Training in progress, step 100

all_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 10.0,
-    "eval_accuracy": 0.75,
-    "eval_loss": 1.1400723457336426,
-    "eval_runtime": 0.2791,
-    "eval_samples_per_second": 57.318,
-    "eval_steps_per_second": 7.165,
-    "train_loss": 0.12750993132591248,
-    "train_runtime": 23.8317,
-    "train_samples_per_second": 16.784,
-    "train_steps_per_second": 4.196
+    "epoch": 20.0,
+    "eval_accuracy": 0.8125,
+    "eval_loss": 0.9370558857917786,
+    "eval_runtime": 0.2629,
+    "eval_samples_per_second": 60.867,
+    "eval_steps_per_second": 7.608,
+    "train_loss": 4.7035366296768185e-05,
+    "train_runtime": 37.8678,
+    "train_samples_per_second": 21.126,
+    "train_steps_per_second": 5.282
 }

config.json

@@ -12,7 +12,8 @@
     "0": "building_site",
     "1": "factory",
     "2": "office",
-    "3": "warehouse"
+    "3": "other",
+    "4": "warehouse"
   },
   "image_size": 224,
   "initializer_range": 0.02,
@@ -21,7 +22,8 @@
     "building_site": "0",
     "factory": "1",
     "office": "2",
-    "warehouse": "3"
+    "other": "3",
+    "warehouse": "4"
   },
   "layer_norm_eps": 1e-12,
   "model_type": "vit",

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 10.0,
-    "eval_accuracy": 0.75,
-    "eval_loss": 1.1400723457336426,
-    "eval_runtime": 0.2791,
-    "eval_samples_per_second": 57.318,
-    "eval_steps_per_second": 7.165
+    "epoch": 20.0,
+    "eval_accuracy": 0.8125,
+    "eval_loss": 0.9370558857917786,
+    "eval_runtime": 0.2629,
+    "eval_samples_per_second": 60.867,
+    "eval_steps_per_second": 7.608
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2181d534ec629ffbe070e0797695e57c54f6cc1c77131a2ff0a29517ccf0d466
-size 343274861
+oid sha256:839d44f3b0ed1721303fcbcd42108a6817b90722fd775469dda429b09dd95954
+size 343277933

runs/Aug16_11-32-01_surveily-ag-02/events.out.tfevents.1692178575.surveily-ag-02.4189.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f90bb586adce7c1339b859fa5a9b30591b23b8766e73ae05653551ca0d271e7
+size 411

runs/Aug16_12-11-02_surveily-ag-02/events.out.tfevents.1692180666.surveily-ag-02.5572.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e0f98ff4aaaa8250f754d7c5efaae39f79fc4845b0b35a013e23442ed27ea34
+size 4163

runs/Aug16_12-12-33_surveily-ag-02/events.out.tfevents.1692180756.surveily-ag-02.5572.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85c5bedc401cbf0d26d53f93358ec439dcde9765c0ee3d9bac8bf270ba0208dd
+size 4317

runs/Aug16_12-12-49_surveily-ag-02/events.out.tfevents.1692180772.surveily-ag-02.5572.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4ed7fdf13026e0523478864c916474eb99903c76ffd42c486f2efa28098c50bc
+size 6020

train_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 10.0,
-    "train_loss": 0.12750993132591248,
-    "train_runtime": 23.8317,
-    "train_samples_per_second": 16.784,
-    "train_steps_per_second": 4.196
+    "epoch": 20.0,
+    "train_loss": 4.7035366296768185e-05,
+    "train_runtime": 37.8678,
+    "train_samples_per_second": 21.126,
+    "train_steps_per_second": 5.282
 }

trainer_state.json

@@ -1,130 +1,163 @@
 {
-  "best_metric": 1.1400723457336426,
-  "best_model_checkpoint": "./vit-base/checkpoint-100",
-  "epoch": 10.0,
-  "global_step": 100,
+  "best_metric": 0.9370558857917786,
+  "best_model_checkpoint": "./vit-base/checkpoint-200",
+  "epoch": 20.0,
+  "global_step": 200,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 1.0,
-      "learning_rate": 0.00018,
-      "loss": 1.1117,
+      "learning_rate": 0.00019,
+      "loss": 0.0001,
       "step": 10
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.00016,
-      "loss": 0.1567,
-      "step": 20
-    },
-    {
-      "epoch": 2.0,
-      "eval_accuracy": 0.75,
-      "eval_loss": 0.9874135255813599,
-      "eval_runtime": 0.2906,
-      "eval_samples_per_second": 55.053,
-      "eval_steps_per_second": 6.882,
+      "learning_rate": 0.00018,
+      "loss": 0.0001,
       "step": 20
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.00014,
-      "loss": 0.0041,
+      "learning_rate": 0.00017,
+      "loss": 0.0001,
       "step": 30
     },
     {
       "epoch": 4.0,
-      "learning_rate": 0.00012,
-      "loss": 0.0007,
-      "step": 40
-    },
-    {
-      "epoch": 4.0,
-      "eval_accuracy": 0.75,
-      "eval_loss": 0.9884828329086304,
-      "eval_runtime": 0.2687,
-      "eval_samples_per_second": 59.539,
-      "eval_steps_per_second": 7.442,
+      "learning_rate": 0.00016,
+      "loss": 0.0001,
       "step": 40
     },
     {
       "epoch": 5.0,
-      "learning_rate": 0.0001,
-      "loss": 0.0005,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.0001,
       "step": 50
     },
     {
       "epoch": 6.0,
-      "learning_rate": 8e-05,
-      "loss": 0.0004,
-      "step": 60
-    },
-    {
-      "epoch": 6.0,
-      "eval_accuracy": 0.75,
-      "eval_loss": 1.1330695152282715,
-      "eval_runtime": 0.2768,
-      "eval_samples_per_second": 57.807,
-      "eval_steps_per_second": 7.226,
+      "learning_rate": 0.00014,
+      "loss": 0.0001,
       "step": 60
     },
     {
       "epoch": 7.0,
-      "learning_rate": 6e-05,
-      "loss": 0.0003,
+      "learning_rate": 0.00013000000000000002,
+      "loss": 0.0,
       "step": 70
     },
     {
       "epoch": 8.0,
-      "learning_rate": 4e-05,
-      "loss": 0.0003,
-      "step": 80
-    },
-    {
-      "epoch": 8.0,
-      "eval_accuracy": 0.75,
-      "eval_loss": 1.141401767730713,
-      "eval_runtime": 0.2772,
-      "eval_samples_per_second": 57.724,
-      "eval_steps_per_second": 7.216,
+      "learning_rate": 0.00012,
+      "loss": 0.0,
       "step": 80
     },
     {
       "epoch": 9.0,
-      "learning_rate": 2e-05,
-      "loss": 0.0003,
+      "learning_rate": 0.00011000000000000002,
+      "loss": 0.0,
       "step": 90
     },
     {
       "epoch": 10.0,
-      "learning_rate": 0.0,
-      "loss": 0.0003,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
       "step": 100
     },
     {
       "epoch": 10.0,
-      "eval_accuracy": 0.75,
-      "eval_loss": 1.1400723457336426,
-      "eval_runtime": 0.2691,
-      "eval_samples_per_second": 59.468,
-      "eval_steps_per_second": 7.433,
+      "eval_accuracy": 0.8125,
+      "eval_loss": 0.9386414885520935,
+      "eval_runtime": 0.2708,
+      "eval_samples_per_second": 59.076,
+      "eval_steps_per_second": 7.385,
       "step": 100
     },
     {
-      "epoch": 10.0,
-      "step": 100,
-      "total_flos": 3.09973514747904e+16,
-      "train_loss": 0.12750993132591248,
-      "train_runtime": 23.8317,
-      "train_samples_per_second": 16.784,
-      "train_steps_per_second": 4.196
+      "epoch": 11.0,
+      "learning_rate": 9e-05,
+      "loss": 0.0,
+      "step": 110
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 8e-05,
+      "loss": 0.0,
+      "step": 120
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 7e-05,
+      "loss": 0.0,
+      "step": 130
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 6e-05,
+      "loss": 0.0,
+      "step": 140
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 5e-05,
+      "loss": 0.0,
+      "step": 150
+    },
+    {
+      "epoch": 16.0,
+      "learning_rate": 4e-05,
+      "loss": 0.0,
+      "step": 160
+    },
+    {
+      "epoch": 17.0,
+      "learning_rate": 3e-05,
+      "loss": 0.0,
+      "step": 170
+    },
+    {
+      "epoch": 18.0,
+      "learning_rate": 2e-05,
+      "loss": 0.0,
+      "step": 180
+    },
+    {
+      "epoch": 19.0,
+      "learning_rate": 1e-05,
+      "loss": 0.0,
+      "step": 190
+    },
+    {
+      "epoch": 20.0,
+      "learning_rate": 0.0,
+      "loss": 0.0,
+      "step": 200
+    },
+    {
+      "epoch": 20.0,
+      "eval_accuracy": 0.8125,
+      "eval_loss": 0.9370558857917786,
+      "eval_runtime": 0.2545,
+      "eval_samples_per_second": 62.865,
+      "eval_steps_per_second": 7.858,
+      "step": 200
+    },
+    {
+      "epoch": 20.0,
+      "step": 200,
+      "total_flos": 6.19947029495808e+16,
+      "train_loss": 4.7035366296768185e-05,
+      "train_runtime": 37.8678,
+      "train_samples_per_second": 21.126,
+      "train_steps_per_second": 5.282
     }
   ],
-  "max_steps": 100,
-  "num_train_epochs": 10,
-  "total_flos": 3.09973514747904e+16,
+  "max_steps": 200,
+  "num_train_epochs": 20,
+  "total_flos": 6.19947029495808e+16,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eee70433826a0e6b7af7b9141e54aae8a5a48c5c19567df138c53083fcbeadc2
+oid sha256:5663d838179ea6da484c8fa0c3148604d031a3ce7df08fd3003b32f22b7222d1
 size 3963