Model save

all_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.25,
-    "eval_accuracy": 0.1956521739130435,
-    "eval_loss": 1.5383708477020264,
-    "eval_runtime": 100.417,
-    "eval_samples_per_second": 0.916,
-    "eval_steps_per_second": 0.229
+    "epoch": 3.23,
+    "eval_accuracy": 0.36507936507936506,
+    "eval_loss": 1.390939474105835,
+    "eval_runtime": 50.4581,
+    "eval_samples_per_second": 1.249,
+    "eval_steps_per_second": 0.317
 }

runs/Jul25_23-36-15_33b58b9b2125/events.out.tfevents.1690329587.33b58b9b2125.471.2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:93f3a7d6ed3575787bf64547ac610a623209ab706ec58bf0fcc494cad8f0fe45
-size 405
+oid sha256:61b0e5deb564424fdb232e0ee799d98376651bff2dc33365d626a0b2c4911ec3
+size 722

test_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.25,
-    "eval_accuracy": 0.1956521739130435,
-    "eval_loss": 1.5383708477020264,
-    "eval_runtime": 100.417,
-    "eval_samples_per_second": 0.916,
-    "eval_steps_per_second": 0.229
+    "epoch": 3.23,
+    "eval_accuracy": 0.36507936507936506,
+    "eval_loss": 1.390939474105835,
+    "eval_runtime": 50.4581,
+    "eval_samples_per_second": 1.249,
+    "eval_steps_per_second": 0.317
 }

trainer_state.json

@@ -1,127 +1,151 @@
 {
-  "best_metric": 0.4166666666666667,
-  "best_model_checkpoint": "videomae-base-finetuned-ASBD_ESBD/checkpoint-42",
-  "epoch": 3.25,
-  "global_step": 84,
+  "best_metric": 0.34328358208955223,
+  "best_model_checkpoint": "videomae-base-finetuned-ASBD_ESBD/checkpoint-62",
+  "epoch": 3.225,
+  "global_step": 120,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.12,
-      "learning_rate": 4.933333333333334e-05,
-      "loss": 1.5615,
+      "epoch": 0.08,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 1.4881,
       "step": 10
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.266666666666667e-05,
-      "loss": 1.4616,
+      "epoch": 0.17,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 1.3349,
       "step": 20
     },
     {
       "epoch": 0.25,
-      "eval_accuracy": 0.2569444444444444,
-      "eval_loss": 1.3287016153335571,
-      "eval_runtime": 164.0785,
-      "eval_samples_per_second": 0.878,
-      "eval_steps_per_second": 0.219,
-      "step": 21
-    },
-    {
-      "epoch": 1.11,
-      "learning_rate": 3.6e-05,
-      "loss": 1.3015,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 1.377,
       "step": 30
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 2.9333333333333336e-05,
-      "loss": 1.215,
-      "step": 40
+      "epoch": 0.26,
+      "eval_accuracy": 0.19402985074626866,
+      "eval_loss": 1.4631586074829102,
+      "eval_runtime": 90.4696,
+      "eval_samples_per_second": 0.741,
+      "eval_steps_per_second": 0.188,
+      "step": 31
     },
     {
-      "epoch": 1.25,
-      "eval_accuracy": 0.4166666666666667,
-      "eval_loss": 1.2314302921295166,
-      "eval_runtime": 134.8195,
-      "eval_samples_per_second": 1.068,
-      "eval_steps_per_second": 0.267,
-      "step": 42
+      "epoch": 1.07,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 1.3908,
+      "step": 40
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 2.2666666666666668e-05,
-      "loss": 1.2456,
+      "epoch": 1.16,
+      "learning_rate": 3.240740740740741e-05,
+      "loss": 1.2884,
       "step": 50
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 1.6000000000000003e-05,
-      "loss": 1.1371,
+      "epoch": 1.24,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 1.2476,
       "step": 60
     },
     {
-      "epoch": 2.25,
-      "eval_accuracy": 0.3541666666666667,
-      "eval_loss": 1.2707536220550537,
-      "eval_runtime": 134.9477,
-      "eval_samples_per_second": 1.067,
-      "eval_steps_per_second": 0.267,
-      "step": 63
+      "epoch": 1.26,
+      "eval_accuracy": 0.34328358208955223,
+      "eval_loss": 1.310468316078186,
+      "eval_runtime": 85.9108,
+      "eval_samples_per_second": 0.78,
+      "eval_steps_per_second": 0.198,
+      "step": 62
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 9.333333333333334e-06,
-      "loss": 1.1823,
+      "epoch": 2.07,
+      "learning_rate": 2.314814814814815e-05,
+      "loss": 1.0522,
       "step": 70
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 2.666666666666667e-06,
-      "loss": 1.043,
+      "epoch": 2.15,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 1.2907,
       "step": 80
     },
     {
-      "epoch": 3.25,
-      "eval_accuracy": 0.3680555555555556,
-      "eval_loss": 1.2449548244476318,
-      "eval_runtime": 140.7456,
-      "eval_samples_per_second": 1.023,
-      "eval_steps_per_second": 0.256,
-      "step": 84
+      "epoch": 2.23,
+      "learning_rate": 1.388888888888889e-05,
+      "loss": 1.2034,
+      "step": 90
     },
     {
-      "epoch": 3.25,
-      "step": 84,
-      "total_flos": 4.186851697156424e+17,
-      "train_loss": 1.2573059399922688,
-      "train_runtime": 1343.3508,
-      "train_samples_per_second": 0.25,
-      "train_steps_per_second": 0.063
+      "epoch": 2.26,
+      "eval_accuracy": 0.2835820895522388,
+      "eval_loss": 1.3650459051132202,
+      "eval_runtime": 87.6565,
+      "eval_samples_per_second": 0.764,
+      "eval_steps_per_second": 0.194,
+      "step": 93
     },
     {
-      "epoch": 3.25,
-      "eval_accuracy": 0.1956521739130435,
-      "eval_loss": 1.5383708477020264,
-      "eval_runtime": 120.4967,
-      "eval_samples_per_second": 0.764,
-      "eval_steps_per_second": 0.191,
-      "step": 84
+      "epoch": 3.06,
+      "learning_rate": 9.259259259259259e-06,
+      "loss": 1.199,
+      "step": 100
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 4.6296296296296296e-06,
+      "loss": 1.0395,
+      "step": 110
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.0,
+      "loss": 0.9454,
+      "step": 120
+    },
+    {
+      "epoch": 3.23,
+      "eval_accuracy": 0.3283582089552239,
+      "eval_loss": 1.347509741783142,
+      "eval_runtime": 86.3368,
+      "eval_samples_per_second": 0.776,
+      "eval_steps_per_second": 0.197,
+      "step": 120
+    },
+    {
+      "epoch": 3.23,
+      "step": 120,
+      "total_flos": 5.869068896906772e+17,
+      "train_loss": 1.238083585103353,
+      "train_runtime": 1251.5665,
+      "train_samples_per_second": 0.384,
+      "train_steps_per_second": 0.096
+    },
+    {
+      "epoch": 3.23,
+      "eval_accuracy": 0.36507936507936506,
+      "eval_loss": 1.390939474105835,
+      "eval_runtime": 57.4152,
+      "eval_samples_per_second": 1.097,
+      "eval_steps_per_second": 0.279,
+      "step": 120
     },
     {
-      "epoch": 3.25,
-      "eval_accuracy": 0.1956521739130435,
-      "eval_loss": 1.5383708477020264,
-      "eval_runtime": 100.417,
-      "eval_samples_per_second": 0.916,
-      "eval_steps_per_second": 0.229,
-      "step": 84
+      "epoch": 3.23,
+      "eval_accuracy": 0.36507936507936506,
+      "eval_loss": 1.390939474105835,
+      "eval_runtime": 50.4581,
+      "eval_samples_per_second": 1.249,
+      "eval_steps_per_second": 0.317,
+      "step": 120
     }
   ],
-  "max_steps": 84,
+  "max_steps": 120,
   "num_train_epochs": 9223372036854775807,
-  "total_flos": 4.186851697156424e+17,
+  "total_flos": 5.869068896906772e+17,
   "trial_name": null,
   "trial_params": null
 }