evaluation

[Xie-Shaofeng]

您好，我在运行测试脚本bash scripts/train train_vg1.2.sh后，得到权重文件output/train/vg1.2_5e/20250823190/checkpoint_0.pth，然后运行测试脚本bash scripts/eval/eval_vg1.2_densecap.sh output/train/vg1.2_5e/20250823190/checkpoint_0.pth输出了如下的结果： 2025-08-24 18:01:25,810 [INFO] load checkpoint from /root/autodl-tmp/ControlCap/ckpts/checkpoint_0.pth 2025-08-24 18:01:25,824 [INFO] Evaluating on val. 2025-08-24 18:01:25,825 [INFO] dataset_ratios not specified, datasets will be concatenated (map-style datasets) or chained (webdataset.DataPipeline). 2025-08-24 18:01:25,825 [INFO] Loaded 3682892 records for train split from the dataset. 2025-08-24 18:01:25,825 [INFO] Loaded 5000 records for val split from the dataset. 2025-08-24 18:01:25,825 [INFO] Empty train splits. 2025-08-24 18:01:25,825 [INFO] Empty train splits. Loading checkpoint shards: 100%|███████████████████████████████████████████████| 2/2 [02:38<00:00, 79.22s/it] Evaluation [ 0/2500] eta: 4:42:57 time: 6.7910 data: 2.9157 max mem: 17443 Evaluation [ 10/2500] eta: 1:18:03 time: 1.8810 data: 0.2698 max mem: 18689 Evaluation [2490/2500] eta: 0:00:14 time: 1.4257 data: 0.0039 max mem: 21590 Evaluation [2499/2500] eta: 0:00:01 time: 1.5855 data: 0.0568 max mem: 21590 Evaluation Total time: 1:00:39 (1.4558 s / it) 2025-08-24 19:03:36,920 [WARNING] Merging results. 100%|██████████████████████████████████████████████████████████████████████████| 2/2 [00:00<00:00, 2.86it/s] 2025-08-24 19:04:04,962 [INFO] :Get (0/237996) predictions. 2025-08-24 19:04:04,963 [INFO] :Save result to (/root/autodl-tmp/ControlCap/output/eval/vg1.2/20250824175/result/val.json). 2025-08-24 19:04:05,699 [INFO] :Begin evaluation (/root/autodl-tmp/ControlCap/output/eval/vg1.2/20250824175/result/val.json). loading annotations into memory... Done (t=3.53s) creating index... index created! loading annotations into memory... Done (t=3.53s) creating index... index created! 100%|██████████████████████████████████████████████████████████████████| 5000/5000 [00:04<00:00, 1001.58it/s] 100%|███████████████████████████████████████████████████████████████████| 5000/5000 [00:24<00:00, 202.84it/s] huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks... To disable this warning, you can either: - Avoid using tokenizers before the fork if possible - Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false) Traceback (most recent call last): File "train.py", line 98, in main() File "train.py", line 94, in main runner.train() File "/root/autodl-tmp/ControlCap/controlcap/runners/runner.py", line 113, in train val_log = self.eval_epoch( File "/root/miniconda3/envs/controlcap/lib/python3.8/site-packages/torch/utils/_contextlib.py", line 115, in decorate_context return func(*args, **kwargs) File "/root/miniconda3/envs/controlcap/lib/python3.8/site-packages/lavis/runners/runner_base.py", line 472, in eval_epoch return self.task.after_evaluation( File "/root/autodl-tmp/ControlCap/controlcap/tasks/task.py", line 348, in after_evaluation metrics = self.report_metrics_densecap(result_file=result_file) File "/root/miniconda3/envs/controlcap/lib/python3.8/site-packages/lavis/common/dist_utils.py", line 112, in wrapper return func(*args, **kwargs) File "/root/autodl-tmp/ControlCap/controlcap/tasks/task.py", line 448, in report_metrics_densecap metrics = ev.evaluate() File "/root/autodl-tmp/ControlCap/controlcap/common/evaluation/eval_densecap.py", line 169, in evaluate meteor, meteors = self.score_captions() File "/root/autodl-tmp/ControlCap/controlcap/common/evaluation/eval_densecap.py", line 92, in score_captions meteor_scorer = Meteor() File "/root/autodl-tmp/ControlCap/controlcap/common/evaluation/meteor/meteor.py", line 48, in init self.meteor_p = subprocess.Popen(meteor_cmd, File "/root/miniconda3/envs/controlcap/lib/python3.8/subprocess.py", line 858, in init self._execute_child(args, executable, preexec_fn, close_fds, File "/root/miniconda3/envs/controlcap/lib/python3.8/subprocess.py", line 1720, in _execute_child raise child_exception_type(errno_num, err_msg, err_filename) FileNotFoundError: [Errno 2] No such file or directory: 'java' 但是我好像不能安装java环境，想请教您这个问题怎么解决呢？

[callsys]

你好！我没有想到比较好的解决办法。似乎这个evaluation package依赖于java