Quick Start

Installation

pip install xlm-core

For existing model implementations, also install:

pip install xlm-models

CLI Usage

XLM uses a simple CLI with three main arguments:

xlm job_type=<JOB> job_name=<NAME> experiment=<CONFIG>

Argument	Description
job_type	One of prepare_data, train, eval, or generate
job_name	A descriptive name for your run
experiment	Path to your Hydra experiment config

Example: ILM on LM1B

A complete workflow demonstrating the Insertion Language Model on the LM1B dataset:

1. Prepare Data

xlm job_type=prepare_data job_name=lm1b_prepare experiment=lm1b_ilm

2. Train

# Quick debug run (overfit a single batch)
xlm job_type=train job_name=lm1b_ilm experiment=lm1b_ilm debug=overfit

# Full training
xlm job_type=train job_name=lm1b_ilm experiment=lm1b_ilm

3. Evaluate

xlm job_type=eval job_name=lm1b_ilm experiment=lm1b_ilm \
    +eval.ckpt_path=<CHECKPOINT_PATH>

4. Generate

xlm job_type=generate job_name=lm1b_ilm experiment=lm1b_ilm \
    +generation.ckpt_path=<CHECKPOINT_PATH>

Tip: Add debug=[overfit,print_predictions] to print generated samples to the console:

xlm job_type=generate job_name=lm1b_ilm experiment=lm1b_ilm \
    +generation.ckpt_path=<CHECKPOINT_PATH> \
    debug=[overfit,print_predictions]

5. Push to Hugging Face Hub

xlm job_type=push_to_hub job_name=lm1b_ilm_hub experiment=lm1b_ilm \
    +hub_checkpoint_path=<CHECKPOINT_PATH> \
    +hub.repo_id=<YOUR_REPO_ID>