diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6433 @@
+{
+  "best_metric": 0.925627350509402,
+  "best_model_checkpoint": "checkpoints/BEE-spoke-data-bert-plus-L8-v1.0-allNLI_matryoshka-synthetic-text-similarity-Mar-07_22-56/checkpoint-3600",
+  "epoch": 0.35376884422110555,
+  "eval_steps": 300,
+  "global_step": 4400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.8406219482421875,
+      "learning_rate": 8.038585209003216e-08,
+      "loss": 0.1372,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.115014553070068,
+      "learning_rate": 1.6077170418006432e-07,
+      "loss": 0.1134,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3239176273345947,
+      "learning_rate": 2.4115755627009647e-07,
+      "loss": 0.0902,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.643126964569092,
+      "learning_rate": 3.2154340836012864e-07,
+      "loss": 0.0907,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8138389587402344,
+      "learning_rate": 4.019292604501608e-07,
+      "loss": 0.087,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0662753582000732,
+      "learning_rate": 4.823151125401929e-07,
+      "loss": 0.0713,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2391765117645264,
+      "learning_rate": 5.627009646302252e-07,
+      "loss": 0.0724,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7336483001708984,
+      "learning_rate": 6.430868167202573e-07,
+      "loss": 0.0711,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5051072239875793,
+      "learning_rate": 7.234726688102894e-07,
+      "loss": 0.0386,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8692330121994019,
+      "learning_rate": 8.038585209003216e-07,
+      "loss": 0.07,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.106194257736206,
+      "learning_rate": 8.842443729903538e-07,
+      "loss": 0.0908,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4909865856170654,
+      "learning_rate": 9.646302250803859e-07,
+      "loss": 0.0632,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8705946207046509,
+      "learning_rate": 1.045016077170418e-06,
+      "loss": 0.0436,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9496204853057861,
+      "learning_rate": 1.1254019292604503e-06,
+      "loss": 0.0481,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2997841835021973,
+      "learning_rate": 1.2057877813504825e-06,
+      "loss": 0.0552,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0006277561187744,
+      "learning_rate": 1.2861736334405146e-06,
+      "loss": 0.0249,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1965364217758179,
+      "learning_rate": 1.3665594855305468e-06,
+      "loss": 0.0261,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8660438656806946,
+      "learning_rate": 1.4469453376205788e-06,
+      "loss": 0.0383,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8808113932609558,
+      "learning_rate": 1.527331189710611e-06,
+      "loss": 0.0254,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.918907880783081,
+      "learning_rate": 1.6077170418006432e-06,
+      "loss": 0.0656,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.735292673110962,
+      "learning_rate": 1.6881028938906755e-06,
+      "loss": 0.0518,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8118743896484375,
+      "learning_rate": 1.7684887459807077e-06,
+      "loss": 0.0908,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0685555934906006,
+      "learning_rate": 1.84887459807074e-06,
+      "loss": 0.0332,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8618605732917786,
+      "learning_rate": 1.9292604501607717e-06,
+      "loss": 0.0456,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7448190450668335,
+      "learning_rate": 2.0096463022508037e-06,
+      "loss": 0.0349,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6762810945510864,
+      "learning_rate": 2.090032154340836e-06,
+      "loss": 0.025,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.172909736633301,
+      "learning_rate": 2.170418006430868e-06,
+      "loss": 0.0334,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.036996364593506,
+      "learning_rate": 2.2508038585209006e-06,
+      "loss": 0.0548,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.498236894607544,
+      "learning_rate": 2.3311897106109326e-06,
+      "loss": 0.0492,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8583147525787354,
+      "learning_rate": 2.411575562700965e-06,
+      "loss": 0.05,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.241025686264038,
+      "learning_rate": 2.491961414790997e-06,
+      "loss": 0.0264,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9725112915039062,
+      "learning_rate": 2.572347266881029e-06,
+      "loss": 0.0321,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9365035891532898,
+      "learning_rate": 2.6527331189710615e-06,
+      "loss": 0.0149,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9742757081985474,
+      "learning_rate": 2.7331189710610936e-06,
+      "loss": 0.0186,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9915574193000793,
+      "learning_rate": 2.813504823151126e-06,
+      "loss": 0.0308,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.514364719390869,
+      "learning_rate": 2.8938906752411576e-06,
+      "loss": 0.0288,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.717421293258667,
+      "learning_rate": 2.97427652733119e-06,
+      "loss": 0.0371,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.118988990783691,
+      "learning_rate": 3.054662379421222e-06,
+      "loss": 0.0494,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1433117389678955,
+      "learning_rate": 3.135048231511254e-06,
+      "loss": 0.0515,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.93222975730896,
+      "learning_rate": 3.2154340836012865e-06,
+      "loss": 0.0276,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6795055270195007,
+      "learning_rate": 3.2958199356913185e-06,
+      "loss": 0.0318,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.950875997543335,
+      "learning_rate": 3.376205787781351e-06,
+      "loss": 0.0335,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8429243564605713,
+      "learning_rate": 3.456591639871383e-06,
+      "loss": 0.0434,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2567358016967773,
+      "learning_rate": 3.5369774919614154e-06,
+      "loss": 0.0324,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.407845139503479,
+      "learning_rate": 3.617363344051447e-06,
+      "loss": 0.0309,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4068539142608643,
+      "learning_rate": 3.69774919614148e-06,
+      "loss": 0.0454,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8196914196014404,
+      "learning_rate": 3.7781350482315114e-06,
+      "loss": 0.0255,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0097465515136719,
+      "learning_rate": 3.8585209003215434e-06,
+      "loss": 0.0263,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.995619773864746,
+      "learning_rate": 3.938906752411576e-06,
+      "loss": 0.0331,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.168858528137207,
+      "learning_rate": 4.0192926045016075e-06,
+      "loss": 0.0416,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8158152103424072,
+      "learning_rate": 4.09967845659164e-06,
+      "loss": 0.0385,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8158726692199707,
+      "learning_rate": 4.180064308681672e-06,
+      "loss": 0.0385,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6891674995422363,
+      "learning_rate": 4.260450160771704e-06,
+      "loss": 0.0182,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5722671747207642,
+      "learning_rate": 4.340836012861736e-06,
+      "loss": 0.0323,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8213624954223633,
+      "learning_rate": 4.421221864951769e-06,
+      "loss": 0.0351,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2886526584625244,
+      "learning_rate": 4.501607717041801e-06,
+      "loss": 0.0401,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6343109607696533,
+      "learning_rate": 4.581993569131833e-06,
+      "loss": 0.0343,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7094342708587646,
+      "learning_rate": 4.662379421221865e-06,
+      "loss": 0.0261,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2275326251983643,
+      "learning_rate": 4.742765273311897e-06,
+      "loss": 0.0183,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9329387545585632,
+      "learning_rate": 4.82315112540193e-06,
+      "loss": 0.0651,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 0.03527999296784401,
+      "eval_pearson_cosine": 0.8237974983947806,
+      "eval_pearson_dot": 0.79982187310788,
+      "eval_pearson_euclidean": 0.8049934987510711,
+      "eval_pearson_manhattan": 0.802979519881207,
+      "eval_pearson_max": 0.8237974983947806,
+      "eval_runtime": 426.4216,
+      "eval_samples_per_second": 1.173,
+      "eval_spearman_cosine": 0.8749033396133584,
+      "eval_spearman_dot": 0.8475867183468735,
+      "eval_spearman_euclidean": 0.8703331453325813,
+      "eval_spearman_manhattan": 0.8672266689066755,
+      "eval_spearman_max": 0.8749033396133584,
+      "eval_steps_per_second": 1.173,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6371708512306213,
+      "learning_rate": 4.903536977491961e-06,
+      "loss": 0.019,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.065721035003662,
+      "learning_rate": 4.983922829581994e-06,
+      "loss": 0.0193,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1992969512939453,
+      "learning_rate": 5.064308681672026e-06,
+      "loss": 0.0579,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.422349214553833,
+      "learning_rate": 5.144694533762058e-06,
+      "loss": 0.0342,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.09092378616333,
+      "learning_rate": 5.22508038585209e-06,
+      "loss": 0.0321,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6249920129776,
+      "learning_rate": 5.305466237942123e-06,
+      "loss": 0.05,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4493677616119385,
+      "learning_rate": 5.385852090032154e-06,
+      "loss": 0.0371,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1457599401474,
+      "learning_rate": 5.466237942122187e-06,
+      "loss": 0.0191,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7196341156959534,
+      "learning_rate": 5.546623794212219e-06,
+      "loss": 0.0653,
+      "step": 345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7609484791755676,
+      "learning_rate": 5.627009646302252e-06,
+      "loss": 0.0234,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9668811559677124,
+      "learning_rate": 5.707395498392283e-06,
+      "loss": 0.043,
+      "step": 355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4503982067108154,
+      "learning_rate": 5.787781350482315e-06,
+      "loss": 0.0423,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6571695804595947,
+      "learning_rate": 5.868167202572348e-06,
+      "loss": 0.0463,
+      "step": 365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0262566804885864,
+      "learning_rate": 5.94855305466238e-06,
+      "loss": 0.0362,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1059563159942627,
+      "learning_rate": 6.028938906752412e-06,
+      "loss": 0.0291,
+      "step": 375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3710355758666992,
+      "learning_rate": 6.109324758842444e-06,
+      "loss": 0.0286,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9536277055740356,
+      "learning_rate": 6.189710610932477e-06,
+      "loss": 0.048,
+      "step": 385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9812152981758118,
+      "learning_rate": 6.270096463022508e-06,
+      "loss": 0.0211,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8256497383117676,
+      "learning_rate": 6.350482315112541e-06,
+      "loss": 0.0248,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5534074306488037,
+      "learning_rate": 6.430868167202573e-06,
+      "loss": 0.0406,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0551486015319824,
+      "learning_rate": 6.511254019292606e-06,
+      "loss": 0.0299,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8460248708724976,
+      "learning_rate": 6.591639871382637e-06,
+      "loss": 0.0182,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7704716920852661,
+      "learning_rate": 6.672025723472669e-06,
+      "loss": 0.0223,
+      "step": 415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.103775024414062,
+      "learning_rate": 6.752411575562702e-06,
+      "loss": 0.0296,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.514302134513855,
+      "learning_rate": 6.832797427652733e-06,
+      "loss": 0.0302,
+      "step": 425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5787880420684814,
+      "learning_rate": 6.913183279742766e-06,
+      "loss": 0.0229,
+      "step": 430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.478020429611206,
+      "learning_rate": 6.993569131832798e-06,
+      "loss": 0.03,
+      "step": 435
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8203696608543396,
+      "learning_rate": 7.073954983922831e-06,
+      "loss": 0.0333,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.000648856163025,
+      "learning_rate": 7.154340836012862e-06,
+      "loss": 0.0177,
+      "step": 445
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5386018753051758,
+      "learning_rate": 7.234726688102894e-06,
+      "loss": 0.0351,
+      "step": 450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0292292833328247,
+      "learning_rate": 7.315112540192927e-06,
+      "loss": 0.0395,
+      "step": 455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5625340938568115,
+      "learning_rate": 7.39549839228296e-06,
+      "loss": 0.0299,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6210525035858154,
+      "learning_rate": 7.475884244372991e-06,
+      "loss": 0.0321,
+      "step": 465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8112176656723022,
+      "learning_rate": 7.556270096463023e-06,
+      "loss": 0.0316,
+      "step": 470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0322524309158325,
+      "learning_rate": 7.636655948553056e-06,
+      "loss": 0.0253,
+      "step": 475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.308305263519287,
+      "learning_rate": 7.717041800643087e-06,
+      "loss": 0.0387,
+      "step": 480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2584173679351807,
+      "learning_rate": 7.79742765273312e-06,
+      "loss": 0.0353,
+      "step": 485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6924129128456116,
+      "learning_rate": 7.877813504823153e-06,
+      "loss": 0.0157,
+      "step": 490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4199492931365967,
+      "learning_rate": 7.958199356913184e-06,
+      "loss": 0.0401,
+      "step": 495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.35832083225250244,
+      "learning_rate": 8.038585209003215e-06,
+      "loss": 0.0227,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6043404340744019,
+      "learning_rate": 8.118971061093248e-06,
+      "loss": 0.0199,
+      "step": 505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0943973064422607,
+      "learning_rate": 8.19935691318328e-06,
+      "loss": 0.0213,
+      "step": 510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8544149398803711,
+      "learning_rate": 8.279742765273314e-06,
+      "loss": 0.0226,
+      "step": 515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.021656036376953,
+      "learning_rate": 8.360128617363345e-06,
+      "loss": 0.0574,
+      "step": 520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3117215633392334,
+      "learning_rate": 8.440514469453378e-06,
+      "loss": 0.019,
+      "step": 525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2881953716278076,
+      "learning_rate": 8.520900321543409e-06,
+      "loss": 0.0267,
+      "step": 530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.115152597427368,
+      "learning_rate": 8.601286173633442e-06,
+      "loss": 0.0233,
+      "step": 535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7420244216918945,
+      "learning_rate": 8.681672025723473e-06,
+      "loss": 0.0216,
+      "step": 540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2626619338989258,
+      "learning_rate": 8.762057877813506e-06,
+      "loss": 0.0182,
+      "step": 545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5143030881881714,
+      "learning_rate": 8.842443729903538e-06,
+      "loss": 0.0261,
+      "step": 550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1002925634384155,
+      "learning_rate": 8.92282958199357e-06,
+      "loss": 0.022,
+      "step": 555
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9754141569137573,
+      "learning_rate": 9.003215434083602e-06,
+      "loss": 0.0252,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3871146440505981,
+      "learning_rate": 9.083601286173634e-06,
+      "loss": 0.0456,
+      "step": 565
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.471267580986023,
+      "learning_rate": 9.163987138263667e-06,
+      "loss": 0.0384,
+      "step": 570
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8994916677474976,
+      "learning_rate": 9.244372990353698e-06,
+      "loss": 0.036,
+      "step": 575
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7490404844284058,
+      "learning_rate": 9.32475884244373e-06,
+      "loss": 0.0221,
+      "step": 580
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3971872329711914,
+      "learning_rate": 9.405144694533763e-06,
+      "loss": 0.0302,
+      "step": 585
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9706817865371704,
+      "learning_rate": 9.485530546623795e-06,
+      "loss": 0.0291,
+      "step": 590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0623135566711426,
+      "learning_rate": 9.565916398713827e-06,
+      "loss": 0.0332,
+      "step": 595
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.330348253250122,
+      "learning_rate": 9.64630225080386e-06,
+      "loss": 0.031,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 0.03203292191028595,
+      "eval_pearson_cosine": 0.8607192041230748,
+      "eval_pearson_dot": 0.8492954726274867,
+      "eval_pearson_euclidean": 0.8425777905216318,
+      "eval_pearson_manhattan": 0.8394006469079365,
+      "eval_pearson_max": 0.8607192041230748,
+      "eval_runtime": 425.6285,
+      "eval_samples_per_second": 1.175,
+      "eval_spearman_cosine": 0.8971169644678579,
+      "eval_spearman_dot": 0.8784386497545991,
+      "eval_spearman_euclidean": 0.894069240276961,
+      "eval_spearman_manhattan": 0.8926532346129386,
+      "eval_spearman_max": 0.8971169644678579,
+      "eval_steps_per_second": 1.175,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.334221124649048,
+      "learning_rate": 9.726688102893891e-06,
+      "loss": 0.0273,
+      "step": 605
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7424949407577515,
+      "learning_rate": 9.807073954983923e-06,
+      "loss": 0.0647,
+      "step": 610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4894630908966064,
+      "learning_rate": 9.887459807073955e-06,
+      "loss": 0.0361,
+      "step": 615
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4873441159725189,
+      "learning_rate": 9.967845659163988e-06,
+      "loss": 0.0264,
+      "step": 620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5973559617996216,
+      "learning_rate": 1.0048231511254021e-05,
+      "loss": 0.029,
+      "step": 625
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1919150352478027,
+      "learning_rate": 1.0128617363344052e-05,
+      "loss": 0.0328,
+      "step": 630
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3969812393188477,
+      "learning_rate": 1.0209003215434084e-05,
+      "loss": 0.043,
+      "step": 635
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.33707332611084,
+      "learning_rate": 1.0289389067524116e-05,
+      "loss": 0.0448,
+      "step": 640
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3516573905944824,
+      "learning_rate": 1.0369774919614148e-05,
+      "loss": 0.0328,
+      "step": 645
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.014158248901367,
+      "learning_rate": 1.045016077170418e-05,
+      "loss": 0.0326,
+      "step": 650
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1477378606796265,
+      "learning_rate": 1.0530546623794213e-05,
+      "loss": 0.025,
+      "step": 655
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8221299648284912,
+      "learning_rate": 1.0610932475884246e-05,
+      "loss": 0.0425,
+      "step": 660
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.68177330493927,
+      "learning_rate": 1.0691318327974279e-05,
+      "loss": 0.025,
+      "step": 665
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4552159309387207,
+      "learning_rate": 1.0771704180064308e-05,
+      "loss": 0.0398,
+      "step": 670
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.352478265762329,
+      "learning_rate": 1.0852090032154341e-05,
+      "loss": 0.0155,
+      "step": 675
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9538177251815796,
+      "learning_rate": 1.0932475884244374e-05,
+      "loss": 0.0299,
+      "step": 680
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0173323154449463,
+      "learning_rate": 1.1012861736334405e-05,
+      "loss": 0.0268,
+      "step": 685
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0908379554748535,
+      "learning_rate": 1.1093247588424438e-05,
+      "loss": 0.0308,
+      "step": 690
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6585212349891663,
+      "learning_rate": 1.1173633440514471e-05,
+      "loss": 0.025,
+      "step": 695
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.959600567817688,
+      "learning_rate": 1.1254019292604504e-05,
+      "loss": 0.0219,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2033023834228516,
+      "learning_rate": 1.1334405144694535e-05,
+      "loss": 0.0315,
+      "step": 705
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.735729455947876,
+      "learning_rate": 1.1414790996784566e-05,
+      "loss": 0.0249,
+      "step": 710
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.850893497467041,
+      "learning_rate": 1.14951768488746e-05,
+      "loss": 0.0248,
+      "step": 715
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.902856171131134,
+      "learning_rate": 1.157556270096463e-05,
+      "loss": 0.0683,
+      "step": 720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6975336670875549,
+      "learning_rate": 1.1655948553054663e-05,
+      "loss": 0.0315,
+      "step": 725
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3982033729553223,
+      "learning_rate": 1.1736334405144696e-05,
+      "loss": 0.0479,
+      "step": 730
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.775419235229492,
+      "learning_rate": 1.1816720257234729e-05,
+      "loss": 0.068,
+      "step": 735
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9351024627685547,
+      "learning_rate": 1.189710610932476e-05,
+      "loss": 0.0254,
+      "step": 740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5352950096130371,
+      "learning_rate": 1.1977491961414791e-05,
+      "loss": 0.0193,
+      "step": 745
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5528688430786133,
+      "learning_rate": 1.2057877813504824e-05,
+      "loss": 0.039,
+      "step": 750
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6287634372711182,
+      "learning_rate": 1.2138263665594855e-05,
+      "loss": 0.017,
+      "step": 755
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0720481872558594,
+      "learning_rate": 1.2218649517684888e-05,
+      "loss": 0.0261,
+      "step": 760
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8808945417404175,
+      "learning_rate": 1.2299035369774921e-05,
+      "loss": 0.0224,
+      "step": 765
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.175379753112793,
+      "learning_rate": 1.2379421221864954e-05,
+      "loss": 0.0157,
+      "step": 770
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7339066863059998,
+      "learning_rate": 1.2459807073954987e-05,
+      "loss": 0.0242,
+      "step": 775
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5107578039169312,
+      "learning_rate": 1.2540192926045016e-05,
+      "loss": 0.0245,
+      "step": 780
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5132850408554077,
+      "learning_rate": 1.2620578778135049e-05,
+      "loss": 0.0191,
+      "step": 785
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7692009210586548,
+      "learning_rate": 1.2700964630225082e-05,
+      "loss": 0.019,
+      "step": 790
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1860913038253784,
+      "learning_rate": 1.2781350482315113e-05,
+      "loss": 0.0121,
+      "step": 795
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9162280559539795,
+      "learning_rate": 1.2861736334405146e-05,
+      "loss": 0.0195,
+      "step": 800
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2460323572158813,
+      "learning_rate": 1.2942122186495179e-05,
+      "loss": 0.0223,
+      "step": 805
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7504421472549438,
+      "learning_rate": 1.3022508038585212e-05,
+      "loss": 0.0325,
+      "step": 810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3735928535461426,
+      "learning_rate": 1.3102893890675241e-05,
+      "loss": 0.0213,
+      "step": 815
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1271485090255737,
+      "learning_rate": 1.3183279742765274e-05,
+      "loss": 0.0219,
+      "step": 820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8849589228630066,
+      "learning_rate": 1.3263665594855307e-05,
+      "loss": 0.0143,
+      "step": 825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7527520060539246,
+      "learning_rate": 1.3344051446945338e-05,
+      "loss": 0.0241,
+      "step": 830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0900169610977173,
+      "learning_rate": 1.3424437299035371e-05,
+      "loss": 0.0414,
+      "step": 835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1919902563095093,
+      "learning_rate": 1.3504823151125404e-05,
+      "loss": 0.0234,
+      "step": 840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8385984301567078,
+      "learning_rate": 1.3585209003215437e-05,
+      "loss": 0.0239,
+      "step": 845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9777655005455017,
+      "learning_rate": 1.3665594855305466e-05,
+      "loss": 0.0145,
+      "step": 850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2803690433502197,
+      "learning_rate": 1.3745980707395499e-05,
+      "loss": 0.0274,
+      "step": 855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1460819244384766,
+      "learning_rate": 1.3826366559485532e-05,
+      "loss": 0.0373,
+      "step": 860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8751824498176575,
+      "learning_rate": 1.3906752411575563e-05,
+      "loss": 0.0204,
+      "step": 865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.566417932510376,
+      "learning_rate": 1.3987138263665596e-05,
+      "loss": 0.0139,
+      "step": 870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6487017869949341,
+      "learning_rate": 1.4067524115755629e-05,
+      "loss": 0.0458,
+      "step": 875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9419571161270142,
+      "learning_rate": 1.4147909967845662e-05,
+      "loss": 0.0215,
+      "step": 880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7554964423179626,
+      "learning_rate": 1.4228295819935693e-05,
+      "loss": 0.0219,
+      "step": 885
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.240870475769043,
+      "learning_rate": 1.4308681672025724e-05,
+      "loss": 0.0291,
+      "step": 890
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2490875720977783,
+      "learning_rate": 1.4389067524115757e-05,
+      "loss": 0.0256,
+      "step": 895
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9301912188529968,
+      "learning_rate": 1.4469453376205788e-05,
+      "loss": 0.0303,
+      "step": 900
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 0.0409804992377758,
+      "eval_pearson_cosine": 0.8657545592954407,
+      "eval_pearson_dot": 0.8604614808110218,
+      "eval_pearson_euclidean": 0.849931010036339,
+      "eval_pearson_manhattan": 0.84985721810182,
+      "eval_pearson_max": 0.8657545592954407,
+      "eval_runtime": 425.4053,
+      "eval_samples_per_second": 1.175,
+      "eval_spearman_cosine": 0.9034730538922157,
+      "eval_spearman_dot": 0.8936158304633218,
+      "eval_spearman_euclidean": 0.9027816591266363,
+      "eval_spearman_manhattan": 0.9029234516938067,
+      "eval_spearman_max": 0.9034730538922157,
+      "eval_steps_per_second": 1.175,
+      "step": 900
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6353436708450317,
+      "learning_rate": 1.454983922829582e-05,
+      "loss": 0.0306,
+      "step": 905
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5071334838867188,
+      "learning_rate": 1.4630225080385854e-05,
+      "loss": 0.0529,
+      "step": 910
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.511518716812134,
+      "learning_rate": 1.4710610932475886e-05,
+      "loss": 0.0351,
+      "step": 915
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8150439858436584,
+      "learning_rate": 1.479099678456592e-05,
+      "loss": 0.0304,
+      "step": 920
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.909027636051178,
+      "learning_rate": 1.4871382636655949e-05,
+      "loss": 0.0306,
+      "step": 925
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.117823362350464,
+      "learning_rate": 1.4951768488745982e-05,
+      "loss": 0.0444,
+      "step": 930
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0820565223693848,
+      "learning_rate": 1.5032154340836015e-05,
+      "loss": 0.0147,
+      "step": 935
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0721031427383423,
+      "learning_rate": 1.5112540192926046e-05,
+      "loss": 0.0379,
+      "step": 940
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6273934841156006,
+      "learning_rate": 1.5192926045016079e-05,
+      "loss": 0.0202,
+      "step": 945
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4175567626953125,
+      "learning_rate": 1.527331189710611e-05,
+      "loss": 0.0265,
+      "step": 950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7840615510940552,
+      "learning_rate": 1.5353697749196143e-05,
+      "loss": 0.0194,
+      "step": 955
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6450772285461426,
+      "learning_rate": 1.5434083601286174e-05,
+      "loss": 0.0271,
+      "step": 960
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1013175249099731,
+      "learning_rate": 1.5514469453376205e-05,
+      "loss": 0.035,
+      "step": 965
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6244901418685913,
+      "learning_rate": 1.559485530546624e-05,
+      "loss": 0.0183,
+      "step": 970
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0193443298339844,
+      "learning_rate": 1.567524115755627e-05,
+      "loss": 0.0234,
+      "step": 975
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1802617311477661,
+      "learning_rate": 1.5755627009646305e-05,
+      "loss": 0.0302,
+      "step": 980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1835167407989502,
+      "learning_rate": 1.5836012861736336e-05,
+      "loss": 0.0253,
+      "step": 985
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4406228065490723,
+      "learning_rate": 1.5916398713826368e-05,
+      "loss": 0.0316,
+      "step": 990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9229764938354492,
+      "learning_rate": 1.59967845659164e-05,
+      "loss": 0.0408,
+      "step": 995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6133905649185181,
+      "learning_rate": 1.607717041800643e-05,
+      "loss": 0.0163,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.388720154762268,
+      "learning_rate": 1.6157556270096464e-05,
+      "loss": 0.0221,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9162605404853821,
+      "learning_rate": 1.6237942122186496e-05,
+      "loss": 0.0292,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0816845893859863,
+      "learning_rate": 1.631832797427653e-05,
+      "loss": 0.0284,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.974607527256012,
+      "learning_rate": 1.639871382636656e-05,
+      "loss": 0.0245,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7858002185821533,
+      "learning_rate": 1.6479099678456592e-05,
+      "loss": 0.0208,
+      "step": 1025
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8541093468666077,
+      "learning_rate": 1.6559485530546627e-05,
+      "loss": 0.0402,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9254925847053528,
+      "learning_rate": 1.6639871382636655e-05,
+      "loss": 0.0277,
+      "step": 1035
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1855261325836182,
+      "learning_rate": 1.672025723472669e-05,
+      "loss": 0.0289,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1872082948684692,
+      "learning_rate": 1.680064308681672e-05,
+      "loss": 0.0255,
+      "step": 1045
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9865713119506836,
+      "learning_rate": 1.6881028938906755e-05,
+      "loss": 0.0331,
+      "step": 1050
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5213806629180908,
+      "learning_rate": 1.6961414790996786e-05,
+      "loss": 0.0385,
+      "step": 1055
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5804238319396973,
+      "learning_rate": 1.7041800643086817e-05,
+      "loss": 0.0447,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2709295749664307,
+      "learning_rate": 1.7122186495176852e-05,
+      "loss": 0.0184,
+      "step": 1065
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6574845314025879,
+      "learning_rate": 1.7202572347266883e-05,
+      "loss": 0.0201,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.590877056121826,
+      "learning_rate": 1.7282958199356914e-05,
+      "loss": 0.0456,
+      "step": 1075
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.060797929763794,
+      "learning_rate": 1.7363344051446945e-05,
+      "loss": 0.0375,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.38614028692245483,
+      "learning_rate": 1.744372990353698e-05,
+      "loss": 0.0371,
+      "step": 1085
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6158710718154907,
+      "learning_rate": 1.752411575562701e-05,
+      "loss": 0.0403,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5008140802383423,
+      "learning_rate": 1.7604501607717042e-05,
+      "loss": 0.0296,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41157564520835876,
+      "learning_rate": 1.7684887459807077e-05,
+      "loss": 0.0255,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3567136526107788,
+      "learning_rate": 1.7765273311897108e-05,
+      "loss": 0.0217,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.698455810546875,
+      "learning_rate": 1.784565916398714e-05,
+      "loss": 0.0188,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4908955097198486,
+      "learning_rate": 1.792604501607717e-05,
+      "loss": 0.0188,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2162156105041504,
+      "learning_rate": 1.8006430868167205e-05,
+      "loss": 0.02,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3919101655483246,
+      "learning_rate": 1.8086816720257236e-05,
+      "loss": 0.0133,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4801468551158905,
+      "learning_rate": 1.8167202572347267e-05,
+      "loss": 0.0303,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5365796685218811,
+      "learning_rate": 1.8247588424437302e-05,
+      "loss": 0.0275,
+      "step": 1135
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.777245819568634,
+      "learning_rate": 1.8327974276527333e-05,
+      "loss": 0.0273,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5682530403137207,
+      "learning_rate": 1.8408360128617364e-05,
+      "loss": 0.0261,
+      "step": 1145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0468765497207642,
+      "learning_rate": 1.8488745980707395e-05,
+      "loss": 0.0288,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2527729272842407,
+      "learning_rate": 1.856913183279743e-05,
+      "loss": 0.0239,
+      "step": 1155
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5002691149711609,
+      "learning_rate": 1.864951768488746e-05,
+      "loss": 0.0219,
+      "step": 1160
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9675867557525635,
+      "learning_rate": 1.8729903536977496e-05,
+      "loss": 0.0476,
+      "step": 1165
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8520388007164001,
+      "learning_rate": 1.8810289389067527e-05,
+      "loss": 0.0258,
+      "step": 1170
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5552634596824646,
+      "learning_rate": 1.8890675241157558e-05,
+      "loss": 0.0237,
+      "step": 1175
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6992123126983643,
+      "learning_rate": 1.897106109324759e-05,
+      "loss": 0.0265,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1183786392211914,
+      "learning_rate": 1.905144694533762e-05,
+      "loss": 0.027,
+      "step": 1185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0854079723358154,
+      "learning_rate": 1.9131832797427655e-05,
+      "loss": 0.0252,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8095922470092773,
+      "learning_rate": 1.9212218649517686e-05,
+      "loss": 0.037,
+      "step": 1195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6434860229492188,
+      "learning_rate": 1.929260450160772e-05,
+      "loss": 0.0277,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 0.030386893078684807,
+      "eval_pearson_cosine": 0.8622394577970579,
+      "eval_pearson_dot": 0.8570136493582737,
+      "eval_pearson_euclidean": 0.849571905885398,
+      "eval_pearson_manhattan": 0.8505210712126229,
+      "eval_pearson_max": 0.8622394577970579,
+      "eval_runtime": 425.8405,
+      "eval_samples_per_second": 1.174,
+      "eval_spearman_cosine": 0.8911063804255216,
+      "eval_spearman_dot": 0.8816442945771783,
+      "eval_spearman_euclidean": 0.8895569342277367,
+      "eval_spearman_manhattan": 0.890007464029856,
+      "eval_spearman_max": 0.8911063804255216,
+      "eval_steps_per_second": 1.174,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8423680067062378,
+      "learning_rate": 1.9372990353697752e-05,
+      "loss": 0.0313,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2395281791687012,
+      "learning_rate": 1.9453376205787783e-05,
+      "loss": 0.0297,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.149891972541809,
+      "learning_rate": 1.9533762057877814e-05,
+      "loss": 0.0208,
+      "step": 1215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4256904125213623,
+      "learning_rate": 1.9614147909967845e-05,
+      "loss": 0.0437,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7631718516349792,
+      "learning_rate": 1.969453376205788e-05,
+      "loss": 0.0335,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0927650928497314,
+      "learning_rate": 1.977491961414791e-05,
+      "loss": 0.0349,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7310745120048523,
+      "learning_rate": 1.9855305466237946e-05,
+      "loss": 0.0305,
+      "step": 1235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.624340534210205,
+      "learning_rate": 1.9935691318327977e-05,
+      "loss": 0.0456,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.09121835231781,
+      "learning_rate": 1.999821316894488e-05,
+      "loss": 0.0119,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6216334104537964,
+      "learning_rate": 1.998927901366926e-05,
+      "loss": 0.0419,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0775346755981445,
+      "learning_rate": 1.998034485839364e-05,
+      "loss": 0.0317,
+      "step": 1255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.044417381286621,
+      "learning_rate": 1.997141070311802e-05,
+      "loss": 0.0245,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.581361174583435,
+      "learning_rate": 1.9962476547842403e-05,
+      "loss": 0.0246,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4875144362449646,
+      "learning_rate": 1.9953542392566784e-05,
+      "loss": 0.0218,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2498222589492798,
+      "learning_rate": 1.9944608237291165e-05,
+      "loss": 0.04,
+      "step": 1275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.562926173210144,
+      "learning_rate": 1.993567408201555e-05,
+      "loss": 0.0268,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5474754571914673,
+      "learning_rate": 1.992673992673993e-05,
+      "loss": 0.0514,
+      "step": 1285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6975357532501221,
+      "learning_rate": 1.9917805771464308e-05,
+      "loss": 0.0345,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0237133502960205,
+      "learning_rate": 1.9908871616188692e-05,
+      "loss": 0.0523,
+      "step": 1295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.604995846748352,
+      "learning_rate": 1.9899937460913073e-05,
+      "loss": 0.0279,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3883243799209595,
+      "learning_rate": 1.9891003305637454e-05,
+      "loss": 0.0227,
+      "step": 1305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5331462621688843,
+      "learning_rate": 1.9882069150361835e-05,
+      "loss": 0.0326,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.44855189323425293,
+      "learning_rate": 1.9873134995086216e-05,
+      "loss": 0.0352,
+      "step": 1315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.213827610015869,
+      "learning_rate": 1.9864200839810597e-05,
+      "loss": 0.0432,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5794016718864441,
+      "learning_rate": 1.985526668453498e-05,
+      "loss": 0.0245,
+      "step": 1325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6042608618736267,
+      "learning_rate": 1.984633252925936e-05,
+      "loss": 0.0168,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.935172200202942,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.0347,
+      "step": 1335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.000755786895752,
+      "learning_rate": 1.982846421870812e-05,
+      "loss": 0.0302,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.054868698120117,
+      "learning_rate": 1.9819530063432503e-05,
+      "loss": 0.0622,
+      "step": 1345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.873913049697876,
+      "learning_rate": 1.9810595908156884e-05,
+      "loss": 0.0178,
+      "step": 1350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6479474306106567,
+      "learning_rate": 1.9801661752881268e-05,
+      "loss": 0.0204,
+      "step": 1355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0146560668945312,
+      "learning_rate": 1.979272759760565e-05,
+      "loss": 0.0312,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7326614856719971,
+      "learning_rate": 1.978379344233003e-05,
+      "loss": 0.0444,
+      "step": 1365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8839195370674133,
+      "learning_rate": 1.977485928705441e-05,
+      "loss": 0.0301,
+      "step": 1370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.485784113407135,
+      "learning_rate": 1.9765925131778792e-05,
+      "loss": 0.0349,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7656601071357727,
+      "learning_rate": 1.9756990976503173e-05,
+      "loss": 0.0282,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6562867164611816,
+      "learning_rate": 1.9748056821227554e-05,
+      "loss": 0.033,
+      "step": 1385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.971908688545227,
+      "learning_rate": 1.9739122665951935e-05,
+      "loss": 0.0358,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4427505731582642,
+      "learning_rate": 1.9730188510676316e-05,
+      "loss": 0.029,
+      "step": 1395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.177832841873169,
+      "learning_rate": 1.9721254355400697e-05,
+      "loss": 0.0296,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5108907222747803,
+      "learning_rate": 1.971232020012508e-05,
+      "loss": 0.0239,
+      "step": 1405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7967244386672974,
+      "learning_rate": 1.9703386044849463e-05,
+      "loss": 0.0336,
+      "step": 1410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.008859395980835,
+      "learning_rate": 1.9694451889573844e-05,
+      "loss": 0.0197,
+      "step": 1415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1244670152664185,
+      "learning_rate": 1.968551773429822e-05,
+      "loss": 0.0407,
+      "step": 1420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6256637573242188,
+      "learning_rate": 1.9676583579022606e-05,
+      "loss": 0.0464,
+      "step": 1425
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6180126667022705,
+      "learning_rate": 1.9667649423746987e-05,
+      "loss": 0.0185,
+      "step": 1430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8708995580673218,
+      "learning_rate": 1.9658715268471368e-05,
+      "loss": 0.0206,
+      "step": 1435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.933128833770752,
+      "learning_rate": 1.964978111319575e-05,
+      "loss": 0.0427,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.588142991065979,
+      "learning_rate": 1.964084695792013e-05,
+      "loss": 0.0242,
+      "step": 1445
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.38248518109321594,
+      "learning_rate": 1.963191280264451e-05,
+      "loss": 0.0252,
+      "step": 1450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8211073279380798,
+      "learning_rate": 1.9622978647368892e-05,
+      "loss": 0.0159,
+      "step": 1455
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4323325455188751,
+      "learning_rate": 1.9614044492093273e-05,
+      "loss": 0.0383,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7777843475341797,
+      "learning_rate": 1.9605110336817658e-05,
+      "loss": 0.0399,
+      "step": 1465
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7342023253440857,
+      "learning_rate": 1.9596176181542035e-05,
+      "loss": 0.0232,
+      "step": 1470
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.43286553025245667,
+      "learning_rate": 1.9587242026266416e-05,
+      "loss": 0.03,
+      "step": 1475
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4914606809616089,
+      "learning_rate": 1.95783078709908e-05,
+      "loss": 0.0253,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6109514236450195,
+      "learning_rate": 1.9569373715715182e-05,
+      "loss": 0.0244,
+      "step": 1485
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.099238395690918,
+      "learning_rate": 1.9560439560439563e-05,
+      "loss": 0.0219,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.672016978263855,
+      "learning_rate": 1.9551505405163944e-05,
+      "loss": 0.0156,
+      "step": 1495
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9910028576850891,
+      "learning_rate": 1.9542571249888325e-05,
+      "loss": 0.025,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 0.029246093705296516,
+      "eval_pearson_cosine": 0.8662779121023975,
+      "eval_pearson_dot": 0.8609972822698294,
+      "eval_pearson_euclidean": 0.8525960773878268,
+      "eval_pearson_manhattan": 0.8525524127391249,
+      "eval_pearson_max": 0.8662779121023975,
+      "eval_runtime": 425.5831,
+      "eval_samples_per_second": 1.175,
+      "eval_spearman_cosine": 0.8936378145512582,
+      "eval_spearman_dot": 0.8876585266341065,
+      "eval_spearman_euclidean": 0.891318445273781,
+      "eval_spearman_manhattan": 0.889811431245725,
+      "eval_spearman_max": 0.8936378145512582,
+      "eval_steps_per_second": 1.175,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7004162073135376,
+      "learning_rate": 1.9533637094612706e-05,
+      "loss": 0.0248,
+      "step": 1505
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0799355506896973,
+      "learning_rate": 1.9524702939337087e-05,
+      "loss": 0.0268,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8577656745910645,
+      "learning_rate": 1.9515768784061468e-05,
+      "loss": 0.0363,
+      "step": 1515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.811858892440796,
+      "learning_rate": 1.950683462878585e-05,
+      "loss": 0.0408,
+      "step": 1520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2648321390151978,
+      "learning_rate": 1.949790047351023e-05,
+      "loss": 0.0273,
+      "step": 1525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6332646012306213,
+      "learning_rate": 1.948896631823461e-05,
+      "loss": 0.0234,
+      "step": 1530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1494219303131104,
+      "learning_rate": 1.9480032162958992e-05,
+      "loss": 0.0396,
+      "step": 1535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5537444353103638,
+      "learning_rate": 1.9471098007683377e-05,
+      "loss": 0.0392,
+      "step": 1540
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7606208324432373,
+      "learning_rate": 1.9462163852407754e-05,
+      "loss": 0.0468,
+      "step": 1545
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3266965448856354,
+      "learning_rate": 1.945322969713214e-05,
+      "loss": 0.0164,
+      "step": 1550
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4910846948623657,
+      "learning_rate": 1.944429554185652e-05,
+      "loss": 0.0181,
+      "step": 1555
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.621263861656189,
+      "learning_rate": 1.94353613865809e-05,
+      "loss": 0.021,
+      "step": 1560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8153178691864014,
+      "learning_rate": 1.9426427231305282e-05,
+      "loss": 0.019,
+      "step": 1565
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9617846012115479,
+      "learning_rate": 1.9417493076029663e-05,
+      "loss": 0.0186,
+      "step": 1570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6482746005058289,
+      "learning_rate": 1.9408558920754044e-05,
+      "loss": 0.0226,
+      "step": 1575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6985549926757812,
+      "learning_rate": 1.9399624765478425e-05,
+      "loss": 0.0267,
+      "step": 1580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9519606828689575,
+      "learning_rate": 1.9390690610202806e-05,
+      "loss": 0.0187,
+      "step": 1585
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9981301426887512,
+      "learning_rate": 1.9381756454927187e-05,
+      "loss": 0.0195,
+      "step": 1590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5453444719314575,
+      "learning_rate": 1.937282229965157e-05,
+      "loss": 0.0185,
+      "step": 1595
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3481849431991577,
+      "learning_rate": 1.936388814437595e-05,
+      "loss": 0.0439,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.49602463841438293,
+      "learning_rate": 1.935495398910033e-05,
+      "loss": 0.0158,
+      "step": 1605
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.477524071931839,
+      "learning_rate": 1.9346019833824715e-05,
+      "loss": 0.0276,
+      "step": 1610
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8374057412147522,
+      "learning_rate": 1.9337085678549096e-05,
+      "loss": 0.0259,
+      "step": 1615
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9524512887001038,
+      "learning_rate": 1.9328151523273477e-05,
+      "loss": 0.0403,
+      "step": 1620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6074572205543518,
+      "learning_rate": 1.9319217367997858e-05,
+      "loss": 0.0217,
+      "step": 1625
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7999153137207031,
+      "learning_rate": 1.931028321272224e-05,
+      "loss": 0.0234,
+      "step": 1630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9614180326461792,
+      "learning_rate": 1.930134905744662e-05,
+      "loss": 0.0231,
+      "step": 1635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1887145042419434,
+      "learning_rate": 1.9292414902171e-05,
+      "loss": 0.0414,
+      "step": 1640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8795381784439087,
+      "learning_rate": 1.9283480746895382e-05,
+      "loss": 0.0208,
+      "step": 1645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3685334920883179,
+      "learning_rate": 1.9274546591619766e-05,
+      "loss": 0.054,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7518516778945923,
+      "learning_rate": 1.9265612436344144e-05,
+      "loss": 0.0274,
+      "step": 1655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.029801607131958,
+      "learning_rate": 1.9256678281068525e-05,
+      "loss": 0.0174,
+      "step": 1660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7995644807815552,
+      "learning_rate": 1.924774412579291e-05,
+      "loss": 0.037,
+      "step": 1665
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5098012685775757,
+      "learning_rate": 1.923880997051729e-05,
+      "loss": 0.0216,
+      "step": 1670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5958812832832336,
+      "learning_rate": 1.9229875815241668e-05,
+      "loss": 0.0142,
+      "step": 1675
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7016432285308838,
+      "learning_rate": 1.9220941659966053e-05,
+      "loss": 0.0416,
+      "step": 1680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5232900977134705,
+      "learning_rate": 1.9212007504690434e-05,
+      "loss": 0.0264,
+      "step": 1685
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4362630248069763,
+      "learning_rate": 1.9203073349414815e-05,
+      "loss": 0.022,
+      "step": 1690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.451072335243225,
+      "learning_rate": 1.9194139194139196e-05,
+      "loss": 0.0262,
+      "step": 1695
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5889037251472473,
+      "learning_rate": 1.9185205038863577e-05,
+      "loss": 0.0267,
+      "step": 1700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.49965912103652954,
+      "learning_rate": 1.9176270883587958e-05,
+      "loss": 0.0159,
+      "step": 1705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.920385479927063,
+      "learning_rate": 1.916733672831234e-05,
+      "loss": 0.0357,
+      "step": 1710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5249290466308594,
+      "learning_rate": 1.915840257303672e-05,
+      "loss": 0.0213,
+      "step": 1715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5141476988792419,
+      "learning_rate": 1.91494684177611e-05,
+      "loss": 0.0196,
+      "step": 1720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5003548264503479,
+      "learning_rate": 1.9140534262485485e-05,
+      "loss": 0.0157,
+      "step": 1725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4535447657108307,
+      "learning_rate": 1.9131600107209863e-05,
+      "loss": 0.019,
+      "step": 1730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4489072263240814,
+      "learning_rate": 1.9122665951934247e-05,
+      "loss": 0.0168,
+      "step": 1735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5539095997810364,
+      "learning_rate": 1.911373179665863e-05,
+      "loss": 0.02,
+      "step": 1740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3125584125518799,
+      "learning_rate": 1.910479764138301e-05,
+      "loss": 0.0202,
+      "step": 1745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6378400921821594,
+      "learning_rate": 1.909586348610739e-05,
+      "loss": 0.0289,
+      "step": 1750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.118848443031311,
+      "learning_rate": 1.908692933083177e-05,
+      "loss": 0.0242,
+      "step": 1755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8065999746322632,
+      "learning_rate": 1.9077995175556153e-05,
+      "loss": 0.0277,
+      "step": 1760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9087148904800415,
+      "learning_rate": 1.9069061020280534e-05,
+      "loss": 0.0239,
+      "step": 1765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6891593933105469,
+      "learning_rate": 1.9060126865004915e-05,
+      "loss": 0.0217,
+      "step": 1770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7434559464454651,
+      "learning_rate": 1.9051192709729296e-05,
+      "loss": 0.0145,
+      "step": 1775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.585472047328949,
+      "learning_rate": 1.904225855445368e-05,
+      "loss": 0.0287,
+      "step": 1780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.561167597770691,
+      "learning_rate": 1.9033324399178058e-05,
+      "loss": 0.0206,
+      "step": 1785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9487882852554321,
+      "learning_rate": 1.902439024390244e-05,
+      "loss": 0.0212,
+      "step": 1790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7181324362754822,
+      "learning_rate": 1.9015456088626823e-05,
+      "loss": 0.0241,
+      "step": 1795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5083106160163879,
+      "learning_rate": 1.9006521933351204e-05,
+      "loss": 0.0159,
+      "step": 1800
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 0.030221089720726013,
+      "eval_pearson_cosine": 0.8857723004603547,
+      "eval_pearson_dot": 0.874834502431298,
+      "eval_pearson_euclidean": 0.8749650474368084,
+      "eval_pearson_manhattan": 0.8746243487782971,
+      "eval_pearson_max": 0.8857723004603547,
+      "eval_runtime": 426.1522,
+      "eval_samples_per_second": 1.173,
+      "eval_spearman_cosine": 0.9000108960435841,
+      "eval_spearman_dot": 0.8859179436717747,
+      "eval_spearman_euclidean": 0.9029959319837279,
+      "eval_spearman_manhattan": 0.9019226476905907,
+      "eval_spearman_max": 0.9029959319837279,
+      "eval_steps_per_second": 1.173,
+      "step": 1800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.43535247445106506,
+      "learning_rate": 1.8997587778075582e-05,
+      "loss": 0.0153,
+      "step": 1805
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1544151306152344,
+      "learning_rate": 1.8988653622799966e-05,
+      "loss": 0.0241,
+      "step": 1810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9339686632156372,
+      "learning_rate": 1.8979719467524347e-05,
+      "loss": 0.0439,
+      "step": 1815
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8600350618362427,
+      "learning_rate": 1.897078531224873e-05,
+      "loss": 0.0268,
+      "step": 1820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7343426942825317,
+      "learning_rate": 1.896185115697311e-05,
+      "loss": 0.0134,
+      "step": 1825
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.237486481666565,
+      "learning_rate": 1.895291700169749e-05,
+      "loss": 0.0387,
+      "step": 1830
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8042169213294983,
+      "learning_rate": 1.8943982846421875e-05,
+      "loss": 0.0218,
+      "step": 1835
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5267782211303711,
+      "learning_rate": 1.8935048691146253e-05,
+      "loss": 0.0277,
+      "step": 1840
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7091891169548035,
+      "learning_rate": 1.8926114535870634e-05,
+      "loss": 0.0196,
+      "step": 1845
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3862394094467163,
+      "learning_rate": 1.8917180380595018e-05,
+      "loss": 0.0173,
+      "step": 1850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8204643726348877,
+      "learning_rate": 1.89082462253194e-05,
+      "loss": 0.0174,
+      "step": 1855
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6407317519187927,
+      "learning_rate": 1.8899312070043777e-05,
+      "loss": 0.0152,
+      "step": 1860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6078154444694519,
+      "learning_rate": 1.889037791476816e-05,
+      "loss": 0.0287,
+      "step": 1865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8564119338989258,
+      "learning_rate": 1.8881443759492542e-05,
+      "loss": 0.0185,
+      "step": 1870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7499640583992004,
+      "learning_rate": 1.8872509604216923e-05,
+      "loss": 0.04,
+      "step": 1875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5337907671928406,
+      "learning_rate": 1.8863575448941304e-05,
+      "loss": 0.0157,
+      "step": 1880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.525449275970459,
+      "learning_rate": 1.8854641293665685e-05,
+      "loss": 0.0252,
+      "step": 1885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.069151759147644,
+      "learning_rate": 1.8845707138390066e-05,
+      "loss": 0.0261,
+      "step": 1890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5704050660133362,
+      "learning_rate": 1.8836772983114447e-05,
+      "loss": 0.0228,
+      "step": 1895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8183658719062805,
+      "learning_rate": 1.882783882783883e-05,
+      "loss": 0.0293,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5661863088607788,
+      "learning_rate": 1.881890467256321e-05,
+      "loss": 0.0214,
+      "step": 1905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24864475429058075,
+      "learning_rate": 1.8809970517287594e-05,
+      "loss": 0.0169,
+      "step": 1910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.622249960899353,
+      "learning_rate": 1.880103636201197e-05,
+      "loss": 0.0267,
+      "step": 1915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6661076545715332,
+      "learning_rate": 1.8792102206736356e-05,
+      "loss": 0.0287,
+      "step": 1920
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.737390398979187,
+      "learning_rate": 1.8783168051460737e-05,
+      "loss": 0.0179,
+      "step": 1925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6081177592277527,
+      "learning_rate": 1.8774233896185118e-05,
+      "loss": 0.0194,
+      "step": 1930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8305076360702515,
+      "learning_rate": 1.87652997409095e-05,
+      "loss": 0.0333,
+      "step": 1935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7024769186973572,
+      "learning_rate": 1.875636558563388e-05,
+      "loss": 0.0272,
+      "step": 1940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7568246722221375,
+      "learning_rate": 1.874743143035826e-05,
+      "loss": 0.0182,
+      "step": 1945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3228060007095337,
+      "learning_rate": 1.8738497275082642e-05,
+      "loss": 0.0223,
+      "step": 1950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.272178053855896,
+      "learning_rate": 1.8729563119807023e-05,
+      "loss": 0.0299,
+      "step": 1955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.30096063017845154,
+      "learning_rate": 1.8720628964531404e-05,
+      "loss": 0.017,
+      "step": 1960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5879131555557251,
+      "learning_rate": 1.871169480925579e-05,
+      "loss": 0.0265,
+      "step": 1965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6623579263687134,
+      "learning_rate": 1.8702760653980166e-05,
+      "loss": 0.024,
+      "step": 1970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7402109503746033,
+      "learning_rate": 1.8693826498704547e-05,
+      "loss": 0.0248,
+      "step": 1975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.46998223662376404,
+      "learning_rate": 1.8684892343428932e-05,
+      "loss": 0.0367,
+      "step": 1980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8414738178253174,
+      "learning_rate": 1.8675958188153313e-05,
+      "loss": 0.0352,
+      "step": 1985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5672045350074768,
+      "learning_rate": 1.866702403287769e-05,
+      "loss": 0.0286,
+      "step": 1990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4115860164165497,
+      "learning_rate": 1.8658089877602075e-05,
+      "loss": 0.0235,
+      "step": 1995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6932913064956665,
+      "learning_rate": 1.8649155722326456e-05,
+      "loss": 0.0128,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6872785687446594,
+      "learning_rate": 1.8640221567050837e-05,
+      "loss": 0.0205,
+      "step": 2005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.590305745601654,
+      "learning_rate": 1.8631287411775218e-05,
+      "loss": 0.0207,
+      "step": 2010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5067042708396912,
+      "learning_rate": 1.86223532564996e-05,
+      "loss": 0.038,
+      "step": 2015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.46255382895469666,
+      "learning_rate": 1.861341910122398e-05,
+      "loss": 0.0237,
+      "step": 2020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4533158540725708,
+      "learning_rate": 1.860448494594836e-05,
+      "loss": 0.0197,
+      "step": 2025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.919363796710968,
+      "learning_rate": 1.8595550790672742e-05,
+      "loss": 0.023,
+      "step": 2030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.42132195830345154,
+      "learning_rate": 1.8586616635397127e-05,
+      "loss": 0.0119,
+      "step": 2035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9792714715003967,
+      "learning_rate": 1.8577682480121508e-05,
+      "loss": 0.022,
+      "step": 2040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0622044801712036,
+      "learning_rate": 1.8568748324845885e-05,
+      "loss": 0.0191,
+      "step": 2045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3776015341281891,
+      "learning_rate": 1.855981416957027e-05,
+      "loss": 0.0173,
+      "step": 2050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7663435935974121,
+      "learning_rate": 1.855088001429465e-05,
+      "loss": 0.0257,
+      "step": 2055
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5264787673950195,
+      "learning_rate": 1.8541945859019032e-05,
+      "loss": 0.023,
+      "step": 2060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0220805406570435,
+      "learning_rate": 1.8533011703743413e-05,
+      "loss": 0.0215,
+      "step": 2065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.47214239835739136,
+      "learning_rate": 1.8524077548467794e-05,
+      "loss": 0.0368,
+      "step": 2070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40313026309013367,
+      "learning_rate": 1.8515143393192175e-05,
+      "loss": 0.0199,
+      "step": 2075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7139260768890381,
+      "learning_rate": 1.8506209237916556e-05,
+      "loss": 0.0182,
+      "step": 2080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0911624431610107,
+      "learning_rate": 1.8497275082640937e-05,
+      "loss": 0.0215,
+      "step": 2085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5224277973175049,
+      "learning_rate": 1.8488340927365318e-05,
+      "loss": 0.0203,
+      "step": 2090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6372708678245544,
+      "learning_rate": 1.84794067720897e-05,
+      "loss": 0.0363,
+      "step": 2095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0599619150161743,
+      "learning_rate": 1.847047261681408e-05,
+      "loss": 0.0343,
+      "step": 2100
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 0.023791618645191193,
+      "eval_pearson_cosine": 0.8906459396898687,
+      "eval_pearson_dot": 0.8866654230263398,
+      "eval_pearson_euclidean": 0.8789258913375184,
+      "eval_pearson_manhattan": 0.8774504409979331,
+      "eval_pearson_max": 0.8906459396898687,
+      "eval_runtime": 425.9801,
+      "eval_samples_per_second": 1.174,
+      "eval_spearman_cosine": 0.9069881079524318,
+      "eval_spearman_dot": 0.8997297109188437,
+      "eval_spearman_euclidean": 0.9081995367981472,
+      "eval_spearman_manhattan": 0.90537923751695,
+      "eval_spearman_max": 0.9081995367981472,
+      "eval_steps_per_second": 1.174,
+      "step": 2100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0371475219726562,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 0.0205,
+      "step": 2105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7635193467140198,
+      "learning_rate": 1.8452604306262846e-05,
+      "loss": 0.0236,
+      "step": 2110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4593610167503357,
+      "learning_rate": 1.8443670150987227e-05,
+      "loss": 0.0223,
+      "step": 2115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9037809371948242,
+      "learning_rate": 1.8434735995711608e-05,
+      "loss": 0.0348,
+      "step": 2120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4961464405059814,
+      "learning_rate": 1.842580184043599e-05,
+      "loss": 0.0278,
+      "step": 2125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0911996364593506,
+      "learning_rate": 1.841686768516037e-05,
+      "loss": 0.0124,
+      "step": 2130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8265781998634338,
+      "learning_rate": 1.840793352988475e-05,
+      "loss": 0.0336,
+      "step": 2135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2838935852050781,
+      "learning_rate": 1.8398999374609132e-05,
+      "loss": 0.0271,
+      "step": 2140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8582805395126343,
+      "learning_rate": 1.8390065219333513e-05,
+      "loss": 0.0466,
+      "step": 2145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2625269889831543,
+      "learning_rate": 1.8381131064057894e-05,
+      "loss": 0.0338,
+      "step": 2150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7238175272941589,
+      "learning_rate": 1.8372196908782275e-05,
+      "loss": 0.0149,
+      "step": 2155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4642048478126526,
+      "learning_rate": 1.8363262753506656e-05,
+      "loss": 0.017,
+      "step": 2160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6727991700172424,
+      "learning_rate": 1.835432859823104e-05,
+      "loss": 0.0302,
+      "step": 2165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1912602186203003,
+      "learning_rate": 1.834539444295542e-05,
+      "loss": 0.0244,
+      "step": 2170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8099949955940247,
+      "learning_rate": 1.83364602876798e-05,
+      "loss": 0.013,
+      "step": 2175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6302483081817627,
+      "learning_rate": 1.8327526132404184e-05,
+      "loss": 0.0245,
+      "step": 2180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.44703614711761475,
+      "learning_rate": 1.8318591977128565e-05,
+      "loss": 0.0179,
+      "step": 2185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4171731472015381,
+      "learning_rate": 1.8309657821852946e-05,
+      "loss": 0.0155,
+      "step": 2190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9061782360076904,
+      "learning_rate": 1.8300723666577327e-05,
+      "loss": 0.0222,
+      "step": 2195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.616051197052002,
+      "learning_rate": 1.8291789511301708e-05,
+      "loss": 0.0167,
+      "step": 2200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4685133695602417,
+      "learning_rate": 1.828285535602609e-05,
+      "loss": 0.0259,
+      "step": 2205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9447664022445679,
+      "learning_rate": 1.827392120075047e-05,
+      "loss": 0.0351,
+      "step": 2210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0720714330673218,
+      "learning_rate": 1.826498704547485e-05,
+      "loss": 0.0164,
+      "step": 2215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7534575462341309,
+      "learning_rate": 1.8256052890199235e-05,
+      "loss": 0.0161,
+      "step": 2220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8038508892059326,
+      "learning_rate": 1.8247118734923613e-05,
+      "loss": 0.0205,
+      "step": 2225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0298678874969482,
+      "learning_rate": 1.8238184579647994e-05,
+      "loss": 0.019,
+      "step": 2230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3235474824905396,
+      "learning_rate": 1.822925042437238e-05,
+      "loss": 0.0265,
+      "step": 2235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6873222589492798,
+      "learning_rate": 1.822031626909676e-05,
+      "loss": 0.0258,
+      "step": 2240
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4005903899669647,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.0185,
+      "step": 2245
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.47836175560951233,
+      "learning_rate": 1.820244795854552e-05,
+      "loss": 0.0229,
+      "step": 2250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.653409481048584,
+      "learning_rate": 1.8193513803269903e-05,
+      "loss": 0.0155,
+      "step": 2255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2501707077026367,
+      "learning_rate": 1.8184579647994284e-05,
+      "loss": 0.0253,
+      "step": 2260
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5146343111991882,
+      "learning_rate": 1.8175645492718665e-05,
+      "loss": 0.0355,
+      "step": 2265
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7069603204727173,
+      "learning_rate": 1.8166711337443046e-05,
+      "loss": 0.0245,
+      "step": 2270
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3391192853450775,
+      "learning_rate": 1.8157777182167427e-05,
+      "loss": 0.0125,
+      "step": 2275
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3187044858932495,
+      "learning_rate": 1.8148843026891808e-05,
+      "loss": 0.0314,
+      "step": 2280
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3739178478717804,
+      "learning_rate": 1.813990887161619e-05,
+      "loss": 0.0254,
+      "step": 2285
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9314795732498169,
+      "learning_rate": 1.8130974716340573e-05,
+      "loss": 0.0264,
+      "step": 2290
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9376474022865295,
+      "learning_rate": 1.8122040561064954e-05,
+      "loss": 0.0172,
+      "step": 2295
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9479859471321106,
+      "learning_rate": 1.8113106405789332e-05,
+      "loss": 0.0321,
+      "step": 2300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8741356730461121,
+      "learning_rate": 1.8104172250513716e-05,
+      "loss": 0.0216,
+      "step": 2305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7016437649726868,
+      "learning_rate": 1.8095238095238097e-05,
+      "loss": 0.0181,
+      "step": 2310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6568303108215332,
+      "learning_rate": 1.808630393996248e-05,
+      "loss": 0.0374,
+      "step": 2315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4468625783920288,
+      "learning_rate": 1.807736978468686e-05,
+      "loss": 0.021,
+      "step": 2320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8952460885047913,
+      "learning_rate": 1.806843562941124e-05,
+      "loss": 0.0215,
+      "step": 2325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5987122654914856,
+      "learning_rate": 1.805950147413562e-05,
+      "loss": 0.0182,
+      "step": 2330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8986082077026367,
+      "learning_rate": 1.8050567318860003e-05,
+      "loss": 0.0224,
+      "step": 2335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4435628652572632,
+      "learning_rate": 1.8041633163584384e-05,
+      "loss": 0.0246,
+      "step": 2340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5219196081161499,
+      "learning_rate": 1.8032699008308765e-05,
+      "loss": 0.0222,
+      "step": 2345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8987548351287842,
+      "learning_rate": 1.802376485303315e-05,
+      "loss": 0.0416,
+      "step": 2350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3122328817844391,
+      "learning_rate": 1.8014830697757527e-05,
+      "loss": 0.0162,
+      "step": 2355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.45380455255508423,
+      "learning_rate": 1.8005896542481908e-05,
+      "loss": 0.0348,
+      "step": 2360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7789851427078247,
+      "learning_rate": 1.7996962387206292e-05,
+      "loss": 0.0207,
+      "step": 2365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7289481163024902,
+      "learning_rate": 1.7988028231930673e-05,
+      "loss": 0.0154,
+      "step": 2370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5287164449691772,
+      "learning_rate": 1.7979094076655054e-05,
+      "loss": 0.0306,
+      "step": 2375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.236247181892395,
+      "learning_rate": 1.7970159921379435e-05,
+      "loss": 0.0346,
+      "step": 2380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6785498857498169,
+      "learning_rate": 1.7961225766103816e-05,
+      "loss": 0.0191,
+      "step": 2385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7532624006271362,
+      "learning_rate": 1.7952291610828197e-05,
+      "loss": 0.0247,
+      "step": 2390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8474083542823792,
+      "learning_rate": 1.794335745555258e-05,
+      "loss": 0.028,
+      "step": 2395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8650068640708923,
+      "learning_rate": 1.793442330027696e-05,
+      "loss": 0.0234,
+      "step": 2400
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 0.02186727523803711,
+      "eval_pearson_cosine": 0.9016899300379176,
+      "eval_pearson_dot": 0.8979110419976675,
+      "eval_pearson_euclidean": 0.892515905459987,
+      "eval_pearson_manhattan": 0.8905747154959989,
+      "eval_pearson_max": 0.9016899300379176,
+      "eval_runtime": 426.0444,
+      "eval_samples_per_second": 1.174,
+      "eval_spearman_cosine": 0.9066876267505071,
+      "eval_spearman_dot": 0.9012341329365315,
+      "eval_spearman_euclidean": 0.9076169104676419,
+      "eval_spearman_manhattan": 0.9052689330757322,
+      "eval_spearman_max": 0.9076169104676419,
+      "eval_steps_per_second": 1.174,
+      "step": 2400
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7486416101455688,
+      "learning_rate": 1.7925489145001344e-05,
+      "loss": 0.0158,
+      "step": 2405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.20279864966869354,
+      "learning_rate": 1.791655498972572e-05,
+      "loss": 0.0141,
+      "step": 2410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4669971466064453,
+      "learning_rate": 1.7907620834450103e-05,
+      "loss": 0.019,
+      "step": 2415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6175631880760193,
+      "learning_rate": 1.7898686679174487e-05,
+      "loss": 0.0148,
+      "step": 2420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5767547488212585,
+      "learning_rate": 1.7889752523898868e-05,
+      "loss": 0.0252,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5371144413948059,
+      "learning_rate": 1.7880818368623246e-05,
+      "loss": 0.0121,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6145671606063843,
+      "learning_rate": 1.787188421334763e-05,
+      "loss": 0.0195,
+      "step": 2435
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.262970209121704,
+      "learning_rate": 1.786295005807201e-05,
+      "loss": 0.0298,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3572503328323364,
+      "learning_rate": 1.7854015902796392e-05,
+      "loss": 0.0152,
+      "step": 2445
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6049114465713501,
+      "learning_rate": 1.7845081747520773e-05,
+      "loss": 0.0153,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.40164369344711304,
+      "learning_rate": 1.7836147592245154e-05,
+      "loss": 0.0171,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.33443671464920044,
+      "learning_rate": 1.7827213436969535e-05,
+      "loss": 0.0174,
+      "step": 2460
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37140563130378723,
+      "learning_rate": 1.7818279281693916e-05,
+      "loss": 0.0147,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.44236427545547485,
+      "learning_rate": 1.7809345126418297e-05,
+      "loss": 0.019,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1675889492034912,
+      "learning_rate": 1.7800410971142682e-05,
+      "loss": 0.0196,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7172742486000061,
+      "learning_rate": 1.7791476815867063e-05,
+      "loss": 0.0262,
+      "step": 2480
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9804342985153198,
+      "learning_rate": 1.778254266059144e-05,
+      "loss": 0.0337,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.989030361175537,
+      "learning_rate": 1.7773608505315825e-05,
+      "loss": 0.0238,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.754467785358429,
+      "learning_rate": 1.7764674350040206e-05,
+      "loss": 0.0224,
+      "step": 2495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9083816409111023,
+      "learning_rate": 1.7755740194764587e-05,
+      "loss": 0.0165,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5486024022102356,
+      "learning_rate": 1.7746806039488968e-05,
+      "loss": 0.0173,
+      "step": 2505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4576135575771332,
+      "learning_rate": 1.773787188421335e-05,
+      "loss": 0.0177,
+      "step": 2510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9265009760856628,
+      "learning_rate": 1.772893772893773e-05,
+      "loss": 0.0216,
+      "step": 2515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.506006121635437,
+      "learning_rate": 1.772000357366211e-05,
+      "loss": 0.0135,
+      "step": 2520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2627745270729065,
+      "learning_rate": 1.7711069418386492e-05,
+      "loss": 0.0244,
+      "step": 2525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7158712148666382,
+      "learning_rate": 1.7702135263110873e-05,
+      "loss": 0.0238,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8237607479095459,
+      "learning_rate": 1.7693201107835258e-05,
+      "loss": 0.0272,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.266671359539032,
+      "learning_rate": 1.7684266952559635e-05,
+      "loss": 0.0202,
+      "step": 2540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6004403829574585,
+      "learning_rate": 1.7675332797284016e-05,
+      "loss": 0.0414,
+      "step": 2545
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9549585580825806,
+      "learning_rate": 1.76663986420084e-05,
+      "loss": 0.0209,
+      "step": 2550
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9587855935096741,
+      "learning_rate": 1.7657464486732782e-05,
+      "loss": 0.0239,
+      "step": 2555
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5804222226142883,
+      "learning_rate": 1.7648530331457163e-05,
+      "loss": 0.0183,
+      "step": 2560
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.506087601184845,
+      "learning_rate": 1.7639596176181544e-05,
+      "loss": 0.0202,
+      "step": 2565
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8496732711791992,
+      "learning_rate": 1.7630662020905925e-05,
+      "loss": 0.0181,
+      "step": 2570
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3317871391773224,
+      "learning_rate": 1.7621727865630306e-05,
+      "loss": 0.0218,
+      "step": 2575
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3938535451889038,
+      "learning_rate": 1.7612793710354687e-05,
+      "loss": 0.0112,
+      "step": 2580
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3293476104736328,
+      "learning_rate": 1.7603859555079068e-05,
+      "loss": 0.0181,
+      "step": 2585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28195351362228394,
+      "learning_rate": 1.7594925399803452e-05,
+      "loss": 0.0135,
+      "step": 2590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7966617941856384,
+      "learning_rate": 1.758599124452783e-05,
+      "loss": 0.0145,
+      "step": 2595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5484746694564819,
+      "learning_rate": 1.757705708925221e-05,
+      "loss": 0.0207,
+      "step": 2600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.576448380947113,
+      "learning_rate": 1.7568122933976596e-05,
+      "loss": 0.0127,
+      "step": 2605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28207531571388245,
+      "learning_rate": 1.7559188778700977e-05,
+      "loss": 0.0132,
+      "step": 2610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29697030782699585,
+      "learning_rate": 1.7550254623425354e-05,
+      "loss": 0.0127,
+      "step": 2615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.122375249862671,
+      "learning_rate": 1.754132046814974e-05,
+      "loss": 0.0297,
+      "step": 2620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5798119902610779,
+      "learning_rate": 1.753238631287412e-05,
+      "loss": 0.0171,
+      "step": 2625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6546443104743958,
+      "learning_rate": 1.75234521575985e-05,
+      "loss": 0.0436,
+      "step": 2630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8825979828834534,
+      "learning_rate": 1.7514518002322882e-05,
+      "loss": 0.0138,
+      "step": 2635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.46849188208580017,
+      "learning_rate": 1.7505583847047263e-05,
+      "loss": 0.0328,
+      "step": 2640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.324487566947937,
+      "learning_rate": 1.7496649691771644e-05,
+      "loss": 0.0241,
+      "step": 2645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5214132070541382,
+      "learning_rate": 1.7487715536496025e-05,
+      "loss": 0.0344,
+      "step": 2650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.791459858417511,
+      "learning_rate": 1.7478781381220406e-05,
+      "loss": 0.0167,
+      "step": 2655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.888538658618927,
+      "learning_rate": 1.746984722594479e-05,
+      "loss": 0.0216,
+      "step": 2660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.31034013628959656,
+      "learning_rate": 1.746091307066917e-05,
+      "loss": 0.0327,
+      "step": 2665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4363711476325989,
+      "learning_rate": 1.745197891539355e-05,
+      "loss": 0.0203,
+      "step": 2670
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0624545812606812,
+      "learning_rate": 1.7443044760117934e-05,
+      "loss": 0.0216,
+      "step": 2675
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4411804676055908,
+      "learning_rate": 1.7434110604842315e-05,
+      "loss": 0.0294,
+      "step": 2680
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7424313426017761,
+      "learning_rate": 1.7425176449566696e-05,
+      "loss": 0.0233,
+      "step": 2685
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6678118705749512,
+      "learning_rate": 1.7416242294291077e-05,
+      "loss": 0.0129,
+      "step": 2690
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.43497857451438904,
+      "learning_rate": 1.7407308139015458e-05,
+      "loss": 0.0272,
+      "step": 2695
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3903897702693939,
+      "learning_rate": 1.739837398373984e-05,
+      "loss": 0.0209,
+      "step": 2700
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 0.027375079691410065,
+      "eval_pearson_cosine": 0.9001051843718942,
+      "eval_pearson_dot": 0.9006651139179915,
+      "eval_pearson_euclidean": 0.8930651682478097,
+      "eval_pearson_manhattan": 0.891765654050478,
+      "eval_pearson_max": 0.9006651139179915,
+      "eval_runtime": 426.2183,
+      "eval_samples_per_second": 1.173,
+      "eval_spearman_cosine": 0.9177515750063,
+      "eval_spearman_dot": 0.9167129468517873,
+      "eval_spearman_euclidean": 0.9167275389101556,
+      "eval_spearman_manhattan": 0.9159309277237108,
+      "eval_spearman_max": 0.9177515750063,
+      "eval_steps_per_second": 1.173,
+      "step": 2700
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3496129512786865,
+      "learning_rate": 1.738943982846422e-05,
+      "loss": 0.0259,
+      "step": 2705
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1606944799423218,
+      "learning_rate": 1.73805056731886e-05,
+      "loss": 0.0181,
+      "step": 2710
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5323413610458374,
+      "learning_rate": 1.7371571517912982e-05,
+      "loss": 0.0137,
+      "step": 2715
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5756454467773438,
+      "learning_rate": 1.7362637362637363e-05,
+      "loss": 0.0181,
+      "step": 2720
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3727392256259918,
+      "learning_rate": 1.7353703207361744e-05,
+      "loss": 0.0123,
+      "step": 2725
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.441544771194458,
+      "learning_rate": 1.7344769052086125e-05,
+      "loss": 0.0157,
+      "step": 2730
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6806672811508179,
+      "learning_rate": 1.733583489681051e-05,
+      "loss": 0.0196,
+      "step": 2735
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5537866353988647,
+      "learning_rate": 1.732690074153489e-05,
+      "loss": 0.0163,
+      "step": 2740
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5595599412918091,
+      "learning_rate": 1.731796658625927e-05,
+      "loss": 0.0229,
+      "step": 2745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.48301759362220764,
+      "learning_rate": 1.7309032430983652e-05,
+      "loss": 0.0197,
+      "step": 2750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2349361181259155,
+      "learning_rate": 1.7300098275708034e-05,
+      "loss": 0.0198,
+      "step": 2755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3702278137207031,
+      "learning_rate": 1.7291164120432415e-05,
+      "loss": 0.0165,
+      "step": 2760
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5885152220726013,
+      "learning_rate": 1.7282229965156796e-05,
+      "loss": 0.0126,
+      "step": 2765
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2845672369003296,
+      "learning_rate": 1.7273295809881177e-05,
+      "loss": 0.0232,
+      "step": 2770
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0059493780136108,
+      "learning_rate": 1.7264361654605558e-05,
+      "loss": 0.0102,
+      "step": 2775
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.621356189250946,
+      "learning_rate": 1.725542749932994e-05,
+      "loss": 0.0149,
+      "step": 2780
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8701572418212891,
+      "learning_rate": 1.724649334405432e-05,
+      "loss": 0.0147,
+      "step": 2785
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.35073745250701904,
+      "learning_rate": 1.7237559188778704e-05,
+      "loss": 0.0176,
+      "step": 2790
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6895755529403687,
+      "learning_rate": 1.7228625033503085e-05,
+      "loss": 0.0212,
+      "step": 2795
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9424052834510803,
+      "learning_rate": 1.7219690878227463e-05,
+      "loss": 0.0213,
+      "step": 2800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.45782536268234253,
+      "learning_rate": 1.7210756722951847e-05,
+      "loss": 0.0122,
+      "step": 2805
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9347144365310669,
+      "learning_rate": 1.720182256767623e-05,
+      "loss": 0.0221,
+      "step": 2810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4957163333892822,
+      "learning_rate": 1.719288841240061e-05,
+      "loss": 0.0154,
+      "step": 2815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.40040886402130127,
+      "learning_rate": 1.718395425712499e-05,
+      "loss": 0.018,
+      "step": 2820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4707423150539398,
+      "learning_rate": 1.717502010184937e-05,
+      "loss": 0.0122,
+      "step": 2825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6660599708557129,
+      "learning_rate": 1.7166085946573752e-05,
+      "loss": 0.0103,
+      "step": 2830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5606631636619568,
+      "learning_rate": 1.7157151791298134e-05,
+      "loss": 0.0246,
+      "step": 2835
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0515074729919434,
+      "learning_rate": 1.7148217636022515e-05,
+      "loss": 0.0173,
+      "step": 2840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8896968364715576,
+      "learning_rate": 1.71392834807469e-05,
+      "loss": 0.0357,
+      "step": 2845
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8183419108390808,
+      "learning_rate": 1.7130349325471277e-05,
+      "loss": 0.0125,
+      "step": 2850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28626278042793274,
+      "learning_rate": 1.7121415170195658e-05,
+      "loss": 0.011,
+      "step": 2855
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5537832975387573,
+      "learning_rate": 1.7112481014920042e-05,
+      "loss": 0.0155,
+      "step": 2860
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3938451111316681,
+      "learning_rate": 1.7103546859644423e-05,
+      "loss": 0.0163,
+      "step": 2865
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27609241008758545,
+      "learning_rate": 1.7094612704368804e-05,
+      "loss": 0.0193,
+      "step": 2870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3002891540527344,
+      "learning_rate": 1.7085678549093185e-05,
+      "loss": 0.0142,
+      "step": 2875
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8881809711456299,
+      "learning_rate": 1.7076744393817566e-05,
+      "loss": 0.0296,
+      "step": 2880
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.99460768699646,
+      "learning_rate": 1.7067810238541947e-05,
+      "loss": 0.0157,
+      "step": 2885
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.14095938205719,
+      "learning_rate": 1.705887608326633e-05,
+      "loss": 0.0164,
+      "step": 2890
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2840583026409149,
+      "learning_rate": 1.704994192799071e-05,
+      "loss": 0.0139,
+      "step": 2895
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3837971091270447,
+      "learning_rate": 1.704100777271509e-05,
+      "loss": 0.016,
+      "step": 2900
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4202588200569153,
+      "learning_rate": 1.703207361743947e-05,
+      "loss": 0.0161,
+      "step": 2905
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9836485385894775,
+      "learning_rate": 1.7023139462163853e-05,
+      "loss": 0.0336,
+      "step": 2910
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28357017040252686,
+      "learning_rate": 1.7014205306888234e-05,
+      "loss": 0.0151,
+      "step": 2915
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3464276194572449,
+      "learning_rate": 1.7005271151612618e-05,
+      "loss": 0.0151,
+      "step": 2920
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6251973509788513,
+      "learning_rate": 1.6996336996336996e-05,
+      "loss": 0.0275,
+      "step": 2925
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1420838832855225,
+      "learning_rate": 1.698740284106138e-05,
+      "loss": 0.0339,
+      "step": 2930
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.43490684032440186,
+      "learning_rate": 1.697846868578576e-05,
+      "loss": 0.0124,
+      "step": 2935
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.576763391494751,
+      "learning_rate": 1.6969534530510142e-05,
+      "loss": 0.0162,
+      "step": 2940
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8549007773399353,
+      "learning_rate": 1.6960600375234523e-05,
+      "loss": 0.0217,
+      "step": 2945
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3743274509906769,
+      "learning_rate": 1.6951666219958904e-05,
+      "loss": 0.0165,
+      "step": 2950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7152208089828491,
+      "learning_rate": 1.6942732064683285e-05,
+      "loss": 0.0158,
+      "step": 2955
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5613319873809814,
+      "learning_rate": 1.6933797909407666e-05,
+      "loss": 0.0232,
+      "step": 2960
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3946327269077301,
+      "learning_rate": 1.6924863754132047e-05,
+      "loss": 0.0156,
+      "step": 2965
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5846483707427979,
+      "learning_rate": 1.691592959885643e-05,
+      "loss": 0.0118,
+      "step": 2970
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6644983291625977,
+      "learning_rate": 1.6906995443580813e-05,
+      "loss": 0.0156,
+      "step": 2975
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7113365530967712,
+      "learning_rate": 1.689806128830519e-05,
+      "loss": 0.0131,
+      "step": 2980
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7063427567481995,
+      "learning_rate": 1.688912713302957e-05,
+      "loss": 0.0196,
+      "step": 2985
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8692206740379333,
+      "learning_rate": 1.6880192977753956e-05,
+      "loss": 0.02,
+      "step": 2990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6556200385093689,
+      "learning_rate": 1.6871258822478337e-05,
+      "loss": 0.0273,
+      "step": 2995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9636979103088379,
+      "learning_rate": 1.6862324667202718e-05,
+      "loss": 0.0192,
+      "step": 3000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 0.0198129303753376,
+      "eval_pearson_cosine": 0.9064648118738065,
+      "eval_pearson_dot": 0.9033815480817863,
+      "eval_pearson_euclidean": 0.8939116848670458,
+      "eval_pearson_manhattan": 0.8927147647429707,
+      "eval_pearson_max": 0.9064648118738065,
+      "eval_runtime": 426.1029,
+      "eval_samples_per_second": 1.173,
+      "eval_spearman_cosine": 0.9222772251089003,
+      "eval_spearman_dot": 0.9168591554366217,
+      "eval_spearman_euclidean": 0.9222438169752679,
+      "eval_spearman_manhattan": 0.9220269521078085,
+      "eval_spearman_max": 0.9222772251089003,
+      "eval_steps_per_second": 1.173,
+      "step": 3000
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5286144614219666,
+      "learning_rate": 1.68533905119271e-05,
+      "loss": 0.0228,
+      "step": 3005
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7542079091072083,
+      "learning_rate": 1.684445635665148e-05,
+      "loss": 0.0156,
+      "step": 3010
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6064092516899109,
+      "learning_rate": 1.683552220137586e-05,
+      "loss": 0.0258,
+      "step": 3015
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7690675258636475,
+      "learning_rate": 1.6826588046100242e-05,
+      "loss": 0.0201,
+      "step": 3020
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4456731975078583,
+      "learning_rate": 1.6817653890824623e-05,
+      "loss": 0.0159,
+      "step": 3025
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6881704330444336,
+      "learning_rate": 1.6808719735549008e-05,
+      "loss": 0.0191,
+      "step": 3030
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.515715479850769,
+      "learning_rate": 1.6799785580273385e-05,
+      "loss": 0.0272,
+      "step": 3035
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9520334601402283,
+      "learning_rate": 1.6790851424997766e-05,
+      "loss": 0.0164,
+      "step": 3040
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4020353555679321,
+      "learning_rate": 1.678191726972215e-05,
+      "loss": 0.0224,
+      "step": 3045
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.868922472000122,
+      "learning_rate": 1.6772983114446532e-05,
+      "loss": 0.0343,
+      "step": 3050
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5014618635177612,
+      "learning_rate": 1.676404895917091e-05,
+      "loss": 0.0318,
+      "step": 3055
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.820292592048645,
+      "learning_rate": 1.6755114803895294e-05,
+      "loss": 0.0264,
+      "step": 3060
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6107310056686401,
+      "learning_rate": 1.6746180648619675e-05,
+      "loss": 0.0188,
+      "step": 3065
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3835718631744385,
+      "learning_rate": 1.6737246493344056e-05,
+      "loss": 0.012,
+      "step": 3070
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.46898066997528076,
+      "learning_rate": 1.6728312338068437e-05,
+      "loss": 0.0333,
+      "step": 3075
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6556499004364014,
+      "learning_rate": 1.6719378182792818e-05,
+      "loss": 0.0184,
+      "step": 3080
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8477081060409546,
+      "learning_rate": 1.67104440275172e-05,
+      "loss": 0.0327,
+      "step": 3085
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6761387586593628,
+      "learning_rate": 1.670150987224158e-05,
+      "loss": 0.0178,
+      "step": 3090
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4892427623271942,
+      "learning_rate": 1.669257571696596e-05,
+      "loss": 0.0266,
+      "step": 3095
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7155781388282776,
+      "learning_rate": 1.6683641561690342e-05,
+      "loss": 0.0242,
+      "step": 3100
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.25152644515037537,
+      "learning_rate": 1.6674707406414727e-05,
+      "loss": 0.0102,
+      "step": 3105
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8297644853591919,
+      "learning_rate": 1.6665773251139104e-05,
+      "loss": 0.0209,
+      "step": 3110
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.36103129386901855,
+      "learning_rate": 1.665683909586349e-05,
+      "loss": 0.0166,
+      "step": 3115
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8184224367141724,
+      "learning_rate": 1.664790494058787e-05,
+      "loss": 0.0215,
+      "step": 3120
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5235499739646912,
+      "learning_rate": 1.663897078531225e-05,
+      "loss": 0.0223,
+      "step": 3125
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6001405715942383,
+      "learning_rate": 1.6630036630036632e-05,
+      "loss": 0.0177,
+      "step": 3130
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5021240711212158,
+      "learning_rate": 1.6621102474761013e-05,
+      "loss": 0.0199,
+      "step": 3135
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2629213035106659,
+      "learning_rate": 1.6612168319485394e-05,
+      "loss": 0.0375,
+      "step": 3140
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6333101987838745,
+      "learning_rate": 1.6603234164209775e-05,
+      "loss": 0.0138,
+      "step": 3145
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5239858627319336,
+      "learning_rate": 1.6594300008934156e-05,
+      "loss": 0.0211,
+      "step": 3150
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5263399481773376,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.0281,
+      "step": 3155
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7547926902770996,
+      "learning_rate": 1.657643169838292e-05,
+      "loss": 0.0325,
+      "step": 3160
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.098939061164856,
+      "learning_rate": 1.65674975431073e-05,
+      "loss": 0.0136,
+      "step": 3165
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5231286883354187,
+      "learning_rate": 1.655856338783168e-05,
+      "loss": 0.0241,
+      "step": 3170
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6786376237869263,
+      "learning_rate": 1.6549629232556064e-05,
+      "loss": 0.0264,
+      "step": 3175
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.402637243270874,
+      "learning_rate": 1.6540695077280446e-05,
+      "loss": 0.0214,
+      "step": 3180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.169063687324524,
+      "learning_rate": 1.6531760922004823e-05,
+      "loss": 0.038,
+      "step": 3185
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9597774147987366,
+      "learning_rate": 1.6522826766729208e-05,
+      "loss": 0.0289,
+      "step": 3190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5611427426338196,
+      "learning_rate": 1.651389261145359e-05,
+      "loss": 0.0155,
+      "step": 3195
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9891091585159302,
+      "learning_rate": 1.650495845617797e-05,
+      "loss": 0.0261,
+      "step": 3200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.445168137550354,
+      "learning_rate": 1.649602430090235e-05,
+      "loss": 0.0192,
+      "step": 3205
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0663169622421265,
+      "learning_rate": 1.6487090145626732e-05,
+      "loss": 0.0252,
+      "step": 3210
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9698647856712341,
+      "learning_rate": 1.6478155990351116e-05,
+      "loss": 0.0223,
+      "step": 3215
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3728877305984497,
+      "learning_rate": 1.6469221835075494e-05,
+      "loss": 0.0194,
+      "step": 3220
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.654509961605072,
+      "learning_rate": 1.6460287679799875e-05,
+      "loss": 0.0255,
+      "step": 3225
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8245360851287842,
+      "learning_rate": 1.645135352452426e-05,
+      "loss": 0.0151,
+      "step": 3230
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6417776942253113,
+      "learning_rate": 1.644241936924864e-05,
+      "loss": 0.0153,
+      "step": 3235
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.23088638484477997,
+      "learning_rate": 1.6433485213973018e-05,
+      "loss": 0.0258,
+      "step": 3240
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4798796474933624,
+      "learning_rate": 1.6424551058697402e-05,
+      "loss": 0.0152,
+      "step": 3245
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42295366525650024,
+      "learning_rate": 1.6415616903421783e-05,
+      "loss": 0.0153,
+      "step": 3250
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3053552210330963,
+      "learning_rate": 1.6406682748146165e-05,
+      "loss": 0.0098,
+      "step": 3255
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.37020963430404663,
+      "learning_rate": 1.6397748592870546e-05,
+      "loss": 0.0151,
+      "step": 3260
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5717061758041382,
+      "learning_rate": 1.6388814437594927e-05,
+      "loss": 0.0157,
+      "step": 3265
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5504093766212463,
+      "learning_rate": 1.6379880282319308e-05,
+      "loss": 0.0397,
+      "step": 3270
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6604989767074585,
+      "learning_rate": 1.637094612704369e-05,
+      "loss": 0.0158,
+      "step": 3275
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7319514155387878,
+      "learning_rate": 1.636201197176807e-05,
+      "loss": 0.0118,
+      "step": 3280
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5383831262588501,
+      "learning_rate": 1.6353077816492454e-05,
+      "loss": 0.0196,
+      "step": 3285
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9546700716018677,
+      "learning_rate": 1.6344143661216835e-05,
+      "loss": 0.0204,
+      "step": 3290
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.929811954498291,
+      "learning_rate": 1.6335209505941213e-05,
+      "loss": 0.0286,
+      "step": 3295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.545769453048706,
+      "learning_rate": 1.6326275350665597e-05,
+      "loss": 0.0178,
+      "step": 3300
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 0.02022329904139042,
+      "eval_pearson_cosine": 0.9200674333472122,
+      "eval_pearson_dot": 0.9176731158869308,
+      "eval_pearson_euclidean": 0.9061361146439217,
+      "eval_pearson_manhattan": 0.9047424977345989,
+      "eval_pearson_max": 0.9200674333472122,
+      "eval_runtime": 426.1218,
+      "eval_samples_per_second": 1.173,
+      "eval_spearman_cosine": 0.9299045516182064,
+      "eval_spearman_dot": 0.9254067416269663,
+      "eval_spearman_euclidean": 0.9302610970443881,
+      "eval_spearman_manhattan": 0.9284666578666315,
+      "eval_spearman_max": 0.9302610970443881,
+      "eval_steps_per_second": 1.173,
+      "step": 3300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5426648259162903,
+      "learning_rate": 1.6317341195389978e-05,
+      "loss": 0.0123,
+      "step": 3305
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4093814492225647,
+      "learning_rate": 1.630840704011436e-05,
+      "loss": 0.0176,
+      "step": 3310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9546383619308472,
+      "learning_rate": 1.629947288483874e-05,
+      "loss": 0.0195,
+      "step": 3315
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6893852949142456,
+      "learning_rate": 1.629053872956312e-05,
+      "loss": 0.0253,
+      "step": 3320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3756393790245056,
+      "learning_rate": 1.6281604574287502e-05,
+      "loss": 0.0106,
+      "step": 3325
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.563850998878479,
+      "learning_rate": 1.6272670419011883e-05,
+      "loss": 0.017,
+      "step": 3330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3767915964126587,
+      "learning_rate": 1.6263736263736265e-05,
+      "loss": 0.0231,
+      "step": 3335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3786735236644745,
+      "learning_rate": 1.6254802108460646e-05,
+      "loss": 0.0121,
+      "step": 3340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2710884213447571,
+      "learning_rate": 1.624586795318503e-05,
+      "loss": 0.0121,
+      "step": 3345
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4059496819972992,
+      "learning_rate": 1.6236933797909408e-05,
+      "loss": 0.0138,
+      "step": 3350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5236977338790894,
+      "learning_rate": 1.622799964263379e-05,
+      "loss": 0.023,
+      "step": 3355
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2646647691726685,
+      "learning_rate": 1.6219065487358173e-05,
+      "loss": 0.026,
+      "step": 3360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5769827365875244,
+      "learning_rate": 1.6210131332082554e-05,
+      "loss": 0.0145,
+      "step": 3365
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8552572727203369,
+      "learning_rate": 1.6201197176806932e-05,
+      "loss": 0.0132,
+      "step": 3370
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.690263271331787,
+      "learning_rate": 1.6192263021531316e-05,
+      "loss": 0.04,
+      "step": 3375
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6928017735481262,
+      "learning_rate": 1.6183328866255697e-05,
+      "loss": 0.0184,
+      "step": 3380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.31537023186683655,
+      "learning_rate": 1.6174394710980078e-05,
+      "loss": 0.0156,
+      "step": 3385
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.826603889465332,
+      "learning_rate": 1.616546055570446e-05,
+      "loss": 0.0253,
+      "step": 3390
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0481098890304565,
+      "learning_rate": 1.615652640042884e-05,
+      "loss": 0.019,
+      "step": 3395
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.719532310962677,
+      "learning_rate": 1.614759224515322e-05,
+      "loss": 0.0152,
+      "step": 3400
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7920011281967163,
+      "learning_rate": 1.6138658089877602e-05,
+      "loss": 0.0142,
+      "step": 3405
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9310560822486877,
+      "learning_rate": 1.6129723934601983e-05,
+      "loss": 0.0187,
+      "step": 3410
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4473055303096771,
+      "learning_rate": 1.6120789779326368e-05,
+      "loss": 0.0317,
+      "step": 3415
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3835084140300751,
+      "learning_rate": 1.611185562405075e-05,
+      "loss": 0.0143,
+      "step": 3420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6586607098579407,
+      "learning_rate": 1.6102921468775127e-05,
+      "loss": 0.027,
+      "step": 3425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.510007917881012,
+      "learning_rate": 1.609398731349951e-05,
+      "loss": 0.0213,
+      "step": 3430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7229037284851074,
+      "learning_rate": 1.6085053158223892e-05,
+      "loss": 0.014,
+      "step": 3435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.35988929867744446,
+      "learning_rate": 1.6076119002948273e-05,
+      "loss": 0.0166,
+      "step": 3440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9715976119041443,
+      "learning_rate": 1.6067184847672654e-05,
+      "loss": 0.0202,
+      "step": 3445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5804651975631714,
+      "learning_rate": 1.6058250692397035e-05,
+      "loss": 0.0177,
+      "step": 3450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6329138278961182,
+      "learning_rate": 1.6049316537121416e-05,
+      "loss": 0.0184,
+      "step": 3455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.46104130148887634,
+      "learning_rate": 1.6040382381845797e-05,
+      "loss": 0.0208,
+      "step": 3460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.040407657623291,
+      "learning_rate": 1.6031448226570178e-05,
+      "loss": 0.0281,
+      "step": 3465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7342686057090759,
+      "learning_rate": 1.6022514071294563e-05,
+      "loss": 0.0163,
+      "step": 3470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7275007367134094,
+      "learning_rate": 1.601357991601894e-05,
+      "loss": 0.0193,
+      "step": 3475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.30385661125183105,
+      "learning_rate": 1.600464576074332e-05,
+      "loss": 0.0153,
+      "step": 3480
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5035312175750732,
+      "learning_rate": 1.5995711605467706e-05,
+      "loss": 0.0189,
+      "step": 3485
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2530502676963806,
+      "learning_rate": 1.5986777450192087e-05,
+      "loss": 0.017,
+      "step": 3490
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.068692922592163,
+      "learning_rate": 1.5977843294916468e-05,
+      "loss": 0.0178,
+      "step": 3495
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40073007345199585,
+      "learning_rate": 1.596890913964085e-05,
+      "loss": 0.0165,
+      "step": 3500
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.783413887023926,
+      "learning_rate": 1.595997498436523e-05,
+      "loss": 0.0235,
+      "step": 3505
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.46688321232795715,
+      "learning_rate": 1.595104082908961e-05,
+      "loss": 0.0188,
+      "step": 3510
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6832770705223083,
+      "learning_rate": 1.5942106673813992e-05,
+      "loss": 0.0202,
+      "step": 3515
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.39423632621765137,
+      "learning_rate": 1.5933172518538373e-05,
+      "loss": 0.0118,
+      "step": 3520
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8265882730484009,
+      "learning_rate": 1.5924238363262754e-05,
+      "loss": 0.0175,
+      "step": 3525
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8301231265068054,
+      "learning_rate": 1.5915304207987135e-05,
+      "loss": 0.0396,
+      "step": 3530
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.45745956897735596,
+      "learning_rate": 1.5906370052711516e-05,
+      "loss": 0.0195,
+      "step": 3535
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.464941382408142,
+      "learning_rate": 1.5897435897435897e-05,
+      "loss": 0.0207,
+      "step": 3540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7965016961097717,
+      "learning_rate": 1.588850174216028e-05,
+      "loss": 0.0178,
+      "step": 3545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6353152394294739,
+      "learning_rate": 1.5879567586884663e-05,
+      "loss": 0.019,
+      "step": 3550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.38456305861473083,
+      "learning_rate": 1.587063343160904e-05,
+      "loss": 0.0169,
+      "step": 3555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2627872228622437,
+      "learning_rate": 1.5861699276333425e-05,
+      "loss": 0.0226,
+      "step": 3560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5265514254570007,
+      "learning_rate": 1.5852765121057806e-05,
+      "loss": 0.0141,
+      "step": 3565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5938783288002014,
+      "learning_rate": 1.5843830965782187e-05,
+      "loss": 0.0178,
+      "step": 3570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.39797863364219666,
+      "learning_rate": 1.5834896810506568e-05,
+      "loss": 0.0123,
+      "step": 3575
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6035088300704956,
+      "learning_rate": 1.582596265523095e-05,
+      "loss": 0.0162,
+      "step": 3580
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.34465670585632324,
+      "learning_rate": 1.581702849995533e-05,
+      "loss": 0.0135,
+      "step": 3585
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.40939584374427795,
+      "learning_rate": 1.580809434467971e-05,
+      "loss": 0.0123,
+      "step": 3590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4972923994064331,
+      "learning_rate": 1.5799160189404092e-05,
+      "loss": 0.0178,
+      "step": 3595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3403087556362152,
+      "learning_rate": 1.5790226034128477e-05,
+      "loss": 0.0104,
+      "step": 3600
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 0.01895288936793804,
+      "eval_pearson_cosine": 0.9183723675584331,
+      "eval_pearson_dot": 0.9165141926322535,
+      "eval_pearson_euclidean": 0.9111139441737433,
+      "eval_pearson_manhattan": 0.9105498844088082,
+      "eval_pearson_max": 0.9183723675584331,
+      "eval_runtime": 425.5624,
+      "eval_samples_per_second": 1.175,
+      "eval_spearman_cosine": 0.925627350509402,
+      "eval_spearman_dot": 0.9219613838455353,
+      "eval_spearman_euclidean": 0.9258841515366061,
+      "eval_spearman_manhattan": 0.9260336241344965,
+      "eval_spearman_max": 0.9260336241344965,
+      "eval_steps_per_second": 1.175,
+      "step": 3600
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2765352129936218,
+      "learning_rate": 1.5781291878852854e-05,
+      "loss": 0.0154,
+      "step": 3605
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1084377765655518,
+      "learning_rate": 1.5772357723577235e-05,
+      "loss": 0.0214,
+      "step": 3610
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.42401209473609924,
+      "learning_rate": 1.576342356830162e-05,
+      "loss": 0.0111,
+      "step": 3615
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4498884081840515,
+      "learning_rate": 1.5754489413026e-05,
+      "loss": 0.0144,
+      "step": 3620
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7784323692321777,
+      "learning_rate": 1.574555525775038e-05,
+      "loss": 0.0164,
+      "step": 3625
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.792809247970581,
+      "learning_rate": 1.5736621102474763e-05,
+      "loss": 0.0226,
+      "step": 3630
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5559731721878052,
+      "learning_rate": 1.5727686947199144e-05,
+      "loss": 0.0157,
+      "step": 3635
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5122278928756714,
+      "learning_rate": 1.5718752791923525e-05,
+      "loss": 0.0114,
+      "step": 3640
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.579097330570221,
+      "learning_rate": 1.5709818636647906e-05,
+      "loss": 0.0176,
+      "step": 3645
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5125916004180908,
+      "learning_rate": 1.5700884481372287e-05,
+      "loss": 0.0112,
+      "step": 3650
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6206502318382263,
+      "learning_rate": 1.569195032609667e-05,
+      "loss": 0.0129,
+      "step": 3655
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.578705906867981,
+      "learning_rate": 1.568301617082105e-05,
+      "loss": 0.0133,
+      "step": 3660
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.273260235786438,
+      "learning_rate": 1.567408201554543e-05,
+      "loss": 0.0238,
+      "step": 3665
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6316016912460327,
+      "learning_rate": 1.5665147860269814e-05,
+      "loss": 0.0107,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6251145005226135,
+      "learning_rate": 1.5656213704994195e-05,
+      "loss": 0.0115,
+      "step": 3675
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.48484379053115845,
+      "learning_rate": 1.5647279549718573e-05,
+      "loss": 0.0171,
+      "step": 3680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4522140920162201,
+      "learning_rate": 1.5638345394442958e-05,
+      "loss": 0.0222,
+      "step": 3685
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6267346739768982,
+      "learning_rate": 1.562941123916734e-05,
+      "loss": 0.0214,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9123010635375977,
+      "learning_rate": 1.562047708389172e-05,
+      "loss": 0.0214,
+      "step": 3695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2564802169799805,
+      "learning_rate": 1.56115429286161e-05,
+      "loss": 0.0209,
+      "step": 3700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4140920341014862,
+      "learning_rate": 1.5602608773340482e-05,
+      "loss": 0.0194,
+      "step": 3705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2953025102615356,
+      "learning_rate": 1.5593674618064863e-05,
+      "loss": 0.0175,
+      "step": 3710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.19866344332695007,
+      "learning_rate": 1.5584740462789244e-05,
+      "loss": 0.017,
+      "step": 3715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5271120071411133,
+      "learning_rate": 1.5575806307513625e-05,
+      "loss": 0.0137,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6619897484779358,
+      "learning_rate": 1.5566872152238006e-05,
+      "loss": 0.0184,
+      "step": 3725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7720787525177002,
+      "learning_rate": 1.555793799696239e-05,
+      "loss": 0.0198,
+      "step": 3730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2014611959457397,
+      "learning_rate": 1.5549003841686768e-05,
+      "loss": 0.0176,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42123496532440186,
+      "learning_rate": 1.554006968641115e-05,
+      "loss": 0.0136,
+      "step": 3740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8967819213867188,
+      "learning_rate": 1.5531135531135533e-05,
+      "loss": 0.0202,
+      "step": 3745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4060193598270416,
+      "learning_rate": 1.5522201375859914e-05,
+      "loss": 0.0183,
+      "step": 3750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6353384256362915,
+      "learning_rate": 1.5513267220584295e-05,
+      "loss": 0.0141,
+      "step": 3755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3105901777744293,
+      "learning_rate": 1.5504333065308677e-05,
+      "loss": 0.0123,
+      "step": 3760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.33913782238960266,
+      "learning_rate": 1.5495398910033058e-05,
+      "loss": 0.0176,
+      "step": 3765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.664908766746521,
+      "learning_rate": 1.548646475475744e-05,
+      "loss": 0.0135,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7868657112121582,
+      "learning_rate": 1.547753059948182e-05,
+      "loss": 0.0122,
+      "step": 3775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6869320869445801,
+      "learning_rate": 1.54685964442062e-05,
+      "loss": 0.0135,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1669894456863403,
+      "learning_rate": 1.5459662288930585e-05,
+      "loss": 0.024,
+      "step": 3785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.32140159606933594,
+      "learning_rate": 1.5450728133654963e-05,
+      "loss": 0.0109,
+      "step": 3790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5477967262268066,
+      "learning_rate": 1.5441793978379344e-05,
+      "loss": 0.0132,
+      "step": 3795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2863134741783142,
+      "learning_rate": 1.5432859823103728e-05,
+      "loss": 0.0225,
+      "step": 3800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7895782589912415,
+      "learning_rate": 1.542392566782811e-05,
+      "loss": 0.0265,
+      "step": 3805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.36137330532073975,
+      "learning_rate": 1.5414991512552487e-05,
+      "loss": 0.0114,
+      "step": 3810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5505108833312988,
+      "learning_rate": 1.540605735727687e-05,
+      "loss": 0.0244,
+      "step": 3815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2253755331039429,
+      "learning_rate": 1.5397123202001252e-05,
+      "loss": 0.0253,
+      "step": 3820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.25381264090538025,
+      "learning_rate": 1.5388189046725633e-05,
+      "loss": 0.0215,
+      "step": 3825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.59675133228302,
+      "learning_rate": 1.5379254891450014e-05,
+      "loss": 0.0229,
+      "step": 3830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4737885296344757,
+      "learning_rate": 1.5370320736174395e-05,
+      "loss": 0.0172,
+      "step": 3835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7271224856376648,
+      "learning_rate": 1.536138658089878e-05,
+      "loss": 0.0155,
+      "step": 3840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3435806930065155,
+      "learning_rate": 1.5352452425623158e-05,
+      "loss": 0.0115,
+      "step": 3845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.37383875250816345,
+      "learning_rate": 1.534351827034754e-05,
+      "loss": 0.0175,
+      "step": 3850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2769816219806671,
+      "learning_rate": 1.5334584115071923e-05,
+      "loss": 0.012,
+      "step": 3855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.421824187040329,
+      "learning_rate": 1.5325649959796304e-05,
+      "loss": 0.0168,
+      "step": 3860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.48985883593559265,
+      "learning_rate": 1.5316715804520682e-05,
+      "loss": 0.0121,
+      "step": 3865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4765077829360962,
+      "learning_rate": 1.5307781649245066e-05,
+      "loss": 0.0181,
+      "step": 3870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.48202696442604065,
+      "learning_rate": 1.5298847493969447e-05,
+      "loss": 0.0157,
+      "step": 3875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.033970594406128,
+      "learning_rate": 1.5289913338693828e-05,
+      "loss": 0.0136,
+      "step": 3880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.33205658197402954,
+      "learning_rate": 1.528097918341821e-05,
+      "loss": 0.012,
+      "step": 3885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4964057207107544,
+      "learning_rate": 1.527204502814259e-05,
+      "loss": 0.0109,
+      "step": 3890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9996275901794434,
+      "learning_rate": 1.526311087286697e-05,
+      "loss": 0.0153,
+      "step": 3895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5008103251457214,
+      "learning_rate": 1.5254176717591354e-05,
+      "loss": 0.0194,
+      "step": 3900
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 0.019112512469291687,
+      "eval_pearson_cosine": 0.9187946359889192,
+      "eval_pearson_dot": 0.916935359224543,
+      "eval_pearson_euclidean": 0.9078268066209557,
+      "eval_pearson_manhattan": 0.9068373107836385,
+      "eval_pearson_max": 0.9187946359889192,
+      "eval_runtime": 426.0704,
+      "eval_samples_per_second": 1.174,
+      "eval_spearman_cosine": 0.9320740642962572,
+      "eval_spearman_dot": 0.9297467269869079,
+      "eval_spearman_euclidean": 0.9318172632690529,
+      "eval_spearman_manhattan": 0.9317702230808924,
+      "eval_spearman_max": 0.9320740642962572,
+      "eval_steps_per_second": 1.174,
+      "step": 3900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5863137245178223,
+      "learning_rate": 1.5245242562315733e-05,
+      "loss": 0.0155,
+      "step": 3905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9501858353614807,
+      "learning_rate": 1.5236308407040114e-05,
+      "loss": 0.023,
+      "step": 3910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0070725679397583,
+      "learning_rate": 1.5227374251764497e-05,
+      "loss": 0.0159,
+      "step": 3915
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1535245180130005,
+      "learning_rate": 1.5218440096488878e-05,
+      "loss": 0.0271,
+      "step": 3920
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3500936031341553,
+      "learning_rate": 1.5209505941213261e-05,
+      "loss": 0.0185,
+      "step": 3925
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5063101649284363,
+      "learning_rate": 1.5200571785937642e-05,
+      "loss": 0.0092,
+      "step": 3930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.48119908571243286,
+      "learning_rate": 1.5191637630662021e-05,
+      "loss": 0.0156,
+      "step": 3935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7710624933242798,
+      "learning_rate": 1.5182703475386404e-05,
+      "loss": 0.0182,
+      "step": 3940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5304814577102661,
+      "learning_rate": 1.5173769320110785e-05,
+      "loss": 0.0127,
+      "step": 3945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5894374251365662,
+      "learning_rate": 1.5164835164835166e-05,
+      "loss": 0.012,
+      "step": 3950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.48509135842323303,
+      "learning_rate": 1.5155901009559549e-05,
+      "loss": 0.0164,
+      "step": 3955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.00001060962677,
+      "learning_rate": 1.5146966854283928e-05,
+      "loss": 0.0105,
+      "step": 3960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5003077983856201,
+      "learning_rate": 1.513803269900831e-05,
+      "loss": 0.0191,
+      "step": 3965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4220547378063202,
+      "learning_rate": 1.5129098543732692e-05,
+      "loss": 0.0182,
+      "step": 3970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.353935182094574,
+      "learning_rate": 1.5120164388457073e-05,
+      "loss": 0.0105,
+      "step": 3975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7862482070922852,
+      "learning_rate": 1.5111230233181452e-05,
+      "loss": 0.0153,
+      "step": 3980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4707240164279938,
+      "learning_rate": 1.5102296077905835e-05,
+      "loss": 0.0106,
+      "step": 3985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5706127882003784,
+      "learning_rate": 1.5093361922630216e-05,
+      "loss": 0.0193,
+      "step": 3990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5268707275390625,
+      "learning_rate": 1.5084427767354597e-05,
+      "loss": 0.0115,
+      "step": 3995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42871159315109253,
+      "learning_rate": 1.507549361207898e-05,
+      "loss": 0.0111,
+      "step": 4000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.47120198607444763,
+      "learning_rate": 1.5066559456803361e-05,
+      "loss": 0.0148,
+      "step": 4005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6913847327232361,
+      "learning_rate": 1.505762530152774e-05,
+      "loss": 0.0156,
+      "step": 4010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.40243229269981384,
+      "learning_rate": 1.5048691146252123e-05,
+      "loss": 0.021,
+      "step": 4015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4630833566188812,
+      "learning_rate": 1.5039756990976504e-05,
+      "loss": 0.0142,
+      "step": 4020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7278119325637817,
+      "learning_rate": 1.5030822835700887e-05,
+      "loss": 0.0221,
+      "step": 4025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5856547951698303,
+      "learning_rate": 1.5021888680425268e-05,
+      "loss": 0.0127,
+      "step": 4030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3612341582775116,
+      "learning_rate": 1.5012954525149647e-05,
+      "loss": 0.0142,
+      "step": 4035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3373230993747711,
+      "learning_rate": 1.500402036987403e-05,
+      "loss": 0.013,
+      "step": 4040
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3555310070514679,
+      "learning_rate": 1.4995086214598411e-05,
+      "loss": 0.0188,
+      "step": 4045
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27012017369270325,
+      "learning_rate": 1.4986152059322792e-05,
+      "loss": 0.016,
+      "step": 4050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7879682183265686,
+      "learning_rate": 1.4977217904047175e-05,
+      "loss": 0.019,
+      "step": 4055
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4750874638557434,
+      "learning_rate": 1.4968283748771554e-05,
+      "loss": 0.0156,
+      "step": 4060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5031574368476868,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.0206,
+      "step": 4065
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5057447552680969,
+      "learning_rate": 1.4950415438220318e-05,
+      "loss": 0.0144,
+      "step": 4070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0315029621124268,
+      "learning_rate": 1.4941481282944699e-05,
+      "loss": 0.0179,
+      "step": 4075
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5417433977127075,
+      "learning_rate": 1.493254712766908e-05,
+      "loss": 0.0126,
+      "step": 4080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.667726993560791,
+      "learning_rate": 1.4923612972393463e-05,
+      "loss": 0.0143,
+      "step": 4085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8723002672195435,
+      "learning_rate": 1.4914678817117842e-05,
+      "loss": 0.012,
+      "step": 4090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8944787979125977,
+      "learning_rate": 1.4905744661842223e-05,
+      "loss": 0.0205,
+      "step": 4095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5145695805549622,
+      "learning_rate": 1.4896810506566606e-05,
+      "loss": 0.0118,
+      "step": 4100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9632664322853088,
+      "learning_rate": 1.4887876351290987e-05,
+      "loss": 0.0233,
+      "step": 4105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4608353078365326,
+      "learning_rate": 1.487894219601537e-05,
+      "loss": 0.0192,
+      "step": 4110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8873416781425476,
+      "learning_rate": 1.4870008040739749e-05,
+      "loss": 0.0196,
+      "step": 4115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7420912384986877,
+      "learning_rate": 1.486107388546413e-05,
+      "loss": 0.0114,
+      "step": 4120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3144592344760895,
+      "learning_rate": 1.4852139730188513e-05,
+      "loss": 0.0136,
+      "step": 4125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4143444001674652,
+      "learning_rate": 1.4843205574912894e-05,
+      "loss": 0.0102,
+      "step": 4130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3450114727020264,
+      "learning_rate": 1.4834271419637275e-05,
+      "loss": 0.0273,
+      "step": 4135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6474159955978394,
+      "learning_rate": 1.4825337264361657e-05,
+      "loss": 0.0125,
+      "step": 4140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3440127372741699,
+      "learning_rate": 1.4816403109086037e-05,
+      "loss": 0.016,
+      "step": 4145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6236218214035034,
+      "learning_rate": 1.4807468953810418e-05,
+      "loss": 0.019,
+      "step": 4150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9320889115333557,
+      "learning_rate": 1.47985347985348e-05,
+      "loss": 0.015,
+      "step": 4155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.36292609572410583,
+      "learning_rate": 1.4789600643259182e-05,
+      "loss": 0.0155,
+      "step": 4160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.278648018836975,
+      "learning_rate": 1.4780666487983561e-05,
+      "loss": 0.0229,
+      "step": 4165
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.24334418773651123,
+      "learning_rate": 1.4771732332707944e-05,
+      "loss": 0.0099,
+      "step": 4170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7839992046356201,
+      "learning_rate": 1.4762798177432325e-05,
+      "loss": 0.0167,
+      "step": 4175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3292075395584106,
+      "learning_rate": 1.4753864022156706e-05,
+      "loss": 0.0168,
+      "step": 4180
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4192187488079071,
+      "learning_rate": 1.4744929866881089e-05,
+      "loss": 0.0177,
+      "step": 4185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5623775124549866,
+      "learning_rate": 1.4735995711605468e-05,
+      "loss": 0.0103,
+      "step": 4190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5713854432106018,
+      "learning_rate": 1.4727061556329849e-05,
+      "loss": 0.011,
+      "step": 4195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6574892997741699,
+      "learning_rate": 1.4718127401054232e-05,
+      "loss": 0.0169,
+      "step": 4200
+    },
+    {
+      "epoch": 0.34,
+      "eval_loss": 0.01563582755625248,
+      "eval_pearson_cosine": 0.93071357698338,
+      "eval_pearson_dot": 0.9303618828257327,
+      "eval_pearson_euclidean": 0.9158033535152222,
+      "eval_pearson_manhattan": 0.9136122746167228,
+      "eval_pearson_max": 0.93071357698338,
+      "eval_runtime": 426.316,
+      "eval_samples_per_second": 1.173,
+      "eval_spearman_cosine": 0.9357050868203473,
+      "eval_spearman_dot": 0.9326605306421225,
+      "eval_spearman_euclidean": 0.9346635786543147,
+      "eval_spearman_manhattan": 0.9334011736046943,
+      "eval_spearman_max": 0.9357050868203473,
+      "eval_steps_per_second": 1.173,
+      "step": 4200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5762566328048706,
+      "learning_rate": 1.4709193245778613e-05,
+      "loss": 0.0296,
+      "step": 4205
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6294378638267517,
+      "learning_rate": 1.4700259090502995e-05,
+      "loss": 0.0158,
+      "step": 4210
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5651215314865112,
+      "learning_rate": 1.4691324935227376e-05,
+      "loss": 0.0131,
+      "step": 4215
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6574472784996033,
+      "learning_rate": 1.4682390779951756e-05,
+      "loss": 0.0166,
+      "step": 4220
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4540180563926697,
+      "learning_rate": 1.4673456624676139e-05,
+      "loss": 0.024,
+      "step": 4225
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3493760824203491,
+      "learning_rate": 1.466452246940052e-05,
+      "loss": 0.0149,
+      "step": 4230
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0167644023895264,
+      "learning_rate": 1.46555883141249e-05,
+      "loss": 0.0149,
+      "step": 4235
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9241525530815125,
+      "learning_rate": 1.4646654158849283e-05,
+      "loss": 0.0184,
+      "step": 4240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6745620965957642,
+      "learning_rate": 1.4637720003573663e-05,
+      "loss": 0.0158,
+      "step": 4245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4359780251979828,
+      "learning_rate": 1.4628785848298044e-05,
+      "loss": 0.0098,
+      "step": 4250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4030389189720154,
+      "learning_rate": 1.4619851693022426e-05,
+      "loss": 0.0184,
+      "step": 4255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5261355638504028,
+      "learning_rate": 1.4610917537746807e-05,
+      "loss": 0.0141,
+      "step": 4260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.38097327947616577,
+      "learning_rate": 1.4601983382471187e-05,
+      "loss": 0.0083,
+      "step": 4265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.41486430168151855,
+      "learning_rate": 1.459304922719557e-05,
+      "loss": 0.0202,
+      "step": 4270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5645167827606201,
+      "learning_rate": 1.458411507191995e-05,
+      "loss": 0.0126,
+      "step": 4275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4771600365638733,
+      "learning_rate": 1.4575180916644332e-05,
+      "loss": 0.0103,
+      "step": 4280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8267874717712402,
+      "learning_rate": 1.4566246761368714e-05,
+      "loss": 0.0182,
+      "step": 4285
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7216476202011108,
+      "learning_rate": 1.4557312606093095e-05,
+      "loss": 0.0155,
+      "step": 4290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0095362663269043,
+      "learning_rate": 1.4548378450817478e-05,
+      "loss": 0.0199,
+      "step": 4295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.36790329217910767,
+      "learning_rate": 1.4539444295541857e-05,
+      "loss": 0.0231,
+      "step": 4300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4778492748737335,
+      "learning_rate": 1.4530510140266239e-05,
+      "loss": 0.0101,
+      "step": 4305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3736858665943146,
+      "learning_rate": 1.4521575984990621e-05,
+      "loss": 0.0157,
+      "step": 4310
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5705660581588745,
+      "learning_rate": 1.4512641829715002e-05,
+      "loss": 0.0133,
+      "step": 4315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7104588747024536,
+      "learning_rate": 1.4503707674439382e-05,
+      "loss": 0.0209,
+      "step": 4320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4687543511390686,
+      "learning_rate": 1.4494773519163764e-05,
+      "loss": 0.0072,
+      "step": 4325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5142909288406372,
+      "learning_rate": 1.4485839363888145e-05,
+      "loss": 0.0127,
+      "step": 4330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3397541046142578,
+      "learning_rate": 1.4476905208612526e-05,
+      "loss": 0.0122,
+      "step": 4335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4727453589439392,
+      "learning_rate": 1.446797105333691e-05,
+      "loss": 0.0212,
+      "step": 4340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7717282176017761,
+      "learning_rate": 1.445903689806129e-05,
+      "loss": 0.0277,
+      "step": 4345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4375796914100647,
+      "learning_rate": 1.445010274278567e-05,
+      "loss": 0.0119,
+      "step": 4350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.45401817560195923,
+      "learning_rate": 1.4441168587510052e-05,
+      "loss": 0.0128,
+      "step": 4355
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4388526976108551,
+      "learning_rate": 1.4432234432234433e-05,
+      "loss": 0.0181,
+      "step": 4360
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4506477415561676,
+      "learning_rate": 1.4423300276958814e-05,
+      "loss": 0.0118,
+      "step": 4365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7086319923400879,
+      "learning_rate": 1.4414366121683197e-05,
+      "loss": 0.0111,
+      "step": 4370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.185771182179451,
+      "learning_rate": 1.4405431966407576e-05,
+      "loss": 0.0151,
+      "step": 4375
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26904481649398804,
+      "learning_rate": 1.4396497811131958e-05,
+      "loss": 0.0116,
+      "step": 4380
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8145022392272949,
+      "learning_rate": 1.438756365585634e-05,
+      "loss": 0.0097,
+      "step": 4385
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1931172609329224,
+      "learning_rate": 1.4378629500580721e-05,
+      "loss": 0.0158,
+      "step": 4390
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.37602362036705017,
+      "learning_rate": 1.4369695345305104e-05,
+      "loss": 0.0145,
+      "step": 4395
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5753958225250244,
+      "learning_rate": 1.4360761190029483e-05,
+      "loss": 0.0156,
+      "step": 4400
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 12437,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 400,
+  "total_flos": 0.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}