Merge pull request #20 from wli51/dev-fix-multi-gpu-update-trainer

PR integrating engine into new trainers and finalizing the multi-gpu fix

Author: wli51

CHANGELOG.md

@@ -7,6 +7,26 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ---
 
+## [0.4.2] - 2025-11-17
+
+### Added
+
+#### training infrastructure (`virtual_stain_flow/engine/...`)
+- Abstract away the forward pass and multiple loss accumulation from trainers
+#### logging trainer (`virutal_stain_flow/trainers/logging_trainer.py`)
+- New logging trainer for single generator model training using `engine`
+
+### Refactors
+
+#### abstract trainer (`virutal_stain_flow/trainers/AbstractTrainer.py`)
+- Add progress bar
+
+### Removes
+
+#### Obselete trainer classes (`virutal_stain_flow/trainers/logging_trainers/...`)
+
+---
+
 ## [0.4.1] - 2025-10-16
 
 ### Added

examples/0.download_data.py

@@ -0,0 +1,93 @@
+"""
+Download JUMP pilot plate data from AWS S3 bucket.
+
+This script downloads high-resolution cell imaging data from the JUMP (JCP
+Understanding Morphology & Plasticity) public dataset. The data includes
+multiple channels of brightfield and fluorescence microscopy images.
+
+Data Source:
+    S3 bucket: s3://cellpainting-gallery/cpg0000-jump-pilot/
+    Access: Public (no AWS credentials required with --no-sign-request)
+    Reference: https://github.com/jump-cellpainting/2024_Chandrasekaran_NatureMethods_CPJUMP1
+
+Defaults:
+    Batch: 2020_11_04_CPJUMP1
+    Plate: BR00117010__2020-11-08T18_18_00-Measurement1
+    Output: ./jump_pilot_subset/
+
+Usage:
+    # Download with defaults
+    python 0.download_data.py
+    
+    # Download to custom directory
+    python 0.download_data.py --outdir /path/to/data
+    
+    # Download different plate
+    python 0.download_data.py --batch 2020_11_04_CPJUMP1 --plate PLATE_ID
+
+Requirements:
+    - AWS CLI installed: https://aws.amazon.com/cli/
+    - Internet connection
+    - Sufficient disk space (~10GB for full plate)
+
+Note: Initial download may take minutes to hours depending on internet speed.
+"""
+
+import subprocess
+from pathlib import Path
+
+BASE = "s3://cellpainting-gallery/cpg0000-jump-pilot/source_4/images"
+
+def download_jump_plate(batch, plate, outdir="jump_pilot_subset"):
+    out = Path(outdir) / batch / plate
+    out.mkdir(parents=True, exist_ok=True)
+
+    s3_prefix = f"{BASE}/{batch}/images/{plate}/Images/"
+    print("Downloading from:", s3_prefix)
+
+    subprocess.run(
+        [
+            "aws", "s3", "cp", "--recursive",
+            s3_prefix, str(out),
+            "--no-sign-request",
+        ],
+        check=True,
+    )
+
+
+def main():
+    import argparse
+    parser = argparse.ArgumentParser(
+        description="Download JUMP pilot plate data from AWS S3",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        epilog="""
+Examples:
+  python 0.download_data.py
+  python 0.download_data.py --outdir ~/data/jump
+  python 0.download_data.py --batch 2020_11_04_CPJUMP1 --plate BR00117010__2020-11-08T18_18_00-Measurement1
+        """
+    )
+    parser.add_argument(
+        "--batch",
+        type=str,
+        default='2020_11_04_CPJUMP1',
+        help="JUMP batch ID (default: 2020_11_04_CPJUMP1)"
+    )
+    parser.add_argument(
+        "--plate",
+        type=str,
+        default='BR00117010__2020-11-08T18_18_00-Measurement1',
+        help="Plate ID within batch (default: BR00117010__2020-11-08T18_18_00-Measurement1)"
+    )
+    parser.add_argument(
+        "--outdir",
+        type=str,
+        default="jump_pilot_subset",
+        help="Output directory for downloaded data (default: ./jump_pilot_subset)"
+    )
+    args = parser.parse_args()
+    download_jump_plate(args.batch, args.plate, outdir=args.outdir)
+
+
+if __name__ == "__main__":
+    main()