Pretrained Model

[rayush7]

Hey @dtransposed

Thanks for the code. Could you please share the pretrained model for this code? I don't have the resources to train it myself.

[dtransposed]

Hello!

First of all, please specify which code you have in mind. There are several research papers implemented, you have not specified which you have on your mind.
Secondly, I do not keep the pre-trained model forever. Additionally, I am rather reluctant to retraining the model once I finish it and post the code here. There are many reasons, e.g. I usually need to download dataset once again and use up my own computing power and time. So basically I go through the same process as you do.

Please tell me which code you are interested in and we can think of a solution of how to simplify the problem so that your machine can handle it. Sharing the specs would also help.

[rayush7]

Hi @dtransposed

I am sorry for carelessly posting my issue without any particular details.

I am referring to the implementation of the paper on "Action Recognition using Visual Attention". I understand that keeping the pretrained model is not viable. Just reaching out to you if you can help me to run the code on my machine.

The Machine I am using has the following specs:

1. Intel Core i7 Six Core Processor 8750H (2.2GHz, 4.1GHz Turbo)
2. 32GB Corsair 2400MHz SODIMM DDR4 (2 x 16GB)
3. NVIDIA GeForce GTX 1060 - 6.0GB GDDR5 Video RAM

I am trying to run both ALSTM and ConvALSTM model but I am getting a out of memory error. Even trying smaller batch size doesn't help. Please suggest me if you have any solution in mind.

Thanks