CPU-First Premium Transformer Library for low-resource devices
Project description
🚀 Pranjit Transformer
A CPU-first, lightweight transformer library optimized for training and inference on low-resource devices. Built with PyTorch, featuring modern architecture components like RoPE embeddings, RMS normalization, and gated feed-forward networks.
✨ Features
- 🖥️ CPU-Optimized: Designed for efficient training on standard hardware
- 🔧 User-Controlled: Full control over model configuration and hyperparameters
- 🧩 Hybrid Tokenizer: Word-level and character-level tokenization support
- ⚡ Modern Architecture: RoPE embeddings, RMS normalization, gated FFN
- 📦 Clean API: Simple and intuitive interface for quick experimentation
- 💾 Model I/O: Easy save/load functionality for checkpoints
📦 Installation
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file pranjit_transformer-0.1.0.tar.gz.
File metadata
- Download URL: pranjit_transformer-0.1.0.tar.gz
- Upload date:
- Size: 3.0 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.11.6
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
ccdedaba50adbdbcf7ff9ae8745ae14300d344f715ddef7f5ee818cd5374fda8
|
|
| MD5 |
382369333586bae17e727cba8cb6652f
|
|
| BLAKE2b-256 |
4be86eb180b276ae9d619b8c5fcfa3371baafb42d00a18ded0af6d63ce5be392
|
File details
Details for the file pranjit_transformer-0.1.0-py3-none-any.whl.
File metadata
- Download URL: pranjit_transformer-0.1.0-py3-none-any.whl
- Upload date:
- Size: 3.0 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.11.6
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
28991ed29d9cc4e2be5eeba9a2a00880594efc10865dbfc355da14696381d2d2
|
|
| MD5 |
5382ae637804c3744436fba10dd4c3c9
|
|
| BLAKE2b-256 |
d3a177e6b221fb4a07fc5438993d7cc6a2d1ea298bcb3d18fc30e7bc2c95cb3d
|
