls towering and glistening under a soft, muted light. As it soars, the scene transitions into a surreal cloudscape, where fluffy clouds drift lazily across a pastel sky. The motion is fluid and dynamic, capturing the ethereal beauty of the clouds. Suddenly, the view shifts to a serene forest, where tall trees sway gently in a whispering breeze, their leaves a muted green. The journey continues, seamlessly flowing into a tranquil desert, where rolling dunes stretch endlessly, bathed in the soft glow of a setting sun. The muted color palette ties each landscape together, creating a harmonious and dreamlike visual experience.",
+    "valid_video_csv": [
+      "assets/video_gen_eval_top100.csv"
+    ],
+    "negative_prompt": "",
+    "resolution": "480p",
+    "model_path": "./ckpts",
+    "aspect_ratio": "16:9",
+    "fixed_size": "480x864",
+    "num_inference_steps": 40,
+    "video_length": 121,
+    "sr": false,
+    "save_pre_sr_video": false,
+    "rewrite": false,
+    "cfg_distilled": false,
+    "enable_step_distill": false,
+    "sparse_attn": false,
+    "offloading": false,
+    "group_offloading": null,
+    "overlap_group_offloading": false,
+    "dtype": "bf16",
+    "seed": 789,
+    "image_path": null,
+    "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_085_seed_789_The camera glides through a majestic glacial canyo.mp4",
+    "original_noise_dir": null,
+    "use_sageattn": false,
+    "sage_blocks_range": "0-53",
+    "enable_torch_compile": false,
+    "enable_cache": false,
+    "cache_type": "deepcache",
+    "no_cache_block_id": "53",
+    "cache_start_step": 11,
+    "cache_end_step": 45,
+    "total_steps": 50,
+    "cache_step_interval": 4,
+    "save_generation_config": true,
+    "checkpoint_path": "/root/yw008-1-sage-videoscore2-token-credit-video-adv/checkpoint-100/transformer",
+    "lora_path": null,
+    "use_fp8_gemm": false,
+    "quant_type": "fp8-per-token-sgl",
+    "include_patterns": "double_blocks"
+  }
+}

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_096_seed_789_A mesmerizing dancer clad in a flowing crimson dre_config.json

@@ -0,0 +1,49 @@
+{
+  "timestamp": "2026-06-13T06:06:12.653077",
+  "task": "t2v",
+  "transformer_version": "480p_t2v",
+  "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_096_seed_789_A mesmerizing dancer clad in a flowing crimson dre.mp4",
+  "arguments": {
+    "prompt": "A mesmerizing dancer, clad in a flowing crimson dress, twirls elegantly on a dimly lit stage, her movements fluid and captivating. The camera circles her, capturing each graceful step and spin, as vibrant, multicolored lights cast dynamic patterns across the stage. Wisps of smoke swirl around her, adding an ethereal quality to the performance. Her expressive face reflects passion and intensity, while the lights shift from deep blues to fiery oranges, enhancing the dramatic atmosphere. The camera captures her every move, from delicate hand gestures to powerful leaps, against the backdrop of swirling smoke and vibrant, ever-changing lights.",
+    "valid_video_csv": [
+      "assets/video_gen_eval_top100.csv"
+    ],
+    "negative_prompt": "",
+    "resolution": "480p",
+    "model_path": "./ckpts",
+    "aspect_ratio": "16:9",
+    "fixed_size": "480x864",
+    "num_inference_steps": 40,
+    "video_length": 121,
+    "sr": false,
+    "save_pre_sr_video": false,
+    "rewrite": false,
+    "cfg_distilled": false,
+    "enable_step_distill": false,
+    "sparse_attn": false,
+    "offloading": false,
+    "group_offloading": null,
+    "overlap_group_offloading": false,
+    "dtype": "bf16",
+    "seed": 789,
+    "image_path": null,
+    "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_096_seed_789_A mesmerizing dancer clad in a flowing crimson dre.mp4",
+    "original_noise_dir": null,
+    "use_sageattn": false,
+    "sage_blocks_range": "0-53",
+    "enable_torch_compile": false,
+    "enable_cache": false,
+    "cache_type": "deepcache",
+    "no_cache_block_id": "53",
+    "cache_start_step": 11,
+    "cache_end_step": 45,
+    "total_steps": 50,
+    "cache_step_interval": 4,
+    "save_generation_config": true,
+    "checkpoint_path": "/root/yw008-1-sage-videoscore2-token-credit-video-adv/checkpoint-100/transformer",
+    "lora_path": null,
+    "use_fp8_gemm": false,
+    "quant_type": "fp8-per-token-sgl",
+    "include_patterns": "double_blocks"
+  }
+}

outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_099_seed_789_In a sunlit studio a painter with tousled hair and_config.json

@@ -0,0 +1,49 @@
+{
+  "timestamp": "2026-06-13T06:12:22.870809",
+  "task": "t2v",
+  "transformer_version": "480p_t2v",
+  "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_099_seed_789_In a sunlit studio a painter with tousled hair and.mp4",
+  "arguments": {
+    "prompt": "In a sunlit studio, a painter with tousled hair and a focused expression stands before a large canvas, surrounded by an array of vibrant paints and brushes. The camera captures the artist's hand as it deftly moves across the canvas, blending rich blues and fiery oranges with a palette knife, creating a mesmerizing dance of colors. As the scene unfolds, the painter steps back, revealing the gradual emergence of a breathtaking landscape, with rolling hills and a golden sky. The camera zooms in on the intricate details, showcasing the texture and depth of the brushstrokes, as the artwork comes to life in a symphony of hues.",
+    "valid_video_csv": [
+      "assets/video_gen_eval_top100.csv"
+    ],
+    "negative_prompt": "",
+    "resolution": "480p",
+    "model_path": "./ckpts",
+    "aspect_ratio": "16:9",
+    "fixed_size": "480x864",
+    "num_inference_steps": 40,
+    "video_length": 121,
+    "sr": false,
+    "save_pre_sr_video": false,
+    "rewrite": false,
+    "cfg_distilled": false,
+    "enable_step_distill": false,
+    "sparse_attn": false,
+    "offloading": false,
+    "group_offloading": null,
+    "overlap_group_offloading": false,
+    "dtype": "bf16",
+    "seed": 789,
+    "image_path": null,
+    "output_path": "./outputs/yw008-1-sage-videoscore2-token-credit-video-adv/videogen_val3/step_000100/sample_099_seed_789_In a sunlit studio a painter with tousled hair and.mp4",
+    "original_noise_dir": null,
+    "use_sageattn": false,
+    "sage_blocks_range": "0-53",
+    "enable_torch_compile": false,
+    "enable_cache": false,
+    "cache_type": "deepcache",
+    "no_cache_block_id": "53",
+    "cache_start_step": 11,
+    "cache_end_step": 45,
+    "total_steps": 50,
+    "cache_step_interval": 4,
+    "save_generation_config": true,
+    "checkpoint_path": "/root/yw008-1-sage-videoscore2-token-credit-video-adv/checkpoint-100/transformer",
+    "lora_path": null,
+    "use_fp8_gemm": false,
+    "quant_type": "fp8-per-token-sgl",
+    "include_patterns": "double_blocks"
+  }
+}