v1.2.1

18.11.2025 - Version 1.2.1

Fixes

• Added additional verifications for h5 files and embeddings calculation to improve reliability in autoeval_pipeline.
• Fixed issues with example embedding functions in the plm_eval Jupyter notebook for accurate demonstrations.
• Using Pydantic for BiotrainerSeqRecord class for validation and consistency.
• Exported OutputData in the output_files module for better modularity.

Features

• Introduced a new input_data configuration option, allowing sequence records to be directly provided in code for
  flexibility.