google-research / bert

TensorFlow code and pre-trained models for BERT

Home Page:https://arxiv.org/abs/1810.04805

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Internal: Blas GEMM launch failed when running classifier for URLs

loginName1 opened this issue · comments

System information:

  • os: Windows 11
  • gpu: Nvidia GeForce RTX 3080 TI (12GB)
  • Tensor flow: tensorflow-gpu v 1.14.0
  • cuda: v 10.0 (but i have other version installed: 12.4 and 9.0 which dont have the required .dll file, all of them are in my PATH)
  • python: 3.7 (for the purposes of protobuf)

The error:

2024-04-22 15:35:14.641625: E tensorflow/stream_executor/cuda/cuda_blas.cc:428] failed to run cuBLAS routine: CUBLAS_STATUS_EXECUTION_FAILED
ERROR:tensorflow:Error recorded from training_loop: 2 root error(s) found.
  (0) Internal: Blas GEMM launch failed : a.shape=(4096, 2), b.shape=(2, 768), m=4096, n=768, k=2
         [[node bert/embeddings/MatMul (defined at D:\Faks\UM-Mag 23-25\Drugi semester\JT\google-bert\modeling.py:487) ]]
         [[loss/Mean/_4861]]
  (1) Internal: Blas GEMM launch failed : a.shape=(4096, 2), b.shape=(2, 768), m=4096, n=768, k=2
         [[node bert/embeddings/MatMul (defined at D:\Faks\UM-Mag 23-25\Drugi semester\JT\google-bert\modeling.py:487) ]]
0 successful operations.
0 derived errors ignored.

What i've tried:
I tried checking nvidia-smi.exe to see if i had something running on the GPU while training but got the following result:

+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 552.22                 Driver Version: 552.22         CUDA Version: 12.4     |
|-----------------------------------------+------------------------+----------------------+
| GPU  Name                     TCC/WDDM  | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  NVIDIA GeForce RTX 3080 Ti   WDDM  |   00000000:0A:00.0 Off |                  N/A |
|  0%   36C    P8             24W /  350W |    1598MiB /  12288MiB |      3%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI        PID   Type   Process name                              GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|    0   N/A  N/A      9072    C+G   C:\Windows\explorer.exe                     N/A      |
|    0   N/A  N/A     10652    C+G   ...al\Discord\app-1.0.9042\Discord.exe      N/A      |
|    0   N/A  N/A     10688    C+G   ...ekyb3d8bbwe\PhoneExperienceHost.exe      N/A      |
|    0   N/A  N/A     10820    C+G   ...nt.CBS_cw5n1h2txyewy\SearchHost.exe      N/A      |
|    0   N/A  N/A     10844    C+G   ...2txyewy\StartMenuExperienceHost.exe      N/A      |
|    0   N/A  N/A     14572    C+G   ...\cef\cef.win7x64\steamwebhelper.exe      N/A      |
|    0   N/A  N/A     14744    C+G   ...GeForce Experience\NVIDIA Share.exe      N/A      |
|    0   N/A  N/A     14824    C+G   ...1.0_x64__8wekyb3d8bbwe\Video.UI.exe      N/A      |
|    0   N/A  N/A     15072    C+G   ...t.LockApp_cw5n1h2txyewy\LockApp.exe      N/A      |
|    0   N/A  N/A     17164    C+G   ...CBS_cw5n1h2txyewy\TextInputHost.exe      N/A      |
|    0   N/A  N/A     18100    C+G   ...les\Microsoft OneDrive\OneDrive.exe      N/A      |
|    0   N/A  N/A     18936    C+G   ...5n1h2txyewy\ShellExperienceHost.exe      N/A      |
|    0   N/A  N/A     19308    C+G   ...air\Corsair iCUE5 Software\iCUE.exe      N/A      |
|    0   N/A  N/A     19848    C+G   ...crosoft\Edge\Application\msedge.exe      N/A      |
|    0   N/A  N/A     23840    C+G   ..._x64__kzf8qxf38zg5c\Skype\Skype.exe      N/A      |
|    0   N/A  N/A     24040    C+G   ...lf\0.248.120.19\OverwolfBrowser.exe      N/A      |
|    0   N/A  N/A     25276    C+G   ...on\123.0.2420.97\msedgewebview2.exe      N/A      |
|    0   N/A  N/A     25604    C+G   ...ejd91yc\AdobeNotificationClient.exe      N/A      |
|    0   N/A  N/A     25636    C+G   ...509_x64__8wekyb3d8bbwe\ms-teams.exe      N/A      |
|    0   N/A  N/A     25920    C+G   ...ktop\EA Desktop\EACefSubProcess.exe      N/A      |
|    0   N/A  N/A     25972    C+G   ...\GOG Galaxy\GalaxyClient Helper.exe      N/A      |
|    0   N/A  N/A     26180    C+G   ...EA Desktop\EA Desktop\EADesktop.exe      N/A      |
|    0   N/A  N/A     28536    C+G   ...cks-services\BlueStacksServices.exe      N/A      |
|    0   N/A  N/A     29332    C+G   ...aam7r\AcrobatNotificationClient.exe      N/A      |
|    0   N/A  N/A     31120    C+G   ...on\123.0.2420.97\msedgewebview2.exe      N/A      |
|    0   N/A  N/A     31468    C+G   ..._x64__kzf8qxf38zg5c\Skype\Skype.exe      N/A      |
|    0   N/A  N/A     32340    C+G   ...m Files\Mozilla Firefox\firefox.exe      N/A      |
|    0   N/A  N/A     33104    C+G   ...on\HEX\Creative Cloud UI Helper.exe      N/A      |
|    0   N/A  N/A     33208    C+G   ...on\123.0.2420.97\msedgewebview2.exe      N/A      |
|    0   N/A  N/A     35396    C+G   ...wekyb3d8bbwe\XboxGameBarWidgets.exe      N/A      |
|    0   N/A  N/A     39264    C+G   ...m Files\Mozilla Firefox\firefox.exe      N/A      |
+-----------------------------------------------------------------------------------------+

Then i tried googling for other similar issues and found this and when following this answers instructions, adding the lines to the modeling.py after the imports i received the same error.

I didn't find any other possible solutions and i'm unsure as to what i'm doing wrong. Did i add the memory growth lines in the wrong file or did i go about solving the issue completely wrong? Any help is appreciated.

I am running the 3.7 kernel in an virtual environment and the data i am feeding the model is properly formatted. I am using the BERT base uncased model downloaded from this repository.