Training in progress, epoch 0

all_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 8.88888888888889,
-    "total_flos": 7.753465317556224e+16,
-    "train_loss": 3.181795922267553e+30,
-    "train_runtime": 46.0692,
-    "train_samples_per_second": 187.544,
-    "train_steps_per_second": 1.302
+    "epoch": 2.6666666666666665,
+    "total_flos": 2.326039595266867e+16,
+    "train_loss": 3.1840518987395357e+30,
+    "train_runtime": 15.2988,
+    "train_samples_per_second": 169.425,
+    "train_steps_per_second": 1.177
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1a28aa672236f18658afc6ad148fb337d606300bac7df8e79600f13d84e83adb
+oid sha256:db3b90d629a2f1c91871139cb963554956161f5281accc8990448df997f7cac3
 size 44766388

runs/Jun06_03-58-56_b446df81307f/events.out.tfevents.1717646337.b446df81307f.157.4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:27447a041c7420789c2a32b8b6ec917918c14e44c8c3e459d21f901c49f04d1d
+size 5232

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 8.88888888888889,
-    "total_flos": 7.753465317556224e+16,
-    "train_loss": 3.181795922267553e+30,
-    "train_runtime": 46.0692,
-    "train_samples_per_second": 187.544,
-    "train_steps_per_second": 1.302
+    "epoch": 2.6666666666666665,
+    "total_flos": 2.326039595266867e+16,
+    "train_loss": 3.1840518987395357e+30,
+    "train_runtime": 15.2988,
+    "train_samples_per_second": 169.425,
+    "train_steps_per_second": 1.177
 }

trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": 0.3958333333333333,
   "best_model_checkpoint": "resnet-18/checkpoint-6",
-  "epoch": 8.88888888888889,
+  "epoch": 2.6666666666666665,
   "eval_steps": 500,
-  "global_step": 60,
+  "global_step": 18,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -12,15 +12,15 @@
       "epoch": 0.8888888888888888,
       "eval_accuracy": 0.3958333333333333,
       "eval_loss": 2.8539976205603575e+30,
-      "eval_runtime": 0.2573,
-      "eval_samples_per_second": 373.047,
-      "eval_steps_per_second": 11.658,
+      "eval_runtime": 0.2601,
+      "eval_samples_per_second": 369.032,
+      "eval_steps_per_second": 11.532,
       "step": 6
     },
     {
       "epoch": 1.4814814814814814,
-      "grad_norm": 22.85268211364746,
-      "learning_rate": 4.62962962962963e-05,
+      "grad_norm": 25.099891662597656,
+      "learning_rate": 2.5e-05,
       "loss": 3.195940475249626e+30,
       "step": 10
     },
@@ -28,123 +28,34 @@
       "epoch": 1.925925925925926,
       "eval_accuracy": 0.3958333333333333,
       "eval_loss": 2.8539976205603575e+30,
-      "eval_runtime": 0.2547,
-      "eval_samples_per_second": 376.954,
-      "eval_steps_per_second": 11.78,
+      "eval_runtime": 0.2646,
+      "eval_samples_per_second": 362.761,
+      "eval_steps_per_second": 11.336,
       "step": 13
     },
     {
-      "epoch": 2.962962962962963,
-      "grad_norm": 26.933372497558594,
-      "learning_rate": 3.7037037037037037e-05,
-      "loss": 3.1627329754809595e+30,
-      "step": 20
-    },
-    {
-      "epoch": 2.962962962962963,
-      "eval_accuracy": 0.3958333333333333,
-      "eval_loss": 2.8539976205603575e+30,
-      "eval_runtime": 0.2507,
-      "eval_samples_per_second": 382.993,
-      "eval_steps_per_second": 11.969,
-      "step": 20
-    },
-    {
-      "epoch": 4.0,
-      "eval_accuracy": 0.3958333333333333,
-      "eval_loss": 2.8539976205603575e+30,
-      "eval_runtime": 0.3363,
-      "eval_samples_per_second": 285.439,
-      "eval_steps_per_second": 8.92,
-      "step": 27
-    },
-    {
-      "epoch": 4.444444444444445,
-      "grad_norm": 28.446868896484375,
-      "learning_rate": 2.777777777777778e-05,
-      "loss": 3.162728139777681e+30,
-      "step": 30
-    },
-    {
-      "epoch": 4.888888888888889,
-      "eval_accuracy": 0.3958333333333333,
-      "eval_loss": 2.8539976205603575e+30,
-      "eval_runtime": 0.2484,
-      "eval_samples_per_second": 386.545,
-      "eval_steps_per_second": 12.08,
-      "step": 33
-    },
-    {
-      "epoch": 5.925925925925926,
-      "grad_norm": 29.070621490478516,
-      "learning_rate": 1.8518518518518518e-05,
-      "loss": 3.203328220933291e+30,
-      "step": 40
-    },
-    {
-      "epoch": 5.925925925925926,
-      "eval_accuracy": 0.3958333333333333,
-      "eval_loss": 2.8539976205603575e+30,
-      "eval_runtime": 0.2586,
-      "eval_samples_per_second": 371.285,
-      "eval_steps_per_second": 11.603,
-      "step": 40
-    },
-    {
-      "epoch": 6.962962962962963,
-      "eval_accuracy": 0.3958333333333333,
-      "eval_loss": 2.8539976205603575e+30,
-      "eval_runtime": 0.2547,
-      "eval_samples_per_second": 376.967,
-      "eval_steps_per_second": 11.78,
-      "step": 47
-    },
-    {
-      "epoch": 7.407407407407407,
-      "grad_norm": 29.25043487548828,
-      "learning_rate": 9.259259259259259e-06,
-      "loss": 3.2033323312810776e+30,
-      "step": 50
-    },
-    {
-      "epoch": 8.0,
-      "eval_accuracy": 0.3958333333333333,
-      "eval_loss": 2.8539976205603575e+30,
-      "eval_runtime": 0.2704,
-      "eval_samples_per_second": 355.065,
-      "eval_steps_per_second": 11.096,
-      "step": 54
-    },
-    {
-      "epoch": 8.88888888888889,
-      "grad_norm": 29.65780258178711,
-      "learning_rate": 0.0,
-      "loss": 3.162713390882682e+30,
-      "step": 60
-    },
-    {
-      "epoch": 8.88888888888889,
+      "epoch": 2.6666666666666665,
       "eval_accuracy": 0.3958333333333333,
       "eval_loss": 2.8539976205603575e+30,
-      "eval_runtime": 0.3257,
-      "eval_samples_per_second": 294.758,
-      "eval_steps_per_second": 9.211,
-      "step": 60
+      "eval_runtime": 0.2658,
+      "eval_samples_per_second": 361.215,
+      "eval_steps_per_second": 11.288,
+      "step": 18
     },
     {
-      "epoch": 8.88888888888889,
-      "step": 60,
-      "total_flos": 7.753465317556224e+16,
-      "train_loss": 3.181795922267553e+30,
-      "train_runtime": 46.0692,
-      "train_samples_per_second": 187.544,
-      "train_steps_per_second": 1.302
+      "epoch": 2.6666666666666665,
+      "step": 18,
+      "total_flos": 2.326039595266867e+16,
+      "train_loss": 3.1840518987395357e+30,
+      "train_runtime": 15.2988,
+      "train_samples_per_second": 169.425,
+      "train_steps_per_second": 1.177
     }
   ],
   "logging_steps": 10,
-  "max_steps": 60,
+  "max_steps": 18,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 10,
+  "num_train_epochs": 3,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -158,7 +69,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 7.753465317556224e+16,
+  "total_flos": 2.326039595266867e+16,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5d2122e108f1986add0878418492368ef0309c0bf3dc4e3aa32e19fe51429f8a
+oid sha256:a4510e39fc69d06cfadbe8aeabda68062b30b7ff3b10f556d9764d556cb7aeb3
 size 5112