End of training

README.md

@@ -18,7 +18,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [microsoft/swinv2-tiny-patch4-window8-256](https://huggingface.co/microsoft/swinv2-tiny-patch4-window8-256) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.8446
+- Loss: 1.8351
 - Accuracy: 0.6538
 
 ## Model description

all_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 29.8,
+    "eval_accuracy": 0.6538461538461539,
+    "eval_loss": 1.8350580930709839,
+    "eval_runtime": 1.9929,
+    "eval_samples_per_second": 52.185,
+    "eval_steps_per_second": 2.007,
+    "total_flos": 3.1009435460901274e+17,
+    "train_loss": 1.446416116754214,
+    "train_runtime": 581.9968,
+    "train_samples_per_second": 23.969,
+    "train_steps_per_second": 0.155
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 29.8,
+    "eval_accuracy": 0.6538461538461539,
+    "eval_loss": 1.8350580930709839,
+    "eval_runtime": 1.9929,
+    "eval_samples_per_second": 52.185,
+    "eval_steps_per_second": 2.007
+}

runs/Feb04_02-12-51_ad369e174003/events.out.tfevents.1738635783.ad369e174003.1285.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:567b4445b1a44260c1690f7432567aa3e315bb64f583e24ed18c83b0e741e5a5
+size 405

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 29.8,
+    "total_flos": 3.1009435460901274e+17,
+    "train_loss": 1.446416116754214,
+    "train_runtime": 581.9968,
+    "train_samples_per_second": 23.969,
+    "train_steps_per_second": 0.155
+}

trainer_state.json

@@ -0,0 +1,361 @@
+{
+  "best_metric": 0.6538461538461539,
+  "best_model_checkpoint": "swinv2-tiny-patch4-window8-256-dmae-humeda-DAV37/checkpoint-54",
+  "epoch": 29.8,
+  "eval_steps": 500,
+  "global_step": 90,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.8,
+      "eval_accuracy": 0.1346153846153846,
+      "eval_loss": 1.6284154653549194,
+      "eval_runtime": 1.4528,
+      "eval_samples_per_second": 71.585,
+      "eval_steps_per_second": 2.753,
+      "step": 3
+    },
+    {
+      "epoch": 1.8,
+      "eval_accuracy": 0.2403846153846154,
+      "eval_loss": 1.5965616703033447,
+      "eval_runtime": 1.4255,
+      "eval_samples_per_second": 72.959,
+      "eval_steps_per_second": 2.806,
+      "step": 6
+    },
+    {
+      "epoch": 2.8,
+      "eval_accuracy": 0.3942307692307692,
+      "eval_loss": 1.5076172351837158,
+      "eval_runtime": 1.9058,
+      "eval_samples_per_second": 54.57,
+      "eval_steps_per_second": 2.099,
+      "step": 9
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 30.521873474121094,
+      "learning_rate": 4.983095894354858e-05,
+      "loss": 6.28,
+      "step": 12
+    },
+    {
+      "epoch": 3.8,
+      "eval_accuracy": 0.46153846153846156,
+      "eval_loss": 1.2911909818649292,
+      "eval_runtime": 1.4733,
+      "eval_samples_per_second": 70.589,
+      "eval_steps_per_second": 2.715,
+      "step": 12
+    },
+    {
+      "epoch": 4.8,
+      "eval_accuracy": 0.5096153846153846,
+      "eval_loss": 1.2136619091033936,
+      "eval_runtime": 1.4295,
+      "eval_samples_per_second": 72.754,
+      "eval_steps_per_second": 2.798,
+      "step": 15
+    },
+    {
+      "epoch": 5.8,
+      "eval_accuracy": 0.5384615384615384,
+      "eval_loss": 1.1916911602020264,
+      "eval_runtime": 1.4167,
+      "eval_samples_per_second": 73.411,
+      "eval_steps_per_second": 2.823,
+      "step": 18
+    },
+    {
+      "epoch": 6.8,
+      "eval_accuracy": 0.5673076923076923,
+      "eval_loss": 1.1497925519943237,
+      "eval_runtime": 1.4207,
+      "eval_samples_per_second": 73.205,
+      "eval_steps_per_second": 2.816,
+      "step": 21
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 21.03725814819336,
+      "learning_rate": 4.588719528532342e-05,
+      "loss": 2.9539,
+      "step": 24
+    },
+    {
+      "epoch": 7.8,
+      "eval_accuracy": 0.5865384615384616,
+      "eval_loss": 1.202553629875183,
+      "eval_runtime": 1.8718,
+      "eval_samples_per_second": 55.562,
+      "eval_steps_per_second": 2.137,
+      "step": 24
+    },
+    {
+      "epoch": 8.8,
+      "eval_accuracy": 0.5961538461538461,
+      "eval_loss": 1.2710736989974976,
+      "eval_runtime": 1.845,
+      "eval_samples_per_second": 56.369,
+      "eval_steps_per_second": 2.168,
+      "step": 27
+    },
+    {
+      "epoch": 9.8,
+      "eval_accuracy": 0.625,
+      "eval_loss": 1.3533679246902466,
+      "eval_runtime": 1.412,
+      "eval_samples_per_second": 73.654,
+      "eval_steps_per_second": 2.833,
+      "step": 30
+    },
+    {
+      "epoch": 10.8,
+      "eval_accuracy": 0.625,
+      "eval_loss": 1.3210114240646362,
+      "eval_runtime": 1.4139,
+      "eval_samples_per_second": 73.555,
+      "eval_steps_per_second": 2.829,
+      "step": 33
+    },
+    {
+      "epoch": 11.8,
+      "grad_norm": 31.268625259399414,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.9643,
+      "step": 36
+    },
+    {
+      "epoch": 11.8,
+      "eval_accuracy": 0.6346153846153846,
+      "eval_loss": 1.394026279449463,
+      "eval_runtime": 1.4476,
+      "eval_samples_per_second": 71.841,
+      "eval_steps_per_second": 2.763,
+      "step": 36
+    },
+    {
+      "epoch": 12.8,
+      "eval_accuracy": 0.6346153846153846,
+      "eval_loss": 1.4859095811843872,
+      "eval_runtime": 1.9438,
+      "eval_samples_per_second": 53.502,
+      "eval_steps_per_second": 2.058,
+      "step": 39
+    },
+    {
+      "epoch": 13.8,
+      "eval_accuracy": 0.6346153846153846,
+      "eval_loss": 1.4965322017669678,
+      "eval_runtime": 1.4426,
+      "eval_samples_per_second": 72.093,
+      "eval_steps_per_second": 2.773,
+      "step": 42
+    },
+    {
+      "epoch": 14.8,
+      "eval_accuracy": 0.625,
+      "eval_loss": 1.5463168621063232,
+      "eval_runtime": 1.4453,
+      "eval_samples_per_second": 71.956,
+      "eval_steps_per_second": 2.768,
+      "step": 45
+    },
+    {
+      "epoch": 15.8,
+      "grad_norm": 22.09276008605957,
+      "learning_rate": 2.6453620722761896e-05,
+      "loss": 0.3275,
+      "step": 48
+    },
+    {
+      "epoch": 15.8,
+      "eval_accuracy": 0.6346153846153846,
+      "eval_loss": 1.5884944200515747,
+      "eval_runtime": 1.5085,
+      "eval_samples_per_second": 68.944,
+      "eval_steps_per_second": 2.652,
+      "step": 48
+    },
+    {
+      "epoch": 16.8,
+      "eval_accuracy": 0.6442307692307693,
+      "eval_loss": 1.6466037034988403,
+      "eval_runtime": 1.9209,
+      "eval_samples_per_second": 54.142,
+      "eval_steps_per_second": 2.082,
+      "step": 51
+    },
+    {
+      "epoch": 17.8,
+      "eval_accuracy": 0.6538461538461539,
+      "eval_loss": 1.8350580930709839,
+      "eval_runtime": 1.4271,
+      "eval_samples_per_second": 72.873,
+      "eval_steps_per_second": 2.803,
+      "step": 54
+    },
+    {
+      "epoch": 18.8,
+      "eval_accuracy": 0.6442307692307693,
+      "eval_loss": 1.8326284885406494,
+      "eval_runtime": 1.4279,
+      "eval_samples_per_second": 72.832,
+      "eval_steps_per_second": 2.801,
+      "step": 57
+    },
+    {
+      "epoch": 19.8,
+      "grad_norm": 31.147811889648438,
+      "learning_rate": 1.509800584902108e-05,
+      "loss": 0.1501,
+      "step": 60
+    },
+    {
+      "epoch": 19.8,
+      "eval_accuracy": 0.6346153846153846,
+      "eval_loss": 1.7521249055862427,
+      "eval_runtime": 1.4343,
+      "eval_samples_per_second": 72.507,
+      "eval_steps_per_second": 2.789,
+      "step": 60
+    },
+    {
+      "epoch": 20.8,
+      "eval_accuracy": 0.6538461538461539,
+      "eval_loss": 1.780649185180664,
+      "eval_runtime": 1.8229,
+      "eval_samples_per_second": 57.051,
+      "eval_steps_per_second": 2.194,
+      "step": 63
+    },
+    {
+      "epoch": 21.8,
+      "eval_accuracy": 0.6538461538461539,
+      "eval_loss": 1.766915202140808,
+      "eval_runtime": 1.7174,
+      "eval_samples_per_second": 60.557,
+      "eval_steps_per_second": 2.329,
+      "step": 66
+    },
+    {
+      "epoch": 22.8,
+      "eval_accuracy": 0.6346153846153846,
+      "eval_loss": 1.8873956203460693,
+      "eval_runtime": 1.4404,
+      "eval_samples_per_second": 72.202,
+      "eval_steps_per_second": 2.777,
+      "step": 69
+    },
+    {
+      "epoch": 23.8,
+      "grad_norm": 9.280792236328125,
+      "learning_rate": 5.848888922025553e-06,
+      "loss": 0.09,
+      "step": 72
+    },
+    {
+      "epoch": 23.8,
+      "eval_accuracy": 0.6538461538461539,
+      "eval_loss": 1.8826991319656372,
+      "eval_runtime": 1.431,
+      "eval_samples_per_second": 72.678,
+      "eval_steps_per_second": 2.795,
+      "step": 72
+    },
+    {
+      "epoch": 24.8,
+      "eval_accuracy": 0.6538461538461539,
+      "eval_loss": 1.8329722881317139,
+      "eval_runtime": 1.9352,
+      "eval_samples_per_second": 53.742,
+      "eval_steps_per_second": 2.067,
+      "step": 75
+    },
+    {
+      "epoch": 25.8,
+      "eval_accuracy": 0.6538461538461539,
+      "eval_loss": 1.8330668210983276,
+      "eval_runtime": 1.5124,
+      "eval_samples_per_second": 68.765,
+      "eval_steps_per_second": 2.645,
+      "step": 78
+    },
+    {
+      "epoch": 26.8,
+      "eval_accuracy": 0.6538461538461539,
+      "eval_loss": 1.841007947921753,
+      "eval_runtime": 1.438,
+      "eval_samples_per_second": 72.324,
+      "eval_steps_per_second": 2.782,
+      "step": 81
+    },
+    {
+      "epoch": 27.8,
+      "grad_norm": 13.664288520812988,
+      "learning_rate": 6.738782355044049e-07,
+      "loss": 0.0595,
+      "step": 84
+    },
+    {
+      "epoch": 27.8,
+      "eval_accuracy": 0.6442307692307693,
+      "eval_loss": 1.8440735340118408,
+      "eval_runtime": 1.4316,
+      "eval_samples_per_second": 72.647,
+      "eval_steps_per_second": 2.794,
+      "step": 84
+    },
+    {
+      "epoch": 28.8,
+      "eval_accuracy": 0.6538461538461539,
+      "eval_loss": 1.8444327116012573,
+      "eval_runtime": 1.7905,
+      "eval_samples_per_second": 58.086,
+      "eval_steps_per_second": 2.234,
+      "step": 87
+    },
+    {
+      "epoch": 29.8,
+      "eval_accuracy": 0.6538461538461539,
+      "eval_loss": 1.844649076461792,
+      "eval_runtime": 1.7539,
+      "eval_samples_per_second": 59.298,
+      "eval_steps_per_second": 2.281,
+      "step": 90
+    },
+    {
+      "epoch": 29.8,
+      "step": 90,
+      "total_flos": 3.1009435460901274e+17,
+      "train_loss": 1.446416116754214,
+      "train_runtime": 581.9968,
+      "train_samples_per_second": 23.969,
+      "train_steps_per_second": 0.155
+    }
+  ],
+  "logging_steps": 12,
+  "max_steps": 90,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 30,
+  "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": 3.1009435460901274e+17,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}