End of training

README.md

@@ -23,11 +23,11 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [microsoft/swin-large-patch4-window7-224-in22k](https://huggingface.co/microsoft/swin-large-patch4-window7-224-in22k) on the medmnist-v2 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3214
-- Accuracy: 0.8846
-- Precision: 0.8506
-- Recall: 0.8609
-- F1: 0.8555
+- Loss: 0.3595
+- Accuracy: 0.8526
+- Precision: 0.8162
+- Recall: 0.8014
+- F1: 0.8082
 
 ## Model description
 

all_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 9.142857142857142,
+    "eval_accuracy": 0.8525641025641025,
+    "eval_f1": 0.808241141574475,
+    "eval_loss": 0.359468936920166,
+    "eval_precision": 0.8162393162393162,
+    "eval_recall": 0.8013784461152882,
+    "eval_runtime": 1.7938,
+    "eval_samples_per_second": 86.967,
+    "eval_steps_per_second": 5.575,
+    "total_flos": 8.847492793597624e+17,
+    "train_loss": 0.48110649585723875,
+    "train_runtime": 140.8237,
+    "train_samples_per_second": 38.772,
+    "train_steps_per_second": 0.568
+}

eval_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 9.142857142857142,
+    "eval_accuracy": 0.8525641025641025,
+    "eval_f1": 0.808241141574475,
+    "eval_loss": 0.359468936920166,
+    "eval_precision": 0.8162393162393162,
+    "eval_recall": 0.8013784461152882,
+    "eval_runtime": 1.7938,
+    "eval_samples_per_second": 86.967,
+    "eval_steps_per_second": 5.575
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 9.142857142857142,
+    "total_flos": 8.847492793597624e+17,
+    "train_loss": 0.48110649585723875,
+    "train_runtime": 140.8237,
+    "train_samples_per_second": 38.772,
+    "train_steps_per_second": 0.568
+}

trainer_state.json

@@ -0,0 +1,218 @@
+{
+  "best_metric": 0.8974358974358975,
+  "best_model_checkpoint": "swin-large-patch4-window7-224-in22k-finetuned-lora-medmnistv2/checkpoint-35",
+  "epoch": 9.142857142857142,
+  "eval_steps": 500,
+  "global_step": 80,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.9142857142857143,
+      "eval_accuracy": 0.7435897435897436,
+      "eval_f1": 0.47082767978290374,
+      "eval_loss": 0.506903886795044,
+      "eval_precision": 0.8701298701298701,
+      "eval_recall": 0.5238095238095238,
+      "eval_runtime": 1.0819,
+      "eval_samples_per_second": 72.096,
+      "eval_steps_per_second": 4.622,
+      "step": 8
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 1.5360957384109497,
+      "learning_rate": 0.004375,
+      "loss": 0.6976,
+      "step": 10
+    },
+    {
+      "epoch": 1.9428571428571428,
+      "eval_accuracy": 0.8589743589743589,
+      "eval_f1": 0.8234204568841326,
+      "eval_loss": 0.4591017961502075,
+      "eval_precision": 0.8189935064935066,
+      "eval_recall": 0.8283208020050126,
+      "eval_runtime": 0.8911,
+      "eval_samples_per_second": 87.528,
+      "eval_steps_per_second": 5.611,
+      "step": 17
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.8164573311805725,
+      "learning_rate": 0.00375,
+      "loss": 0.5351,
+      "step": 20
+    },
+    {
+      "epoch": 2.9714285714285715,
+      "eval_accuracy": 0.8846153846153846,
+      "eval_f1": 0.8461538461538461,
+      "eval_loss": 0.37451839447021484,
+      "eval_precision": 0.8666666666666667,
+      "eval_recall": 0.8308270676691729,
+      "eval_runtime": 0.9058,
+      "eval_samples_per_second": 86.111,
+      "eval_steps_per_second": 5.52,
+      "step": 26
+    },
+    {
+      "epoch": 3.4285714285714284,
+      "grad_norm": 0.974304735660553,
+      "learning_rate": 0.003125,
+      "loss": 0.4998,
+      "step": 30
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.8974358974358975,
+      "eval_f1": 0.8696741854636592,
+      "eval_loss": 0.32434844970703125,
+      "eval_precision": 0.8696741854636592,
+      "eval_recall": 0.8696741854636592,
+      "eval_runtime": 0.9239,
+      "eval_samples_per_second": 84.422,
+      "eval_steps_per_second": 5.412,
+      "step": 35
+    },
+    {
+      "epoch": 4.571428571428571,
+      "grad_norm": 0.9340265989303589,
+      "learning_rate": 0.0025,
+      "loss": 0.4569,
+      "step": 40
+    },
+    {
+      "epoch": 4.914285714285715,
+      "eval_accuracy": 0.8589743589743589,
+      "eval_f1": 0.811965811965812,
+      "eval_loss": 0.40701231360435486,
+      "eval_precision": 0.8305555555555555,
+      "eval_recall": 0.7982456140350878,
+      "eval_runtime": 0.8814,
+      "eval_samples_per_second": 88.5,
+      "eval_steps_per_second": 5.673,
+      "step": 43
+    },
+    {
+      "epoch": 5.714285714285714,
+      "grad_norm": 1.012099266052246,
+      "learning_rate": 0.001875,
+      "loss": 0.4182,
+      "step": 50
+    },
+    {
+      "epoch": 5.942857142857143,
+      "eval_accuracy": 0.8717948717948718,
+      "eval_f1": 0.831896551724138,
+      "eval_loss": 0.3801339566707611,
+      "eval_precision": 0.8438893844781445,
+      "eval_recall": 0.8220551378446115,
+      "eval_runtime": 0.8409,
+      "eval_samples_per_second": 92.758,
+      "eval_steps_per_second": 5.946,
+      "step": 52
+    },
+    {
+      "epoch": 6.857142857142857,
+      "grad_norm": 0.7531532645225525,
+      "learning_rate": 0.00125,
+      "loss": 0.4432,
+      "step": 60
+    },
+    {
+      "epoch": 6.9714285714285715,
+      "eval_accuracy": 0.8717948717948718,
+      "eval_f1": 0.8370927318295739,
+      "eval_loss": 0.3070574402809143,
+      "eval_precision": 0.8370927318295739,
+      "eval_recall": 0.8370927318295739,
+      "eval_runtime": 0.856,
+      "eval_samples_per_second": 91.124,
+      "eval_steps_per_second": 5.841,
+      "step": 61
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 1.1633806228637695,
+      "learning_rate": 0.000625,
+      "loss": 0.3988,
+      "step": 70
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.8717948717948718,
+      "eval_f1": 0.8417207792207791,
+      "eval_loss": 0.320502907037735,
+      "eval_precision": 0.833201581027668,
+      "eval_recall": 0.8521303258145363,
+      "eval_runtime": 0.8567,
+      "eval_samples_per_second": 91.048,
+      "eval_steps_per_second": 5.836,
+      "step": 70
+    },
+    {
+      "epoch": 8.914285714285715,
+      "eval_accuracy": 0.8846153846153846,
+      "eval_f1": 0.8555258283597448,
+      "eval_loss": 0.3238745331764221,
+      "eval_precision": 0.8506493506493507,
+      "eval_recall": 0.8609022556390977,
+      "eval_runtime": 0.8616,
+      "eval_samples_per_second": 90.526,
+      "eval_steps_per_second": 5.803,
+      "step": 78
+    },
+    {
+      "epoch": 9.142857142857142,
+      "grad_norm": 0.8995063900947571,
+      "learning_rate": 0.0,
+      "loss": 0.3993,
+      "step": 80
+    },
+    {
+      "epoch": 9.142857142857142,
+      "eval_accuracy": 0.8846153846153846,
+      "eval_f1": 0.8555258283597448,
+      "eval_loss": 0.32138964533805847,
+      "eval_precision": 0.8506493506493507,
+      "eval_recall": 0.8609022556390977,
+      "eval_runtime": 0.8772,
+      "eval_samples_per_second": 88.917,
+      "eval_steps_per_second": 5.7,
+      "step": 80
+    },
+    {
+      "epoch": 9.142857142857142,
+      "step": 80,
+      "total_flos": 8.847492793597624e+17,
+      "train_loss": 0.48110649585723875,
+      "train_runtime": 140.8237,
+      "train_samples_per_second": 38.772,
+      "train_steps_per_second": 0.568
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 80,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 8.847492793597624e+17,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}