fix: remove dataset download, it's already in repo

This commit is contained in:
Christian Medina
2026-06-30 17:11:29 -04:00
parent 28d0a956fd
commit 7e8ddabf03

View File

@@ -51,8 +51,9 @@ echo ""
echo "=== Dataset Preparation ===" echo "=== Dataset Preparation ==="
echo "Checking for combined_20k.jsonl..." echo "Checking for combined_20k.jsonl..."
if [ ! -f "dataset/combined_20k.jsonl" ]; then if [ ! -f "dataset/combined_20k.jsonl" ]; then
echo "Dataset not found. Downloading..." echo "ERROR: Dataset not found!"
curl -L -o dataset/combined_20k.jsonl https://gitea.cyaren.com/cmedina/agenx-lora-training/raw/branch/main/dataset/combined_20k.jsonl echo "The file should be in the repo at: dataset/combined_20k.jsonl"
exit 1
fi fi
echo "Preparing dataset..." echo "Preparing dataset..."