Metadata-Version: 2.4
Name: xlm-core
Version: 0.1.1
Summary: XLM Framework
Author: Dhruvesh Patel, Benjamin Rozonoyer, Sai Sreenivas Chintha, Durga Prasad Maram
Project-URL: Source Code, https://github.com/dhruvdcoder/xlm-core
Keywords: AI,ML,Machine Learning,Deep Learning,Non-Autoregressive Language Models
Classifier: Development Status :: 4 - Beta
Classifier: Intended Audience :: Developers
Classifier: Intended Audience :: Science/Research
Classifier: Programming Language :: Python :: 3.11
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Requires-Python: >=3.11
Requires-Dist: lightning==2.5.2
Requires-Dist: fsspec
Requires-Dist: torch
Requires-Dist: jupyter
Requires-Dist: hydra-core
Requires-Dist: hydra-colorlog
Requires-Dist: hydra-joblib-launcher
Requires-Dist: hydra-submitit-launcher
Requires-Dist: datasets<4.0.0,>=3.3.2
Requires-Dist: transformers
Requires-Dist: more-itertools
Requires-Dist: torchdata>=0.9.0
Requires-Dist: rich
Requires-Dist: python-dotenv
Requires-Dist: jaxtyping
Requires-Dist: tensorboard
Requires-Dist: torch-ema
Requires-Dist: pydot
Requires-Dist: tabulate
Requires-Dist: pandas
Requires-Dist: simple_slurm
Dynamic: author
Dynamic: classifier
Dynamic: description
Dynamic: keywords
Dynamic: project-url
Dynamic: requires-dist
Dynamic: requires-python
Dynamic: summary


    XLM is a unified framework for developing and comparing small non-autoregressive language models. It uses PyTorch as the deep learning framework, PyTorch Lightning for training utilities, and Hydra for configuration management. XLM provides core components for flexible data handling and training, useful architectural implementations for non-autoregressive workflows, and support for arbitrary runtime code injection. Custom model implementations that leverage the core components of xlm can be found in the xlm-models package. The package also includes a few preconfigured synthetic planning and language-modeling datasets.

    Usage:
        pip install xlm-core

    Command usage:
        xlm job_type=[JOB_TYPE] job_name=[JOB_NAME] experiment=[CONFIG_PATH]
       
        The job_type argument can be one of train ,eval and generate. The experiment argument should point to the root hydra config file.
