Training in progress, epoch 0

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 2.6666666666666665,
-    "eval_accuracy": 0.6595744680851063,
-    "eval_loss": 0.6631675362586975,
-    "eval_runtime": 0.68,
-    "eval_samples_per_second": 138.226,
-    "eval_steps_per_second": 4.411,
-    "total_flos": 5.5925975218176e+16,
-    "train_loss": 0.6220952669779459,
-    "train_runtime": 39.3173,
-    "train_samples_per_second": 63.865,
-    "train_steps_per_second": 0.458
+    "epoch": 17.77777777777778,
+    "eval_accuracy": 0.8404255319148937,
+    "eval_loss": 0.4538363516330719,
+    "eval_runtime": 0.7115,
+    "eval_samples_per_second": 132.115,
+    "eval_steps_per_second": 4.216,
+    "total_flos": 3.7037909187824026e+17,
+    "train_loss": 0.427582456668218,
+    "train_runtime": 256.3024,
+    "train_samples_per_second": 65.313,
+    "train_steps_per_second": 0.468
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 2.6666666666666665,
-    "eval_accuracy": 0.6595744680851063,
-    "eval_loss": 0.6631675362586975,
-    "eval_runtime": 0.68,
-    "eval_samples_per_second": 138.226,
-    "eval_steps_per_second": 4.411
+    "epoch": 17.77777777777778,
+    "eval_accuracy": 0.8404255319148937,
+    "eval_loss": 0.4538363516330719,
+    "eval_runtime": 0.7115,
+    "eval_samples_per_second": 132.115,
+    "eval_steps_per_second": 4.216
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6e25e93da128415a502632494938479473e4d52fd586b6e04e15152456d390f5
+oid sha256:3f1d0fa58c1a87cc835d76ca4d8435a9103ba44909676556d7f0b317a933e4f2
 size 110342832

runs/May16_09-15-36_959f1d94c84b/events.out.tfevents.1715851233.959f1d94c84b.2038.3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb0f2a33b8421cd8674ddb7ab807e468091d1e119455dd3e48f46ff8e70eace3
+size 405

runs/May16_09-15-36_959f1d94c84b/events.out.tfevents.1715851570.959f1d94c84b.2038.4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef5c20e7186b3b4803858e34af0de1365be58138d90ed38523cecb9623d4d46d
+size 5502

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 2.6666666666666665,
-    "total_flos": 5.5925975218176e+16,
-    "train_loss": 0.6220952669779459,
-    "train_runtime": 39.3173,
-    "train_samples_per_second": 63.865,
-    "train_steps_per_second": 0.458
+    "epoch": 17.77777777777778,
+    "total_flos": 3.7037909187824026e+17,
+    "train_loss": 0.427582456668218,
+    "train_runtime": 256.3024,
+    "train_samples_per_second": 65.313,
+    "train_steps_per_second": 0.468
 }

trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 0.6595744680851063,
-  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-phones/checkpoint-6",
-  "epoch": 2.6666666666666665,
+  "best_metric": 0.8404255319148937,
+  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-phones/checkpoint-81",
+  "epoch": 17.77777777777778,
   "eval_steps": 500,
-  "global_step": 18,
+  "global_step": 120,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -11,53 +11,265 @@
     {
       "epoch": 0.8888888888888888,
       "eval_accuracy": 0.6595744680851063,
-      "eval_loss": 0.6631675362586975,
-      "eval_runtime": 0.7139,
-      "eval_samples_per_second": 131.668,
-      "eval_steps_per_second": 4.202,
+      "eval_loss": 0.6158677935600281,
+      "eval_runtime": 0.616,
+      "eval_samples_per_second": 152.588,
+      "eval_steps_per_second": 4.87,
       "step": 6
     },
     {
       "epoch": 1.4814814814814814,
-      "grad_norm": 6.819284439086914,
-      "learning_rate": 2.5e-05,
-      "loss": 0.6636,
+      "grad_norm": 8.271566390991211,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.5542,
       "step": 10
     },
     {
       "epoch": 1.925925925925926,
-      "eval_accuracy": 0.6595744680851063,
-      "eval_loss": 0.6286473870277405,
-      "eval_runtime": 0.661,
-      "eval_samples_per_second": 142.213,
-      "eval_steps_per_second": 4.539,
+      "eval_accuracy": 0.7127659574468085,
+      "eval_loss": 0.5690343379974365,
+      "eval_runtime": 0.5828,
+      "eval_samples_per_second": 161.284,
+      "eval_steps_per_second": 5.147,
       "step": 13
     },
     {
-      "epoch": 2.6666666666666665,
-      "eval_accuracy": 0.6595744680851063,
-      "eval_loss": 0.6195614337921143,
-      "eval_runtime": 0.5956,
-      "eval_samples_per_second": 157.837,
-      "eval_steps_per_second": 5.037,
-      "step": 18
-    },
-    {
-      "epoch": 2.6666666666666665,
-      "step": 18,
-      "total_flos": 5.5925975218176e+16,
-      "train_loss": 0.6220952669779459,
-      "train_runtime": 39.3173,
-      "train_samples_per_second": 63.865,
-      "train_steps_per_second": 0.458
+      "epoch": 2.962962962962963,
+      "grad_norm": 9.217803001403809,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 0.5334,
+      "step": 20
+    },
+    {
+      "epoch": 2.962962962962963,
+      "eval_accuracy": 0.7127659574468085,
+      "eval_loss": 0.5524399280548096,
+      "eval_runtime": 0.6208,
+      "eval_samples_per_second": 151.413,
+      "eval_steps_per_second": 4.832,
+      "step": 20
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.7021276595744681,
+      "eval_loss": 0.5923256874084473,
+      "eval_runtime": 0.6139,
+      "eval_samples_per_second": 153.118,
+      "eval_steps_per_second": 4.887,
+      "step": 27
+    },
+    {
+      "epoch": 4.444444444444445,
+      "grad_norm": 8.293265342712402,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.4658,
+      "step": 30
+    },
+    {
+      "epoch": 4.888888888888889,
+      "eval_accuracy": 0.7659574468085106,
+      "eval_loss": 0.4978613257408142,
+      "eval_runtime": 0.7172,
+      "eval_samples_per_second": 131.065,
+      "eval_steps_per_second": 4.183,
+      "step": 33
+    },
+    {
+      "epoch": 5.925925925925926,
+      "grad_norm": 7.5387701988220215,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.469,
+      "step": 40
+    },
+    {
+      "epoch": 5.925925925925926,
+      "eval_accuracy": 0.776595744680851,
+      "eval_loss": 0.4834465980529785,
+      "eval_runtime": 0.6072,
+      "eval_samples_per_second": 154.816,
+      "eval_steps_per_second": 4.941,
+      "step": 40
+    },
+    {
+      "epoch": 6.962962962962963,
+      "eval_accuracy": 0.7446808510638298,
+      "eval_loss": 0.5267057418823242,
+      "eval_runtime": 0.604,
+      "eval_samples_per_second": 155.621,
+      "eval_steps_per_second": 4.967,
+      "step": 47
+    },
+    {
+      "epoch": 7.407407407407407,
+      "grad_norm": 7.728250026702881,
+      "learning_rate": 3.240740740740741e-05,
+      "loss": 0.4469,
+      "step": 50
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.723404255319149,
+      "eval_loss": 0.5309192538261414,
+      "eval_runtime": 0.6111,
+      "eval_samples_per_second": 153.821,
+      "eval_steps_per_second": 4.909,
+      "step": 54
+    },
+    {
+      "epoch": 8.88888888888889,
+      "grad_norm": 12.034634590148926,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.436,
+      "step": 60
+    },
+    {
+      "epoch": 8.88888888888889,
+      "eval_accuracy": 0.7872340425531915,
+      "eval_loss": 0.4591914117336273,
+      "eval_runtime": 0.6041,
+      "eval_samples_per_second": 155.611,
+      "eval_steps_per_second": 4.966,
+      "step": 60
+    },
+    {
+      "epoch": 9.925925925925926,
+      "eval_accuracy": 0.8191489361702128,
+      "eval_loss": 0.4601520001888275,
+      "eval_runtime": 0.6315,
+      "eval_samples_per_second": 148.848,
+      "eval_steps_per_second": 4.75,
+      "step": 67
+    },
+    {
+      "epoch": 10.37037037037037,
+      "grad_norm": 8.522727012634277,
+      "learning_rate": 2.314814814814815e-05,
+      "loss": 0.3978,
+      "step": 70
+    },
+    {
+      "epoch": 10.962962962962964,
+      "eval_accuracy": 0.8085106382978723,
+      "eval_loss": 0.479958176612854,
+      "eval_runtime": 0.7165,
+      "eval_samples_per_second": 131.191,
+      "eval_steps_per_second": 4.187,
+      "step": 74
+    },
+    {
+      "epoch": 11.851851851851851,
+      "grad_norm": 8.708232879638672,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 0.3916,
+      "step": 80
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.8404255319148937,
+      "eval_loss": 0.4538363516330719,
+      "eval_runtime": 0.7058,
+      "eval_samples_per_second": 133.179,
+      "eval_steps_per_second": 4.25,
+      "step": 81
+    },
+    {
+      "epoch": 12.88888888888889,
+      "eval_accuracy": 0.7872340425531915,
+      "eval_loss": 0.4807003438472748,
+      "eval_runtime": 0.7274,
+      "eval_samples_per_second": 129.222,
+      "eval_steps_per_second": 4.124,
+      "step": 87
+    },
+    {
+      "epoch": 13.333333333333334,
+      "grad_norm": 10.32046127319336,
+      "learning_rate": 1.388888888888889e-05,
+      "loss": 0.3932,
+      "step": 90
+    },
+    {
+      "epoch": 13.925925925925926,
+      "eval_accuracy": 0.8297872340425532,
+      "eval_loss": 0.4472525715827942,
+      "eval_runtime": 0.6791,
+      "eval_samples_per_second": 138.408,
+      "eval_steps_per_second": 4.417,
+      "step": 94
+    },
+    {
+      "epoch": 14.814814814814815,
+      "grad_norm": 7.876950263977051,
+      "learning_rate": 9.259259259259259e-06,
+      "loss": 0.3362,
+      "step": 100
+    },
+    {
+      "epoch": 14.962962962962964,
+      "eval_accuracy": 0.8085106382978723,
+      "eval_loss": 0.4526437520980835,
+      "eval_runtime": 0.6231,
+      "eval_samples_per_second": 150.861,
+      "eval_steps_per_second": 4.815,
+      "step": 101
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.8191489361702128,
+      "eval_loss": 0.4435100257396698,
+      "eval_runtime": 0.6246,
+      "eval_samples_per_second": 150.502,
+      "eval_steps_per_second": 4.803,
+      "step": 108
+    },
+    {
+      "epoch": 16.296296296296298,
+      "grad_norm": 7.970451354980469,
+      "learning_rate": 4.6296296296296296e-06,
+      "loss": 0.3631,
+      "step": 110
+    },
+    {
+      "epoch": 16.88888888888889,
+      "eval_accuracy": 0.8297872340425532,
+      "eval_loss": 0.4465709328651428,
+      "eval_runtime": 0.6157,
+      "eval_samples_per_second": 152.667,
+      "eval_steps_per_second": 4.872,
+      "step": 114
+    },
+    {
+      "epoch": 17.77777777777778,
+      "grad_norm": 7.780444622039795,
+      "learning_rate": 0.0,
+      "loss": 0.3441,
+      "step": 120
+    },
+    {
+      "epoch": 17.77777777777778,
+      "eval_accuracy": 0.8297872340425532,
+      "eval_loss": 0.45035284757614136,
+      "eval_runtime": 0.6091,
+      "eval_samples_per_second": 154.338,
+      "eval_steps_per_second": 4.926,
+      "step": 120
+    },
+    {
+      "epoch": 17.77777777777778,
+      "step": 120,
+      "total_flos": 3.7037909187824026e+17,
+      "train_loss": 0.427582456668218,
+      "train_runtime": 256.3024,
+      "train_samples_per_second": 65.313,
+      "train_steps_per_second": 0.468
     }
   ],
   "logging_steps": 10,
-  "max_steps": 18,
+  "max_steps": 120,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 3,
+  "num_train_epochs": 20,
   "save_steps": 500,
-  "total_flos": 5.5925975218176e+16,
+  "total_flos": 3.7037909187824026e+17,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null