enable using stitched dataset

castorini · Aug 10, 2022 · 4ba5f42 · 4ba5f42
1 parent fc00927
commit 4ba5f42
Show file tree

Hide file tree

Showing 2 changed files with 2 additions and 2 deletions.
diff --git a/README.md b/README.md
@@ -55,7 +55,7 @@ For detailed explanation, please refer to [How to generate a dataset for custom
 ### Training and Running a Model
 
 1. Source the relevant environment variables for training the `res8` model: `source envs/res8.env`.
-2. Train the model: `python -m training.run.train -i datasets/fire/positive datasets/fire/negative --model res8 --workspace workspaces/fire-res8`.
+2. Train the model: `python -m training.run.train -i datasets/fire/positive datasets/fire/negative --model res8 --workspace workspaces/fire-res8`. It's recommended to also use `--use-stitched-datasets` if the training datasets are small.
 3. For the CLI demo, run `python -m training.run.demo --model res8 --workspace workspaces/fire-res8`.
 
 `train_model.sh` is also available which encaspulates individual command into a single bash script

diff --git a/train_model.sh b/train_model.sh
@@ -23,4 +23,4 @@ done
 source ${ENV_FILE_PATH}
 
 echo ">>> training a model for ${VOCAB}; model will be stored at ${WORKSPACE_PATH}"
-time python -m training.run.train --model ${MODEL_TYPE} --workspace "${WORKSPACE_PATH}" ${DATASET_ARGUMENT}
+time python -m training.run.train --model ${MODEL_TYPE} --workspace "${WORKSPACE_PATH}" ${DATASET_ARGUMENT} --use-stitched-datasets