🍻 cheers

all_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 4.0,
-    "eval_accuracy": 0.7126063204481018,
-    "eval_loss": 0.9990608096122742,
-    "eval_runtime": 155.7648,
-    "eval_samples_per_second": 185.677,
-    "eval_steps_per_second": 23.214,
-    "train_loss": 0.9726295792995491,
-    "train_runtime": 14146.7049,
-    "train_samples_per_second": 65.422,
-    "train_steps_per_second": 2.045
+    "eval_accuracy": 0.7155106839084434,
+    "eval_loss": 0.9953664541244507,
+    "eval_runtime": 153.0701,
+    "eval_samples_per_second": 188.946,
+    "eval_steps_per_second": 23.623,
+    "train_loss": 0.9035214643197474,
+    "train_runtime": 13914.0149,
+    "train_samples_per_second": 66.516,
+    "train_steps_per_second": 2.079
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 4.0,
-    "eval_accuracy": 0.7126063204481018,
-    "eval_loss": 0.9990608096122742,
-    "eval_runtime": 155.7648,
-    "eval_samples_per_second": 185.677,
-    "eval_steps_per_second": 23.214
+    "eval_accuracy": 0.7155106839084434,
+    "eval_loss": 0.9953664541244507,
+    "eval_runtime": 153.0701,
+    "eval_samples_per_second": 188.946,
+    "eval_steps_per_second": 23.623
 }

runs/Jun09_16-17-56_adrian-development/events.out.tfevents.1686334369.adrian-development.7991.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:592626b8cebeedc1bb9fe80ce040a64910094eb0f5325b096d5c36f8bdfec215
+size 369

train_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 4.0,
-    "train_loss": 0.9726295792995491,
-    "train_runtime": 14146.7049,
-    "train_samples_per_second": 65.422,
-    "train_steps_per_second": 2.045
+    "train_loss": 0.9035214643197474,
+    "train_runtime": 13914.0149,
+    "train_samples_per_second": 66.516,
+    "train_steps_per_second": 2.079
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.9990608096122742,
-  "best_model_checkpoint": "Transformers/vit-base-clothing-leafs-example-full-simple/checkpoint-21000",
+  "best_metric": 0.9953664541244507,
+  "best_model_checkpoint": "Transformers/vit-base-clothing-leafs-example-full-simple/checkpoint-20000",
   "epoch": 4.0,
   "global_step": 28924,
   "is_hyper_param_search": false,
@@ -9,432 +9,432 @@
   "log_history": [
     {
       "epoch": 0.14,
-      "learning_rate": 1.930853270640299e-05,
-      "loss": 2.1442,
+      "learning_rate": 2.8963836260544875e-05,
+      "loss": 1.9495,
       "step": 1000
     },
     {
       "epoch": 0.14,
-      "eval_accuracy": 0.6229513864877948,
-      "eval_loss": 1.6088389158248901,
-      "eval_runtime": 166.6999,
-      "eval_samples_per_second": 173.497,
-      "eval_steps_per_second": 21.692,
+      "eval_accuracy": 0.6307309314708527,
+      "eval_loss": 1.4553368091583252,
+      "eval_runtime": 164.2927,
+      "eval_samples_per_second": 176.039,
+      "eval_steps_per_second": 22.009,
       "step": 1000
     },
     {
       "epoch": 0.28,
-      "learning_rate": 1.861844834739317e-05,
-      "loss": 1.4492,
+      "learning_rate": 2.7926635320149357e-05,
+      "loss": 1.3079,
       "step": 2000
     },
     {
       "epoch": 0.28,
-      "eval_accuracy": 0.6612267478044396,
-      "eval_loss": 1.3310657739639282,
-      "eval_runtime": 159.1133,
-      "eval_samples_per_second": 181.77,
-      "eval_steps_per_second": 22.726,
+      "eval_accuracy": 0.6676924140792476,
+      "eval_loss": 1.2347172498703003,
+      "eval_runtime": 157.0046,
+      "eval_samples_per_second": 184.211,
+      "eval_steps_per_second": 23.031,
       "step": 2000
     },
     {
       "epoch": 0.41,
-      "learning_rate": 1.7926981053796158e-05,
-      "loss": 1.2669,
+      "learning_rate": 2.6890471580694235e-05,
+      "loss": 1.178,
       "step": 3000
     },
     {
       "epoch": 0.41,
-      "eval_accuracy": 0.6771315953253578,
-      "eval_loss": 1.2137295007705688,
-      "eval_runtime": 158.2416,
-      "eval_samples_per_second": 182.771,
-      "eval_steps_per_second": 22.851,
+      "eval_accuracy": 0.6757831408616278,
+      "eval_loss": 1.1606581211090088,
+      "eval_runtime": 154.1382,
+      "eval_samples_per_second": 187.637,
+      "eval_steps_per_second": 23.459,
       "step": 3000
     },
     {
       "epoch": 0.55,
-      "learning_rate": 1.7235513760199145e-05,
-      "loss": 1.1901,
+      "learning_rate": 2.5853270640298713e-05,
+      "loss": 1.1324,
       "step": 4000
     },
     {
       "epoch": 0.55,
-      "eval_accuracy": 0.6789641103658115,
-      "eval_loss": 1.1688439846038818,
-      "eval_runtime": 159.3121,
-      "eval_samples_per_second": 181.543,
-      "eval_steps_per_second": 22.698,
+      "eval_accuracy": 0.682387110158357,
+      "eval_loss": 1.1307393312454224,
+      "eval_runtime": 154.2255,
+      "eval_samples_per_second": 187.531,
+      "eval_steps_per_second": 23.446,
       "step": 4000
     },
     {
       "epoch": 0.69,
-      "learning_rate": 1.6544737933895728e-05,
-      "loss": 1.1535,
+      "learning_rate": 2.4816069699903195e-05,
+      "loss": 1.0928,
       "step": 5000
     },
     {
       "epoch": 0.69,
-      "eval_accuracy": 0.6876080492358758,
-      "eval_loss": 1.122076153755188,
-      "eval_runtime": 159.944,
-      "eval_samples_per_second": 180.826,
-      "eval_steps_per_second": 22.608,
+      "eval_accuracy": 0.6908927460065002,
+      "eval_loss": 1.0955638885498047,
+      "eval_runtime": 152.3863,
+      "eval_samples_per_second": 189.794,
+      "eval_steps_per_second": 23.729,
       "step": 5000
     },
     {
       "epoch": 0.83,
-      "learning_rate": 1.5853270640298714e-05,
-      "loss": 1.1027,
+      "learning_rate": 2.3778868759507677e-05,
+      "loss": 1.0679,
       "step": 6000
     },
     {
       "epoch": 0.83,
-      "eval_accuracy": 0.6927252610469539,
-      "eval_loss": 1.092621922492981,
-      "eval_runtime": 160.5515,
-      "eval_samples_per_second": 180.142,
-      "eval_steps_per_second": 22.522,
+      "eval_accuracy": 0.6912385035613028,
+      "eval_loss": 1.0789850950241089,
+      "eval_runtime": 153.243,
+      "eval_samples_per_second": 188.733,
+      "eval_steps_per_second": 23.597,
       "step": 6000
     },
     {
       "epoch": 0.97,
-      "learning_rate": 1.5162494813995299e-05,
-      "loss": 1.0715,
+      "learning_rate": 2.274166781911216e-05,
+      "loss": 1.0488,
       "step": 7000
     },
     {
       "epoch": 0.97,
-      "eval_accuracy": 0.6932438973791577,
-      "eval_loss": 1.0782523155212402,
-      "eval_runtime": 160.1635,
-      "eval_samples_per_second": 180.578,
-      "eval_steps_per_second": 22.577,
+      "eval_accuracy": 0.7013691999170182,
+      "eval_loss": 1.0485832691192627,
+      "eval_runtime": 155.3074,
+      "eval_samples_per_second": 186.224,
+      "eval_steps_per_second": 23.283,
       "step": 7000
     },
     {
       "epoch": 1.11,
-      "learning_rate": 1.4471027520398286e-05,
-      "loss": 1.0052,
+      "learning_rate": 2.1705504079657033e-05,
+      "loss": 0.9548,
       "step": 8000
     },
     {
       "epoch": 1.11,
-      "eval_accuracy": 0.7011963211396169,
-      "eval_loss": 1.0529263019561768,
-      "eval_runtime": 159.1353,
-      "eval_samples_per_second": 181.745,
-      "eval_steps_per_second": 22.723,
+      "eval_accuracy": 0.70161123020538,
+      "eval_loss": 1.0449484586715698,
+      "eval_runtime": 154.3409,
+      "eval_samples_per_second": 187.39,
+      "eval_steps_per_second": 23.429,
       "step": 8000
     },
     {
       "epoch": 1.24,
-      "learning_rate": 1.378025169409487e-05,
-      "loss": 0.9777,
+      "learning_rate": 2.0668303139261512e-05,
+      "loss": 0.9352,
       "step": 9000
     },
     {
       "epoch": 1.24,
-      "eval_accuracy": 0.7021990180485443,
-      "eval_loss": 1.044997215270996,
-      "eval_runtime": 157.3717,
-      "eval_samples_per_second": 183.782,
-      "eval_steps_per_second": 22.977,
+      "eval_accuracy": 0.7042044118663993,
+      "eval_loss": 1.034779667854309,
+      "eval_runtime": 155.2786,
+      "eval_samples_per_second": 186.259,
+      "eval_steps_per_second": 23.287,
       "step": 9000
     },
     {
       "epoch": 1.38,
-      "learning_rate": 1.3088784400497859e-05,
-      "loss": 0.9798,
+      "learning_rate": 1.963213939980639e-05,
+      "loss": 0.9164,
       "step": 10000
     },
     {
       "epoch": 1.38,
-      "eval_accuracy": 0.7012654726505774,
-      "eval_loss": 1.044914960861206,
-      "eval_runtime": 157.2779,
-      "eval_samples_per_second": 183.891,
-      "eval_steps_per_second": 22.991,
+      "eval_accuracy": 0.7033745937348731,
+      "eval_loss": 1.0340265035629272,
+      "eval_runtime": 154.4587,
+      "eval_samples_per_second": 187.247,
+      "eval_steps_per_second": 23.411,
       "step": 10000
     },
     {
       "epoch": 1.52,
-      "learning_rate": 1.2397317106900844e-05,
-      "loss": 0.9473,
+      "learning_rate": 1.8595975660351264e-05,
+      "loss": 0.9267,
       "step": 11000
     },
     {
       "epoch": 1.52,
-      "eval_accuracy": 0.7041006845999586,
-      "eval_loss": 1.0282418727874756,
-      "eval_runtime": 159.9465,
-      "eval_samples_per_second": 180.823,
-      "eval_steps_per_second": 22.608,
+      "eval_accuracy": 0.7089412903671946,
+      "eval_loss": 1.0178329944610596,
+      "eval_runtime": 153.0672,
+      "eval_samples_per_second": 188.95,
+      "eval_steps_per_second": 23.624,
       "step": 11000
     },
     {
       "epoch": 1.66,
-      "learning_rate": 1.1706541280597427e-05,
-      "loss": 0.9463,
+      "learning_rate": 1.7558774719955746e-05,
+      "loss": 0.9058,
       "step": 12000
     },
     {
       "epoch": 1.66,
-      "eval_accuracy": 0.6971855335039071,
-      "eval_loss": 1.04224693775177,
-      "eval_runtime": 160.2925,
-      "eval_samples_per_second": 180.433,
-      "eval_steps_per_second": 22.559,
+      "eval_accuracy": 0.7063481087061753,
+      "eval_loss": 1.0159752368927002,
+      "eval_runtime": 153.6019,
+      "eval_samples_per_second": 188.292,
+      "eval_steps_per_second": 23.541,
       "step": 12000
     },
     {
       "epoch": 1.8,
-      "learning_rate": 1.1015765454294012e-05,
-      "loss": 0.9508,
+      "learning_rate": 1.6521573779560228e-05,
+      "loss": 0.9028,
       "step": 13000
     },
     {
       "epoch": 1.8,
-      "eval_accuracy": 0.7025793513588272,
-      "eval_loss": 1.0288690328598022,
-      "eval_runtime": 164.2839,
-      "eval_samples_per_second": 176.049,
-      "eval_steps_per_second": 22.011,
+      "eval_accuracy": 0.7110504114514902,
+      "eval_loss": 1.0084210634231567,
+      "eval_runtime": 154.4213,
+      "eval_samples_per_second": 187.293,
+      "eval_steps_per_second": 23.416,
       "step": 13000
     },
     {
       "epoch": 1.94,
-      "learning_rate": 1.0324298160697e-05,
-      "loss": 0.9447,
+      "learning_rate": 1.548437283916471e-05,
+      "loss": 0.9093,
       "step": 14000
     },
     {
       "epoch": 1.94,
-      "eval_accuracy": 0.7070396238157803,
-      "eval_loss": 1.0159052610397339,
-      "eval_runtime": 159.9439,
-      "eval_samples_per_second": 180.826,
-      "eval_steps_per_second": 22.608,
+      "eval_accuracy": 0.7136090173570292,
+      "eval_loss": 1.00092351436615,
+      "eval_runtime": 151.8217,
+      "eval_samples_per_second": 190.5,
+      "eval_steps_per_second": 23.817,
       "step": 14000
     },
     {
       "epoch": 2.07,
-      "learning_rate": 9.632830867099987e-06,
-      "loss": 0.8831,
+      "learning_rate": 1.4447171898769188e-05,
+      "loss": 0.8346,
       "step": 15000
     },
     {
       "epoch": 2.07,
-      "eval_accuracy": 0.7081806237466288,
-      "eval_loss": 1.0163402557373047,
-      "eval_runtime": 159.7911,
-      "eval_samples_per_second": 180.999,
-      "eval_steps_per_second": 22.63,
+      "eval_accuracy": 0.7117073508056151,
+      "eval_loss": 1.0151883363723755,
+      "eval_runtime": 153.6777,
+      "eval_samples_per_second": 188.199,
+      "eval_steps_per_second": 23.53,
       "step": 15000
     },
     {
       "epoch": 2.21,
-      "learning_rate": 8.941363573502974e-06,
-      "loss": 0.8581,
+      "learning_rate": 1.3411008159314064e-05,
+      "loss": 0.7897,
       "step": 16000
     },
     {
       "epoch": 2.21,
-      "eval_accuracy": 0.7075582601479843,
-      "eval_loss": 1.0116521120071411,
-      "eval_runtime": 157.9996,
-      "eval_samples_per_second": 183.051,
-      "eval_steps_per_second": 22.886,
+      "eval_accuracy": 0.7141276536892331,
+      "eval_loss": 1.0071992874145508,
+      "eval_runtime": 153.402,
+      "eval_samples_per_second": 188.537,
+      "eval_steps_per_second": 23.572,
       "step": 16000
     },
     {
       "epoch": 2.35,
-      "learning_rate": 8.249896279905962e-06,
-      "loss": 0.8453,
+      "learning_rate": 1.237484441985894e-05,
+      "loss": 0.7869,
       "step": 17000
     },
     {
       "epoch": 2.35,
-      "eval_accuracy": 0.7117765023165756,
-      "eval_loss": 1.004479169845581,
-      "eval_runtime": 163.1784,
-      "eval_samples_per_second": 177.242,
-      "eval_steps_per_second": 22.16,
+      "eval_accuracy": 0.7082843510130696,
+      "eval_loss": 1.0088136196136475,
+      "eval_runtime": 153.7395,
+      "eval_samples_per_second": 188.123,
+      "eval_steps_per_second": 23.52,
       "step": 17000
     },
     {
       "epoch": 2.49,
-      "learning_rate": 7.559120453602545e-06,
-      "loss": 0.8422,
+      "learning_rate": 1.133764347946342e-05,
+      "loss": 0.7853,
       "step": 18000
     },
     {
       "epoch": 2.49,
-      "eval_accuracy": 0.7068321692828988,
-      "eval_loss": 1.0124915838241577,
-      "eval_runtime": 158.2344,
-      "eval_samples_per_second": 182.779,
-      "eval_steps_per_second": 22.852,
+      "eval_accuracy": 0.7162021990180485,
+      "eval_loss": 0.9981487989425659,
+      "eval_runtime": 153.6335,
+      "eval_samples_per_second": 188.253,
+      "eval_steps_per_second": 23.537,
       "step": 18000
     },
     {
       "epoch": 2.63,
-      "learning_rate": 6.86834462729913e-06,
-      "loss": 0.8451,
+      "learning_rate": 1.0300442539067902e-05,
+      "loss": 0.7732,
       "step": 19000
     },
     {
       "epoch": 2.63,
-      "eval_accuracy": 0.7045155936657216,
-      "eval_loss": 1.0194756984710693,
-      "eval_runtime": 159.3072,
-      "eval_samples_per_second": 181.549,
-      "eval_steps_per_second": 22.698,
+      "eval_accuracy": 0.7148537445543185,
+      "eval_loss": 1.003010630607605,
+      "eval_runtime": 152.6681,
+      "eval_samples_per_second": 189.444,
+      "eval_steps_per_second": 23.685,
       "step": 19000
     },
     {
       "epoch": 2.77,
-      "learning_rate": 6.176877333702116e-06,
-      "loss": 0.8348,
+      "learning_rate": 9.264278799612779e-06,
+      "loss": 0.779,
       "step": 20000
     },
     {
       "epoch": 2.77,
-      "eval_accuracy": 0.7130903810248254,
-      "eval_loss": 0.9999533891677856,
-      "eval_runtime": 158.3242,
-      "eval_samples_per_second": 182.676,
-      "eval_steps_per_second": 22.839,
+      "eval_accuracy": 0.7155106839084434,
+      "eval_loss": 0.9953664541244507,
+      "eval_runtime": 154.6075,
+      "eval_samples_per_second": 187.067,
+      "eval_steps_per_second": 23.388,
       "step": 20000
     },
     {
       "epoch": 2.9,
-      "learning_rate": 5.485410040105103e-06,
-      "loss": 0.8319,
+      "learning_rate": 8.227077859217259e-06,
+      "loss": 0.7655,
       "step": 21000
     },
     {
       "epoch": 2.9,
-      "eval_accuracy": 0.7126063204481018,
-      "eval_loss": 0.9990608096122742,
-      "eval_runtime": 158.4722,
-      "eval_samples_per_second": 182.505,
-      "eval_steps_per_second": 22.818,
+      "eval_accuracy": 0.7178618352811009,
+      "eval_loss": 0.9971597790718079,
+      "eval_runtime": 153.1084,
+      "eval_samples_per_second": 188.899,
+      "eval_steps_per_second": 23.617,
       "step": 21000
     },
     {
       "epoch": 3.04,
-      "learning_rate": 4.793942746508091e-06,
-      "loss": 0.8165,
+      "learning_rate": 7.1898769188217404e-06,
+      "loss": 0.74,
       "step": 22000
     },
     {
       "epoch": 3.04,
-      "eval_accuracy": 0.7117765023165756,
-      "eval_loss": 1.0050508975982666,
-      "eval_runtime": 159.8886,
-      "eval_samples_per_second": 180.888,
-      "eval_steps_per_second": 22.616,
+      "eval_accuracy": 0.7138164718899108,
+      "eval_loss": 1.0114420652389526,
+      "eval_runtime": 152.2988,
+      "eval_samples_per_second": 189.903,
+      "eval_steps_per_second": 23.743,
       "step": 22000
     },
     {
       "epoch": 3.18,
-      "learning_rate": 4.103166920204675e-06,
-      "loss": 0.7483,
+      "learning_rate": 6.1526759784262205e-06,
+      "loss": 0.6824,
       "step": 23000
     },
     {
       "epoch": 3.18,
-      "eval_accuracy": 0.713297835557707,
-      "eval_loss": 1.007352352142334,
-      "eval_runtime": 158.6943,
-      "eval_samples_per_second": 182.25,
-      "eval_steps_per_second": 22.786,
+      "eval_accuracy": 0.7130212295138649,
+      "eval_loss": 1.017103910446167,
+      "eval_runtime": 152.678,
+      "eval_samples_per_second": 189.431,
+      "eval_steps_per_second": 23.684,
       "step": 23000
     },
     {
       "epoch": 3.32,
-      "learning_rate": 3.411699626607662e-06,
-      "loss": 0.7689,
+      "learning_rate": 5.116512238971097e-06,
+      "loss": 0.68,
       "step": 24000
     },
     {
       "epoch": 3.32,
-      "eval_accuracy": 0.7097711084987207,
-      "eval_loss": 1.0143715143203735,
-      "eval_runtime": 159.7158,
-      "eval_samples_per_second": 181.084,
-      "eval_steps_per_second": 22.64,
+      "eval_accuracy": 0.7177581080146601,
+      "eval_loss": 1.011143684387207,
+      "eval_runtime": 152.6609,
+      "eval_samples_per_second": 189.453,
+      "eval_steps_per_second": 23.686,
       "step": 24000
     },
     {
       "epoch": 3.46,
-      "learning_rate": 2.7209238003042458e-06,
-      "loss": 0.7574,
+      "learning_rate": 4.079311298575578e-06,
+      "loss": 0.6787,
       "step": 25000
     },
     {
       "epoch": 3.46,
-      "eval_accuracy": 0.7122605628932992,
-      "eval_loss": 1.0074282884597778,
-      "eval_runtime": 155.8079,
-      "eval_samples_per_second": 185.626,
-      "eval_steps_per_second": 23.208,
+      "eval_accuracy": 0.7151303505981605,
+      "eval_loss": 1.0124437808990479,
+      "eval_runtime": 152.3072,
+      "eval_samples_per_second": 189.893,
+      "eval_steps_per_second": 23.741,
       "step": 25000
     },
     {
       "epoch": 3.6,
-      "learning_rate": 2.029456506707233e-06,
-      "loss": 0.7625,
+      "learning_rate": 3.042110358180058e-06,
+      "loss": 0.6808,
       "step": 26000
     },
     {
       "epoch": 3.6,
-      "eval_accuracy": 0.7105317751192863,
-      "eval_loss": 1.0111548900604248,
-      "eval_runtime": 156.0076,
-      "eval_samples_per_second": 185.388,
-      "eval_steps_per_second": 23.178,
+      "eval_accuracy": 0.7149920475762396,
+      "eval_loss": 1.0181164741516113,
+      "eval_runtime": 153.0013,
+      "eval_samples_per_second": 189.031,
+      "eval_steps_per_second": 23.634,
       "step": 26000
     },
     {
       "epoch": 3.73,
-      "learning_rate": 1.33798921311022e-06,
-      "loss": 0.7667,
+      "learning_rate": 2.0059466187249345e-06,
+      "loss": 0.6561,
       "step": 27000
     },
     {
       "epoch": 3.73,
-      "eval_accuracy": 0.7106700781412074,
-      "eval_loss": 1.0088056325912476,
-      "eval_runtime": 157.0208,
-      "eval_samples_per_second": 184.192,
-      "eval_steps_per_second": 23.029,
+      "eval_accuracy": 0.7168245626166931,
+      "eval_loss": 1.0144038200378418,
+      "eval_runtime": 153.8199,
+      "eval_samples_per_second": 188.025,
+      "eval_steps_per_second": 23.508,
       "step": 27000
     },
     {
       "epoch": 3.87,
-      "learning_rate": 6.465219195132071e-07,
-      "loss": 0.7472,
+      "learning_rate": 9.697828792698105e-07,
+      "loss": 0.6611,
       "step": 28000
     },
     {
       "epoch": 3.87,
-      "eval_accuracy": 0.7116727750501348,
-      "eval_loss": 1.008212924003601,
-      "eval_runtime": 155.8357,
-      "eval_samples_per_second": 185.593,
-      "eval_steps_per_second": 23.204,
+      "eval_accuracy": 0.7155106839084434,
+      "eval_loss": 1.0154051780700684,
+      "eval_runtime": 153.3953,
+      "eval_samples_per_second": 188.546,
+      "eval_steps_per_second": 23.573,
       "step": 28000
     },
     {
       "epoch": 4.0,
       "step": 28924,
       "total_flos": 7.174773947150918e+19,
-      "train_loss": 0.9726295792995491,
-      "train_runtime": 14146.7049,
-      "train_samples_per_second": 65.422,
-      "train_steps_per_second": 2.045
+      "train_loss": 0.9035214643197474,
+      "train_runtime": 13914.0149,
+      "train_samples_per_second": 66.516,
+      "train_steps_per_second": 2.079
     }
   ],
   "max_steps": 28924,