File size: 3,057 Bytes
23888ab 877be9d 23888ab 80f1681 ebcf22e a03bcff 8b2754f 215291a fa081cd a03bcff eb5f95c 215291a eb5f95c 215291a eb5f95c 215291a eb5f95c 215291a eb5f95c 215291a eb5f95c a03bcff 9d38843 171701a e2d70c9 a03bcff 76d6031 86312d8 76d6031 1b8be13 538572f 816c4bb fa081cd 0a81102 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 |
---
title: Summarizer For Learning
emoji: ✍️
colorFrom: yellow
colorTo: yellow
sdk: gradio
sdk_version: 4.8.0
app_file: app.py
pinned: false
---
<h1 style="text-align: center;">✍️ Summarizer for Learning Documentation<h1>
## 1. User Manual
<details>
<summary>1. Insert your <a href="https://huggingface.co/settings/tokens" target="_blank">🤗 Hugging Face Access Token</a> in the textbox located as shown.</summary>
<img src="images/1.png" alt="" width="800">
</details>
<details>
<summary>2. Insert a YouTube link in the textbox located as shown. The YouTube video must be able to have its caption on.</summary>
<img src="images/2.png" alt="" width="800">
- Or you can use the example YouTube links the bottom of the page to try out the app:
- <img src="images/2-1.png" alt="" width="800">
</details>
<details>
<summary>3. Set the Settings according to your need.</summary>
<img src="images/3.png" alt="" width="800">
</details>
<details>
<summary>4. Set the Advanced Settings according to your need. Slide the Temperate or/and tick the `do_sample` checkbox to your need.</summary>
<img src="images/4.png" alt="" width="800">
</details>
<details>
<summary>5. Summarize it by clicking the Summarize button.</summary>
<img src="images/5.png" alt="" width="800">
</details>
<details>
<summary>6. To use the `Key sentence`, `Keywords`, `Fill in the blank`, or `Flashcards` feature, click on the desired tab and click the Generate button after the summarization process is finished.</summary>
<img src="images/6.png" alt="" width="800">
<details>
<summary>The result should similiar like these:</summary>
- <img src="images/6-1.png" alt="" width="800">
- <img src="images/6-2.png" alt="" width="800">
- <img src="images/6-3.png" alt="" width="800">
- <img src="images/6-4.png" alt="" width="800">
</details>
</details>
## 2. How It Works
<p align="center">
<img src="images/hiw.png" alt="" width="1150">
</p>
## 3. Concept
The goal of the app was for learners to **(1)retain** long-text information better, and also make the important points **(2)memorable**. To do that the app takes a long text then summarize it, extract its keywords, and quiz the learner.
## 4. Additional Note
The app is a pet project and was made for me to learn more about summarization in NLP. The summarization methodology was derived from [Amin Kamali's summarizer app](https://huggingface.co/spaces/smakamali/summarize_youtube). I am thankful for Amin for publishing his material online for me to learn from it.
Apart from the that, some of the features in the app are still under development due to some contraints. The LLM used for this summarization is `tiiuae/falcon-7b-instruct` that is primarily intended for a chat-based scenario or for text generation. Because of that, small part of the summarized output may include information not inlucuded in the original text. This problem will be looked into further.
<!---Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference--> |