can anyone provide pip list with versions. I am having a lot of isssues
If you specify a bit more we might be able to help.
What problem do you have?
I imagine you mean for facebook mmf. You have here the requirements .
Also check installation instructions.
They recommend python 3.7+ and pytorch 1.6.
I am using this code
https://www.drivendata.co/blog/hateful-memes-benchmark/
it keeps throwing
this error
init() got an unexpected keyword argument ‘default_save_path’
That’s an example notebook from the12th of may, so yes maybe you are using a different version of the pytorch-lightning library, which released a new version with lots of changes recently.
If I am not mistaken that error is thrown in the Trainer initialization.
In the dict called trainer_params try to change the parameter “default_save_path” to “default_root_dir”.
You can also try to install a previous pytorch-lightning version.
You can check the parameters of the Trainer class here.
I am getting this error can you please help me out
EarlyStopping mode auto is unknown, fallback to auto mode.
INFO:lightning:EarlyStopping mode auto is unknown, fallback to auto mode.
EarlyStopping mode set to min for monitoring avg_val_loss.
INFO:lightning:EarlyStopping mode set to min for monitoring avg_val_loss.
GPU available: False, used: False
INFO:lightning:GPU available: False, used: False
TPU available: False, using: 0 TPU cores
INFO:lightning:TPU available: False, using: 0 TPU cores
| Name | Type | Params
0 | model | LanguageAndVisionConcat | 58 M
INFO:lightning:
| Name | Type | Params
0 | model | LanguageAndVisionConcat | 58 M
Validation sanity check:
0/? [00:00<?, ?it/s]
@Sahithi23 You need to care more the posts, so we can help better.
In what cell of that notebook you have the error? You should quote that code and the error output in good format.
I imagine you have solved the previous error and this error is happening during fit call.
You should adapt some hyperparameters to your execution settings.
If I am not mistaken you are not using GPU or TPU, so running on CPU?!
This could take a very loooong time.
I do not see the error in what you posted, I just see pytorch-lightning doing the validation sanity check, even this (that is just a few steps) could take very long on CPU.
I recommend you to go to Google Colab 
can you suggest me how to stop performing validation sanity check and get into execution
Add this parameter to Trainer on init:
num_sanity_val_steps = 0
You can see this on docs.
not working