Francesco

when using likwid, the compilation flag `-DLIKWID_PERFMON` and the include directories are not forwarded correctly in the process of creating the `special-core` executable. Therefore, likwid is not able to properly...

WIP: talian translation: Chapter 3

6

Italian translation of chapter 3, as per #45

Import Error: File is not valid

2

Hello, I am getting a `File is not valid` error when I try to import some plugins. By inspecting with a debugger, I have pinned down the issue to the...

Node out of memory because of testing batch size

10

Our current implementation for both [torch](https://github.com/fedbiomed/fedbiomed/blob/6c568b892e8222e04d15c23e39a331ab028cc2f9/fedbiomed/common/data/_torch_data_manager.py#L128) and [sklearn](https://github.com/fedbiomed/fedbiomed/blob/6c568b892e8222e04d15c23e39a331ab028cc2f9/fedbiomed/common/data/_sklearn_data_manager.py#L356) sets the batch size of the testing loader to be equal to the full testing dataset. This leads to an out-of-memory problem...

Implement POC for Federated Preprocessing Use Case 1

4

Implement POC based on document in #800 - [x] Implement code infrastructure for federated queries - [x] Implement mean on `MedicalFolderDataset` and `TabularDataset` - [x] Implement std on `MedicalFolderDataset` and...

As reported by @mvesin, the following command should fail ```bash EDITOR=emacs ${FEDBIOMED_DIR}/scripts/fedbiomed_run node --config config-n1.ini training-plan view ```

Potentially dead code for secure aggregation in Aggregator class

2

Our base `Aggregator` class has a [`secure_aggregation`](https://github.com/fedbiomed/fedbiomed/blob/fbbcd6c16963d0a5ea7af26450db1cfd361f8230/fedbiomed/researcher/aggregators/aggregator.py#L34) method and a [`_secagg_crypter`](https://github.com/fedbiomed/fedbiomed/blob/fbbcd6c16963d0a5ea7af26450db1cfd361f8230/fedbiomed/researcher/aggregators/aggregator.py#L32) variable. To my knowledge, neither of those are never used, because we use instead the `aggregate` method of the...

Allow specifying number of samples for validation

1

As a researcher, I would like to specify the number of sample for validation instead of a ratio. This is required for two reasons, both related to the use case...

The error happens while trying to convert the torch-based`TorchDataManager` to sklearn. Specifically, it happens in the `TorchDataManager.to_sklearn()` function at this point: ```python inputs = next(iter(loader))[0].numpy() ``` The reason is that,...

Slow loading of large csv file leads to crash of node

2

Issue originally reported on Discord [support channel](https://discord.com/channels/927917604188852235/927919698862047292/1168869258999119904) > Hello, I am facing some issues when loading large CSV files from CLI. The procedure never finishes, and the dataset is never...