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: 0.3549
-- Accuracy: 0.8691
+- Loss: 0.3590
+- Accuracy: 0.8716
 
 ## Model description
 

all_results.json

@@ -1,13 +1,13 @@
 {
-    "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
+    "epoch": 9.8,
+    "eval_accuracy": 0.8715686274509804,
+    "eval_loss": 0.3589998185634613,
+    "eval_runtime": 34.1863,
+    "eval_samples_per_second": 59.673,
+    "eval_steps_per_second": 1.872,
+    "total_flos": 7.501244449451129e+18,
+    "train_loss": 0.7515069055557251,
+    "train_runtime": 2638.4717,
+    "train_samples_per_second": 30.923,
+    "train_steps_per_second": 0.095
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "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
+    "epoch": 9.8,
+    "eval_accuracy": 0.8715686274509804,
+    "eval_loss": 0.3589998185634613,
+    "eval_runtime": 34.1863,
+    "eval_samples_per_second": 59.673,
+    "eval_steps_per_second": 1.872
 }

runs/Feb02_03-29-40_ae322b9c3506/events.out.tfevents.1706847275.ae322b9c3506.3394.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d4f149d568f278686d72523c6682cd0f8c8c5e5643377e506e196ea43949239
+size 411

train_results.json

@@ -1,8 +1,8 @@
 {
-    "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
+    "epoch": 9.8,
+    "total_flos": 7.501244449451129e+18,
+    "train_loss": 0.7515069055557251,
+    "train_runtime": 2638.4717,
+    "train_samples_per_second": 30.923,
+    "train_steps_per_second": 0.095
 }

trainer_state.json

@@ -1,265 +1,268 @@
 {
-  "best_metric": 0.8568627450980392,
-  "best_model_checkpoint": "resnet-101-finetuned-CivilEng11k/checkpoint-238",
-  "epoch": 15.0,
+  "best_metric": 0.8715686274509804,
+  "best_model_checkpoint": "resnet-101-finetuned-CivilEng11k/checkpoint-204",
+  "epoch": 9.803921568627452,
   "eval_steps": 500,
-  "global_step": 255,
+  "global_step": 250,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.88,
-      "learning_rate": 0.00017307692307692304,
-      "loss": 2.6258,
-      "step": 15
+      "epoch": 0.39,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 2.6064,
+      "step": 10
     },
     {
-      "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": 0.78,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 2.4679,
+      "step": 20
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 0.0002947598253275109,
-      "loss": 2.2924,
+      "epoch": 0.98,
+      "eval_accuracy": 0.5367647058823529,
+      "eval_loss": 1.9448940753936768,
+      "eval_runtime": 41.2734,
+      "eval_samples_per_second": 49.426,
+      "eval_steps_per_second": 1.551,
+      "step": 25
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00029333333333333327,
+      "loss": 1.9417,
       "step": 30
     },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00028,
+      "loss": 1.3729,
+      "step": 40
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.0002666666666666666,
+      "loss": 1.2247,
+      "step": 50
+    },
     {
       "epoch": 2.0,
-      "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,
+      "eval_accuracy": 0.6799019607843138,
+      "eval_loss": 1.1000670194625854,
+      "eval_runtime": 42.9676,
+      "eval_samples_per_second": 47.478,
+      "eval_steps_per_second": 1.489,
       "step": 51
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 0.0002554585152838428,
-      "loss": 1.0363,
+      "epoch": 2.35,
+      "learning_rate": 0.00025333333333333333,
+      "loss": 1.062,
       "step": 60
     },
     {
-      "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": 2.75,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 0.8434,
+      "step": 70
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 0.0002358078602620087,
-      "loss": 0.7086,
-      "step": 75
+      "epoch": 2.98,
+      "eval_accuracy": 0.7887254901960784,
+      "eval_loss": 0.5851422548294067,
+      "eval_runtime": 40.7115,
+      "eval_samples_per_second": 50.109,
+      "eval_steps_per_second": 1.572,
+      "step": 76
     },
     {
-      "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": 3.14,
+      "learning_rate": 0.00022666666666666663,
+      "loss": 0.6665,
+      "step": 80
     },
     {
-      "epoch": 5.29,
-      "learning_rate": 0.00021615720524017465,
-      "loss": 0.5185,
+      "epoch": 3.53,
+      "learning_rate": 0.00021333333333333333,
+      "loss": 0.5464,
       "step": 90
     },
     {
-      "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": 3.92,
+      "learning_rate": 0.00019999999999999998,
+      "loss": 0.491,
+      "step": 100
     },
     {
-      "epoch": 6.18,
-      "learning_rate": 0.00019650655021834058,
-      "loss": 0.4238,
-      "step": 105
+      "epoch": 4.0,
+      "eval_accuracy": 0.8338235294117647,
+      "eval_loss": 0.43349361419677734,
+      "eval_runtime": 41.9829,
+      "eval_samples_per_second": 48.591,
+      "eval_steps_per_second": 1.524,
+      "step": 102
     },
     {
-      "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": 4.31,
+      "learning_rate": 0.00018666666666666666,
+      "loss": 0.4475,
+      "step": 110
     },
     {
-      "epoch": 7.06,
-      "learning_rate": 0.00017685589519650654,
-      "loss": 0.3846,
+      "epoch": 4.71,
+      "learning_rate": 0.0001733333333333333,
+      "loss": 0.4512,
       "step": 120
     },
     {
-      "epoch": 7.94,
-      "learning_rate": 0.00015720524017467246,
-      "loss": 0.3436,
-      "step": 135
+      "epoch": 4.98,
+      "eval_accuracy": 0.8401960784313726,
+      "eval_loss": 0.40097689628601074,
+      "eval_runtime": 40.6062,
+      "eval_samples_per_second": 50.239,
+      "eval_steps_per_second": 1.576,
+      "step": 127
     },
     {
-      "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": 5.1,
+      "learning_rate": 0.00015999999999999999,
+      "loss": 0.4392,
+      "step": 130
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 0.00014666666666666664,
+      "loss": 0.4147,
+      "step": 140
     },
     {
-      "epoch": 8.82,
-      "learning_rate": 0.00013755458515283842,
-      "loss": 0.3185,
+      "epoch": 5.88,
+      "learning_rate": 0.0001333333333333333,
+      "loss": 0.3891,
       "step": 150
     },
     {
-      "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,
+      "epoch": 6.0,
+      "eval_accuracy": 0.8544117647058823,
+      "eval_loss": 0.38171347975730896,
+      "eval_runtime": 43.7502,
+      "eval_samples_per_second": 46.628,
+      "eval_steps_per_second": 1.463,
       "step": 153
     },
     {
-      "epoch": 9.71,
-      "learning_rate": 0.00011790393013100435,
-      "loss": 0.3116,
-      "step": 165
+      "epoch": 6.27,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 0.3902,
+      "step": 160
     },
     {
-      "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,
+      "epoch": 6.67,
+      "learning_rate": 0.00010666666666666667,
+      "loss": 0.3533,
       "step": 170
     },
     {
-      "epoch": 10.59,
-      "learning_rate": 9.825327510917029e-05,
-      "loss": 0.2983,
+      "epoch": 6.98,
+      "eval_accuracy": 0.8602941176470589,
+      "eval_loss": 0.3679962158203125,
+      "eval_runtime": 38.6694,
+      "eval_samples_per_second": 52.755,
+      "eval_steps_per_second": 1.655,
+      "step": 178
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 9.333333333333333e-05,
+      "loss": 0.3553,
       "step": 180
     },
     {
-      "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": 7.45,
+      "learning_rate": 7.999999999999999e-05,
+      "loss": 0.352,
+      "step": 190
     },
     {
-      "epoch": 11.47,
-      "learning_rate": 7.860262008733623e-05,
-      "loss": 0.2703,
-      "step": 195
+      "epoch": 7.84,
+      "learning_rate": 6.666666666666666e-05,
+      "loss": 0.3407,
+      "step": 200
     },
     {
-      "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,
+      "epoch": 8.0,
+      "eval_accuracy": 0.8715686274509804,
+      "eval_loss": 0.3589998185634613,
+      "eval_runtime": 31.761,
+      "eval_samples_per_second": 64.23,
+      "eval_steps_per_second": 2.015,
       "step": 204
     },
     {
-      "epoch": 12.35,
-      "learning_rate": 5.8951965065502174e-05,
-      "loss": 0.2618,
+      "epoch": 8.24,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 0.351,
       "step": 210
     },
     {
-      "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": 8.63,
+      "learning_rate": 3.9999999999999996e-05,
+      "loss": 0.3244,
+      "step": 220
     },
     {
-      "epoch": 13.24,
-      "learning_rate": 3.9301310043668116e-05,
-      "loss": 0.2613,
-      "step": 225
+      "epoch": 8.98,
+      "eval_accuracy": 0.8661764705882353,
+      "eval_loss": 0.36056193709373474,
+      "eval_runtime": 34.9574,
+      "eval_samples_per_second": 58.357,
+      "eval_steps_per_second": 1.831,
+      "step": 229
     },
     {
-      "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": 9.02,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.3148,
+      "step": 230
     },
     {
-      "epoch": 14.12,
-      "learning_rate": 1.9650655021834058e-05,
-      "loss": 0.2514,
+      "epoch": 9.41,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.3156,
       "step": 240
     },
     {
-      "epoch": 15.0,
+      "epoch": 9.8,
       "learning_rate": 0.0,
-      "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
+      "loss": 0.3258,
+      "step": 250
+    },
+    {
+      "epoch": 9.8,
+      "eval_accuracy": 0.8691176470588236,
+      "eval_loss": 0.3549394905567169,
+      "eval_runtime": 38.198,
+      "eval_samples_per_second": 53.406,
+      "eval_steps_per_second": 1.675,
+      "step": 250
+    },
+    {
+      "epoch": 9.8,
+      "step": 250,
+      "total_flos": 7.501244449451129e+18,
+      "train_loss": 0.7515069055557251,
+      "train_runtime": 2638.4717,
+      "train_samples_per_second": 30.923,
+      "train_steps_per_second": 0.095
     }
   ],
-  "logging_steps": 15,
-  "max_steps": 255,
-  "num_train_epochs": 15,
+  "logging_steps": 10,
+  "max_steps": 250,
+  "num_train_epochs": 10,
   "save_steps": 500,
-  "total_flos": 4.700892636385505e+18,
+  "total_flos": 7.501244449451129e+18,
   "trial_name": null,
   "trial_params": null
 }