🍻 cheers

README.md

@@ -24,10 +24,10 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.8171064604185623
+      value: 0.827115559599636
     - name: F1
       type: f1
-      value: 0.7841031149301826
+      value: 0.7978723404255319
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -37,12 +37,12 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the stool-image dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.4538
-- Auroc: 0.8897
-- Accuracy: 0.8171
-- Sensitivity: 0.8111
-- Specificty: 0.8213
-- F1: 0.7841
+- Loss: 0.3761
+- Auroc: 0.9189
+- Accuracy: 0.8271
+- Sensitivity: 0.8333
+- Specificty: 0.8228
+- F1: 0.7979
 
 ## Model description
 
@@ -67,19 +67,15 @@ The following hyperparameters were used during training:
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 100
+- num_epochs: 2
 - mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Auroc  | Accuracy | Sensitivity | Specificty | F1     |
 |:-------------:|:-----:|:----:|:---------------:|:------:|:--------:|:-----------:|:----------:|:------:|
-| 0.5303        | 0.98  | 100  | 0.4327          | 0.8826 | 0.7942   | 0.7191      | 0.8607     | 0.7665 |
-| 0.3909        | 1.96  | 200  | 0.5196          | 0.8675 | 0.8047   | 0.8539      | 0.7612     | 0.8042 |
-| 0.5328        | 2.94  | 300  | 0.4421          | 0.8864 | 0.8074   | 0.7528      | 0.8557     | 0.7859 |
-| 0.4834        | 3.92  | 400  | 0.4721          | 0.8596 | 0.7757   | 0.7135      | 0.8308     | 0.7493 |
-| 0.4209        | 4.9   | 500  | 0.4797          | 0.8625 | 0.7863   | 0.6798      | 0.8806     | 0.7492 |
-| 0.4567        | 5.88  | 600  | 0.5150          | 0.8688 | 0.7942   | 0.6011      | 0.9652     | 0.7329 |
+| 0.3755        | 0.98  | 100  | 0.4455          | 0.8677 | 0.7757   | 0.6966      | 0.8458     | 0.7447 |
+| 0.309         | 1.96  | 200  | 0.4360          | 0.8754 | 0.8127   | 0.7528      | 0.8657     | 0.7906 |
 
 
 ### Framework versions

all_results.json

@@ -1,16 +1,17 @@
 {
-    "epoch": 5.88,
-    "eval_accuracy": 0.8171064604185623,
-    "eval_auroc": 0.8896644410203732,
-    "eval_f1": 0.7841031149301826,
-    "eval_loss": 0.45378199219703674,
-    "eval_runtime": 63.9212,
-    "eval_samples_per_second": 17.193,
-    "eval_sensitivity": 0.8111111111111111,
-    "eval_specificty": 0.8212634822804314,
-    "eval_steps_per_second": 2.159,
-    "train_loss": 0.4638279656569163,
-    "train_runtime": 457.0802,
-    "train_samples_per_second": 355.517,
-    "train_steps_per_second": 22.316
+    "epoch": 2.0,
+    "eval_accuracy": 0.9320388349514563,
+    "eval_auroc": 0.9441448382126347,
+    "eval_eval_specificty": 0.8863636363636364,
+    "eval_f1": 0.9176470588235294,
+    "eval_loss": 0.37609395384788513,
+    "eval_runtime": 70.0505,
+    "eval_samples_per_second": 15.689,
+    "eval_sensitivity": 0.8863636363636364,
+    "eval_specificty": 0.8228043143297381,
+    "eval_steps_per_second": 1.97,
+    "train_loss": 0.4438221442933176,
+    "train_runtime": 152.958,
+    "train_samples_per_second": 21.248,
+    "train_steps_per_second": 1.334
 }

eval_results.json

@@ -1,12 +1,13 @@
 {
-    "epoch": 5.88,
-    "eval_accuracy": 0.8171064604185623,
-    "eval_auroc": 0.8896644410203732,
-    "eval_f1": 0.7841031149301826,
-    "eval_loss": 0.45378199219703674,
-    "eval_runtime": 63.9212,
-    "eval_samples_per_second": 17.193,
-    "eval_sensitivity": 0.8111111111111111,
-    "eval_specificty": 0.8212634822804314,
-    "eval_steps_per_second": 2.159
+    "epoch": 2.0,
+    "eval_accuracy": 0.9320388349514563,
+    "eval_auroc": 0.9441448382126347,
+    "eval_eval_specificty": 0.8863636363636364,
+    "eval_f1": 0.9176470588235294,
+    "eval_loss": 0.37609395384788513,
+    "eval_runtime": 70.0505,
+    "eval_samples_per_second": 15.689,
+    "eval_sensitivity": 0.8863636363636364,
+    "eval_specificty": 0.8228043143297381,
+    "eval_steps_per_second": 1.97
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bce917709538915003c0c8fe4efef00dae13bd7d04517b013edb5b0d28d71e17
+oid sha256:49e65a419570bcc528d2259195bf6cef77c24c492a4522dbe45e38c726a6a8a1
 size 343223968

runs/Jan05_17-26-06_DESKTOP-BDBS5RV/events.out.tfevents.1704443688.DESKTOP-BDBS5RV

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:09b62303b6c67ec8004a1373b7d8162f85881f93156735bc5dc4c83946ad955c
-size 40
+oid sha256:371e35b19c483e082c64794a04cb3d55c6f3e5fe391e4302a4aa9cb8102a4a95
+size 567

runs/Jan06_12-28-05_DESKTOP-BDBS5RV/events.out.tfevents.1704511685.DESKTOP-BDBS5RV

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7fe0d6bce62a3e37322c977b8cff2445b9932eb02a04b027b55d5b5cae9e816a
+size 8853

runs/Jan06_12-28-05_DESKTOP-BDBS5RV/events.out.tfevents.1704511909.DESKTOP-BDBS5RV

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0e59688d86c65619588f4e21bee0cdc0790862a16c985c09988785780ba640c
+size 40

train_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 5.88,
-    "train_loss": 0.4638279656569163,
-    "train_runtime": 457.0802,
-    "train_samples_per_second": 355.517,
-    "train_steps_per_second": 22.316
+    "epoch": 2.0,
+    "train_loss": 0.4438221442933176,
+    "train_runtime": 152.958,
+    "train_samples_per_second": 21.248,
+    "train_steps_per_second": 1.334
 }

trainer_state.json

@@ -1,467 +1,175 @@
 {
-  "best_metric": 0.8863966683436749,
-  "best_model_checkpoint": "./stool-condition-classification\\checkpoint-300",
-  "epoch": 5.882352941176471,
+  "best_metric": 0.8754262395885739,
+  "best_model_checkpoint": "./stool-condition-classification\\checkpoint-200",
+  "epoch": 2.0,
   "eval_steps": 100,
-  "global_step": 600,
+  "global_step": 204,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.1,
-      "learning_rate": 0.00019980392156862745,
-      "loss": 0.6109,
+      "learning_rate": 0.00019019607843137254,
+      "loss": 0.6003,
       "step": 10
     },
     {
       "epoch": 0.2,
-      "learning_rate": 0.00019960784313725492,
-      "loss": 0.5518,
+      "learning_rate": 0.0001803921568627451,
+      "loss": 0.5057,
       "step": 20
     },
     {
       "epoch": 0.29,
-      "learning_rate": 0.00019941176470588236,
-      "loss": 0.4854,
+      "learning_rate": 0.00017058823529411766,
+      "loss": 0.492,
       "step": 30
     },
     {
       "epoch": 0.39,
-      "learning_rate": 0.0001992156862745098,
-      "loss": 0.5984,
+      "learning_rate": 0.00016078431372549022,
+      "loss": 0.5363,
       "step": 40
     },
     {
       "epoch": 0.49,
-      "learning_rate": 0.00019901960784313727,
-      "loss": 0.5832,
+      "learning_rate": 0.00015098039215686275,
+      "loss": 0.5434,
       "step": 50
     },
     {
       "epoch": 0.59,
-      "learning_rate": 0.00019882352941176472,
-      "loss": 0.5199,
+      "learning_rate": 0.0001411764705882353,
+      "loss": 0.4924,
       "step": 60
     },
     {
       "epoch": 0.69,
-      "learning_rate": 0.00019862745098039218,
-      "loss": 0.5347,
+      "learning_rate": 0.00013137254901960784,
+      "loss": 0.4675,
       "step": 70
     },
     {
       "epoch": 0.78,
-      "learning_rate": 0.00019843137254901963,
-      "loss": 0.5159,
+      "learning_rate": 0.00012156862745098039,
+      "loss": 0.4547,
       "step": 80
     },
     {
       "epoch": 0.88,
-      "learning_rate": 0.00019823529411764707,
-      "loss": 0.5069,
+      "learning_rate": 0.00011176470588235294,
+      "loss": 0.4565,
       "step": 90
     },
     {
       "epoch": 0.98,
-      "learning_rate": 0.00019803921568627454,
-      "loss": 0.5303,
+      "learning_rate": 0.00010196078431372549,
+      "loss": 0.3755,
       "step": 100
     },
     {
       "epoch": 0.98,
-      "eval_accuracy": 0.7941952506596306,
-      "eval_auroc": 0.8826094247861815,
-      "eval_f1": 0.7664670658682635,
-      "eval_loss": 0.43268465995788574,
-      "eval_runtime": 19.5335,
-      "eval_samples_per_second": 19.403,
-      "eval_sensitivity": 0.7191011235955056,
-      "eval_specificty": 0.8606965174129353,
-      "eval_steps_per_second": 2.457,
+      "eval_accuracy": 0.7757255936675461,
+      "eval_auroc": 0.8676561015148975,
+      "eval_f1": 0.7447447447447447,
+      "eval_loss": 0.4455398917198181,
+      "eval_runtime": 19.6185,
+      "eval_samples_per_second": 19.318,
+      "eval_sensitivity": 0.6966292134831461,
+      "eval_specificty": 0.845771144278607,
+      "eval_steps_per_second": 2.447,
       "step": 100
     },
     {
       "epoch": 1.08,
-      "learning_rate": 0.00019784313725490198,
-      "loss": 0.514,
+      "learning_rate": 9.215686274509804e-05,
+      "loss": 0.4789,
       "step": 110
     },
     {
       "epoch": 1.18,
-      "learning_rate": 0.00019764705882352942,
-      "loss": 0.3919,
+      "learning_rate": 8.23529411764706e-05,
+      "loss": 0.3842,
       "step": 120
     },
     {
       "epoch": 1.27,
-      "learning_rate": 0.0001974509803921569,
-      "loss": 0.5246,
+      "learning_rate": 7.254901960784314e-05,
+      "loss": 0.5317,
       "step": 130
     },
     {
       "epoch": 1.37,
-      "learning_rate": 0.00019725490196078433,
-      "loss": 0.5262,
+      "learning_rate": 6.274509803921569e-05,
+      "loss": 0.4017,
       "step": 140
     },
     {
       "epoch": 1.47,
-      "learning_rate": 0.00019705882352941177,
-      "loss": 0.4092,
+      "learning_rate": 5.294117647058824e-05,
+      "loss": 0.3627,
       "step": 150
     },
     {
       "epoch": 1.57,
-      "learning_rate": 0.00019686274509803922,
-      "loss": 0.6038,
+      "learning_rate": 4.313725490196079e-05,
+      "loss": 0.4424,
       "step": 160
     },
     {
       "epoch": 1.67,
-      "learning_rate": 0.00019666666666666666,
-      "loss": 0.5079,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.3926,
       "step": 170
     },
     {
       "epoch": 1.76,
-      "learning_rate": 0.00019647058823529413,
-      "loss": 0.4992,
+      "learning_rate": 2.3529411764705884e-05,
+      "loss": 0.4004,
       "step": 180
     },
     {
       "epoch": 1.86,
-      "learning_rate": 0.00019627450980392157,
-      "loss": 0.4788,
+      "learning_rate": 1.3725490196078432e-05,
+      "loss": 0.3161,
       "step": 190
     },
     {
       "epoch": 1.96,
-      "learning_rate": 0.000196078431372549,
-      "loss": 0.3909,
+      "learning_rate": 3.92156862745098e-06,
+      "loss": 0.309,
       "step": 200
     },
     {
       "epoch": 1.96,
-      "eval_accuracy": 0.8047493403693932,
-      "eval_auroc": 0.8674604505562076,
-      "eval_f1": 0.8042328042328041,
-      "eval_loss": 0.5196136832237244,
-      "eval_runtime": 19.2864,
-      "eval_samples_per_second": 19.651,
-      "eval_sensitivity": 0.8539325842696629,
-      "eval_specificty": 0.7611940298507462,
-      "eval_steps_per_second": 2.489,
-      "step": 200
-    },
-    {
-      "epoch": 2.06,
-      "learning_rate": 0.00019588235294117648,
-      "loss": 0.4421,
-      "step": 210
-    },
-    {
-      "epoch": 2.16,
-      "learning_rate": 0.00019568627450980392,
-      "loss": 0.4479,
-      "step": 220
-    },
-    {
-      "epoch": 2.25,
-      "learning_rate": 0.00019549019607843136,
-      "loss": 0.4655,
-      "step": 230
-    },
-    {
-      "epoch": 2.35,
-      "learning_rate": 0.00019529411764705883,
-      "loss": 0.4479,
-      "step": 240
-    },
-    {
-      "epoch": 2.45,
-      "learning_rate": 0.00019509803921568628,
-      "loss": 0.3894,
-      "step": 250
-    },
-    {
-      "epoch": 2.55,
-      "learning_rate": 0.00019490196078431372,
-      "loss": 0.4315,
-      "step": 260
-    },
-    {
-      "epoch": 2.65,
-      "learning_rate": 0.0001947058823529412,
-      "loss": 0.4783,
-      "step": 270
-    },
-    {
-      "epoch": 2.75,
-      "learning_rate": 0.00019450980392156863,
-      "loss": 0.3792,
-      "step": 280
-    },
-    {
-      "epoch": 2.84,
-      "learning_rate": 0.0001943137254901961,
-      "loss": 0.5301,
-      "step": 290
-    },
-    {
-      "epoch": 2.94,
-      "learning_rate": 0.00019411764705882354,
-      "loss": 0.5328,
-      "step": 300
-    },
-    {
-      "epoch": 2.94,
-      "eval_accuracy": 0.8073878627968337,
-      "eval_auroc": 0.8863966683436749,
-      "eval_f1": 0.7859237536656891,
-      "eval_loss": 0.4421083927154541,
-      "eval_runtime": 19.1397,
-      "eval_samples_per_second": 19.802,
+      "eval_accuracy": 0.8126649076517151,
+      "eval_auroc": 0.8754262395885739,
+      "eval_f1": 0.7905604719764012,
+      "eval_loss": 0.4360436201095581,
+      "eval_runtime": 19.2175,
+      "eval_samples_per_second": 19.722,
       "eval_sensitivity": 0.7528089887640449,
-      "eval_specificty": 0.8557213930348259,
-      "eval_steps_per_second": 2.508,
-      "step": 300
-    },
-    {
-      "epoch": 3.04,
-      "learning_rate": 0.00019392156862745098,
-      "loss": 0.4437,
-      "step": 310
-    },
-    {
-      "epoch": 3.14,
-      "learning_rate": 0.00019372549019607845,
-      "loss": 0.3859,
-      "step": 320
-    },
-    {
-      "epoch": 3.24,
-      "learning_rate": 0.0001935294117647059,
-      "loss": 0.4995,
-      "step": 330
-    },
-    {
-      "epoch": 3.33,
-      "learning_rate": 0.00019333333333333333,
-      "loss": 0.4282,
-      "step": 340
-    },
-    {
-      "epoch": 3.43,
-      "learning_rate": 0.0001931372549019608,
-      "loss": 0.4742,
-      "step": 350
-    },
-    {
-      "epoch": 3.53,
-      "learning_rate": 0.00019294117647058825,
-      "loss": 0.4219,
-      "step": 360
-    },
-    {
-      "epoch": 3.63,
-      "learning_rate": 0.0001927450980392157,
-      "loss": 0.4531,
-      "step": 370
-    },
-    {
-      "epoch": 3.73,
-      "learning_rate": 0.00019254901960784316,
-      "loss": 0.545,
-      "step": 380
-    },
-    {
-      "epoch": 3.82,
-      "learning_rate": 0.0001923529411764706,
-      "loss": 0.4216,
-      "step": 390
-    },
-    {
-      "epoch": 3.92,
-      "learning_rate": 0.00019215686274509807,
-      "loss": 0.4834,
-      "step": 400
-    },
-    {
-      "epoch": 3.92,
-      "eval_accuracy": 0.7757255936675461,
-      "eval_auroc": 0.8596483872770976,
-      "eval_f1": 0.7492625368731564,
-      "eval_loss": 0.4720683693885803,
-      "eval_runtime": 19.0669,
-      "eval_samples_per_second": 19.877,
-      "eval_sensitivity": 0.7134831460674157,
-      "eval_specificty": 0.8308457711442786,
-      "eval_steps_per_second": 2.517,
-      "step": 400
-    },
-    {
-      "epoch": 4.02,
-      "learning_rate": 0.0001919607843137255,
-      "loss": 0.3281,
-      "step": 410
-    },
-    {
-      "epoch": 4.12,
-      "learning_rate": 0.00019176470588235295,
-      "loss": 0.5168,
-      "step": 420
-    },
-    {
-      "epoch": 4.22,
-      "learning_rate": 0.00019156862745098042,
-      "loss": 0.5219,
-      "step": 430
-    },
-    {
-      "epoch": 4.31,
-      "learning_rate": 0.00019137254901960786,
-      "loss": 0.4131,
-      "step": 440
-    },
-    {
-      "epoch": 4.41,
-      "learning_rate": 0.0001911764705882353,
-      "loss": 0.4494,
-      "step": 450
-    },
-    {
-      "epoch": 4.51,
-      "learning_rate": 0.00019098039215686277,
-      "loss": 0.4206,
-      "step": 460
-    },
-    {
-      "epoch": 4.61,
-      "learning_rate": 0.00019078431372549022,
-      "loss": 0.3932,
-      "step": 470
-    },
-    {
-      "epoch": 4.71,
-      "learning_rate": 0.00019058823529411766,
-      "loss": 0.4607,
-      "step": 480
-    },
-    {
-      "epoch": 4.8,
-      "learning_rate": 0.0001903921568627451,
-      "loss": 0.3892,
-      "step": 490
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 0.00019019607843137254,
-      "loss": 0.4209,
-      "step": 500
-    },
-    {
-      "epoch": 4.9,
-      "eval_accuracy": 0.7862796833773087,
-      "eval_auroc": 0.8624853261780983,
-      "eval_f1": 0.7492260061919503,
-      "eval_loss": 0.47969380021095276,
-      "eval_runtime": 18.951,
-      "eval_samples_per_second": 19.999,
-      "eval_sensitivity": 0.6797752808988764,
-      "eval_specificty": 0.8805970149253731,
-      "eval_steps_per_second": 2.533,
-      "step": 500
-    },
-    {
-      "epoch": 5.0,
-      "learning_rate": 0.00019,
-      "loss": 0.3681,
-      "step": 510
-    },
-    {
-      "epoch": 5.1,
-      "learning_rate": 0.00018980392156862745,
-      "loss": 0.397,
-      "step": 520
-    },
-    {
-      "epoch": 5.2,
-      "learning_rate": 0.0001896078431372549,
-      "loss": 0.3833,
-      "step": 530
-    },
-    {
-      "epoch": 5.29,
-      "learning_rate": 0.00018941176470588236,
-      "loss": 0.4202,
-      "step": 540
-    },
-    {
-      "epoch": 5.39,
-      "learning_rate": 0.0001892156862745098,
-      "loss": 0.4304,
-      "step": 550
-    },
-    {
-      "epoch": 5.49,
-      "learning_rate": 0.00018901960784313725,
-      "loss": 0.3883,
-      "step": 560
-    },
-    {
-      "epoch": 5.59,
-      "learning_rate": 0.00018882352941176472,
-      "loss": 0.4737,
-      "step": 570
-    },
-    {
-      "epoch": 5.69,
-      "learning_rate": 0.00018862745098039216,
-      "loss": 0.4661,
-      "step": 580
-    },
-    {
-      "epoch": 5.78,
-      "learning_rate": 0.0001884313725490196,
-      "loss": 0.3469,
-      "step": 590
-    },
-    {
-      "epoch": 5.88,
-      "learning_rate": 0.00018823529411764707,
-      "loss": 0.4567,
-      "step": 600
-    },
-    {
-      "epoch": 5.88,
-      "eval_accuracy": 0.7941952506596306,
-      "eval_auroc": 0.8687741069931243,
-      "eval_f1": 0.7328767123287672,
-      "eval_loss": 0.514951765537262,
-      "eval_runtime": 18.7983,
-      "eval_samples_per_second": 20.161,
-      "eval_sensitivity": 0.601123595505618,
-      "eval_specificty": 0.9651741293532339,
-      "eval_steps_per_second": 2.553,
-      "step": 600
+      "eval_specificty": 0.8656716417910447,
+      "eval_steps_per_second": 2.498,
+      "step": 200
     },
     {
-      "epoch": 5.88,
-      "step": 600,
-      "total_flos": 7.412108806635725e+17,
-      "train_loss": 0.4638279656569163,
-      "train_runtime": 457.0802,
-      "train_samples_per_second": 355.517,
-      "train_steps_per_second": 22.316
+      "epoch": 2.0,
+      "step": 204,
+      "total_flos": 2.51848966247424e+17,
+      "train_loss": 0.4438221442933176,
+      "train_runtime": 152.958,
+      "train_samples_per_second": 21.248,
+      "train_steps_per_second": 1.334
     }
   ],
   "logging_steps": 10,
-  "max_steps": 10200,
+  "max_steps": 204,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 100,
+  "num_train_epochs": 2,
   "save_steps": 100,
-  "total_flos": 7.412108806635725e+17,
+  "total_flos": 2.51848966247424e+17,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:00021469e2fc2da6e352c64ba5c1eff722ad7949ad69e804b54677deb2c22f06
+oid sha256:1fca6169b13c98d2024034ca2a018b9c8536fee89f8afeb3b4867f7e487a0d5b
 size 4283