Project description

Pytora

A minimal pytorch implementation for lora via pytorch parametrizations

pip install pytora

from pytora import apply_lora

apply_lora(
    model,
    lora_r = 4,
    lora_alpha = 1,
    lora_dropout = 0.0
)

A simple example use case with a pretrained huggingface model can be found in demo.ipynb

These details have not been verified by PyPI

0.0.4

May 25, 2024

This version

0.0.3

May 14, 2024

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Uploaded May 14, 2024 Source

Uploaded May 14, 2024 Python 3

Hashes for pytora-0.0.3.tar.gz
Algorithm	Hash digest
SHA256	`da0c9b6b49d407a67e9a83dad382882a04f4bd9948fb5e0eb401f0fd9c722d7d`
MD5	`d218435d38276850671b652ea160bad1`
BLAKE2b-256	`4160729c8060e954e225315652999af5a61889e0791cca0615e5a415ac43bf47`

Hashes for pytora-0.0.3-py3-none-any.whl
Algorithm	Hash digest
SHA256	`56fef73b00d3a34fa0551a0ffb9784bfe03debd7319c175fbf80e7ce0cb2e750`
MD5	`574a443775e2dc5f0fa28fd2a5a40cf2`
BLAKE2b-256	`fa5415f58d6cefab891a5e0cbe3c96eac652b3e7be25145179d619d5d94c3c9d`