End of training

README.md

@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [microsoft/resnet-101](https://huggingface.co/microsoft/resnet-101) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3944
-- Accuracy: 0.8471
+- Loss: 0.3785
+- Accuracy: 0.8569
 
 ## Model description
 

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 9.8,
-    "eval_accuracy": 0.7700980392156863,
-    "eval_loss": 0.6762834191322327,
-    "eval_runtime": 27.2812,
-    "eval_samples_per_second": 74.777,
-    "eval_steps_per_second": 2.346,
-    "total_flos": 3.072509726495183e+18,
-    "train_loss": 1.3191132202148437,
-    "train_runtime": 2261.7516,
-    "train_samples_per_second": 36.074,
-    "train_steps_per_second": 0.111
+    "epoch": 15.0,
+    "eval_accuracy": 0.8568627450980392,
+    "eval_loss": 0.3784649968147278,
+    "eval_runtime": 25.0793,
+    "eval_samples_per_second": 81.342,
+    "eval_steps_per_second": 2.552,
+    "total_flos": 4.700892636385505e+18,
+    "train_loss": 0.7059793472290039,
+    "train_runtime": 3152.3203,
+    "train_samples_per_second": 38.824,
+    "train_steps_per_second": 0.081
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 9.8,
-    "eval_accuracy": 0.7700980392156863,
-    "eval_loss": 0.6762834191322327,
-    "eval_runtime": 27.2812,
-    "eval_samples_per_second": 74.777,
-    "eval_steps_per_second": 2.346
+    "epoch": 15.0,
+    "eval_accuracy": 0.8568627450980392,
+    "eval_loss": 0.3784649968147278,
+    "eval_runtime": 25.0793,
+    "eval_samples_per_second": 81.342,
+    "eval_steps_per_second": 2.552
 }

runs/Jan25_03-08-49_173fbb9c3dc5/events.out.tfevents.1706155344.173fbb9c3dc5.3455.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e928d6d3f08ac14fc46e55c05215b33f65006a2e4178439b77125e41794eb08
+size 411

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 9.8,
-    "total_flos": 3.072509726495183e+18,
-    "train_loss": 1.3191132202148437,
-    "train_runtime": 2261.7516,
-    "train_samples_per_second": 36.074,
-    "train_steps_per_second": 0.111
+    "epoch": 15.0,
+    "total_flos": 4.700892636385505e+18,
+    "train_loss": 0.7059793472290039,
+    "train_runtime": 3152.3203,
+    "train_samples_per_second": 38.824,
+    "train_steps_per_second": 0.081
 }

trainer_state.json

@@ -1,268 +1,265 @@
 {
-  "best_metric": 0.7700980392156863,
-  "best_model_checkpoint": "resnet-101-finetuned-CivilEng11k/checkpoint-229",
-  "epoch": 9.803921568627452,
+  "best_metric": 0.8568627450980392,
+  "best_model_checkpoint": "resnet-101-finetuned-CivilEng11k/checkpoint-238",
+  "epoch": 15.0,
   "eval_steps": 500,
-  "global_step": 250,
+  "global_step": 255,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.39,
-      "learning_rate": 4e-05,
-      "loss": 2.6157,
-      "step": 10
+      "epoch": 0.88,
+      "learning_rate": 0.00017307692307692304,
+      "loss": 2.6258,
+      "step": 15
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 8e-05,
-      "loss": 2.5769,
-      "step": 20
+      "epoch": 1.0,
+      "eval_accuracy": 0.5504901960784314,
+      "eval_loss": 2.5047667026519775,
+      "eval_runtime": 36.2965,
+      "eval_samples_per_second": 56.204,
+      "eval_steps_per_second": 1.763,
+      "step": 17
     },
     {
-      "epoch": 0.98,
-      "eval_accuracy": 0.5269607843137255,
-      "eval_loss": 2.49281907081604,
-      "eval_runtime": 40.5904,
-      "eval_samples_per_second": 50.258,
-      "eval_steps_per_second": 1.577,
-      "step": 25
-    },
-    {
-      "epoch": 1.18,
-      "learning_rate": 9.777777777777778e-05,
-      "loss": 2.4933,
+      "epoch": 1.76,
+      "learning_rate": 0.0002947598253275109,
+      "loss": 2.2924,
       "step": 30
     },
-    {
-      "epoch": 1.57,
-      "learning_rate": 9.333333333333334e-05,
-      "loss": 2.3726,
-      "step": 40
-    },
-    {
-      "epoch": 1.96,
-      "learning_rate": 8.888888888888889e-05,
-      "loss": 2.2271,
-      "step": 50
-    },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.5284313725490196,
-      "eval_loss": 2.184375047683716,
-      "eval_runtime": 41.3607,
-      "eval_samples_per_second": 49.322,
-      "eval_steps_per_second": 1.547,
+      "eval_accuracy": 0.5274509803921569,
+      "eval_loss": 1.4727771282196045,
+      "eval_runtime": 37.3662,
+      "eval_samples_per_second": 54.595,
+      "eval_steps_per_second": 1.713,
+      "step": 34
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.00027510917030567684,
+      "loss": 1.452,
+      "step": 45
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.6950980392156862,
+      "eval_loss": 1.0708895921707153,
+      "eval_runtime": 36.9842,
+      "eval_samples_per_second": 55.159,
+      "eval_steps_per_second": 1.73,
       "step": 51
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 8.444444444444444e-05,
-      "loss": 2.004,
+      "epoch": 3.53,
+      "learning_rate": 0.0002554585152838428,
+      "loss": 1.0363,
       "step": 60
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 8e-05,
-      "loss": 1.6261,
-      "step": 70
+      "epoch": 4.0,
+      "eval_accuracy": 0.7710784313725491,
+      "eval_loss": 0.6555339694023132,
+      "eval_runtime": 36.72,
+      "eval_samples_per_second": 55.556,
+      "eval_steps_per_second": 1.743,
+      "step": 68
     },
     {
-      "epoch": 2.98,
-      "eval_accuracy": 0.5269607843137255,
-      "eval_loss": 1.4098289012908936,
-      "eval_runtime": 40.6734,
-      "eval_samples_per_second": 50.156,
-      "eval_steps_per_second": 1.574,
-      "step": 76
+      "epoch": 4.41,
+      "learning_rate": 0.0002358078602620087,
+      "loss": 0.7086,
+      "step": 75
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 7.555555555555556e-05,
-      "loss": 1.4818,
-      "step": 80
+      "epoch": 5.0,
+      "eval_accuracy": 0.8122549019607843,
+      "eval_loss": 0.4875487685203552,
+      "eval_runtime": 37.4307,
+      "eval_samples_per_second": 54.501,
+      "eval_steps_per_second": 1.71,
+      "step": 85
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 7.111111111111112e-05,
-      "loss": 1.3531,
+      "epoch": 5.29,
+      "learning_rate": 0.00021615720524017465,
+      "loss": 0.5185,
       "step": 90
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 6.666666666666667e-05,
-      "loss": 1.2715,
-      "step": 100
-    },
-    {
-      "epoch": 4.0,
-      "eval_accuracy": 0.5799019607843138,
-      "eval_loss": 1.204033613204956,
-      "eval_runtime": 41.3797,
-      "eval_samples_per_second": 49.3,
-      "eval_steps_per_second": 1.547,
+      "epoch": 6.0,
+      "eval_accuracy": 0.8357843137254902,
+      "eval_loss": 0.43364840745925903,
+      "eval_runtime": 37.5452,
+      "eval_samples_per_second": 54.334,
+      "eval_steps_per_second": 1.705,
       "step": 102
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 6.222222222222222e-05,
-      "loss": 1.1843,
-      "step": 110
+      "epoch": 6.18,
+      "learning_rate": 0.00019650655021834058,
+      "loss": 0.4238,
+      "step": 105
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 5.7777777777777776e-05,
-      "loss": 1.1368,
-      "step": 120
+      "epoch": 7.0,
+      "eval_accuracy": 0.8509803921568627,
+      "eval_loss": 0.3962140381336212,
+      "eval_runtime": 39.9756,
+      "eval_samples_per_second": 51.031,
+      "eval_steps_per_second": 1.601,
+      "step": 119
     },
     {
-      "epoch": 4.98,
-      "eval_accuracy": 0.6852941176470588,
-      "eval_loss": 1.0043939352035522,
-      "eval_runtime": 40.9737,
-      "eval_samples_per_second": 49.788,
-      "eval_steps_per_second": 1.562,
-      "step": 127
+      "epoch": 7.06,
+      "learning_rate": 0.00017685589519650654,
+      "loss": 0.3846,
+      "step": 120
     },
     {
-      "epoch": 5.1,
-      "learning_rate": 5.333333333333333e-05,
-      "loss": 1.0742,
-      "step": 130
+      "epoch": 7.94,
+      "learning_rate": 0.00015720524017467246,
+      "loss": 0.3436,
+      "step": 135
     },
     {
-      "epoch": 5.49,
-      "learning_rate": 4.888888888888889e-05,
-      "loss": 0.9825,
-      "step": 140
+      "epoch": 8.0,
+      "eval_accuracy": 0.8514705882352941,
+      "eval_loss": 0.3823707103729248,
+      "eval_runtime": 40.0175,
+      "eval_samples_per_second": 50.978,
+      "eval_steps_per_second": 1.599,
+      "step": 136
     },
     {
-      "epoch": 5.88,
-      "learning_rate": 4.4444444444444447e-05,
-      "loss": 0.9366,
+      "epoch": 8.82,
+      "learning_rate": 0.00013755458515283842,
+      "loss": 0.3185,
       "step": 150
     },
     {
-      "epoch": 6.0,
-      "eval_accuracy": 0.7455882352941177,
-      "eval_loss": 0.8462627530097961,
-      "eval_runtime": 40.2496,
-      "eval_samples_per_second": 50.684,
-      "eval_steps_per_second": 1.59,
+      "epoch": 9.0,
+      "eval_accuracy": 0.8514705882352941,
+      "eval_loss": 0.38265544176101685,
+      "eval_runtime": 39.6707,
+      "eval_samples_per_second": 51.423,
+      "eval_steps_per_second": 1.613,
       "step": 153
     },
     {
-      "epoch": 6.27,
-      "learning_rate": 4e-05,
-      "loss": 0.899,
-      "step": 160
+      "epoch": 9.71,
+      "learning_rate": 0.00011790393013100435,
+      "loss": 0.3116,
+      "step": 165
     },
     {
-      "epoch": 6.67,
-      "learning_rate": 3.555555555555556e-05,
-      "loss": 0.8249,
+      "epoch": 10.0,
+      "eval_accuracy": 0.8534313725490196,
+      "eval_loss": 0.37894687056541443,
+      "eval_runtime": 38.7692,
+      "eval_samples_per_second": 52.619,
+      "eval_steps_per_second": 1.651,
       "step": 170
     },
     {
-      "epoch": 6.98,
-      "eval_accuracy": 0.7686274509803922,
-      "eval_loss": 0.7511501908302307,
-      "eval_runtime": 47.4571,
-      "eval_samples_per_second": 42.986,
-      "eval_steps_per_second": 1.349,
-      "step": 178
-    },
-    {
-      "epoch": 7.06,
-      "learning_rate": 3.111111111111111e-05,
-      "loss": 0.8197,
+      "epoch": 10.59,
+      "learning_rate": 9.825327510917029e-05,
+      "loss": 0.2983,
       "step": 180
     },
     {
-      "epoch": 7.45,
-      "learning_rate": 2.6666666666666667e-05,
-      "loss": 0.7562,
-      "step": 190
+      "epoch": 11.0,
+      "eval_accuracy": 0.8558823529411764,
+      "eval_loss": 0.3759206235408783,
+      "eval_runtime": 38.4006,
+      "eval_samples_per_second": 53.124,
+      "eval_steps_per_second": 1.667,
+      "step": 187
     },
     {
-      "epoch": 7.84,
-      "learning_rate": 2.2222222222222223e-05,
-      "loss": 0.7635,
-      "step": 200
+      "epoch": 11.47,
+      "learning_rate": 7.860262008733623e-05,
+      "loss": 0.2703,
+      "step": 195
     },
     {
-      "epoch": 8.0,
-      "eval_accuracy": 0.7612745098039215,
-      "eval_loss": 0.7078944444656372,
-      "eval_runtime": 40.9903,
-      "eval_samples_per_second": 49.768,
-      "eval_steps_per_second": 1.561,
+      "epoch": 12.0,
+      "eval_accuracy": 0.8480392156862745,
+      "eval_loss": 0.3838962912559509,
+      "eval_runtime": 38.2843,
+      "eval_samples_per_second": 53.286,
+      "eval_steps_per_second": 1.672,
       "step": 204
     },
     {
-      "epoch": 8.24,
-      "learning_rate": 1.777777777777778e-05,
-      "loss": 0.7415,
+      "epoch": 12.35,
+      "learning_rate": 5.8951965065502174e-05,
+      "loss": 0.2618,
       "step": 210
     },
     {
-      "epoch": 8.63,
-      "learning_rate": 1.3333333333333333e-05,
-      "loss": 0.7213,
-      "step": 220
+      "epoch": 13.0,
+      "eval_accuracy": 0.8539215686274509,
+      "eval_loss": 0.38307902216911316,
+      "eval_runtime": 38.0757,
+      "eval_samples_per_second": 53.577,
+      "eval_steps_per_second": 1.681,
+      "step": 221
     },
     {
-      "epoch": 8.98,
-      "eval_accuracy": 0.7700980392156863,
-      "eval_loss": 0.6762834191322327,
-      "eval_runtime": 40.8595,
-      "eval_samples_per_second": 49.927,
-      "eval_steps_per_second": 1.566,
-      "step": 229
+      "epoch": 13.24,
+      "learning_rate": 3.9301310043668116e-05,
+      "loss": 0.2613,
+      "step": 225
     },
     {
-      "epoch": 9.02,
-      "learning_rate": 8.88888888888889e-06,
-      "loss": 0.6951,
-      "step": 230
+      "epoch": 14.0,
+      "eval_accuracy": 0.8568627450980392,
+      "eval_loss": 0.3784649968147278,
+      "eval_runtime": 38.1613,
+      "eval_samples_per_second": 53.457,
+      "eval_steps_per_second": 1.677,
+      "step": 238
     },
     {
-      "epoch": 9.41,
-      "learning_rate": 4.444444444444445e-06,
-      "loss": 0.7295,
+      "epoch": 14.12,
+      "learning_rate": 1.9650655021834058e-05,
+      "loss": 0.2514,
       "step": 240
     },
     {
-      "epoch": 9.8,
+      "epoch": 15.0,
       "learning_rate": 0.0,
-      "loss": 0.6905,
-      "step": 250
-    },
-    {
-      "epoch": 9.8,
-      "eval_accuracy": 0.7676470588235295,
-      "eval_loss": 0.6683045625686646,
-      "eval_runtime": 39.5198,
-      "eval_samples_per_second": 51.62,
-      "eval_steps_per_second": 1.619,
-      "step": 250
-    },
-    {
-      "epoch": 9.8,
-      "step": 250,
-      "total_flos": 3.072509726495183e+18,
-      "train_loss": 1.3191132202148437,
-      "train_runtime": 2261.7516,
-      "train_samples_per_second": 36.074,
-      "train_steps_per_second": 0.111
+      "loss": 0.2428,
+      "step": 255
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.8470588235294118,
+      "eval_loss": 0.39441215991973877,
+      "eval_runtime": 38.4101,
+      "eval_samples_per_second": 53.111,
+      "eval_steps_per_second": 1.666,
+      "step": 255
+    },
+    {
+      "epoch": 15.0,
+      "step": 255,
+      "total_flos": 4.700892636385505e+18,
+      "train_loss": 0.7059793472290039,
+      "train_runtime": 3152.3203,
+      "train_samples_per_second": 38.824,
+      "train_steps_per_second": 0.081
     }
   ],
-  "logging_steps": 10,
-  "max_steps": 250,
-  "num_train_epochs": 10,
+  "logging_steps": 15,
+  "max_steps": 255,
+  "num_train_epochs": 15,
   "save_steps": 500,
-  "total_flos": 3.072509726495183e+18,
+  "total_flos": 4.700892636385505e+18,
   "trial_name": null,
   "trial_params": null
 }