Update README.md
Browse files
README.md
CHANGED
|
@@ -1,16 +1,63 @@
|
|
| 1 |
---
|
| 2 |
-
title:
|
| 3 |
-
emoji:
|
| 4 |
colorFrom: yellow
|
| 5 |
colorTo: indigo
|
| 6 |
sdk: streamlit
|
| 7 |
-
sdk_version: 1.
|
| 8 |
app_file: app.py
|
| 9 |
pinned: false
|
| 10 |
license: mit
|
| 11 |
-
short_description: Torch Transformers Diffusion SFT
|
| 12 |
---
|
| 13 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 14 |
# TorchTransformers Diffusion CV SFT Titans 🚀
|
| 15 |
|
| 16 |
A Streamlit app blending `torch`, `transformers`, and `diffusers` for vision and NLP fun! Snap PDFs 📄, turn them into double-page spreads 🖼️, extract text with GPT 🤖, and craft emoji-packed Markdown outlines 📝—all with a witty UI and CPU-friendly SFT.
|
|
@@ -64,40 +111,6 @@ Run: `pip install -r requirements.txt`, `streamlit run app.py`. Snap, process, s
|
|
| 64 |
- 🖼️ **Image Process**: OCR images with GPT vision.
|
| 65 |
- 📚 **MD Gallery**: Summarize Markdown files into emoji outlines.
|
| 66 |
|
| 67 |
-
## Tutorial: Single to Double Page Emoji Outlines
|
| 68 |
-
|
| 69 |
-
### Single Page Outline: Key Functions in `app.py`
|
| 70 |
-
|
| 71 |
-
| **Function** | **Purpose** 🎯 | **How It Works** 🛠️ | **Emoji Insight** 😎 |
|
| 72 |
-
|----------------------------|---------------------------------------------|--------------------------------------------------|-------------------------------|
|
| 73 |
-
| `generate_filename` | Unique file names 📅 | Adds timestamp to sequence | 🕰️ Time’s your file buddy! |
|
| 74 |
-
| `pdf_url_to_filename` | Safe PDF names 🖋️ | Cleans URLs to underscores | 🚫 No URL mess! |
|
| 75 |
-
| `get_download_link` | Downloadable files ⬇️ | Base64-encodes for HTML links | 📦 Grab it, go! |
|
| 76 |
-
| `download_pdf` | Web PDF snatcher 🌐 | Fetches PDFs with `requests` | 📚 PDF pirate ahoy! |
|
| 77 |
-
| `process_pdf_snapshot` | PDF to images 🖼️ | Async snapshots (single/double/all) with `fitz` | 📸 Double-page dazzle! |
|
| 78 |
-
| `process_ocr` | Image text extractor 🔍 | Async GOT-OCR2_0 with `transformers` | 👀 Text ninja strikes! |
|
| 79 |
-
| `process_image_gen` | Prompt to image 🎨 | Async Stable Diffusion with `diffusers` | 🖌️ Art from words—bam! |
|
| 80 |
-
| `process_image_with_prompt`| GPT image analysis 🤖 | Base64 to GPT vision | 🧠 GPT sees all! |
|
| 81 |
-
| `process_text_with_prompt` | GPT text summarizer ✍️ | Text to GPT for outlining | 📝 Summarize like a pro! |
|
| 82 |
-
| `update_gallery` | File showcase 🖼️📖 | Sidebar display with delete options | 🌟 Your creations shine! |
|
| 83 |
-
|
| 84 |
-
### Double Page Outline: Libraries in `requirements.txt`
|
| 85 |
-
|
| 86 |
-
| **Library** | **Single Page Purpose** 🎯 | **Double Page Usage** 🛠️ | **Emoji Insight** 😎 |
|
| 87 |
-
|---------------|-------------------------------------------|----------------------------------------------------|-------------------------------|
|
| 88 |
-
| `streamlit` | App UI 🌐 | Tabs like “PDF Process 📄” and “MD Gallery 📚” | 🎬 App star—lights, action! |
|
| 89 |
-
| `pandas` | Data crunching 📈 | Ready for OCR/metadata tables | 📊 Table tamer awaits! |
|
| 90 |
-
| `torch` | ML engine 🔥 | Powers `transformers` and `diffusers` | 🔥 AI’s fiery heart! |
|
| 91 |
-
| `requests` | Web grabber 🌍 | Downloads PDFs in `download_pdf` | 🌐 Web loot collector! |
|
| 92 |
-
| `aiofiles` | Fast file ops ⚡ | Async writes in `process_ocr` | ✈️ File speed demon! |
|
| 93 |
-
| `pillow` | Image magic 🖌️ | PDF to image in `process_pdf_snapshot` | 🖼️ Pixel Picasso! |
|
| 94 |
-
| `PyMuPDF` | PDF handler 📜 | Snapshots in `process_pdf_snapshot` | 📜 PDF scroll master! |
|
| 95 |
-
| `transformers`| AI models 🗣️ | GOT-OCR2_0 in `process_ocr` | 🤖 Brain in a box! |
|
| 96 |
-
| `diffusers` | Image gen 🎨 | Stable Diffusion in `process_image_gen` | 🎨 Art generator supreme! |
|
| 97 |
-
| `openai` | GPT vision/text 🤖 | Image/text processing in GPT functions | 🌌 All-seeing AI oracle! |
|
| 98 |
-
| `glob2` | File finder 🔍 | Gallery files in `update_gallery` | 🕵️ File sleuth! |
|
| 99 |
-
| `pytz` | Time zones ⏰ | Timestamps in `generate_filename` | ⏳ Time wizard! |
|
| 100 |
-
|
| 101 |
## Automation Instructions: Witty & Funny Steps 😂
|
| 102 |
|
| 103 |
1. **Load PDFs** 📚
|
|
@@ -169,4 +182,4 @@ Run: pip install -r requirements.txt, streamlit run app.py. Snap, process, outli
|
|
| 169 |
- View it in a Markdown renderer (e.g., GitHub, VS Code) to see tables and Mermaid diagram rendered.
|
| 170 |
- Follow the automation steps to process PDFs and generate outlines—perfect for learners exploring AI vision and text summarization!
|
| 171 |
|
| 172 |
-
This README now serves as both a project overview and a tutorial, making it a fun, educational asset for all! 🚀
|
|
|
|
| 1 |
---
|
| 2 |
+
title: 🧠Torch 📚Transformers 🖼️Diffusion CV 🧠SFT
|
| 3 |
+
emoji: 🧠📚
|
| 4 |
colorFrom: yellow
|
| 5 |
colorTo: indigo
|
| 6 |
sdk: streamlit
|
| 7 |
+
sdk_version: 1.44.1
|
| 8 |
app_file: app.py
|
| 9 |
pinned: false
|
| 10 |
license: mit
|
| 11 |
+
short_description: 📷Torch 📚Transformers 🖼️Diffusion CV 🧠SFT
|
| 12 |
---
|
| 13 |
|
| 14 |
+
# Features:
|
| 15 |
+
1. Camera Snap 📷
|
| 16 |
+
2. Test OCR 🔍
|
| 17 |
+
3. MD Gallery 📚
|
| 18 |
+
4. Download PDFs 📥
|
| 19 |
+
5. Build Titan 🌱
|
| 20 |
+
6. Test Image Gen 🎨
|
| 21 |
+
7. PDF Process 📄
|
| 22 |
+
8. Image Process 🖼️
|
| 23 |
+
9. Character Editor 🧑🎨
|
| 24 |
+
10. Character Gallery 🖼️
|
| 25 |
+
|
| 26 |
+
## Tutorial: Single to Double Page Emoji Outlines
|
| 27 |
+
|
| 28 |
+
### Single Page Outline: Key Functions in `app.py`
|
| 29 |
+
|
| 30 |
+
| **Function** | **Purpose** 🎯 | **How It Works** 🛠️ | **Emoji Insight** 😎 |
|
| 31 |
+
|----------------------------|---------------------------------------------|--------------------------------------------------|-------------------------------|
|
| 32 |
+
| `generate_filename` | Unique file names 📅 | Adds timestamp to sequence | 🕰️ Time’s your file buddy! |
|
| 33 |
+
| `pdf_url_to_filename` | Safe PDF names 🖋️ | Cleans URLs to underscores | 🚫 No URL mess! |
|
| 34 |
+
| `get_download_link` | Downloadable files ⬇️ | Base64-encodes for HTML links | 📦 Grab it, go! |
|
| 35 |
+
| `download_pdf` | Web PDF snatcher 🌐 | Fetches PDFs with `requests` | 📚 PDF pirate ahoy! |
|
| 36 |
+
| `process_pdf_snapshot` | PDF to images 🖼️ | Async snapshots (single/double/all) with `fitz` | 📸 Double-page dazzle! |
|
| 37 |
+
| `process_ocr` | Image text extractor 🔍 | Async GOT-OCR2_0 with `transformers` | 👀 Text ninja strikes! |
|
| 38 |
+
| `process_image_gen` | Prompt to image 🎨 | Async Stable Diffusion with `diffusers` | 🖌️ Art from words—bam! |
|
| 39 |
+
| `process_image_with_prompt`| GPT image analysis 🤖 | Base64 to GPT vision | 🧠 GPT sees all! |
|
| 40 |
+
| `process_text_with_prompt` | GPT text summarizer ✍️ | Text to GPT for outlining | 📝 Summarize like a pro! |
|
| 41 |
+
| `update_gallery` | File showcase 🖼️📖 | Sidebar display with delete options | 🌟 Your creations shine! |
|
| 42 |
+
|
| 43 |
+
### Double Page Outline: Libraries in `requirements.txt`
|
| 44 |
+
|
| 45 |
+
| **Library** | **Single Page Purpose** 🎯 | **Double Page Usage** 🛠️ | **Emoji Insight** 😎 |
|
| 46 |
+
|---------------|-------------------------------------------|----------------------------------------------------|-------------------------------|
|
| 47 |
+
| `streamlit` | App UI 🌐 | Tabs like “PDF Process 📄” and “MD Gallery 📚” | 🎬 App star—lights, action! |
|
| 48 |
+
| `pandas` | Data crunching 📈 | Ready for OCR/metadata tables | 📊 Table tamer awaits! |
|
| 49 |
+
| `torch` | ML engine 🔥 | Powers `transformers` and `diffusers` | 🔥 AI’s fiery heart! |
|
| 50 |
+
| `requests` | Web grabber 🌍 | Downloads PDFs in `download_pdf` | 🌐 Web loot collector! |
|
| 51 |
+
| `aiofiles` | Fast file ops ⚡ | Async writes in `process_ocr` | ✈️ File speed demon! |
|
| 52 |
+
| `pillow` | Image magic 🖌️ | PDF to image in `process_pdf_snapshot` | 🖼️ Pixel Picasso! |
|
| 53 |
+
| `PyMuPDF` | PDF handler 📜 | Snapshots in `process_pdf_snapshot` | 📜 PDF scroll master! |
|
| 54 |
+
| `transformers`| AI models 🗣️ | GOT-OCR2_0 in `process_ocr` | 🤖 Brain in a box! |
|
| 55 |
+
| `diffusers` | Image gen 🎨 | Stable Diffusion in `process_image_gen` | 🎨 Art generator supreme! |
|
| 56 |
+
| `openai` | GPT vision/text 🤖 | Image/text processing in GPT functions | 🌌 All-seeing AI oracle! |
|
| 57 |
+
| `glob2` | File finder 🔍 | Gallery files in `update_gallery` | 🕵️ File sleuth! |
|
| 58 |
+
| `pytz` | Time zones ⏰ | Timestamps in `generate_filename` | ⏳ Time wizard! |
|
| 59 |
+
|
| 60 |
+
|
| 61 |
# TorchTransformers Diffusion CV SFT Titans 🚀
|
| 62 |
|
| 63 |
A Streamlit app blending `torch`, `transformers`, and `diffusers` for vision and NLP fun! Snap PDFs 📄, turn them into double-page spreads 🖼️, extract text with GPT 🤖, and craft emoji-packed Markdown outlines 📝—all with a witty UI and CPU-friendly SFT.
|
|
|
|
| 111 |
- 🖼️ **Image Process**: OCR images with GPT vision.
|
| 112 |
- 📚 **MD Gallery**: Summarize Markdown files into emoji outlines.
|
| 113 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 114 |
## Automation Instructions: Witty & Funny Steps 😂
|
| 115 |
|
| 116 |
1. **Load PDFs** 📚
|
|
|
|
| 182 |
- View it in a Markdown renderer (e.g., GitHub, VS Code) to see tables and Mermaid diagram rendered.
|
| 183 |
- Follow the automation steps to process PDFs and generate outlines—perfect for learners exploring AI vision and text summarization!
|
| 184 |
|
| 185 |
+
This README now serves as both a project overview and a tutorial, making it a fun, educational asset for all! 🚀
|