End of training

all_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.0,
-    "train_loss": 1.9830793414199561,
-    "train_runtime": 617.3135,
+    "epoch": 2.0,
+    "train_loss": 2.2499421772203947,
+    "train_runtime": 531.0436,
     "train_samples": 303,
-    "train_samples_per_second": 1.473,
-    "train_steps_per_second": 0.092
+    "train_samples_per_second": 1.141,
+    "train_steps_per_second": 0.072
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.0,
-    "train_loss": 1.9830793414199561,
-    "train_runtime": 617.3135,
+    "epoch": 2.0,
+    "train_loss": 2.2499421772203947,
+    "train_runtime": 531.0436,
     "train_samples": 303,
-    "train_samples_per_second": 1.473,
-    "train_steps_per_second": 0.092
+    "train_samples_per_second": 1.141,
+    "train_steps_per_second": 0.072
 }

trainer_state.json

@@ -1,8 +1,8 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 3.0,
-  "global_step": 57,
+  "epoch": 2.0,
+  "global_step": 38,
   "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": 3.1639,
-      "eval_samples_per_second": 18.016,
-      "eval_steps_per_second": 1.264,
+      "eval_runtime": 3.4086,
+      "eval_samples_per_second": 16.722,
+      "eval_steps_per_second": 1.173,
       "step": 1
     },
     {
@@ -32,14 +32,14 @@
       "epoch": 0.11,
       "eval_accuracy": 0.062219502243982046,
       "eval_loss": 2.654296875,
-      "eval_runtime": 3.1507,
-      "eval_samples_per_second": 18.091,
-      "eval_steps_per_second": 1.27,
+      "eval_runtime": 2.4357,
+      "eval_samples_per_second": 23.402,
+      "eval_steps_per_second": 1.642,
       "step": 2
     },
     {
       "epoch": 0.16,
-      "learning_rate": 2.997722274649974e-05,
+      "learning_rate": 2.994876739510005e-05,
       "loss": 2.6003,
       "step": 3
     },
@@ -47,834 +47,549 @@
       "epoch": 0.16,
       "eval_accuracy": 0.06265016546534294,
       "eval_loss": 2.6015625,
-      "eval_runtime": 3.124,
-      "eval_samples_per_second": 18.246,
-      "eval_steps_per_second": 1.28,
+      "eval_runtime": 2.3986,
+      "eval_samples_per_second": 23.763,
+      "eval_steps_per_second": 1.668,
       "step": 3
     },
     {
       "epoch": 0.21,
-      "learning_rate": 2.9908960159769243e-05,
+      "learning_rate": 2.9795419551040836e-05,
       "loss": 2.5603,
       "step": 4
     },
     {
       "epoch": 0.21,
-      "eval_accuracy": 0.06265016546534294,
+      "eval_accuracy": 0.0627181649213473,
       "eval_loss": 2.5703125,
-      "eval_runtime": 3.1396,
-      "eval_samples_per_second": 18.155,
-      "eval_steps_per_second": 1.274,
+      "eval_runtime": 2.5872,
+      "eval_samples_per_second": 22.032,
+      "eval_steps_per_second": 1.546,
       "step": 4
     },
     {
       "epoch": 0.26,
-      "learning_rate": 2.9795419551040836e-05,
-      "loss": 2.6072,
+      "learning_rate": 2.9541003989089956e-05,
+      "loss": 2.6067,
       "step": 5
     },
     {
       "epoch": 0.26,
-      "eval_accuracy": 0.06296749626002993,
+      "eval_accuracy": 0.06289949680402557,
       "eval_loss": 2.55078125,
-      "eval_runtime": 2.3207,
-      "eval_samples_per_second": 24.561,
-      "eval_steps_per_second": 1.724,
+      "eval_runtime": 2.7942,
+      "eval_samples_per_second": 20.4,
+      "eval_steps_per_second": 1.432,
       "step": 5
     },
     {
       "epoch": 0.32,
-      "learning_rate": 2.9636945739411533e-05,
-      "loss": 2.5444,
+      "learning_rate": 2.9187258625509518e-05,
+      "loss": 2.5439,
       "step": 6
     },
     {
       "epoch": 0.32,
-      "eval_accuracy": 0.06283149734802121,
+      "eval_accuracy": 0.0626954984360125,
       "eval_loss": 2.546875,
-      "eval_runtime": 3.1226,
-      "eval_samples_per_second": 18.254,
-      "eval_steps_per_second": 1.281,
+      "eval_runtime": 2.8026,
+      "eval_samples_per_second": 20.338,
+      "eval_steps_per_second": 1.427,
       "step": 6
     },
     {
       "epoch": 0.37,
-      "learning_rate": 2.9434020004638757e-05,
-      "loss": 2.4467,
+      "learning_rate": 2.873659989982586e-05,
+      "loss": 2.4459,
       "step": 7
     },
     {
       "epoch": 0.37,
-      "eval_accuracy": 0.06292216328936036,
-      "eval_loss": 2.55078125,
-      "eval_runtime": 2.1195,
-      "eval_samples_per_second": 26.894,
-      "eval_steps_per_second": 1.887,
+      "eval_accuracy": 0.06289949680402557,
+      "eval_loss": 2.548828125,
+      "eval_runtime": 2.3868,
+      "eval_samples_per_second": 23.881,
+      "eval_steps_per_second": 1.676,
       "step": 7
     },
     {
       "epoch": 0.42,
-      "learning_rate": 2.9187258625509518e-05,
-      "loss": 2.5452,
+      "learning_rate": 2.8192106268097336e-05,
+      "loss": 2.5439,
       "step": 8
     },
     {
       "epoch": 0.42,
-      "eval_accuracy": 0.06287683031869079,
-      "eval_loss": 2.55078125,
-      "eval_runtime": 2.717,
-      "eval_samples_per_second": 20.979,
-      "eval_steps_per_second": 1.472,
+      "eval_accuracy": 0.06278616437735164,
+      "eval_loss": 2.548828125,
+      "eval_runtime": 3.0044,
+      "eval_samples_per_second": 18.972,
+      "eval_steps_per_second": 1.331,
       "step": 8
     },
     {
       "epoch": 0.47,
-      "learning_rate": 2.8897411008222026e-05,
-      "loss": 2.6128,
+      "learning_rate": 2.7557497173937928e-05,
+      "loss": 2.6125,
       "step": 9
     },
     {
       "epoch": 0.47,
-      "eval_accuracy": 0.06308082868670384,
-      "eval_loss": 2.544921875,
-      "eval_runtime": 2.9144,
-      "eval_samples_per_second": 19.558,
-      "eval_steps_per_second": 1.372,
+      "eval_accuracy": 0.06317149462804297,
+      "eval_loss": 2.54296875,
+      "eval_runtime": 2.9895,
+      "eval_samples_per_second": 19.067,
+      "eval_steps_per_second": 1.338,
       "step": 9
     },
     {
       "epoch": 0.53,
-      "learning_rate": 2.8565357410463664e-05,
-      "loss": 2.4568,
+      "learning_rate": 2.6837107640945904e-05,
+      "loss": 2.4583,
       "step": 10
     },
     {
       "epoch": 0.53,
-      "eval_accuracy": 0.06265016546534294,
+      "eval_accuracy": 0.06262749898000816,
       "eval_loss": 2.5390625,
-      "eval_runtime": 3.1199,
-      "eval_samples_per_second": 18.27,
-      "eval_steps_per_second": 1.282,
+      "eval_runtime": 3.2063,
+      "eval_samples_per_second": 17.778,
+      "eval_steps_per_second": 1.248,
       "step": 10
     },
     {
       "epoch": 0.58,
-      "learning_rate": 2.8192106268097336e-05,
-      "loss": 2.5098,
+      "learning_rate": 2.6035858660096975e-05,
+      "loss": 2.5088,
       "step": 11
     },
     {
       "epoch": 0.58,
-      "eval_accuracy": 0.06278616437735164,
-      "eval_loss": 2.53515625,
-      "eval_runtime": 3.1268,
-      "eval_samples_per_second": 18.23,
-      "eval_steps_per_second": 1.279,
+      "eval_accuracy": 0.06287683031869079,
+      "eval_loss": 2.533203125,
+      "eval_runtime": 2.5988,
+      "eval_samples_per_second": 21.933,
+      "eval_steps_per_second": 1.539,
       "step": 11
     },
     {
       "epoch": 0.63,
-      "learning_rate": 2.7778791132574908e-05,
-      "loss": 2.6047,
+      "learning_rate": 2.5159223574386117e-05,
+      "loss": 2.6033,
       "step": 12
     },
     {
       "epoch": 0.63,
-      "eval_accuracy": 0.0631261616573734,
-      "eval_loss": 2.5234375,
-      "eval_runtime": 3.1183,
-      "eval_samples_per_second": 18.279,
-      "eval_steps_per_second": 1.283,
+      "eval_accuracy": 0.06308082868670384,
+      "eval_loss": 2.521484375,
+      "eval_runtime": 2.595,
+      "eval_samples_per_second": 21.965,
+      "eval_steps_per_second": 1.541,
       "step": 12
     },
     {
       "epoch": 0.68,
-      "learning_rate": 2.7326667228378677e-05,
-      "loss": 2.5022,
+      "learning_rate": 2.4213190690345018e-05,
+      "loss": 2.5037,
       "step": 13
     },
     {
       "epoch": 0.68,
-      "eval_accuracy": 0.0630128292306995,
+      "eval_accuracy": 0.0632848270547169,
       "eval_loss": 2.515625,
-      "eval_runtime": 2.5192,
-      "eval_samples_per_second": 22.626,
-      "eval_steps_per_second": 1.588,
+      "eval_runtime": 3.1996,
+      "eval_samples_per_second": 17.815,
+      "eval_steps_per_second": 1.25,
       "step": 13
     },
     {
       "epoch": 0.74,
-      "learning_rate": 2.6837107640945904e-05,
-      "loss": 2.605,
+      "learning_rate": 2.320422237183641e-05,
+      "loss": 2.6033,
       "step": 14
     },
     {
       "epoch": 0.74,
-      "eval_accuracy": 0.06326216056938211,
+      "eval_accuracy": 0.06339815948139081,
       "eval_loss": 2.5078125,
-      "eval_runtime": 3.1482,
-      "eval_samples_per_second": 18.105,
-      "eval_steps_per_second": 1.271,
+      "eval_runtime": 2.4053,
+      "eval_samples_per_second": 23.698,
+      "eval_steps_per_second": 1.663,
       "step": 14
     },
     {
       "epoch": 0.79,
-      "learning_rate": 2.6311599146653446e-05,
-      "loss": 2.6055,
+      "learning_rate": 2.2139210895556104e-05,
+      "loss": 2.6023,
       "step": 15
     },
     {
       "epoch": 0.79,
-      "eval_accuracy": 0.06335282651072124,
+      "eval_accuracy": 0.06339815948139081,
       "eval_loss": 2.501953125,
-      "eval_runtime": 3.116,
-      "eval_samples_per_second": 18.293,
-      "eval_steps_per_second": 1.284,
+      "eval_runtime": 2.804,
+      "eval_samples_per_second": 20.328,
+      "eval_steps_per_second": 1.427,
       "step": 15
     },
     {
       "epoch": 0.84,
-      "learning_rate": 2.575173769752677e-05,
-      "loss": 2.5061,
+      "learning_rate": 2.1025431369794546e-05,
+      "loss": 2.5034,
       "step": 16
     },
     {
       "epoch": 0.84,
-      "eval_accuracy": 0.06323949408404733,
+      "eval_accuracy": 0.06333016002538647,
       "eval_loss": 2.49609375,
-      "eval_runtime": 2.7142,
-      "eval_samples_per_second": 21.001,
-      "eval_steps_per_second": 1.474,
+      "eval_runtime": 2.1988,
+      "eval_samples_per_second": 25.923,
+      "eval_steps_per_second": 1.819,
       "step": 16
     },
     {
       "epoch": 0.89,
-      "learning_rate": 2.5159223574386117e-05,
-      "loss": 2.4348,
+      "learning_rate": 1.9870492038070255e-05,
+      "loss": 2.4353,
       "step": 17
     },
     {
       "epoch": 0.89,
-      "eval_accuracy": 0.06310349517203863,
+      "eval_accuracy": 0.0631488281427082,
       "eval_loss": 2.490234375,
-      "eval_runtime": 3.1245,
-      "eval_samples_per_second": 18.243,
-      "eval_steps_per_second": 1.28,
+      "eval_runtime": 2.5999,
+      "eval_samples_per_second": 21.924,
+      "eval_steps_per_second": 1.538,
       "step": 17
     },
     {
       "epoch": 0.95,
-      "learning_rate": 2.4535856223149525e-05,
-      "loss": 2.6284,
+      "learning_rate": 1.8682282307111988e-05,
+      "loss": 2.6262,
       "step": 18
     },
     {
       "epoch": 0.95,
-      "eval_accuracy": 0.06319416111337776,
-      "eval_loss": 2.48828125,
-      "eval_runtime": 3.1284,
-      "eval_samples_per_second": 18.22,
-      "eval_steps_per_second": 1.279,
+      "eval_accuracy": 0.06333016002538647,
+      "eval_loss": 2.486328125,
+      "eval_runtime": 2.9949,
+      "eval_samples_per_second": 19.032,
+      "eval_steps_per_second": 1.336,
       "step": 18
     },
     {
       "epoch": 1.0,
-      "learning_rate": 2.3883528789974703e-05,
-      "loss": 2.5574,
+      "learning_rate": 1.746891885421101e-05,
+      "loss": 2.5613,
       "step": 19
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.06308082868670384,
-      "eval_loss": 2.486328125,
-      "eval_runtime": 2.5211,
-      "eval_samples_per_second": 22.61,
-      "eval_steps_per_second": 1.587,
+      "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": 2.320422237183641e-05,
-      "loss": 2.0814,
+      "learning_rate": 1.623869018208499e-05,
+      "loss": 2.115,
       "step": 20
     },
     {
       "epoch": 1.05,
-      "eval_accuracy": 0.0632848270547169,
-      "eval_loss": 2.484375,
-      "eval_runtime": 3.128,
-      "eval_samples_per_second": 18.222,
-      "eval_steps_per_second": 1.279,
+      "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": 2.25e-05,
-      "loss": 2.0636,
+      "learning_rate": 1.5e-05,
+      "loss": 2.0885,
       "step": 21
     },
     {
       "epoch": 1.11,
       "eval_accuracy": 0.06351149190806474,
-      "eval_loss": 2.484375,
-      "eval_runtime": 3.1237,
-      "eval_samples_per_second": 18.248,
-      "eval_steps_per_second": 1.281,
+      "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": 2.25e-05,
-      "loss": 1.9459,
+      "learning_rate": 1.5e-05,
+      "loss": 1.9777,
       "step": 22
     },
     {
       "epoch": 1.16,
       "eval_accuracy": 0.06351149190806474,
-      "eval_loss": 2.484375,
-      "eval_runtime": 2.9479,
-      "eval_samples_per_second": 19.336,
-      "eval_steps_per_second": 1.357,
+      "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": 2.177300037466334e-05,
-      "loss": 2.0527,
+      "learning_rate": 1.3761309817915017e-05,
+      "loss": 2.1053,
       "step": 23
     },
     {
       "epoch": 1.21,
-      "eval_accuracy": 0.06344349245206038,
-      "eval_loss": 2.48828125,
-      "eval_runtime": 2.921,
-      "eval_samples_per_second": 19.514,
-      "eval_steps_per_second": 1.369,
+      "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": 2.1025431369794546e-05,
-      "loss": 1.8881,
+      "learning_rate": 1.2531081145788989e-05,
+      "loss": 1.9324,
       "step": 24
     },
     {
       "epoch": 1.26,
-      "eval_accuracy": 0.06348882542272995,
-      "eval_loss": 2.49609375,
-      "eval_runtime": 3.1227,
-      "eval_samples_per_second": 18.253,
-      "eval_steps_per_second": 1.281,
+      "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": 2.025956332789132e-05,
-      "loss": 1.8668,
+      "learning_rate": 1.1317717692888014e-05,
+      "loss": 1.9122,
       "step": 25
     },
     {
       "epoch": 1.32,
-      "eval_accuracy": 0.06362482433473865,
-      "eval_loss": 2.51171875,
-      "eval_runtime": 3.1172,
-      "eval_samples_per_second": 18.286,
-      "eval_steps_per_second": 1.283,
+      "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": 1.9477722165025422e-05,
-      "loss": 2.0375,
+      "learning_rate": 1.0129507961929749e-05,
+      "loss": 2.1058,
       "step": 26
     },
     {
       "epoch": 1.37,
-      "eval_accuracy": 0.06357949136406908,
-      "eval_loss": 2.529296875,
-      "eval_runtime": 2.719,
-      "eval_samples_per_second": 20.963,
-      "eval_steps_per_second": 1.471,
+      "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": 1.8682282307111988e-05,
-      "loss": 1.9402,
+      "learning_rate": 8.974568630205462e-06,
+      "loss": 1.9803,
       "step": 27
     },
     {
       "epoch": 1.42,
-      "eval_accuracy": 0.06319416111337776,
-      "eval_loss": 2.544921875,
-      "eval_runtime": 3.1228,
-      "eval_samples_per_second": 18.253,
-      "eval_steps_per_second": 1.281,
+      "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": 1.7875659478856077e-05,
-      "loss": 1.6086,
+      "learning_rate": 7.860789104443897e-06,
+      "loss": 1.6486,
       "step": 28
     },
     {
       "epoch": 1.47,
-      "eval_accuracy": 0.06333016002538647,
-      "eval_loss": 2.55859375,
-      "eval_runtime": 2.917,
-      "eval_samples_per_second": 19.541,
-      "eval_steps_per_second": 1.371,
+      "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": 1.7060303367276123e-05,
-      "loss": 1.8185,
+      "learning_rate": 6.795777628163599e-06,
+      "loss": 1.8983,
       "step": 29
     },
     {
       "epoch": 1.53,
-      "eval_accuracy": 0.06319416111337776,
-      "eval_loss": 2.564453125,
-      "eval_runtime": 2.7052,
-      "eval_samples_per_second": 21.07,
-      "eval_steps_per_second": 1.479,
+      "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": 1.623869018208499e-05,
-      "loss": 1.7324,
+      "learning_rate": 5.786809309654983e-06,
+      "loss": 1.8105,
       "step": 30
     },
     {
       "epoch": 1.58,
-      "eval_accuracy": 0.0630128292306995,
-      "eval_loss": 2.560546875,
-      "eval_runtime": 3.1208,
-      "eval_samples_per_second": 18.264,
-      "eval_steps_per_second": 1.282,
+      "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": 1.5413315135522434e-05,
-      "loss": 1.9285,
+      "learning_rate": 4.840776425613887e-06,
+      "loss": 1.9916,
       "step": 31
     },
     {
       "epoch": 1.63,
-      "eval_accuracy": 0.06283149734802121,
-      "eval_loss": 2.552734375,
-      "eval_runtime": 3.1183,
-      "eval_samples_per_second": 18.279,
-      "eval_steps_per_second": 1.283,
+      "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": 1.4586684864477572e-05,
-      "loss": 1.8031,
+      "learning_rate": 3.964141339903026e-06,
+      "loss": 1.886,
       "step": 32
     },
     {
       "epoch": 1.68,
-      "eval_accuracy": 0.06305816220136906,
-      "eval_loss": 2.544921875,
-      "eval_runtime": 2.9132,
-      "eval_samples_per_second": 19.566,
-      "eval_steps_per_second": 1.373,
+      "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": 1.3761309817915017e-05,
-      "loss": 1.7321,
+      "learning_rate": 3.162892359054098e-06,
+      "loss": 1.8013,
       "step": 33
     },
     {
       "epoch": 1.74,
-      "eval_accuracy": 0.0630128292306995,
-      "eval_loss": 2.53515625,
-      "eval_runtime": 2.7189,
-      "eval_samples_per_second": 20.964,
-      "eval_steps_per_second": 1.471,
+      "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": 1.2939696632723877e-05,
-      "loss": 1.7802,
+      "learning_rate": 2.442502826062072e-06,
+      "loss": 1.8435,
       "step": 34
     },
     {
       "epoch": 1.79,
-      "eval_accuracy": 0.0631488281427082,
+      "eval_accuracy": 0.06344349245206038,
       "eval_loss": 2.525390625,
-      "eval_runtime": 3.1376,
-      "eval_samples_per_second": 18.167,
-      "eval_steps_per_second": 1.275,
+      "eval_runtime": 3.2047,
+      "eval_samples_per_second": 17.786,
+      "eval_steps_per_second": 1.248,
       "step": 34
     },
     {
       "epoch": 1.84,
-      "learning_rate": 1.2124340521143929e-05,
-      "loss": 2.0637,
+      "learning_rate": 1.8078937319026655e-06,
+      "loss": 2.1229,
       "step": 35
     },
     {
       "epoch": 1.84,
-      "eval_accuracy": 0.06321682759871254,
-      "eval_loss": 2.515625,
-      "eval_runtime": 3.126,
-      "eval_samples_per_second": 18.234,
-      "eval_steps_per_second": 1.28,
+      "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": 1.1317717692888014e-05,
-      "loss": 1.8159,
+      "learning_rate": 1.2634001001741375e-06,
+      "loss": 1.8739,
       "step": 36
     },
     {
       "epoch": 1.89,
-      "eval_accuracy": 0.0632848270547169,
-      "eval_loss": 2.5078125,
-      "eval_runtime": 2.5304,
-      "eval_samples_per_second": 22.526,
-      "eval_steps_per_second": 1.581,
+      "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": 1.0522277834974586e-05,
-      "loss": 1.7142,
+      "learning_rate": 8.127413744904805e-07,
+      "loss": 1.7528,
       "step": 37
     },
     {
       "epoch": 1.95,
-      "eval_accuracy": 0.06344349245206038,
-      "eval_loss": 2.50390625,
-      "eval_runtime": 2.9235,
-      "eval_samples_per_second": 19.497,
-      "eval_steps_per_second": 1.368,
+      "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": 9.740436672108686e-06,
-      "loss": 1.8793,
+      "learning_rate": 4.589960109100444e-07,
+      "loss": 1.9462,
       "step": 38
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.06337549299605603,
-      "eval_loss": 2.5,
-      "eval_runtime": 3.1269,
-      "eval_samples_per_second": 18.229,
-      "eval_steps_per_second": 1.279,
+      "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": 2.05,
-      "learning_rate": 8.974568630205462e-06,
-      "loss": 1.6914,
-      "step": 39
-    },
-    {
-      "epoch": 2.05,
-      "eval_accuracy": 0.06357949136406908,
-      "eval_loss": 2.501953125,
-      "eval_runtime": 2.3189,
-      "eval_samples_per_second": 24.581,
-      "eval_steps_per_second": 1.725,
-      "step": 39
-    },
-    {
-      "epoch": 2.11,
-      "learning_rate": 8.226999625336663e-06,
-      "loss": 1.411,
-      "step": 40
-    },
-    {
-      "epoch": 2.11,
-      "eval_accuracy": 0.06389682215875607,
-      "eval_loss": 2.50390625,
-      "eval_runtime": 3.1258,
-      "eval_samples_per_second": 18.235,
-      "eval_steps_per_second": 1.28,
-      "step": 40
-    },
-    {
-      "epoch": 2.16,
-      "learning_rate": 7.500000000000004e-06,
-      "loss": 1.4182,
-      "step": 41
-    },
-    {
-      "epoch": 2.16,
-      "eval_accuracy": 0.0638514891880865,
-      "eval_loss": 2.509765625,
-      "eval_runtime": 3.1169,
-      "eval_samples_per_second": 18.287,
-      "eval_steps_per_second": 1.283,
-      "step": 41
-    },
-    {
-      "epoch": 2.21,
-      "learning_rate": 6.795777628163599e-06,
-      "loss": 1.6223,
-      "step": 42
-    },
-    {
-      "epoch": 2.21,
-      "eval_accuracy": 0.06378348973208214,
-      "eval_loss": 2.517578125,
-      "eval_runtime": 2.7171,
-      "eval_samples_per_second": 20.978,
-      "eval_steps_per_second": 1.472,
-      "step": 42
-    },
-    {
-      "epoch": 2.26,
-      "learning_rate": 6.116471210025302e-06,
-      "loss": 1.623,
-      "step": 43
-    },
-    {
-      "epoch": 2.26,
-      "eval_accuracy": 0.06344349245206038,
-      "eval_loss": 2.52734375,
-      "eval_runtime": 2.7121,
-      "eval_samples_per_second": 21.017,
-      "eval_steps_per_second": 1.475,
-      "step": 43
-    },
-    {
-      "epoch": 2.32,
-      "learning_rate": 5.464143776850483e-06,
-      "loss": 1.5748,
-      "step": 44
-    },
-    {
-      "epoch": 2.32,
-      "eval_accuracy": 0.06344349245206038,
-      "eval_loss": 2.537109375,
-      "eval_runtime": 2.9216,
-      "eval_samples_per_second": 19.51,
-      "eval_steps_per_second": 1.369,
-      "step": 44
-    },
-    {
-      "epoch": 2.37,
-      "learning_rate": 4.840776425613887e-06,
-      "loss": 1.7166,
-      "step": 45
-    },
-    {
-      "epoch": 2.37,
-      "eval_accuracy": 0.0631488281427082,
-      "eval_loss": 2.546875,
-      "eval_runtime": 3.1159,
-      "eval_samples_per_second": 18.293,
-      "eval_steps_per_second": 1.284,
-      "step": 45
-    },
-    {
-      "epoch": 2.42,
-      "learning_rate": 4.248262302473233e-06,
-      "loss": 1.3432,
-      "step": 46
-    },
-    {
-      "epoch": 2.42,
-      "eval_accuracy": 0.0629901627453647,
-      "eval_loss": 2.556640625,
-      "eval_runtime": 2.9081,
-      "eval_samples_per_second": 19.6,
-      "eval_steps_per_second": 1.375,
-      "step": 46
-    },
-    {
-      "epoch": 2.47,
-      "learning_rate": 3.688400853346558e-06,
-      "loss": 1.5325,
-      "step": 47
-    },
-    {
-      "epoch": 2.47,
-      "eval_accuracy": 0.06305816220136906,
-      "eval_loss": 2.564453125,
-      "eval_runtime": 2.9197,
-      "eval_samples_per_second": 19.522,
-      "eval_steps_per_second": 1.37,
-      "step": 47
-    },
-    {
-      "epoch": 2.53,
-      "learning_rate": 3.162892359054098e-06,
-      "loss": 1.5076,
-      "step": 48
-    },
-    {
-      "epoch": 2.53,
-      "eval_accuracy": 0.062854163833356,
-      "eval_loss": 2.572265625,
-      "eval_runtime": 3.1077,
-      "eval_samples_per_second": 18.341,
-      "eval_steps_per_second": 1.287,
-      "step": 48
-    },
-    {
-      "epoch": 2.58,
-      "learning_rate": 2.673332771621324e-06,
-      "loss": 1.6636,
-      "step": 49
-    },
-    {
-      "epoch": 2.58,
-      "eval_accuracy": 0.06274083140668207,
-      "eval_loss": 2.578125,
-      "eval_runtime": 2.5045,
-      "eval_samples_per_second": 22.759,
-      "eval_steps_per_second": 1.597,
-      "step": 49
-    },
-    {
-      "epoch": 2.63,
-      "learning_rate": 2.221208867425096e-06,
-      "loss": 1.2897,
-      "step": 50
-    },
-    {
-      "epoch": 2.63,
-      "eval_accuracy": 0.06267283195067773,
-      "eval_loss": 2.583984375,
-      "eval_runtime": 3.1236,
-      "eval_samples_per_second": 18.248,
-      "eval_steps_per_second": 1.281,
-      "step": 50
-    },
-    {
-      "epoch": 2.68,
-      "learning_rate": 1.8078937319026655e-06,
-      "loss": 1.4559,
-      "step": 51
-    },
-    {
-      "epoch": 2.68,
-      "eval_accuracy": 0.06265016546534294,
-      "eval_loss": 2.587890625,
-      "eval_runtime": 2.523,
-      "eval_samples_per_second": 22.592,
-      "eval_steps_per_second": 1.585,
-      "step": 51
-    },
-    {
-      "epoch": 2.74,
-      "learning_rate": 1.4346425895363385e-06,
-      "loss": 1.3904,
-      "step": 52
-    },
-    {
-      "epoch": 2.74,
-      "eval_accuracy": 0.06267283195067773,
-      "eval_loss": 2.58984375,
-      "eval_runtime": 3.1182,
-      "eval_samples_per_second": 18.28,
-      "eval_steps_per_second": 1.283,
-      "step": 52
-    },
-    {
-      "epoch": 2.79,
-      "learning_rate": 1.1025889917779735e-06,
-      "loss": 1.4961,
-      "step": 53
-    },
-    {
-      "epoch": 2.79,
-      "eval_accuracy": 0.0625594995240038,
-      "eval_loss": 2.591796875,
-      "eval_runtime": 2.3139,
-      "eval_samples_per_second": 24.633,
-      "eval_steps_per_second": 1.729,
-      "step": 53
-    },
-    {
-      "epoch": 2.84,
-      "learning_rate": 8.127413744904805e-07,
-      "loss": 1.5276,
-      "step": 54
-    },
-    {
-      "epoch": 2.84,
-      "eval_accuracy": 0.06253683303866903,
-      "eval_loss": 2.59375,
-      "eval_runtime": 2.5424,
-      "eval_samples_per_second": 22.42,
-      "eval_steps_per_second": 1.573,
-      "step": 54
-    },
-    {
-      "epoch": 2.89,
-      "learning_rate": 5.659799953612438e-07,
-      "loss": 1.3479,
-      "step": 55
-    },
-    {
-      "epoch": 2.89,
-      "eval_accuracy": 0.062468833582664675,
-      "eval_loss": 2.595703125,
-      "eval_runtime": 3.124,
-      "eval_samples_per_second": 18.246,
-      "eval_steps_per_second": 1.28,
-      "step": 55
-    },
-    {
-      "epoch": 2.95,
-      "learning_rate": 3.630542605884657e-07,
-      "loss": 1.4094,
-      "step": 56
-    },
-    {
-      "epoch": 2.95,
-      "eval_accuracy": 0.06242350061199511,
-      "eval_loss": 2.595703125,
-      "eval_runtime": 2.3282,
-      "eval_samples_per_second": 24.482,
-      "eval_steps_per_second": 1.718,
-      "step": 56
-    },
-    {
-      "epoch": 3.0,
-      "learning_rate": 2.0458044895916516e-07,
-      "loss": 1.5486,
-      "step": 57
-    },
-    {
-      "epoch": 3.0,
-      "eval_accuracy": 0.06235550115599075,
-      "eval_loss": 2.595703125,
-      "eval_runtime": 2.9026,
-      "eval_samples_per_second": 19.637,
-      "eval_steps_per_second": 1.378,
-      "step": 57
-    },
-    {
-      "epoch": 3.0,
-      "step": 57,
-      "total_flos": 3114896719872.0,
-      "train_loss": 1.9830793414199561,
-      "train_runtime": 617.3135,
-      "train_samples_per_second": 1.473,
-      "train_steps_per_second": 0.092
+      "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
     }
   ],
-  "max_steps": 57,
-  "num_train_epochs": 3,
-  "total_flos": 3114896719872.0,
+  "max_steps": 38,
+  "num_train_epochs": 2,
+  "total_flos": 2058056761344.0,
   "trial_name": null,
   "trial_params": null
 }