ValueError: array must not contain infs or NaNs

[TianheWu]

Using GPU 2
number of train scenes: 160
number of test scenes: 40
100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1875/1875 [06:53<00:00, 4.54it/s]
Traceback (most recent call last):
File "/home/wth/IQT/train.py", line 156, in
loss, rho_s, rho_p = train_epoch(config, epoch, model_transformer, model_backbone, save_output, criterion, optimizer, scheduler, train_loader)
File "/home/wth/IQT/trainer.py", line 87, in train_epoch
rho_p, _ = pearsonr(np.squeeze(pred_epoch), np.squeeze(labels_epoch))
File "/home/wth/anaconda3/envs/wthEnv/lib/python3.9/site-packages/scipy/stats/stats.py", line 4045, in pearsonr
normxm = linalg.norm(xm)
File "/home/wth/anaconda3/envs/wthEnv/lib/python3.9/site-packages/scipy/linalg/misc.py", line 145, in norm
a = np.asarray_chkfinite(a)
File "/home/wth/anaconda3/envs/wthEnv/lib/python3.9/site-packages/numpy/lib/function_base.py", line 488, in asarray_chkfinite
raise ValueError(
ValueError: array must not contain infs or NaNs

I using scipy 1.7.2. I don't know how to solve this problem.

[anse3832]

Maybe the variable "pred_epoch" has infs or NaNs when calculating correlation coefficients. Did you check for that?

[870715478]

Using GPU 2 number of train scenes: 160 number of test scenes: 40 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1875/1875 [06:53<00:00, 4.54it/s] Traceback (most recent call last): File "/home/wth/IQT/train.py", line 156, in loss, rho_s, rho_p = train_epoch(config, epoch, model_transformer, model_backbone, save_output, criterion, optimizer, scheduler, train_loader) File "/home/wth/IQT/trainer.py", line 87, in train_epoch rho_p, _ = pearsonr(np.squeeze(pred_epoch), np.squeeze(labels_epoch)) File "/home/wth/anaconda3/envs/wthEnv/lib/python3.9/site-packages/scipy/stats/stats.py", line 4045, in pearsonr normxm = linalg.norm(xm) File "/home/wth/anaconda3/envs/wthEnv/lib/python3.9/site-packages/scipy/linalg/misc.py", line 145, in norm a = np.asarray_chkfinite(a) File "/home/wth/anaconda3/envs/wthEnv/lib/python3.9/site-packages/numpy/lib/function_base.py", line 488, in asarray_chkfinite raise ValueError( ValueError: array must not contain infs or NaNs

I using scipy 1.7.2. I don't know how to solve this problem.

Can datasets be used directly? Need to deal with it? Like the format or something?

[TianheWu]

I solved this problem by redownload the pre-trained model.

…

---Original--- From: ***@***.***&gt; Date: Tue, Jan 4, 2022 20:55 PM To: ***@***.***&gt;; Cc: "Tianhe ***@***.******@***.***&gt;; Subject: Re: [anse3832/IQT] ValueError: array must not contain infs or NaNs (Issue #2) Using GPU 2 number of train scenes: 160 number of test scenes: 40 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1875/1875 [06:53<00:00, 4.54it/s] Traceback (most recent call last): File "/home/wth/IQT/train.py", line 156, in loss, rho_s, rho_p = train_epoch(config, epoch, model_transformer, model_backbone, save_output, criterion, optimizer, scheduler, train_loader) File "/home/wth/IQT/trainer.py", line 87, in train_epoch rho_p, _ = pearsonr(np.squeeze(pred_epoch), np.squeeze(labels_epoch)) File "/home/wth/anaconda3/envs/wthEnv/lib/python3.9/site-packages/scipy/stats/stats.py", line 4045, in pearsonr normxm = linalg.norm(xm) File "/home/wth/anaconda3/envs/wthEnv/lib/python3.9/site-packages/scipy/linalg/misc.py", line 145, in norm a = np.asarray_chkfinite(a) File "/home/wth/anaconda3/envs/wthEnv/lib/python3.9/site-packages/numpy/lib/function_base.py", line 488, in asarray_chkfinite raise ValueError( ValueError: array must not contain infs or NaNs I using scipy 1.7.2. I don't know how to solve this problem. Can datasets be used directly? Need to deal with it? Like the format or something? — Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you authored the thread.Message ID: ***@***.***&gt;