🍻 test2

all_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 50.0,
+    "epoch": 10.0,
     "eval_accuracy": 0.75,
-    "eval_loss": 0.7729406356811523,
-    "eval_runtime": 0.1206,
-    "eval_samples_per_second": 33.163,
-    "eval_steps_per_second": 8.291,
-    "train_loss": 0.04134651497006416,
-    "train_runtime": 38.0787,
-    "train_samples_per_second": 26.261,
-    "train_steps_per_second": 2.626
+    "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
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 50.0,
+    "epoch": 10.0,
     "eval_accuracy": 0.75,
-    "eval_loss": 0.7729406356811523,
-    "eval_runtime": 0.1206,
-    "eval_samples_per_second": 33.163,
-    "eval_steps_per_second": 8.291
+    "eval_loss": 1.1400723457336426,
+    "eval_runtime": 0.2791,
+    "eval_samples_per_second": 57.318,
+    "eval_steps_per_second": 7.165
 }

runs/Aug16_10-30-38_surveily-ag-02/events.out.tfevents.1692174704.surveily-ag-02.37321.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d42c6a3e7201ec2299e58dd0deca8f6d6b35e641d9c61621778827c49e06f2f2
+size 405

train_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 50.0,
-    "train_loss": 0.04134651497006416,
-    "train_runtime": 38.0787,
-    "train_samples_per_second": 26.261,
-    "train_steps_per_second": 2.626
+    "epoch": 10.0,
+    "train_loss": 0.12750993132591248,
+    "train_runtime": 23.8317,
+    "train_samples_per_second": 16.784,
+    "train_steps_per_second": 4.196
 }

trainer_state.json

@@ -1,94 +1,130 @@
 {
-  "best_metric": 0.7729406356811523,
+  "best_metric": 1.1400723457336426,
   "best_model_checkpoint": "./vit-base/checkpoint-100",
-  "epoch": 50.0,
+  "epoch": 10.0,
   "global_step": 100,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 5.0,
-      "learning_rate": 0.000182,
-      "loss": 0.4121,
+      "epoch": 1.0,
+      "learning_rate": 0.00018,
+      "loss": 1.1117,
       "step": 10
     },
     {
-      "epoch": 10.0,
-      "learning_rate": 0.000162,
-      "loss": 0.0007,
+      "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,
       "step": 20
     },
     {
-      "epoch": 15.0,
-      "learning_rate": 0.000142,
-      "loss": 0.0002,
+      "epoch": 3.0,
+      "learning_rate": 0.00014,
+      "loss": 0.0041,
       "step": 30
     },
     {
-      "epoch": 20.0,
-      "learning_rate": 0.000122,
-      "loss": 0.0001,
+      "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,
       "step": 40
     },
     {
-      "epoch": 25.0,
-      "learning_rate": 0.00010200000000000001,
-      "loss": 0.0001,
+      "epoch": 5.0,
+      "learning_rate": 0.0001,
+      "loss": 0.0005,
       "step": 50
     },
     {
-      "epoch": 30.0,
-      "learning_rate": 8.2e-05,
-      "loss": 0.0001,
+      "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,
       "step": 60
     },
     {
-      "epoch": 35.0,
-      "learning_rate": 6.2e-05,
-      "loss": 0.0001,
+      "epoch": 7.0,
+      "learning_rate": 6e-05,
+      "loss": 0.0003,
       "step": 70
     },
     {
-      "epoch": 40.0,
-      "learning_rate": 4.2e-05,
-      "loss": 0.0001,
+      "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,
       "step": 80
     },
     {
-      "epoch": 45.0,
-      "learning_rate": 2.2000000000000003e-05,
-      "loss": 0.0001,
+      "epoch": 9.0,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
       "step": 90
     },
     {
-      "epoch": 50.0,
-      "learning_rate": 2.0000000000000003e-06,
-      "loss": 0.0,
+      "epoch": 10.0,
+      "learning_rate": 0.0,
+      "loss": 0.0003,
       "step": 100
     },
     {
-      "epoch": 50.0,
+      "epoch": 10.0,
       "eval_accuracy": 0.75,
-      "eval_loss": 0.7729406356811523,
-      "eval_runtime": 0.1172,
-      "eval_samples_per_second": 34.137,
-      "eval_steps_per_second": 8.534,
+      "eval_loss": 1.1400723457336426,
+      "eval_runtime": 0.2691,
+      "eval_samples_per_second": 59.468,
+      "eval_steps_per_second": 7.433,
       "step": 100
     },
     {
-      "epoch": 50.0,
+      "epoch": 10.0,
       "step": 100,
-      "total_flos": 7.7493378686976e+16,
-      "train_loss": 0.04134651497006416,
-      "train_runtime": 38.0787,
-      "train_samples_per_second": 26.261,
-      "train_steps_per_second": 2.626
+      "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
     }
   ],
   "max_steps": 100,
-  "num_train_epochs": 50,
-  "total_flos": 7.7493378686976e+16,
+  "num_train_epochs": 10,
+  "total_flos": 3.09973514747904e+16,
   "trial_name": null,
   "trial_params": null
 }