sschmidtjulia/vit-multitask61

ViT experiments: Multitask

Overview

This repository is a compact, custom PyTorch implementation of ViT for Multitask. The tiny configuration is intended for code review, smoke tests, and small controlled experiments rather than as a production-ready pretrained release.

Quick check

python predict.py --help

Inspect the script's __main__ block for its generated smoke-test example. Because this is a custom implementation, generic automatic loading APIs require an explicit adapter before use.

Acknowledgments

Thanks to the maintainers of the open-source libraries this project leans on.

Repository status

• Single-file implementation; the __main__ block runs a smoke test.
• config.json and training_args.json hold the generated settings.
• model.safetensors is freshly initialized — plumbing, not results.
• No benchmark numbers are reported on purpose.

Training details

• Framework: PyTorch
• Precision: fp32
• Init scheme: xavier
• Hardware: a single consumer GPU is enough
• Typical run: minutes, not hours

Default experiment recipe

Defaults: rmsprop optimizer, onecycle schedule. Treat these as starting points; a real run should sweep learning rate and batch size before drawing conclusions.

Architecture

• Architecture: ViT
• Scale: tiny
• Attention: multi query
• Fusion: bilinear
• Activation: swish
• Normalization: instancenorm

Limitations

• weights are random-init, so outputs carry no signal
• robustness, fairness and domain transfer are untested
• the API may change without notice while the design settles

Files

• predict.py — primary artifact
• README.md — this documentation
• config.json — architecture configuration
• training_args.json — default experiment settings
• model.safetensors — initialization checkpoint

License

Released under unknown. Review the source-data terms separately when this repository is used with external datasets.