error while reproducing #3
Assignees
Comments
|
Hi, yes we are aware of this error. The eval script currently does not work as we are updating the code to improve usability. |
|
hi, can you give update on this, i just want to run for super resolution. |
|
I'm going to work on this specific issue next week. I will close the issue whenever the problem is fixed. |
|
Ok sure, and can you tell me that what would be the changes in the code if I want to run only for super resolution ?
… On 25 Oct 2024, at 1:22 PM, Gabriel della Maggiora ***@***.***> wrote:
I'm going to work on this specific issue next week. I will close the issue whenever the problem is fixed.
—
Reply to this email directly, view it on GitHub <#3 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BBQXTAIQZO4ABYRDABGL2XDZ5H2E3AVCNFSM6AAAAABQKR5MN2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDIMZXGEZTAMRXGI>.
You are receiving this because you authored the thread.
|
|
I'm assuming you downloaded BioSR. What needs to be done is first fix the data loader (you can change this in the train.py script and load the data there, skipping the data loader). For this experiment, the parameter diff_inp needs to be set to true. And for the evaluation (the eval script) needs to be experiment-specific as it is currently set up for an experiment that is not related to super-resolution. The module issue is a pathing issue that might be fixed by changing the imports. Again, I will fix this next week, as I don't have the time for it today. Please be patient. |
|
The issue should be fixed now. Please read the updated readme and prepare the environment accordingly. Download the BioSR data first, update the config file, train the model using the |
|
I have edited the config file but now it's coming module not found for
tensorflow.keras
…On Mon, Oct 28, 2024, 6:32 PM Gabriel della Maggiora < ***@***.***> wrote:
The issue should be fixed now. Please read the updated readme and prepare
the environment accordingly. Download the BioSR data first, update the
config file, train the model using the train script, and use the eval
script over the test set to obtain the results.
—
Reply to this email directly, view it on GitHub
<#3 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/BBQXTAMU6UU3EQGHSRD5223Z5YRULAVCNFSM6AAAAABQKR5MN2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDINBRGUZDQMJQGA>
.
You are receiving this because you authored the thread.Message ID:
***@***.***>
|
|
Then, you are missing the tensorflow module. Please see the updated instructions. |
|
Maybe the tensorflow 2.14.0 is set in the docker but as 2.14 is not
currently available so tensorflow addons is not working too as it's
integrated upto 2.14 version
…On Tue, Oct 29, 2024, 2:14 PM Gabriel della Maggiora < ***@***.***> wrote:
Then, you are missing the tensorflow module. Please see the updated
instructions.
—
Reply to this email directly, view it on GitHub
<#3 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/BBQXTAOELIJOU5W5Z5WBHA3Z55DHRAVCNFSM6AAAAABQKR5MN2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDINBTGU4DQMRVGQ>
.
You are receiving this because you authored the thread.Message ID:
***@***.***>
|
|
Hlw, sorry to disturb you but can you once look into this
On Tue, Oct 29, 2024, 5:03 PM Avinash Gohite ***@***.***>
wrote:
… Maybe the tensorflow 2.14.0 is set in the docker but as 2.14 is not
currently available so tensorflow addons is not working too as it's
integrated upto 2.14 version
On Tue, Oct 29, 2024, 2:14 PM Gabriel della Maggiora <
***@***.***> wrote:
> Then, you are missing the tensorflow module. Please see the updated
> instructions.
>
> —
> Reply to this email directly, view it on GitHub
> <#3 (comment)>, or
> unsubscribe
> <https://github.com/notifications/unsubscribe-auth/BBQXTAOELIJOU5W5Z5WBHA3Z55DHRAVCNFSM6AAAAABQKR5MN2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDINBTGU4DQMRVGQ>
> .
> You are receiving this because you authored the thread.Message ID:
> ***@***.***>
>
|
|
Hi, it was just a dependency issue. I updated the requirements file. It should be fixed now. |
|
Hlw can you tell me how much time it takes to load this step cat bio_sr_part* > biosr.npz. I have downloaded the dataset and now concatenating this since last night , still processing
… On 1 Nov 2024, at 3:15 PM, Gabriel della Maggiora ***@***.***> wrote:
Hi, it was just a dependency issue. I updated the requirements file. It should be fixed now.
—
Reply to this email directly, view it on GitHub <#3 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BBQXTAPH6BH57QB74MW6PZDZ6NETHAVCNFSM6AAAAABQKR5MN2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDINJRGYYDGNBXHA>.
You are receiving this because you authored the thread.
|
|
Hi, finally I loaded the dataset properly, and modified the config file and then created environment, installed requirements but still this is not running
python train.py --config-path configs/biosr.yaml bcoz of dependencies issue ie, Neptune, then if I m installing Neptune then it is saying about tensor flow again ……..I m frustrated here , 😫 brother plz do run once and tell me.
… On 4 Nov 2024, at 11:10 PM, Avinash Gohite ***@***.***> wrote:
On Mon, Nov 4, 2024, 3:24 PM Avinash Gohite ***@***.*** ***@***.***>> wrote:
> Hlw can you tell me how much time it takes to load this step cat bio_sr_part* > biosr.npz. I have downloaded the dataset and now concatenating this since last night , still processing
>
>
>> On 1 Nov 2024, at 3:15 PM, Gabriel della Maggiora ***@***.*** ***@***.***>> wrote:
>>
>>
>> Hi, it was just a dependency issue. I updated the requirements file. It should be fixed now.
>>
>> —
>> Reply to this email directly, view it on GitHub <#3 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BBQXTAPH6BH57QB74MW6PZDZ6NETHAVCNFSM6AAAAABQKR5MN2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDINJRGYYDGNBXHA>.
>> You are receiving this because you authored the thread.
>>
>
|
|
I don't know what could be wrong I installed the environment from scratch and it worked for me :( |
|
Hlw,
Finally I m doing it in GitHub codespace and the requirements.txt file ran successfully, now for the bios.yaml I m running the training.py but it seems to the error as follow
(cvdm_env) @insanemate033-gif ➜ /workspaces/cvdm (main) $ python train.py --config-path configs/biosr.yaml
2024-12-12 06:45:00.228431: I tensorflow/core/util/port.cc:111] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
2024-12-12 06:45:00.466731: E tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:9342] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
2024-12-12 06:45:00.467014: E tensorflow/compiler/xla/stream_executor/cuda/cuda_fft.cc:609] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
2024-12-12 06:45:00.467649: E tensorflow/compiler/xla/stream_executor/cuda/cuda_blas.cc:1518] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
2024-12-12 06:45:00.531448: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
/opt/conda/envs/cvdm_env/lib/python3.10/site-packages/tensorflow_addons/utils/tfa_eol_msg.py:23: UserWarning:
TensorFlow Addons (TFA) has ended development and introduction of new features.
TFA has entered a minimal maintenance and release mode until a planned end of life in May 2024.
Please modify downstream libraries to take dependencies from other repositories in our TensorFlow community (e.g. Keras, Keras-CV, and Keras-NLP).
For more information see: tensorflow/addons#2807
warnings.warn(
Traceback (most recent call last):
File "/workspaces/cvdm/train.py", line 17, in <module>
from cvdm.diffusion_models.joint_model import instantiate_cvdm
File "/workspaces/cvdm/cvdm/diffusion_models/joint_model.py", line 13, in <module>
from cvdm.utils.training_utils import taylor_expand_gamma, time_grad
File "/workspaces/cvdm/cvdm/utils/training_utils.py", line 10, in <module>
from cvdm.data.biosr_dataloader import BioSRDataloader
ModuleNotFoundError: No module named ‘cvdm.data.biosr_dataloader'
I want to run this , can you look it into through my system I will give you access through any desk , plz help me brother.
… On 5 Nov 2024, at 3:24 AM, Gabriel della Maggiora ***@***.***> wrote:
Closed #3 <#3> as completed.
—
Reply to this email directly, view it on GitHub <#3 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BBQXTAPCFIQG4J4PK5VYV3LZ67UJZAVCNFSM6AAAAABQKR5MN2VHI2DSMVQWIX3LMV45UABCJFZXG5LFIV3GK3TUJZXXI2LGNFRWC5DJN5XDWMJVGEYTAOBXGMZTKNY>.
You are receiving this because you authored the thread.
|
|
Hlw,
Finally I m doing it in GitHub codespace and the requirements.txt file ran successfully, now for the bios.yaml I m running the training.py but it seems to the error as follow
(cvdm_env) @insanemate033-gif ➜ /workspaces/cvdm (main) $ python train.py --config-path configs/biosr.yaml
2024-12-12 06:45:00.228431: I tensorflow/core/util/port.cc:111 <http://port.cc:111/>] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
2024-12-12 06:45:00.466731: E tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:9342 <http://cuda_dnn.cc:9342/>] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
2024-12-12 06:45:00.467014: E tensorflow/compiler/xla/stream_executor/cuda/cuda_fft.cc:609 <http://cuda_fft.cc:609/>] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
2024-12-12 06:45:00.467649: E tensorflow/compiler/xla/stream_executor/cuda/cuda_blas.cc:1518 <http://cuda_blas.cc:1518/>] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
2024-12-12 06:45:00.531448: I tensorflow/core/platform/cpu_feature_guard.cc:182 <http://cpu_feature_guard.cc:182/>] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
/opt/conda/envs/cvdm_env/lib/python3.10/site-packages/tensorflow_addons/utils/tfa_eol_msg.py:23: UserWarning:
TensorFlow Addons (TFA) has ended development and introduction of new features.
TFA has entered a minimal maintenance and release mode until a planned end of life in May 2024.
Please modify downstream libraries to take dependencies from other repositories in our TensorFlow community (e.g. Keras, Keras-CV, and Keras-NLP).
For more information see: tensorflow/addons#2807
warnings.warn(
Traceback (most recent call last):
File "/workspaces/cvdm/train.py", line 17, in <module>
from cvdm.diffusion_models.joint_model import instantiate_cvdm
File "/workspaces/cvdm/cvdm/diffusion_models/joint_model.py", line 13, in <module>
from cvdm.utils.training_utils import taylor_expand_gamma, time_grad
File "/workspaces/cvdm/cvdm/utils/training_utils.py", line 10, in <module>
from cvdm.data.biosr_dataloader import BioSRDataloader
ModuleNotFoundError: No module named ‘cvdm.data.biosr_dataloader’
I want to run this , can you look it into through my system I will give you access through any desk , plz help me brother. Just go through it once. When you are free
… On 5 Nov 2024, at 3:24 AM, Gabriel della Maggiora ***@***.***> wrote:
Closed #3 <#3> as completed.
—
Reply to this email directly, view it on GitHub <#3 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BBQXTAPCFIQG4J4PK5VYV3LZ67UJZAVCNFSM6AAAAABQKR5MN2VHI2DSMVQWIX3LMV45UABCJFZXG5LFIV3GK3TUJZXXI2LGNFRWC5DJN5XDWMJVGEYTAOBXGMZTKNY>.
You are receiving this because you authored the thread.
|
|
Hlw
On Thu, Dec 12, 2024, 12:36 PM Avinash Gohite ***@***.***>
wrote:
… Hlw,
Finally I m doing it in GitHub codespace and the requirements.txt file ran
successfully, now for the bios.yaml I m running the training.py but it
seems to the error as follow
(cvdm_env) @insanemate033-gif ➜ /workspaces/cvdm (main) $ python train.py
--config-path configs/biosr.yaml
2024-12-12 06:45:00.228431: I tensorflow/core/util/port.cc:111] oneDNN
custom operations are on. You may see slightly different numerical results
due to floating-point round-off errors from different computation orders.
To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
2024-12-12 06:45:00.466731: E tensorflow/compiler/xla/stream_executor/cuda/
cuda_dnn.cc:9342] Unable to register cuDNN factory: Attempting to
register factory for plugin cuDNN when one has already been registered
2024-12-12 06:45:00.467014: E tensorflow/compiler/xla/stream_executor/cuda/
cuda_fft.cc:609] Unable to register cuFFT factory: Attempting to register
factory for plugin cuFFT when one has already been registered
2024-12-12 06:45:00.467649: E tensorflow/compiler/xla/stream_executor/cuda/
cuda_blas.cc:1518] Unable to register cuBLAS factory: Attempting to
register factory for plugin cuBLAS when one has already been registered
2024-12-12 06:45:00.531448: I tensorflow/core/platform/
cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use
available CPU instructions in performance-critical operations.
To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in
other operations, rebuild TensorFlow with the appropriate compiler flags.
/opt/conda/envs/cvdm_env/lib/python3.10/site-packages/tensorflow_addons/utils/tfa_eol_msg.py:23:
UserWarning:
TensorFlow Addons (TFA) has ended development and introduction of new
features.
TFA has entered a minimal maintenance and release mode until a planned end
of life in May 2024.
Please modify downstream libraries to take dependencies from other
repositories in our TensorFlow community (e.g. Keras, Keras-CV, and
Keras-NLP).
For more information see: tensorflow/addons#2807
warnings.warn(
Traceback (most recent call last):
File "/workspaces/cvdm/train.py", line 17, in <module>
from cvdm.diffusion_models.joint_model import instantiate_cvdm
File "/workspaces/cvdm/cvdm/diffusion_models/joint_model.py", line 13,
in <module>
from cvdm.utils.training_utils import taylor_expand_gamma, time_grad
File "/workspaces/cvdm/cvdm/utils/training_utils.py", line 10, in
<module>
from cvdm.data.biosr_dataloader import BioSRDataloader
ModuleNotFoundError: No module named ‘cvdm.data.biosr_dataloader’
I want to run this , can you look it into through my system I will give
you access through any desk , plz help me brother. Just go through it once.
When you are free
On 5 Nov 2024, at 3:24 AM, Gabriel della Maggiora <
***@***.***> wrote:
Closed #3 <#3> as completed.
—
Reply to this email directly, view it on GitHub
<#3 (comment)>, or unsubscribe
<https://github.com/notifications/unsubscribe-auth/BBQXTAPCFIQG4J4PK5VYV3LZ67UJZAVCNFSM6AAAAABQKR5MN2VHI2DSMVQWIX3LMV45UABCJFZXG5LFIV3GK3TUJZXXI2LGNFRWC5DJN5XDWMJVGEYTAOBXGMZTKNY>
.
You are receiving this because you authored the thread.Message ID:
***@***.***>
|
|
Sorry, I can't help you here because what you are having is a pathing issue. Check if you updated the code to the last version. If you did the class that python is missing should be in that folder, what you need to do is fix the pathing of the project such that python finds the missing class. |
|
Actually I m only focusing on the application part I want the results from
my datasets, that is why I want to reproduce this, i am not able to update
the code as there is no folder biosr_dataloader. Can you just look it on my
system through any desk so it would work
…On Thu, Dec 19, 2024, 5:21 PM Gabriel della Maggiora < ***@***.***> wrote:
Sorry, I can't help you here because what you are having is a pathing
issue. Check if you updated the code to the last version. If you did the
class that python is missing should be in that folder, what you need to do
is fix the pathing of the project such that python finds the missing class.
—
Reply to this email directly, view it on GitHub
<#3 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/BBQXTAM4MEO4H22DDRUZWHL2GKXKVAVCNFSM6AAAAABQKR5MN2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKNJTGUZDENJXGI>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
Hi @insanemate033-gif we thank you for your interest, however what you are requesting goes beyond the scope of the open source code license. Furthermore, I think you are misusing the issues functionality here. What you are raising is not an "issue" as you are not providing any steps to reproduce it. Unless you can provide a concrete description and steps to reproduce an issue, this thread should be closed. Requests to adopt the codebase to your data cannot constitute an issue. Unless you follow the policy I'm going to report your activity. |
|
Hi @insanemate033-gif. I provided the dataloader and added the pathing for you. I tested the code in a new environment, and it trained as expected (with the biosr.npz). Please see if this fixes your issue. If you want to train with your own data, you will have to create the dataloader yourself. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
module not found error for cvdm
The text was updated successfully, but these errors were encountered: