Model save

all_results.json

@@ -1,27 +1,27 @@
 {
-    "epoch": 0.35853976531942633,
-    "test_accuracy": 0.7754491017964071,
-    "test_loss": 0.5873367786407471,
-    "test_macro_f1": 0.7741620846115503,
-    "test_macro_precision": 0.7884283981425227,
-    "test_macro_recall": 0.7754491017964072,
-    "test_runtime": 1.48,
-    "test_samples_per_second": 3385.149,
-    "test_steps_per_second": 53.379,
-    "test_weighted_f1": 0.7741620846115502,
-    "total_flos": 9038998352627712.0,
-    "train_loss": 0.7395022045482289,
-    "train_runtime": 205.3102,
+    "epoch": 3.0,
+    "test_accuracy": 0.8421157684630739,
+    "test_loss": 0.43828198313713074,
+    "test_macro_f1": 0.8425829568493537,
+    "test_macro_precision": 0.8480652996607976,
+    "test_macro_recall": 0.8421157684630739,
+    "test_runtime": 9.3912,
+    "test_samples_per_second": 533.48,
+    "test_steps_per_second": 8.412,
+    "test_weighted_f1": 0.8425829568493536,
+    "total_flos": 3.099754961058632e+17,
+    "train_loss": 0.4458606608895207,
+    "train_runtime": 5493.4622,
     "train_samples": 392702,
-    "train_samples_per_second": 5738.176,
-    "train_steps_per_second": 89.659,
-    "validation_accuracy": 0.7759036144578313,
-    "validation_loss": 0.5791348218917847,
-    "validation_macro_f1": 0.7750845837086177,
-    "validation_macro_precision": 0.7882453231618829,
-    "validation_macro_recall": 0.7759036144578313,
-    "validation_runtime": 0.6774,
-    "validation_samples_per_second": 3675.962,
-    "validation_steps_per_second": 57.575,
-    "validation_weighted_f1": 0.7750845837086177
+    "train_samples_per_second": 214.456,
+    "train_steps_per_second": 3.351,
+    "validation_accuracy": 0.8349397590361446,
+    "validation_loss": 0.44213181734085083,
+    "validation_macro_f1": 0.8357775851206196,
+    "validation_macro_precision": 0.841060220470884,
+    "validation_macro_recall": 0.8349397590361446,
+    "validation_runtime": 4.581,
+    "validation_samples_per_second": 543.555,
+    "validation_steps_per_second": 8.514,
+    "validation_weighted_f1": 0.8357775851206194
 }

per_language_test_metrics.csv

@@ -0,0 +1,16 @@
+language,accuracy,macro_f1,weighted_f1
+en,0.8421157684630739,0.8425829568493537,0.8425829568493536
+es,0.7495009980039921,0.7453574588534669,0.745357458853467
+fr,0.7489021956087825,0.7451661982804695,0.7451661982804695
+ru,0.7253493013972055,0.7162195026232362,0.7162195026232363
+de,0.7227544910179641,0.7135089464658995,0.7135089464658994
+bg,0.7129740518962076,0.7021542465700179,0.7021542465700179
+vi,0.6902195608782435,0.6751038144652277,0.6751038144652278
+zh,0.6926147704590818,0.6750201845330862,0.6750201845330862
+th,0.6726546906187625,0.6501680609668395,0.6501680609668395
+el,0.6726546906187625,0.649670141857844,0.649670141857844
+tr,0.6429141716566866,0.6148982403711342,0.614898240371134
+hi,0.6099800399201597,0.5697032867533204,0.5697032867533204
+ar,0.6003992015968064,0.5490620317671846,0.5490620317671846
+ur,0.5568862275449101,0.49464629258512066,0.4946462925851207
+sw,0.5343313373253493,0.4724317512711995,0.47243175127119946

test_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 0.35853976531942633,
-    "test_accuracy": 0.7754491017964071,
-    "test_loss": 0.5873367786407471,
-    "test_macro_f1": 0.7741620846115503,
-    "test_macro_precision": 0.7884283981425227,
-    "test_macro_recall": 0.7754491017964072,
-    "test_runtime": 1.48,
-    "test_samples_per_second": 3385.149,
-    "test_steps_per_second": 53.379,
-    "test_weighted_f1": 0.7741620846115502
+    "epoch": 3.0,
+    "test_accuracy": 0.8421157684630739,
+    "test_loss": 0.43828198313713074,
+    "test_macro_f1": 0.8425829568493537,
+    "test_macro_precision": 0.8480652996607976,
+    "test_macro_recall": 0.8421157684630739,
+    "test_runtime": 9.3912,
+    "test_samples_per_second": 533.48,
+    "test_steps_per_second": 8.412,
+    "test_weighted_f1": 0.8425829568493536
 }

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 0.35853976531942633,
-    "total_flos": 9038998352627712.0,
-    "train_loss": 0.7395022045482289,
-    "train_runtime": 205.3102,
+    "epoch": 3.0,
+    "total_flos": 3.099754961058632e+17,
+    "train_loss": 0.4458606608895207,
+    "train_runtime": 5493.4622,
     "train_samples": 392702,
-    "train_samples_per_second": 5738.176,
-    "train_steps_per_second": 89.659
+    "train_samples_per_second": 214.456,
+    "train_steps_per_second": 3.351
 }

trainer_state.json

@@ -1,491 +1,402 @@
 {
-  "best_global_step": 1900,
-  "best_metric": 0.8105035074694252,
-  "best_model_checkpoint": null,
-  "epoch": 0.35853976531942633,
-  "eval_steps": 100,
-  "global_step": 2200,
+  "best_global_step": 14000,
+  "best_metric": 0.8463211795322131,
+  "best_model_checkpoint": "xnli-xlm-roberta-base/checkpoint-10000",
+  "epoch": 3.0,
+  "eval_steps": 1000,
+  "global_step": 18408,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
-    {
-      "epoch": 0.016297262059973925,
-      "grad_norm": 2.553225040435791,
-      "learning_rate": 1.7918552036199097e-06,
-      "loss": 1.1077,
-      "step": 100
-    },
-    {
-      "epoch": 0.016297262059973925,
-      "eval_accuracy": 0.3333333333333333,
-      "eval_loss": 1.104527473449707,
-      "eval_macro_f1": 0.16666666666666666,
-      "eval_macro_precision": 0.1111111111111111,
-      "eval_macro_recall": 0.3333333333333333,
-      "eval_runtime": 0.667,
-      "eval_samples_per_second": 3733.376,
-      "eval_steps_per_second": 58.475,
-      "eval_weighted_f1": 0.16666666666666666,
-      "step": 100
-    },
-    {
-      "epoch": 0.03259452411994785,
-      "grad_norm": 1.907884120941162,
-      "learning_rate": 3.6018099547511313e-06,
-      "loss": 1.1039,
-      "step": 200
-    },
-    {
-      "epoch": 0.03259452411994785,
-      "eval_accuracy": 0.3333333333333333,
-      "eval_loss": 1.0994290113449097,
-      "eval_macro_f1": 0.16666666666666666,
-      "eval_macro_precision": 0.1111111111111111,
-      "eval_macro_recall": 0.3333333333333333,
-      "eval_runtime": 0.6647,
-      "eval_samples_per_second": 3745.923,
-      "eval_steps_per_second": 58.671,
-      "eval_weighted_f1": 0.16666666666666666,
-      "step": 200
-    },
-    {
-      "epoch": 0.04889178617992177,
-      "grad_norm": 8.018975257873535,
-      "learning_rate": 5.411764705882353e-06,
-      "loss": 1.0902,
-      "step": 300
-    },
-    {
-      "epoch": 0.04889178617992177,
-      "eval_accuracy": 0.4393574297188755,
-      "eval_loss": 1.0789587497711182,
-      "eval_macro_f1": 0.43193114861778614,
-      "eval_macro_precision": 0.46277646586041393,
-      "eval_macro_recall": 0.43935742971887554,
-      "eval_runtime": 0.6642,
-      "eval_samples_per_second": 3748.76,
-      "eval_steps_per_second": 58.716,
-      "eval_weighted_f1": 0.43193114861778614,
-      "step": 300
-    },
-    {
-      "epoch": 0.0651890482398957,
-      "grad_norm": 11.130058288574219,
-      "learning_rate": 7.221719457013575e-06,
-      "loss": 1.0247,
-      "step": 400
-    },
-    {
-      "epoch": 0.0651890482398957,
-      "eval_accuracy": 0.4751004016064257,
-      "eval_loss": 1.0407814979553223,
-      "eval_macro_f1": 0.43127519152885446,
-      "eval_macro_precision": 0.5305172899247589,
-      "eval_macro_recall": 0.47510040160642575,
-      "eval_runtime": 0.6659,
-      "eval_samples_per_second": 3739.322,
-      "eval_steps_per_second": 58.568,
-      "eval_weighted_f1": 0.4312751915288544,
-      "step": 400
-    },
-    {
-      "epoch": 0.08148631029986962,
-      "grad_norm": 8.967389106750488,
-      "learning_rate": 9.031674208144797e-06,
-      "loss": 0.9446,
-      "step": 500
-    },
-    {
-      "epoch": 0.08148631029986962,
-      "eval_accuracy": 0.6052208835341365,
-      "eval_loss": 0.8834017515182495,
-      "eval_macro_f1": 0.6044827304056648,
-      "eval_macro_precision": 0.6564437205440745,
-      "eval_macro_recall": 0.6052208835341366,
-      "eval_runtime": 0.6636,
-      "eval_samples_per_second": 3752.135,
-      "eval_steps_per_second": 58.768,
-      "eval_weighted_f1": 0.6044827304056647,
-      "step": 500
-    },
-    {
-      "epoch": 0.09778357235984354,
-      "grad_norm": 18.94449806213379,
-      "learning_rate": 1.0841628959276019e-05,
-      "loss": 0.8576,
-      "step": 600
-    },
-    {
-      "epoch": 0.09778357235984354,
-      "eval_accuracy": 0.6919678714859437,
-      "eval_loss": 0.7354851961135864,
-      "eval_macro_f1": 0.6922967262959062,
-      "eval_macro_precision": 0.69700080086924,
-      "eval_macro_recall": 0.6919678714859437,
-      "eval_runtime": 0.6678,
-      "eval_samples_per_second": 3728.877,
-      "eval_steps_per_second": 58.404,
-      "eval_weighted_f1": 0.6922967262959063,
-      "step": 600
-    },
-    {
-      "epoch": 0.11408083441981746,
-      "grad_norm": 15.024765968322754,
-      "learning_rate": 1.265158371040724e-05,
-      "loss": 0.7851,
-      "step": 700
-    },
-    {
-      "epoch": 0.11408083441981746,
-      "eval_accuracy": 0.7216867469879518,
-      "eval_loss": 0.6741242408752441,
-      "eval_macro_f1": 0.7208840894640302,
-      "eval_macro_precision": 0.730993053217949,
-      "eval_macro_recall": 0.7216867469879519,
-      "eval_runtime": 0.667,
-      "eval_samples_per_second": 3733.401,
-      "eval_steps_per_second": 58.475,
-      "eval_weighted_f1": 0.7208840894640302,
-      "step": 700
-    },
-    {
-      "epoch": 0.1303780964797914,
-      "grad_norm": 12.894227981567383,
-      "learning_rate": 1.4461538461538462e-05,
-      "loss": 0.7377,
-      "step": 800
-    },
-    {
-      "epoch": 0.1303780964797914,
-      "eval_accuracy": 0.742570281124498,
-      "eval_loss": 0.6462946534156799,
-      "eval_macro_f1": 0.7438220814732438,
-      "eval_macro_precision": 0.7481277304403244,
-      "eval_macro_recall": 0.742570281124498,
-      "eval_runtime": 0.6663,
-      "eval_samples_per_second": 3737.302,
-      "eval_steps_per_second": 58.536,
-      "eval_weighted_f1": 0.7438220814732438,
-      "step": 800
-    },
-    {
-      "epoch": 0.14667535853976532,
-      "grad_norm": 16.079246520996094,
-      "learning_rate": 1.6271493212669684e-05,
-      "loss": 0.699,
-      "step": 900
-    },
-    {
-      "epoch": 0.14667535853976532,
-      "eval_accuracy": 0.7469879518072289,
-      "eval_loss": 0.6353606581687927,
-      "eval_macro_f1": 0.7483067221406298,
-      "eval_macro_precision": 0.7601105361776005,
-      "eval_macro_recall": 0.7469879518072289,
-      "eval_runtime": 0.6656,
-      "eval_samples_per_second": 3740.748,
-      "eval_steps_per_second": 58.59,
-      "eval_weighted_f1": 0.7483067221406298,
-      "step": 900
-    },
     {
       "epoch": 0.16297262059973924,
-      "grad_norm": 14.99389362335205,
+      "grad_norm": 10.51728343963623,
       "learning_rate": 1.8081447963800906e-05,
-      "loss": 0.6922,
+      "loss": 0.9575,
       "step": 1000
     },
     {
       "epoch": 0.16297262059973924,
-      "eval_accuracy": 0.7449799196787149,
-      "eval_loss": 0.6129863858222961,
-      "eval_macro_f1": 0.7459488107225619,
-      "eval_macro_precision": 0.7535338260093193,
-      "eval_macro_recall": 0.7449799196787148,
-      "eval_runtime": 0.6673,
-      "eval_samples_per_second": 3731.717,
-      "eval_steps_per_second": 58.449,
-      "eval_weighted_f1": 0.745948810722562,
+      "eval_accuracy": 0.7164658634538152,
+      "eval_loss": 0.6793871521949768,
+      "eval_macro_f1": 0.7179694771560841,
+      "eval_macro_precision": 0.7238898844951175,
+      "eval_macro_recall": 0.7164658634538151,
+      "eval_runtime": 4.5345,
+      "eval_samples_per_second": 549.125,
+      "eval_steps_per_second": 8.601,
+      "eval_weighted_f1": 0.7179694771560841,
       "step": 1000
     },
-    {
-      "epoch": 0.17926988265971316,
-      "grad_norm": 18.338092803955078,
-      "learning_rate": 1.9891402714932128e-05,
-      "loss": 0.6572,
-      "step": 1100
-    },
-    {
-      "epoch": 0.17926988265971316,
-      "eval_accuracy": 0.7305220883534137,
-      "eval_loss": 0.6956562399864197,
-      "eval_macro_f1": 0.7300827453358979,
-      "eval_macro_precision": 0.7581368909493604,
-      "eval_macro_recall": 0.7305220883534136,
-      "eval_runtime": 0.6807,
-      "eval_samples_per_second": 3657.801,
-      "eval_steps_per_second": 57.291,
-      "eval_weighted_f1": 0.7300827453358979,
-      "step": 1100
-    },
-    {
-      "epoch": 0.19556714471968709,
-      "grad_norm": 11.609926223754883,
-      "learning_rate": 1.989134832110039e-05,
-      "loss": 0.6592,
-      "step": 1200
-    },
-    {
-      "epoch": 0.19556714471968709,
-      "eval_accuracy": 0.7526104417670683,
-      "eval_loss": 0.614703893661499,
-      "eval_macro_f1": 0.7539523536061177,
-      "eval_macro_precision": 0.7695772179728464,
-      "eval_macro_recall": 0.7526104417670684,
-      "eval_runtime": 0.6686,
-      "eval_samples_per_second": 3724.239,
-      "eval_steps_per_second": 58.331,
-      "eval_weighted_f1": 0.7539523536061176,
-      "step": 1200
-    },
-    {
-      "epoch": 0.211864406779661,
-      "grad_norm": 11.125432968139648,
-      "learning_rate": 1.977576142865399e-05,
-      "loss": 0.6382,
-      "step": 1300
-    },
-    {
-      "epoch": 0.211864406779661,
-      "eval_accuracy": 0.763855421686747,
-      "eval_loss": 0.5744169354438782,
-      "eval_macro_f1": 0.7642595462026183,
-      "eval_macro_precision": 0.7745668320768545,
-      "eval_macro_recall": 0.7638554216867469,
-      "eval_runtime": 0.6666,
-      "eval_samples_per_second": 3735.42,
-      "eval_steps_per_second": 58.507,
-      "eval_weighted_f1": 0.7642595462026183,
-      "step": 1300
-    },
-    {
-      "epoch": 0.22816166883963493,
-      "grad_norm": 18.14826011657715,
-      "learning_rate": 1.9660174536207594e-05,
-      "loss": 0.6136,
-      "step": 1400
-    },
-    {
-      "epoch": 0.22816166883963493,
-      "eval_accuracy": 0.7562248995983936,
-      "eval_loss": 0.6171919107437134,
-      "eval_macro_f1": 0.7578385667633957,
-      "eval_macro_precision": 0.7795494771340636,
-      "eval_macro_recall": 0.7562248995983936,
-      "eval_runtime": 0.6641,
-      "eval_samples_per_second": 3749.623,
-      "eval_steps_per_second": 58.729,
-      "eval_weighted_f1": 0.7578385667633957,
-      "step": 1400
-    },
-    {
-      "epoch": 0.24445893089960888,
-      "grad_norm": 13.110569953918457,
-      "learning_rate": 1.9544587643761198e-05,
-      "loss": 0.6097,
-      "step": 1500
-    },
-    {
-      "epoch": 0.24445893089960888,
-      "eval_accuracy": 0.7485943775100402,
-      "eval_loss": 0.6241039037704468,
-      "eval_macro_f1": 0.7494034450273651,
-      "eval_macro_precision": 0.7724661665486536,
-      "eval_macro_recall": 0.7485943775100402,
-      "eval_runtime": 0.6629,
-      "eval_samples_per_second": 3755.99,
-      "eval_steps_per_second": 58.829,
-      "eval_weighted_f1": 0.7494034450273649,
-      "step": 1500
-    },
-    {
-      "epoch": 0.2607561929595828,
-      "grad_norm": 14.377798080444336,
-      "learning_rate": 1.9429000751314802e-05,
-      "loss": 0.5981,
-      "step": 1600
-    },
-    {
-      "epoch": 0.2607561929595828,
-      "eval_accuracy": 0.7755020080321285,
-      "eval_loss": 0.5857760310173035,
-      "eval_macro_f1": 0.7756304467061907,
-      "eval_macro_precision": 0.7868413877422515,
-      "eval_macro_recall": 0.7755020080321285,
-      "eval_runtime": 0.6611,
-      "eval_samples_per_second": 3766.404,
-      "eval_steps_per_second": 58.992,
-      "eval_weighted_f1": 0.7756304467061907,
-      "step": 1600
-    },
-    {
-      "epoch": 0.2770534550195567,
-      "grad_norm": 10.924996376037598,
-      "learning_rate": 1.9313413858868406e-05,
-      "loss": 0.6077,
-      "step": 1700
-    },
-    {
-      "epoch": 0.2770534550195567,
-      "eval_accuracy": 0.7738955823293173,
-      "eval_loss": 0.5644207000732422,
-      "eval_macro_f1": 0.7726011626210361,
-      "eval_macro_precision": 0.7775631481368256,
-      "eval_macro_recall": 0.7738955823293173,
-      "eval_runtime": 0.6716,
-      "eval_samples_per_second": 3707.806,
-      "eval_steps_per_second": 58.074,
-      "eval_weighted_f1": 0.7726011626210361,
-      "step": 1700
-    },
-    {
-      "epoch": 0.29335071707953064,
-      "grad_norm": 10.792520523071289,
-      "learning_rate": 1.919782696642201e-05,
-      "loss": 0.5685,
-      "step": 1800
-    },
-    {
-      "epoch": 0.29335071707953064,
-      "eval_accuracy": 0.7726907630522089,
-      "eval_loss": 0.5637930631637573,
-      "eval_macro_f1": 0.771470756049217,
-      "eval_macro_precision": 0.7776570624573044,
-      "eval_macro_recall": 0.7726907630522089,
-      "eval_runtime": 0.6689,
-      "eval_samples_per_second": 3722.561,
-      "eval_steps_per_second": 58.305,
-      "eval_weighted_f1": 0.771470756049217,
-      "step": 1800
-    },
-    {
-      "epoch": 0.30964797913950454,
-      "grad_norm": 11.029982566833496,
-      "learning_rate": 1.9082240073975613e-05,
-      "loss": 0.5776,
-      "step": 1900
-    },
-    {
-      "epoch": 0.30964797913950454,
-      "eval_accuracy": 0.810843373493976,
-      "eval_loss": 0.5153859853744507,
-      "eval_macro_f1": 0.8105035074694252,
-      "eval_macro_precision": 0.8121165203086332,
-      "eval_macro_recall": 0.8108433734939758,
-      "eval_runtime": 0.6653,
-      "eval_samples_per_second": 3742.528,
-      "eval_steps_per_second": 58.618,
-      "eval_weighted_f1": 0.8105035074694252,
-      "step": 1900
-    },
     {
       "epoch": 0.3259452411994785,
-      "grad_norm": 16.579381942749023,
+      "grad_norm": 6.724602699279785,
       "learning_rate": 1.8966653181529214e-05,
-      "loss": 0.5557,
+      "loss": 0.6483,
       "step": 2000
     },
     {
       "epoch": 0.3259452411994785,
-      "eval_accuracy": 0.8028112449799196,
-      "eval_loss": 0.5254373550415039,
-      "eval_macro_f1": 0.8021145838057344,
-      "eval_macro_precision": 0.8057436512211583,
-      "eval_macro_recall": 0.8028112449799196,
-      "eval_runtime": 0.6693,
-      "eval_samples_per_second": 3720.577,
-      "eval_steps_per_second": 58.274,
-      "eval_weighted_f1": 0.8021145838057343,
+      "eval_accuracy": 0.7967871485943775,
+      "eval_loss": 0.5350644588470459,
+      "eval_macro_f1": 0.7972864852344191,
+      "eval_macro_precision": 0.8018465694414347,
+      "eval_macro_recall": 0.7967871485943775,
+      "eval_runtime": 4.5631,
+      "eval_samples_per_second": 545.684,
+      "eval_steps_per_second": 8.547,
+      "eval_weighted_f1": 0.7972864852344191,
       "step": 2000
     },
     {
-      "epoch": 0.34224250325945244,
-      "grad_norm": 11.838844299316406,
-      "learning_rate": 1.885106628908282e-05,
-      "loss": 0.5533,
-      "step": 2100
+      "epoch": 0.48891786179921776,
+      "grad_norm": 6.1638946533203125,
+      "learning_rate": 1.781078425706525e-05,
+      "loss": 0.5616,
+      "step": 3000
     },
     {
-      "epoch": 0.34224250325945244,
+      "epoch": 0.48891786179921776,
       "eval_accuracy": 0.7947791164658634,
-      "eval_loss": 0.5396072864532471,
-      "eval_macro_f1": 0.7957208629368709,
-      "eval_macro_precision": 0.799954081725605,
+      "eval_loss": 0.509804368019104,
+      "eval_macro_f1": 0.7951670773482286,
+      "eval_macro_precision": 0.8064539329297533,
       "eval_macro_recall": 0.7947791164658634,
-      "eval_runtime": 0.6759,
-      "eval_samples_per_second": 3684.172,
-      "eval_steps_per_second": 57.704,
-      "eval_weighted_f1": 0.7957208629368708,
-      "step": 2100
-    },
-    {
-      "epoch": 0.35853976531942633,
-      "grad_norm": 14.574402809143066,
-      "learning_rate": 1.873547939663642e-05,
-      "loss": 0.5877,
-      "step": 2200
-    },
-    {
-      "epoch": 0.35853976531942633,
-      "eval_accuracy": 0.7759036144578313,
-      "eval_loss": 0.5791348218917847,
-      "eval_macro_f1": 0.7750845837086177,
-      "eval_macro_precision": 0.7882453231618829,
-      "eval_macro_recall": 0.7759036144578313,
-      "eval_runtime": 0.6819,
-      "eval_samples_per_second": 3651.488,
-      "eval_steps_per_second": 57.192,
-      "eval_weighted_f1": 0.7750845837086177,
-      "step": 2200
-    },
-    {
-      "epoch": 0.35853976531942633,
-      "step": 2200,
-      "total_flos": 9038998352627712.0,
-      "train_loss": 0.7395022045482289,
-      "train_runtime": 205.3102,
-      "train_samples_per_second": 5738.176,
-      "train_steps_per_second": 89.659
+      "eval_runtime": 4.6567,
+      "eval_samples_per_second": 534.712,
+      "eval_steps_per_second": 8.375,
+      "eval_weighted_f1": 0.7951670773482287,
+      "step": 3000
+    },
+    {
+      "epoch": 0.651890482398957,
+      "grad_norm": 5.407867908477783,
+      "learning_rate": 1.6654915332601286e-05,
+      "loss": 0.5226,
+      "step": 4000
+    },
+    {
+      "epoch": 0.651890482398957,
+      "eval_accuracy": 0.8004016064257028,
+      "eval_loss": 0.49269911646842957,
+      "eval_macro_f1": 0.8015680893541245,
+      "eval_macro_precision": 0.8123398526926758,
+      "eval_macro_recall": 0.8004016064257028,
+      "eval_runtime": 4.6437,
+      "eval_samples_per_second": 536.212,
+      "eval_steps_per_second": 8.399,
+      "eval_weighted_f1": 0.8015680893541245,
+      "step": 4000
+    },
+    {
+      "epoch": 0.8148631029986962,
+      "grad_norm": 8.23536205291748,
+      "learning_rate": 1.549904640813732e-05,
+      "loss": 0.4966,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8148631029986962,
+      "eval_accuracy": 0.8124497991967872,
+      "eval_loss": 0.480131596326828,
+      "eval_macro_f1": 0.8128197553956924,
+      "eval_macro_precision": 0.8202401283021422,
+      "eval_macro_recall": 0.8124497991967873,
+      "eval_runtime": 4.5688,
+      "eval_samples_per_second": 545.006,
+      "eval_steps_per_second": 8.536,
+      "eval_weighted_f1": 0.8128197553956925,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9778357235984355,
+      "grad_norm": 9.736433029174805,
+      "learning_rate": 1.4343177483673352e-05,
+      "loss": 0.4799,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9778357235984355,
+      "eval_accuracy": 0.8092369477911646,
+      "eval_loss": 0.4666927456855774,
+      "eval_macro_f1": 0.8108832750227813,
+      "eval_macro_precision": 0.8254810350822899,
+      "eval_macro_recall": 0.8092369477911646,
+      "eval_runtime": 4.542,
+      "eval_samples_per_second": 548.221,
+      "eval_steps_per_second": 8.587,
+      "eval_weighted_f1": 0.8108832750227812,
+      "step": 6000
+    },
+    {
+      "epoch": 1.1408083441981747,
+      "grad_norm": 8.493484497070312,
+      "learning_rate": 1.3187308559209387e-05,
+      "loss": 0.4251,
+      "step": 7000
+    },
+    {
+      "epoch": 1.1408083441981747,
+      "eval_accuracy": 0.8180722891566266,
+      "eval_loss": 0.45433178544044495,
+      "eval_macro_f1": 0.8191191956693326,
+      "eval_macro_precision": 0.8269376625195844,
+      "eval_macro_recall": 0.8180722891566266,
+      "eval_runtime": 4.4157,
+      "eval_samples_per_second": 563.894,
+      "eval_steps_per_second": 8.832,
+      "eval_weighted_f1": 0.8191191956693327,
+      "step": 7000
+    },
+    {
+      "epoch": 1.303780964797914,
+      "grad_norm": 9.21502685546875,
+      "learning_rate": 1.2031439634745421e-05,
+      "loss": 0.4186,
+      "step": 8000
+    },
+    {
+      "epoch": 1.303780964797914,
+      "eval_accuracy": 0.8176706827309237,
+      "eval_loss": 0.46454402804374695,
+      "eval_macro_f1": 0.8173393684861902,
+      "eval_macro_precision": 0.8258790100555838,
+      "eval_macro_recall": 0.8176706827309238,
+      "eval_runtime": 4.6094,
+      "eval_samples_per_second": 540.202,
+      "eval_steps_per_second": 8.461,
+      "eval_weighted_f1": 0.8173393684861902,
+      "step": 8000
+    },
+    {
+      "epoch": 1.4667535853976532,
+      "grad_norm": 7.096667289733887,
+      "learning_rate": 1.0875570710281456e-05,
+      "loss": 0.4129,
+      "step": 9000
+    },
+    {
+      "epoch": 1.4667535853976532,
+      "eval_accuracy": 0.8257028112449799,
+      "eval_loss": 0.44398775696754456,
+      "eval_macro_f1": 0.8268290877462207,
+      "eval_macro_precision": 0.8358894256319416,
+      "eval_macro_recall": 0.8257028112449799,
+      "eval_runtime": 4.7765,
+      "eval_samples_per_second": 521.303,
+      "eval_steps_per_second": 8.165,
+      "eval_weighted_f1": 0.8268290877462207,
+      "step": 9000
+    },
+    {
+      "epoch": 1.6297262059973925,
+      "grad_norm": 6.999719142913818,
+      "learning_rate": 9.719701785817489e-06,
+      "loss": 0.4053,
+      "step": 10000
+    },
+    {
+      "epoch": 1.6297262059973925,
+      "eval_accuracy": 0.8309236947791164,
+      "eval_loss": 0.4421282708644867,
+      "eval_macro_f1": 0.8314919406349293,
+      "eval_macro_precision": 0.836586446573221,
+      "eval_macro_recall": 0.8309236947791164,
+      "eval_runtime": 4.5642,
+      "eval_samples_per_second": 545.553,
+      "eval_steps_per_second": 8.545,
+      "eval_weighted_f1": 0.8314919406349294,
+      "step": 10000
+    },
+    {
+      "epoch": 1.7926988265971318,
+      "grad_norm": 7.709704399108887,
+      "learning_rate": 8.563832861353523e-06,
+      "loss": 0.3971,
+      "step": 11000
+    },
+    {
+      "epoch": 1.7926988265971318,
+      "eval_accuracy": 0.8248995983935743,
+      "eval_loss": 0.4437423646450043,
+      "eval_macro_f1": 0.825940758187211,
+      "eval_macro_precision": 0.8357952618825585,
+      "eval_macro_recall": 0.8248995983935744,
+      "eval_runtime": 4.5323,
+      "eval_samples_per_second": 549.394,
+      "eval_steps_per_second": 8.605,
+      "eval_weighted_f1": 0.8259407581872109,
+      "step": 11000
+    },
+    {
+      "epoch": 1.9556714471968708,
+      "grad_norm": 5.534423828125,
+      "learning_rate": 7.407963936889557e-06,
+      "loss": 0.3947,
+      "step": 12000
+    },
+    {
+      "epoch": 1.9556714471968708,
+      "eval_accuracy": 0.834136546184739,
+      "eval_loss": 0.4154554307460785,
+      "eval_macro_f1": 0.8350152412898564,
+      "eval_macro_precision": 0.8416677147812893,
+      "eval_macro_recall": 0.8341365461847389,
+      "eval_runtime": 4.5044,
+      "eval_samples_per_second": 552.797,
+      "eval_steps_per_second": 8.658,
+      "eval_weighted_f1": 0.8350152412898562,
+      "step": 12000
+    },
+    {
+      "epoch": 2.1186440677966103,
+      "grad_norm": 6.122025489807129,
+      "learning_rate": 6.252095012425592e-06,
+      "loss": 0.3437,
+      "step": 13000
+    },
+    {
+      "epoch": 2.1186440677966103,
+      "eval_accuracy": 0.8236947791164658,
+      "eval_loss": 0.4527803659439087,
+      "eval_macro_f1": 0.8249615968184654,
+      "eval_macro_precision": 0.8339980704173974,
+      "eval_macro_recall": 0.8236947791164658,
+      "eval_runtime": 4.5511,
+      "eval_samples_per_second": 547.117,
+      "eval_steps_per_second": 8.569,
+      "eval_weighted_f1": 0.8249615968184654,
+      "step": 13000
+    },
+    {
+      "epoch": 2.2816166883963493,
+      "grad_norm": 10.657154083251953,
+      "learning_rate": 5.096226087961625e-06,
+      "loss": 0.3296,
+      "step": 14000
+    },
+    {
+      "epoch": 2.2816166883963493,
+      "eval_accuracy": 0.8461847389558232,
+      "eval_loss": 0.4210074543952942,
+      "eval_macro_f1": 0.8463211795322131,
+      "eval_macro_precision": 0.8503350839590045,
+      "eval_macro_recall": 0.8461847389558232,
+      "eval_runtime": 4.5213,
+      "eval_samples_per_second": 550.73,
+      "eval_steps_per_second": 8.626,
+      "eval_weighted_f1": 0.846321179532213,
+      "step": 14000
+    },
+    {
+      "epoch": 2.444589308996089,
+      "grad_norm": 7.961305141448975,
+      "learning_rate": 3.940357163497659e-06,
+      "loss": 0.3219,
+      "step": 15000
+    },
+    {
+      "epoch": 2.444589308996089,
+      "eval_accuracy": 0.8337349397590361,
+      "eval_loss": 0.4486163258552551,
+      "eval_macro_f1": 0.8349888002365645,
+      "eval_macro_precision": 0.8443748814531276,
+      "eval_macro_recall": 0.8337349397590362,
+      "eval_runtime": 4.5481,
+      "eval_samples_per_second": 547.482,
+      "eval_steps_per_second": 8.575,
+      "eval_weighted_f1": 0.8349888002365645,
+      "step": 15000
+    },
+    {
+      "epoch": 2.607561929595828,
+      "grad_norm": 7.545483589172363,
+      "learning_rate": 2.7844882390336937e-06,
+      "loss": 0.3241,
+      "step": 16000
+    },
+    {
+      "epoch": 2.607561929595828,
+      "eval_accuracy": 0.8333333333333334,
+      "eval_loss": 0.46160411834716797,
+      "eval_macro_f1": 0.8337624688566216,
+      "eval_macro_precision": 0.8404933913764827,
+      "eval_macro_recall": 0.8333333333333334,
+      "eval_runtime": 4.5462,
+      "eval_samples_per_second": 547.708,
+      "eval_steps_per_second": 8.579,
+      "eval_weighted_f1": 0.8337624688566218,
+      "step": 16000
+    },
+    {
+      "epoch": 2.770534550195567,
+      "grad_norm": 7.061276435852051,
+      "learning_rate": 1.6286193145697278e-06,
+      "loss": 0.326,
+      "step": 17000
+    },
+    {
+      "epoch": 2.770534550195567,
+      "eval_accuracy": 0.8345381526104417,
+      "eval_loss": 0.4480307400226593,
+      "eval_macro_f1": 0.8351454373443175,
+      "eval_macro_precision": 0.8421626579409135,
+      "eval_macro_recall": 0.8345381526104418,
+      "eval_runtime": 4.5496,
+      "eval_samples_per_second": 547.299,
+      "eval_steps_per_second": 8.572,
+      "eval_weighted_f1": 0.8351454373443173,
+      "step": 17000
+    },
+    {
+      "epoch": 2.9335071707953064,
+      "grad_norm": 6.060895919799805,
+      "learning_rate": 4.72750390105762e-07,
+      "loss": 0.3128,
+      "step": 18000
+    },
+    {
+      "epoch": 2.9335071707953064,
+      "eval_accuracy": 0.834136546184739,
+      "eval_loss": 0.45053839683532715,
+      "eval_macro_f1": 0.8350207810106501,
+      "eval_macro_precision": 0.8411309111788473,
+      "eval_macro_recall": 0.8341365461847391,
+      "eval_runtime": 4.5608,
+      "eval_samples_per_second": 545.951,
+      "eval_steps_per_second": 8.551,
+      "eval_weighted_f1": 0.8350207810106502,
+      "step": 18000
+    },
+    {
+      "epoch": 3.0,
+      "step": 18408,
+      "total_flos": 3.099754961058632e+17,
+      "train_loss": 0.4458606608895207,
+      "train_runtime": 5493.4622,
+      "train_samples_per_second": 214.456,
+      "train_steps_per_second": 3.351
     }
   ],
-  "logging_steps": 100,
+  "logging_steps": 1000,
   "max_steps": 18408,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
-  "save_steps": 1000,
+  "save_steps": 10000,
   "stateful_callbacks": {
-    "EarlyStoppingCallback": {
-      "args": {
-        "early_stopping_patience": 3,
-        "early_stopping_threshold": 0.0
-      },
-      "attributes": {
-        "early_stopping_patience_counter": 1
-      }
-    },
     "TrainerControl": {
       "args": {
         "should_epoch_stop": false,
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }
   },
-  "total_flos": 9038998352627712.0,
+  "total_flos": 3.099754961058632e+17,
   "train_batch_size": 64,
   "trial_name": null,
   "trial_params": null

validation_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 0.35853976531942633,
-    "validation_accuracy": 0.7759036144578313,
-    "validation_loss": 0.5791348218917847,
-    "validation_macro_f1": 0.7750845837086177,
-    "validation_macro_precision": 0.7882453231618829,
-    "validation_macro_recall": 0.7759036144578313,
-    "validation_runtime": 0.6774,
-    "validation_samples_per_second": 3675.962,
-    "validation_steps_per_second": 57.575,
-    "validation_weighted_f1": 0.7750845837086177
+    "epoch": 3.0,
+    "validation_accuracy": 0.8349397590361446,
+    "validation_loss": 0.44213181734085083,
+    "validation_macro_f1": 0.8357775851206196,
+    "validation_macro_precision": 0.841060220470884,
+    "validation_macro_recall": 0.8349397590361446,
+    "validation_runtime": 4.581,
+    "validation_samples_per_second": 543.555,
+    "validation_steps_per_second": 8.514,
+    "validation_weighted_f1": 0.8357775851206194
 }