🍻 cheers

all_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 4.0,
-    "eval_accuracy": 0.7164788050618907,
-    "eval_loss": 0.9746271967887878,
-    "eval_runtime": 155.7191,
-    "eval_samples_per_second": 185.732,
-    "eval_steps_per_second": 23.221,
-    "train_loss": 0.9355616505553886,
-    "train_runtime": 13547.1303,
-    "train_samples_per_second": 68.318,
-    "train_steps_per_second": 2.135
+    "eval_accuracy": 0.7055528663301293,
+    "eval_loss": 1.0334995985031128,
+    "eval_runtime": 152.4169,
+    "eval_samples_per_second": 189.756,
+    "eval_steps_per_second": 23.724,
+    "train_loss": 1.1119355370668336,
+    "train_runtime": 13795.8304,
+    "train_samples_per_second": 67.086,
+    "train_steps_per_second": 2.097
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 4.0,
-    "eval_accuracy": 0.7164788050618907,
-    "eval_loss": 0.9746271967887878,
-    "eval_runtime": 155.7191,
-    "eval_samples_per_second": 185.732,
-    "eval_steps_per_second": 23.221
+    "eval_accuracy": 0.7055528663301293,
+    "eval_loss": 1.0334995985031128,
+    "eval_runtime": 152.4169,
+    "eval_samples_per_second": 189.756,
+    "eval_steps_per_second": 23.724
 }

runs/Jun07_08-09-43_adrian-development/events.out.tfevents.1686132156.adrian-development.7920.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7fb59c16ce322830415205bda7399949db2853bf92a2f3da3f48b2beb5f7819c
+size 369

train_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 4.0,
-    "train_loss": 0.9355616505553886,
-    "train_runtime": 13547.1303,
-    "train_samples_per_second": 68.318,
-    "train_steps_per_second": 2.135
+    "train_loss": 1.1119355370668336,
+    "train_runtime": 13795.8304,
+    "train_samples_per_second": 67.086,
+    "train_steps_per_second": 2.097
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.9746271967887878,
-  "best_model_checkpoint": "Transformers/vit-base-clothing-leafs-example-full-simple/checkpoint-21000",
+  "best_metric": 1.0334995985031128,
+  "best_model_checkpoint": "Transformers/vit-base-clothing-leafs-example-full-simple/checkpoint-27000",
   "epoch": 4.0,
   "global_step": 28924,
   "is_hyper_param_search": false,
@@ -9,432 +9,432 @@
   "log_history": [
     {
       "epoch": 0.14,
-      "learning_rate": 2.4135665883003737e-05,
-      "loss": 2.0439,
+      "learning_rate": 9.654266353201494e-06,
+      "loss": 2.4588,
       "step": 1000
     },
     {
       "epoch": 0.14,
-      "eval_accuracy": 0.6367125371689372,
-      "eval_loss": 1.4945471286773682,
-      "eval_runtime": 155.4022,
-      "eval_samples_per_second": 186.111,
-      "eval_steps_per_second": 23.269,
+      "eval_accuracy": 0.5605767236014106,
+      "eval_loss": 1.9567219018936157,
+      "eval_runtime": 155.8077,
+      "eval_samples_per_second": 185.626,
+      "eval_steps_per_second": 23.208,
       "step": 1000
     },
     {
       "epoch": 0.28,
-      "learning_rate": 2.327133176600747e-05,
-      "loss": 1.3784,
+      "learning_rate": 9.308878440049787e-06,
+      "loss": 1.7412,
       "step": 2000
     },
     {
       "epoch": 0.28,
-      "eval_accuracy": 0.6662748081045571,
-      "eval_loss": 1.2658541202545166,
-      "eval_runtime": 152.9449,
-      "eval_samples_per_second": 189.101,
-      "eval_steps_per_second": 23.642,
+      "eval_accuracy": 0.6285180831201161,
+      "eval_loss": 1.5931317806243896,
+      "eval_runtime": 148.2867,
+      "eval_samples_per_second": 195.041,
+      "eval_steps_per_second": 24.385,
       "step": 2000
     },
     {
       "epoch": 0.41,
-      "learning_rate": 2.2408726317245195e-05,
-      "loss": 1.2223,
+      "learning_rate": 8.963490526898079e-06,
+      "loss": 1.4825,
       "step": 3000
     },
     {
       "epoch": 0.41,
-      "eval_accuracy": 0.6780305649678445,
-      "eval_loss": 1.176424503326416,
-      "eval_runtime": 155.5666,
-      "eval_samples_per_second": 185.914,
-      "eval_steps_per_second": 23.244,
+      "eval_accuracy": 0.6542770209529079,
+      "eval_loss": 1.4156450033187866,
+      "eval_runtime": 148.2362,
+      "eval_samples_per_second": 195.108,
+      "eval_steps_per_second": 24.394,
       "step": 3000
     },
     {
       "epoch": 0.55,
-      "learning_rate": 2.1544392200248927e-05,
-      "loss": 1.153,
+      "learning_rate": 8.61810261374637e-06,
+      "loss": 1.3499,
       "step": 4000
     },
     {
       "epoch": 0.55,
-      "eval_accuracy": 0.687227715925593,
-      "eval_loss": 1.115257740020752,
-      "eval_runtime": 154.7152,
-      "eval_samples_per_second": 186.937,
-      "eval_steps_per_second": 23.372,
+      "eval_accuracy": 0.6701472927183459,
+      "eval_loss": 1.3057234287261963,
+      "eval_runtime": 148.7022,
+      "eval_samples_per_second": 194.496,
+      "eval_steps_per_second": 24.317,
       "step": 4000
     },
     {
       "epoch": 0.69,
-      "learning_rate": 2.0680058083252663e-05,
-      "loss": 1.1037,
+      "learning_rate": 8.272368966947864e-06,
+      "loss": 1.2777,
       "step": 5000
     },
     {
       "epoch": 0.69,
-      "eval_accuracy": 0.6912385035613028,
-      "eval_loss": 1.0900990962982178,
-      "eval_runtime": 154.22,
-      "eval_samples_per_second": 187.537,
-      "eval_steps_per_second": 23.447,
+      "eval_accuracy": 0.6749187469746214,
+      "eval_loss": 1.2455776929855347,
+      "eval_runtime": 148.0641,
+      "eval_samples_per_second": 195.334,
+      "eval_steps_per_second": 24.422,
       "step": 5000
     },
     {
       "epoch": 0.83,
-      "learning_rate": 1.9816588300373394e-05,
-      "loss": 1.0827,
+      "learning_rate": 7.926635320149357e-06,
+      "loss": 1.2325,
       "step": 6000
     },
     {
       "epoch": 0.83,
-      "eval_accuracy": 0.6933822004010788,
-      "eval_loss": 1.068312406539917,
-      "eval_runtime": 150.0826,
-      "eval_samples_per_second": 192.707,
-      "eval_steps_per_second": 24.093,
+      "eval_accuracy": 0.6833206555563239,
+      "eval_loss": 1.1947654485702515,
+      "eval_runtime": 150.1766,
+      "eval_samples_per_second": 192.587,
+      "eval_steps_per_second": 24.078,
       "step": 6000
     },
     {
       "epoch": 0.97,
-      "learning_rate": 1.8953118517494124e-05,
-      "loss": 1.0646,
+      "learning_rate": 7.5812474069976495e-06,
+      "loss": 1.1932,
       "step": 7000
     },
     {
       "epoch": 0.97,
-      "eval_accuracy": 0.7026485028697877,
-      "eval_loss": 1.037534475326538,
-      "eval_runtime": 148.9104,
-      "eval_samples_per_second": 194.224,
-      "eval_steps_per_second": 24.283,
+      "eval_accuracy": 0.6858101099509024,
+      "eval_loss": 1.1682132482528687,
+      "eval_runtime": 153.9643,
+      "eval_samples_per_second": 187.849,
+      "eval_steps_per_second": 23.486,
       "step": 7000
     },
     {
       "epoch": 1.11,
-      "learning_rate": 1.8088784400497857e-05,
-      "loss": 0.9782,
+      "learning_rate": 7.235513760199143e-06,
+      "loss": 1.1182,
       "step": 8000
     },
     {
       "epoch": 1.11,
-      "eval_accuracy": 0.6986722909895581,
-      "eval_loss": 1.0466938018798828,
-      "eval_runtime": 149.1228,
-      "eval_samples_per_second": 193.948,
-      "eval_steps_per_second": 24.248,
+      "eval_accuracy": 0.6895788672982505,
+      "eval_loss": 1.1419979333877563,
+      "eval_runtime": 152.1866,
+      "eval_samples_per_second": 190.043,
+      "eval_steps_per_second": 23.76,
       "step": 8000
     },
     {
       "epoch": 1.24,
-      "learning_rate": 1.7225314617618587e-05,
-      "loss": 0.9627,
+      "learning_rate": 6.889780113400637e-06,
+      "loss": 1.086,
       "step": 9000
     },
     {
       "epoch": 1.24,
-      "eval_accuracy": 0.7060023511513727,
-      "eval_loss": 1.0178191661834717,
-      "eval_runtime": 149.6361,
-      "eval_samples_per_second": 193.282,
-      "eval_steps_per_second": 24.165,
+      "eval_accuracy": 0.6893368370098887,
+      "eval_loss": 1.1294597387313843,
+      "eval_runtime": 153.4946,
+      "eval_samples_per_second": 188.424,
+      "eval_steps_per_second": 23.558,
       "step": 9000
     },
     {
       "epoch": 1.38,
-      "learning_rate": 1.6360980500622323e-05,
-      "loss": 0.9411,
+      "learning_rate": 6.54404646660213e-06,
+      "loss": 1.0761,
       "step": 10000
     },
     {
       "epoch": 1.38,
-      "eval_accuracy": 0.7056911693520503,
-      "eval_loss": 1.0125459432601929,
-      "eval_runtime": 148.9456,
-      "eval_samples_per_second": 194.178,
-      "eval_steps_per_second": 24.277,
+      "eval_accuracy": 0.6903741096742964,
+      "eval_loss": 1.1167794466018677,
+      "eval_runtime": 152.2341,
+      "eval_samples_per_second": 189.984,
+      "eval_steps_per_second": 23.753,
       "step": 10000
     },
     {
       "epoch": 1.52,
-      "learning_rate": 1.5496646383626052e-05,
-      "loss": 0.9333,
+      "learning_rate": 6.198312819803624e-06,
+      "loss": 1.0525,
       "step": 11000
     },
     {
       "epoch": 1.52,
-      "eval_accuracy": 0.7067630177719383,
-      "eval_loss": 1.0095747709274292,
-      "eval_runtime": 149.6511,
-      "eval_samples_per_second": 193.263,
-      "eval_steps_per_second": 24.163,
+      "eval_accuracy": 0.6951801396860522,
+      "eval_loss": 1.0984091758728027,
+      "eval_runtime": 151.8562,
+      "eval_samples_per_second": 190.456,
+      "eval_steps_per_second": 23.812,
       "step": 11000
     },
     {
       "epoch": 1.66,
-      "learning_rate": 1.4633176600746785e-05,
-      "loss": 0.9333,
+      "learning_rate": 5.852924906651915e-06,
+      "loss": 1.047,
       "step": 12000
     },
     {
       "epoch": 1.66,
-      "eval_accuracy": 0.7047576239540834,
-      "eval_loss": 1.0087416172027588,
-      "eval_runtime": 149.4203,
-      "eval_samples_per_second": 193.561,
-      "eval_steps_per_second": 24.2,
+      "eval_accuracy": 0.6963211396169007,
+      "eval_loss": 1.0861539840698242,
+      "eval_runtime": 154.131,
+      "eval_samples_per_second": 187.646,
+      "eval_steps_per_second": 23.461,
       "step": 12000
     },
     {
       "epoch": 1.8,
-      "learning_rate": 1.3768842483750519e-05,
-      "loss": 0.9227,
+      "learning_rate": 5.5071912598534095e-06,
+      "loss": 1.044,
       "step": 13000
     },
     {
       "epoch": 1.8,
-      "eval_accuracy": 0.7118456538275362,
-      "eval_loss": 0.9883840084075928,
-      "eval_runtime": 149.2315,
-      "eval_samples_per_second": 193.806,
-      "eval_steps_per_second": 24.231,
+      "eval_accuracy": 0.6946960791093285,
+      "eval_loss": 1.084086537361145,
+      "eval_runtime": 153.6622,
+      "eval_samples_per_second": 188.218,
+      "eval_steps_per_second": 23.532,
       "step": 13000
     },
     {
       "epoch": 1.94,
-      "learning_rate": 1.290537270087125e-05,
-      "loss": 0.9131,
+      "learning_rate": 5.161803346701701e-06,
+      "loss": 1.0231,
       "step": 14000
     },
     {
       "epoch": 1.94,
-      "eval_accuracy": 0.7127100477145426,
-      "eval_loss": 0.986507773399353,
-      "eval_runtime": 149.078,
-      "eval_samples_per_second": 194.006,
-      "eval_steps_per_second": 24.256,
+      "eval_accuracy": 0.6982228061683148,
+      "eval_loss": 1.0656354427337646,
+      "eval_runtime": 154.8778,
+      "eval_samples_per_second": 186.741,
+      "eval_steps_per_second": 23.347,
       "step": 14000
     },
     {
       "epoch": 2.07,
-      "learning_rate": 1.2041038583874984e-05,
-      "loss": 0.8703,
+      "learning_rate": 4.8160696999031944e-06,
+      "loss": 1.0118,
       "step": 15000
     },
     {
       "epoch": 2.07,
-      "eval_accuracy": 0.7143696839775949,
-      "eval_loss": 0.9831659197807312,
-      "eval_runtime": 150.612,
-      "eval_samples_per_second": 192.03,
-      "eval_steps_per_second": 24.009,
+      "eval_accuracy": 0.698395684945716,
+      "eval_loss": 1.065485954284668,
+      "eval_runtime": 152.9666,
+      "eval_samples_per_second": 189.074,
+      "eval_steps_per_second": 23.639,
       "step": 15000
     },
     {
       "epoch": 2.21,
-      "learning_rate": 1.1177568800995714e-05,
-      "loss": 0.8064,
+      "learning_rate": 4.470681786751487e-06,
+      "loss": 0.9659,
       "step": 16000
     },
     {
       "epoch": 2.21,
-      "eval_accuracy": 0.7136435931125095,
-      "eval_loss": 0.9858880043029785,
-      "eval_runtime": 149.8659,
-      "eval_samples_per_second": 192.986,
-      "eval_steps_per_second": 24.128,
+      "eval_accuracy": 0.7020261392711431,
+      "eval_loss": 1.0567070245742798,
+      "eval_runtime": 154.2161,
+      "eval_samples_per_second": 187.542,
+      "eval_steps_per_second": 23.448,
       "step": 16000
     },
     {
       "epoch": 2.35,
-      "learning_rate": 1.0313234683999447e-05,
-      "loss": 0.8131,
+      "learning_rate": 4.124948139952981e-06,
+      "loss": 0.9602,
       "step": 17000
     },
     {
       "epoch": 2.35,
-      "eval_accuracy": 0.7127100477145426,
-      "eval_loss": 0.9879944920539856,
-      "eval_runtime": 150.5293,
-      "eval_samples_per_second": 192.135,
-      "eval_steps_per_second": 24.022,
+      "eval_accuracy": 0.7003665030080908,
+      "eval_loss": 1.0593522787094116,
+      "eval_runtime": 154.104,
+      "eval_samples_per_second": 187.679,
+      "eval_steps_per_second": 23.465,
       "step": 17000
     },
     {
       "epoch": 2.49,
-      "learning_rate": 9.448900567003181e-06,
-      "loss": 0.8064,
+      "learning_rate": 3.779214493154474e-06,
+      "loss": 0.9757,
       "step": 18000
     },
     {
       "epoch": 2.49,
-      "eval_accuracy": 0.7113270174953322,
-      "eval_loss": 0.9872538447380066,
-      "eval_runtime": 149.9196,
-      "eval_samples_per_second": 192.917,
-      "eval_steps_per_second": 24.12,
+      "eval_accuracy": 0.7006431090519328,
+      "eval_loss": 1.053895115852356,
+      "eval_runtime": 153.596,
+      "eval_samples_per_second": 188.299,
+      "eval_steps_per_second": 23.542,
       "step": 18000
     },
     {
       "epoch": 2.63,
-      "learning_rate": 8.584566450006915e-06,
-      "loss": 0.812,
+      "learning_rate": 3.433826580002766e-06,
+      "loss": 0.9631,
       "step": 19000
     },
     {
       "epoch": 2.63,
-      "eval_accuracy": 0.7153723808865223,
-      "eval_loss": 0.9793804883956909,
-      "eval_runtime": 151.3957,
-      "eval_samples_per_second": 191.036,
-      "eval_steps_per_second": 23.884,
+      "eval_accuracy": 0.7040661088444783,
+      "eval_loss": 1.0456914901733398,
+      "eval_runtime": 152.8682,
+      "eval_samples_per_second": 189.196,
+      "eval_steps_per_second": 23.654,
       "step": 19000
     },
     {
       "epoch": 2.77,
-      "learning_rate": 7.721096667127644e-06,
-      "loss": 0.7977,
+      "learning_rate": 3.0880929332042597e-06,
+      "loss": 0.951,
       "step": 20000
     },
     {
       "epoch": 2.77,
-      "eval_accuracy": 0.7151649263536408,
-      "eval_loss": 0.9755117297172546,
-      "eval_runtime": 144.4577,
-      "eval_samples_per_second": 200.211,
-      "eval_steps_per_second": 25.032,
+      "eval_accuracy": 0.7039278058225572,
+      "eval_loss": 1.0407930612564087,
+      "eval_runtime": 152.6055,
+      "eval_samples_per_second": 189.521,
+      "eval_steps_per_second": 23.695,
       "step": 20000
     },
     {
       "epoch": 2.9,
-      "learning_rate": 6.856762550131379e-06,
-      "loss": 0.8118,
+      "learning_rate": 2.742359286405753e-06,
+      "loss": 0.942,
       "step": 21000
     },
     {
       "epoch": 2.9,
-      "eval_accuracy": 0.7164788050618907,
-      "eval_loss": 0.9746271967887878,
-      "eval_runtime": 155.3329,
-      "eval_samples_per_second": 186.194,
-      "eval_steps_per_second": 23.279,
+      "eval_accuracy": 0.7021298665375838,
+      "eval_loss": 1.042394757270813,
+      "eval_runtime": 154.2947,
+      "eval_samples_per_second": 187.446,
+      "eval_steps_per_second": 23.436,
       "step": 21000
     },
     {
       "epoch": 3.04,
-      "learning_rate": 5.992428433135113e-06,
-      "loss": 0.7741,
+      "learning_rate": 2.396625639607247e-06,
+      "loss": 0.9415,
       "step": 22000
     },
     {
       "epoch": 3.04,
-      "eval_accuracy": 0.7189336837009889,
-      "eval_loss": 0.9803725481033325,
-      "eval_runtime": 153.5991,
-      "eval_samples_per_second": 188.295,
-      "eval_steps_per_second": 23.542,
+      "eval_accuracy": 0.7035128967567942,
+      "eval_loss": 1.0409824848175049,
+      "eval_runtime": 151.55,
+      "eval_samples_per_second": 190.841,
+      "eval_steps_per_second": 23.86,
       "step": 22000
     },
     {
       "epoch": 3.18,
-      "learning_rate": 5.128958650255843e-06,
-      "loss": 0.7202,
+      "learning_rate": 2.0512377264555388e-06,
+      "loss": 0.9128,
       "step": 23000
     },
     {
       "epoch": 3.18,
-      "eval_accuracy": 0.71371274462347,
-      "eval_loss": 0.9932627081871033,
-      "eval_runtime": 153.8948,
-      "eval_samples_per_second": 187.934,
-      "eval_steps_per_second": 23.497,
+      "eval_accuracy": 0.7044118663992808,
+      "eval_loss": 1.0388602018356323,
+      "eval_runtime": 153.9577,
+      "eval_samples_per_second": 187.857,
+      "eval_steps_per_second": 23.487,
       "step": 23000
     },
     {
       "epoch": 3.32,
-      "learning_rate": 4.264624533259577e-06,
-      "loss": 0.7186,
+      "learning_rate": 1.7055040796570323e-06,
+      "loss": 0.9071,
       "step": 24000
     },
     {
       "epoch": 3.32,
-      "eval_accuracy": 0.7173086231934168,
-      "eval_loss": 0.9890537261962891,
-      "eval_runtime": 154.048,
-      "eval_samples_per_second": 187.747,
-      "eval_steps_per_second": 23.473,
+      "eval_accuracy": 0.7033745937348731,
+      "eval_loss": 1.0366332530975342,
+      "eval_runtime": 155.1151,
+      "eval_samples_per_second": 186.455,
+      "eval_steps_per_second": 23.312,
       "step": 24000
     },
     {
       "epoch": 3.46,
-      "learning_rate": 3.4011547503803073e-06,
-      "loss": 0.7164,
+      "learning_rate": 1.3601161665053245e-06,
+      "loss": 0.91,
       "step": 25000
     },
     {
       "epoch": 3.46,
-      "eval_accuracy": 0.7176198049927391,
-      "eval_loss": 0.9812184572219849,
-      "eval_runtime": 153.9568,
-      "eval_samples_per_second": 187.858,
-      "eval_steps_per_second": 23.487,
+      "eval_accuracy": 0.7045847451766821,
+      "eval_loss": 1.036346435546875,
+      "eval_runtime": 152.4669,
+      "eval_samples_per_second": 189.694,
+      "eval_steps_per_second": 23.717,
       "step": 25000
     },
     {
       "epoch": 3.6,
-      "learning_rate": 2.536820633384041e-06,
-      "loss": 0.7047,
+      "learning_rate": 1.0147282533536165e-06,
+      "loss": 0.8962,
       "step": 26000
     },
     {
       "epoch": 3.6,
-      "eval_accuracy": 0.7157527141968052,
-      "eval_loss": 0.9885823130607605,
-      "eval_runtime": 155.4293,
-      "eval_samples_per_second": 186.078,
-      "eval_steps_per_second": 23.265,
+      "eval_accuracy": 0.7042044118663993,
+      "eval_loss": 1.0368608236312866,
+      "eval_runtime": 151.6795,
+      "eval_samples_per_second": 190.678,
+      "eval_steps_per_second": 23.84,
       "step": 26000
     },
     {
       "epoch": 3.73,
-      "learning_rate": 1.6724865163877752e-06,
-      "loss": 0.7241,
+      "learning_rate": 6.6899460655511e-07,
+      "loss": 0.8979,
       "step": 27000
     },
     {
       "epoch": 3.73,
-      "eval_accuracy": 0.7153032293755619,
-      "eval_loss": 0.9851787686347961,
-      "eval_runtime": 155.7881,
-      "eval_samples_per_second": 185.65,
-      "eval_steps_per_second": 23.211,
+      "eval_accuracy": 0.7055528663301293,
+      "eval_loss": 1.0334995985031128,
+      "eval_runtime": 151.884,
+      "eval_samples_per_second": 190.422,
+      "eval_steps_per_second": 23.808,
       "step": 27000
     },
     {
       "epoch": 3.87,
-      "learning_rate": 8.081523993915088e-07,
-      "loss": 0.699,
+      "learning_rate": 3.2326095975660353e-07,
+      "loss": 0.906,
       "step": 28000
     },
     {
       "epoch": 3.87,
-      "eval_accuracy": 0.7164788050618907,
-      "eval_loss": 0.984923243522644,
-      "eval_runtime": 156.4441,
-      "eval_samples_per_second": 184.871,
-      "eval_steps_per_second": 23.114,
+      "eval_accuracy": 0.7039278058225572,
+      "eval_loss": 1.03609037399292,
+      "eval_runtime": 152.3393,
+      "eval_samples_per_second": 189.853,
+      "eval_steps_per_second": 23.736,
       "step": 28000
     },
     {
       "epoch": 4.0,
       "step": 28924,
       "total_flos": 7.174773947150918e+19,
-      "train_loss": 0.9355616505553886,
-      "train_runtime": 13547.1303,
-      "train_samples_per_second": 68.318,
-      "train_steps_per_second": 2.135
+      "train_loss": 1.1119355370668336,
+      "train_runtime": 13795.8304,
+      "train_samples_per_second": 67.086,
+      "train_steps_per_second": 2.097
     }
   ],
   "max_steps": 28924,