Training in progress, epoch 1

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 10.0,
+    "total_flos": 4.769009900476416e+17,
+    "train_loss": 1.3194919480217828,
+    "train_runtime": 7630.326,
+    "train_samples_per_second": 2.942,
+    "train_steps_per_second": 0.012
+}

config.json

@@ -1,26 +1,13 @@
 {
   "_name_or_path": "google/vit-base-patch16-224",
   "architectures": [
-    "ResNetForImageClassification"
+    "ViTForImageClassification"
   ],
   "attention_probs_dropout_prob": 0.0,
-  "depths": [
-    3,
-    4,
-    6,
-    3
-  ],
-  "downsample_in_first_stage": false,
-  "embedding_size": 64,
+  "encoder_stride": 16,
   "hidden_act": "gelu",
   "hidden_dropout_prob": 0.0,
   "hidden_size": 768,
-  "hidden_sizes": [
-    256,
-    512,
-    1024,
-    2048
-  ],
   "id2label": {
     "0": "buildings",
     "1": "forest",
@@ -41,8 +28,7 @@
     "street": 5
   },
   "layer_norm_eps": 1e-12,
-  "layer_type": "bottleneck",
-  "model_type": "resnet",
+  "model_type": "vit",
   "num_attention_heads": 12,
   "num_channels": 3,
   "num_hidden_layers": 12,

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4abac6db2cf7f131e44462640ba38f0af9da739cf0b62b02fca2cc5af7a773e0
-size 94405441
+oid sha256:c7ebf6795980ce7cc67e7dc73455c42937d82ed37f161d63e6d87f793f7b8efb
+size 343279281

runs/Nov06_19-15-31_a77448885f3d/1667762138.3268747/events.out.tfevents.1667762138.a77448885f3d.77.3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c062c99830492c329dc2ffb42afec79225477fc4cec6d0f622b3f779114ba619
+size 5571

runs/Nov06_19-15-31_a77448885f3d/events.out.tfevents.1667762138.a77448885f3d.77.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e898b46ff7970509ecad2683ec73a68beca972ff66e5b0afd3598229fa5df88
+size 4274

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 10.0,
+    "total_flos": 4.769009900476416e+17,
+    "train_loss": 1.3194919480217828,
+    "train_runtime": 7630.326,
+    "train_samples_per_second": 2.942,
+    "train_steps_per_second": 0.012
+}

trainer_state.json

@@ -0,0 +1,169 @@
+{
+  "best_metric": 0.5742596303718548,
+  "best_model_checkpoint": "vit-base-patch16-224-finetuned-imageclassification/checkpoint-90",
+  "epoch": 10.0,
+  "global_step": 90,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.18147405922957024,
+      "eval_loss": 2.0663375854492188,
+      "eval_runtime": 5315.8097,
+      "eval_samples_per_second": 1.69,
+      "eval_steps_per_second": 0.027,
+      "step": 9
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.876543209876543e-05,
+      "loss": 1.7049,
+      "step": 10
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.34769539078156314,
+      "eval_loss": 1.633109211921692,
+      "eval_runtime": 62.6786,
+      "eval_samples_per_second": 143.302,
+      "eval_steps_per_second": 2.25,
+      "step": 18
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 8.641975308641975e-05,
+      "loss": 1.5848,
+      "step": 20
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.3907815631262525,
+      "eval_loss": 1.6548928022384644,
+      "eval_runtime": 59.6679,
+      "eval_samples_per_second": 150.533,
+      "eval_steps_per_second": 2.363,
+      "step": 27
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 7.407407407407407e-05,
+      "loss": 1.4471,
+      "step": 30
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.47873524827432645,
+      "eval_loss": 1.3122586011886597,
+      "eval_runtime": 64.0678,
+      "eval_samples_per_second": 140.195,
+      "eval_steps_per_second": 2.201,
+      "step": 36
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 6.17283950617284e-05,
+      "loss": 1.3214,
+      "step": 40
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.5012246715653529,
+      "eval_loss": 1.2882213592529297,
+      "eval_runtime": 63.7848,
+      "eval_samples_per_second": 140.817,
+      "eval_steps_per_second": 2.211,
+      "step": 45
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 4.938271604938271e-05,
+      "loss": 1.2432,
+      "step": 50
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.5526608773101759,
+      "eval_loss": 1.1481624841690063,
+      "eval_runtime": 63.6972,
+      "eval_samples_per_second": 141.011,
+      "eval_steps_per_second": 2.214,
+      "step": 54
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 1.1932,
+      "step": 60
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.5577822311289246,
+      "eval_loss": 1.1259204149246216,
+      "eval_runtime": 64.3997,
+      "eval_samples_per_second": 139.473,
+      "eval_steps_per_second": 2.189,
+      "step": 63
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 2.4691358024691357e-05,
+      "loss": 1.1336,
+      "step": 70
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.56312625250501,
+      "eval_loss": 1.1054401397705078,
+      "eval_runtime": 60.6277,
+      "eval_samples_per_second": 148.15,
+      "eval_steps_per_second": 2.326,
+      "step": 72
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 1.2345679012345678e-05,
+      "loss": 1.1301,
+      "step": 80
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.5655755956357159,
+      "eval_loss": 1.0953378677368164,
+      "eval_runtime": 63.7804,
+      "eval_samples_per_second": 140.827,
+      "eval_steps_per_second": 2.211,
+      "step": 81
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 0.0,
+      "loss": 1.1172,
+      "step": 90
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.5742596303718548,
+      "eval_loss": 1.0771620273590088,
+      "eval_runtime": 64.1958,
+      "eval_samples_per_second": 139.916,
+      "eval_steps_per_second": 2.196,
+      "step": 90
+    },
+    {
+      "epoch": 10.0,
+      "step": 90,
+      "total_flos": 4.769009900476416e+17,
+      "train_loss": 1.3194919480217828,
+      "train_runtime": 7630.326,
+      "train_samples_per_second": 2.942,
+      "train_steps_per_second": 0.012
+    }
+  ],
+  "max_steps": 90,
+  "num_train_epochs": 10,
+  "total_flos": 4.769009900476416e+17,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ffd89709a7eb60b1ea5efb67fa48dac5cf8beb3bf9f47f61506ffc0eb51632d9
+oid sha256:2d3e4e86369a939107e7ddeda00ff61e05c6d3a5dee77cf209e7104ea10ea3d4
 size 3439