v0.2.0

Summary of Changes

• Adds a new data_formatter_template field to format data while training from a JSON with custom fields. Eliminating the need to do preprocessing and format data to alpaca style. Find details in README
• Update evaluation_strategy flag to eval_strategy
• Add evaluation data format scripts to use as reference

What's Changed

• fix: check if output dir exists by @anhuong in #160
• tests for fixing full fine tuning by @Ssukriti in #162
• Evaluation Data Format Scripts by @alex-jw-brooks in #115
• Refactor tests explicit params by @Ssukriti in #163
• update eval_strategy flag used in transformers by @anhuong in #168
• remove unused python39 from dockerfile by @jbusche in #167
• Add formatting function alpaca by @Ssukriti in #161

Pip package: pip install fms-hf-tuning==0.2.0

Full Changelog: v0.1.0...v0.2.0