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 [facebook/deit-base-patch16-224](https://huggingface.co/facebook/deit-base-patch16-224) on the medmnist-v2 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3454
-- Accuracy: 0.8718
-- Precision: 0.8698
-- Recall: 0.7920
-- F1: 0.8194
+- Loss: 0.3832
+- Accuracy: 0.8333
+- Precision: 0.8079
+- Recall: 0.7431
+- F1: 0.7653
 
 ## Model description
 

all_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 9.142857142857142,
+    "eval_accuracy": 0.8333333333333334,
+    "eval_f1": 0.7652777777777777,
+    "eval_loss": 0.38321784138679504,
+    "eval_precision": 0.807936507936508,
+    "eval_recall": 0.7431077694235589,
+    "eval_runtime": 0.7783,
+    "eval_samples_per_second": 200.43,
+    "eval_steps_per_second": 12.848,
+    "total_flos": 3.8966228771394355e+17,
+    "train_loss": 0.5176919877529145,
+    "train_runtime": 49.4572,
+    "train_samples_per_second": 110.398,
+    "train_steps_per_second": 1.618
+}

eval_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 9.142857142857142,
+    "eval_accuracy": 0.8333333333333334,
+    "eval_f1": 0.7652777777777777,
+    "eval_loss": 0.38321784138679504,
+    "eval_precision": 0.807936507936508,
+    "eval_recall": 0.7431077694235589,
+    "eval_runtime": 0.7783,
+    "eval_samples_per_second": 200.43,
+    "eval_steps_per_second": 12.848
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 9.142857142857142,
+    "total_flos": 3.8966228771394355e+17,
+    "train_loss": 0.5176919877529145,
+    "train_runtime": 49.4572,
+    "train_samples_per_second": 110.398,
+    "train_steps_per_second": 1.618
+}

trainer_state.json

@@ -0,0 +1,218 @@
+{
+  "best_metric": 0.8717948717948718,
+  "best_model_checkpoint": "deit-base-patch16-224-finetuned-lora-medmnistv2/checkpoint-43",
+  "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": 0.5025883913040161,
+      "eval_precision": 0.8701298701298701,
+      "eval_recall": 0.5238095238095238,
+      "eval_runtime": 0.3964,
+      "eval_samples_per_second": 196.791,
+      "eval_steps_per_second": 12.615,
+      "step": 8
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 2.9126269817352295,
+      "learning_rate": 0.004375,
+      "loss": 0.6168,
+      "step": 10
+    },
+    {
+      "epoch": 1.9428571428571428,
+      "eval_accuracy": 0.8461538461538461,
+      "eval_f1": 0.7833333333333333,
+      "eval_loss": 0.4762480854988098,
+      "eval_precision": 0.8285714285714285,
+      "eval_recall": 0.7593984962406015,
+      "eval_runtime": 0.4194,
+      "eval_samples_per_second": 185.972,
+      "eval_steps_per_second": 11.921,
+      "step": 17
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 4.246772289276123,
+      "learning_rate": 0.00375,
+      "loss": 0.5954,
+      "step": 20
+    },
+    {
+      "epoch": 2.9714285714285715,
+      "eval_accuracy": 0.7307692307692307,
+      "eval_f1": 0.42222222222222217,
+      "eval_loss": 0.5305488109588623,
+      "eval_precision": 0.36538461538461536,
+      "eval_recall": 0.5,
+      "eval_runtime": 0.3621,
+      "eval_samples_per_second": 215.403,
+      "eval_steps_per_second": 13.808,
+      "step": 26
+    },
+    {
+      "epoch": 3.4285714285714284,
+      "grad_norm": 0.7361263036727905,
+      "learning_rate": 0.003125,
+      "loss": 0.5934,
+      "step": 30
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.7692307692307693,
+      "eval_f1": 0.5846153846153846,
+      "eval_loss": 0.4789685606956482,
+      "eval_precision": 0.7835616438356164,
+      "eval_recall": 0.5864661654135338,
+      "eval_runtime": 0.3661,
+      "eval_samples_per_second": 213.032,
+      "eval_steps_per_second": 13.656,
+      "step": 35
+    },
+    {
+      "epoch": 4.571428571428571,
+      "grad_norm": 2.504427909851074,
+      "learning_rate": 0.0025,
+      "loss": 0.526,
+      "step": 40
+    },
+    {
+      "epoch": 4.914285714285715,
+      "eval_accuracy": 0.8717948717948718,
+      "eval_f1": 0.8194444444444444,
+      "eval_loss": 0.3693440556526184,
+      "eval_precision": 0.8698412698412699,
+      "eval_recall": 0.7919799498746867,
+      "eval_runtime": 0.3764,
+      "eval_samples_per_second": 207.212,
+      "eval_steps_per_second": 13.283,
+      "step": 43
+    },
+    {
+      "epoch": 5.714285714285714,
+      "grad_norm": 1.3234041929244995,
+      "learning_rate": 0.001875,
+      "loss": 0.4651,
+      "step": 50
+    },
+    {
+      "epoch": 5.942857142857143,
+      "eval_accuracy": 0.7948717948717948,
+      "eval_f1": 0.7533596837944665,
+      "eval_loss": 0.47893068194389343,
+      "eval_precision": 0.7433962264150944,
+      "eval_recall": 0.7694235588972431,
+      "eval_runtime": 0.4225,
+      "eval_samples_per_second": 184.594,
+      "eval_steps_per_second": 11.833,
+      "step": 52
+    },
+    {
+      "epoch": 6.857142857142857,
+      "grad_norm": 0.49008217453956604,
+      "learning_rate": 0.00125,
+      "loss": 0.493,
+      "step": 60
+    },
+    {
+      "epoch": 6.9714285714285715,
+      "eval_accuracy": 0.8205128205128205,
+      "eval_f1": 0.7564674397859055,
+      "eval_loss": 0.4186874330043793,
+      "eval_precision": 0.7791706846673095,
+      "eval_recall": 0.7418546365914787,
+      "eval_runtime": 0.367,
+      "eval_samples_per_second": 212.543,
+      "eval_steps_per_second": 13.625,
+      "step": 61
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 1.9268220663070679,
+      "learning_rate": 0.000625,
+      "loss": 0.4337,
+      "step": 70
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.8589743589743589,
+      "eval_f1": 0.8051328639563933,
+      "eval_loss": 0.3599574863910675,
+      "eval_precision": 0.841733870967742,
+      "eval_recall": 0.7832080200501252,
+      "eval_runtime": 0.3708,
+      "eval_samples_per_second": 210.383,
+      "eval_steps_per_second": 13.486,
+      "step": 70
+    },
+    {
+      "epoch": 8.914285714285715,
+      "eval_accuracy": 0.8717948717948718,
+      "eval_f1": 0.8260481712756467,
+      "eval_loss": 0.3467828929424286,
+      "eval_precision": 0.854387656702025,
+      "eval_recall": 0.8070175438596491,
+      "eval_runtime": 0.4353,
+      "eval_samples_per_second": 179.205,
+      "eval_steps_per_second": 11.487,
+      "step": 78
+    },
+    {
+      "epoch": 9.142857142857142,
+      "grad_norm": 0.5700221657752991,
+      "learning_rate": 0.0,
+      "loss": 0.418,
+      "step": 80
+    },
+    {
+      "epoch": 9.142857142857142,
+      "eval_accuracy": 0.8717948717948718,
+      "eval_f1": 0.8194444444444444,
+      "eval_loss": 0.3453872501850128,
+      "eval_precision": 0.8698412698412699,
+      "eval_recall": 0.7919799498746867,
+      "eval_runtime": 0.4424,
+      "eval_samples_per_second": 176.325,
+      "eval_steps_per_second": 11.303,
+      "step": 80
+    },
+    {
+      "epoch": 9.142857142857142,
+      "step": 80,
+      "total_flos": 3.8966228771394355e+17,
+      "train_loss": 0.5176919877529145,
+      "train_runtime": 49.4572,
+      "train_samples_per_second": 110.398,
+      "train_steps_per_second": 1.618
+    }
+  ],
+  "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": 3.8966228771394355e+17,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}