# =============================================================================
# Healthcare ML Pipeline Configuration
# =============================================================================
# Copy this file to .env and fill in your values.
# DO NOT commit .env to version control!

# -----------------------------------------------------------------------------
# RunPod Configuration (Required)
# -----------------------------------------------------------------------------
RUNPOD_API_KEY=your_runpod_api_key_here
RUNPOD_ENDPOINT=your_endpoint_id_here
RUNPOD_API_BASE=https://api.runpod.ai/v2

# -----------------------------------------------------------------------------
# AWS Configuration (Required for model storage)
# -----------------------------------------------------------------------------
AWS_ACCESS_KEY_ID=
AWS_SECRET_ACCESS_KEY=
AWS_SESSION_TOKEN=           # Optional - for assumed role sessions
AWS_REGION=us-east-1
S3_BUCKET=your-model-bucket

# -----------------------------------------------------------------------------
# Model Training Defaults (Optional - sensible defaults provided)
# -----------------------------------------------------------------------------
BASE_MODEL=emilyalsentzer/Bio_ClinicalBERT
MAX_SAMPLES=10000
EPOCHS=3
BATCH_SIZE=16
EVAL_SPLIT=0.1
LEARNING_RATE=2e-5

# -----------------------------------------------------------------------------
# Pipeline Runtime Settings (Optional)
# -----------------------------------------------------------------------------
POLL_INTERVAL_SECONDS=10     # How often to check training status
TRAINING_TIMEOUT_SECONDS=3600  # Max training time (1 hour default)