changes.md

TODO

1. Finish Run - Stancedata 6 pairs, not baseline. (2 unfrozen) DONE COPIED

   • rename 5-fold-hyper-tune as stance_results_run1
   • rename stancemodels as stance_models_run1

2. (finetune target) DOBASELINE=True in run_full_experiment and run again. DONE

   • stance_results_baseline
   • stance_models_baseline

3. 8 unfrozen bert layers, baseline-false, stance data DONE

   • stance_results_run2
   • stance_models_run2

4. Sentiment run DONE COPIED

   • sentiment_results_run1
   • sentiment_models_run1

5. Updated (finetune source+target) DOBASELINE=True DONE

   • stance_results_baseline2
   • stance_models_baseline2

6. Sentiment (finetune source+target) DOBASELINE=True DONE (change in run full exp to data & models )

   • sentiment_results_baseline2
   • sentiment_models_baseline2