The examples run only once, then they begin to generate error

[kmkarakaya]

Describe the bug

When I deploy an app on Spaces, the examples run only once, then if I click one of them they begin to generate error.

After building an app on spaces, each example run perfectly ONLY for ONCE. When I click any of them for the second or more times on the screen I only see ERROR message.

However, I can enter a new prompt manually and it works fine.

If I re-build the app, again the examples run only ONCE.

I checked the log but there seems no problem.

As far as I understood, there is a problem to fetch the cached example and its result.

Thank you in advanced

Is there an existing issue for this?

• I have searched the existing issues

Reproduction

kmkarakaya/Auto_Review_Generation_in_Turkish

Screenshot

Logs

2022-07-17 13:40:02.394081: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory
2022-07-17 13:40:02.394233: I tensorflow/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
2022-07-17 13:40:03.758419: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcuda.so.1'; dlerror: libcuda.so.1: cannot open shared object file: No such file or directory
2022-07-17 13:40:03.758556: W tensorflow/stream_executor/cuda/cuda_driver.cc:269] failed call to cuInit: UNKNOWN ERROR (303)
2022-07-17 13:40:03.758695: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:156] kernel driver does not appear to be running on this host (s-kmkarakaya-auto--review--generation--in--turkish-ece09-bmmkbs): /proc/driver/nvidia/version does not exist
2022-07-17 13:40:03.759272: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 AVX512F FMA
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.

Downloading:   0%|          | 0.00/869 [00:00<?, ?B/s]
Downloading: 100%|██████████| 869/869 [00:00<00:00, 766kB/s]

Downloading:   0%|          | 0.00/480M [00:00<?, ?B/s]
Downloading:   2%|▏         | 8.74M/480M [00:00<00:05, 91.6MB/s]
Downloading:   4%|▎         | 17.5M/480M [00:00<00:05, 87.8MB/s]
Downloading:   5%|▌         | 25.9M/480M [00:00<00:05, 85.4MB/s]
Downloading:   7%|▋         | 34.6M/480M [00:00<00:05, 87.8MB/s]
Downloading:   9%|▉         | 43.0M/480M [00:00<00:05, 87.0MB/s]
Downloading:  11%|█         | 51.5M/480M [00:00<00:05, 87.7MB/s]
Downloading:  13%|█▎        | 60.2M/480M [00:00<00:04, 88.8MB/s]
Downloading:  14%|█▍        | 69.4M/480M [00:00<00:04, 91.1MB/s]
Downloading:  16%|█▋        | 78.1M/480M [00:00<00:04, 90.8MB/s]
Downloading:  18%|█▊        | 86.7M/480M [00:01<00:05, 79.8MB/s]
Downloading:  20%|██        | 96.1M/480M [00:01<00:04, 85.0MB/s]
Downloading:  22%|██▏       | 105M/480M [00:01<00:04, 86.6MB/s] 
Downloading:  24%|██▎       | 114M/480M [00:01<00:04, 88.3MB/s]
Downloading:  25%|██▌       | 122M/480M [00:01<00:04, 87.8MB/s]
Downloading:  27%|██▋       | 131M/480M [00:01<00:04, 88.1MB/s]
Downloading:  29%|██▉       | 140M/480M [00:01<00:03, 90.6MB/s]
Downloading:  31%|███       | 148M/480M [00:01<00:03, 90.6MB/s]
Downloading:  33%|███▎      | 158M/480M [00:01<00:03, 91.8MB/s]
Downloading:  35%|███▍      | 167M/480M [00:01<00:03, 92.8MB/s]
Downloading:  37%|███▋      | 176M/480M [00:02<00:03, 94.0MB/s]
Downloading:  39%|███▊      | 185M/480M [00:02<00:03, 95.0MB/s]
Downloading:  40%|████      | 194M/480M [00:02<00:03, 95.7MB/s]
Downloading:  42%|████▏     | 204M/480M [00:02<00:03, 96.2MB/s]
Downloading:  44%|████▍     | 213M/480M [00:02<00:02, 96.6MB/s]
Downloading:  46%|████▋     | 222M/480M [00:02<00:02, 97.1MB/s]
Downloading:  48%|████▊     | 232M/480M [00:02<00:02, 97.6MB/s]
Downloading:  50%|█████     | 241M/480M [00:02<00:02, 97.8MB/s]
Downloading:  52%|█████▏    | 251M/480M [00:02<00:02, 98.1MB/s]
Downloading:  54%|█████▍    | 260M/480M [00:02<00:02, 98.5MB/s]
Downloading:  56%|█████▌    | 269M/480M [00:03<00:02, 98.5MB/s]
Downloading:  58%|█████▊    | 279M/480M [00:03<00:02, 98.4MB/s]
Downloading:  60%|██████    | 288M/480M [00:03<00:02, 98.3MB/s]
Downloading:  62%|██████▏   | 298M/480M [00:03<00:01, 98.3MB/s]
Downloading:  64%|██████▍   | 307M/480M [00:03<00:01, 98.1MB/s]
Downloading:  66%|██████▌   | 316M/480M [00:03<00:01, 98.1MB/s]
Downloading:  68%|██████▊   | 326M/480M [00:03<00:01, 98.1MB/s]
Downloading:  70%|██████▉   | 335M/480M [00:03<00:01, 98.2MB/s]
Downloading:  72%|███████▏  | 344M/480M [00:03<00:01, 98.2MB/s]
Downloading:  74%|███████▎  | 354M/480M [00:03<00:01, 97.6MB/s]
Downloading:  76%|███████▌  | 363M/480M [00:04<00:01, 97.3MB/s]
Downloading:  78%|███████▊  | 372M/480M [00:04<00:01, 97.0MB/s]
Downloading:  80%|███████▉  | 382M/480M [00:04<00:01, 97.2MB/s]
Downloading:  81%|████████▏ | 391M/480M [00:04<00:00, 97.3MB/s]
Downloading:  83%|████████▎ | 400M/480M [00:04<00:00, 97.7MB/s]
Downloading:  85%|████████▌ | 410M/480M [00:04<00:00, 97.8MB/s]
Downloading:  87%|████████▋ | 419M/480M [00:04<00:00, 98.1MB/s]
Downloading:  89%|████████▉ | 429M/480M [00:04<00:00, 98.2MB/s]
Downloading:  91%|█████████▏| 438M/480M [00:04<00:00, 98.3MB/s]
Downloading:  93%|█████████▎| 447M/480M [00:04<00:00, 97.7MB/s]
Downloading:  95%|█████████▌| 457M/480M [00:05<00:00, 97.8MB/s]
Downloading:  97%|█████████▋| 466M/480M [00:05<00:00, 97.9MB/s]
Downloading:  99%|█████████▉| 475M/480M [00:05<00:00, 98.1MB/s]
Downloading: 100%|██████████| 480M/480M [00:05<00:00, 94.6MB/s]
All model checkpoint layers were used when initializing TFGPT2LMHeadModel.

All the layers of TFGPT2LMHeadModel were initialized from the model checkpoint at kmkarakaya/turkishReviews-ds.
If your task is similar to the task the model of the checkpoint was trained on, you can already use TFGPT2LMHeadModel for predictions without further training.

Downloading:   0%|          | 0.00/772 [00:00<?, ?B/s]
Downloading: 100%|██████████| 772/772 [00:00<00:00, 812kB/s]

Downloading:   0%|          | 0.00/975k [00:00<?, ?B/s]
Downloading: 100%|██████████| 975k/975k [00:00<00:00, 80.7MB/s]

Downloading:   0%|          | 0.00/630k [00:00<?, ?B/s]
Downloading: 100%|██████████| 630k/630k [00:00<00:00, 78.3MB/s]

Downloading:   0%|          | 0.00/2.41M [00:00<?, ?B/s]
Downloading: 100%|██████████| 2.41M/2.41M [00:00<00:00, 86.4MB/s]

Downloading:   0%|          | 0.00/25.0 [00:00<?, ?B/s]
Downloading: 100%|██████████| 25.0/25.0 [00:00<00:00, 22.4kB/s]

Downloading:   0%|          | 0.00/470 [00:00<?, ?B/s]
Downloading: 100%|██████████| 470/470 [00:00<00:00, 428kB/s]
/home/user/.local/lib/python3.8/site-packages/gradio/deprecation.py:40: UserWarning: `enable_queue` is deprecated in `Interface()`, please use it within `launch()` instead.
  warnings.warn(value)
/home/user/.local/lib/python3.8/site-packages/gradio/interface.py:311: UserWarning: Currently, only the 'default' theme is supported.
  warnings.warn("Currently, only the 'default' theme is supported.")
Caching examples at: '/home/user/app/gradio_cached_examples/17/log.csv'
The attention mask and the pad token id were not set. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
Setting `pad_token_id` to 0 (first `eos_token_id`) to generate sequence
The attention mask and the pad token id were not set. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
Setting `pad_token_id` to 0 (first `eos_token_id`) to generate sequence
The attention mask and the pad token id were not set. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
Setting `pad_token_id` to 0 (first `eos_token_id`) to generate sequence
The attention mask and the pad token id were not set. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
Setting `pad_token_id` to 0 (first `eos_token_id`) to generate sequence
Running on local URL:  http://localhost:7860/

To create a public link, set `share=True` in `launch()`.
<IPython.core.display.HTML object>
The attention mask and the pad token id were not set. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
Setting `pad_token_id` to 0 (first `eos_token_id`) to generate sequence
The attention mask and the pad token id were not set. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
Setting `pad_token_id` to 0 (first `eos_token_id`) to generate sequence

System Info

I use the Spaces Gradio, 
Chrome Version 103.0.5060.114 (Official Build) (64-bit)

Severity

serious, but I can work around it

[abidlabs]

Hi @kmkarakaya, I can't seem to reproduce the issue in your Space. Did you figure it out?

[kmkarakaya]

Hi @abidlabs, NO. I just set cache_examples = False to avoid the issue. Now I removed the "cache_examples = False" parameter. If you re-try the examples few times you can experience the issue. Regards

[abidlabs]

Hmm yes I see the issue, thanks we'll take a look at it.

[dhansmair]

I'm having the same issue with a huggingface space

[abidlabs]

@dhansmair @kmkarakaya this should be fixed now! We'll release a new version of gradio shortly so you can update your Spaces. Thanks for bringing this to our attention.

[abidlabs]

Should be fixed in gradio==3.1.1!

[kmkarakaya]

Unfortunately, even I set up gradio==3.1.1 in the requirements.txt, the ERROR msg still appears.

[freddyaboulton]

Hi @kmkarakaya ! Can you post the logs? Might help us figure out what's happening.

[abidlabs]

Ah I think I know the issue. Setting the version in requirements.txt doesn't work on Spaces. You have to set the sdk_version in the README. I've opened a PR in your repo to make the change: https://huggingface.co/spaces/kmkarakaya/Auto_Review_Generation_in_Turkish/discussions/2

[kmkarakaya]

@abidlabs yes, this time it works smoothly. Thank you for your consideration and help.