https://github.com/humanpose1/automatic_flashcard_for_arabic
A simple python script using langchain + genanki to create flashcard to learn arabic
https://github.com/humanpose1/automatic_flashcard_for_arabic
Last synced: 11 months ago
JSON representation
A simple python script using langchain + genanki to create flashcard to learn arabic
- Host: GitHub
- URL: https://github.com/humanpose1/automatic_flashcard_for_arabic
- Owner: humanpose1
- License: mit
- Created: 2025-04-04T21:29:16.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2025-04-04T22:31:37.000Z (about 1 year ago)
- Last Synced: 2025-04-12T11:51:43.505Z (about 1 year ago)
- Language: Python
- Size: 215 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Create flashcards automatically with AI.
We use uv to manage dependancy.
use the right version of python
```bash
uv venv --python 3.13
```
We installed everything like this:
```bash
uv add ipykernel ipython torch torchvision numpy matplotlib playwright langchain-core langchain-groq langchain beautifulsoup4 genanki pydub transformers nest_asyncio
```
The `uv.lock` and `pyproject.toml` is here you can download as:
```bash
uv sync --locked
```
install firefox and chrome with playwright
```bash
uv run playwright install chrome firefox
```
We scrape media arabic sentence from [aljazeera learning platform](https://learning.aljazeera.net/en):
```bash
uv run python scrape_arabic_text.py -o out/article_all.json
```
Then to compute flashcard using llm you can use:
```bash
uv run python automatic_translation_all.py -i out/article_all.json -o out/article_all_with_llm.json
```
The model runs locally using QWen2.5 72B. If you want a faster model you can use Groq (by default it uses mistral-saba-24b)
You can find an diagram of the workflow to create the flashcard. It is a very simple workflow.
```bash
uv run python automatic_translation_all.py -i out/article_all.json -o out/article_all_with_llm.json --use-groq
```
Do not forget to export
```bash
export GROQ_API_KEY=XXX
```
Then to create flashcard: use:
```bash
uv run python create_anki_flashcard.py -i out/articles_all_with_llm.json -o out/test.apkg --title FlashCard_Aljazeera_Learning
```