Model save

README.md

@@ -22,7 +22,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.981203007518797
+      value: 0.9887218045112782
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -32,8 +32,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [facebook/convnext-tiny-224](https://huggingface.co/facebook/convnext-tiny-224) on the imagefolder dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0595
-- Accuracy: 0.9812
+- Loss: 0.0732
+- Accuracy: 0.9887
 
 ## Model description
 
@@ -58,19 +58,30 @@ 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: 20
+- num_epochs: 50
 - mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Accuracy |
 |:-------------:|:-----:|:----:|:---------------:|:--------:|
-| 0.2926        | 2.94  | 100  | 0.1524          | 0.9436   |
-| 0.1844        | 5.88  | 200  | 0.1718          | 0.9436   |
-| 0.1189        | 8.82  | 300  | 0.0886          | 0.9662   |
-| 0.0682        | 11.76 | 400  | 0.0978          | 0.9699   |
-| 0.0439        | 14.71 | 500  | 0.0812          | 0.9737   |
-| 0.0544        | 17.65 | 600  | 0.0595          | 0.9812   |
+| 0.2908        | 2.94  | 100  | 0.1524          | 0.9511   |
+| 0.1806        | 5.88  | 200  | 0.1269          | 0.9586   |
+| 0.1135        | 8.82  | 300  | 0.0720          | 0.9774   |
+| 0.1061        | 11.76 | 400  | 0.1519          | 0.9624   |
+| 0.0816        | 14.71 | 500  | 0.1845          | 0.9398   |
+| 0.0815        | 17.65 | 600  | 0.0966          | 0.9737   |
+| 0.0741        | 20.59 | 700  | 0.1029          | 0.9812   |
+| 0.0423        | 23.53 | 800  | 0.1519          | 0.9699   |
+| 0.0468        | 26.47 | 900  | 0.0757          | 0.9850   |
+| 0.0249        | 29.41 | 1000 | 0.0859          | 0.9850   |
+| 0.0443        | 32.35 | 1100 | 0.0878          | 0.9774   |
+| 0.0291        | 35.29 | 1200 | 0.0487          | 0.9887   |
+| 0.0263        | 38.24 | 1300 | 0.0643          | 0.9887   |
+| 0.0239        | 41.18 | 1400 | 0.1042          | 0.9774   |
+| 0.0331        | 44.12 | 1500 | 0.0679          | 0.9887   |
+| 0.0103        | 47.06 | 1600 | 0.0723          | 0.9887   |
+| 0.0131        | 50.0  | 1700 | 0.0732          | 0.9887   |
 
 
 ### Framework versions

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 20.0,
-    "eval_accuracy": 0.9624060150375939,
-    "eval_loss": 0.0946127399802208,
-    "eval_runtime": 2.633,
-    "eval_samples_per_second": 101.024,
-    "eval_steps_per_second": 12.913,
-    "total_flos": 1.6434612963230515e+18,
-    "train_loss": 0.5515072485103326,
-    "train_runtime": 465.2023,
-    "train_samples_per_second": 45.658,
-    "train_steps_per_second": 1.462
+    "eval_accuracy": 0.015037593984962405,
+    "eval_loss": 2.07399845123291,
+    "eval_runtime": 3.7505,
+    "eval_samples_per_second": 70.923,
+    "eval_steps_per_second": 9.065,
+    "total_flos": 5.337848001643315e+17,
+    "train_loss": 0.17142059330773704,
+    "train_runtime": 289.6781,
+    "train_samples_per_second": 73.323,
+    "train_steps_per_second": 2.347
 }

eval_results.json

@@ -1,8 +1,7 @@
 {
-    "epoch": 20.0,
-    "eval_accuracy": 0.9624060150375939,
-    "eval_loss": 0.0946127399802208,
-    "eval_runtime": 2.633,
-    "eval_samples_per_second": 101.024,
-    "eval_steps_per_second": 12.913
+    "eval_accuracy": 0.015037593984962405,
+    "eval_loss": 2.07399845123291,
+    "eval_runtime": 3.7505,
+    "eval_samples_per_second": 70.923,
+    "eval_steps_per_second": 9.065
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4fa14359c275d65dc97a7f3b269406a2a20534033e6cd3c3319d98d6a78eba7f
+oid sha256:a925714ef2935d8483bd6ac2cc822604c804a1a4e0e9f6c1009b76d6af332472
 size 111323316

runs/Dec01_13-41-24_610b2a9400b8/events.out.tfevents.1701438092.610b2a9400b8.3653.8

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:44278824c45c5f66cfc1d834fe8aa9456abef81f190b89f154cc7d318c6c92ec
+size 37403

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
-    "total_flos": 1.6434612963230515e+18,
-    "train_loss": 0.5515072485103326,
-    "train_runtime": 465.2023,
-    "train_samples_per_second": 45.658,
-    "train_steps_per_second": 1.462
+    "total_flos": 5.337848001643315e+17,
+    "train_loss": 0.17142059330773704,
+    "train_runtime": 289.6781,
+    "train_samples_per_second": 73.323,
+    "train_steps_per_second": 2.347
 }

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.0946127399802208,
+  "best_metric": 0.05951835587620735,
   "best_model_checkpoint": "finetuned-Leukemia-cell/checkpoint-600",
   "epoch": 20.0,
   "eval_steps": 100,
@@ -10,481 +10,481 @@
   "log_history": [
     {
       "epoch": 0.29,
-      "learning_rate": 0.00019794117647058826,
-      "loss": 1.9655,
+      "learning_rate": 0.00019705882352941177,
+      "loss": 1.5799,
       "step": 10
     },
     {
       "epoch": 0.59,
-      "learning_rate": 0.000195,
-      "loss": 1.7836,
+      "learning_rate": 0.00019441176470588235,
+      "loss": 1.1614,
       "step": 20
     },
     {
       "epoch": 0.88,
-      "learning_rate": 0.0001920588235294118,
-      "loss": 1.5744,
+      "learning_rate": 0.00019147058823529414,
+      "loss": 0.8677,
       "step": 30
     },
     {
       "epoch": 1.18,
-      "learning_rate": 0.00018911764705882353,
-      "loss": 1.3903,
+      "learning_rate": 0.00018882352941176472,
+      "loss": 0.7303,
       "step": 40
     },
     {
       "epoch": 1.47,
-      "learning_rate": 0.00018617647058823532,
-      "loss": 1.3596,
+      "learning_rate": 0.00018588235294117648,
+      "loss": 0.4837,
       "step": 50
     },
     {
       "epoch": 1.76,
-      "learning_rate": 0.00018323529411764706,
-      "loss": 1.3915,
+      "learning_rate": 0.00018294117647058825,
+      "loss": 0.4625,
       "step": 60
     },
     {
       "epoch": 2.06,
       "learning_rate": 0.00018029411764705885,
-      "loss": 1.2525,
+      "loss": 0.3693,
       "step": 70
     },
     {
       "epoch": 2.35,
       "learning_rate": 0.00017735294117647059,
-      "loss": 1.167,
+      "loss": 0.331,
       "step": 80
     },
     {
       "epoch": 2.65,
       "learning_rate": 0.00017441176470588235,
-      "loss": 1.0339,
+      "loss": 0.2914,
       "step": 90
     },
     {
       "epoch": 2.94,
       "learning_rate": 0.00017147058823529412,
-      "loss": 0.9733,
+      "loss": 0.2926,
       "step": 100
     },
     {
       "epoch": 2.94,
-      "eval_accuracy": 0.7255639097744361,
-      "eval_loss": 0.8894439339637756,
-      "eval_runtime": 2.491,
-      "eval_samples_per_second": 106.783,
-      "eval_steps_per_second": 13.649,
+      "eval_accuracy": 0.943609022556391,
+      "eval_loss": 0.15237735211849213,
+      "eval_runtime": 1.9186,
+      "eval_samples_per_second": 138.643,
+      "eval_steps_per_second": 17.721,
       "step": 100
     },
     {
       "epoch": 3.24,
       "learning_rate": 0.00016852941176470588,
-      "loss": 0.9561,
+      "loss": 0.2343,
       "step": 110
     },
     {
       "epoch": 3.53,
       "learning_rate": 0.00016558823529411765,
-      "loss": 0.9493,
+      "loss": 0.2078,
       "step": 120
     },
     {
       "epoch": 3.82,
       "learning_rate": 0.0001626470588235294,
-      "loss": 0.8258,
+      "loss": 0.2448,
       "step": 130
     },
     {
       "epoch": 4.12,
       "learning_rate": 0.0001597058823529412,
-      "loss": 0.8111,
+      "loss": 0.2146,
       "step": 140
     },
     {
       "epoch": 4.41,
       "learning_rate": 0.00015676470588235294,
-      "loss": 0.9361,
+      "loss": 0.2086,
       "step": 150
     },
     {
       "epoch": 4.71,
       "learning_rate": 0.00015382352941176473,
-      "loss": 0.7733,
+      "loss": 0.1924,
       "step": 160
     },
     {
       "epoch": 5.0,
       "learning_rate": 0.00015088235294117647,
-      "loss": 0.5113,
+      "loss": 0.0882,
       "step": 170
     },
     {
       "epoch": 5.29,
       "learning_rate": 0.00014794117647058826,
-      "loss": 0.7088,
+      "loss": 0.1726,
       "step": 180
     },
     {
       "epoch": 5.59,
       "learning_rate": 0.000145,
-      "loss": 0.6652,
+      "loss": 0.1193,
       "step": 190
     },
     {
       "epoch": 5.88,
       "learning_rate": 0.00014205882352941177,
-      "loss": 0.7184,
+      "loss": 0.1844,
       "step": 200
     },
     {
       "epoch": 5.88,
-      "eval_accuracy": 0.7293233082706767,
-      "eval_loss": 0.7875903248786926,
-      "eval_runtime": 2.5205,
-      "eval_samples_per_second": 105.533,
-      "eval_steps_per_second": 13.489,
+      "eval_accuracy": 0.943609022556391,
+      "eval_loss": 0.1717682033777237,
+      "eval_runtime": 2.2294,
+      "eval_samples_per_second": 119.314,
+      "eval_steps_per_second": 15.251,
       "step": 200
     },
     {
       "epoch": 6.18,
       "learning_rate": 0.00013911764705882353,
-      "loss": 0.9143,
+      "loss": 0.31,
       "step": 210
     },
     {
       "epoch": 6.47,
       "learning_rate": 0.0001361764705882353,
-      "loss": 0.7709,
+      "loss": 0.1387,
       "step": 220
     },
     {
       "epoch": 6.76,
       "learning_rate": 0.00013323529411764706,
-      "loss": 0.5797,
+      "loss": 0.2172,
       "step": 230
     },
     {
       "epoch": 7.06,
       "learning_rate": 0.00013029411764705883,
-      "loss": 0.5424,
+      "loss": 0.1062,
       "step": 240
     },
     {
       "epoch": 7.35,
       "learning_rate": 0.0001273529411764706,
-      "loss": 0.6525,
+      "loss": 0.1055,
       "step": 250
     },
     {
       "epoch": 7.65,
       "learning_rate": 0.00012441176470588236,
-      "loss": 0.3516,
+      "loss": 0.087,
       "step": 260
     },
     {
       "epoch": 7.94,
       "learning_rate": 0.00012147058823529412,
-      "loss": 0.4829,
+      "loss": 0.0817,
       "step": 270
     },
     {
       "epoch": 8.24,
       "learning_rate": 0.00011852941176470589,
-      "loss": 0.5154,
+      "loss": 0.0727,
       "step": 280
     },
     {
       "epoch": 8.53,
       "learning_rate": 0.00011558823529411764,
-      "loss": 0.3485,
+      "loss": 0.0561,
       "step": 290
     },
     {
       "epoch": 8.82,
       "learning_rate": 0.00011264705882352942,
-      "loss": 0.5299,
+      "loss": 0.1189,
       "step": 300
     },
     {
       "epoch": 8.82,
-      "eval_accuracy": 0.8609022556390977,
-      "eval_loss": 0.5183172821998596,
-      "eval_runtime": 2.8712,
-      "eval_samples_per_second": 92.644,
-      "eval_steps_per_second": 11.842,
+      "eval_accuracy": 0.9661654135338346,
+      "eval_loss": 0.08863785862922668,
+      "eval_runtime": 1.8695,
+      "eval_samples_per_second": 142.285,
+      "eval_steps_per_second": 18.187,
       "step": 300
     },
     {
       "epoch": 9.12,
       "learning_rate": 0.0001097058823529412,
-      "loss": 0.5263,
+      "loss": 0.0705,
       "step": 310
     },
     {
       "epoch": 9.41,
       "learning_rate": 0.00010676470588235295,
-      "loss": 0.4048,
+      "loss": 0.147,
       "step": 320
     },
     {
       "epoch": 9.71,
       "learning_rate": 0.00010382352941176472,
-      "loss": 0.4386,
+      "loss": 0.0612,
       "step": 330
     },
     {
       "epoch": 10.0,
       "learning_rate": 0.00010088235294117648,
-      "loss": 0.4728,
+      "loss": 0.0926,
       "step": 340
     },
     {
       "epoch": 10.29,
       "learning_rate": 9.794117647058824e-05,
-      "loss": 0.379,
+      "loss": 0.0977,
       "step": 350
     },
     {
       "epoch": 10.59,
       "learning_rate": 9.5e-05,
-      "loss": 0.4361,
+      "loss": 0.1104,
       "step": 360
     },
     {
       "epoch": 10.88,
       "learning_rate": 9.205882352941177e-05,
-      "loss": 0.2929,
+      "loss": 0.0834,
       "step": 370
     },
     {
       "epoch": 11.18,
       "learning_rate": 8.911764705882354e-05,
-      "loss": 0.2848,
+      "loss": 0.0831,
       "step": 380
     },
     {
       "epoch": 11.47,
       "learning_rate": 8.61764705882353e-05,
-      "loss": 0.4282,
+      "loss": 0.0464,
       "step": 390
     },
     {
       "epoch": 11.76,
       "learning_rate": 8.323529411764707e-05,
-      "loss": 0.3991,
+      "loss": 0.0682,
       "step": 400
     },
     {
       "epoch": 11.76,
-      "eval_accuracy": 0.8947368421052632,
-      "eval_loss": 0.31206753849983215,
-      "eval_runtime": 2.9634,
-      "eval_samples_per_second": 89.761,
-      "eval_steps_per_second": 11.473,
+      "eval_accuracy": 0.9699248120300752,
+      "eval_loss": 0.09778111428022385,
+      "eval_runtime": 2.3942,
+      "eval_samples_per_second": 111.103,
+      "eval_steps_per_second": 14.201,
       "step": 400
     },
     {
       "epoch": 12.06,
       "learning_rate": 8.029411764705883e-05,
-      "loss": 0.3552,
+      "loss": 0.0504,
       "step": 410
     },
     {
       "epoch": 12.35,
       "learning_rate": 7.73529411764706e-05,
-      "loss": 0.2886,
+      "loss": 0.0379,
       "step": 420
     },
     {
       "epoch": 12.65,
       "learning_rate": 7.441176470588236e-05,
-      "loss": 0.2809,
+      "loss": 0.0261,
       "step": 430
     },
     {
       "epoch": 12.94,
       "learning_rate": 7.147058823529412e-05,
-      "loss": 0.3627,
+      "loss": 0.0838,
       "step": 440
     },
     {
       "epoch": 13.24,
       "learning_rate": 6.852941176470589e-05,
-      "loss": 0.3137,
+      "loss": 0.0262,
       "step": 450
     },
     {
       "epoch": 13.53,
       "learning_rate": 6.558823529411765e-05,
-      "loss": 0.198,
+      "loss": 0.0232,
       "step": 460
     },
     {
       "epoch": 13.82,
       "learning_rate": 6.264705882352942e-05,
-      "loss": 0.3268,
+      "loss": 0.0403,
       "step": 470
     },
     {
       "epoch": 14.12,
       "learning_rate": 5.970588235294118e-05,
-      "loss": 0.256,
+      "loss": 0.049,
       "step": 480
     },
     {
       "epoch": 14.41,
       "learning_rate": 5.676470588235294e-05,
-      "loss": 0.2544,
+      "loss": 0.0606,
       "step": 490
     },
     {
       "epoch": 14.71,
       "learning_rate": 5.382352941176471e-05,
-      "loss": 0.2263,
+      "loss": 0.0439,
       "step": 500
     },
     {
       "epoch": 14.71,
-      "eval_accuracy": 0.9548872180451128,
-      "eval_loss": 0.13367173075675964,
-      "eval_runtime": 2.4729,
-      "eval_samples_per_second": 107.566,
-      "eval_steps_per_second": 13.749,
+      "eval_accuracy": 0.9736842105263158,
+      "eval_loss": 0.08120405673980713,
+      "eval_runtime": 1.897,
+      "eval_samples_per_second": 140.22,
+      "eval_steps_per_second": 17.923,
       "step": 500
     },
     {
       "epoch": 15.0,
       "learning_rate": 5.088235294117647e-05,
-      "loss": 0.4011,
+      "loss": 0.0495,
       "step": 510
     },
     {
       "epoch": 15.29,
       "learning_rate": 4.794117647058824e-05,
-      "loss": 0.2051,
+      "loss": 0.0337,
       "step": 520
     },
     {
       "epoch": 15.59,
       "learning_rate": 4.5e-05,
-      "loss": 0.2563,
+      "loss": 0.0987,
       "step": 530
     },
     {
       "epoch": 15.88,
       "learning_rate": 4.205882352941177e-05,
-      "loss": 0.2428,
+      "loss": 0.0218,
       "step": 540
     },
     {
       "epoch": 16.18,
       "learning_rate": 3.911764705882353e-05,
-      "loss": 0.189,
+      "loss": 0.0324,
       "step": 550
     },
     {
       "epoch": 16.47,
       "learning_rate": 3.61764705882353e-05,
-      "loss": 0.19,
+      "loss": 0.0364,
       "step": 560
     },
     {
       "epoch": 16.76,
       "learning_rate": 3.3235294117647056e-05,
-      "loss": 0.1696,
+      "loss": 0.0284,
       "step": 570
     },
     {
       "epoch": 17.06,
       "learning_rate": 3.0294117647058824e-05,
-      "loss": 0.1336,
+      "loss": 0.0306,
       "step": 580
     },
     {
       "epoch": 17.35,
       "learning_rate": 2.7352941176470593e-05,
-      "loss": 0.1532,
+      "loss": 0.0535,
       "step": 590
     },
     {
       "epoch": 17.65,
       "learning_rate": 2.4411764705882354e-05,
-      "loss": 0.1782,
+      "loss": 0.0544,
       "step": 600
     },
     {
       "epoch": 17.65,
-      "eval_accuracy": 0.9624060150375939,
-      "eval_loss": 0.0946127399802208,
-      "eval_runtime": 2.8201,
-      "eval_samples_per_second": 94.322,
-      "eval_steps_per_second": 12.056,
+      "eval_accuracy": 0.981203007518797,
+      "eval_loss": 0.05951835587620735,
+      "eval_runtime": 1.8559,
+      "eval_samples_per_second": 143.327,
+      "eval_steps_per_second": 18.32,
       "step": 600
     },
     {
       "epoch": 17.94,
       "learning_rate": 2.1470588235294116e-05,
-      "loss": 0.1511,
+      "loss": 0.0121,
       "step": 610
     },
     {
       "epoch": 18.24,
       "learning_rate": 1.8529411764705884e-05,
-      "loss": 0.0837,
+      "loss": 0.0049,
       "step": 620
     },
     {
       "epoch": 18.53,
       "learning_rate": 1.558823529411765e-05,
-      "loss": 0.1084,
+      "loss": 0.0086,
       "step": 630
     },
     {
       "epoch": 18.82,
       "learning_rate": 1.2647058823529412e-05,
-      "loss": 0.1313,
+      "loss": 0.0102,
       "step": 640
     },
     {
       "epoch": 19.12,
       "learning_rate": 9.705882352941177e-06,
-      "loss": 0.1135,
+      "loss": 0.0342,
       "step": 650
     },
     {
       "epoch": 19.41,
       "learning_rate": 6.7647058823529414e-06,
-      "loss": 0.0844,
+      "loss": 0.0367,
       "step": 660
     },
     {
       "epoch": 19.71,
       "learning_rate": 3.823529411764706e-06,
-      "loss": 0.1765,
+      "loss": 0.0313,
       "step": 670
     },
     {
       "epoch": 20.0,
       "learning_rate": 8.823529411764706e-07,
-      "loss": 0.1745,
+      "loss": 0.0465,
       "step": 680
     },
     {
       "epoch": 20.0,
       "step": 680,
-      "total_flos": 1.6434612963230515e+18,
-      "train_loss": 0.5515072485103326,
-      "train_runtime": 465.2023,
-      "train_samples_per_second": 45.658,
-      "train_steps_per_second": 1.462
+      "total_flos": 5.337848001643315e+17,
+      "train_loss": 0.17142059330773704,
+      "train_runtime": 289.6781,
+      "train_samples_per_second": 73.323,
+      "train_steps_per_second": 2.347
     }
   ],
   "logging_steps": 10,
   "max_steps": 680,
   "num_train_epochs": 20,
   "save_steps": 100,
-  "total_flos": 1.6434612963230515e+18,
+  "total_flos": 5.337848001643315e+17,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7453b4d852d9cd2e5e3323177f837efc3e7f9c2f92828be34bc2ab59e6d6d25a
+oid sha256:68354e0c5b01561e938df0004ef20ef191e70a1505432fac27a2b4cc3d56354e
 size 4600