fionafio1984/ml-retrieval

Retrieval (Coca)

Overview

Working implementation of Coca for Retrieval using a xlarge configuration. The repository focuses on transparent code and repeatable smoke tests; benchmark claims are deliberately omitted.

Expect rough edges — this is a working sketch, not a polished release.

Default experiment recipe

The generated recipe uses adafactor with a step schedule.

These values exist so the script runs end-to-end. A meaningful experiment needs its own tuning budget and at least three seeds.

Changelog

• 2026-10-09 — initial upload
• 2026-10-22 — config and docs sync

Citation

@misc{retrievalcoca},
  title = {Retrieval (Coca)},
  year = {{2026}},
  publisher = {{Hugging Face}},
}

Quick check

Run python train.py --help for the built-in smoke test. Because this is a custom implementation, generic automatic loading APIs require an explicit adapter before use.

Repository status

One Python module contains the model and a runnable example. Generated configuration lives in config.json; the recipe in training_args.json. The checkpoint is an initialization artifact and should not be read as a trained model. Scores are deliberately absent.

Limitations

The initialization checkpoint has not been trained or audited for robustness, fairness, or domain transfer. The implementation should be treated as an experimental starting point. Results from a future trained checkpoint must be documented separately from the defaults shipped here.

Files

• train.py — primary artifact
• README.md — this documentation
• config.json — architecture configuration
• training_args.json — default experiment settings
• model.safetensors — initialization checkpoint
• USE_POLICY.md — use and redistribution notes

License

Distributed under the bsd-2-clause license. External datasets keep their own terms — check them separately.