End of training

all_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 2.0,
-    "train_loss": 2.2499421772203947,
-    "train_runtime": 531.0436,
+    "epoch": 1.0,
+    "train_loss": 2.5612600226151314,
+    "train_runtime": 437.2347,
     "train_samples": 303,
-    "train_samples_per_second": 1.141,
-    "train_steps_per_second": 0.072
+    "train_samples_per_second": 0.693,
+    "train_steps_per_second": 0.043
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 2.0,
-    "train_loss": 2.2499421772203947,
-    "train_runtime": 531.0436,
+    "epoch": 1.0,
+    "train_loss": 2.5612600226151314,
+    "train_runtime": 437.2347,
     "train_samples": 303,
-    "train_samples_per_second": 1.141,
-    "train_steps_per_second": 0.072
+    "train_samples_per_second": 0.693,
+    "train_steps_per_second": 0.043
 }

trainer_state.json

@@ -1,8 +1,8 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.0,
-  "global_step": 38,
+  "epoch": 1.0,
+  "global_step": 19,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -17,9 +17,9 @@
       "epoch": 0.05,
       "eval_accuracy": 0.062219502243982046,
       "eval_loss": 2.654296875,
-      "eval_runtime": 3.4086,
-      "eval_samples_per_second": 16.722,
-      "eval_steps_per_second": 1.173,
+      "eval_runtime": 2.3869,
+      "eval_samples_per_second": 23.881,
+      "eval_steps_per_second": 1.676,
       "step": 1
     },
     {
@@ -32,14 +32,14 @@
       "epoch": 0.11,
       "eval_accuracy": 0.062219502243982046,
       "eval_loss": 2.654296875,
-      "eval_runtime": 2.4357,
-      "eval_samples_per_second": 23.402,
-      "eval_steps_per_second": 1.642,
+      "eval_runtime": 3.2001,
+      "eval_samples_per_second": 17.812,
+      "eval_steps_per_second": 1.25,
       "step": 2
     },
     {
       "epoch": 0.16,
-      "learning_rate": 2.994876739510005e-05,
+      "learning_rate": 2.9795419551040836e-05,
       "loss": 2.6003,
       "step": 3
     },
@@ -47,549 +47,264 @@
       "epoch": 0.16,
       "eval_accuracy": 0.06265016546534294,
       "eval_loss": 2.6015625,
-      "eval_runtime": 2.3986,
-      "eval_samples_per_second": 23.763,
-      "eval_steps_per_second": 1.668,
+      "eval_runtime": 3.1764,
+      "eval_samples_per_second": 17.945,
+      "eval_steps_per_second": 1.259,
       "step": 3
     },
     {
       "epoch": 0.21,
-      "learning_rate": 2.9795419551040836e-05,
+      "learning_rate": 2.9187258625509518e-05,
       "loss": 2.5603,
       "step": 4
     },
     {
       "epoch": 0.21,
-      "eval_accuracy": 0.0627181649213473,
+      "eval_accuracy": 0.06260483249467337,
       "eval_loss": 2.5703125,
-      "eval_runtime": 2.5872,
-      "eval_samples_per_second": 22.032,
-      "eval_steps_per_second": 1.546,
+      "eval_runtime": 2.9857,
+      "eval_samples_per_second": 19.091,
+      "eval_steps_per_second": 1.34,
       "step": 4
     },
     {
       "epoch": 0.26,
-      "learning_rate": 2.9541003989089956e-05,
-      "loss": 2.6067,
+      "learning_rate": 2.8192106268097336e-05,
+      "loss": 2.606,
       "step": 5
     },
     {
       "epoch": 0.26,
-      "eval_accuracy": 0.06289949680402557,
+      "eval_accuracy": 0.06292216328936036,
       "eval_loss": 2.55078125,
-      "eval_runtime": 2.7942,
-      "eval_samples_per_second": 20.4,
-      "eval_steps_per_second": 1.432,
+      "eval_runtime": 2.5694,
+      "eval_samples_per_second": 22.184,
+      "eval_steps_per_second": 1.557,
       "step": 5
     },
     {
       "epoch": 0.32,
-      "learning_rate": 2.9187258625509518e-05,
+      "learning_rate": 2.6837107640945904e-05,
       "loss": 2.5439,
       "step": 6
     },
     {
       "epoch": 0.32,
-      "eval_accuracy": 0.0626954984360125,
-      "eval_loss": 2.546875,
-      "eval_runtime": 2.8026,
-      "eval_samples_per_second": 20.338,
-      "eval_steps_per_second": 1.427,
+      "eval_accuracy": 0.06292216328936036,
+      "eval_loss": 2.544921875,
+      "eval_runtime": 3.1779,
+      "eval_samples_per_second": 17.937,
+      "eval_steps_per_second": 1.259,
       "step": 6
     },
     {
       "epoch": 0.37,
-      "learning_rate": 2.873659989982586e-05,
-      "loss": 2.4459,
+      "learning_rate": 2.5159223574386117e-05,
+      "loss": 2.4449,
       "step": 7
     },
     {
       "epoch": 0.37,
-      "eval_accuracy": 0.06289949680402557,
-      "eval_loss": 2.548828125,
-      "eval_runtime": 2.3868,
-      "eval_samples_per_second": 23.881,
-      "eval_steps_per_second": 1.676,
+      "eval_accuracy": 0.06287683031869079,
+      "eval_loss": 2.546875,
+      "eval_runtime": 2.3628,
+      "eval_samples_per_second": 24.124,
+      "eval_steps_per_second": 1.693,
       "step": 7
     },
     {
       "epoch": 0.42,
-      "learning_rate": 2.8192106268097336e-05,
-      "loss": 2.5439,
+      "learning_rate": 2.320422237183641e-05,
+      "loss": 2.5422,
       "step": 8
     },
     {
       "epoch": 0.42,
-      "eval_accuracy": 0.06278616437735164,
-      "eval_loss": 2.548828125,
-      "eval_runtime": 3.0044,
-      "eval_samples_per_second": 18.972,
-      "eval_steps_per_second": 1.331,
+      "eval_accuracy": 0.0629901627453647,
+      "eval_loss": 2.546875,
+      "eval_runtime": 3.1712,
+      "eval_samples_per_second": 17.974,
+      "eval_steps_per_second": 1.261,
       "step": 8
     },
     {
       "epoch": 0.47,
-      "learning_rate": 2.7557497173937928e-05,
-      "loss": 2.6125,
+      "learning_rate": 2.1025431369794546e-05,
+      "loss": 2.6101,
       "step": 9
     },
     {
       "epoch": 0.47,
-      "eval_accuracy": 0.06317149462804297,
-      "eval_loss": 2.54296875,
-      "eval_runtime": 2.9895,
-      "eval_samples_per_second": 19.067,
-      "eval_steps_per_second": 1.338,
+      "eval_accuracy": 0.06319416111337776,
+      "eval_loss": 2.541015625,
+      "eval_runtime": 2.3922,
+      "eval_samples_per_second": 23.827,
+      "eval_steps_per_second": 1.672,
       "step": 9
     },
     {
       "epoch": 0.53,
-      "learning_rate": 2.6837107640945904e-05,
-      "loss": 2.4583,
+      "learning_rate": 1.8682282307111988e-05,
+      "loss": 2.4482,
       "step": 10
     },
     {
       "epoch": 0.53,
-      "eval_accuracy": 0.06262749898000816,
-      "eval_loss": 2.5390625,
-      "eval_runtime": 3.2063,
-      "eval_samples_per_second": 17.778,
-      "eval_steps_per_second": 1.248,
+      "eval_accuracy": 0.0629901627453647,
+      "eval_loss": 2.53515625,
+      "eval_runtime": 2.3705,
+      "eval_samples_per_second": 24.045,
+      "eval_steps_per_second": 1.687,
       "step": 10
     },
     {
       "epoch": 0.58,
-      "learning_rate": 2.6035858660096975e-05,
-      "loss": 2.5088,
+      "learning_rate": 1.623869018208499e-05,
+      "loss": 2.501,
       "step": 11
     },
     {
       "epoch": 0.58,
-      "eval_accuracy": 0.06287683031869079,
-      "eval_loss": 2.533203125,
-      "eval_runtime": 2.5988,
-      "eval_samples_per_second": 21.933,
-      "eval_steps_per_second": 1.539,
+      "eval_accuracy": 0.0631261616573734,
+      "eval_loss": 2.529296875,
+      "eval_runtime": 2.7727,
+      "eval_samples_per_second": 20.558,
+      "eval_steps_per_second": 1.443,
       "step": 11
     },
     {
       "epoch": 0.63,
-      "learning_rate": 2.5159223574386117e-05,
-      "loss": 2.6033,
+      "learning_rate": 1.3761309817915017e-05,
+      "loss": 2.5967,
       "step": 12
     },
     {
       "epoch": 0.63,
-      "eval_accuracy": 0.06308082868670384,
+      "eval_accuracy": 0.06337549299605603,
       "eval_loss": 2.521484375,
-      "eval_runtime": 2.595,
-      "eval_samples_per_second": 21.965,
-      "eval_steps_per_second": 1.541,
+      "eval_runtime": 2.168,
+      "eval_samples_per_second": 26.292,
+      "eval_steps_per_second": 1.845,
       "step": 12
     },
     {
       "epoch": 0.68,
-      "learning_rate": 2.4213190690345018e-05,
-      "loss": 2.5037,
+      "learning_rate": 1.1317717692888014e-05,
+      "loss": 2.4998,
       "step": 13
     },
     {
       "epoch": 0.68,
-      "eval_accuracy": 0.0632848270547169,
-      "eval_loss": 2.515625,
-      "eval_runtime": 3.1996,
-      "eval_samples_per_second": 17.815,
-      "eval_steps_per_second": 1.25,
+      "eval_accuracy": 0.06346615893739517,
+      "eval_loss": 2.513671875,
+      "eval_runtime": 3.1858,
+      "eval_samples_per_second": 17.892,
+      "eval_steps_per_second": 1.256,
       "step": 13
     },
     {
       "epoch": 0.74,
-      "learning_rate": 2.320422237183641e-05,
-      "loss": 2.6033,
+      "learning_rate": 8.974568630205462e-06,
+      "loss": 2.5957,
       "step": 14
     },
     {
       "epoch": 0.74,
-      "eval_accuracy": 0.06339815948139081,
-      "eval_loss": 2.5078125,
-      "eval_runtime": 2.4053,
-      "eval_samples_per_second": 23.698,
-      "eval_steps_per_second": 1.663,
+      "eval_accuracy": 0.06364749082007344,
+      "eval_loss": 2.509765625,
+      "eval_runtime": 2.1678,
+      "eval_samples_per_second": 26.294,
+      "eval_steps_per_second": 1.845,
       "step": 14
     },
     {
       "epoch": 0.79,
-      "learning_rate": 2.2139210895556104e-05,
-      "loss": 2.6023,
+      "learning_rate": 6.795777628163599e-06,
+      "loss": 2.5967,
       "step": 15
     },
     {
       "epoch": 0.79,
-      "eval_accuracy": 0.06339815948139081,
-      "eval_loss": 2.501953125,
-      "eval_runtime": 2.804,
-      "eval_samples_per_second": 20.328,
-      "eval_steps_per_second": 1.427,
+      "eval_accuracy": 0.06387415567342128,
+      "eval_loss": 2.50390625,
+      "eval_runtime": 3.188,
+      "eval_samples_per_second": 17.879,
+      "eval_steps_per_second": 1.255,
       "step": 15
     },
     {
       "epoch": 0.84,
-      "learning_rate": 2.1025431369794546e-05,
-      "loss": 2.5034,
+      "learning_rate": 4.840776425613887e-06,
+      "loss": 2.5022,
       "step": 16
     },
     {
       "epoch": 0.84,
-      "eval_accuracy": 0.06333016002538647,
-      "eval_loss": 2.49609375,
-      "eval_runtime": 2.1988,
-      "eval_samples_per_second": 25.923,
-      "eval_steps_per_second": 1.819,
+      "eval_accuracy": 0.06373815676141258,
+      "eval_loss": 2.5,
+      "eval_runtime": 3.178,
+      "eval_samples_per_second": 17.936,
+      "eval_steps_per_second": 1.259,
       "step": 16
     },
     {
       "epoch": 0.89,
-      "learning_rate": 1.9870492038070255e-05,
-      "loss": 2.4353,
+      "learning_rate": 3.162892359054098e-06,
+      "loss": 2.4314,
       "step": 17
     },
     {
       "epoch": 0.89,
-      "eval_accuracy": 0.0631488281427082,
-      "eval_loss": 2.490234375,
-      "eval_runtime": 2.5999,
-      "eval_samples_per_second": 21.924,
-      "eval_steps_per_second": 1.538,
+      "eval_accuracy": 0.06371549027607779,
+      "eval_loss": 2.498046875,
+      "eval_runtime": 2.968,
+      "eval_samples_per_second": 19.205,
+      "eval_steps_per_second": 1.348,
       "step": 17
     },
     {
       "epoch": 0.95,
-      "learning_rate": 1.8682282307111988e-05,
-      "loss": 2.6262,
+      "learning_rate": 1.8078937319026655e-06,
+      "loss": 2.6279,
       "step": 18
     },
     {
       "epoch": 0.95,
-      "eval_accuracy": 0.06333016002538647,
-      "eval_loss": 2.486328125,
-      "eval_runtime": 2.9949,
-      "eval_samples_per_second": 19.032,
-      "eval_steps_per_second": 1.336,
+      "eval_accuracy": 0.06362482433473865,
+      "eval_loss": 2.49609375,
+      "eval_runtime": 3.1736,
+      "eval_samples_per_second": 17.961,
+      "eval_steps_per_second": 1.26,
       "step": 18
     },
     {
       "epoch": 1.0,
-      "learning_rate": 1.746891885421101e-05,
-      "loss": 2.5613,
+      "learning_rate": 8.127413744904805e-07,
+      "loss": 2.571,
       "step": 19
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.0631261616573734,
-      "eval_loss": 2.484375,
-      "eval_runtime": 3.2109,
-      "eval_samples_per_second": 17.752,
-      "eval_steps_per_second": 1.246,
-      "step": 19
-    },
-    {
-      "epoch": 1.05,
-      "learning_rate": 1.623869018208499e-05,
-      "loss": 2.115,
-      "step": 20
-    },
-    {
-      "epoch": 1.05,
-      "eval_accuracy": 0.06317149462804297,
-      "eval_loss": 2.48046875,
-      "eval_runtime": 3.1997,
-      "eval_samples_per_second": 17.814,
-      "eval_steps_per_second": 1.25,
-      "step": 20
-    },
-    {
-      "epoch": 1.11,
-      "learning_rate": 1.5e-05,
-      "loss": 2.0885,
-      "step": 21
-    },
-    {
-      "epoch": 1.11,
-      "eval_accuracy": 0.06351149190806474,
-      "eval_loss": 2.48046875,
-      "eval_runtime": 2.7957,
-      "eval_samples_per_second": 20.388,
-      "eval_steps_per_second": 1.431,
-      "step": 21
-    },
-    {
-      "epoch": 1.16,
-      "learning_rate": 1.5e-05,
-      "loss": 1.9777,
-      "step": 22
-    },
-    {
-      "epoch": 1.16,
-      "eval_accuracy": 0.06351149190806474,
-      "eval_loss": 2.48046875,
-      "eval_runtime": 2.6375,
-      "eval_samples_per_second": 21.612,
-      "eval_steps_per_second": 1.517,
-      "step": 22
-    },
-    {
-      "epoch": 1.21,
-      "learning_rate": 1.3761309817915017e-05,
-      "loss": 2.1053,
-      "step": 23
-    },
-    {
-      "epoch": 1.21,
-      "eval_accuracy": 0.06335282651072124,
-      "eval_loss": 2.48046875,
-      "eval_runtime": 2.3919,
-      "eval_samples_per_second": 23.831,
-      "eval_steps_per_second": 1.672,
-      "step": 23
-    },
-    {
-      "epoch": 1.26,
-      "learning_rate": 1.2531081145788989e-05,
-      "loss": 1.9324,
-      "step": 24
-    },
-    {
-      "epoch": 1.26,
-      "eval_accuracy": 0.06364749082007344,
-      "eval_loss": 2.482421875,
-      "eval_runtime": 2.7945,
-      "eval_samples_per_second": 20.397,
-      "eval_steps_per_second": 1.431,
-      "step": 24
-    },
-    {
-      "epoch": 1.32,
-      "learning_rate": 1.1317717692888014e-05,
-      "loss": 1.9122,
-      "step": 25
-    },
-    {
-      "epoch": 1.32,
-      "eval_accuracy": 0.06369282379074301,
-      "eval_loss": 2.48828125,
-      "eval_runtime": 2.7869,
-      "eval_samples_per_second": 20.453,
-      "eval_steps_per_second": 1.435,
-      "step": 25
-    },
-    {
-      "epoch": 1.37,
-      "learning_rate": 1.0129507961929749e-05,
-      "loss": 2.1058,
-      "step": 26
-    },
-    {
-      "epoch": 1.37,
-      "eval_accuracy": 0.06389682215875607,
-      "eval_loss": 2.49609375,
-      "eval_runtime": 2.4103,
-      "eval_samples_per_second": 23.648,
-      "eval_steps_per_second": 1.66,
-      "step": 26
-    },
-    {
-      "epoch": 1.42,
-      "learning_rate": 8.974568630205462e-06,
-      "loss": 1.9803,
-      "step": 27
-    },
-    {
-      "epoch": 1.42,
-      "eval_accuracy": 0.06389682215875607,
-      "eval_loss": 2.501953125,
-      "eval_runtime": 2.7926,
-      "eval_samples_per_second": 20.411,
-      "eval_steps_per_second": 1.432,
-      "step": 27
-    },
-    {
-      "epoch": 1.47,
-      "learning_rate": 7.860789104443897e-06,
-      "loss": 1.6486,
-      "step": 28
-    },
-    {
-      "epoch": 1.47,
-      "eval_accuracy": 0.06378348973208214,
-      "eval_loss": 2.509765625,
-      "eval_runtime": 3.2131,
-      "eval_samples_per_second": 17.74,
-      "eval_steps_per_second": 1.245,
-      "step": 28
-    },
-    {
-      "epoch": 1.53,
-      "learning_rate": 6.795777628163599e-06,
-      "loss": 1.8983,
-      "step": 29
-    },
-    {
-      "epoch": 1.53,
-      "eval_accuracy": 0.06357949136406908,
-      "eval_loss": 2.515625,
-      "eval_runtime": 3.1898,
-      "eval_samples_per_second": 17.869,
-      "eval_steps_per_second": 1.254,
-      "step": 29
-    },
-    {
-      "epoch": 1.58,
-      "learning_rate": 5.786809309654983e-06,
-      "loss": 1.8105,
-      "step": 30
-    },
-    {
-      "epoch": 1.58,
-      "eval_accuracy": 0.06335282651072124,
-      "eval_loss": 2.521484375,
-      "eval_runtime": 3.1961,
-      "eval_samples_per_second": 17.834,
-      "eval_steps_per_second": 1.252,
-      "step": 30
-    },
-    {
-      "epoch": 1.63,
-      "learning_rate": 4.840776425613887e-06,
-      "loss": 1.9916,
-      "step": 31
-    },
-    {
-      "epoch": 1.63,
-      "eval_accuracy": 0.06344349245206038,
-      "eval_loss": 2.5234375,
-      "eval_runtime": 2.7902,
-      "eval_samples_per_second": 20.428,
-      "eval_steps_per_second": 1.434,
-      "step": 31
-    },
-    {
-      "epoch": 1.68,
-      "learning_rate": 3.964141339903026e-06,
-      "loss": 1.886,
-      "step": 32
-    },
-    {
-      "epoch": 1.68,
-      "eval_accuracy": 0.06346615893739517,
-      "eval_loss": 2.525390625,
-      "eval_runtime": 2.6003,
-      "eval_samples_per_second": 21.921,
-      "eval_steps_per_second": 1.538,
-      "step": 32
-    },
-    {
-      "epoch": 1.74,
-      "learning_rate": 3.162892359054098e-06,
-      "loss": 1.8013,
-      "step": 33
-    },
-    {
-      "epoch": 1.74,
-      "eval_accuracy": 0.06344349245206038,
-      "eval_loss": 2.52734375,
-      "eval_runtime": 2.3881,
-      "eval_samples_per_second": 23.868,
-      "eval_steps_per_second": 1.675,
-      "step": 33
-    },
-    {
-      "epoch": 1.79,
-      "learning_rate": 2.442502826062072e-06,
-      "loss": 1.8435,
-      "step": 34
-    },
-    {
-      "epoch": 1.79,
-      "eval_accuracy": 0.06344349245206038,
-      "eval_loss": 2.525390625,
-      "eval_runtime": 3.2047,
-      "eval_samples_per_second": 17.786,
-      "eval_steps_per_second": 1.248,
-      "step": 34
-    },
-    {
-      "epoch": 1.84,
-      "learning_rate": 1.8078937319026655e-06,
-      "loss": 2.1229,
-      "step": 35
-    },
-    {
-      "epoch": 1.84,
-      "eval_accuracy": 0.06348882542272995,
-      "eval_loss": 2.5234375,
-      "eval_runtime": 2.4087,
-      "eval_samples_per_second": 23.664,
-      "eval_steps_per_second": 1.661,
-      "step": 35
-    },
-    {
-      "epoch": 1.89,
-      "learning_rate": 1.2634001001741375e-06,
-      "loss": 1.8739,
-      "step": 36
-    },
-    {
-      "epoch": 1.89,
       "eval_accuracy": 0.06357949136406908,
-      "eval_loss": 2.5234375,
-      "eval_runtime": 3.196,
-      "eval_samples_per_second": 17.835,
-      "eval_steps_per_second": 1.252,
-      "step": 36
-    },
-    {
-      "epoch": 1.95,
-      "learning_rate": 8.127413744904805e-07,
-      "loss": 1.7528,
-      "step": 37
-    },
-    {
-      "epoch": 1.95,
-      "eval_accuracy": 0.06369282379074301,
-      "eval_loss": 2.5234375,
-      "eval_runtime": 2.1908,
-      "eval_samples_per_second": 26.018,
-      "eval_steps_per_second": 1.826,
-      "step": 37
-    },
-    {
-      "epoch": 2.0,
-      "learning_rate": 4.589960109100444e-07,
-      "loss": 1.9462,
-      "step": 38
+      "eval_loss": 2.49609375,
+      "eval_runtime": 2.9674,
+      "eval_samples_per_second": 19.208,
+      "eval_steps_per_second": 1.348,
+      "step": 19
     },
     {
-      "epoch": 2.0,
-      "eval_accuracy": 0.06360215784940387,
-      "eval_loss": 2.521484375,
-      "eval_runtime": 2.6164,
-      "eval_samples_per_second": 21.786,
-      "eval_steps_per_second": 1.529,
-      "step": 38
-    },
-    {
-      "epoch": 2.0,
-      "step": 38,
-      "total_flos": 2058056761344.0,
-      "train_loss": 2.2499421772203947,
-      "train_runtime": 531.0436,
-      "train_samples_per_second": 1.141,
-      "train_steps_per_second": 0.072
+      "epoch": 1.0,
+      "step": 19,
+      "total_flos": 1001216802816.0,
+      "train_loss": 2.5612600226151314,
+      "train_runtime": 437.2347,
+      "train_samples_per_second": 0.693,
+      "train_steps_per_second": 0.043
     }
   ],
-  "max_steps": 38,
-  "num_train_epochs": 2,
-  "total_flos": 2058056761344.0,
+  "max_steps": 19,
+  "num_train_epochs": 1,
+  "total_flos": 1001216802816.0,
   "trial_name": null,
   "trial_params": null
 }