You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
> This template contains some useful structure and convention for new research
4
-
> projects that will help you get started more quickly, and will make your
5
-
> code more accessible, maintainable, and reproducible. This will make your
6
-
> work more likely to be adopted by others!
7
-
>
8
-
> I highly recommend taking a second to read Patrick Mineault's
9
-
> [Good Research Code Handbook](https://goodresearch.dev/index.html) if you are
10
-
> in the process of starting a new project for many tips that will help you
11
-
> beyond the initial setup phase.
12
-
>
13
-
> > *Note:* You should delete everything within markdown blockquotes `>` before
14
-
> going live with your project.
15
-
>
16
-
> `### DELETE ME >>>`
17
-
18
-
19
-
<divalign="center">
20
1
21
2
22
3
<!-- TITLE -->
23
-
# Project Template `> REPLACE ME`
24
-
A new project template for research projects. `> REPLACE ME`
25
-
26
-
<!-- BADGES -->
27
-
> <divalign="left">
28
-
>
29
-
> `### <<< DELETE ME`***Badges****(optional)*
30
-
>
31
-
> If you have an arXiv paper, you can add and update the `[arXiv]` badge by
32
-
> replacing `1234.56789` with the arXiv ID of your paper and the arXiv
33
-
> subject `cs.LG` with the main subject. Else, delete it (or comment out).
34
-
>
35
-
> if your paper is published at a conference, you can add and update the
36
-
> `[Conference]` badge by replacing `Conference`, `Year`, and the link fields.
37
-
> Else, delete it (or comment out).
38
-
>
39
-
> It is also useful to add a CI build/test [status badge](https://docs.github.com/en/actions/monitoring-and-troubleshooting-workflows/adding-a-workflow-status-badge)
40
-
> to your project. A base Continuous Integration pipeline has been defined in
# Positional Encoding Benchmark for Time Series Classification
5
+
6
+
This repository provides a comprehensive benchmark for evaluating different positional encoding techniques in Transformer models, specifically for time series classification tasks. The project includes implementations of several positional encoding methods and Transformer architectures to test their effectiveness on various time series datasets.
7
+
58
8
59
9
60
10
<!-- DESCRIPTION -->
61
11
## Description
62
-
> `### <<< DELETE ME:`***Description***
63
-
>
64
-
> Fill in a succinct description of your project.
65
-
>
66
-
> `### DELETE ME >>>`
67
-
68
-
69
-
<!-- SETUP -->
70
-
## Setup
71
-
72
-
> `### <<< DELETE ME:`***Setup***
73
-
>
74
-
> Below are some base instructions for setting up a conda environment. See this
> Add any instructions necessary to setup the project. The best time to create
79
-
> this is ***as you are developing*** the project, while you remember the steps
80
-
> you have taken.
81
-
>
82
-
> *Brownie points*: try to follow your setup instructions to replicate the setup
83
-
> from scratch on another machine to ensure that it is truly reproducible.
84
-
>
85
-
> `### DELETE ME >>>`
86
-
87
-
88
-
### Conda Virtual Environment
89
-
90
-
Create the Conda virtual environment using the [environment file](environment.yml):
91
-
```bash
92
-
conda env create -f environment.yml
93
12
94
-
# dynamically set python path for the environment
95
-
conda activate YOUR_PROJECT
96
-
conda env config vars set PYTHONPATH=$(pwd):$(pwd)/src
97
-
```
13
+
14
+
This project aims to analyze how positional encodings impact Transformer-based models in time series classification. The benchmark includes both fixed and learnable encoding techniques and explores advanced approaches like relative positional encoding. The project evaluates performance on a diverse set of datasets from different domains, such as human activity recognition, financial data, EEG recordings, etc.
98
15
99
16
100
-
<!-- USAGE -->
101
-
## Usage
102
-
> `### <<< DELETE ME:`***Usage***
103
-
>
104
-
> Provide information on how to run your project. Delete the below example.
@@ -126,23 +57,6 @@ Please make sure to update tests as appropriate.
126
57
<!-- CITATION -->
127
58
## Citation
128
59
129
-
> `### <<< DELETE ME:`***Citation***
130
-
>
131
-
> Adding a citation to your README will make it easier for others to cite your
132
-
> work. Add your bibtext citation to the README below. GitHub also will
133
-
> automatically detect [Citation File Format (`.cff`) files](https://citation-file-format.github.io/),
134
-
> rendering a "Cite this repository" button. See [GitHub's tutorial](https://docs.github.com/en/repositories/managing-your-repositorys-settings-and-features/customizing-your-repository/about-citation-files)
135
-
> for more information. The example from this tutorial is included in
136
-
> [CITATION.cff](CITATION.cff), and should be modified or deleted.
0 commit comments