🍻 cheers

all_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 2.21,
-    "eval_accuracy": 0.704446442154761,
-    "eval_loss": 1.016680121421814,
-    "eval_runtime": 152.7892,
-    "eval_samples_per_second": 189.294,
-    "eval_steps_per_second": 23.667,
-    "train_loss": 0.965315860748291,
-    "train_runtime": 7707.001,
-    "train_samples_per_second": 150.108,
-    "train_steps_per_second": 4.691
+    "epoch": 2.49,
+    "eval_accuracy": 0.7163750777954498,
+    "eval_loss": 0.9792525172233582,
+    "eval_runtime": 155.2753,
+    "eval_samples_per_second": 186.263,
+    "eval_steps_per_second": 23.288,
+    "train_loss": 0.9232679036458333,
+    "train_runtime": 8603.1753,
+    "train_samples_per_second": 134.472,
+    "train_steps_per_second": 4.203
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 2.21,
-    "eval_accuracy": 0.704446442154761,
-    "eval_loss": 1.016680121421814,
-    "eval_runtime": 152.7892,
-    "eval_samples_per_second": 189.294,
-    "eval_steps_per_second": 23.667
+    "epoch": 2.49,
+    "eval_accuracy": 0.7163750777954498,
+    "eval_loss": 0.9792525172233582,
+    "eval_runtime": 155.2753,
+    "eval_samples_per_second": 186.263,
+    "eval_steps_per_second": 23.288
 }

runs/May28_21-20-54_adrian-development/events.out.tfevents.1685310471.adrian-development.89081.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e447789ec67dc46db5a3c4c34da1588a275dcfbb741973a9a3dc72541925923
+size 369

train_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 2.21,
-    "train_loss": 0.965315860748291,
-    "train_runtime": 7707.001,
-    "train_samples_per_second": 150.108,
-    "train_steps_per_second": 4.691
+    "epoch": 2.49,
+    "train_loss": 0.9232679036458333,
+    "train_runtime": 8603.1753,
+    "train_samples_per_second": 134.472,
+    "train_steps_per_second": 4.203
 }

trainer_state.json

@@ -1,265 +1,295 @@
 {
-  "best_metric": 1.016680121421814,
+  "best_metric": 0.9792525172233582,
   "best_model_checkpoint": "Transformers/vit-base-clothing-leafs-example-full-simple/checkpoint-13000",
-  "epoch": 2.212695339510441,
-  "global_step": 16000,
+  "epoch": 2.489282256949246,
+  "global_step": 18000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.14,
-      "learning_rate": 9.723413082561195e-05,
-      "loss": 1.5425,
+      "learning_rate": 4.8617065412805976e-05,
+      "loss": 1.7133,
       "step": 1000
     },
     {
       "epoch": 0.14,
-      "eval_accuracy": 0.6361593250812531,
-      "eval_loss": 1.2916964292526245,
-      "eval_runtime": 156.1905,
-      "eval_samples_per_second": 185.171,
-      "eval_steps_per_second": 23.151,
+      "eval_accuracy": 0.6511306272042044,
+      "eval_loss": 1.298416018486023,
+      "eval_runtime": 153.5861,
+      "eval_samples_per_second": 188.311,
+      "eval_steps_per_second": 23.544,
       "step": 1000
     },
     {
       "epoch": 0.28,
-      "learning_rate": 9.44682616512239e-05,
-      "loss": 1.2,
+      "learning_rate": 4.723551376019915e-05,
+      "loss": 1.2132,
       "step": 2000
     },
     {
       "epoch": 0.28,
-      "eval_accuracy": 0.663439596155176,
-      "eval_loss": 1.1727856397628784,
-      "eval_runtime": 154.4853,
-      "eval_samples_per_second": 187.215,
-      "eval_steps_per_second": 23.407,
+      "eval_accuracy": 0.6832515040453634,
+      "eval_loss": 1.146278977394104,
+      "eval_runtime": 154.4414,
+      "eval_samples_per_second": 187.268,
+      "eval_steps_per_second": 23.413,
       "step": 2000
     },
     {
       "epoch": 0.41,
-      "learning_rate": 9.170515834601023e-05,
-      "loss": 1.1415,
+      "learning_rate": 4.5853962107592315e-05,
+      "loss": 1.1346,
       "step": 3000
     },
     {
       "epoch": 0.41,
-      "eval_accuracy": 0.6729133531567665,
-      "eval_loss": 1.1330761909484863,
-      "eval_runtime": 154.2347,
-      "eval_samples_per_second": 187.519,
-      "eval_steps_per_second": 23.445,
+      "eval_accuracy": 0.6904086854297766,
+      "eval_loss": 1.0889592170715332,
+      "eval_runtime": 155.3927,
+      "eval_samples_per_second": 186.122,
+      "eval_steps_per_second": 23.27,
       "step": 3000
     },
     {
       "epoch": 0.55,
-      "learning_rate": 8.894205504079658e-05,
-      "loss": 1.1054,
+      "learning_rate": 4.447102752039829e-05,
+      "loss": 1.0888,
       "step": 4000
     },
     {
       "epoch": 0.55,
-      "eval_accuracy": 0.68145356476039,
-      "eval_loss": 1.092808723449707,
-      "eval_runtime": 156.3316,
-      "eval_samples_per_second": 185.004,
-      "eval_steps_per_second": 23.13,
+      "eval_accuracy": 0.6949035336422101,
+      "eval_loss": 1.0718365907669067,
+      "eval_runtime": 154.1959,
+      "eval_samples_per_second": 187.567,
+      "eval_steps_per_second": 23.451,
       "step": 4000
     },
     {
       "epoch": 0.69,
-      "learning_rate": 8.617618586640852e-05,
-      "loss": 1.0592,
+      "learning_rate": 4.308809293320426e-05,
+      "loss": 1.0576,
       "step": 5000
     },
     {
       "epoch": 0.69,
-      "eval_accuracy": 0.6865016250605076,
-      "eval_loss": 1.0824187994003296,
-      "eval_runtime": 155.4175,
-      "eval_samples_per_second": 186.092,
-      "eval_steps_per_second": 23.266,
+      "eval_accuracy": 0.7007814120738538,
+      "eval_loss": 1.0322985649108887,
+      "eval_runtime": 156.7369,
+      "eval_samples_per_second": 184.526,
+      "eval_steps_per_second": 23.071,
       "step": 5000
     },
     {
       "epoch": 0.83,
-      "learning_rate": 8.341031669202047e-05,
-      "loss": 1.0516,
+      "learning_rate": 4.170654128059743e-05,
+      "loss": 1.031,
       "step": 6000
     },
     {
       "epoch": 0.83,
-      "eval_accuracy": 0.6868128068598299,
-      "eval_loss": 1.07091224193573,
-      "eval_runtime": 153.8441,
-      "eval_samples_per_second": 187.996,
-      "eval_steps_per_second": 23.504,
+      "eval_accuracy": 0.7021644422930641,
+      "eval_loss": 1.0212959051132202,
+      "eval_runtime": 155.5747,
+      "eval_samples_per_second": 185.904,
+      "eval_steps_per_second": 23.243,
       "step": 6000
     },
     {
       "epoch": 0.97,
-      "learning_rate": 8.06499792559812e-05,
-      "loss": 1.0335,
+      "learning_rate": 4.03236066934034e-05,
+      "loss": 1.0086,
       "step": 7000
     },
     {
       "epoch": 0.97,
-      "eval_accuracy": 0.6915496853606251,
-      "eval_loss": 1.0542349815368652,
-      "eval_runtime": 154.6828,
-      "eval_samples_per_second": 186.976,
-      "eval_steps_per_second": 23.377,
+      "eval_accuracy": 0.700677684807413,
+      "eval_loss": 1.0171376466751099,
+      "eval_runtime": 156.0917,
+      "eval_samples_per_second": 185.289,
+      "eval_steps_per_second": 23.166,
       "step": 7000
     },
     {
       "epoch": 1.11,
-      "learning_rate": 7.788411008159315e-05,
-      "loss": 0.9049,
+      "learning_rate": 3.894067210620938e-05,
+      "loss": 0.8678,
       "step": 8000
     },
     {
       "epoch": 1.11,
-      "eval_accuracy": 0.6960791093285388,
-      "eval_loss": 1.0505256652832031,
-      "eval_runtime": 154.9548,
-      "eval_samples_per_second": 186.648,
-      "eval_steps_per_second": 23.336,
+      "eval_accuracy": 0.7091141691445958,
+      "eval_loss": 0.9995917081832886,
+      "eval_runtime": 156.6769,
+      "eval_samples_per_second": 184.596,
+      "eval_steps_per_second": 23.079,
       "step": 8000
     },
     {
       "epoch": 1.24,
-      "learning_rate": 7.51182409072051e-05,
-      "loss": 0.855,
+      "learning_rate": 3.755773751901535e-05,
+      "loss": 0.8331,
       "step": 9000
     },
     {
       "epoch": 1.24,
-      "eval_accuracy": 0.7003319272526105,
-      "eval_loss": 1.0373600721359253,
-      "eval_runtime": 153.7792,
-      "eval_samples_per_second": 188.075,
-      "eval_steps_per_second": 23.514,
+      "eval_accuracy": 0.7084226540349907,
+      "eval_loss": 1.0052682161331177,
+      "eval_runtime": 156.5024,
+      "eval_samples_per_second": 184.802,
+      "eval_steps_per_second": 23.105,
       "step": 9000
     },
     {
       "epoch": 1.38,
-      "learning_rate": 7.235237173281704e-05,
-      "loss": 0.8416,
+      "learning_rate": 3.617480293182133e-05,
+      "loss": 0.837,
       "step": 10000
     },
     {
       "epoch": 1.38,
-      "eval_accuracy": 0.6987760182559989,
-      "eval_loss": 1.0450818538665771,
-      "eval_runtime": 153.737,
-      "eval_samples_per_second": 188.126,
-      "eval_steps_per_second": 23.521,
+      "eval_accuracy": 0.7132286840467464,
+      "eval_loss": 0.9893946051597595,
+      "eval_runtime": 155.3223,
+      "eval_samples_per_second": 186.206,
+      "eval_steps_per_second": 23.281,
       "step": 10000
     },
     {
       "epoch": 1.52,
-      "learning_rate": 6.958650255842899e-05,
-      "loss": 0.8389,
+      "learning_rate": 3.4793251279214494e-05,
+      "loss": 0.8405,
       "step": 11000
     },
     {
       "epoch": 1.52,
-      "eval_accuracy": 0.7001936242306894,
-      "eval_loss": 1.0445177555084229,
-      "eval_runtime": 154.5258,
-      "eval_samples_per_second": 187.166,
-      "eval_steps_per_second": 23.401,
+      "eval_accuracy": 0.7134707143351082,
+      "eval_loss": 0.9912254810333252,
+      "eval_runtime": 155.9637,
+      "eval_samples_per_second": 185.441,
+      "eval_steps_per_second": 23.185,
       "step": 11000
     },
     {
       "epoch": 1.66,
-      "learning_rate": 6.682063338404095e-05,
-      "loss": 0.8565,
+      "learning_rate": 3.341169962660766e-05,
+      "loss": 0.8273,
       "step": 12000
     },
     {
       "epoch": 1.66,
-      "eval_accuracy": 0.7031671392019916,
-      "eval_loss": 1.0185155868530273,
-      "eval_runtime": 155.9035,
-      "eval_samples_per_second": 185.512,
-      "eval_steps_per_second": 23.194,
+      "eval_accuracy": 0.7159601687296867,
+      "eval_loss": 0.9918663501739502,
+      "eval_runtime": 155.3075,
+      "eval_samples_per_second": 186.224,
+      "eval_steps_per_second": 23.283,
       "step": 12000
     },
     {
       "epoch": 1.8,
-      "learning_rate": 6.406029594800165e-05,
-      "loss": 0.8621,
+      "learning_rate": 3.202876503941364e-05,
+      "loss": 0.8371,
       "step": 13000
     },
     {
       "epoch": 1.8,
-      "eval_accuracy": 0.704446442154761,
-      "eval_loss": 1.016680121421814,
-      "eval_runtime": 153.8181,
-      "eval_samples_per_second": 188.027,
-      "eval_steps_per_second": 23.508,
+      "eval_accuracy": 0.7163750777954498,
+      "eval_loss": 0.9792525172233582,
+      "eval_runtime": 155.0944,
+      "eval_samples_per_second": 186.48,
+      "eval_steps_per_second": 23.315,
       "step": 13000
     },
     {
       "epoch": 1.94,
-      "learning_rate": 6.129442677361361e-05,
-      "loss": 0.8402,
+      "learning_rate": 3.064583045221961e-05,
+      "loss": 0.8225,
       "step": 14000
     },
     {
       "epoch": 1.94,
-      "eval_accuracy": 0.7033054422239126,
-      "eval_loss": 1.0216416120529175,
-      "eval_runtime": 155.4829,
-      "eval_samples_per_second": 186.014,
-      "eval_steps_per_second": 23.257,
+      "eval_accuracy": 0.7162367747735288,
+      "eval_loss": 0.9806181788444519,
+      "eval_runtime": 155.9339,
+      "eval_samples_per_second": 185.476,
+      "eval_steps_per_second": 23.189,
       "step": 14000
     },
     {
       "epoch": 2.07,
-      "learning_rate": 5.852855759922557e-05,
-      "loss": 0.7132,
+      "learning_rate": 2.9264278799612784e-05,
+      "loss": 0.7055,
       "step": 15000
     },
     {
       "epoch": 2.07,
-      "eval_accuracy": 0.7061752299287739,
-      "eval_loss": 1.047763466835022,
-      "eval_runtime": 154.3004,
-      "eval_samples_per_second": 187.44,
-      "eval_steps_per_second": 23.435,
+      "eval_accuracy": 0.7169628656386142,
+      "eval_loss": 1.0149892568588257,
+      "eval_runtime": 155.1376,
+      "eval_samples_per_second": 186.428,
+      "eval_steps_per_second": 23.308,
       "step": 15000
     },
     {
       "epoch": 2.21,
-      "learning_rate": 5.576268842483751e-05,
-      "loss": 0.5989,
+      "learning_rate": 2.7881344212418754e-05,
+      "loss": 0.6012,
       "step": 16000
     },
     {
       "epoch": 2.21,
-      "eval_accuracy": 0.7044810179102413,
-      "eval_loss": 1.0767056941986084,
-      "eval_runtime": 155.7802,
-      "eval_samples_per_second": 185.659,
-      "eval_steps_per_second": 23.212,
+      "eval_accuracy": 0.7123988659152203,
+      "eval_loss": 1.0372015237808228,
+      "eval_runtime": 155.042,
+      "eval_samples_per_second": 186.543,
+      "eval_steps_per_second": 23.323,
       "step": 16000
     },
     {
-      "epoch": 2.21,
-      "step": 16000,
-      "total_flos": 3.968921951752827e+19,
-      "train_loss": 0.965315860748291,
-      "train_runtime": 7707.001,
-      "train_samples_per_second": 150.108,
-      "train_steps_per_second": 4.691
+      "epoch": 2.35,
+      "learning_rate": 2.649840962522473e-05,
+      "loss": 0.6035,
+      "step": 17000
+    },
+    {
+      "epoch": 2.35,
+      "eval_accuracy": 0.71236429015974,
+      "eval_loss": 1.0361833572387695,
+      "eval_runtime": 156.1808,
+      "eval_samples_per_second": 185.183,
+      "eval_steps_per_second": 23.153,
+      "step": 17000
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.51154750380307e-05,
+      "loss": 0.596,
+      "step": 18000
+    },
+    {
+      "epoch": 2.49,
+      "eval_accuracy": 0.7120531083604177,
+      "eval_loss": 1.0460479259490967,
+      "eval_runtime": 157.7455,
+      "eval_samples_per_second": 183.346,
+      "eval_steps_per_second": 22.923,
+      "step": 18000
+    },
+    {
+      "epoch": 2.49,
+      "step": 18000,
+      "total_flos": 4.465066266677883e+19,
+      "train_loss": 0.9232679036458333,
+      "train_runtime": 8603.1753,
+      "train_samples_per_second": 134.472,
+      "train_steps_per_second": 4.203
     }
   ],
   "max_steps": 36155,
   "num_train_epochs": 5,
-  "total_flos": 3.968921951752827e+19,
+  "total_flos": 4.465066266677883e+19,
   "trial_name": null,
   "trial_params": null
 }