https://github.com/starxsky/ane-gpt-new
New ANE GPT
https://github.com/starxsky/ane-gpt-new
ane coreml gpt-2 nlp pytorch transformers
Last synced: about 1 year ago
JSON representation
New ANE GPT
- Host: GitHub
- URL: https://github.com/starxsky/ane-gpt-new
- Owner: StarxSky
- License: other
- Created: 2022-12-13T10:52:17.000Z (over 3 years ago)
- Default Branch: main
- Last Pushed: 2022-12-19T10:30:19.000Z (over 3 years ago)
- Last Synced: 2025-04-06T21:24:37.474Z (about 1 year ago)
- Topics: ane, coreml, gpt-2, nlp, pytorch, transformers
- Language: Python
- Homepage:
- Size: 1.05 MB
- Stars: 4
- Watchers: 1
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- Contributing: CONTRIBUTING.md
- License: LICENSE.md
- Code of conduct: CODE_OF_CONDUCT.md
Awesome Lists containing this project
README
# ANE-GPT
### Azure Pipline
[](https://dev.azure.com/starxsky/ANE-GPT/_build/latest?definitionId=7&branchName=main)
- Optimizations and improvements based on Apple's project ["ML-ANE-Transformers"](https://github.com/Apple/ml-ane-transformers).
- The adaptation of the GPT-2 model in ANE will be carried out later in the project.
- In the original official example I made a slight modification to make it a model of GPT-2 architecture,
and can be optimized and accelerated on the ANE engine.
# USE
# clone this repo
>> git clone https://github.com/StarxSky/ANE-GPT-New
#First you need to install the required packages
>> cd ANE-GPT-New-main
# Strat Install
>> pip install -r requirements.txt
## Start Traing the ANE-GPT
# run 'Train.py'
>> python Train.py
# [LICENSE](https://github.com/StarxSky/ANE-GPT-New/blob/main/LICENSE.md)
- This repository uses Apple's original license,Apple reserves the rights to the original code