details

README.md

@@ -82,11 +82,11 @@ A fine-tuned version of [google/long-t5-tglobal-base](https://huggingface.co/goo
 
 - between different checkpoints, about 20 epochs in total
 - all training was done at 16384 token input / 1024 max output
-- early checkpoints of this model were trained on a "smaller" subsection of the dataset as it was filtered for summaries of 1024 **characters**. This was subsequently caught and adjusted to **1024** tokens, and then trained further for at least five epochs. 
+- early checkpoints of this model were trained on a "smaller" subsection of the dataset as it was filtered for summaries of **1024 characters**. This was subsequently caught and adjusted to **1024 tokens**, and then trained further for at least five epochs. 
 
 ## Intended uses & limitations
 
-- At time of writing, the model is not _fully converged_ despite training for 20+ epochs. This checkpoint is serviceable enough (see examples).
+- At the time of writing, the model is not _fully converged_ despite training for 20+ epochs. This checkpoint is serviceable enough (see examples).
 - I plan to update this page with newer checkpoints and post some metrics over time.
 - Compare performance to [LED-base](https://huggingface.co/pszemraj/led-base-book-summary) trained on the same dataset.
 
@@ -98,7 +98,7 @@ A fine-tuned version of [google/long-t5-tglobal-base](https://huggingface.co/goo
 
 ### Training hyperparameters
 
-The following hyperparameters were used during the final training round:
+The following hyperparameters were used during the **final** training round\*:
 - learning_rate: 0.0004
 - train_batch_size: 2
 - eval_batch_size: 1
@@ -111,6 +111,8 @@ The following hyperparameters were used during the final training round:
 - lr_scheduler_warmup_ratio: 0.02
 - num_epochs: 2
 
+\*_Prior training sessions used roughly similar parameters, multiple sessions were required as this takes eons to train
+
 ### Training results