diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/last-checkpoint/trainer_state.json"
@@ -0,0 +1,5381 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.280224929709466,
+  "eval_steps": 250,
+  "global_step": 7000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004686035613870665,
+      "grad_norm": 0.2652921974658966,
+      "learning_rate": 9.999414245548266e-06,
+      "loss": 3.4545,
+      "step": 10
+    },
+    {
+      "epoch": 0.00937207122774133,
+      "grad_norm": 0.9133270382881165,
+      "learning_rate": 9.998828491096533e-06,
+      "loss": 3.4005,
+      "step": 20
+    },
+    {
+      "epoch": 0.014058106841611996,
+      "grad_norm": 1.7206453084945679,
+      "learning_rate": 9.9982427366448e-06,
+      "loss": 3.2559,
+      "step": 30
+    },
+    {
+      "epoch": 0.01874414245548266,
+      "grad_norm": 1.3775238990783691,
+      "learning_rate": 9.997656982193065e-06,
+      "loss": 3.0152,
+      "step": 40
+    },
+    {
+      "epoch": 0.023430178069353328,
+      "grad_norm": 1.5043880939483643,
+      "learning_rate": 9.997071227741332e-06,
+      "loss": 2.904,
+      "step": 50
+    },
+    {
+      "epoch": 0.028116213683223992,
+      "grad_norm": 1.3289176225662231,
+      "learning_rate": 9.996485473289597e-06,
+      "loss": 2.7989,
+      "step": 60
+    },
+    {
+      "epoch": 0.03280224929709466,
+      "grad_norm": 1.5171725749969482,
+      "learning_rate": 9.995899718837864e-06,
+      "loss": 2.7384,
+      "step": 70
+    },
+    {
+      "epoch": 0.03748828491096532,
+      "grad_norm": 2.6390421390533447,
+      "learning_rate": 9.99531396438613e-06,
+      "loss": 2.6937,
+      "step": 80
+    },
+    {
+      "epoch": 0.04217432052483599,
+      "grad_norm": 2.382201671600342,
+      "learning_rate": 9.994728209934396e-06,
+      "loss": 2.5994,
+      "step": 90
+    },
+    {
+      "epoch": 0.046860356138706656,
+      "grad_norm": 3.3837928771972656,
+      "learning_rate": 9.994142455482663e-06,
+      "loss": 2.5344,
+      "step": 100
+    },
+    {
+      "epoch": 0.05154639175257732,
+      "grad_norm": 3.612391710281372,
+      "learning_rate": 9.993556701030928e-06,
+      "loss": 2.4489,
+      "step": 110
+    },
+    {
+      "epoch": 0.056232427366447985,
+      "grad_norm": 4.58568811416626,
+      "learning_rate": 9.992970946579195e-06,
+      "loss": 2.2835,
+      "step": 120
+    },
+    {
+      "epoch": 0.06091846298031865,
+      "grad_norm": 4.746246814727783,
+      "learning_rate": 9.99238519212746e-06,
+      "loss": 2.2496,
+      "step": 130
+    },
+    {
+      "epoch": 0.06560449859418932,
+      "grad_norm": 3.7526397705078125,
+      "learning_rate": 9.991799437675727e-06,
+      "loss": 2.0891,
+      "step": 140
+    },
+    {
+      "epoch": 0.07029053420805999,
+      "grad_norm": 3.860639810562134,
+      "learning_rate": 9.991213683223994e-06,
+      "loss": 1.988,
+      "step": 150
+    },
+    {
+      "epoch": 0.07497656982193064,
+      "grad_norm": 4.522952079772949,
+      "learning_rate": 9.99062792877226e-06,
+      "loss": 1.7019,
+      "step": 160
+    },
+    {
+      "epoch": 0.07966260543580131,
+      "grad_norm": 4.140243053436279,
+      "learning_rate": 9.990042174320525e-06,
+      "loss": 1.5405,
+      "step": 170
+    },
+    {
+      "epoch": 0.08434864104967198,
+      "grad_norm": 4.114628791809082,
+      "learning_rate": 9.989456419868792e-06,
+      "loss": 1.324,
+      "step": 180
+    },
+    {
+      "epoch": 0.08903467666354264,
+      "grad_norm": 4.97472620010376,
+      "learning_rate": 9.988870665417057e-06,
+      "loss": 1.2381,
+      "step": 190
+    },
+    {
+      "epoch": 0.09372071227741331,
+      "grad_norm": 4.743481636047363,
+      "learning_rate": 9.988284910965324e-06,
+      "loss": 1.1454,
+      "step": 200
+    },
+    {
+      "epoch": 0.09840674789128398,
+      "grad_norm": 4.455767631530762,
+      "learning_rate": 9.98769915651359e-06,
+      "loss": 1.1716,
+      "step": 210
+    },
+    {
+      "epoch": 0.10309278350515463,
+      "grad_norm": 4.199008941650391,
+      "learning_rate": 9.987113402061856e-06,
+      "loss": 1.1001,
+      "step": 220
+    },
+    {
+      "epoch": 0.1077788191190253,
+      "grad_norm": 4.948373794555664,
+      "learning_rate": 9.986527647610123e-06,
+      "loss": 1.0628,
+      "step": 230
+    },
+    {
+      "epoch": 0.11246485473289597,
+      "grad_norm": 4.091549396514893,
+      "learning_rate": 9.985941893158388e-06,
+      "loss": 0.9783,
+      "step": 240
+    },
+    {
+      "epoch": 0.11715089034676664,
+      "grad_norm": 3.7287604808807373,
+      "learning_rate": 9.985356138706655e-06,
+      "loss": 1.0084,
+      "step": 250
+    },
+    {
+      "epoch": 0.11715089034676664,
+      "eval_loss": 0.15787309408187866,
+      "eval_pearson_cosine": 0.6837765056882574,
+      "eval_pearson_dot": 0.3479516218854144,
+      "eval_pearson_euclidean": 0.662064077308699,
+      "eval_pearson_manhattan": 0.6615032406288108,
+      "eval_runtime": 39.6668,
+      "eval_samples_per_second": 37.815,
+      "eval_spearman_cosine": 0.6994144273095011,
+      "eval_spearman_dot": 0.34419950655316756,
+      "eval_spearman_euclidean": 0.6693727884988351,
+      "eval_spearman_manhattan": 0.66929953399832,
+      "eval_steps_per_second": 37.815,
+      "step": 250
+    },
+    {
+      "epoch": 0.1218369259606373,
+      "grad_norm": 4.118359565734863,
+      "learning_rate": 9.984770384254922e-06,
+      "loss": 1.0095,
+      "step": 260
+    },
+    {
+      "epoch": 0.12652296157450796,
+      "grad_norm": 4.312327861785889,
+      "learning_rate": 9.984184629803187e-06,
+      "loss": 0.8829,
+      "step": 270
+    },
+    {
+      "epoch": 0.13120899718837864,
+      "grad_norm": 4.199273586273193,
+      "learning_rate": 9.983598875351454e-06,
+      "loss": 0.9286,
+      "step": 280
+    },
+    {
+      "epoch": 0.1358950328022493,
+      "grad_norm": 4.146886348724365,
+      "learning_rate": 9.98301312089972e-06,
+      "loss": 0.935,
+      "step": 290
+    },
+    {
+      "epoch": 0.14058106841611998,
+      "grad_norm": 4.277599811553955,
+      "learning_rate": 9.982427366447985e-06,
+      "loss": 0.9089,
+      "step": 300
+    },
+    {
+      "epoch": 0.14526710402999063,
+      "grad_norm": 4.278558731079102,
+      "learning_rate": 9.981841611996253e-06,
+      "loss": 0.8515,
+      "step": 310
+    },
+    {
+      "epoch": 0.14995313964386128,
+      "grad_norm": 4.297761917114258,
+      "learning_rate": 9.981255857544518e-06,
+      "loss": 0.9178,
+      "step": 320
+    },
+    {
+      "epoch": 0.15463917525773196,
+      "grad_norm": 4.595494747161865,
+      "learning_rate": 9.980670103092784e-06,
+      "loss": 0.8835,
+      "step": 330
+    },
+    {
+      "epoch": 0.15932521087160262,
+      "grad_norm": 3.567511558532715,
+      "learning_rate": 9.98008434864105e-06,
+      "loss": 0.8519,
+      "step": 340
+    },
+    {
+      "epoch": 0.1640112464854733,
+      "grad_norm": 4.046640872955322,
+      "learning_rate": 9.979498594189316e-06,
+      "loss": 0.8736,
+      "step": 350
+    },
+    {
+      "epoch": 0.16869728209934395,
+      "grad_norm": 4.551926136016846,
+      "learning_rate": 9.978912839737583e-06,
+      "loss": 0.8348,
+      "step": 360
+    },
+    {
+      "epoch": 0.1733833177132146,
+      "grad_norm": 4.367603778839111,
+      "learning_rate": 9.97832708528585e-06,
+      "loss": 0.7652,
+      "step": 370
+    },
+    {
+      "epoch": 0.1780693533270853,
+      "grad_norm": 4.336862564086914,
+      "learning_rate": 9.977741330834115e-06,
+      "loss": 0.8035,
+      "step": 380
+    },
+    {
+      "epoch": 0.18275538894095594,
+      "grad_norm": 3.89697265625,
+      "learning_rate": 9.977155576382382e-06,
+      "loss": 0.7218,
+      "step": 390
+    },
+    {
+      "epoch": 0.18744142455482662,
+      "grad_norm": 4.190406799316406,
+      "learning_rate": 9.976569821930647e-06,
+      "loss": 0.8163,
+      "step": 400
+    },
+    {
+      "epoch": 0.19212746016869728,
+      "grad_norm": 4.282820224761963,
+      "learning_rate": 9.975984067478914e-06,
+      "loss": 0.7987,
+      "step": 410
+    },
+    {
+      "epoch": 0.19681349578256796,
+      "grad_norm": 3.7369320392608643,
+      "learning_rate": 9.97539831302718e-06,
+      "loss": 0.8085,
+      "step": 420
+    },
+    {
+      "epoch": 0.2014995313964386,
+      "grad_norm": 4.697251319885254,
+      "learning_rate": 9.974812558575446e-06,
+      "loss": 0.82,
+      "step": 430
+    },
+    {
+      "epoch": 0.20618556701030927,
+      "grad_norm": 3.3654944896698,
+      "learning_rate": 9.974226804123713e-06,
+      "loss": 0.7689,
+      "step": 440
+    },
+    {
+      "epoch": 0.21087160262417995,
+      "grad_norm": 4.78535270690918,
+      "learning_rate": 9.973641049671978e-06,
+      "loss": 0.704,
+      "step": 450
+    },
+    {
+      "epoch": 0.2155576382380506,
+      "grad_norm": 3.661980628967285,
+      "learning_rate": 9.973055295220243e-06,
+      "loss": 0.7202,
+      "step": 460
+    },
+    {
+      "epoch": 0.22024367385192128,
+      "grad_norm": 3.4837870597839355,
+      "learning_rate": 9.97246954076851e-06,
+      "loss": 0.7008,
+      "step": 470
+    },
+    {
+      "epoch": 0.22492970946579194,
+      "grad_norm": 4.407717704772949,
+      "learning_rate": 9.971883786316777e-06,
+      "loss": 0.7182,
+      "step": 480
+    },
+    {
+      "epoch": 0.2296157450796626,
+      "grad_norm": 3.2862961292266846,
+      "learning_rate": 9.971298031865042e-06,
+      "loss": 0.7351,
+      "step": 490
+    },
+    {
+      "epoch": 0.23430178069353327,
+      "grad_norm": 3.018371343612671,
+      "learning_rate": 9.97071227741331e-06,
+      "loss": 0.7072,
+      "step": 500
+    },
+    {
+      "epoch": 0.23430178069353327,
+      "eval_loss": 0.13636364042758942,
+      "eval_pearson_cosine": 0.7226338282187044,
+      "eval_pearson_dot": 0.40017542467791145,
+      "eval_pearson_euclidean": 0.7214409735618545,
+      "eval_pearson_manhattan": 0.7207106754958978,
+      "eval_runtime": 38.9491,
+      "eval_samples_per_second": 38.512,
+      "eval_spearman_cosine": 0.737511621673308,
+      "eval_spearman_dot": 0.39096188952446115,
+      "eval_spearman_euclidean": 0.7270501587145487,
+      "eval_spearman_manhattan": 0.7263223218978012,
+      "eval_steps_per_second": 38.512,
+      "step": 500
+    },
+    {
+      "epoch": 0.23898781630740393,
+      "grad_norm": 4.259694576263428,
+      "learning_rate": 9.970126522961575e-06,
+      "loss": 0.7365,
+      "step": 510
+    },
+    {
+      "epoch": 0.2436738519212746,
+      "grad_norm": 4.129746437072754,
+      "learning_rate": 9.969540768509842e-06,
+      "loss": 0.7005,
+      "step": 520
+    },
+    {
+      "epoch": 0.24835988753514526,
+      "grad_norm": 3.5990495681762695,
+      "learning_rate": 9.968955014058108e-06,
+      "loss": 0.6892,
+      "step": 530
+    },
+    {
+      "epoch": 0.2530459231490159,
+      "grad_norm": 4.118338584899902,
+      "learning_rate": 9.968369259606374e-06,
+      "loss": 0.7235,
+      "step": 540
+    },
+    {
+      "epoch": 0.25773195876288657,
+      "grad_norm": 3.2775630950927734,
+      "learning_rate": 9.96778350515464e-06,
+      "loss": 0.6567,
+      "step": 550
+    },
+    {
+      "epoch": 0.2624179943767573,
+      "grad_norm": 3.7363176345825195,
+      "learning_rate": 9.967197750702906e-06,
+      "loss": 0.6724,
+      "step": 560
+    },
+    {
+      "epoch": 0.26710402999062793,
+      "grad_norm": 3.317964792251587,
+      "learning_rate": 9.966611996251173e-06,
+      "loss": 0.6603,
+      "step": 570
+    },
+    {
+      "epoch": 0.2717900656044986,
+      "grad_norm": 4.1027703285217285,
+      "learning_rate": 9.966026241799438e-06,
+      "loss": 0.6384,
+      "step": 580
+    },
+    {
+      "epoch": 0.27647610121836924,
+      "grad_norm": 3.434147357940674,
+      "learning_rate": 9.965440487347705e-06,
+      "loss": 0.6555,
+      "step": 590
+    },
+    {
+      "epoch": 0.28116213683223995,
+      "grad_norm": 3.4407315254211426,
+      "learning_rate": 9.964854732895972e-06,
+      "loss": 0.6695,
+      "step": 600
+    },
+    {
+      "epoch": 0.2858481724461106,
+      "grad_norm": 3.800232172012329,
+      "learning_rate": 9.964268978444237e-06,
+      "loss": 0.6727,
+      "step": 610
+    },
+    {
+      "epoch": 0.29053420805998126,
+      "grad_norm": 3.191505193710327,
+      "learning_rate": 9.963683223992502e-06,
+      "loss": 0.6513,
+      "step": 620
+    },
+    {
+      "epoch": 0.2952202436738519,
+      "grad_norm": 4.152797222137451,
+      "learning_rate": 9.963097469540769e-06,
+      "loss": 0.7115,
+      "step": 630
+    },
+    {
+      "epoch": 0.29990627928772257,
+      "grad_norm": 3.339895248413086,
+      "learning_rate": 9.962511715089036e-06,
+      "loss": 0.6515,
+      "step": 640
+    },
+    {
+      "epoch": 0.3045923149015933,
+      "grad_norm": 3.2941200733184814,
+      "learning_rate": 9.961925960637301e-06,
+      "loss": 0.6024,
+      "step": 650
+    },
+    {
+      "epoch": 0.30927835051546393,
+      "grad_norm": 3.0853567123413086,
+      "learning_rate": 9.961340206185568e-06,
+      "loss": 0.6179,
+      "step": 660
+    },
+    {
+      "epoch": 0.3139643861293346,
+      "grad_norm": 3.827177047729492,
+      "learning_rate": 9.960754451733833e-06,
+      "loss": 0.6678,
+      "step": 670
+    },
+    {
+      "epoch": 0.31865042174320524,
+      "grad_norm": 3.8214621543884277,
+      "learning_rate": 9.9601686972821e-06,
+      "loss": 0.6248,
+      "step": 680
+    },
+    {
+      "epoch": 0.3233364573570759,
+      "grad_norm": 4.431512832641602,
+      "learning_rate": 9.959582942830366e-06,
+      "loss": 0.6562,
+      "step": 690
+    },
+    {
+      "epoch": 0.3280224929709466,
+      "grad_norm": 3.046006679534912,
+      "learning_rate": 9.958997188378632e-06,
+      "loss": 0.5865,
+      "step": 700
+    },
+    {
+      "epoch": 0.33270852858481725,
+      "grad_norm": 3.6602745056152344,
+      "learning_rate": 9.9584114339269e-06,
+      "loss": 0.611,
+      "step": 710
+    },
+    {
+      "epoch": 0.3373945641986879,
+      "grad_norm": 3.815204381942749,
+      "learning_rate": 9.957825679475165e-06,
+      "loss": 0.619,
+      "step": 720
+    },
+    {
+      "epoch": 0.34208059981255856,
+      "grad_norm": 3.5464768409729004,
+      "learning_rate": 9.957239925023431e-06,
+      "loss": 0.563,
+      "step": 730
+    },
+    {
+      "epoch": 0.3467666354264292,
+      "grad_norm": 3.254077911376953,
+      "learning_rate": 9.956654170571697e-06,
+      "loss": 0.6541,
+      "step": 740
+    },
+    {
+      "epoch": 0.3514526710402999,
+      "grad_norm": 3.465792655944824,
+      "learning_rate": 9.956068416119962e-06,
+      "loss": 0.6207,
+      "step": 750
+    },
+    {
+      "epoch": 0.3514526710402999,
+      "eval_loss": 0.11937286704778671,
+      "eval_pearson_cosine": 0.7370582778898296,
+      "eval_pearson_dot": 0.4517357649062319,
+      "eval_pearson_euclidean": 0.7300011163349307,
+      "eval_pearson_manhattan": 0.7294921991619532,
+      "eval_runtime": 40.2019,
+      "eval_samples_per_second": 37.312,
+      "eval_spearman_cosine": 0.7509292468775599,
+      "eval_spearman_dot": 0.44621372678317905,
+      "eval_spearman_euclidean": 0.7400885968162902,
+      "eval_spearman_manhattan": 0.7397623919578791,
+      "eval_steps_per_second": 37.312,
+      "step": 750
+    },
+    {
+      "epoch": 0.3561387066541706,
+      "grad_norm": 3.754749298095703,
+      "learning_rate": 9.95548266166823e-06,
+      "loss": 0.5565,
+      "step": 760
+    },
+    {
+      "epoch": 0.36082474226804123,
+      "grad_norm": 4.001673698425293,
+      "learning_rate": 9.954896907216496e-06,
+      "loss": 0.6017,
+      "step": 770
+    },
+    {
+      "epoch": 0.3655107778819119,
+      "grad_norm": 3.2223927974700928,
+      "learning_rate": 9.954311152764761e-06,
+      "loss": 0.6053,
+      "step": 780
+    },
+    {
+      "epoch": 0.3701968134957826,
+      "grad_norm": 3.135162353515625,
+      "learning_rate": 9.953725398313028e-06,
+      "loss": 0.5703,
+      "step": 790
+    },
+    {
+      "epoch": 0.37488284910965325,
+      "grad_norm": 3.6593549251556396,
+      "learning_rate": 9.953139643861293e-06,
+      "loss": 0.5683,
+      "step": 800
+    },
+    {
+      "epoch": 0.3795688847235239,
+      "grad_norm": 3.2421703338623047,
+      "learning_rate": 9.95255388940956e-06,
+      "loss": 0.6306,
+      "step": 810
+    },
+    {
+      "epoch": 0.38425492033739456,
+      "grad_norm": 3.2221267223358154,
+      "learning_rate": 9.951968134957827e-06,
+      "loss": 0.5942,
+      "step": 820
+    },
+    {
+      "epoch": 0.3889409559512652,
+      "grad_norm": 3.55708646774292,
+      "learning_rate": 9.951382380506092e-06,
+      "loss": 0.5695,
+      "step": 830
+    },
+    {
+      "epoch": 0.3936269915651359,
+      "grad_norm": 3.310382843017578,
+      "learning_rate": 9.950796626054359e-06,
+      "loss": 0.6249,
+      "step": 840
+    },
+    {
+      "epoch": 0.3983130271790066,
+      "grad_norm": 3.126347303390503,
+      "learning_rate": 9.950210871602624e-06,
+      "loss": 0.5515,
+      "step": 850
+    },
+    {
+      "epoch": 0.4029990627928772,
+      "grad_norm": 3.691459894180298,
+      "learning_rate": 9.949625117150891e-06,
+      "loss": 0.6074,
+      "step": 860
+    },
+    {
+      "epoch": 0.4076850984067479,
+      "grad_norm": 3.8297834396362305,
+      "learning_rate": 9.949039362699158e-06,
+      "loss": 0.6538,
+      "step": 870
+    },
+    {
+      "epoch": 0.41237113402061853,
+      "grad_norm": 3.097637414932251,
+      "learning_rate": 9.948453608247423e-06,
+      "loss": 0.5832,
+      "step": 880
+    },
+    {
+      "epoch": 0.41705716963448924,
+      "grad_norm": 4.003639221191406,
+      "learning_rate": 9.94786785379569e-06,
+      "loss": 0.5407,
+      "step": 890
+    },
+    {
+      "epoch": 0.4217432052483599,
+      "grad_norm": 3.648155689239502,
+      "learning_rate": 9.947282099343956e-06,
+      "loss": 0.5298,
+      "step": 900
+    },
+    {
+      "epoch": 0.42642924086223055,
+      "grad_norm": 3.071843147277832,
+      "learning_rate": 9.94669634489222e-06,
+      "loss": 0.4969,
+      "step": 910
+    },
+    {
+      "epoch": 0.4311152764761012,
+      "grad_norm": 3.8248066902160645,
+      "learning_rate": 9.94611059044049e-06,
+      "loss": 0.5894,
+      "step": 920
+    },
+    {
+      "epoch": 0.43580131208997186,
+      "grad_norm": 3.3905293941497803,
+      "learning_rate": 9.945524835988755e-06,
+      "loss": 0.5924,
+      "step": 930
+    },
+    {
+      "epoch": 0.44048734770384257,
+      "grad_norm": 3.8499722480773926,
+      "learning_rate": 9.94493908153702e-06,
+      "loss": 0.5636,
+      "step": 940
+    },
+    {
+      "epoch": 0.4451733833177132,
+      "grad_norm": 3.613633155822754,
+      "learning_rate": 9.944353327085287e-06,
+      "loss": 0.5631,
+      "step": 950
+    },
+    {
+      "epoch": 0.4498594189315839,
+      "grad_norm": 2.895521879196167,
+      "learning_rate": 9.943767572633552e-06,
+      "loss": 0.5791,
+      "step": 960
+    },
+    {
+      "epoch": 0.45454545454545453,
+      "grad_norm": 3.783475160598755,
+      "learning_rate": 9.943181818181819e-06,
+      "loss": 0.583,
+      "step": 970
+    },
+    {
+      "epoch": 0.4592314901593252,
+      "grad_norm": 3.0767059326171875,
+      "learning_rate": 9.942596063730086e-06,
+      "loss": 0.5625,
+      "step": 980
+    },
+    {
+      "epoch": 0.4639175257731959,
+      "grad_norm": 3.0430691242218018,
+      "learning_rate": 9.942010309278351e-06,
+      "loss": 0.6107,
+      "step": 990
+    },
+    {
+      "epoch": 0.46860356138706655,
+      "grad_norm": 3.4637937545776367,
+      "learning_rate": 9.941424554826618e-06,
+      "loss": 0.5767,
+      "step": 1000
+    },
+    {
+      "epoch": 0.46860356138706655,
+      "eval_loss": 0.11473686993122101,
+      "eval_pearson_cosine": 0.7508291149999735,
+      "eval_pearson_dot": 0.5170270771997139,
+      "eval_pearson_euclidean": 0.7400425527843311,
+      "eval_pearson_manhattan": 0.7394957204847508,
+      "eval_runtime": 40.3307,
+      "eval_samples_per_second": 37.192,
+      "eval_spearman_cosine": 0.7636438911546342,
+      "eval_spearman_dot": 0.5181197481922067,
+      "eval_spearman_euclidean": 0.7511102060810355,
+      "eval_spearman_manhattan": 0.750245305109835,
+      "eval_steps_per_second": 37.192,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4732895970009372,
+      "grad_norm": 3.62758207321167,
+      "learning_rate": 9.940838800374883e-06,
+      "loss": 0.5533,
+      "step": 1010
+    },
+    {
+      "epoch": 0.47797563261480785,
+      "grad_norm": 2.8225605487823486,
+      "learning_rate": 9.94025304592315e-06,
+      "loss": 0.562,
+      "step": 1020
+    },
+    {
+      "epoch": 0.48266166822867856,
+      "grad_norm": 3.6451101303100586,
+      "learning_rate": 9.939667291471417e-06,
+      "loss": 0.5854,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4873477038425492,
+      "grad_norm": 3.448962450027466,
+      "learning_rate": 9.939081537019682e-06,
+      "loss": 0.6032,
+      "step": 1040
+    },
+    {
+      "epoch": 0.49203373945641987,
+      "grad_norm": 3.0601136684417725,
+      "learning_rate": 9.938495782567949e-06,
+      "loss": 0.5636,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4967197750702905,
+      "grad_norm": 3.154737710952759,
+      "learning_rate": 9.937910028116214e-06,
+      "loss": 0.5563,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5014058106841612,
+      "grad_norm": 3.4425132274627686,
+      "learning_rate": 9.93732427366448e-06,
+      "loss": 0.5956,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5060918462980318,
+      "grad_norm": 3.6232352256774902,
+      "learning_rate": 9.936738519212746e-06,
+      "loss": 0.5745,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5107778819119025,
+      "grad_norm": 3.701143980026245,
+      "learning_rate": 9.936152764761013e-06,
+      "loss": 0.583,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5154639175257731,
+      "grad_norm": 3.6037821769714355,
+      "learning_rate": 9.935567010309279e-06,
+      "loss": 0.5367,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5201499531396439,
+      "grad_norm": 3.333034038543701,
+      "learning_rate": 9.934981255857546e-06,
+      "loss": 0.5675,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5248359887535146,
+      "grad_norm": 3.335798501968384,
+      "learning_rate": 9.93439550140581e-06,
+      "loss": 0.582,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5295220243673852,
+      "grad_norm": 2.90682315826416,
+      "learning_rate": 9.933809746954078e-06,
+      "loss": 0.5866,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5342080599812559,
+      "grad_norm": 3.268754482269287,
+      "learning_rate": 9.933223992502345e-06,
+      "loss": 0.5832,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5388940955951266,
+      "grad_norm": 3.0026419162750244,
+      "learning_rate": 9.93263823805061e-06,
+      "loss": 0.5303,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5435801312089972,
+      "grad_norm": 3.3386096954345703,
+      "learning_rate": 9.932052483598877e-06,
+      "loss": 0.525,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5482661668228679,
+      "grad_norm": 4.069448947906494,
+      "learning_rate": 9.931466729147142e-06,
+      "loss": 0.5738,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5529522024367385,
+      "grad_norm": 2.9254677295684814,
+      "learning_rate": 9.930880974695409e-06,
+      "loss": 0.5529,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5576382380506092,
+      "grad_norm": 3.216367721557617,
+      "learning_rate": 9.930295220243674e-06,
+      "loss": 0.5646,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5623242736644799,
+      "grad_norm": 3.7063798904418945,
+      "learning_rate": 9.929709465791941e-06,
+      "loss": 0.5793,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5670103092783505,
+      "grad_norm": 3.1322641372680664,
+      "learning_rate": 9.929123711340208e-06,
+      "loss": 0.5415,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5716963448922212,
+      "grad_norm": 2.747403621673584,
+      "learning_rate": 9.928537956888473e-06,
+      "loss": 0.5468,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5763823805060918,
+      "grad_norm": 3.0426528453826904,
+      "learning_rate": 9.927952202436738e-06,
+      "loss": 0.5755,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5810684161199625,
+      "grad_norm": 2.9256389141082764,
+      "learning_rate": 9.927366447985005e-06,
+      "loss": 0.5237,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5857544517338332,
+      "grad_norm": 3.5383565425872803,
+      "learning_rate": 9.92678069353327e-06,
+      "loss": 0.5026,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5857544517338332,
+      "eval_loss": 0.10466309636831284,
+      "eval_pearson_cosine": 0.7506514546227478,
+      "eval_pearson_dot": 0.5487479231049726,
+      "eval_pearson_euclidean": 0.7459287854123886,
+      "eval_pearson_manhattan": 0.7455307169513361,
+      "eval_runtime": 39.6161,
+      "eval_samples_per_second": 37.863,
+      "eval_spearman_cosine": 0.7634978803745991,
+      "eval_spearman_dot": 0.5530948044816919,
+      "eval_spearman_euclidean": 0.7564382757079575,
+      "eval_spearman_manhattan": 0.7557623143729565,
+      "eval_steps_per_second": 37.863,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5904404873477038,
+      "grad_norm": 4.084426403045654,
+      "learning_rate": 9.926194939081537e-06,
+      "loss": 0.5576,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5951265229615745,
+      "grad_norm": 3.5379457473754883,
+      "learning_rate": 9.925609184629804e-06,
+      "loss": 0.547,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5998125585754451,
+      "grad_norm": 2.8720273971557617,
+      "learning_rate": 9.92502343017807e-06,
+      "loss": 0.5385,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6044985941893158,
+      "grad_norm": 2.877453565597534,
+      "learning_rate": 9.924437675726336e-06,
+      "loss": 0.5114,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6091846298031866,
+      "grad_norm": 3.771329164505005,
+      "learning_rate": 9.923851921274602e-06,
+      "loss": 0.5133,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6138706654170571,
+      "grad_norm": 2.411574125289917,
+      "learning_rate": 9.923266166822869e-06,
+      "loss": 0.5207,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6185567010309279,
+      "grad_norm": 3.733808755874634,
+      "learning_rate": 9.922680412371136e-06,
+      "loss": 0.5043,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6232427366447985,
+      "grad_norm": 2.5523786544799805,
+      "learning_rate": 9.9220946579194e-06,
+      "loss": 0.5142,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6279287722586692,
+      "grad_norm": 2.743069887161255,
+      "learning_rate": 9.921508903467668e-06,
+      "loss": 0.5184,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6326148078725399,
+      "grad_norm": 3.086749792098999,
+      "learning_rate": 9.920923149015933e-06,
+      "loss": 0.5135,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6373008434864105,
+      "grad_norm": 3.3106496334075928,
+      "learning_rate": 9.9203373945642e-06,
+      "loss": 0.5107,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6419868791002812,
+      "grad_norm": 3.2946176528930664,
+      "learning_rate": 9.919751640112467e-06,
+      "loss": 0.5112,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6466729147141518,
+      "grad_norm": 3.7036850452423096,
+      "learning_rate": 9.919165885660732e-06,
+      "loss": 0.5112,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6513589503280225,
+      "grad_norm": 3.8145229816436768,
+      "learning_rate": 9.918580131208997e-06,
+      "loss": 0.555,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6560449859418932,
+      "grad_norm": 3.3321220874786377,
+      "learning_rate": 9.917994376757264e-06,
+      "loss": 0.4917,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6607310215557638,
+      "grad_norm": 2.517955780029297,
+      "learning_rate": 9.91740862230553e-06,
+      "loss": 0.527,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6654170571696345,
+      "grad_norm": 3.0029234886169434,
+      "learning_rate": 9.916822867853796e-06,
+      "loss": 0.5192,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6701030927835051,
+      "grad_norm": 2.7767951488494873,
+      "learning_rate": 9.916237113402063e-06,
+      "loss": 0.5526,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6747891283973758,
+      "grad_norm": 3.2969696521759033,
+      "learning_rate": 9.915651358950328e-06,
+      "loss": 0.5444,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6794751640112465,
+      "grad_norm": 3.046128034591675,
+      "learning_rate": 9.915065604498595e-06,
+      "loss": 0.479,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6841611996251171,
+      "grad_norm": 3.239370822906494,
+      "learning_rate": 9.91447985004686e-06,
+      "loss": 0.4626,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6888472352389878,
+      "grad_norm": 3.3885092735290527,
+      "learning_rate": 9.913894095595127e-06,
+      "loss": 0.5115,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6935332708528584,
+      "grad_norm": 3.279419422149658,
+      "learning_rate": 9.913308341143394e-06,
+      "loss": 0.5054,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6982193064667291,
+      "grad_norm": 3.451601028442383,
+      "learning_rate": 9.91272258669166e-06,
+      "loss": 0.5124,
+      "step": 1490
+    },
+    {
+      "epoch": 0.7029053420805998,
+      "grad_norm": 2.9452903270721436,
+      "learning_rate": 9.912136832239926e-06,
+      "loss": 0.5192,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7029053420805998,
+      "eval_loss": 0.11661049723625183,
+      "eval_pearson_cosine": 0.7522095786209633,
+      "eval_pearson_dot": 0.5055151019348045,
+      "eval_pearson_euclidean": 0.7488737825239156,
+      "eval_pearson_manhattan": 0.7487340190517244,
+      "eval_runtime": 39.6015,
+      "eval_samples_per_second": 37.877,
+      "eval_spearman_cosine": 0.7673251411045983,
+      "eval_spearman_dot": 0.5052800577910852,
+      "eval_spearman_euclidean": 0.7594003970324229,
+      "eval_spearman_manhattan": 0.7590967348946761,
+      "eval_steps_per_second": 37.877,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7075913776944704,
+      "grad_norm": 3.0128912925720215,
+      "learning_rate": 9.911551077788192e-06,
+      "loss": 0.4896,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7122774133083412,
+      "grad_norm": 2.955695152282715,
+      "learning_rate": 9.910965323336459e-06,
+      "loss": 0.5161,
+      "step": 1520
+    },
+    {
+      "epoch": 0.7169634489222118,
+      "grad_norm": 2.8971164226531982,
+      "learning_rate": 9.910379568884726e-06,
+      "loss": 0.4868,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7216494845360825,
+      "grad_norm": 2.9848475456237793,
+      "learning_rate": 9.90979381443299e-06,
+      "loss": 0.4919,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7263355201499532,
+      "grad_norm": 3.1218748092651367,
+      "learning_rate": 9.909208059981256e-06,
+      "loss": 0.4954,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7310215557638238,
+      "grad_norm": 3.0474631786346436,
+      "learning_rate": 9.908622305529523e-06,
+      "loss": 0.5007,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7357075913776945,
+      "grad_norm": 2.8136134147644043,
+      "learning_rate": 9.908036551077788e-06,
+      "loss": 0.4998,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7403936269915652,
+      "grad_norm": 2.537595272064209,
+      "learning_rate": 9.907450796626055e-06,
+      "loss": 0.4759,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7450796626054358,
+      "grad_norm": 3.2701878547668457,
+      "learning_rate": 9.906865042174322e-06,
+      "loss": 0.5106,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7497656982193065,
+      "grad_norm": 3.289370536804199,
+      "learning_rate": 9.906279287722587e-06,
+      "loss": 0.513,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7544517338331771,
+      "grad_norm": 3.6762607097625732,
+      "learning_rate": 9.905693533270854e-06,
+      "loss": 0.4854,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7591377694470478,
+      "grad_norm": 3.083775281906128,
+      "learning_rate": 9.90510777881912e-06,
+      "loss": 0.4831,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7638238050609185,
+      "grad_norm": 3.9659175872802734,
+      "learning_rate": 9.904522024367386e-06,
+      "loss": 0.4634,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7685098406747891,
+      "grad_norm": 3.014983654022217,
+      "learning_rate": 9.903936269915653e-06,
+      "loss": 0.4979,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7731958762886598,
+      "grad_norm": 2.3927905559539795,
+      "learning_rate": 9.903350515463918e-06,
+      "loss": 0.528,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7778819119025304,
+      "grad_norm": 3.1232478618621826,
+      "learning_rate": 9.902764761012185e-06,
+      "loss": 0.5355,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7825679475164011,
+      "grad_norm": 3.5926144123077393,
+      "learning_rate": 9.90217900656045e-06,
+      "loss": 0.4849,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7872539831302718,
+      "grad_norm": 3.9513533115386963,
+      "learning_rate": 9.901593252108717e-06,
+      "loss": 0.505,
+      "step": 1680
+    },
+    {
+      "epoch": 0.7919400187441424,
+      "grad_norm": 2.9392502307891846,
+      "learning_rate": 9.901007497656983e-06,
+      "loss": 0.5358,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7966260543580131,
+      "grad_norm": 2.7503395080566406,
+      "learning_rate": 9.90042174320525e-06,
+      "loss": 0.5267,
+      "step": 1700
+    },
+    {
+      "epoch": 0.8013120899718837,
+      "grad_norm": 2.9122848510742188,
+      "learning_rate": 9.899835988753515e-06,
+      "loss": 0.508,
+      "step": 1710
+    },
+    {
+      "epoch": 0.8059981255857545,
+      "grad_norm": 3.3603107929229736,
+      "learning_rate": 9.899250234301782e-06,
+      "loss": 0.4904,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8106841611996252,
+      "grad_norm": 2.9802725315093994,
+      "learning_rate": 9.898664479850047e-06,
+      "loss": 0.5019,
+      "step": 1730
+    },
+    {
+      "epoch": 0.8153701968134958,
+      "grad_norm": 3.4380977153778076,
+      "learning_rate": 9.898078725398314e-06,
+      "loss": 0.4894,
+      "step": 1740
+    },
+    {
+      "epoch": 0.8200562324273665,
+      "grad_norm": 3.106285572052002,
+      "learning_rate": 9.897492970946579e-06,
+      "loss": 0.5046,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8200562324273665,
+      "eval_loss": 0.11102449893951416,
+      "eval_pearson_cosine": 0.7554957932917077,
+      "eval_pearson_dot": 0.5302646386582808,
+      "eval_pearson_euclidean": 0.7581196621481752,
+      "eval_pearson_manhattan": 0.7582295870999474,
+      "eval_runtime": 39.7307,
+      "eval_samples_per_second": 37.754,
+      "eval_spearman_cosine": 0.7675167675754492,
+      "eval_spearman_dot": 0.5391446720291554,
+      "eval_spearman_euclidean": 0.767156398991665,
+      "eval_spearman_manhattan": 0.7675093547587875,
+      "eval_steps_per_second": 37.754,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8247422680412371,
+      "grad_norm": 3.348632574081421,
+      "learning_rate": 9.896907216494846e-06,
+      "loss": 0.4758,
+      "step": 1760
+    },
+    {
+      "epoch": 0.8294283036551078,
+      "grad_norm": 3.240816831588745,
+      "learning_rate": 9.896321462043113e-06,
+      "loss": 0.4631,
+      "step": 1770
+    },
+    {
+      "epoch": 0.8341143392689785,
+      "grad_norm": 3.1542975902557373,
+      "learning_rate": 9.895735707591378e-06,
+      "loss": 0.492,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8388003748828491,
+      "grad_norm": 2.405806064605713,
+      "learning_rate": 9.895149953139645e-06,
+      "loss": 0.5083,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8434864104967198,
+      "grad_norm": 3.3465261459350586,
+      "learning_rate": 9.89456419868791e-06,
+      "loss": 0.4179,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8481724461105904,
+      "grad_norm": 3.382632255554199,
+      "learning_rate": 9.893978444236177e-06,
+      "loss": 0.4982,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8528584817244611,
+      "grad_norm": 2.5863983631134033,
+      "learning_rate": 9.893392689784444e-06,
+      "loss": 0.5203,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8575445173383318,
+      "grad_norm": 2.830101251602173,
+      "learning_rate": 9.89280693533271e-06,
+      "loss": 0.4997,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8622305529522024,
+      "grad_norm": 3.096226453781128,
+      "learning_rate": 9.892221180880976e-06,
+      "loss": 0.493,
+      "step": 1840
+    },
+    {
+      "epoch": 0.8669165885660731,
+      "grad_norm": 3.1559505462646484,
+      "learning_rate": 9.891635426429241e-06,
+      "loss": 0.4968,
+      "step": 1850
+    },
+    {
+      "epoch": 0.8716026241799437,
+      "grad_norm": 3.1153781414031982,
+      "learning_rate": 9.891049671977507e-06,
+      "loss": 0.5037,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8762886597938144,
+      "grad_norm": 3.222080945968628,
+      "learning_rate": 9.890463917525774e-06,
+      "loss": 0.538,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8809746954076851,
+      "grad_norm": 2.716965436935425,
+      "learning_rate": 9.88987816307404e-06,
+      "loss": 0.4525,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8856607310215557,
+      "grad_norm": 3.6638436317443848,
+      "learning_rate": 9.889292408622306e-06,
+      "loss": 0.4959,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8903467666354264,
+      "grad_norm": 2.795483350753784,
+      "learning_rate": 9.888706654170573e-06,
+      "loss": 0.5156,
+      "step": 1900
+    },
+    {
+      "epoch": 0.895032802249297,
+      "grad_norm": 3.4826247692108154,
+      "learning_rate": 9.888120899718838e-06,
+      "loss": 0.5136,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8997188378631678,
+      "grad_norm": 3.1624643802642822,
+      "learning_rate": 9.887535145267105e-06,
+      "loss": 0.4488,
+      "step": 1920
+    },
+    {
+      "epoch": 0.9044048734770385,
+      "grad_norm": 3.1892471313476562,
+      "learning_rate": 9.886949390815372e-06,
+      "loss": 0.4611,
+      "step": 1930
+    },
+    {
+      "epoch": 0.9090909090909091,
+      "grad_norm": 3.075989007949829,
+      "learning_rate": 9.886363636363637e-06,
+      "loss": 0.4376,
+      "step": 1940
+    },
+    {
+      "epoch": 0.9137769447047798,
+      "grad_norm": 3.629077672958374,
+      "learning_rate": 9.885777881911904e-06,
+      "loss": 0.5039,
+      "step": 1950
+    },
+    {
+      "epoch": 0.9184629803186504,
+      "grad_norm": 3.424208402633667,
+      "learning_rate": 9.885192127460169e-06,
+      "loss": 0.5169,
+      "step": 1960
+    },
+    {
+      "epoch": 0.9231490159325211,
+      "grad_norm": 2.9203758239746094,
+      "learning_rate": 9.884606373008436e-06,
+      "loss": 0.4944,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9278350515463918,
+      "grad_norm": 3.0997154712677,
+      "learning_rate": 9.884020618556703e-06,
+      "loss": 0.4586,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9325210871602624,
+      "grad_norm": 2.3105781078338623,
+      "learning_rate": 9.883434864104968e-06,
+      "loss": 0.5109,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9372071227741331,
+      "grad_norm": 3.348557472229004,
+      "learning_rate": 9.882849109653235e-06,
+      "loss": 0.5055,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9372071227741331,
+      "eval_loss": 0.10619169473648071,
+      "eval_pearson_cosine": 0.7545903634746969,
+      "eval_pearson_dot": 0.5637526434800861,
+      "eval_pearson_euclidean": 0.7502010625678288,
+      "eval_pearson_manhattan": 0.7501144315992683,
+      "eval_runtime": 39.4727,
+      "eval_samples_per_second": 38.001,
+      "eval_spearman_cosine": 0.7726348299355251,
+      "eval_spearman_dot": 0.571016162108862,
+      "eval_spearman_euclidean": 0.7650905718991688,
+      "eval_spearman_manhattan": 0.7650111430170369,
+      "eval_steps_per_second": 38.001,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9418931583880038,
+      "grad_norm": 2.4951772689819336,
+      "learning_rate": 9.8822633552015e-06,
+      "loss": 0.4542,
+      "step": 2010
+    },
+    {
+      "epoch": 0.9465791940018744,
+      "grad_norm": 2.8216359615325928,
+      "learning_rate": 9.881677600749765e-06,
+      "loss": 0.4823,
+      "step": 2020
+    },
+    {
+      "epoch": 0.9512652296157451,
+      "grad_norm": 3.510946035385132,
+      "learning_rate": 9.881091846298032e-06,
+      "loss": 0.464,
+      "step": 2030
+    },
+    {
+      "epoch": 0.9559512652296157,
+      "grad_norm": 3.307612419128418,
+      "learning_rate": 9.8805060918463e-06,
+      "loss": 0.4641,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9606373008434864,
+      "grad_norm": 3.4705116748809814,
+      "learning_rate": 9.879920337394564e-06,
+      "loss": 0.4643,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9653233364573571,
+      "grad_norm": 3.5456154346466064,
+      "learning_rate": 9.879334582942831e-06,
+      "loss": 0.4673,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9700093720712277,
+      "grad_norm": 3.027689218521118,
+      "learning_rate": 9.878748828491097e-06,
+      "loss": 0.4997,
+      "step": 2070
+    },
+    {
+      "epoch": 0.9746954076850984,
+      "grad_norm": 3.0385146141052246,
+      "learning_rate": 9.878163074039364e-06,
+      "loss": 0.4941,
+      "step": 2080
+    },
+    {
+      "epoch": 0.979381443298969,
+      "grad_norm": 2.4981963634490967,
+      "learning_rate": 9.87757731958763e-06,
+      "loss": 0.4553,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9840674789128397,
+      "grad_norm": 2.5605556964874268,
+      "learning_rate": 9.876991565135896e-06,
+      "loss": 0.4515,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9887535145267105,
+      "grad_norm": 3.263235330581665,
+      "learning_rate": 9.876405810684163e-06,
+      "loss": 0.5033,
+      "step": 2110
+    },
+    {
+      "epoch": 0.993439550140581,
+      "grad_norm": 3.3904616832733154,
+      "learning_rate": 9.875820056232428e-06,
+      "loss": 0.4285,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9981255857544518,
+      "grad_norm": 3.5745186805725098,
+      "learning_rate": 9.875234301780695e-06,
+      "loss": 0.4793,
+      "step": 2130
+    },
+    {
+      "epoch": 1.0028116213683225,
+      "grad_norm": 2.888436794281006,
+      "learning_rate": 9.874648547328962e-06,
+      "loss": 0.4291,
+      "step": 2140
+    },
+    {
+      "epoch": 1.007497656982193,
+      "grad_norm": 2.378056526184082,
+      "learning_rate": 9.874062792877227e-06,
+      "loss": 0.3839,
+      "step": 2150
+    },
+    {
+      "epoch": 1.0121836925960637,
+      "grad_norm": 3.4162449836730957,
+      "learning_rate": 9.873477038425494e-06,
+      "loss": 0.3953,
+      "step": 2160
+    },
+    {
+      "epoch": 1.0168697282099344,
+      "grad_norm": 3.428281545639038,
+      "learning_rate": 9.872891283973759e-06,
+      "loss": 0.3927,
+      "step": 2170
+    },
+    {
+      "epoch": 1.021555763823805,
+      "grad_norm": 2.3197989463806152,
+      "learning_rate": 9.872305529522024e-06,
+      "loss": 0.3879,
+      "step": 2180
+    },
+    {
+      "epoch": 1.0262417994376758,
+      "grad_norm": 2.783487558364868,
+      "learning_rate": 9.871719775070291e-06,
+      "loss": 0.3992,
+      "step": 2190
+    },
+    {
+      "epoch": 1.0309278350515463,
+      "grad_norm": 3.5884554386138916,
+      "learning_rate": 9.871134020618558e-06,
+      "loss": 0.3921,
+      "step": 2200
+    },
+    {
+      "epoch": 1.035613870665417,
+      "grad_norm": 3.041579008102417,
+      "learning_rate": 9.870548266166823e-06,
+      "loss": 0.4212,
+      "step": 2210
+    },
+    {
+      "epoch": 1.0402999062792877,
+      "grad_norm": 3.096118688583374,
+      "learning_rate": 9.86996251171509e-06,
+      "loss": 0.388,
+      "step": 2220
+    },
+    {
+      "epoch": 1.0449859418931584,
+      "grad_norm": 3.3495426177978516,
+      "learning_rate": 9.869376757263355e-06,
+      "loss": 0.4283,
+      "step": 2230
+    },
+    {
+      "epoch": 1.0496719775070291,
+      "grad_norm": 3.1030259132385254,
+      "learning_rate": 9.868791002811622e-06,
+      "loss": 0.3841,
+      "step": 2240
+    },
+    {
+      "epoch": 1.0543580131208996,
+      "grad_norm": 2.785508155822754,
+      "learning_rate": 9.868205248359888e-06,
+      "loss": 0.4177,
+      "step": 2250
+    },
+    {
+      "epoch": 1.0543580131208996,
+      "eval_loss": 0.09417638927698135,
+      "eval_pearson_cosine": 0.7576901427376015,
+      "eval_pearson_dot": 0.5576948875195811,
+      "eval_pearson_euclidean": 0.750992670186065,
+      "eval_pearson_manhattan": 0.7511405973753114,
+      "eval_runtime": 39.4437,
+      "eval_samples_per_second": 38.029,
+      "eval_spearman_cosine": 0.7708766062647188,
+      "eval_spearman_dot": 0.5635255554893936,
+      "eval_spearman_euclidean": 0.7633148873618476,
+      "eval_spearman_manhattan": 0.7634983981685618,
+      "eval_steps_per_second": 38.029,
+      "step": 2250
+    },
+    {
+      "epoch": 1.0590440487347703,
+      "grad_norm": 2.5779619216918945,
+      "learning_rate": 9.867619493908154e-06,
+      "loss": 0.4064,
+      "step": 2260
+    },
+    {
+      "epoch": 1.063730084348641,
+      "grad_norm": 3.4719700813293457,
+      "learning_rate": 9.867033739456421e-06,
+      "loss": 0.424,
+      "step": 2270
+    },
+    {
+      "epoch": 1.0684161199625117,
+      "grad_norm": 2.8868348598480225,
+      "learning_rate": 9.866447985004687e-06,
+      "loss": 0.3964,
+      "step": 2280
+    },
+    {
+      "epoch": 1.0731021555763824,
+      "grad_norm": 3.0720372200012207,
+      "learning_rate": 9.865862230552954e-06,
+      "loss": 0.4319,
+      "step": 2290
+    },
+    {
+      "epoch": 1.077788191190253,
+      "grad_norm": 2.9588916301727295,
+      "learning_rate": 9.865276476101219e-06,
+      "loss": 0.4045,
+      "step": 2300
+    },
+    {
+      "epoch": 1.0824742268041236,
+      "grad_norm": 2.9125168323516846,
+      "learning_rate": 9.864690721649486e-06,
+      "loss": 0.4147,
+      "step": 2310
+    },
+    {
+      "epoch": 1.0871602624179943,
+      "grad_norm": 3.1080446243286133,
+      "learning_rate": 9.864104967197753e-06,
+      "loss": 0.3792,
+      "step": 2320
+    },
+    {
+      "epoch": 1.091846298031865,
+      "grad_norm": 3.130577802658081,
+      "learning_rate": 9.863519212746018e-06,
+      "loss": 0.3995,
+      "step": 2330
+    },
+    {
+      "epoch": 1.0965323336457358,
+      "grad_norm": 3.0566961765289307,
+      "learning_rate": 9.862933458294283e-06,
+      "loss": 0.4044,
+      "step": 2340
+    },
+    {
+      "epoch": 1.1012183692596063,
+      "grad_norm": 3.1854050159454346,
+      "learning_rate": 9.86234770384255e-06,
+      "loss": 0.3703,
+      "step": 2350
+    },
+    {
+      "epoch": 1.105904404873477,
+      "grad_norm": 3.034507989883423,
+      "learning_rate": 9.861761949390815e-06,
+      "loss": 0.4062,
+      "step": 2360
+    },
+    {
+      "epoch": 1.1105904404873477,
+      "grad_norm": 3.1277084350585938,
+      "learning_rate": 9.861176194939082e-06,
+      "loss": 0.3733,
+      "step": 2370
+    },
+    {
+      "epoch": 1.1152764761012184,
+      "grad_norm": 2.593057155609131,
+      "learning_rate": 9.860590440487349e-06,
+      "loss": 0.4395,
+      "step": 2380
+    },
+    {
+      "epoch": 1.119962511715089,
+      "grad_norm": 3.079909086227417,
+      "learning_rate": 9.860004686035614e-06,
+      "loss": 0.3849,
+      "step": 2390
+    },
+    {
+      "epoch": 1.1246485473289598,
+      "grad_norm": 3.3298873901367188,
+      "learning_rate": 9.859418931583881e-06,
+      "loss": 0.3753,
+      "step": 2400
+    },
+    {
+      "epoch": 1.1293345829428303,
+      "grad_norm": 3.211998224258423,
+      "learning_rate": 9.858833177132146e-06,
+      "loss": 0.4062,
+      "step": 2410
+    },
+    {
+      "epoch": 1.134020618556701,
+      "grad_norm": 3.5661370754241943,
+      "learning_rate": 9.858247422680413e-06,
+      "loss": 0.3734,
+      "step": 2420
+    },
+    {
+      "epoch": 1.1387066541705717,
+      "grad_norm": 3.004592180252075,
+      "learning_rate": 9.85766166822868e-06,
+      "loss": 0.3847,
+      "step": 2430
+    },
+    {
+      "epoch": 1.1433926897844424,
+      "grad_norm": 3.070695400238037,
+      "learning_rate": 9.857075913776945e-06,
+      "loss": 0.3951,
+      "step": 2440
+    },
+    {
+      "epoch": 1.148078725398313,
+      "grad_norm": 2.348797559738159,
+      "learning_rate": 9.856490159325212e-06,
+      "loss": 0.4213,
+      "step": 2450
+    },
+    {
+      "epoch": 1.1527647610121836,
+      "grad_norm": 2.997864246368408,
+      "learning_rate": 9.855904404873478e-06,
+      "loss": 0.39,
+      "step": 2460
+    },
+    {
+      "epoch": 1.1574507966260543,
+      "grad_norm": 2.844785451889038,
+      "learning_rate": 9.855318650421743e-06,
+      "loss": 0.3889,
+      "step": 2470
+    },
+    {
+      "epoch": 1.162136832239925,
+      "grad_norm": 2.3173370361328125,
+      "learning_rate": 9.854732895970011e-06,
+      "loss": 0.387,
+      "step": 2480
+    },
+    {
+      "epoch": 1.1668228678537957,
+      "grad_norm": 2.7600715160369873,
+      "learning_rate": 9.854147141518277e-06,
+      "loss": 0.3932,
+      "step": 2490
+    },
+    {
+      "epoch": 1.1715089034676662,
+      "grad_norm": 2.941807270050049,
+      "learning_rate": 9.853561387066542e-06,
+      "loss": 0.4136,
+      "step": 2500
+    },
+    {
+      "epoch": 1.1715089034676662,
+      "eval_loss": 0.09154797345399857,
+      "eval_pearson_cosine": 0.7611853083767528,
+      "eval_pearson_dot": 0.5553938407700798,
+      "eval_pearson_euclidean": 0.7585919353320634,
+      "eval_pearson_manhattan": 0.7584339332224772,
+      "eval_runtime": 39.5654,
+      "eval_samples_per_second": 37.912,
+      "eval_spearman_cosine": 0.7727273086490055,
+      "eval_spearman_dot": 0.5594934939736275,
+      "eval_spearman_euclidean": 0.7696220909317546,
+      "eval_spearman_manhattan": 0.7696255697874876,
+      "eval_steps_per_second": 37.912,
+      "step": 2500
+    },
+    {
+      "epoch": 1.176194939081537,
+      "grad_norm": 2.902358293533325,
+      "learning_rate": 9.852975632614809e-06,
+      "loss": 0.38,
+      "step": 2510
+    },
+    {
+      "epoch": 1.1808809746954076,
+      "grad_norm": 2.7850334644317627,
+      "learning_rate": 9.852389878163074e-06,
+      "loss": 0.3899,
+      "step": 2520
+    },
+    {
+      "epoch": 1.1855670103092784,
+      "grad_norm": 3.1563310623168945,
+      "learning_rate": 9.851804123711341e-06,
+      "loss": 0.3597,
+      "step": 2530
+    },
+    {
+      "epoch": 1.190253045923149,
+      "grad_norm": 2.9025473594665527,
+      "learning_rate": 9.851218369259608e-06,
+      "loss": 0.376,
+      "step": 2540
+    },
+    {
+      "epoch": 1.1949390815370198,
+      "grad_norm": 3.4233460426330566,
+      "learning_rate": 9.850632614807873e-06,
+      "loss": 0.4129,
+      "step": 2550
+    },
+    {
+      "epoch": 1.1996251171508903,
+      "grad_norm": 2.8708131313323975,
+      "learning_rate": 9.85004686035614e-06,
+      "loss": 0.3979,
+      "step": 2560
+    },
+    {
+      "epoch": 1.204311152764761,
+      "grad_norm": 2.335909843444824,
+      "learning_rate": 9.849461105904405e-06,
+      "loss": 0.3962,
+      "step": 2570
+    },
+    {
+      "epoch": 1.2089971883786317,
+      "grad_norm": 3.7836697101593018,
+      "learning_rate": 9.848875351452672e-06,
+      "loss": 0.4116,
+      "step": 2580
+    },
+    {
+      "epoch": 1.2136832239925024,
+      "grad_norm": 3.2370989322662354,
+      "learning_rate": 9.848289597000939e-06,
+      "loss": 0.3827,
+      "step": 2590
+    },
+    {
+      "epoch": 1.218369259606373,
+      "grad_norm": 3.3212838172912598,
+      "learning_rate": 9.847703842549204e-06,
+      "loss": 0.4082,
+      "step": 2600
+    },
+    {
+      "epoch": 1.2230552952202436,
+      "grad_norm": 3.7373929023742676,
+      "learning_rate": 9.847118088097471e-06,
+      "loss": 0.4193,
+      "step": 2610
+    },
+    {
+      "epoch": 1.2277413308341143,
+      "grad_norm": 2.658543825149536,
+      "learning_rate": 9.846532333645736e-06,
+      "loss": 0.3539,
+      "step": 2620
+    },
+    {
+      "epoch": 1.232427366447985,
+      "grad_norm": 3.5882198810577393,
+      "learning_rate": 9.845946579194002e-06,
+      "loss": 0.3966,
+      "step": 2630
+    },
+    {
+      "epoch": 1.2371134020618557,
+      "grad_norm": 3.1050803661346436,
+      "learning_rate": 9.84536082474227e-06,
+      "loss": 0.4653,
+      "step": 2640
+    },
+    {
+      "epoch": 1.2417994376757264,
+      "grad_norm": 2.8855292797088623,
+      "learning_rate": 9.844775070290535e-06,
+      "loss": 0.3762,
+      "step": 2650
+    },
+    {
+      "epoch": 1.246485473289597,
+      "grad_norm": 3.595693588256836,
+      "learning_rate": 9.8441893158388e-06,
+      "loss": 0.4139,
+      "step": 2660
+    },
+    {
+      "epoch": 1.2511715089034676,
+      "grad_norm": 3.307594060897827,
+      "learning_rate": 9.843603561387068e-06,
+      "loss": 0.3821,
+      "step": 2670
+    },
+    {
+      "epoch": 1.2558575445173383,
+      "grad_norm": 2.86323618888855,
+      "learning_rate": 9.843017806935333e-06,
+      "loss": 0.4311,
+      "step": 2680
+    },
+    {
+      "epoch": 1.260543580131209,
+      "grad_norm": 3.1539525985717773,
+      "learning_rate": 9.8424320524836e-06,
+      "loss": 0.381,
+      "step": 2690
+    },
+    {
+      "epoch": 1.2652296157450795,
+      "grad_norm": 2.8784306049346924,
+      "learning_rate": 9.841846298031867e-06,
+      "loss": 0.4353,
+      "step": 2700
+    },
+    {
+      "epoch": 1.2699156513589505,
+      "grad_norm": 2.426830291748047,
+      "learning_rate": 9.841260543580132e-06,
+      "loss": 0.387,
+      "step": 2710
+    },
+    {
+      "epoch": 1.274601686972821,
+      "grad_norm": 2.350255012512207,
+      "learning_rate": 9.840674789128399e-06,
+      "loss": 0.3409,
+      "step": 2720
+    },
+    {
+      "epoch": 1.2792877225866917,
+      "grad_norm": 2.589404344558716,
+      "learning_rate": 9.840089034676664e-06,
+      "loss": 0.3665,
+      "step": 2730
+    },
+    {
+      "epoch": 1.2839737582005624,
+      "grad_norm": 2.6439192295074463,
+      "learning_rate": 9.839503280224931e-06,
+      "loss": 0.4118,
+      "step": 2740
+    },
+    {
+      "epoch": 1.2886597938144329,
+      "grad_norm": 3.120351791381836,
+      "learning_rate": 9.838917525773196e-06,
+      "loss": 0.4425,
+      "step": 2750
+    },
+    {
+      "epoch": 1.2886597938144329,
+      "eval_loss": 0.09277115762233734,
+      "eval_pearson_cosine": 0.7605294387552419,
+      "eval_pearson_dot": 0.5497643159360166,
+      "eval_pearson_euclidean": 0.7462760938099731,
+      "eval_pearson_manhattan": 0.7461382274954786,
+      "eval_runtime": 39.1884,
+      "eval_samples_per_second": 38.277,
+      "eval_spearman_cosine": 0.7726058957512391,
+      "eval_spearman_dot": 0.551187587856229,
+      "eval_spearman_euclidean": 0.7591610142998378,
+      "eval_spearman_manhattan": 0.7591112847270914,
+      "eval_steps_per_second": 38.277,
+      "step": 2750
+    },
+    {
+      "epoch": 1.2933458294283038,
+      "grad_norm": 2.9216575622558594,
+      "learning_rate": 9.838331771321463e-06,
+      "loss": 0.376,
+      "step": 2760
+    },
+    {
+      "epoch": 1.2980318650421743,
+      "grad_norm": 2.9521021842956543,
+      "learning_rate": 9.83774601686973e-06,
+      "loss": 0.3597,
+      "step": 2770
+    },
+    {
+      "epoch": 1.302717900656045,
+      "grad_norm": 3.4117023944854736,
+      "learning_rate": 9.837160262417995e-06,
+      "loss": 0.3833,
+      "step": 2780
+    },
+    {
+      "epoch": 1.3074039362699157,
+      "grad_norm": 3.0038156509399414,
+      "learning_rate": 9.83657450796626e-06,
+      "loss": 0.3996,
+      "step": 2790
+    },
+    {
+      "epoch": 1.3120899718837864,
+      "grad_norm": 2.500950813293457,
+      "learning_rate": 9.835988753514527e-06,
+      "loss": 0.3619,
+      "step": 2800
+    },
+    {
+      "epoch": 1.316776007497657,
+      "grad_norm": 2.954556465148926,
+      "learning_rate": 9.835402999062794e-06,
+      "loss": 0.4053,
+      "step": 2810
+    },
+    {
+      "epoch": 1.3214620431115276,
+      "grad_norm": 2.647721290588379,
+      "learning_rate": 9.83481724461106e-06,
+      "loss": 0.367,
+      "step": 2820
+    },
+    {
+      "epoch": 1.3261480787253983,
+      "grad_norm": 3.3207929134368896,
+      "learning_rate": 9.834231490159326e-06,
+      "loss": 0.3857,
+      "step": 2830
+    },
+    {
+      "epoch": 1.330834114339269,
+      "grad_norm": 2.4351987838745117,
+      "learning_rate": 9.833645735707592e-06,
+      "loss": 0.4087,
+      "step": 2840
+    },
+    {
+      "epoch": 1.3355201499531397,
+      "grad_norm": 2.845399856567383,
+      "learning_rate": 9.833059981255859e-06,
+      "loss": 0.3628,
+      "step": 2850
+    },
+    {
+      "epoch": 1.3402061855670104,
+      "grad_norm": 2.342569589614868,
+      "learning_rate": 9.832474226804124e-06,
+      "loss": 0.3656,
+      "step": 2860
+    },
+    {
+      "epoch": 1.344892221180881,
+      "grad_norm": 3.218336582183838,
+      "learning_rate": 9.83188847235239e-06,
+      "loss": 0.3445,
+      "step": 2870
+    },
+    {
+      "epoch": 1.3495782567947516,
+      "grad_norm": 3.147611141204834,
+      "learning_rate": 9.831302717900658e-06,
+      "loss": 0.3766,
+      "step": 2880
+    },
+    {
+      "epoch": 1.3542642924086223,
+      "grad_norm": 2.8554821014404297,
+      "learning_rate": 9.830716963448923e-06,
+      "loss": 0.3719,
+      "step": 2890
+    },
+    {
+      "epoch": 1.358950328022493,
+      "grad_norm": 3.0164778232574463,
+      "learning_rate": 9.83013120899719e-06,
+      "loss": 0.3625,
+      "step": 2900
+    },
+    {
+      "epoch": 1.3636363636363638,
+      "grad_norm": 2.705883741378784,
+      "learning_rate": 9.829545454545455e-06,
+      "loss": 0.4185,
+      "step": 2910
+    },
+    {
+      "epoch": 1.3683223992502342,
+      "grad_norm": 2.8652660846710205,
+      "learning_rate": 9.828959700093722e-06,
+      "loss": 0.3563,
+      "step": 2920
+    },
+    {
+      "epoch": 1.373008434864105,
+      "grad_norm": 2.8929593563079834,
+      "learning_rate": 9.828373945641989e-06,
+      "loss": 0.3769,
+      "step": 2930
+    },
+    {
+      "epoch": 1.3776944704779757,
+      "grad_norm": 2.8356854915618896,
+      "learning_rate": 9.827788191190254e-06,
+      "loss": 0.3796,
+      "step": 2940
+    },
+    {
+      "epoch": 1.3823805060918464,
+      "grad_norm": 2.890596866607666,
+      "learning_rate": 9.82720243673852e-06,
+      "loss": 0.4299,
+      "step": 2950
+    },
+    {
+      "epoch": 1.387066541705717,
+      "grad_norm": 3.0824170112609863,
+      "learning_rate": 9.826616682286786e-06,
+      "loss": 0.3807,
+      "step": 2960
+    },
+    {
+      "epoch": 1.3917525773195876,
+      "grad_norm": 2.549909830093384,
+      "learning_rate": 9.826030927835051e-06,
+      "loss": 0.392,
+      "step": 2970
+    },
+    {
+      "epoch": 1.3964386129334583,
+      "grad_norm": 2.3966541290283203,
+      "learning_rate": 9.825445173383318e-06,
+      "loss": 0.3743,
+      "step": 2980
+    },
+    {
+      "epoch": 1.401124648547329,
+      "grad_norm": 2.7320258617401123,
+      "learning_rate": 9.824859418931585e-06,
+      "loss": 0.3664,
+      "step": 2990
+    },
+    {
+      "epoch": 1.4058106841611997,
+      "grad_norm": 1.8689631223678589,
+      "learning_rate": 9.82427366447985e-06,
+      "loss": 0.3708,
+      "step": 3000
+    },
+    {
+      "epoch": 1.4058106841611997,
+      "eval_loss": 0.08185213804244995,
+      "eval_pearson_cosine": 0.7670467055300527,
+      "eval_pearson_dot": 0.5834479149623917,
+      "eval_pearson_euclidean": 0.7481055315337883,
+      "eval_pearson_manhattan": 0.7477726537908893,
+      "eval_runtime": 39.1851,
+      "eval_samples_per_second": 38.28,
+      "eval_spearman_cosine": 0.7783408589060774,
+      "eval_spearman_dot": 0.5847220399435686,
+      "eval_spearman_euclidean": 0.7636547096676689,
+      "eval_spearman_manhattan": 0.7633506525118475,
+      "eval_steps_per_second": 38.28,
+      "step": 3000
+    },
+    {
+      "epoch": 1.4104967197750704,
+      "grad_norm": 3.3057873249053955,
+      "learning_rate": 9.823687910028117e-06,
+      "loss": 0.3923,
+      "step": 3010
+    },
+    {
+      "epoch": 1.415182755388941,
+      "grad_norm": 2.955125331878662,
+      "learning_rate": 9.823102155576383e-06,
+      "loss": 0.3788,
+      "step": 3020
+    },
+    {
+      "epoch": 1.4198687910028116,
+      "grad_norm": 2.7104790210723877,
+      "learning_rate": 9.82251640112465e-06,
+      "loss": 0.4158,
+      "step": 3030
+    },
+    {
+      "epoch": 1.4245548266166823,
+      "grad_norm": 2.648211717605591,
+      "learning_rate": 9.821930646672916e-06,
+      "loss": 0.3619,
+      "step": 3040
+    },
+    {
+      "epoch": 1.429240862230553,
+      "grad_norm": 2.5092930793762207,
+      "learning_rate": 9.821344892221182e-06,
+      "loss": 0.3995,
+      "step": 3050
+    },
+    {
+      "epoch": 1.4339268978444237,
+      "grad_norm": 3.207554340362549,
+      "learning_rate": 9.820759137769448e-06,
+      "loss": 0.3602,
+      "step": 3060
+    },
+    {
+      "epoch": 1.4386129334582942,
+      "grad_norm": 3.2078282833099365,
+      "learning_rate": 9.820173383317714e-06,
+      "loss": 0.3644,
+      "step": 3070
+    },
+    {
+      "epoch": 1.443298969072165,
+      "grad_norm": 2.8371787071228027,
+      "learning_rate": 9.819587628865979e-06,
+      "loss": 0.368,
+      "step": 3080
+    },
+    {
+      "epoch": 1.4479850046860356,
+      "grad_norm": 2.8893673419952393,
+      "learning_rate": 9.819001874414248e-06,
+      "loss": 0.3611,
+      "step": 3090
+    },
+    {
+      "epoch": 1.4526710402999063,
+      "grad_norm": 3.0875115394592285,
+      "learning_rate": 9.818416119962513e-06,
+      "loss": 0.3809,
+      "step": 3100
+    },
+    {
+      "epoch": 1.457357075913777,
+      "grad_norm": 3.2046687602996826,
+      "learning_rate": 9.817830365510778e-06,
+      "loss": 0.4254,
+      "step": 3110
+    },
+    {
+      "epoch": 1.4620431115276475,
+      "grad_norm": 2.6490259170532227,
+      "learning_rate": 9.817244611059045e-06,
+      "loss": 0.3708,
+      "step": 3120
+    },
+    {
+      "epoch": 1.4667291471415183,
+      "grad_norm": 2.928335189819336,
+      "learning_rate": 9.81665885660731e-06,
+      "loss": 0.4054,
+      "step": 3130
+    },
+    {
+      "epoch": 1.471415182755389,
+      "grad_norm": 3.058704137802124,
+      "learning_rate": 9.816073102155577e-06,
+      "loss": 0.3795,
+      "step": 3140
+    },
+    {
+      "epoch": 1.4761012183692597,
+      "grad_norm": 3.042874336242676,
+      "learning_rate": 9.815487347703844e-06,
+      "loss": 0.349,
+      "step": 3150
+    },
+    {
+      "epoch": 1.4807872539831304,
+      "grad_norm": 3.230562210083008,
+      "learning_rate": 9.81490159325211e-06,
+      "loss": 0.3883,
+      "step": 3160
+    },
+    {
+      "epoch": 1.4854732895970009,
+      "grad_norm": 2.735044240951538,
+      "learning_rate": 9.814315838800376e-06,
+      "loss": 0.3573,
+      "step": 3170
+    },
+    {
+      "epoch": 1.4901593252108716,
+      "grad_norm": 3.3735740184783936,
+      "learning_rate": 9.813730084348641e-06,
+      "loss": 0.3969,
+      "step": 3180
+    },
+    {
+      "epoch": 1.4948453608247423,
+      "grad_norm": 2.9641244411468506,
+      "learning_rate": 9.813144329896908e-06,
+      "loss": 0.4103,
+      "step": 3190
+    },
+    {
+      "epoch": 1.499531396438613,
+      "grad_norm": 2.666172742843628,
+      "learning_rate": 9.812558575445175e-06,
+      "loss": 0.3739,
+      "step": 3200
+    },
+    {
+      "epoch": 1.5042174320524837,
+      "grad_norm": 3.061447858810425,
+      "learning_rate": 9.81197282099344e-06,
+      "loss": 0.3759,
+      "step": 3210
+    },
+    {
+      "epoch": 1.5089034676663542,
+      "grad_norm": 2.6585240364074707,
+      "learning_rate": 9.811387066541707e-06,
+      "loss": 0.3521,
+      "step": 3220
+    },
+    {
+      "epoch": 1.513589503280225,
+      "grad_norm": 3.168820858001709,
+      "learning_rate": 9.810801312089973e-06,
+      "loss": 0.3681,
+      "step": 3230
+    },
+    {
+      "epoch": 1.5182755388940956,
+      "grad_norm": 3.069749355316162,
+      "learning_rate": 9.810215557638238e-06,
+      "loss": 0.3696,
+      "step": 3240
+    },
+    {
+      "epoch": 1.522961574507966,
+      "grad_norm": 3.588362693786621,
+      "learning_rate": 9.809629803186505e-06,
+      "loss": 0.3934,
+      "step": 3250
+    },
+    {
+      "epoch": 1.522961574507966,
+      "eval_loss": 0.08483820408582687,
+      "eval_pearson_cosine": 0.7708958105015995,
+      "eval_pearson_dot": 0.5654852409557236,
+      "eval_pearson_euclidean": 0.7542229262005833,
+      "eval_pearson_manhattan": 0.7539007508537807,
+      "eval_runtime": 40.1586,
+      "eval_samples_per_second": 37.352,
+      "eval_spearman_cosine": 0.7814200034454707,
+      "eval_spearman_dot": 0.5667704812278335,
+      "eval_spearman_euclidean": 0.7689048877801009,
+      "eval_spearman_manhattan": 0.7692340606121537,
+      "eval_steps_per_second": 37.352,
+      "step": 3250
+    },
+    {
+      "epoch": 1.527647610121837,
+      "grad_norm": 2.5335776805877686,
+      "learning_rate": 9.809044048734772e-06,
+      "loss": 0.404,
+      "step": 3260
+    },
+    {
+      "epoch": 1.5323336457357075,
+      "grad_norm": 3.6495397090911865,
+      "learning_rate": 9.808458294283037e-06,
+      "loss": 0.4017,
+      "step": 3270
+    },
+    {
+      "epoch": 1.5370196813495782,
+      "grad_norm": 3.3083150386810303,
+      "learning_rate": 9.807872539831304e-06,
+      "loss": 0.3839,
+      "step": 3280
+    },
+    {
+      "epoch": 1.541705716963449,
+      "grad_norm": 2.9375736713409424,
+      "learning_rate": 9.807286785379569e-06,
+      "loss": 0.4067,
+      "step": 3290
+    },
+    {
+      "epoch": 1.5463917525773194,
+      "grad_norm": 2.8293914794921875,
+      "learning_rate": 9.806701030927836e-06,
+      "loss": 0.4506,
+      "step": 3300
+    },
+    {
+      "epoch": 1.5510777881911904,
+      "grad_norm": 2.9611403942108154,
+      "learning_rate": 9.806115276476103e-06,
+      "loss": 0.3703,
+      "step": 3310
+    },
+    {
+      "epoch": 1.5557638238050608,
+      "grad_norm": 2.821136713027954,
+      "learning_rate": 9.805529522024368e-06,
+      "loss": 0.3756,
+      "step": 3320
+    },
+    {
+      "epoch": 1.5604498594189316,
+      "grad_norm": 2.8910715579986572,
+      "learning_rate": 9.804943767572635e-06,
+      "loss": 0.3911,
+      "step": 3330
+    },
+    {
+      "epoch": 1.5651358950328023,
+      "grad_norm": 3.2398719787597656,
+      "learning_rate": 9.8043580131209e-06,
+      "loss": 0.4014,
+      "step": 3340
+    },
+    {
+      "epoch": 1.569821930646673,
+      "grad_norm": 2.7479453086853027,
+      "learning_rate": 9.803772258669167e-06,
+      "loss": 0.3854,
+      "step": 3350
+    },
+    {
+      "epoch": 1.5745079662605437,
+      "grad_norm": 3.2347192764282227,
+      "learning_rate": 9.803186504217432e-06,
+      "loss": 0.3913,
+      "step": 3360
+    },
+    {
+      "epoch": 1.5791940018744142,
+      "grad_norm": 2.865525245666504,
+      "learning_rate": 9.8026007497657e-06,
+      "loss": 0.3685,
+      "step": 3370
+    },
+    {
+      "epoch": 1.5838800374882849,
+      "grad_norm": 3.2018258571624756,
+      "learning_rate": 9.802014995313966e-06,
+      "loss": 0.3376,
+      "step": 3380
+    },
+    {
+      "epoch": 1.5885660731021556,
+      "grad_norm": 2.9291999340057373,
+      "learning_rate": 9.801429240862231e-06,
+      "loss": 0.3673,
+      "step": 3390
+    },
+    {
+      "epoch": 1.5932521087160263,
+      "grad_norm": 2.761934757232666,
+      "learning_rate": 9.800843486410497e-06,
+      "loss": 0.3885,
+      "step": 3400
+    },
+    {
+      "epoch": 1.597938144329897,
+      "grad_norm": 2.509103775024414,
+      "learning_rate": 9.800257731958763e-06,
+      "loss": 0.3523,
+      "step": 3410
+    },
+    {
+      "epoch": 1.6026241799437675,
+      "grad_norm": 2.377322196960449,
+      "learning_rate": 9.79967197750703e-06,
+      "loss": 0.3791,
+      "step": 3420
+    },
+    {
+      "epoch": 1.6073102155576382,
+      "grad_norm": 3.3458712100982666,
+      "learning_rate": 9.799086223055296e-06,
+      "loss": 0.3897,
+      "step": 3430
+    },
+    {
+      "epoch": 1.611996251171509,
+      "grad_norm": 2.7032053470611572,
+      "learning_rate": 9.798500468603563e-06,
+      "loss": 0.339,
+      "step": 3440
+    },
+    {
+      "epoch": 1.6166822867853796,
+      "grad_norm": 3.09551739692688,
+      "learning_rate": 9.797914714151828e-06,
+      "loss": 0.384,
+      "step": 3450
+    },
+    {
+      "epoch": 1.6213683223992503,
+      "grad_norm": 3.2145328521728516,
+      "learning_rate": 9.797328959700095e-06,
+      "loss": 0.3961,
+      "step": 3460
+    },
+    {
+      "epoch": 1.6260543580131208,
+      "grad_norm": 3.6237032413482666,
+      "learning_rate": 9.79674320524836e-06,
+      "loss": 0.348,
+      "step": 3470
+    },
+    {
+      "epoch": 1.6307403936269915,
+      "grad_norm": 2.706770420074463,
+      "learning_rate": 9.796157450796627e-06,
+      "loss": 0.3281,
+      "step": 3480
+    },
+    {
+      "epoch": 1.6354264292408622,
+      "grad_norm": 3.071535348892212,
+      "learning_rate": 9.795571696344894e-06,
+      "loss": 0.3645,
+      "step": 3490
+    },
+    {
+      "epoch": 1.640112464854733,
+      "grad_norm": 2.757957696914673,
+      "learning_rate": 9.794985941893159e-06,
+      "loss": 0.3203,
+      "step": 3500
+    },
+    {
+      "epoch": 1.640112464854733,
+      "eval_loss": 0.07808861136436462,
+      "eval_pearson_cosine": 0.7705724911864849,
+      "eval_pearson_dot": 0.5870542264131444,
+      "eval_pearson_euclidean": 0.7531019492466129,
+      "eval_pearson_manhattan": 0.7529286527249113,
+      "eval_runtime": 39.5374,
+      "eval_samples_per_second": 37.939,
+      "eval_spearman_cosine": 0.781003234465689,
+      "eval_spearman_dot": 0.5891471444366018,
+      "eval_spearman_euclidean": 0.7690560764997629,
+      "eval_spearman_manhattan": 0.7689244287192701,
+      "eval_steps_per_second": 37.939,
+      "step": 3500
+    },
+    {
+      "epoch": 1.6447985004686037,
+      "grad_norm": 2.8622870445251465,
+      "learning_rate": 9.794400187441426e-06,
+      "loss": 0.361,
+      "step": 3510
+    },
+    {
+      "epoch": 1.6494845360824741,
+      "grad_norm": 3.1786670684814453,
+      "learning_rate": 9.793814432989691e-06,
+      "loss": 0.376,
+      "step": 3520
+    },
+    {
+      "epoch": 1.6541705716963448,
+      "grad_norm": 2.9098026752471924,
+      "learning_rate": 9.793228678537958e-06,
+      "loss": 0.383,
+      "step": 3530
+    },
+    {
+      "epoch": 1.6588566073102156,
+      "grad_norm": 3.0556936264038086,
+      "learning_rate": 9.792642924086225e-06,
+      "loss": 0.3966,
+      "step": 3540
+    },
+    {
+      "epoch": 1.6635426429240863,
+      "grad_norm": 2.920565128326416,
+      "learning_rate": 9.79205716963449e-06,
+      "loss": 0.3531,
+      "step": 3550
+    },
+    {
+      "epoch": 1.668228678537957,
+      "grad_norm": 2.669887065887451,
+      "learning_rate": 9.791471415182755e-06,
+      "loss": 0.3784,
+      "step": 3560
+    },
+    {
+      "epoch": 1.6729147141518275,
+      "grad_norm": 2.7581350803375244,
+      "learning_rate": 9.790885660731022e-06,
+      "loss": 0.3562,
+      "step": 3570
+    },
+    {
+      "epoch": 1.6776007497656982,
+      "grad_norm": 3.3164896965026855,
+      "learning_rate": 9.790299906279287e-06,
+      "loss": 0.363,
+      "step": 3580
+    },
+    {
+      "epoch": 1.6822867853795689,
+      "grad_norm": 3.1747140884399414,
+      "learning_rate": 9.789714151827554e-06,
+      "loss": 0.3816,
+      "step": 3590
+    },
+    {
+      "epoch": 1.6869728209934396,
+      "grad_norm": 2.896433115005493,
+      "learning_rate": 9.789128397375821e-06,
+      "loss": 0.3727,
+      "step": 3600
+    },
+    {
+      "epoch": 1.6916588566073103,
+      "grad_norm": 2.8748035430908203,
+      "learning_rate": 9.788542642924087e-06,
+      "loss": 0.3501,
+      "step": 3610
+    },
+    {
+      "epoch": 1.6963448922211808,
+      "grad_norm": 3.0647246837615967,
+      "learning_rate": 9.787956888472353e-06,
+      "loss": 0.3742,
+      "step": 3620
+    },
+    {
+      "epoch": 1.7010309278350515,
+      "grad_norm": 2.368314504623413,
+      "learning_rate": 9.787371134020619e-06,
+      "loss": 0.3509,
+      "step": 3630
+    },
+    {
+      "epoch": 1.7057169634489222,
+      "grad_norm": 2.563969135284424,
+      "learning_rate": 9.786785379568886e-06,
+      "loss": 0.3205,
+      "step": 3640
+    },
+    {
+      "epoch": 1.710402999062793,
+      "grad_norm": 3.2098007202148438,
+      "learning_rate": 9.786199625117153e-06,
+      "loss": 0.3607,
+      "step": 3650
+    },
+    {
+      "epoch": 1.7150890346766636,
+      "grad_norm": 3.1060285568237305,
+      "learning_rate": 9.785613870665418e-06,
+      "loss": 0.3159,
+      "step": 3660
+    },
+    {
+      "epoch": 1.7197750702905341,
+      "grad_norm": 2.471100330352783,
+      "learning_rate": 9.785028116213685e-06,
+      "loss": 0.375,
+      "step": 3670
+    },
+    {
+      "epoch": 1.7244611059044048,
+      "grad_norm": 2.991903305053711,
+      "learning_rate": 9.78444236176195e-06,
+      "loss": 0.3944,
+      "step": 3680
+    },
+    {
+      "epoch": 1.7291471415182755,
+      "grad_norm": 2.6646485328674316,
+      "learning_rate": 9.783856607310215e-06,
+      "loss": 0.4026,
+      "step": 3690
+    },
+    {
+      "epoch": 1.7338331771321462,
+      "grad_norm": 2.392664670944214,
+      "learning_rate": 9.783270852858484e-06,
+      "loss": 0.3606,
+      "step": 3700
+    },
+    {
+      "epoch": 1.738519212746017,
+      "grad_norm": 3.0139496326446533,
+      "learning_rate": 9.782685098406749e-06,
+      "loss": 0.3475,
+      "step": 3710
+    },
+    {
+      "epoch": 1.7432052483598874,
+      "grad_norm": 2.6688876152038574,
+      "learning_rate": 9.782099343955014e-06,
+      "loss": 0.413,
+      "step": 3720
+    },
+    {
+      "epoch": 1.7478912839737581,
+      "grad_norm": 3.372584581375122,
+      "learning_rate": 9.781513589503281e-06,
+      "loss": 0.3544,
+      "step": 3730
+    },
+    {
+      "epoch": 1.7525773195876289,
+      "grad_norm": 2.9981796741485596,
+      "learning_rate": 9.780927835051546e-06,
+      "loss": 0.3747,
+      "step": 3740
+    },
+    {
+      "epoch": 1.7572633552014996,
+      "grad_norm": 3.5128731727600098,
+      "learning_rate": 9.780342080599813e-06,
+      "loss": 0.4052,
+      "step": 3750
+    },
+    {
+      "epoch": 1.7572633552014996,
+      "eval_loss": 0.08239996433258057,
+      "eval_pearson_cosine": 0.7705098430219977,
+      "eval_pearson_dot": 0.5909106805114561,
+      "eval_pearson_euclidean": 0.7628273731383075,
+      "eval_pearson_manhattan": 0.7627845021563395,
+      "eval_runtime": 39.7472,
+      "eval_samples_per_second": 37.738,
+      "eval_spearman_cosine": 0.7816481068106005,
+      "eval_spearman_dot": 0.5989182063745997,
+      "eval_spearman_euclidean": 0.7771176907760753,
+      "eval_spearman_manhattan": 0.7770550572795577,
+      "eval_steps_per_second": 37.738,
+      "step": 3750
+    },
+    {
+      "epoch": 1.7619493908153703,
+      "grad_norm": 3.215078353881836,
+      "learning_rate": 9.77975632614808e-06,
+      "loss": 0.3424,
+      "step": 3760
+    },
+    {
+      "epoch": 1.7666354264292408,
+      "grad_norm": 3.310758113861084,
+      "learning_rate": 9.779170571696345e-06,
+      "loss": 0.3765,
+      "step": 3770
+    },
+    {
+      "epoch": 1.7713214620431117,
+      "grad_norm": 2.7431821823120117,
+      "learning_rate": 9.778584817244612e-06,
+      "loss": 0.3545,
+      "step": 3780
+    },
+    {
+      "epoch": 1.7760074976569822,
+      "grad_norm": 3.1174018383026123,
+      "learning_rate": 9.777999062792877e-06,
+      "loss": 0.3495,
+      "step": 3790
+    },
+    {
+      "epoch": 1.780693533270853,
+      "grad_norm": 2.911485195159912,
+      "learning_rate": 9.777413308341144e-06,
+      "loss": 0.3684,
+      "step": 3800
+    },
+    {
+      "epoch": 1.7853795688847236,
+      "grad_norm": 3.4018588066101074,
+      "learning_rate": 9.776827553889411e-06,
+      "loss": 0.3702,
+      "step": 3810
+    },
+    {
+      "epoch": 1.790065604498594,
+      "grad_norm": 2.8226757049560547,
+      "learning_rate": 9.776241799437677e-06,
+      "loss": 0.3506,
+      "step": 3820
+    },
+    {
+      "epoch": 1.794751640112465,
+      "grad_norm": 3.5894434452056885,
+      "learning_rate": 9.775656044985943e-06,
+      "loss": 0.3928,
+      "step": 3830
+    },
+    {
+      "epoch": 1.7994376757263355,
+      "grad_norm": 2.49806547164917,
+      "learning_rate": 9.775070290534209e-06,
+      "loss": 0.3162,
+      "step": 3840
+    },
+    {
+      "epoch": 1.8041237113402062,
+      "grad_norm": 3.0954976081848145,
+      "learning_rate": 9.774484536082474e-06,
+      "loss": 0.3657,
+      "step": 3850
+    },
+    {
+      "epoch": 1.808809746954077,
+      "grad_norm": 2.629826307296753,
+      "learning_rate": 9.77389878163074e-06,
+      "loss": 0.3017,
+      "step": 3860
+    },
+    {
+      "epoch": 1.8134957825679474,
+      "grad_norm": 3.1166768074035645,
+      "learning_rate": 9.773313027179008e-06,
+      "loss": 0.3778,
+      "step": 3870
+    },
+    {
+      "epoch": 1.8181818181818183,
+      "grad_norm": 3.0856707096099854,
+      "learning_rate": 9.772727272727273e-06,
+      "loss": 0.3949,
+      "step": 3880
+    },
+    {
+      "epoch": 1.8228678537956888,
+      "grad_norm": 2.7912850379943848,
+      "learning_rate": 9.77214151827554e-06,
+      "loss": 0.377,
+      "step": 3890
+    },
+    {
+      "epoch": 1.8275538894095595,
+      "grad_norm": 2.9028005599975586,
+      "learning_rate": 9.771555763823805e-06,
+      "loss": 0.361,
+      "step": 3900
+    },
+    {
+      "epoch": 1.8322399250234302,
+      "grad_norm": 3.034365177154541,
+      "learning_rate": 9.770970009372072e-06,
+      "loss": 0.3776,
+      "step": 3910
+    },
+    {
+      "epoch": 1.8369259606373007,
+      "grad_norm": 2.7740864753723145,
+      "learning_rate": 9.770384254920339e-06,
+      "loss": 0.3322,
+      "step": 3920
+    },
+    {
+      "epoch": 1.8416119962511717,
+      "grad_norm": 3.775836944580078,
+      "learning_rate": 9.769798500468604e-06,
+      "loss": 0.3668,
+      "step": 3930
+    },
+    {
+      "epoch": 1.8462980318650422,
+      "grad_norm": 3.1334052085876465,
+      "learning_rate": 9.769212746016871e-06,
+      "loss": 0.3633,
+      "step": 3940
+    },
+    {
+      "epoch": 1.8509840674789129,
+      "grad_norm": 3.311575174331665,
+      "learning_rate": 9.768626991565136e-06,
+      "loss": 0.3507,
+      "step": 3950
+    },
+    {
+      "epoch": 1.8556701030927836,
+      "grad_norm": 3.3283636569976807,
+      "learning_rate": 9.768041237113403e-06,
+      "loss": 0.3393,
+      "step": 3960
+    },
+    {
+      "epoch": 1.860356138706654,
+      "grad_norm": 2.8960700035095215,
+      "learning_rate": 9.767455482661668e-06,
+      "loss": 0.4164,
+      "step": 3970
+    },
+    {
+      "epoch": 1.865042174320525,
+      "grad_norm": 2.63865327835083,
+      "learning_rate": 9.766869728209935e-06,
+      "loss": 0.3561,
+      "step": 3980
+    },
+    {
+      "epoch": 1.8697282099343955,
+      "grad_norm": 2.750246286392212,
+      "learning_rate": 9.766283973758202e-06,
+      "loss": 0.3668,
+      "step": 3990
+    },
+    {
+      "epoch": 1.8744142455482662,
+      "grad_norm": 2.736009359359741,
+      "learning_rate": 9.765698219306467e-06,
+      "loss": 0.3723,
+      "step": 4000
+    },
+    {
+      "epoch": 1.8744142455482662,
+      "eval_loss": 0.0818546786904335,
+      "eval_pearson_cosine": 0.7719873880440176,
+      "eval_pearson_dot": 0.5710773246097212,
+      "eval_pearson_euclidean": 0.7519982934890663,
+      "eval_pearson_manhattan": 0.7515239555531821,
+      "eval_runtime": 39.1942,
+      "eval_samples_per_second": 38.271,
+      "eval_spearman_cosine": 0.7840046796615474,
+      "eval_spearman_dot": 0.5713176227902256,
+      "eval_spearman_euclidean": 0.7685352754174982,
+      "eval_spearman_manhattan": 0.76785681763944,
+      "eval_steps_per_second": 38.271,
+      "step": 4000
+    },
+    {
+      "epoch": 1.879100281162137,
+      "grad_norm": 3.1088342666625977,
+      "learning_rate": 9.765112464854733e-06,
+      "loss": 0.3162,
+      "step": 4010
+    },
+    {
+      "epoch": 1.8837863167760074,
+      "grad_norm": 2.2759673595428467,
+      "learning_rate": 9.764526710403e-06,
+      "loss": 0.3647,
+      "step": 4020
+    },
+    {
+      "epoch": 1.8884723523898783,
+      "grad_norm": 2.4756667613983154,
+      "learning_rate": 9.763940955951267e-06,
+      "loss": 0.3741,
+      "step": 4030
+    },
+    {
+      "epoch": 1.8931583880037488,
+      "grad_norm": 2.7309701442718506,
+      "learning_rate": 9.763355201499532e-06,
+      "loss": 0.3732,
+      "step": 4040
+    },
+    {
+      "epoch": 1.8978444236176195,
+      "grad_norm": 2.3570773601531982,
+      "learning_rate": 9.762769447047799e-06,
+      "loss": 0.3593,
+      "step": 4050
+    },
+    {
+      "epoch": 1.9025304592314902,
+      "grad_norm": 3.3677115440368652,
+      "learning_rate": 9.762183692596064e-06,
+      "loss": 0.3859,
+      "step": 4060
+    },
+    {
+      "epoch": 1.9072164948453607,
+      "grad_norm": 3.787653684616089,
+      "learning_rate": 9.76159793814433e-06,
+      "loss": 0.3775,
+      "step": 4070
+    },
+    {
+      "epoch": 1.9119025304592316,
+      "grad_norm": 2.686065196990967,
+      "learning_rate": 9.761012183692596e-06,
+      "loss": 0.3531,
+      "step": 4080
+    },
+    {
+      "epoch": 1.9165885660731021,
+      "grad_norm": 2.230189323425293,
+      "learning_rate": 9.760426429240863e-06,
+      "loss": 0.3291,
+      "step": 4090
+    },
+    {
+      "epoch": 1.9212746016869728,
+      "grad_norm": 2.815934896469116,
+      "learning_rate": 9.75984067478913e-06,
+      "loss": 0.3644,
+      "step": 4100
+    },
+    {
+      "epoch": 1.9259606373008435,
+      "grad_norm": 3.123044967651367,
+      "learning_rate": 9.759254920337395e-06,
+      "loss": 0.377,
+      "step": 4110
+    },
+    {
+      "epoch": 1.930646672914714,
+      "grad_norm": 2.2790846824645996,
+      "learning_rate": 9.758669165885662e-06,
+      "loss": 0.3623,
+      "step": 4120
+    },
+    {
+      "epoch": 1.935332708528585,
+      "grad_norm": 2.658747434616089,
+      "learning_rate": 9.758083411433927e-06,
+      "loss": 0.3289,
+      "step": 4130
+    },
+    {
+      "epoch": 1.9400187441424555,
+      "grad_norm": 2.783327102661133,
+      "learning_rate": 9.757497656982194e-06,
+      "loss": 0.3903,
+      "step": 4140
+    },
+    {
+      "epoch": 1.9447047797563262,
+      "grad_norm": 2.442927598953247,
+      "learning_rate": 9.756911902530461e-06,
+      "loss": 0.3539,
+      "step": 4150
+    },
+    {
+      "epoch": 1.9493908153701969,
+      "grad_norm": 3.6274266242980957,
+      "learning_rate": 9.756326148078726e-06,
+      "loss": 0.3407,
+      "step": 4160
+    },
+    {
+      "epoch": 1.9540768509840674,
+      "grad_norm": 3.194626808166504,
+      "learning_rate": 9.755740393626991e-06,
+      "loss": 0.3585,
+      "step": 4170
+    },
+    {
+      "epoch": 1.9587628865979383,
+      "grad_norm": 2.9573676586151123,
+      "learning_rate": 9.755154639175258e-06,
+      "loss": 0.343,
+      "step": 4180
+    },
+    {
+      "epoch": 1.9634489222118088,
+      "grad_norm": 2.608351707458496,
+      "learning_rate": 9.754568884723524e-06,
+      "loss": 0.3632,
+      "step": 4190
+    },
+    {
+      "epoch": 1.9681349578256795,
+      "grad_norm": 2.9371848106384277,
+      "learning_rate": 9.75398313027179e-06,
+      "loss": 0.3341,
+      "step": 4200
+    },
+    {
+      "epoch": 1.9728209934395502,
+      "grad_norm": 3.698261022567749,
+      "learning_rate": 9.753397375820057e-06,
+      "loss": 0.3903,
+      "step": 4210
+    },
+    {
+      "epoch": 1.9775070290534207,
+      "grad_norm": 2.6606619358062744,
+      "learning_rate": 9.752811621368323e-06,
+      "loss": 0.3365,
+      "step": 4220
+    },
+    {
+      "epoch": 1.9821930646672916,
+      "grad_norm": 3.170403480529785,
+      "learning_rate": 9.75222586691659e-06,
+      "loss": 0.3577,
+      "step": 4230
+    },
+    {
+      "epoch": 1.986879100281162,
+      "grad_norm": 2.7663040161132812,
+      "learning_rate": 9.751640112464855e-06,
+      "loss": 0.3687,
+      "step": 4240
+    },
+    {
+      "epoch": 1.9915651358950328,
+      "grad_norm": 2.217230796813965,
+      "learning_rate": 9.751054358013122e-06,
+      "loss": 0.3645,
+      "step": 4250
+    },
+    {
+      "epoch": 1.9915651358950328,
+      "eval_loss": 0.08020295202732086,
+      "eval_pearson_cosine": 0.7676372739184956,
+      "eval_pearson_dot": 0.5685492783866799,
+      "eval_pearson_euclidean": 0.7560020323373777,
+      "eval_pearson_manhattan": 0.7559508884315278,
+      "eval_runtime": 39.1696,
+      "eval_samples_per_second": 38.295,
+      "eval_spearman_cosine": 0.7804455454568248,
+      "eval_spearman_dot": 0.570131474662444,
+      "eval_spearman_euclidean": 0.770297760657173,
+      "eval_spearman_manhattan": 0.7703714920263548,
+      "eval_steps_per_second": 38.295,
+      "step": 4250
+    },
+    {
+      "epoch": 1.9962511715089035,
+      "grad_norm": 2.7864959239959717,
+      "learning_rate": 9.750468603561389e-06,
+      "loss": 0.3761,
+      "step": 4260
+    },
+    {
+      "epoch": 2.000937207122774,
+      "grad_norm": 2.399378776550293,
+      "learning_rate": 9.749882849109654e-06,
+      "loss": 0.3253,
+      "step": 4270
+    },
+    {
+      "epoch": 2.005623242736645,
+      "grad_norm": 3.282205820083618,
+      "learning_rate": 9.74929709465792e-06,
+      "loss": 0.2838,
+      "step": 4280
+    },
+    {
+      "epoch": 2.0103092783505154,
+      "grad_norm": 2.5611684322357178,
+      "learning_rate": 9.748711340206186e-06,
+      "loss": 0.2921,
+      "step": 4290
+    },
+    {
+      "epoch": 2.014995313964386,
+      "grad_norm": 2.7141940593719482,
+      "learning_rate": 9.748125585754453e-06,
+      "loss": 0.274,
+      "step": 4300
+    },
+    {
+      "epoch": 2.019681349578257,
+      "grad_norm": 2.566196918487549,
+      "learning_rate": 9.74753983130272e-06,
+      "loss": 0.2701,
+      "step": 4310
+    },
+    {
+      "epoch": 2.0243673851921273,
+      "grad_norm": 2.6438400745391846,
+      "learning_rate": 9.746954076850985e-06,
+      "loss": 0.2898,
+      "step": 4320
+    },
+    {
+      "epoch": 2.0290534208059983,
+      "grad_norm": 2.603959560394287,
+      "learning_rate": 9.74636832239925e-06,
+      "loss": 0.2588,
+      "step": 4330
+    },
+    {
+      "epoch": 2.0337394564198688,
+      "grad_norm": 3.4404914379119873,
+      "learning_rate": 9.745782567947517e-06,
+      "loss": 0.3006,
+      "step": 4340
+    },
+    {
+      "epoch": 2.0384254920337392,
+      "grad_norm": 2.573943853378296,
+      "learning_rate": 9.745196813495782e-06,
+      "loss": 0.2826,
+      "step": 4350
+    },
+    {
+      "epoch": 2.04311152764761,
+      "grad_norm": 2.327101945877075,
+      "learning_rate": 9.74461105904405e-06,
+      "loss": 0.2842,
+      "step": 4360
+    },
+    {
+      "epoch": 2.0477975632614807,
+      "grad_norm": 3.299482583999634,
+      "learning_rate": 9.744025304592316e-06,
+      "loss": 0.2649,
+      "step": 4370
+    },
+    {
+      "epoch": 2.0524835988753516,
+      "grad_norm": 2.5444982051849365,
+      "learning_rate": 9.743439550140581e-06,
+      "loss": 0.293,
+      "step": 4380
+    },
+    {
+      "epoch": 2.057169634489222,
+      "grad_norm": 2.7384984493255615,
+      "learning_rate": 9.742853795688848e-06,
+      "loss": 0.3027,
+      "step": 4390
+    },
+    {
+      "epoch": 2.0618556701030926,
+      "grad_norm": 2.323399782180786,
+      "learning_rate": 9.742268041237114e-06,
+      "loss": 0.3082,
+      "step": 4400
+    },
+    {
+      "epoch": 2.0665417057169635,
+      "grad_norm": 2.821531057357788,
+      "learning_rate": 9.74168228678538e-06,
+      "loss": 0.3044,
+      "step": 4410
+    },
+    {
+      "epoch": 2.071227741330834,
+      "grad_norm": 3.8503706455230713,
+      "learning_rate": 9.741096532333647e-06,
+      "loss": 0.2886,
+      "step": 4420
+    },
+    {
+      "epoch": 2.075913776944705,
+      "grad_norm": 3.2392382621765137,
+      "learning_rate": 9.740510777881913e-06,
+      "loss": 0.2589,
+      "step": 4430
+    },
+    {
+      "epoch": 2.0805998125585754,
+      "grad_norm": 2.823723316192627,
+      "learning_rate": 9.73992502343018e-06,
+      "loss": 0.2924,
+      "step": 4440
+    },
+    {
+      "epoch": 2.085285848172446,
+      "grad_norm": 2.604548931121826,
+      "learning_rate": 9.739339268978445e-06,
+      "loss": 0.3004,
+      "step": 4450
+    },
+    {
+      "epoch": 2.089971883786317,
+      "grad_norm": 2.7885937690734863,
+      "learning_rate": 9.738753514526712e-06,
+      "loss": 0.2967,
+      "step": 4460
+    },
+    {
+      "epoch": 2.0946579194001873,
+      "grad_norm": 2.909656524658203,
+      "learning_rate": 9.738167760074977e-06,
+      "loss": 0.2602,
+      "step": 4470
+    },
+    {
+      "epoch": 2.0993439550140582,
+      "grad_norm": 3.450695514678955,
+      "learning_rate": 9.737582005623244e-06,
+      "loss": 0.2841,
+      "step": 4480
+    },
+    {
+      "epoch": 2.1040299906279287,
+      "grad_norm": 2.1142079830169678,
+      "learning_rate": 9.736996251171509e-06,
+      "loss": 0.2492,
+      "step": 4490
+    },
+    {
+      "epoch": 2.108716026241799,
+      "grad_norm": 3.1369121074676514,
+      "learning_rate": 9.736410496719776e-06,
+      "loss": 0.3007,
+      "step": 4500
+    },
+    {
+      "epoch": 2.108716026241799,
+      "eval_loss": 0.06621846556663513,
+      "eval_pearson_cosine": 0.768161993291983,
+      "eval_pearson_dot": 0.5972572426991363,
+      "eval_pearson_euclidean": 0.7574105370032385,
+      "eval_pearson_manhattan": 0.7571659262987636,
+      "eval_runtime": 40.2217,
+      "eval_samples_per_second": 37.293,
+      "eval_spearman_cosine": 0.7799039169576075,
+      "eval_spearman_dot": 0.5981151115114701,
+      "eval_spearman_euclidean": 0.7720521764960889,
+      "eval_spearman_manhattan": 0.772146767310716,
+      "eval_steps_per_second": 37.293,
+      "step": 4500
+    },
+    {
+      "epoch": 2.11340206185567,
+      "grad_norm": 2.972123146057129,
+      "learning_rate": 9.735824742268041e-06,
+      "loss": 0.2515,
+      "step": 4510
+    },
+    {
+      "epoch": 2.1180880974695406,
+      "grad_norm": 2.9807615280151367,
+      "learning_rate": 9.735238987816308e-06,
+      "loss": 0.2773,
+      "step": 4520
+    },
+    {
+      "epoch": 2.1227741330834116,
+      "grad_norm": 2.88916015625,
+      "learning_rate": 9.734653233364575e-06,
+      "loss": 0.3021,
+      "step": 4530
+    },
+    {
+      "epoch": 2.127460168697282,
+      "grad_norm": 2.4502129554748535,
+      "learning_rate": 9.73406747891284e-06,
+      "loss": 0.2657,
+      "step": 4540
+    },
+    {
+      "epoch": 2.1321462043111525,
+      "grad_norm": 3.6442370414733887,
+      "learning_rate": 9.733481724461107e-06,
+      "loss": 0.2737,
+      "step": 4550
+    },
+    {
+      "epoch": 2.1368322399250235,
+      "grad_norm": 3.181819438934326,
+      "learning_rate": 9.732895970009372e-06,
+      "loss": 0.3209,
+      "step": 4560
+    },
+    {
+      "epoch": 2.141518275538894,
+      "grad_norm": 2.9747514724731445,
+      "learning_rate": 9.73231021555764e-06,
+      "loss": 0.2731,
+      "step": 4570
+    },
+    {
+      "epoch": 2.146204311152765,
+      "grad_norm": 3.7340550422668457,
+      "learning_rate": 9.731724461105905e-06,
+      "loss": 0.3319,
+      "step": 4580
+    },
+    {
+      "epoch": 2.1508903467666354,
+      "grad_norm": 2.3309686183929443,
+      "learning_rate": 9.731138706654171e-06,
+      "loss": 0.2854,
+      "step": 4590
+    },
+    {
+      "epoch": 2.155576382380506,
+      "grad_norm": 2.801131010055542,
+      "learning_rate": 9.730552952202438e-06,
+      "loss": 0.2759,
+      "step": 4600
+    },
+    {
+      "epoch": 2.160262417994377,
+      "grad_norm": 2.5506978034973145,
+      "learning_rate": 9.729967197750704e-06,
+      "loss": 0.2425,
+      "step": 4610
+    },
+    {
+      "epoch": 2.1649484536082473,
+      "grad_norm": 2.3702712059020996,
+      "learning_rate": 9.72938144329897e-06,
+      "loss": 0.2645,
+      "step": 4620
+    },
+    {
+      "epoch": 2.169634489222118,
+      "grad_norm": 2.560842752456665,
+      "learning_rate": 9.728795688847236e-06,
+      "loss": 0.28,
+      "step": 4630
+    },
+    {
+      "epoch": 2.1743205248359887,
+      "grad_norm": 2.8994758129119873,
+      "learning_rate": 9.728209934395501e-06,
+      "loss": 0.2764,
+      "step": 4640
+    },
+    {
+      "epoch": 2.179006560449859,
+      "grad_norm": 3.6292624473571777,
+      "learning_rate": 9.727624179943768e-06,
+      "loss": 0.2827,
+      "step": 4650
+    },
+    {
+      "epoch": 2.18369259606373,
+      "grad_norm": 2.619485378265381,
+      "learning_rate": 9.727038425492035e-06,
+      "loss": 0.2869,
+      "step": 4660
+    },
+    {
+      "epoch": 2.1883786316776006,
+      "grad_norm": 2.676844358444214,
+      "learning_rate": 9.7264526710403e-06,
+      "loss": 0.2862,
+      "step": 4670
+    },
+    {
+      "epoch": 2.1930646672914715,
+      "grad_norm": 2.7725651264190674,
+      "learning_rate": 9.725866916588567e-06,
+      "loss": 0.2965,
+      "step": 4680
+    },
+    {
+      "epoch": 2.197750702905342,
+      "grad_norm": 2.34023380279541,
+      "learning_rate": 9.725281162136832e-06,
+      "loss": 0.2483,
+      "step": 4690
+    },
+    {
+      "epoch": 2.2024367385192125,
+      "grad_norm": 2.652027130126953,
+      "learning_rate": 9.724695407685099e-06,
+      "loss": 0.2901,
+      "step": 4700
+    },
+    {
+      "epoch": 2.2071227741330834,
+      "grad_norm": 2.3790388107299805,
+      "learning_rate": 9.724109653233366e-06,
+      "loss": 0.2797,
+      "step": 4710
+    },
+    {
+      "epoch": 2.211808809746954,
+      "grad_norm": 2.6680283546447754,
+      "learning_rate": 9.723523898781631e-06,
+      "loss": 0.2446,
+      "step": 4720
+    },
+    {
+      "epoch": 2.216494845360825,
+      "grad_norm": 3.196193218231201,
+      "learning_rate": 9.722938144329898e-06,
+      "loss": 0.3073,
+      "step": 4730
+    },
+    {
+      "epoch": 2.2211808809746953,
+      "grad_norm": 2.9742684364318848,
+      "learning_rate": 9.722352389878163e-06,
+      "loss": 0.2641,
+      "step": 4740
+    },
+    {
+      "epoch": 2.2258669165885663,
+      "grad_norm": 1.8688490390777588,
+      "learning_rate": 9.72176663542643e-06,
+      "loss": 0.2397,
+      "step": 4750
+    },
+    {
+      "epoch": 2.2258669165885663,
+      "eval_loss": 0.06173858791589737,
+      "eval_pearson_cosine": 0.7692502208827889,
+      "eval_pearson_dot": 0.5854569446239672,
+      "eval_pearson_euclidean": 0.7502268637430376,
+      "eval_pearson_manhattan": 0.7501450224167883,
+      "eval_runtime": 40.4876,
+      "eval_samples_per_second": 37.048,
+      "eval_spearman_cosine": 0.7781711434828433,
+      "eval_spearman_dot": 0.5898324527411585,
+      "eval_spearman_euclidean": 0.7652357962017762,
+      "eval_spearman_manhattan": 0.765505305290314,
+      "eval_steps_per_second": 37.048,
+      "step": 4750
+    },
+    {
+      "epoch": 2.2305529522024368,
+      "grad_norm": 3.1260733604431152,
+      "learning_rate": 9.721180880974697e-06,
+      "loss": 0.2686,
+      "step": 4760
+    },
+    {
+      "epoch": 2.2352389878163073,
+      "grad_norm": 2.8164772987365723,
+      "learning_rate": 9.720595126522962e-06,
+      "loss": 0.2664,
+      "step": 4770
+    },
+    {
+      "epoch": 2.239925023430178,
+      "grad_norm": 2.9061129093170166,
+      "learning_rate": 9.72000937207123e-06,
+      "loss": 0.2894,
+      "step": 4780
+    },
+    {
+      "epoch": 2.2446110590440487,
+      "grad_norm": 2.427224636077881,
+      "learning_rate": 9.719423617619495e-06,
+      "loss": 0.2952,
+      "step": 4790
+    },
+    {
+      "epoch": 2.2492970946579196,
+      "grad_norm": 3.4472455978393555,
+      "learning_rate": 9.71883786316776e-06,
+      "loss": 0.3036,
+      "step": 4800
+    },
+    {
+      "epoch": 2.25398313027179,
+      "grad_norm": 3.032599925994873,
+      "learning_rate": 9.718252108716027e-06,
+      "loss": 0.2702,
+      "step": 4810
+    },
+    {
+      "epoch": 2.2586691658856606,
+      "grad_norm": 2.9868416786193848,
+      "learning_rate": 9.717666354264294e-06,
+      "loss": 0.2547,
+      "step": 4820
+    },
+    {
+      "epoch": 2.2633552014995315,
+      "grad_norm": 3.0330114364624023,
+      "learning_rate": 9.717080599812559e-06,
+      "loss": 0.2701,
+      "step": 4830
+    },
+    {
+      "epoch": 2.268041237113402,
+      "grad_norm": 3.1494383811950684,
+      "learning_rate": 9.716494845360826e-06,
+      "loss": 0.2474,
+      "step": 4840
+    },
+    {
+      "epoch": 2.2727272727272725,
+      "grad_norm": 2.392869472503662,
+      "learning_rate": 9.715909090909091e-06,
+      "loss": 0.2571,
+      "step": 4850
+    },
+    {
+      "epoch": 2.2774133083411434,
+      "grad_norm": 2.711582660675049,
+      "learning_rate": 9.715323336457358e-06,
+      "loss": 0.2778,
+      "step": 4860
+    },
+    {
+      "epoch": 2.282099343955014,
+      "grad_norm": 3.1134321689605713,
+      "learning_rate": 9.714737582005625e-06,
+      "loss": 0.2702,
+      "step": 4870
+    },
+    {
+      "epoch": 2.286785379568885,
+      "grad_norm": 3.022188901901245,
+      "learning_rate": 9.71415182755389e-06,
+      "loss": 0.248,
+      "step": 4880
+    },
+    {
+      "epoch": 2.2914714151827553,
+      "grad_norm": 2.618468999862671,
+      "learning_rate": 9.713566073102157e-06,
+      "loss": 0.2878,
+      "step": 4890
+    },
+    {
+      "epoch": 2.296157450796626,
+      "grad_norm": 2.559990882873535,
+      "learning_rate": 9.712980318650422e-06,
+      "loss": 0.259,
+      "step": 4900
+    },
+    {
+      "epoch": 2.3008434864104967,
+      "grad_norm": 3.009366750717163,
+      "learning_rate": 9.712394564198689e-06,
+      "loss": 0.2662,
+      "step": 4910
+    },
+    {
+      "epoch": 2.3055295220243672,
+      "grad_norm": 2.6885673999786377,
+      "learning_rate": 9.711808809746956e-06,
+      "loss": 0.2488,
+      "step": 4920
+    },
+    {
+      "epoch": 2.310215557638238,
+      "grad_norm": 2.8503615856170654,
+      "learning_rate": 9.711223055295221e-06,
+      "loss": 0.2508,
+      "step": 4930
+    },
+    {
+      "epoch": 2.3149015932521086,
+      "grad_norm": 2.469139575958252,
+      "learning_rate": 9.710637300843488e-06,
+      "loss": 0.3025,
+      "step": 4940
+    },
+    {
+      "epoch": 2.319587628865979,
+      "grad_norm": 3.6701242923736572,
+      "learning_rate": 9.710051546391753e-06,
+      "loss": 0.3029,
+      "step": 4950
+    },
+    {
+      "epoch": 2.32427366447985,
+      "grad_norm": 2.860112428665161,
+      "learning_rate": 9.709465791940019e-06,
+      "loss": 0.2924,
+      "step": 4960
+    },
+    {
+      "epoch": 2.3289597000937206,
+      "grad_norm": 1.8148601055145264,
+      "learning_rate": 9.708880037488286e-06,
+      "loss": 0.3011,
+      "step": 4970
+    },
+    {
+      "epoch": 2.3336457357075915,
+      "grad_norm": 3.073507070541382,
+      "learning_rate": 9.708294283036552e-06,
+      "loss": 0.2664,
+      "step": 4980
+    },
+    {
+      "epoch": 2.338331771321462,
+      "grad_norm": 3.759744882583618,
+      "learning_rate": 9.707708528584818e-06,
+      "loss": 0.2977,
+      "step": 4990
+    },
+    {
+      "epoch": 2.3430178069353325,
+      "grad_norm": 2.8136818408966064,
+      "learning_rate": 9.707122774133085e-06,
+      "loss": 0.28,
+      "step": 5000
+    },
+    {
+      "epoch": 2.3430178069353325,
+      "eval_loss": 0.0644521713256836,
+      "eval_pearson_cosine": 0.7653835203754937,
+      "eval_pearson_dot": 0.5924636721600756,
+      "eval_pearson_euclidean": 0.7568919402450831,
+      "eval_pearson_manhattan": 0.7567361379039284,
+      "eval_runtime": 40.0042,
+      "eval_samples_per_second": 37.496,
+      "eval_spearman_cosine": 0.7760073163206256,
+      "eval_spearman_dot": 0.5970251712536361,
+      "eval_spearman_euclidean": 0.7705144458647132,
+      "eval_spearman_manhattan": 0.7705196006726509,
+      "eval_steps_per_second": 37.496,
+      "step": 5000
+    },
+    {
+      "epoch": 2.3477038425492034,
+      "grad_norm": 2.8381240367889404,
+      "learning_rate": 9.70653701968135e-06,
+      "loss": 0.2584,
+      "step": 5010
+    },
+    {
+      "epoch": 2.352389878163074,
+      "grad_norm": 2.48789381980896,
+      "learning_rate": 9.705951265229617e-06,
+      "loss": 0.2518,
+      "step": 5020
+    },
+    {
+      "epoch": 2.357075913776945,
+      "grad_norm": 2.8016576766967773,
+      "learning_rate": 9.705365510777884e-06,
+      "loss": 0.2839,
+      "step": 5030
+    },
+    {
+      "epoch": 2.3617619493908153,
+      "grad_norm": 3.750737428665161,
+      "learning_rate": 9.704779756326149e-06,
+      "loss": 0.2757,
+      "step": 5040
+    },
+    {
+      "epoch": 2.3664479850046862,
+      "grad_norm": 3.028477668762207,
+      "learning_rate": 9.704194001874416e-06,
+      "loss": 0.2542,
+      "step": 5050
+    },
+    {
+      "epoch": 2.3711340206185567,
+      "grad_norm": 2.5787017345428467,
+      "learning_rate": 9.703608247422681e-06,
+      "loss": 0.2623,
+      "step": 5060
+    },
+    {
+      "epoch": 2.375820056232427,
+      "grad_norm": 3.54349422454834,
+      "learning_rate": 9.703022492970948e-06,
+      "loss": 0.2872,
+      "step": 5070
+    },
+    {
+      "epoch": 2.380506091846298,
+      "grad_norm": 3.924848794937134,
+      "learning_rate": 9.702436738519213e-06,
+      "loss": 0.2641,
+      "step": 5080
+    },
+    {
+      "epoch": 2.3851921274601686,
+      "grad_norm": 2.8141496181488037,
+      "learning_rate": 9.70185098406748e-06,
+      "loss": 0.2549,
+      "step": 5090
+    },
+    {
+      "epoch": 2.3898781630740396,
+      "grad_norm": 3.0236399173736572,
+      "learning_rate": 9.701265229615747e-06,
+      "loss": 0.2715,
+      "step": 5100
+    },
+    {
+      "epoch": 2.39456419868791,
+      "grad_norm": 2.611743688583374,
+      "learning_rate": 9.700679475164012e-06,
+      "loss": 0.2572,
+      "step": 5110
+    },
+    {
+      "epoch": 2.3992502343017805,
+      "grad_norm": 2.905355930328369,
+      "learning_rate": 9.700093720712277e-06,
+      "loss": 0.2772,
+      "step": 5120
+    },
+    {
+      "epoch": 2.4039362699156515,
+      "grad_norm": 1.6498035192489624,
+      "learning_rate": 9.699507966260544e-06,
+      "loss": 0.2776,
+      "step": 5130
+    },
+    {
+      "epoch": 2.408622305529522,
+      "grad_norm": 4.166587829589844,
+      "learning_rate": 9.69892221180881e-06,
+      "loss": 0.283,
+      "step": 5140
+    },
+    {
+      "epoch": 2.413308341143393,
+      "grad_norm": 2.960628032684326,
+      "learning_rate": 9.698336457357076e-06,
+      "loss": 0.2562,
+      "step": 5150
+    },
+    {
+      "epoch": 2.4179943767572634,
+      "grad_norm": 3.179250955581665,
+      "learning_rate": 9.697750702905343e-06,
+      "loss": 0.2654,
+      "step": 5160
+    },
+    {
+      "epoch": 2.422680412371134,
+      "grad_norm": 2.3080592155456543,
+      "learning_rate": 9.697164948453609e-06,
+      "loss": 0.2728,
+      "step": 5170
+    },
+    {
+      "epoch": 2.427366447985005,
+      "grad_norm": 2.4087467193603516,
+      "learning_rate": 9.696579194001876e-06,
+      "loss": 0.3017,
+      "step": 5180
+    },
+    {
+      "epoch": 2.4320524835988753,
+      "grad_norm": 2.0054640769958496,
+      "learning_rate": 9.69599343955014e-06,
+      "loss": 0.2592,
+      "step": 5190
+    },
+    {
+      "epoch": 2.436738519212746,
+      "grad_norm": 3.7607529163360596,
+      "learning_rate": 9.695407685098408e-06,
+      "loss": 0.2501,
+      "step": 5200
+    },
+    {
+      "epoch": 2.4414245548266167,
+      "grad_norm": 2.0960841178894043,
+      "learning_rate": 9.694821930646675e-06,
+      "loss": 0.2937,
+      "step": 5210
+    },
+    {
+      "epoch": 2.446110590440487,
+      "grad_norm": 2.9174606800079346,
+      "learning_rate": 9.69423617619494e-06,
+      "loss": 0.2509,
+      "step": 5220
+    },
+    {
+      "epoch": 2.450796626054358,
+      "grad_norm": 2.7972617149353027,
+      "learning_rate": 9.693650421743207e-06,
+      "loss": 0.2648,
+      "step": 5230
+    },
+    {
+      "epoch": 2.4554826616682286,
+      "grad_norm": 2.465677261352539,
+      "learning_rate": 9.693064667291472e-06,
+      "loss": 0.2888,
+      "step": 5240
+    },
+    {
+      "epoch": 2.4601686972820995,
+      "grad_norm": 2.6547045707702637,
+      "learning_rate": 9.692478912839737e-06,
+      "loss": 0.2631,
+      "step": 5250
+    },
+    {
+      "epoch": 2.4601686972820995,
+      "eval_loss": 0.06392496824264526,
+      "eval_pearson_cosine": 0.771214841842891,
+      "eval_pearson_dot": 0.5715473025039159,
+      "eval_pearson_euclidean": 0.75617638488007,
+      "eval_pearson_manhattan": 0.7560991988119383,
+      "eval_runtime": 40.3523,
+      "eval_samples_per_second": 37.173,
+      "eval_spearman_cosine": 0.7797840912495824,
+      "eval_spearman_dot": 0.5730858147900617,
+      "eval_spearman_euclidean": 0.7705235550970289,
+      "eval_spearman_manhattan": 0.7705309049511561,
+      "eval_steps_per_second": 37.173,
+      "step": 5250
+    },
+    {
+      "epoch": 2.46485473289597,
+      "grad_norm": 3.2558162212371826,
+      "learning_rate": 9.691893158388006e-06,
+      "loss": 0.2874,
+      "step": 5260
+    },
+    {
+      "epoch": 2.4695407685098405,
+      "grad_norm": 2.307089328765869,
+      "learning_rate": 9.691307403936271e-06,
+      "loss": 0.2681,
+      "step": 5270
+    },
+    {
+      "epoch": 2.4742268041237114,
+      "grad_norm": 3.001068115234375,
+      "learning_rate": 9.690721649484536e-06,
+      "loss": 0.2815,
+      "step": 5280
+    },
+    {
+      "epoch": 2.478912839737582,
+      "grad_norm": 2.5946176052093506,
+      "learning_rate": 9.690135895032803e-06,
+      "loss": 0.2639,
+      "step": 5290
+    },
+    {
+      "epoch": 2.483598875351453,
+      "grad_norm": 2.6081533432006836,
+      "learning_rate": 9.689550140581068e-06,
+      "loss": 0.2707,
+      "step": 5300
+    },
+    {
+      "epoch": 2.4882849109653233,
+      "grad_norm": 3.037405490875244,
+      "learning_rate": 9.688964386129335e-06,
+      "loss": 0.2844,
+      "step": 5310
+    },
+    {
+      "epoch": 2.492970946579194,
+      "grad_norm": 2.944249391555786,
+      "learning_rate": 9.688378631677602e-06,
+      "loss": 0.2728,
+      "step": 5320
+    },
+    {
+      "epoch": 2.4976569821930648,
+      "grad_norm": 2.7798855304718018,
+      "learning_rate": 9.687792877225867e-06,
+      "loss": 0.3013,
+      "step": 5330
+    },
+    {
+      "epoch": 2.5023430178069352,
+      "grad_norm": 2.005322217941284,
+      "learning_rate": 9.687207122774134e-06,
+      "loss": 0.2523,
+      "step": 5340
+    },
+    {
+      "epoch": 2.5070290534208057,
+      "grad_norm": 2.831803321838379,
+      "learning_rate": 9.6866213683224e-06,
+      "loss": 0.2542,
+      "step": 5350
+    },
+    {
+      "epoch": 2.5117150890346767,
+      "grad_norm": 3.1902036666870117,
+      "learning_rate": 9.686035613870666e-06,
+      "loss": 0.241,
+      "step": 5360
+    },
+    {
+      "epoch": 2.5164011246485476,
+      "grad_norm": 2.875300884246826,
+      "learning_rate": 9.685449859418933e-06,
+      "loss": 0.264,
+      "step": 5370
+    },
+    {
+      "epoch": 2.521087160262418,
+      "grad_norm": 3.253399133682251,
+      "learning_rate": 9.684864104967199e-06,
+      "loss": 0.3013,
+      "step": 5380
+    },
+    {
+      "epoch": 2.5257731958762886,
+      "grad_norm": 2.7400455474853516,
+      "learning_rate": 9.684278350515465e-06,
+      "loss": 0.3005,
+      "step": 5390
+    },
+    {
+      "epoch": 2.530459231490159,
+      "grad_norm": 2.604724884033203,
+      "learning_rate": 9.68369259606373e-06,
+      "loss": 0.2805,
+      "step": 5400
+    },
+    {
+      "epoch": 2.53514526710403,
+      "grad_norm": 2.5421054363250732,
+      "learning_rate": 9.683106841611996e-06,
+      "loss": 0.256,
+      "step": 5410
+    },
+    {
+      "epoch": 2.539831302717901,
+      "grad_norm": 3.7680563926696777,
+      "learning_rate": 9.682521087160265e-06,
+      "loss": 0.2685,
+      "step": 5420
+    },
+    {
+      "epoch": 2.5445173383317714,
+      "grad_norm": 3.140620708465576,
+      "learning_rate": 9.68193533270853e-06,
+      "loss": 0.288,
+      "step": 5430
+    },
+    {
+      "epoch": 2.549203373945642,
+      "grad_norm": 2.77986478805542,
+      "learning_rate": 9.681349578256795e-06,
+      "loss": 0.2827,
+      "step": 5440
+    },
+    {
+      "epoch": 2.5538894095595124,
+      "grad_norm": 3.1461918354034424,
+      "learning_rate": 9.680763823805062e-06,
+      "loss": 0.2907,
+      "step": 5450
+    },
+    {
+      "epoch": 2.5585754451733833,
+      "grad_norm": 2.403411626815796,
+      "learning_rate": 9.680178069353327e-06,
+      "loss": 0.2997,
+      "step": 5460
+    },
+    {
+      "epoch": 2.5632614807872542,
+      "grad_norm": 2.744910955429077,
+      "learning_rate": 9.679592314901594e-06,
+      "loss": 0.2615,
+      "step": 5470
+    },
+    {
+      "epoch": 2.5679475164011247,
+      "grad_norm": 3.292695999145508,
+      "learning_rate": 9.679006560449861e-06,
+      "loss": 0.2679,
+      "step": 5480
+    },
+    {
+      "epoch": 2.572633552014995,
+      "grad_norm": 2.8364174365997314,
+      "learning_rate": 9.678420805998126e-06,
+      "loss": 0.2763,
+      "step": 5490
+    },
+    {
+      "epoch": 2.5773195876288657,
+      "grad_norm": 2.456524610519409,
+      "learning_rate": 9.677835051546393e-06,
+      "loss": 0.2488,
+      "step": 5500
+    },
+    {
+      "epoch": 2.5773195876288657,
+      "eval_loss": 0.06359264999628067,
+      "eval_pearson_cosine": 0.7736336142524749,
+      "eval_pearson_dot": 0.5835126580557244,
+      "eval_pearson_euclidean": 0.7537697181391394,
+      "eval_pearson_manhattan": 0.7536913116270512,
+      "eval_runtime": 39.7446,
+      "eval_samples_per_second": 37.741,
+      "eval_spearman_cosine": 0.7838309998226812,
+      "eval_spearman_dot": 0.5860915978330865,
+      "eval_spearman_euclidean": 0.76851014194766,
+      "eval_spearman_manhattan": 0.7687268040235111,
+      "eval_steps_per_second": 37.741,
+      "step": 5500
+    },
+    {
+      "epoch": 2.5820056232427366,
+      "grad_norm": 2.6862993240356445,
+      "learning_rate": 9.677249297094658e-06,
+      "loss": 0.3049,
+      "step": 5510
+    },
+    {
+      "epoch": 2.5866916588566076,
+      "grad_norm": 2.146784782409668,
+      "learning_rate": 9.676663542642925e-06,
+      "loss": 0.2844,
+      "step": 5520
+    },
+    {
+      "epoch": 2.591377694470478,
+      "grad_norm": 2.9161078929901123,
+      "learning_rate": 9.676077788191192e-06,
+      "loss": 0.2505,
+      "step": 5530
+    },
+    {
+      "epoch": 2.5960637300843485,
+      "grad_norm": 2.771800994873047,
+      "learning_rate": 9.675492033739457e-06,
+      "loss": 0.2741,
+      "step": 5540
+    },
+    {
+      "epoch": 2.600749765698219,
+      "grad_norm": 2.6023566722869873,
+      "learning_rate": 9.674906279287724e-06,
+      "loss": 0.271,
+      "step": 5550
+    },
+    {
+      "epoch": 2.60543580131209,
+      "grad_norm": 3.007667064666748,
+      "learning_rate": 9.67432052483599e-06,
+      "loss": 0.2451,
+      "step": 5560
+    },
+    {
+      "epoch": 2.610121836925961,
+      "grad_norm": 3.014564275741577,
+      "learning_rate": 9.673734770384255e-06,
+      "loss": 0.3036,
+      "step": 5570
+    },
+    {
+      "epoch": 2.6148078725398314,
+      "grad_norm": 2.6932876110076904,
+      "learning_rate": 9.673149015932522e-06,
+      "loss": 0.2642,
+      "step": 5580
+    },
+    {
+      "epoch": 2.619493908153702,
+      "grad_norm": 3.396259307861328,
+      "learning_rate": 9.672563261480789e-06,
+      "loss": 0.26,
+      "step": 5590
+    },
+    {
+      "epoch": 2.624179943767573,
+      "grad_norm": 3.203933000564575,
+      "learning_rate": 9.671977507029054e-06,
+      "loss": 0.2629,
+      "step": 5600
+    },
+    {
+      "epoch": 2.6288659793814433,
+      "grad_norm": 3.077299118041992,
+      "learning_rate": 9.67139175257732e-06,
+      "loss": 0.2566,
+      "step": 5610
+    },
+    {
+      "epoch": 2.633552014995314,
+      "grad_norm": 2.445908308029175,
+      "learning_rate": 9.670805998125586e-06,
+      "loss": 0.2751,
+      "step": 5620
+    },
+    {
+      "epoch": 2.6382380506091847,
+      "grad_norm": 2.4680237770080566,
+      "learning_rate": 9.670220243673853e-06,
+      "loss": 0.2725,
+      "step": 5630
+    },
+    {
+      "epoch": 2.642924086223055,
+      "grad_norm": 3.320460081100464,
+      "learning_rate": 9.669634489222118e-06,
+      "loss": 0.3,
+      "step": 5640
+    },
+    {
+      "epoch": 2.647610121836926,
+      "grad_norm": 2.6988685131073,
+      "learning_rate": 9.669048734770385e-06,
+      "loss": 0.2754,
+      "step": 5650
+    },
+    {
+      "epoch": 2.6522961574507966,
+      "grad_norm": 3.1843039989471436,
+      "learning_rate": 9.668462980318652e-06,
+      "loss": 0.2711,
+      "step": 5660
+    },
+    {
+      "epoch": 2.6569821930646675,
+      "grad_norm": 2.5606629848480225,
+      "learning_rate": 9.667877225866917e-06,
+      "loss": 0.2772,
+      "step": 5670
+    },
+    {
+      "epoch": 2.661668228678538,
+      "grad_norm": 2.2161924839019775,
+      "learning_rate": 9.667291471415184e-06,
+      "loss": 0.2501,
+      "step": 5680
+    },
+    {
+      "epoch": 2.6663542642924085,
+      "grad_norm": 2.6044790744781494,
+      "learning_rate": 9.66670571696345e-06,
+      "loss": 0.2859,
+      "step": 5690
+    },
+    {
+      "epoch": 2.6710402999062794,
+      "grad_norm": 2.9149224758148193,
+      "learning_rate": 9.666119962511716e-06,
+      "loss": 0.2834,
+      "step": 5700
+    },
+    {
+      "epoch": 2.67572633552015,
+      "grad_norm": 3.0715954303741455,
+      "learning_rate": 9.665534208059983e-06,
+      "loss": 0.2802,
+      "step": 5710
+    },
+    {
+      "epoch": 2.680412371134021,
+      "grad_norm": 2.8845441341400146,
+      "learning_rate": 9.664948453608248e-06,
+      "loss": 0.2594,
+      "step": 5720
+    },
+    {
+      "epoch": 2.6850984067478914,
+      "grad_norm": 3.3252668380737305,
+      "learning_rate": 9.664362699156514e-06,
+      "loss": 0.2625,
+      "step": 5730
+    },
+    {
+      "epoch": 2.689784442361762,
+      "grad_norm": 2.8618457317352295,
+      "learning_rate": 9.66377694470478e-06,
+      "loss": 0.2769,
+      "step": 5740
+    },
+    {
+      "epoch": 2.6944704779756328,
+      "grad_norm": 3.196368455886841,
+      "learning_rate": 9.663191190253046e-06,
+      "loss": 0.2557,
+      "step": 5750
+    },
+    {
+      "epoch": 2.6944704779756328,
+      "eval_loss": 0.06136869639158249,
+      "eval_pearson_cosine": 0.7739160712003912,
+      "eval_pearson_dot": 0.6008471845975407,
+      "eval_pearson_euclidean": 0.7570902529490127,
+      "eval_pearson_manhattan": 0.7569554961897609,
+      "eval_runtime": 39.4871,
+      "eval_samples_per_second": 37.987,
+      "eval_spearman_cosine": 0.7830112643640406,
+      "eval_spearman_dot": 0.6040995035629967,
+      "eval_spearman_euclidean": 0.7716512847518021,
+      "eval_spearman_manhattan": 0.7716299413202475,
+      "eval_steps_per_second": 37.987,
+      "step": 5750
+    },
+    {
+      "epoch": 2.6991565135895033,
+      "grad_norm": 2.449201822280884,
+      "learning_rate": 9.662605435801313e-06,
+      "loss": 0.2807,
+      "step": 5760
+    },
+    {
+      "epoch": 2.703842549203374,
+      "grad_norm": 3.0767366886138916,
+      "learning_rate": 9.66201968134958e-06,
+      "loss": 0.2842,
+      "step": 5770
+    },
+    {
+      "epoch": 2.7085285848172447,
+      "grad_norm": 2.7501325607299805,
+      "learning_rate": 9.661433926897845e-06,
+      "loss": 0.281,
+      "step": 5780
+    },
+    {
+      "epoch": 2.713214620431115,
+      "grad_norm": 3.0382957458496094,
+      "learning_rate": 9.660848172446112e-06,
+      "loss": 0.2435,
+      "step": 5790
+    },
+    {
+      "epoch": 2.717900656044986,
+      "grad_norm": 3.02397084236145,
+      "learning_rate": 9.660262417994377e-06,
+      "loss": 0.2797,
+      "step": 5800
+    },
+    {
+      "epoch": 2.7225866916588566,
+      "grad_norm": 2.8019211292266846,
+      "learning_rate": 9.659676663542644e-06,
+      "loss": 0.2728,
+      "step": 5810
+    },
+    {
+      "epoch": 2.7272727272727275,
+      "grad_norm": 2.9764068126678467,
+      "learning_rate": 9.65909090909091e-06,
+      "loss": 0.2796,
+      "step": 5820
+    },
+    {
+      "epoch": 2.731958762886598,
+      "grad_norm": 3.254289150238037,
+      "learning_rate": 9.658505154639176e-06,
+      "loss": 0.2967,
+      "step": 5830
+    },
+    {
+      "epoch": 2.7366447985004685,
+      "grad_norm": 3.0329315662384033,
+      "learning_rate": 9.657919400187443e-06,
+      "loss": 0.2631,
+      "step": 5840
+    },
+    {
+      "epoch": 2.7413308341143394,
+      "grad_norm": 2.720784902572632,
+      "learning_rate": 9.657333645735708e-06,
+      "loss": 0.273,
+      "step": 5850
+    },
+    {
+      "epoch": 2.74601686972821,
+      "grad_norm": 2.6841161251068115,
+      "learning_rate": 9.656747891283973e-06,
+      "loss": 0.2729,
+      "step": 5860
+    },
+    {
+      "epoch": 2.750702905342081,
+      "grad_norm": 3.0265884399414062,
+      "learning_rate": 9.656162136832242e-06,
+      "loss": 0.2762,
+      "step": 5870
+    },
+    {
+      "epoch": 2.7553889409559513,
+      "grad_norm": 2.6341817378997803,
+      "learning_rate": 9.655576382380507e-06,
+      "loss": 0.2918,
+      "step": 5880
+    },
+    {
+      "epoch": 2.760074976569822,
+      "grad_norm": 3.1059679985046387,
+      "learning_rate": 9.654990627928772e-06,
+      "loss": 0.2448,
+      "step": 5890
+    },
+    {
+      "epoch": 2.7647610121836927,
+      "grad_norm": 3.105215311050415,
+      "learning_rate": 9.65440487347704e-06,
+      "loss": 0.3158,
+      "step": 5900
+    },
+    {
+      "epoch": 2.7694470477975632,
+      "grad_norm": 2.5665364265441895,
+      "learning_rate": 9.653819119025304e-06,
+      "loss": 0.2742,
+      "step": 5910
+    },
+    {
+      "epoch": 2.774133083411434,
+      "grad_norm": 3.4417858123779297,
+      "learning_rate": 9.653233364573571e-06,
+      "loss": 0.2666,
+      "step": 5920
+    },
+    {
+      "epoch": 2.7788191190253047,
+      "grad_norm": 3.3958632946014404,
+      "learning_rate": 9.652647610121838e-06,
+      "loss": 0.2861,
+      "step": 5930
+    },
+    {
+      "epoch": 2.783505154639175,
+      "grad_norm": 2.7881741523742676,
+      "learning_rate": 9.652061855670104e-06,
+      "loss": 0.2979,
+      "step": 5940
+    },
+    {
+      "epoch": 2.788191190253046,
+      "grad_norm": 3.3900156021118164,
+      "learning_rate": 9.65147610121837e-06,
+      "loss": 0.2919,
+      "step": 5950
+    },
+    {
+      "epoch": 2.7928772258669166,
+      "grad_norm": 2.6520965099334717,
+      "learning_rate": 9.650890346766636e-06,
+      "loss": 0.3384,
+      "step": 5960
+    },
+    {
+      "epoch": 2.7975632614807875,
+      "grad_norm": 2.6379964351654053,
+      "learning_rate": 9.650304592314903e-06,
+      "loss": 0.2553,
+      "step": 5970
+    },
+    {
+      "epoch": 2.802249297094658,
+      "grad_norm": 3.325380802154541,
+      "learning_rate": 9.64971883786317e-06,
+      "loss": 0.2599,
+      "step": 5980
+    },
+    {
+      "epoch": 2.8069353327085285,
+      "grad_norm": 2.9354546070098877,
+      "learning_rate": 9.649133083411435e-06,
+      "loss": 0.2981,
+      "step": 5990
+    },
+    {
+      "epoch": 2.8116213683223994,
+      "grad_norm": 3.915391445159912,
+      "learning_rate": 9.648547328959702e-06,
+      "loss": 0.2699,
+      "step": 6000
+    },
+    {
+      "epoch": 2.8116213683223994,
+      "eval_loss": 0.063628189265728,
+      "eval_pearson_cosine": 0.7721771737399621,
+      "eval_pearson_dot": 0.5844319866585721,
+      "eval_pearson_euclidean": 0.7572362286473506,
+      "eval_pearson_manhattan": 0.7569847240428516,
+      "eval_runtime": 39.6806,
+      "eval_samples_per_second": 37.802,
+      "eval_spearman_cosine": 0.7795493971324328,
+      "eval_spearman_dot": 0.586396480606851,
+      "eval_spearman_euclidean": 0.7700741181493094,
+      "eval_spearman_manhattan": 0.7699220379699384,
+      "eval_steps_per_second": 37.802,
+      "step": 6000
+    },
+    {
+      "epoch": 2.81630740393627,
+      "grad_norm": 2.758107900619507,
+      "learning_rate": 9.647961574507967e-06,
+      "loss": 0.277,
+      "step": 6010
+    },
+    {
+      "epoch": 2.820993439550141,
+      "grad_norm": 2.5489895343780518,
+      "learning_rate": 9.647375820056232e-06,
+      "loss": 0.2776,
+      "step": 6020
+    },
+    {
+      "epoch": 2.8256794751640113,
+      "grad_norm": 2.8726046085357666,
+      "learning_rate": 9.6467900656045e-06,
+      "loss": 0.2719,
+      "step": 6030
+    },
+    {
+      "epoch": 2.830365510777882,
+      "grad_norm": 2.856821298599243,
+      "learning_rate": 9.646204311152766e-06,
+      "loss": 0.2717,
+      "step": 6040
+    },
+    {
+      "epoch": 2.8350515463917527,
+      "grad_norm": 3.395746946334839,
+      "learning_rate": 9.645618556701031e-06,
+      "loss": 0.2657,
+      "step": 6050
+    },
+    {
+      "epoch": 2.839737582005623,
+      "grad_norm": 2.8443150520324707,
+      "learning_rate": 9.645032802249298e-06,
+      "loss": 0.2889,
+      "step": 6060
+    },
+    {
+      "epoch": 2.844423617619494,
+      "grad_norm": 2.680279016494751,
+      "learning_rate": 9.644447047797563e-06,
+      "loss": 0.3128,
+      "step": 6070
+    },
+    {
+      "epoch": 2.8491096532333646,
+      "grad_norm": 3.375882148742676,
+      "learning_rate": 9.64386129334583e-06,
+      "loss": 0.2784,
+      "step": 6080
+    },
+    {
+      "epoch": 2.853795688847235,
+      "grad_norm": 3.4931676387786865,
+      "learning_rate": 9.643275538894097e-06,
+      "loss": 0.2906,
+      "step": 6090
+    },
+    {
+      "epoch": 2.858481724461106,
+      "grad_norm": 3.0119099617004395,
+      "learning_rate": 9.642689784442362e-06,
+      "loss": 0.3064,
+      "step": 6100
+    },
+    {
+      "epoch": 2.8631677600749765,
+      "grad_norm": 2.9946563243865967,
+      "learning_rate": 9.64210402999063e-06,
+      "loss": 0.2708,
+      "step": 6110
+    },
+    {
+      "epoch": 2.8678537956888475,
+      "grad_norm": 2.687286853790283,
+      "learning_rate": 9.641518275538894e-06,
+      "loss": 0.2827,
+      "step": 6120
+    },
+    {
+      "epoch": 2.872539831302718,
+      "grad_norm": 2.7718350887298584,
+      "learning_rate": 9.640932521087161e-06,
+      "loss": 0.2815,
+      "step": 6130
+    },
+    {
+      "epoch": 2.8772258669165884,
+      "grad_norm": 2.4918980598449707,
+      "learning_rate": 9.640346766635427e-06,
+      "loss": 0.2383,
+      "step": 6140
+    },
+    {
+      "epoch": 2.8819119025304594,
+      "grad_norm": 3.1328234672546387,
+      "learning_rate": 9.639761012183694e-06,
+      "loss": 0.2917,
+      "step": 6150
+    },
+    {
+      "epoch": 2.88659793814433,
+      "grad_norm": 3.1907870769500732,
+      "learning_rate": 9.63917525773196e-06,
+      "loss": 0.2682,
+      "step": 6160
+    },
+    {
+      "epoch": 2.891283973758201,
+      "grad_norm": 2.851372241973877,
+      "learning_rate": 9.638589503280226e-06,
+      "loss": 0.2523,
+      "step": 6170
+    },
+    {
+      "epoch": 2.8959700093720713,
+      "grad_norm": 2.8064658641815186,
+      "learning_rate": 9.638003748828491e-06,
+      "loss": 0.276,
+      "step": 6180
+    },
+    {
+      "epoch": 2.9006560449859418,
+      "grad_norm": 3.0457820892333984,
+      "learning_rate": 9.637417994376758e-06,
+      "loss": 0.2713,
+      "step": 6190
+    },
+    {
+      "epoch": 2.9053420805998127,
+      "grad_norm": 2.8154349327087402,
+      "learning_rate": 9.636832239925025e-06,
+      "loss": 0.2757,
+      "step": 6200
+    },
+    {
+      "epoch": 2.910028116213683,
+      "grad_norm": 2.6084420680999756,
+      "learning_rate": 9.63624648547329e-06,
+      "loss": 0.2786,
+      "step": 6210
+    },
+    {
+      "epoch": 2.914714151827554,
+      "grad_norm": 3.3972620964050293,
+      "learning_rate": 9.635660731021557e-06,
+      "loss": 0.284,
+      "step": 6220
+    },
+    {
+      "epoch": 2.9194001874414246,
+      "grad_norm": 2.3089513778686523,
+      "learning_rate": 9.635074976569822e-06,
+      "loss": 0.2912,
+      "step": 6230
+    },
+    {
+      "epoch": 2.924086223055295,
+      "grad_norm": 2.8329966068267822,
+      "learning_rate": 9.634489222118089e-06,
+      "loss": 0.2872,
+      "step": 6240
+    },
+    {
+      "epoch": 2.928772258669166,
+      "grad_norm": 2.8177921772003174,
+      "learning_rate": 9.633903467666354e-06,
+      "loss": 0.2794,
+      "step": 6250
+    },
+    {
+      "epoch": 2.928772258669166,
+      "eval_loss": 0.06385794281959534,
+      "eval_pearson_cosine": 0.7704242421671665,
+      "eval_pearson_dot": 0.5817298961264266,
+      "eval_pearson_euclidean": 0.7580713117672779,
+      "eval_pearson_manhattan": 0.7582345366438972,
+      "eval_runtime": 39.8682,
+      "eval_samples_per_second": 37.624,
+      "eval_spearman_cosine": 0.7799516394114882,
+      "eval_spearman_dot": 0.5793232888676765,
+      "eval_spearman_euclidean": 0.7745810839892152,
+      "eval_spearman_manhattan": 0.7744580162860011,
+      "eval_steps_per_second": 37.624,
+      "step": 6250
+    },
+    {
+      "epoch": 2.9334582942830365,
+      "grad_norm": 3.0042576789855957,
+      "learning_rate": 9.633317713214621e-06,
+      "loss": 0.3072,
+      "step": 6260
+    },
+    {
+      "epoch": 2.9381443298969074,
+      "grad_norm": 2.3798370361328125,
+      "learning_rate": 9.632731958762888e-06,
+      "loss": 0.2472,
+      "step": 6270
+    },
+    {
+      "epoch": 2.942830365510778,
+      "grad_norm": 3.12076735496521,
+      "learning_rate": 9.632146204311153e-06,
+      "loss": 0.2649,
+      "step": 6280
+    },
+    {
+      "epoch": 2.9475164011246484,
+      "grad_norm": 2.4176595211029053,
+      "learning_rate": 9.63156044985942e-06,
+      "loss": 0.3011,
+      "step": 6290
+    },
+    {
+      "epoch": 2.9522024367385193,
+      "grad_norm": 4.061246871948242,
+      "learning_rate": 9.630974695407685e-06,
+      "loss": 0.2905,
+      "step": 6300
+    },
+    {
+      "epoch": 2.95688847235239,
+      "grad_norm": 3.4516801834106445,
+      "learning_rate": 9.630388940955952e-06,
+      "loss": 0.2757,
+      "step": 6310
+    },
+    {
+      "epoch": 2.9615745079662608,
+      "grad_norm": 2.8717644214630127,
+      "learning_rate": 9.62980318650422e-06,
+      "loss": 0.2566,
+      "step": 6320
+    },
+    {
+      "epoch": 2.9662605435801312,
+      "grad_norm": 2.948293685913086,
+      "learning_rate": 9.629217432052484e-06,
+      "loss": 0.2602,
+      "step": 6330
+    },
+    {
+      "epoch": 2.9709465791940017,
+      "grad_norm": 2.5591063499450684,
+      "learning_rate": 9.62863167760075e-06,
+      "loss": 0.2637,
+      "step": 6340
+    },
+    {
+      "epoch": 2.9756326148078727,
+      "grad_norm": 3.3883564472198486,
+      "learning_rate": 9.628045923149017e-06,
+      "loss": 0.3073,
+      "step": 6350
+    },
+    {
+      "epoch": 2.980318650421743,
+      "grad_norm": 2.641995668411255,
+      "learning_rate": 9.627460168697282e-06,
+      "loss": 0.2711,
+      "step": 6360
+    },
+    {
+      "epoch": 2.985004686035614,
+      "grad_norm": 2.4701485633850098,
+      "learning_rate": 9.626874414245549e-06,
+      "loss": 0.2458,
+      "step": 6370
+    },
+    {
+      "epoch": 2.9896907216494846,
+      "grad_norm": 3.103329658508301,
+      "learning_rate": 9.626288659793816e-06,
+      "loss": 0.2854,
+      "step": 6380
+    },
+    {
+      "epoch": 2.994376757263355,
+      "grad_norm": 2.9376931190490723,
+      "learning_rate": 9.625702905342081e-06,
+      "loss": 0.3098,
+      "step": 6390
+    },
+    {
+      "epoch": 2.999062792877226,
+      "grad_norm": 2.806793212890625,
+      "learning_rate": 9.625117150890348e-06,
+      "loss": 0.2574,
+      "step": 6400
+    },
+    {
+      "epoch": 3.0037488284910965,
+      "grad_norm": 2.9033491611480713,
+      "learning_rate": 9.624531396438613e-06,
+      "loss": 0.2101,
+      "step": 6410
+    },
+    {
+      "epoch": 3.0084348641049674,
+      "grad_norm": 2.3773772716522217,
+      "learning_rate": 9.62394564198688e-06,
+      "loss": 0.1983,
+      "step": 6420
+    },
+    {
+      "epoch": 3.013120899718838,
+      "grad_norm": 2.7549703121185303,
+      "learning_rate": 9.623359887535147e-06,
+      "loss": 0.1925,
+      "step": 6430
+    },
+    {
+      "epoch": 3.0178069353327084,
+      "grad_norm": 2.121821880340576,
+      "learning_rate": 9.622774133083412e-06,
+      "loss": 0.1893,
+      "step": 6440
+    },
+    {
+      "epoch": 3.0224929709465793,
+      "grad_norm": 1.9096688032150269,
+      "learning_rate": 9.622188378631679e-06,
+      "loss": 0.1832,
+      "step": 6450
+    },
+    {
+      "epoch": 3.02717900656045,
+      "grad_norm": 1.793131947517395,
+      "learning_rate": 9.621602624179944e-06,
+      "loss": 0.1903,
+      "step": 6460
+    },
+    {
+      "epoch": 3.0318650421743207,
+      "grad_norm": 2.6187257766723633,
+      "learning_rate": 9.62101686972821e-06,
+      "loss": 0.1846,
+      "step": 6470
+    },
+    {
+      "epoch": 3.036551077788191,
+      "grad_norm": 3.1435282230377197,
+      "learning_rate": 9.620431115276478e-06,
+      "loss": 0.2057,
+      "step": 6480
+    },
+    {
+      "epoch": 3.0412371134020617,
+      "grad_norm": 1.9570343494415283,
+      "learning_rate": 9.619845360824743e-06,
+      "loss": 0.2094,
+      "step": 6490
+    },
+    {
+      "epoch": 3.0459231490159326,
+      "grad_norm": 1.9423924684524536,
+      "learning_rate": 9.619259606373008e-06,
+      "loss": 0.1778,
+      "step": 6500
+    },
+    {
+      "epoch": 3.0459231490159326,
+      "eval_loss": 0.05262026935815811,
+      "eval_pearson_cosine": 0.7738324758503552,
+      "eval_pearson_dot": 0.6192771918549269,
+      "eval_pearson_euclidean": 0.7572697799605308,
+      "eval_pearson_manhattan": 0.7574249862527225,
+      "eval_runtime": 39.2621,
+      "eval_samples_per_second": 38.205,
+      "eval_spearman_cosine": 0.7811305074925675,
+      "eval_spearman_dot": 0.6255090522219068,
+      "eval_spearman_euclidean": 0.7739030959623808,
+      "eval_spearman_manhattan": 0.7738816457691842,
+      "eval_steps_per_second": 38.205,
+      "step": 6500
+    },
+    {
+      "epoch": 3.050609184629803,
+      "grad_norm": 2.544872760772705,
+      "learning_rate": 9.618673851921275e-06,
+      "loss": 0.2161,
+      "step": 6510
+    },
+    {
+      "epoch": 3.055295220243674,
+      "grad_norm": 3.1003711223602295,
+      "learning_rate": 9.61808809746954e-06,
+      "loss": 0.1964,
+      "step": 6520
+    },
+    {
+      "epoch": 3.0599812558575445,
+      "grad_norm": 2.2699685096740723,
+      "learning_rate": 9.617502343017808e-06,
+      "loss": 0.1992,
+      "step": 6530
+    },
+    {
+      "epoch": 3.064667291471415,
+      "grad_norm": 1.655086636543274,
+      "learning_rate": 9.616916588566074e-06,
+      "loss": 0.1834,
+      "step": 6540
+    },
+    {
+      "epoch": 3.069353327085286,
+      "grad_norm": 2.532160758972168,
+      "learning_rate": 9.61633083411434e-06,
+      "loss": 0.1978,
+      "step": 6550
+    },
+    {
+      "epoch": 3.0740393626991565,
+      "grad_norm": 3.1178247928619385,
+      "learning_rate": 9.615745079662607e-06,
+      "loss": 0.2249,
+      "step": 6560
+    },
+    {
+      "epoch": 3.0787253983130274,
+      "grad_norm": 3.530034303665161,
+      "learning_rate": 9.615159325210872e-06,
+      "loss": 0.214,
+      "step": 6570
+    },
+    {
+      "epoch": 3.083411433926898,
+      "grad_norm": 2.2964463233947754,
+      "learning_rate": 9.614573570759139e-06,
+      "loss": 0.1695,
+      "step": 6580
+    },
+    {
+      "epoch": 3.0880974695407684,
+      "grad_norm": 2.613321304321289,
+      "learning_rate": 9.613987816307406e-06,
+      "loss": 0.1888,
+      "step": 6590
+    },
+    {
+      "epoch": 3.0927835051546393,
+      "grad_norm": 3.2321712970733643,
+      "learning_rate": 9.613402061855671e-06,
+      "loss": 0.1721,
+      "step": 6600
+    },
+    {
+      "epoch": 3.0974695407685098,
+      "grad_norm": 2.779963731765747,
+      "learning_rate": 9.612816307403938e-06,
+      "loss": 0.2098,
+      "step": 6610
+    },
+    {
+      "epoch": 3.1021555763823807,
+      "grad_norm": 2.531707763671875,
+      "learning_rate": 9.612230552952203e-06,
+      "loss": 0.1906,
+      "step": 6620
+    },
+    {
+      "epoch": 3.106841611996251,
+      "grad_norm": 3.4784843921661377,
+      "learning_rate": 9.611644798500468e-06,
+      "loss": 0.1856,
+      "step": 6630
+    },
+    {
+      "epoch": 3.1115276476101217,
+      "grad_norm": 2.682420253753662,
+      "learning_rate": 9.611059044048735e-06,
+      "loss": 0.2047,
+      "step": 6640
+    },
+    {
+      "epoch": 3.1162136832239926,
+      "grad_norm": 2.4793782234191895,
+      "learning_rate": 9.610473289597002e-06,
+      "loss": 0.2062,
+      "step": 6650
+    },
+    {
+      "epoch": 3.120899718837863,
+      "grad_norm": 2.4194202423095703,
+      "learning_rate": 9.609887535145267e-06,
+      "loss": 0.184,
+      "step": 6660
+    },
+    {
+      "epoch": 3.125585754451734,
+      "grad_norm": 2.717632293701172,
+      "learning_rate": 9.609301780693534e-06,
+      "loss": 0.2148,
+      "step": 6670
+    },
+    {
+      "epoch": 3.1302717900656045,
+      "grad_norm": 2.401958703994751,
+      "learning_rate": 9.6087160262418e-06,
+      "loss": 0.1912,
+      "step": 6680
+    },
+    {
+      "epoch": 3.134957825679475,
+      "grad_norm": 2.585925340652466,
+      "learning_rate": 9.608130271790066e-06,
+      "loss": 0.1784,
+      "step": 6690
+    },
+    {
+      "epoch": 3.139643861293346,
+      "grad_norm": 3.0962131023406982,
+      "learning_rate": 9.607544517338333e-06,
+      "loss": 0.2255,
+      "step": 6700
+    },
+    {
+      "epoch": 3.1443298969072164,
+      "grad_norm": 2.8932714462280273,
+      "learning_rate": 9.606958762886598e-06,
+      "loss": 0.2022,
+      "step": 6710
+    },
+    {
+      "epoch": 3.1490159325210874,
+      "grad_norm": 3.122772693634033,
+      "learning_rate": 9.606373008434865e-06,
+      "loss": 0.224,
+      "step": 6720
+    },
+    {
+      "epoch": 3.153701968134958,
+      "grad_norm": 2.3057713508605957,
+      "learning_rate": 9.60578725398313e-06,
+      "loss": 0.1793,
+      "step": 6730
+    },
+    {
+      "epoch": 3.1583880037488283,
+      "grad_norm": 2.4614179134368896,
+      "learning_rate": 9.605201499531398e-06,
+      "loss": 0.2021,
+      "step": 6740
+    },
+    {
+      "epoch": 3.1630740393626993,
+      "grad_norm": 2.207731008529663,
+      "learning_rate": 9.604615745079663e-06,
+      "loss": 0.1791,
+      "step": 6750
+    },
+    {
+      "epoch": 3.1630740393626993,
+      "eval_loss": 0.051918212324380875,
+      "eval_pearson_cosine": 0.7727540079103647,
+      "eval_pearson_dot": 0.6115735313924304,
+      "eval_pearson_euclidean": 0.7538142936820122,
+      "eval_pearson_manhattan": 0.7540359717423257,
+      "eval_runtime": 40.2509,
+      "eval_samples_per_second": 37.266,
+      "eval_spearman_cosine": 0.7783352094358659,
+      "eval_spearman_dot": 0.6182056058746139,
+      "eval_spearman_euclidean": 0.7700146198863392,
+      "eval_spearman_manhattan": 0.7703753603565089,
+      "eval_steps_per_second": 37.266,
+      "step": 6750
+    },
+    {
+      "epoch": 3.1677600749765698,
+      "grad_norm": 3.1614291667938232,
+      "learning_rate": 9.60402999062793e-06,
+      "loss": 0.2023,
+      "step": 6760
+    },
+    {
+      "epoch": 3.1724461105904407,
+      "grad_norm": 2.1920371055603027,
+      "learning_rate": 9.603444236176197e-06,
+      "loss": 0.2102,
+      "step": 6770
+    },
+    {
+      "epoch": 3.177132146204311,
+      "grad_norm": 2.640230894088745,
+      "learning_rate": 9.602858481724462e-06,
+      "loss": 0.1945,
+      "step": 6780
+    },
+    {
+      "epoch": 3.1818181818181817,
+      "grad_norm": 2.749765396118164,
+      "learning_rate": 9.602272727272727e-06,
+      "loss": 0.1929,
+      "step": 6790
+    },
+    {
+      "epoch": 3.1865042174320526,
+      "grad_norm": 2.674724578857422,
+      "learning_rate": 9.601686972820994e-06,
+      "loss": 0.1615,
+      "step": 6800
+    },
+    {
+      "epoch": 3.191190253045923,
+      "grad_norm": 2.891061782836914,
+      "learning_rate": 9.601101218369261e-06,
+      "loss": 0.1957,
+      "step": 6810
+    },
+    {
+      "epoch": 3.195876288659794,
+      "grad_norm": 2.452010154724121,
+      "learning_rate": 9.600515463917526e-06,
+      "loss": 0.1728,
+      "step": 6820
+    },
+    {
+      "epoch": 3.2005623242736645,
+      "grad_norm": 2.2987475395202637,
+      "learning_rate": 9.599929709465793e-06,
+      "loss": 0.2001,
+      "step": 6830
+    },
+    {
+      "epoch": 3.205248359887535,
+      "grad_norm": 1.8266733884811401,
+      "learning_rate": 9.599343955014058e-06,
+      "loss": 0.178,
+      "step": 6840
+    },
+    {
+      "epoch": 3.209934395501406,
+      "grad_norm": 2.5674402713775635,
+      "learning_rate": 9.598758200562325e-06,
+      "loss": 0.1998,
+      "step": 6850
+    },
+    {
+      "epoch": 3.2146204311152764,
+      "grad_norm": 2.6671390533447266,
+      "learning_rate": 9.59817244611059e-06,
+      "loss": 0.1953,
+      "step": 6860
+    },
+    {
+      "epoch": 3.2193064667291473,
+      "grad_norm": 2.2907798290252686,
+      "learning_rate": 9.597586691658857e-06,
+      "loss": 0.2099,
+      "step": 6870
+    },
+    {
+      "epoch": 3.223992502343018,
+      "grad_norm": 3.427845001220703,
+      "learning_rate": 9.597000937207124e-06,
+      "loss": 0.2197,
+      "step": 6880
+    },
+    {
+      "epoch": 3.2286785379568883,
+      "grad_norm": 2.7697179317474365,
+      "learning_rate": 9.59641518275539e-06,
+      "loss": 0.2015,
+      "step": 6890
+    },
+    {
+      "epoch": 3.2333645735707592,
+      "grad_norm": 2.18623423576355,
+      "learning_rate": 9.595829428303656e-06,
+      "loss": 0.186,
+      "step": 6900
+    },
+    {
+      "epoch": 3.2380506091846297,
+      "grad_norm": 3.5410492420196533,
+      "learning_rate": 9.595243673851922e-06,
+      "loss": 0.1952,
+      "step": 6910
+    },
+    {
+      "epoch": 3.2427366447985007,
+      "grad_norm": 2.631042242050171,
+      "learning_rate": 9.594657919400188e-06,
+      "loss": 0.1987,
+      "step": 6920
+    },
+    {
+      "epoch": 3.247422680412371,
+      "grad_norm": 2.416839122772217,
+      "learning_rate": 9.594072164948455e-06,
+      "loss": 0.1882,
+      "step": 6930
+    },
+    {
+      "epoch": 3.2521087160262416,
+      "grad_norm": 1.9796010255813599,
+      "learning_rate": 9.59348641049672e-06,
+      "loss": 0.2133,
+      "step": 6940
+    },
+    {
+      "epoch": 3.2567947516401126,
+      "grad_norm": 2.4925854206085205,
+      "learning_rate": 9.592900656044986e-06,
+      "loss": 0.1915,
+      "step": 6950
+    },
+    {
+      "epoch": 3.261480787253983,
+      "grad_norm": 2.323317766189575,
+      "learning_rate": 9.592314901593253e-06,
+      "loss": 0.1951,
+      "step": 6960
+    },
+    {
+      "epoch": 3.266166822867854,
+      "grad_norm": 2.2892370223999023,
+      "learning_rate": 9.591729147141518e-06,
+      "loss": 0.1938,
+      "step": 6970
+    },
+    {
+      "epoch": 3.2708528584817245,
+      "grad_norm": 2.5155346393585205,
+      "learning_rate": 9.591143392689785e-06,
+      "loss": 0.2233,
+      "step": 6980
+    },
+    {
+      "epoch": 3.275538894095595,
+      "grad_norm": 2.344982624053955,
+      "learning_rate": 9.590557638238052e-06,
+      "loss": 0.2029,
+      "step": 6990
+    },
+    {
+      "epoch": 3.280224929709466,
+      "grad_norm": 2.588749408721924,
+      "learning_rate": 9.589971883786317e-06,
+      "loss": 0.201,
+      "step": 7000
+    },
+    {
+      "epoch": 3.280224929709466,
+      "eval_loss": 0.05114726722240448,
+      "eval_pearson_cosine": 0.7755109260376774,
+      "eval_pearson_dot": 0.6038894215473363,
+      "eval_pearson_euclidean": 0.7503042403542959,
+      "eval_pearson_manhattan": 0.7506062835400229,
+      "eval_runtime": 39.6185,
+      "eval_samples_per_second": 37.861,
+      "eval_spearman_cosine": 0.782480390320873,
+      "eval_spearman_dot": 0.6071370021112881,
+      "eval_spearman_euclidean": 0.7669617369314659,
+      "eval_spearman_manhattan": 0.767145329588772,
+      "eval_steps_per_second": 37.861,
+      "step": 7000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 21340,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}