chore: add trained weights using Git LFS

README.md

@@ -65,7 +65,7 @@ FOXES
 │   ├── convert_aia.py           # Raw AIA FITS -> paired 512x512 .npy stacks (itipy)
 │   ├── combine_sxr.py           # Combine raw multi-satellite GOES files into per-satellite CSVs
 │   ├── align_aia_sxr.py         # Match AIA timestamps to GOES CSVs -> per-timestamp SXR .npy
-│   ├── split_train_val_test.py  # Split processed AIA/SXR into train/val/test (training only)
+│   ├── split_train_val_test.py  # Split processed AIA/SXR into train/val/test (use for training a new model only)
 │   └── sxr_normalization.py     # Compute log-space mean/std over SXR .npy files for training
 ├── forecasting
 │   ├── dataset.py            # AIAGOESDataset / AIAGOESDataModule: loads paired AIA + SXR .npy files
@@ -73,7 +73,8 @@ FOXES
 │   ├── inference.py           # Run a checkpoint over a folder of data; writes predictions.csv
 │   ├── inference_config.yaml  # Config for inference.py
 │   ├── evaluation.py          # Compute metrics and generate evaluation plots
-│   └── evaluation_config.yaml # Config for evaluation.py
+│   ├── evaluation_config.yaml # Config for evaluation.py
+│   └── trained_weights_and_normalization/  # 3 released checkpoints + normalized_sxr.npy (see below)
 ├── training
 │   ├── train.py                # Train ViTLocal with PyTorch Lightning + Weights & Biases logging
 │   ├── train_config.yaml       # Config for train.py
@@ -181,14 +182,26 @@ If you don't have ground-truth SXR data (e.g. scoring new/live data), set
 
 ### 2) Run inference
 
+Three released checkpoints are included under
+`forecasting/trained_weights_and_normalization/`, differing only in the
+self-attention mask they were trained with (see `mask_mode` in
+`training/train_config.yaml`) — each checkpoint carries its own mask, so
+nothing else needs to change to switch between them, just `checkpoint_path`:
+
+| Checkpoint | `mask_mode` | Description |
+| --- | --- | --- |
+| `inverted-attention-mask.ckpt` | `inverted` | **Original released FOXES model.** Each patch attends to *distant* patches (the flipped local-attention mask it was actually trained with). |
+| `localized-attention-mask.ckpt` | `local` | True local attention — the opposite of the original model: each patch attends only to its own neighborhood. |
+| `no-attention-mask.ckpt` | `none` | Global ViT — standard full/global attention, no masking at all. |
+
 Edit `forecasting/inference_config.yaml`:
 
 ```yaml
 data:
   aia_dir:         "/path/to/your/aia_data"
   sxr_dir:         "/path/to/your/sxr_data"   # omit/ignore if prediction_only
-  sxr_norm_path:   "/path/to/normalized_sxr.npy"
-  checkpoint_path: "/path/to/checkpoint.ckpt"
+  sxr_norm_path:   "forecasting/trained_weights_and_normalization/normalized_sxr.npy"
+  checkpoint_path: "forecasting/trained_weights_and_normalization/inverted-attention-mask.ckpt"
 
 output_path: "/path/to/predictions.csv"
 ```

data/build_dataset_config.yaml

@@ -10,17 +10,17 @@
 wavelengths: [94, 131, 171, 193, 211, 304, 335]
 
 aia:
-  raw_dir:        "/Volumes/T9/testing_foxes/AIA_raw"       # raw FITS from download/download_sdo.py
-  bad_files_dir:  "/Volumes/T9/testing_foxes/SDO/AIA_bad"       # bad-timestamp FITS get moved here
-  processed_dir:  "/Volumes/T9/testing_foxes/AIA_processed" # paired 512x512 .npy stacks
+  raw_dir:        "./foxes_data/AIA_raw"        # raw FITS from download/download_sdo.py
+  bad_files_dir:  "./foxes_data/AIA_bad"        # bad-timestamp FITS get moved here
+  processed_dir:  "./foxes_data/AIA_processed"  # paired 512x512 .npy stacks
 
 sxr:
-  raw_dir:      "/Volumes/T9/testing_foxes/GOES_raw"        # raw netCDF from download/download_sxr.py
-  combined_dir: "/Volumes/T9/testing_foxes/GOES_combined"   # one interpolated CSV per satellite
+  raw_dir:      "./foxes_data/SXR_raw"         # raw netCDF from download/download_sxr.py
+  combined_dir: "./foxes_data/SXR_combined"    # one interpolated CSV per satellite
 
 output:
-  sxr_dir:         "/Volumes/T9/testing_foxes/SXR_processed" # per-timestamp xrsb_flux .npy, aligned to AIA
-  aia_missing_dir: "/Volumes/T9/testing_foxes/AIA_missing"   # AIA files with no matching SXR value
+  sxr_dir:         "./foxes_data/SXR_processed" # per-timestamp xrsb_flux .npy, aligned to AIA
+  aia_missing_dir: "./foxes_data/AIA_missing"   # AIA files with no matching SXR value
 
 processing:
   max_processes: null    # cap worker processes; null = use all CPUs
@@ -34,7 +34,7 @@ steps:
   convert_aia: true    # raw AIA FITS -> paired .npy stacks
   combine_sxr: true    # combine raw multi-satellite GOES files into per-satellite CSVs
   align:       true    # match AIA timestamps to GOES CSVs -> per-timestamp SXR .npy
-  split:       true   # split aia.processed_dir / output.sxr_dir into train/val/test
+  split:       false  # split aia.processed_dir / output.sxr_dir into train/val/test
 
 # Only needed if you're training (training/train_config.yaml expects
 # aia_dir/sxr_dir to each have train/val/test subfolders) — not needed for
@@ -42,14 +42,14 @@ steps:
 # aia.processed_dir and output.sxr_dir in place, using the same date ranges
 # so a given timestamp lands in the same split for both.
 split:
-  train_range: ["2023-08-01","2023-08-15"]   # e.g. ["2012-01-01", "2022-12-31"]; null with the others -> month-based default
-  val_range:   ["2023-08-16","2023-08-22"]   # e.g. ["2023-01-01", "2023-06-30"]
-  test_range:  ["2023-08-23","2023-08-31"]   # e.g. ["2023-08-01", "2023-08-31"]
+  train_range: null   # e.g. ["2012-01-01", "2022-12-31"]; null with the others -> month-based default
+  val_range:   null   # e.g. ["2023-01-01", "2023-06-30"]
+  test_range:  null   # e.g. ["2023-08-01", "2023-08-31"]
   copy_files:  false  # copy instead of move
 
 # Optional — only needed if you're training, not for running inference
 # against a released checkpoint. Requires steps.split to have run (computes
 # stats from the train split only).
 sxr_normalization:
-  compute:     true
-  output_path: "/Volumes/T9/testing_foxes/normalized_sxr.npy"
+  compute:     false
+  output_path: "./foxes_data/normalized_sxr.npy"

download/hf_download_config.yaml

@@ -18,8 +18,8 @@ repo_id: "griffingoodwin04/FOXES"
 #   {aia_dir}/{local_split}/{filename}   ("validation" maps to local dir "val")
 #   {sxr_dir}/{local_split}/{filename}
 # -----------------------------------------------------------------------------
-aia_dir: "/Volumes/T9/AIA_hg_processed"
-sxr_dir: "/Volumes/T9/SXR_hg_processed"
+aia_dir: "./foxes_data/AIA_hg_processed"
+sxr_dir: "./foxes_data/SXR_hg_processed"
 
 # -----------------------------------------------------------------------------
 # Splits to download  (any subset of: train, validation, test)

download/sdo_download_config.yaml

@@ -10,11 +10,11 @@
 #   python download/download_sdo.py --config download/sdo_download_config.yaml
 # =============================================================================
 
-download_dir: "/Volumes/T9/testing_foxes/AIA_raw"
+download_dir: "./foxes_data/AIA_raw"
 email: "ggoodwin5@gsu.edu"   # must be registered with JSOC: http://jsoc.stanford.edu/ajax/register_email.html
 
 start_date: "2023-08-01 00:00:00"
 end_date:   "2023-12-01 00:00:00"
-cadence: 7200    # minutes between samples
+cadence: 60    # minutes between samples
 
 wavelengths: [94, 131, 171, 193, 211, 304, 335]

download/sxr_download_config.yaml

@@ -11,7 +11,7 @@
 #   python download/download_sxr.py --config download/sxr_download_config.yaml
 # =============================================================================
 
-save_dir: "/Volumes/T9/testing_foxes/GOES_raw"
+save_dir: "./foxes_data/SXR_raw"
 
 start_date: "2023-08-01"
 end_date:   "2023-08-31"

training/train_config.yaml

@@ -91,10 +91,10 @@ callbacks:
   attention_log_every_n_epochs: 1  # how often (in epochs) to log attention maps
 
 data:
-  aia_dir: "/Volumes/T9/testing_foxes/AIA_processed"
-  sxr_dir: "/Volumes/T9/testing_foxes/SXR_processed"
-  sxr_norm_path: "/Volumes/T9/testing_foxes/normalized_sxr.npy"
-  checkpoints_dir: "/Volumes/T9/chkpt/"  # update to actual checkpoint
+  aia_dir: "./foxes_data/AIA_processed"
+  sxr_dir: "./foxes_data/SXR_processed"
+  sxr_norm_path: "./foxes_data/normalized_sxr.npy"
+  checkpoints_dir: "./foxes_data/chkpt/"  # update to actual checkpoint
   num_workers: 8   # DataLoader worker processes; defaults to min(8, cpu_count()) if omitted
 
 wandb: