From 6070c371f3edd0c1f48695a22df5d953f429240a Mon Sep 17 00:00:00 2001
From: YannAhlgrim <yann@ahlgrim.net>
Date: Mon, 1 Jun 2026 11:12:38 +0200
Subject: [PATCH] vitb supervised config

---
 configs/grids/lp_grid.yaml                    |  2 +-
 .../supervised_wilds_vitb16-448_ep600-lp.yaml | 61 +++++++++++++++++++
 2 files changed, 62 insertions(+), 1 deletion(-)
 create mode 100644 configs/supervised_wilds_vitb16-448_ep600-lp.yaml

diff --git a/configs/grids/lp_grid.yaml b/configs/grids/lp_grid.yaml
index 9450176..23c9202 100644
--- a/configs/grids/lp_grid.yaml
+++ b/configs/grids/lp_grid.yaml
@@ -1,4 +1,4 @@
-base_config: configs/supervised_wilds_vith14_ep300-lp.yaml
+base_config: configs/supervised_wilds_vitb16-448_ep600-lp.yaml
 
 constants:
   logging.write_tag: linear_probe
diff --git a/configs/supervised_wilds_vitb16-448_ep600-lp.yaml b/configs/supervised_wilds_vitb16-448_ep600-lp.yaml
new file mode 100644
index 0000000..d51bead
--- /dev/null
+++ b/configs/supervised_wilds_vitb16-448_ep600-lp.yaml
@@ -0,0 +1,61 @@
+meta:
+  model_name: vit_base
+  embed_dim: 768
+  load_checkpoint: true
+  checkpoint_folder: experiment_logs/vitb16.448-bs.16-ep.600/
+  read_checkpoint: jepa-ep100.pth.tar
+  use_bfloat16: true
+  num_classes: 182
+  probe_type: linear
+  mlp_hidden_dim: 1024
+  dropout: 0.3
+
+data:
+  batch_size: 128
+  root_path: ./wilds_data
+  num_workers: 10
+  pin_mem: true
+  crop_size: 448
+  crop_scale: [1.0, 1.0]
+  use_random_resized_crop: false
+  use_horizontal_flip: false
+  use_color_distortion: false
+  color_jitter_strength: 0.0
+  use_gaussian_blur: false
+
+mask:
+  patch_size: 16
+
+optimization:
+  optimizer: adamw # 'adamw', 'sgd', or 'lars'
+  freeze_weights: true # true for linear probing, false for full fine-tuning
+  epochs: 300
+  lr: 0.01
+  weight_decay: 5.0e-4
+  lr_schedule: cosine
+  step_milestones: [15, 30, 45]
+  step_gamma: 0.1
+  start_lr: 0.0
+  final_lr: 0.0
+  warmup: 0
+  momentum: 0.99
+  lars_eta: 0.001
+  lars_eps: 1.0e-8
+  lars_exclude_bias_and_norm: true
+  ipe_scale: 1.0
+  # Sweep suggestions (manual edits):
+  # - lr: 0.01 | 0.05 | 0.001
+  # - weight_decay: 5.0e-4 | 0.0
+  early_stopping:
+    enabled: true
+    patience: 10
+    min_delta: 1.0e-4
+    min_epochs: 15
+    restore_best_weights: true
+
+validation:
+  eval_every: 1
+
+logging:
+  write_tag: linear_probe
+  auto_folder: true