Training in progress, epoch 0

all_results.json

@@ -1,8 +1,13 @@
 {
     "epoch": 28.571428571428573,
+    "eval_accuracy": 0.8873239436619719,
+    "eval_loss": 0.5122641324996948,
+    "eval_runtime": 0.6466,
+    "eval_samples_per_second": 219.608,
+    "eval_steps_per_second": 4.64,
     "total_flos": 2.3842598606630093e+17,
-    "train_loss": 0.4332954502105713,
-    "train_runtime": 131.7536,
-    "train_samples_per_second": 161.665,
-    "train_steps_per_second": 0.379
+    "train_loss": 0.43769028663635257,
+    "train_runtime": 134.3919,
+    "train_samples_per_second": 158.492,
+    "train_steps_per_second": 0.372
 }

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 28.571428571428573,
+    "eval_accuracy": 0.8873239436619719,
+    "eval_loss": 0.5122641324996948,
+    "eval_runtime": 0.6466,
+    "eval_samples_per_second": 219.608,
+    "eval_steps_per_second": 4.64
+}

runs/Jan27_17-12-41_db1093ce036b/events.out.tfevents.1737998264.db1093ce036b.224.6

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c986a248b1092422da5b4d1e3d672d571bf03464b7b4d2a807d8924b4097dac
+size 5786

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 28.571428571428573,
     "total_flos": 2.3842598606630093e+17,
-    "train_loss": 0.4332954502105713,
-    "train_runtime": 131.7536,
-    "train_samples_per_second": 161.665,
-    "train_steps_per_second": 0.379
+    "train_loss": 0.43769028663635257,
+    "train_runtime": 134.3919,
+    "train_samples_per_second": 158.492,
+    "train_steps_per_second": 0.372
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.5166566371917725,
-  "best_model_checkpoint": "vit-msn-small-beta-fia-manually-enhanced-HSV_test_3/checkpoint-24",
+  "best_metric": 0.5122641324996948,
+  "best_model_checkpoint": "vit-msn-small-beta-fia-manually-enhanced-HSV_test_3/checkpoint-15",
   "epoch": 28.571428571428573,
   "eval_steps": 500,
   "global_step": 50,
@@ -10,308 +10,308 @@
   "log_history": [
     {
       "epoch": 0.5714285714285714,
-      "eval_accuracy": 0.8873239436619719,
-      "eval_loss": 0.5373859405517578,
-      "eval_runtime": 0.608,
-      "eval_samples_per_second": 233.567,
-      "eval_steps_per_second": 4.935,
+      "eval_accuracy": 0.8802816901408451,
+      "eval_loss": 0.5166566371917725,
+      "eval_runtime": 0.6288,
+      "eval_samples_per_second": 225.827,
+      "eval_steps_per_second": 4.771,
       "step": 1
     },
     {
       "epoch": 1.7142857142857144,
-      "eval_accuracy": 0.8873239436619719,
-      "eval_loss": 0.5290086269378662,
-      "eval_runtime": 0.6285,
-      "eval_samples_per_second": 225.924,
-      "eval_steps_per_second": 4.773,
+      "eval_accuracy": 0.8802816901408451,
+      "eval_loss": 0.519730806350708,
+      "eval_runtime": 0.6203,
+      "eval_samples_per_second": 228.939,
+      "eval_steps_per_second": 4.837,
       "step": 3
     },
     {
       "epoch": 2.857142857142857,
-      "eval_accuracy": 0.8873239436619719,
-      "eval_loss": 0.5365429520606995,
-      "eval_runtime": 0.6154,
-      "eval_samples_per_second": 230.751,
-      "eval_steps_per_second": 4.875,
+      "eval_accuracy": 0.8802816901408451,
+      "eval_loss": 0.5266488194465637,
+      "eval_runtime": 0.6213,
+      "eval_samples_per_second": 228.546,
+      "eval_steps_per_second": 4.828,
       "step": 5
     },
     {
       "epoch": 4.0,
       "eval_accuracy": 0.8802816901408451,
-      "eval_loss": 0.5444374084472656,
-      "eval_runtime": 0.6664,
-      "eval_samples_per_second": 213.073,
-      "eval_steps_per_second": 4.502,
+      "eval_loss": 0.5390564203262329,
+      "eval_runtime": 0.6545,
+      "eval_samples_per_second": 216.969,
+      "eval_steps_per_second": 4.584,
       "step": 7
     },
     {
       "epoch": 4.571428571428571,
       "eval_accuracy": 0.8802816901408451,
-      "eval_loss": 0.5448340773582458,
-      "eval_runtime": 0.7088,
-      "eval_samples_per_second": 200.338,
-      "eval_steps_per_second": 4.232,
+      "eval_loss": 0.5425485968589783,
+      "eval_runtime": 0.6176,
+      "eval_samples_per_second": 229.935,
+      "eval_steps_per_second": 4.858,
       "step": 8
     },
     {
       "epoch": 5.714285714285714,
-      "grad_norm": 3.9558749198913574,
+      "grad_norm": 5.448883056640625,
       "learning_rate": 5.882352941176471e-06,
-      "loss": 0.4449,
+      "loss": 0.4435,
       "step": 10
     },
     {
       "epoch": 5.714285714285714,
-      "eval_accuracy": 0.8591549295774648,
-      "eval_loss": 0.5697565078735352,
-      "eval_runtime": 0.6385,
-      "eval_samples_per_second": 222.381,
-      "eval_steps_per_second": 4.698,
+      "eval_accuracy": 0.8802816901408451,
+      "eval_loss": 0.5403485298156738,
+      "eval_runtime": 0.6202,
+      "eval_samples_per_second": 228.973,
+      "eval_steps_per_second": 4.837,
       "step": 10
     },
     {
       "epoch": 6.857142857142857,
-      "eval_accuracy": 0.8661971830985915,
-      "eval_loss": 0.5764577984809875,
-      "eval_runtime": 0.6144,
-      "eval_samples_per_second": 231.129,
-      "eval_steps_per_second": 4.883,
+      "eval_accuracy": 0.8802816901408451,
+      "eval_loss": 0.5251158475875854,
+      "eval_runtime": 0.6735,
+      "eval_samples_per_second": 210.854,
+      "eval_steps_per_second": 4.455,
       "step": 12
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.8661971830985915,
-      "eval_loss": 0.5411155223846436,
-      "eval_runtime": 0.62,
-      "eval_samples_per_second": 229.044,
-      "eval_steps_per_second": 4.839,
+      "eval_accuracy": 0.8732394366197183,
+      "eval_loss": 0.5160439610481262,
+      "eval_runtime": 0.6272,
+      "eval_samples_per_second": 226.418,
+      "eval_steps_per_second": 4.783,
       "step": 14
     },
     {
       "epoch": 8.571428571428571,
-      "eval_accuracy": 0.8802816901408451,
-      "eval_loss": 0.5282784104347229,
-      "eval_runtime": 0.6438,
-      "eval_samples_per_second": 220.561,
-      "eval_steps_per_second": 4.66,
+      "eval_accuracy": 0.8873239436619719,
+      "eval_loss": 0.5122641324996948,
+      "eval_runtime": 0.6632,
+      "eval_samples_per_second": 214.109,
+      "eval_steps_per_second": 4.523,
       "step": 15
     },
     {
       "epoch": 9.714285714285714,
       "eval_accuracy": 0.8802816901408451,
-      "eval_loss": 0.5272403955459595,
-      "eval_runtime": 0.7039,
-      "eval_samples_per_second": 201.741,
-      "eval_steps_per_second": 4.262,
+      "eval_loss": 0.5291638374328613,
+      "eval_runtime": 0.7041,
+      "eval_samples_per_second": 201.678,
+      "eval_steps_per_second": 4.261,
       "step": 17
     },
     {
       "epoch": 10.857142857142858,
       "eval_accuracy": 0.8732394366197183,
-      "eval_loss": 0.5336695313453674,
-      "eval_runtime": 0.6047,
-      "eval_samples_per_second": 234.826,
-      "eval_steps_per_second": 4.961,
+      "eval_loss": 0.568649172782898,
+      "eval_runtime": 0.6802,
+      "eval_samples_per_second": 208.773,
+      "eval_steps_per_second": 4.411,
       "step": 19
     },
     {
       "epoch": 11.428571428571429,
-      "grad_norm": 5.938806533813477,
+      "grad_norm": 5.323070049285889,
       "learning_rate": 9.090909090909091e-06,
-      "loss": 0.4343,
+      "loss": 0.4418,
       "step": 20
     },
     {
       "epoch": 12.0,
       "eval_accuracy": 0.8732394366197183,
-      "eval_loss": 0.5399956703186035,
-      "eval_runtime": 0.7179,
-      "eval_samples_per_second": 197.787,
-      "eval_steps_per_second": 4.179,
+      "eval_loss": 0.5459948182106018,
+      "eval_runtime": 0.6906,
+      "eval_samples_per_second": 205.619,
+      "eval_steps_per_second": 4.344,
       "step": 21
     },
     {
       "epoch": 12.571428571428571,
-      "eval_accuracy": 0.8802816901408451,
-      "eval_loss": 0.5316654443740845,
-      "eval_runtime": 0.6548,
-      "eval_samples_per_second": 216.861,
-      "eval_steps_per_second": 4.582,
+      "eval_accuracy": 0.8873239436619719,
+      "eval_loss": 0.5333032608032227,
+      "eval_runtime": 0.6327,
+      "eval_samples_per_second": 224.427,
+      "eval_steps_per_second": 4.741,
       "step": 22
     },
     {
       "epoch": 13.714285714285714,
       "eval_accuracy": 0.8802816901408451,
-      "eval_loss": 0.5166566371917725,
-      "eval_runtime": 0.6957,
-      "eval_samples_per_second": 204.108,
-      "eval_steps_per_second": 4.312,
+      "eval_loss": 0.5152425169944763,
+      "eval_runtime": 0.6626,
+      "eval_samples_per_second": 214.318,
+      "eval_steps_per_second": 4.528,
       "step": 24
     },
     {
       "epoch": 14.857142857142858,
       "eval_accuracy": 0.8732394366197183,
-      "eval_loss": 0.5271970629692078,
-      "eval_runtime": 0.6146,
-      "eval_samples_per_second": 231.056,
-      "eval_steps_per_second": 4.881,
+      "eval_loss": 0.5236279964447021,
+      "eval_runtime": 0.6606,
+      "eval_samples_per_second": 214.94,
+      "eval_steps_per_second": 4.541,
       "step": 26
     },
     {
       "epoch": 16.0,
-      "eval_accuracy": 0.8802816901408451,
-      "eval_loss": 0.5361518859863281,
-      "eval_runtime": 0.679,
-      "eval_samples_per_second": 209.117,
-      "eval_steps_per_second": 4.418,
+      "eval_accuracy": 0.8591549295774648,
+      "eval_loss": 0.5372341275215149,
+      "eval_runtime": 0.6123,
+      "eval_samples_per_second": 231.928,
+      "eval_steps_per_second": 4.9,
       "step": 28
     },
     {
       "epoch": 16.571428571428573,
-      "eval_accuracy": 0.8732394366197183,
-      "eval_loss": 0.5353840589523315,
-      "eval_runtime": 0.6496,
-      "eval_samples_per_second": 218.579,
-      "eval_steps_per_second": 4.618,
+      "eval_accuracy": 0.8591549295774648,
+      "eval_loss": 0.5472158193588257,
+      "eval_runtime": 0.6625,
+      "eval_samples_per_second": 214.333,
+      "eval_steps_per_second": 4.528,
       "step": 29
     },
     {
       "epoch": 17.142857142857142,
-      "grad_norm": 4.4233927726745605,
+      "grad_norm": 5.04396390914917,
       "learning_rate": 6.060606060606061e-06,
-      "loss": 0.4389,
+      "loss": 0.4363,
       "step": 30
     },
     {
       "epoch": 17.714285714285715,
-      "eval_accuracy": 0.8802816901408451,
-      "eval_loss": 0.5263069868087769,
-      "eval_runtime": 0.6615,
-      "eval_samples_per_second": 214.668,
-      "eval_steps_per_second": 4.535,
+      "eval_accuracy": 0.8591549295774648,
+      "eval_loss": 0.5421658754348755,
+      "eval_runtime": 0.6149,
+      "eval_samples_per_second": 230.945,
+      "eval_steps_per_second": 4.879,
       "step": 31
     },
     {
       "epoch": 18.857142857142858,
-      "eval_accuracy": 0.8732394366197183,
-      "eval_loss": 0.5342370271682739,
-      "eval_runtime": 0.6468,
-      "eval_samples_per_second": 219.55,
-      "eval_steps_per_second": 4.638,
+      "eval_accuracy": 0.8802816901408451,
+      "eval_loss": 0.5293453335762024,
+      "eval_runtime": 0.6909,
+      "eval_samples_per_second": 205.543,
+      "eval_steps_per_second": 4.342,
       "step": 33
     },
     {
       "epoch": 20.0,
       "eval_accuracy": 0.8802816901408451,
-      "eval_loss": 0.5446226596832275,
-      "eval_runtime": 0.666,
-      "eval_samples_per_second": 213.222,
-      "eval_steps_per_second": 4.505,
+      "eval_loss": 0.523467481136322,
+      "eval_runtime": 0.6321,
+      "eval_samples_per_second": 224.663,
+      "eval_steps_per_second": 4.746,
       "step": 35
     },
     {
       "epoch": 20.571428571428573,
       "eval_accuracy": 0.8802816901408451,
-      "eval_loss": 0.5494755506515503,
-      "eval_runtime": 0.6103,
-      "eval_samples_per_second": 232.659,
-      "eval_steps_per_second": 4.915,
+      "eval_loss": 0.523999810218811,
+      "eval_runtime": 0.6902,
+      "eval_samples_per_second": 205.737,
+      "eval_steps_per_second": 4.347,
       "step": 36
     },
     {
       "epoch": 21.714285714285715,
       "eval_accuracy": 0.8802816901408451,
-      "eval_loss": 0.5442414879798889,
-      "eval_runtime": 0.6404,
-      "eval_samples_per_second": 221.745,
-      "eval_steps_per_second": 4.685,
+      "eval_loss": 0.5302459001541138,
+      "eval_runtime": 0.6908,
+      "eval_samples_per_second": 205.559,
+      "eval_steps_per_second": 4.343,
       "step": 38
     },
     {
       "epoch": 22.857142857142858,
-      "grad_norm": 3.910562038421631,
+      "grad_norm": 5.993457794189453,
       "learning_rate": 3.0303030303030305e-06,
-      "loss": 0.423,
+      "loss": 0.4371,
       "step": 40
     },
     {
       "epoch": 22.857142857142858,
       "eval_accuracy": 0.8802816901408451,
-      "eval_loss": 0.5322338938713074,
-      "eval_runtime": 0.6332,
-      "eval_samples_per_second": 224.25,
-      "eval_steps_per_second": 4.738,
+      "eval_loss": 0.5324126482009888,
+      "eval_runtime": 0.7047,
+      "eval_samples_per_second": 201.5,
+      "eval_steps_per_second": 4.257,
       "step": 40
     },
     {
       "epoch": 24.0,
       "eval_accuracy": 0.8802816901408451,
-      "eval_loss": 0.5273756384849548,
-      "eval_runtime": 0.6297,
-      "eval_samples_per_second": 225.51,
-      "eval_steps_per_second": 4.764,
+      "eval_loss": 0.5349227786064148,
+      "eval_runtime": 0.6311,
+      "eval_samples_per_second": 225.0,
+      "eval_steps_per_second": 4.754,
       "step": 42
     },
     {
       "epoch": 24.571428571428573,
-      "eval_accuracy": 0.8802816901408451,
-      "eval_loss": 0.5295581817626953,
-      "eval_runtime": 0.6259,
-      "eval_samples_per_second": 226.871,
-      "eval_steps_per_second": 4.793,
+      "eval_accuracy": 0.8732394366197183,
+      "eval_loss": 0.5362741351127625,
+      "eval_runtime": 0.6225,
+      "eval_samples_per_second": 228.126,
+      "eval_steps_per_second": 4.82,
       "step": 43
     },
     {
       "epoch": 25.714285714285715,
-      "eval_accuracy": 0.8802816901408451,
-      "eval_loss": 0.5362545847892761,
-      "eval_runtime": 0.7205,
-      "eval_samples_per_second": 197.076,
-      "eval_steps_per_second": 4.164,
+      "eval_accuracy": 0.8732394366197183,
+      "eval_loss": 0.5341742634773254,
+      "eval_runtime": 0.6561,
+      "eval_samples_per_second": 216.426,
+      "eval_steps_per_second": 4.572,
       "step": 45
     },
     {
       "epoch": 26.857142857142858,
       "eval_accuracy": 0.8732394366197183,
-      "eval_loss": 0.5421485900878906,
-      "eval_runtime": 0.6367,
-      "eval_samples_per_second": 223.028,
-      "eval_steps_per_second": 4.712,
+      "eval_loss": 0.5314902067184448,
+      "eval_runtime": 0.6626,
+      "eval_samples_per_second": 214.302,
+      "eval_steps_per_second": 4.528,
       "step": 47
     },
     {
       "epoch": 28.0,
       "eval_accuracy": 0.8732394366197183,
-      "eval_loss": 0.5466616153717041,
-      "eval_runtime": 0.6369,
-      "eval_samples_per_second": 222.952,
-      "eval_steps_per_second": 4.71,
+      "eval_loss": 0.5318764448165894,
+      "eval_runtime": 0.6393,
+      "eval_samples_per_second": 222.13,
+      "eval_steps_per_second": 4.693,
       "step": 49
     },
     {
       "epoch": 28.571428571428573,
-      "grad_norm": 4.531747817993164,
+      "grad_norm": 4.919415473937988,
       "learning_rate": 0.0,
-      "loss": 0.4254,
+      "loss": 0.4298,
       "step": 50
     },
     {
       "epoch": 28.571428571428573,
-      "eval_accuracy": 0.8732394366197183,
-      "eval_loss": 0.5473288893699646,
-      "eval_runtime": 0.7049,
-      "eval_samples_per_second": 201.437,
-      "eval_steps_per_second": 4.256,
+      "eval_accuracy": 0.8661971830985915,
+      "eval_loss": 0.5321409106254578,
+      "eval_runtime": 0.6159,
+      "eval_samples_per_second": 230.575,
+      "eval_steps_per_second": 4.871,
       "step": 50
     },
     {
       "epoch": 28.571428571428573,
       "step": 50,
       "total_flos": 2.3842598606630093e+17,
-      "train_loss": 0.4332954502105713,
-      "train_runtime": 131.7536,
-      "train_samples_per_second": 161.665,
-      "train_steps_per_second": 0.379
+      "train_loss": 0.43769028663635257,
+      "train_runtime": 134.3919,
+      "train_samples_per_second": 158.492,
+      "train_steps_per_second": 0.372
     }
   ],
   "logging_steps": 10,