Extra epochs training

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 4.99,
-    "eval_accuracy": 0.796875,
-    "eval_loss": 0.776684582233429,
-    "eval_runtime": 45.0682,
-    "eval_samples_per_second": 35.502,
-    "eval_steps_per_second": 0.555,
-    "total_flos": 1.971449904436347e+18,
-    "train_loss": 0.42377444733964637,
-    "train_runtime": 1020.3883,
-    "train_samples_per_second": 59.12,
-    "train_steps_per_second": 0.23
+    "epoch": 6.98,
+    "eval_accuracy": 0.87375,
+    "eval_loss": 0.4391254186630249,
+    "eval_runtime": 43.2356,
+    "eval_samples_per_second": 37.007,
+    "eval_steps_per_second": 0.578,
+    "total_flos": 2.784658737197482e+18,
+    "train_loss": 0.45341705913601676,
+    "train_runtime": 2059.1149,
+    "train_samples_per_second": 41.015,
+    "train_steps_per_second": 0.16
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 4.99,
-    "eval_accuracy": 0.796875,
-    "eval_loss": 0.776684582233429,
-    "eval_runtime": 45.0682,
-    "eval_samples_per_second": 35.502,
-    "eval_steps_per_second": 0.555
+    "epoch": 6.98,
+    "eval_accuracy": 0.87375,
+    "eval_loss": 0.4391254186630249,
+    "eval_runtime": 43.2356,
+    "eval_samples_per_second": 37.007,
+    "eval_steps_per_second": 0.578
 }

runs/Apr22_01-57-25_25733e00f35f/events.out.tfevents.1682128678.25733e00f35f.544.12

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a3d6a7ba72927d8f8501f28022ec30726a0aaad39f498cdd3c21adae8297776e
-size 18526
+oid sha256:12b3f2e0499f66f2356028ac5033c7ff56d5eaced70bb265cb43c2fa977a683c
+size 18880

runs/Apr22_01-57-25_25733e00f35f/events.out.tfevents.1682131388.25733e00f35f.544.15

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1056d1d2a8358a8416e282fbbe0a5d21654f1017027460e8141275c83c610170
+size 411

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 4.99,
-    "total_flos": 1.971449904436347e+18,
-    "train_loss": 0.42377444733964637,
-    "train_runtime": 1020.3883,
-    "train_samples_per_second": 59.12,
-    "train_steps_per_second": 0.23
+    "epoch": 6.98,
+    "total_flos": 2.784658737197482e+18,
+    "train_loss": 0.45341705913601676,
+    "train_runtime": 2059.1149,
+    "train_samples_per_second": 41.015,
+    "train_steps_per_second": 0.16
 }

trainer_state.json

@@ -1,8 +1,8 @@
 {
-  "best_metric": 0.796875,
-  "best_model_checkpoint": "swinv2-tiny-patch4-window8-256-finetuned-thai/checkpoint-141",
-  "epoch": 4.98941798941799,
-  "global_step": 235,
+  "best_metric": 0.87375,
+  "best_model_checkpoint": "swinv2-tiny-patch4-window8-256-finetuned-thai/checkpoint-329",
+  "epoch": 6.978835978835979,
+  "global_step": 329,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -120,89 +120,161 @@
     },
     {
       "epoch": 3.19,
-      "learning_rate": 0.0,
-      "loss": 1.092,
+      "learning_rate": 3.0236486486486488e-05,
+      "loss": 1.0802,
       "step": 150
     },
     {
       "epoch": 3.4,
-      "learning_rate": 0.0,
-      "loss": 1.0781,
+      "learning_rate": 2.8547297297297298e-05,
+      "loss": 1.018,
       "step": 160
     },
     {
       "epoch": 3.61,
-      "learning_rate": 0.0,
-      "loss": 1.0461,
+      "learning_rate": 2.685810810810811e-05,
+      "loss": 1.0061,
       "step": 170
     },
     {
       "epoch": 3.83,
-      "learning_rate": 0.0,
-      "loss": 1.0487,
+      "learning_rate": 2.516891891891892e-05,
+      "loss": 0.9125,
       "step": 180
     },
     {
       "epoch": 3.99,
-      "eval_accuracy": 0.796875,
-      "eval_loss": 0.776684582233429,
-      "eval_runtime": 43.5371,
-      "eval_samples_per_second": 36.75,
-      "eval_steps_per_second": 0.574,
+      "eval_accuracy": 0.840625,
+      "eval_loss": 0.6059834361076355,
+      "eval_runtime": 43.4292,
+      "eval_samples_per_second": 36.842,
+      "eval_steps_per_second": 0.576,
       "step": 188
     },
     {
       "epoch": 4.04,
-      "learning_rate": 0.0,
-      "loss": 1.0316,
+      "learning_rate": 2.347972972972973e-05,
+      "loss": 0.8931,
       "step": 190
     },
     {
       "epoch": 4.25,
-      "learning_rate": 0.0,
-      "loss": 1.06,
+      "learning_rate": 2.179054054054054e-05,
+      "loss": 0.8468,
       "step": 200
     },
     {
       "epoch": 4.46,
-      "learning_rate": 0.0,
-      "loss": 1.0528,
+      "learning_rate": 2.0101351351351353e-05,
+      "loss": 0.8092,
       "step": 210
     },
     {
       "epoch": 4.67,
-      "learning_rate": 0.0,
-      "loss": 1.0339,
+      "learning_rate": 1.8412162162162163e-05,
+      "loss": 0.7857,
       "step": 220
     },
     {
       "epoch": 4.88,
-      "learning_rate": 0.0,
-      "loss": 1.0889,
+      "learning_rate": 1.6722972972972976e-05,
+      "loss": 0.7527,
       "step": 230
     },
     {
       "epoch": 4.99,
-      "eval_accuracy": 0.796875,
-      "eval_loss": 0.776684582233429,
-      "eval_runtime": 43.9225,
-      "eval_samples_per_second": 36.428,
-      "eval_steps_per_second": 0.569,
+      "eval_accuracy": 0.8575,
+      "eval_loss": 0.5213606953620911,
+      "eval_runtime": 44.0663,
+      "eval_samples_per_second": 36.309,
+      "eval_steps_per_second": 0.567,
       "step": 235
     },
     {
-      "epoch": 4.99,
-      "step": 235,
-      "total_flos": 1.971449904436347e+18,
-      "train_loss": 0.42377444733964637,
-      "train_runtime": 1020.3883,
-      "train_samples_per_second": 59.12,
-      "train_steps_per_second": 0.23
+      "epoch": 5.1,
+      "learning_rate": 1.5033783783783784e-05,
+      "loss": 0.7727,
+      "step": 240
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 1.3344594594594596e-05,
+      "loss": 0.731,
+      "step": 250
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 1.1655405405405405e-05,
+      "loss": 0.7291,
+      "step": 260
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 9.966216216216217e-06,
+      "loss": 0.6753,
+      "step": 270
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 8.277027027027027e-06,
+      "loss": 0.6852,
+      "step": 280
+    },
+    {
+      "epoch": 5.98,
+      "eval_accuracy": 0.865625,
+      "eval_loss": 0.45880234241485596,
+      "eval_runtime": 43.8008,
+      "eval_samples_per_second": 36.529,
+      "eval_steps_per_second": 0.571,
+      "step": 282
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 6.587837837837837e-06,
+      "loss": 0.7279,
+      "step": 290
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 4.898648648648649e-06,
+      "loss": 0.6565,
+      "step": 300
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 3.2094594594594594e-06,
+      "loss": 0.6914,
+      "step": 310
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 1.5202702702702704e-06,
+      "loss": 0.6233,
+      "step": 320
+    },
+    {
+      "epoch": 6.98,
+      "eval_accuracy": 0.87375,
+      "eval_loss": 0.4391254186630249,
+      "eval_runtime": 43.5089,
+      "eval_samples_per_second": 36.774,
+      "eval_steps_per_second": 0.575,
+      "step": 329
+    },
+    {
+      "epoch": 6.98,
+      "step": 329,
+      "total_flos": 2.784658737197482e+18,
+      "train_loss": 0.45341705913601676,
+      "train_runtime": 2059.1149,
+      "train_samples_per_second": 41.015,
+      "train_steps_per_second": 0.16
     }
   ],
-  "max_steps": 235,
-  "num_train_epochs": 5,
-  "total_flos": 1.971449904436347e+18,
+  "max_steps": 329,
+  "num_train_epochs": 7,
+  "total_flos": 2.784658737197482e+18,
   "trial_name": null,
   "trial_params": null
 }