Update README.md

README.md

@@ -45,14 +45,27 @@ Due to inconsistencies in the training data, the classifier performance leaves r
 
 ## Training and evaluation data
 
-The dataset is comprised of data from 2 sources:
+The training dataset is comprised of labelled passages from 2 sources:
 - [ClimateWatch NDC Sector data](https://www.climatewatchdata.org/data-explorer/historical-emissions?historical-emissions-data-sources=climate-watch&historical-emissions-gases=all-ghg&historical-emissions-regions=All%20Selected&historical-emissions-sectors=total-including-lucf%2Ctotal-including-lucf&page=1)
-- [IKI TraCS Climate Strategies for Transport Tracker](https://changing-transport.org/wp-content/uploads/20220722_Tracker_Database.xlsx) implemented by GIZ and funded by theInternational Climate Initiative (IKI) of the German Federal Ministry for Economic Affairs and Climate Action (BMWK).
+- [IKI TraCS Climate Strategies for Transport Tracker](https://changing-transport.org/wp-content/uploads/20220722_Tracker_Database.xlsx) implemented by GIZ and funded by theInternational Climate Initiative (IKI) of the German Federal Ministry for Economic Affairs and Climate Action (BMWK). Here we utilized the QA dataset (CW_NDC_data_Sector).
+
+The combined dataset[GIZ/policy_qa_v0_1](https://huggingface.co/datasets/GIZ/policy_qa_v0_1) contains ~85k rows. Each row is duplicated twice, to provide varying sequence lengths (denoted by the values 'small', 'medium', and 'large', which correspond to sequence lengths of 60, 85, and 150 respectively - indicated in the 'strategy' column). This effectively means the dataset is reduced by 1/3 in useful size, and the 'strategy' value should be selected based on the use case. For this training, we utilized the 'medium' samples Furthermore, for each row, the 'context' column contains 3 samples of varying quality. The approach used to assess quality and select samples is described below.
+
+The pre-processing operations used to produce the final training dataset were as follows:
+
+1. Dataset is filtered based on 'medium' value in 'strategy' column (sequence length = 85).
+2. For IKITracs, labels are assigned based on the presence of certain substrings ('_unc' or '_c') based on 'parameter' values which correspond to assessments of 'unconditional' or 'conditional' commitments by human annotaters.
+3. For ClimateWatch, the 'QuestionText' field is searched for the terms 'unconditional' or 'conditional', and labels assigned accordingly.
+3. If 'context_translated' is available and the 'language' is not English, 'context' is replaced with 'context_translated'. 
+4. The dataset is "exploded" - i.e., the text samples in the 'context' column, which are lists, are converted into separate rows - and labels are merged to align with the associated samples.
+5. The 'match_onanswer' and 'answerWordcount' are used conditionally to select hihg quality samples (prefers high % of word matches in 'match_onanswer', but will take lower if there is a high 'answerWordcount')
+6. Data is then augmented using sentence shuffle from the ```albumentations``` library
 
-From the first source, we take
 
 ## Training procedure
 
+The model hyperparameters were tuned using ```optuna``` over 10 trials on a truncated training and validation dataset. The model was then trained over 5 epochs using the best hyperparameters identified.
+
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
@@ -82,3 +95,4 @@ The following hyperparameters were used during training:
 - Pytorch 2.0.1+cu118
 - Datasets 2.13.1
 - Tokenizers 0.13.3
+