snel-repo / autolfads

Generalized modeling of single-trial neural population dynamics.

autolfads

Generalized modeling of single-trial neural population dynamics.

AutoLFADS code is kept in the GCP branch

RADICaL code is kept in the GCP-radical branch

Tutorial content is managed in the tutorial branch and accessible at https://snel-repo.github.io/autolfads/

autolfads's Issues

Missing Data File

The following datasets is missing from the tutorial_package.zip

• Synthetic 2-photon calcium imaging dataset | calcium/lfads_cal_data.h5
• Real EMG dataset | emg/lfads_emg_input.h5

Running the code on local servers

Hi there,

Thanks for sharing this amazing tool. I was wondering if you are going to publish the code that could be run on local GPU servers?

Thanks,
Mohsen

code on customer GPU cluster available?

Hi the authoers,

Thanks for your contribution.
I'm excited to apply this tool to calcium signals. However, I have to use the custom GPU in our lab. Do you have a version of code for that case?
Thank you!

Best,
Mango

Error running Radical on GCP

Hello, I am having a trouble running radical on GCP. I reach the point on the tutorial where I start the radical run but I get this error always.
`aalothman@tutserver:~/autolfads/pbt_opt$ python2 pbt_script_multiVM.py

Traceback (most recent call last):

File "pbt_script_multiVM.py", line 25, in
helper = pbtHelper(bucket_name, data_path, run_path, name, nprocess_gpu)
File "/home/aalothma/autolfads/pbt_opt/pbt_helper_fn.py", line 19, in init
self.server_id = os.environ["HOSTNAME"]
File "/usr/lib/python2.7/UserDict.py", line 40, in getitem
raise KeyError(key)
KeyError: 'HOSTNAME'`

Things I have done:
1- checked that docker is installed correctly on clients
2- deleted client and server three times and created new one. I still get the same error each run.

Things to note:
I only have one client, as opposed to the recommended number in the tutorial.

I appreciate any help that can be provided.

Thank you so much.