Chronos Forecasting Files

🚀 What's New?

• Support for low-rank adaption (LoRA) fine-tuning
• Optimized dataframe operations, making predict_df significantly faster than before

🐛 Bug Fixes and Enhancements

• Removed assert on batch size in multi-GPU environments
• Added option to skip dataframe validation in predict_df
• Chronos-2: Renamed predict_batches_jointly to cross_learning for clarity
• Chronos-2: Added option to specify callbacks in fit
• Chronos-2: Handle prediction_length < 3 for case where future_df is provided

What's Changed

• Chronos-2: Convert assert about batch size to warning by @abdulfatir in https://github.com/amazon-science/chronos-forecasting/pull/392
• Fix link in Chronos-Bolt tutorial notebook entry by @shchur in https://github.com/amazon-science/chronos-forecasting/pull/396
• Optimize convert_df_input_to_list_of_dicts_input and validate_df_inputs by @shchur in https://github.com/amazon-science/chronos-forecasting/pull/395
• Chronos-2: Add LoRA fine-tuning support by @abdulfatir in https://github.com/amazon-science/chronos-forecasting/pull/393
• Chronos-2: Add option to skip dataframe validation in predict_df by @abdulfatir in https://github.com/amazon-science/chronos-forecasting/pull/400
• Fix potential crash due to unbound local in df_utils by @abdulfatir in https://github.com/amazon-science/chronos-forecasting/pull/404
• Chronos-2: Add option to specify callbacks in fit by @abdulfatir in https://github.com/amazon-science/chronos-forecasting/pull/405
• Handle prediction_length below 3 by @shchur in https://github.com/amazon-science/chronos-forecasting/pull/407
• Chronos-2: set default dataloader_num_workers=0 in Trainer by @abdulfatir in https://github.com/amazon-science/chronos-forecasting/pull/409
• Chronos-2: Add option to remove PrinterCallback by @abdulfatir in https://github.com/amazon-science/chronos-forecasting/pull/410
• Bump version to 2.2.0rc3 by @abdulfatir in https://github.com/amazon-science/chronos-forecasting/pull/411
• Chronos-2: Do not catch exceptions in predict_fev when fine-tuning by @abdulfatir in https://github.com/amazon-science/chronos-forecasting/pull/417
• Chronos-2: Rename predict_batches_jointly to cross_learning by @abdulfatir in https://github.com/amazon-science/chronos-forecasting/pull/418
• Chronos-2: Update context_length if fine-tuned with longer than default by @abdulfatir in https://github.com/amazon-science/chronos-forecasting/pull/419
• Chronos-2: Ensure updated chronos_config is saved after fine-tuning by @abdulfatir in https://github.com/amazon-science/chronos-forecasting/pull/420
• Chronos-2: Handle missing 'peft' and lora_config specified by @abdulfatir in https://github.com/amazon-science/chronos-forecasting/pull/421
• Bump version from 2.2.0rc3 to 2.2.0rc4 by @abdulfatir in https://github.com/amazon-science/chronos-forecasting/pull/422
• Bump version from 2.2.0rc4 to 2.2.0 by @abdulfatir in https://github.com/amazon-science/chronos-forecasting/pull/426
• Add validate_inputs to predict_df in BaseChronosPipeline by @abdulfatir in https://github.com/amazon-science/chronos-forecasting/pull/427

Full Changelog: https://github.com/amazon-science/chronos-forecasting/compare/v2.1.0...v2.2.0