search

xuekt98 / BBDM

BBDM: Image-to-image Translation with Brownian Bridge Diffusion Models
MIT License
262 stars 27 forks source link

Train problem #4

Open zhangbaijin opened 1 year ago

zhangbaijin commented 1 year ago

Thanks for your code, i am trying train the code, and my dataset is ./data/Semi/train/, ./data/Semi/val, ./data/Semi/test, and i set the config in the config.yaml, there is problem when i run the code: if self.config.model.only_load_latent_mean_std: AttributeError: 'Namespace' object has no attribute 'only_load_latent_mean_std Thanks for your help~

xuekt98 commented 1 year ago

Thank you for your attention. Could you show the information in detail? Better using screenshot.

hanzh346 commented 1 year ago

Hello,

I can probably compensate a bit here as I encountered the same issue:

create output path results/low_dose_PET/BrownianBridge/ Total Number of parameter: 237.09M Trainable Number of parameter: 237.09M Traceback (most recent call last): File "/home/hang/GitHub/BBDM/main.py", line 130, in main() File "/home/hang/GitHub/BBDM/main.py", line 125, in main CPU_singleGPU_launcher(nconfig) File "/home/hang/GitHub/BBDM/main.py", line 91, in CPU_singleGPU_launcher runner = get_runner(config.runner, config) File "/home/hang/GitHub/BBDM/utils.py", line 45, in get_runner runner = Registers.runnersrunner_name File "/home/hang/GitHub/BBDM/runners/DiffusionBasedModelRunners/BBDMRunner.py", line 19, in init super().init(config) File "/home/hang/GitHub/BBDM/runners/DiffusionBasedModelRunners/DiffusionBaseRunner.py", line 11, in init super().init(config) File "/home/hang/GitHub/BBDM/runners/BaseRunner.py", line 66, in init self.load_model_from_checkpoint() File "/home/hang/GitHub/BBDM/runners/DiffusionBasedModelRunners/BBDMRunner.py", line 33, in load_model_from_checkpoint if self.config.model.only_load_latent_mean_std: AttributeError: 'Namespace' object has no attribute 'only_load_latent_mean_std'

This happened when I run python3 main.py --config configs/Template-BBDM.yaml --train --gpu_ids 0

Thank you for your attention. Could you show the information in detail? Better using screenshot.

hanzh346 commented 1 year ago

Hello,

I can probably compensate a bit here as I encountered the same issue:

create output path results/low_dose_PET/BrownianBridge/ Total Number of parameter: 237.09M Trainable Number of parameter: 237.09M Traceback (most recent call last): File "/home/hang/GitHub/BBDM/main.py", line 130, in main() File "/home/hang/GitHub/BBDM/main.py", line 125, in main CPU_singleGPU_launcher(nconfig) File "/home/hang/GitHub/BBDM/main.py", line 91, in CPU_singleGPU_launcher runner = get_runner(config.runner, config) File "/home/hang/GitHub/BBDM/utils.py", line 45, in get_runner runner = Registers.runnersrunner_name File "/home/hang/GitHub/BBDM/runners/DiffusionBasedModelRunners/BBDMRunner.py", line 19, in init super().init(config) File "/home/hang/GitHub/BBDM/runners/DiffusionBasedModelRunners/DiffusionBaseRunner.py", line 11, in init super().init(config) File "/home/hang/GitHub/BBDM/runners/BaseRunner.py", line 66, in init self.load_model_from_checkpoint() File "/home/hang/GitHub/BBDM/runners/DiffusionBasedModelRunners/BBDMRunner.py", line 33, in load_model_from_checkpoint if self.config.model.only_load_latent_mean_std: AttributeError: 'Namespace' object has no attribute 'only_load_latent_mean_std'

This happened when I run python3 main.py --config configs/Template-BBDM.yaml --train --gpu_ids 0

Thank you for your attention. Could you show the information in detail? Better using screenshot.

I overcome this issue by adding:
latent_before_quant_conv: False normalize_latent: False only_load_latent_mean_std: False into the Template-BBDM.yaml file, under model section.

I think it is just in the old Template-BBDM.yaml, those parameters are not specified

zhangbaijin commented 1 year ago

Thanks a lot~

yunfei920406 commented 1 year ago

Hello, I can probably compensate a bit here as I encountered the same issue: create output path results/low_dose_PET/BrownianBridge/ Total Number of parameter: 237.09M Trainable Number of parameter: 237.09M Traceback (most recent call last): File "/home/hang/GitHub/BBDM/main.py", line 130, in main() File "/home/hang/GitHub/BBDM/main.py", line 125, in main CPU_singleGPU_launcher(nconfig) File "/home/hang/GitHub/BBDM/main.py", line 91, in CPU_singleGPU_launcher runner = get_runner(config.runner, config) File "/home/hang/GitHub/BBDM/utils.py", line 45, in get_runner runner = Registers.runnersrunner_name File "/home/hang/GitHub/BBDM/runners/DiffusionBasedModelRunners/BBDMRunner.py", line 19, in init super().init(config) File "/home/hang/GitHub/BBDM/runners/DiffusionBasedModelRunners/DiffusionBaseRunner.py", line 11, in init super().init(config) File "/home/hang/GitHub/BBDM/runners/BaseRunner.py", line 66, in init self.load_model_from_checkpoint() File "/home/hang/GitHub/BBDM/runners/DiffusionBasedModelRunners/BBDMRunner.py", line 33, in load_model_from_checkpoint if self.config.model.only_load_latent_mean_std: AttributeError: 'Namespace' object has no attribute 'only_load_latent_mean_std' This happened when I run python3 main.py --config configs/Template-BBDM.yaml --train --gpu_ids 0

Thank you for your attention. Could you show the information in detail? Better using screenshot.

I overcome this issue by adding: latent_before_quant_conv: False normalize_latent: False only_load_latent_mean_std: False into the Template-BBDM.yaml file, under model section.

I think it is just in the old Template-BBDM.yaml, those parameters are not specified

I Met some errors! How did you define the "resume_model "in the *****.sh file for train if you just use your own dataset

xuekt98 commented 1 year ago

Hello, I can probably compensate a bit here as I encountered the same issue: create output path results/low_dose_PET/BrownianBridge/ Total Number of parameter: 237.09M Trainable Number of parameter: 237.09M Traceback (most recent call last): File "/home/hang/GitHub/BBDM/main.py", line 130, in main() File "/home/hang/GitHub/BBDM/main.py", line 125, in main CPU_singleGPU_launcher(nconfig) File "/home/hang/GitHub/BBDM/main.py", line 91, in CPU_singleGPU_launcher runner = get_runner(config.runner, config) File "/home/hang/GitHub/BBDM/utils.py", line 45, in get_runner runner = Registers.runnersrunner_name File "/home/hang/GitHub/BBDM/runners/DiffusionBasedModelRunners/BBDMRunner.py", line 19, in init super().init(config) File "/home/hang/GitHub/BBDM/runners/DiffusionBasedModelRunners/DiffusionBaseRunner.py", line 11, in init super().init(config) File "/home/hang/GitHub/BBDM/runners/BaseRunner.py", line 66, in init self.load_model_from_checkpoint() File "/home/hang/GitHub/BBDM/runners/DiffusionBasedModelRunners/BBDMRunner.py", line 33, in load_model_from_checkpoint if self.config.model.only_load_latent_mean_std: AttributeError: 'Namespace' object has no attribute 'only_load_latent_mean_std' This happened when I run python3 main.py --config configs/Template-BBDM.yaml --train --gpu_ids 0

Thank you for your attention. Could you show the information in detail? Better using screenshot.

I overcome this issue by adding: latent_before_quant_conv: False normalize_latent: False only_load_latent_mean_std: False into the Template-BBDM.yaml file, under model section. I think it is just in the old Template-BBDM.yaml, those parameters are not specified

I Met some errors! How did you define the "resume_model "in the *****.sh file for train if you just use your own dataset

just leave this option as default