diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,25794 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 3219,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "cosine_similarity": 0.049044237750423235,
+      "epoch": 0.0009319664492078285,
+      "grad_norm": 5.174747826218702,
+      "learning_rate": 6.211180124223603e-08,
+      "loss": 4.6296,
+      "step": 1
+    },
+    {
+      "cosine_similarity": 0.07432216181333663,
+      "epoch": 0.001863932898415657,
+      "grad_norm": 4.905220517720085,
+      "learning_rate": 1.2422360248447206e-07,
+      "loss": 4.1935,
+      "step": 2
+    },
+    {
+      "cosine_similarity": 0.05477264307342341,
+      "epoch": 0.0027958993476234857,
+      "grad_norm": 5.6070973769499215,
+      "learning_rate": 1.863354037267081e-07,
+      "loss": 4.748,
+      "step": 3
+    },
+    {
+      "cosine_similarity": 0.09865303367717819,
+      "epoch": 0.003727865796831314,
+      "grad_norm": 6.422703512199275,
+      "learning_rate": 2.484472049689441e-07,
+      "loss": 5.0671,
+      "step": 4
+    },
+    {
+      "cosine_similarity": 0.053827668947926656,
+      "epoch": 0.004659832246039142,
+      "grad_norm": 6.010404060241676,
+      "learning_rate": 3.1055900621118013e-07,
+      "loss": 4.9616,
+      "step": 5
+    },
+    {
+      "cosine_similarity": 0.06831189120575962,
+      "epoch": 0.005591798695246971,
+      "grad_norm": 5.531425002735728,
+      "learning_rate": 3.726708074534162e-07,
+      "loss": 4.9479,
+      "step": 6
+    },
+    {
+      "cosine_similarity": 0.10929847841771693,
+      "epoch": 0.0065237651444548,
+      "grad_norm": 6.697057960537303,
+      "learning_rate": 4.347826086956522e-07,
+      "loss": 5.1478,
+      "step": 7
+    },
+    {
+      "cosine_similarity": 0.11403296691687877,
+      "epoch": 0.007455731593662628,
+      "grad_norm": 6.441772796952103,
+      "learning_rate": 4.968944099378882e-07,
+      "loss": 5.0133,
+      "step": 8
+    },
+    {
+      "cosine_similarity": 0.07734445967032566,
+      "epoch": 0.008387698042870456,
+      "grad_norm": 6.054698461271223,
+      "learning_rate": 5.590062111801243e-07,
+      "loss": 4.9878,
+      "step": 9
+    },
+    {
+      "cosine_similarity": 0.028995507276549197,
+      "epoch": 0.009319664492078284,
+      "grad_norm": 5.660826137146643,
+      "learning_rate": 6.211180124223603e-07,
+      "loss": 4.6407,
+      "step": 10
+    },
+    {
+      "cosine_similarity": 0.05338083801507462,
+      "epoch": 0.010251630941286114,
+      "grad_norm": 6.478680811962412,
+      "learning_rate": 6.832298136645964e-07,
+      "loss": 4.8958,
+      "step": 11
+    },
+    {
+      "cosine_similarity": 0.05442403462050067,
+      "epoch": 0.011183597390493943,
+      "grad_norm": 5.629380719444735,
+      "learning_rate": 7.453416149068324e-07,
+      "loss": 4.9872,
+      "step": 12
+    },
+    {
+      "cosine_similarity": 0.1060289915204631,
+      "epoch": 0.012115563839701771,
+      "grad_norm": 7.330936700110883,
+      "learning_rate": 8.074534161490684e-07,
+      "loss": 5.0353,
+      "step": 13
+    },
+    {
+      "cosine_similarity": 0.04972427157935803,
+      "epoch": 0.0130475302889096,
+      "grad_norm": 5.9438816998979105,
+      "learning_rate": 8.695652173913044e-07,
+      "loss": 4.6167,
+      "step": 14
+    },
+    {
+      "cosine_similarity": 0.07169916716531984,
+      "epoch": 0.013979496738117428,
+      "grad_norm": 5.193693889110981,
+      "learning_rate": 9.316770186335404e-07,
+      "loss": 4.1442,
+      "step": 15
+    },
+    {
+      "cosine_similarity": 0.06657293446836592,
+      "epoch": 0.014911463187325256,
+      "grad_norm": 6.594882730330985,
+      "learning_rate": 9.937888198757765e-07,
+      "loss": 4.9836,
+      "step": 16
+    },
+    {
+      "cosine_similarity": 0.07008458388837156,
+      "epoch": 0.015843429636533086,
+      "grad_norm": 5.57838689382849,
+      "learning_rate": 1.0559006211180126e-06,
+      "loss": 4.4986,
+      "step": 17
+    },
+    {
+      "cosine_similarity": 0.06477858184234728,
+      "epoch": 0.016775396085740912,
+      "grad_norm": 5.402642890634636,
+      "learning_rate": 1.1180124223602485e-06,
+      "loss": 4.3296,
+      "step": 18
+    },
+    {
+      "cosine_similarity": 0.09738996952677464,
+      "epoch": 0.017707362534948742,
+      "grad_norm": 5.63871972223129,
+      "learning_rate": 1.1801242236024846e-06,
+      "loss": 4.9793,
+      "step": 19
+    },
+    {
+      "cosine_similarity": 0.06470522593871468,
+      "epoch": 0.01863932898415657,
+      "grad_norm": 5.195696977007632,
+      "learning_rate": 1.2422360248447205e-06,
+      "loss": 4.3063,
+      "step": 20
+    },
+    {
+      "cosine_similarity": 0.07011620842258434,
+      "epoch": 0.0195712954333644,
+      "grad_norm": 6.180261829081403,
+      "learning_rate": 1.3043478260869566e-06,
+      "loss": 4.7375,
+      "step": 21
+    },
+    {
+      "cosine_similarity": 0.08045866520062946,
+      "epoch": 0.02050326188257223,
+      "grad_norm": 4.724180302790162,
+      "learning_rate": 1.3664596273291927e-06,
+      "loss": 4.1236,
+      "step": 22
+    },
+    {
+      "cosine_similarity": 0.06736945641168617,
+      "epoch": 0.021435228331780055,
+      "grad_norm": 4.655479183547909,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 4.0861,
+      "step": 23
+    },
+    {
+      "cosine_similarity": 0.10465911587379916,
+      "epoch": 0.022367194780987885,
+      "grad_norm": 4.942323956797542,
+      "learning_rate": 1.4906832298136647e-06,
+      "loss": 4.3504,
+      "step": 24
+    },
+    {
+      "cosine_similarity": 0.07721782095236539,
+      "epoch": 0.023299161230195712,
+      "grad_norm": 5.149134643372241,
+      "learning_rate": 1.5527950310559006e-06,
+      "loss": 4.1863,
+      "step": 25
+    },
+    {
+      "cosine_similarity": 0.044250163960668595,
+      "epoch": 0.024231127679403542,
+      "grad_norm": 5.46714867492111,
+      "learning_rate": 1.6149068322981367e-06,
+      "loss": 3.6513,
+      "step": 26
+    },
+    {
+      "cosine_similarity": 0.10593045318908996,
+      "epoch": 0.02516309412861137,
+      "grad_norm": 4.903102404610553,
+      "learning_rate": 1.6770186335403729e-06,
+      "loss": 4.3304,
+      "step": 27
+    },
+    {
+      "cosine_similarity": 0.08792687688307571,
+      "epoch": 0.0260950605778192,
+      "grad_norm": 4.827765784060252,
+      "learning_rate": 1.7391304347826088e-06,
+      "loss": 3.8479,
+      "step": 28
+    },
+    {
+      "cosine_similarity": 0.13876792372426938,
+      "epoch": 0.02702702702702703,
+      "grad_norm": 4.221052734713945,
+      "learning_rate": 1.8012422360248449e-06,
+      "loss": 3.962,
+      "step": 29
+    },
+    {
+      "cosine_similarity": 0.13836180304786858,
+      "epoch": 0.027958993476234855,
+      "grad_norm": 3.361326022920078,
+      "learning_rate": 1.8633540372670808e-06,
+      "loss": 3.5766,
+      "step": 30
+    },
+    {
+      "cosine_similarity": 0.14363758674093485,
+      "epoch": 0.028890959925442685,
+      "grad_norm": 4.239101647213156,
+      "learning_rate": 1.925465838509317e-06,
+      "loss": 3.8754,
+      "step": 31
+    },
+    {
+      "cosine_similarity": 0.1301115896302442,
+      "epoch": 0.02982292637465051,
+      "grad_norm": 4.4712872736793505,
+      "learning_rate": 1.987577639751553e-06,
+      "loss": 3.752,
+      "step": 32
+    },
+    {
+      "cosine_similarity": 0.14814068436163,
+      "epoch": 0.03075489282385834,
+      "grad_norm": 4.62711481738544,
+      "learning_rate": 2.049689440993789e-06,
+      "loss": 3.9298,
+      "step": 33
+    },
+    {
+      "cosine_similarity": 0.1951871800617607,
+      "epoch": 0.03168685927306617,
+      "grad_norm": 4.213858192199002,
+      "learning_rate": 2.111801242236025e-06,
+      "loss": 3.6055,
+      "step": 34
+    },
+    {
+      "cosine_similarity": 0.24927353987707482,
+      "epoch": 0.032618825722273995,
+      "grad_norm": 4.260307618842313,
+      "learning_rate": 2.173913043478261e-06,
+      "loss": 3.96,
+      "step": 35
+    },
+    {
+      "cosine_similarity": 0.2695495309346067,
+      "epoch": 0.033550792171481825,
+      "grad_norm": 4.047856333632422,
+      "learning_rate": 2.236024844720497e-06,
+      "loss": 3.7037,
+      "step": 36
+    },
+    {
+      "cosine_similarity": 0.23875804386483473,
+      "epoch": 0.034482758620689655,
+      "grad_norm": 3.974566153381094,
+      "learning_rate": 2.298136645962733e-06,
+      "loss": 3.8045,
+      "step": 37
+    },
+    {
+      "cosine_similarity": 0.16586516683182137,
+      "epoch": 0.035414725069897485,
+      "grad_norm": 4.417864781905701,
+      "learning_rate": 2.3602484472049692e-06,
+      "loss": 3.7716,
+      "step": 38
+    },
+    {
+      "cosine_similarity": 0.2383775822139317,
+      "epoch": 0.036346691519105315,
+      "grad_norm": 3.85744208770018,
+      "learning_rate": 2.422360248447205e-06,
+      "loss": 3.8129,
+      "step": 39
+    },
+    {
+      "cosine_similarity": 0.2491386252006945,
+      "epoch": 0.03727865796831314,
+      "grad_norm": 3.8314641853913405,
+      "learning_rate": 2.484472049689441e-06,
+      "loss": 3.7409,
+      "step": 40
+    },
+    {
+      "cosine_similarity": 0.1729240835387595,
+      "epoch": 0.03821062441752097,
+      "grad_norm": 3.2651566335212516,
+      "learning_rate": 2.546583850931677e-06,
+      "loss": 3.3717,
+      "step": 41
+    },
+    {
+      "cosine_similarity": 0.21686576452239834,
+      "epoch": 0.0391425908667288,
+      "grad_norm": 4.6117226407523395,
+      "learning_rate": 2.6086956521739132e-06,
+      "loss": 3.6157,
+      "step": 42
+    },
+    {
+      "cosine_similarity": 0.24475807613410083,
+      "epoch": 0.04007455731593663,
+      "grad_norm": 3.482276859652839,
+      "learning_rate": 2.670807453416149e-06,
+      "loss": 3.5538,
+      "step": 43
+    },
+    {
+      "cosine_similarity": 0.24579000705911164,
+      "epoch": 0.04100652376514446,
+      "grad_norm": 3.484337922315546,
+      "learning_rate": 2.7329192546583855e-06,
+      "loss": 3.5558,
+      "step": 44
+    },
+    {
+      "cosine_similarity": 0.24496950511618756,
+      "epoch": 0.04193849021435228,
+      "grad_norm": 3.3963089735429195,
+      "learning_rate": 2.795031055900621e-06,
+      "loss": 3.647,
+      "step": 45
+    },
+    {
+      "cosine_similarity": 0.20433759707322022,
+      "epoch": 0.04287045666356011,
+      "grad_norm": 3.4464999084977563,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 2.9435,
+      "step": 46
+    },
+    {
+      "cosine_similarity": 0.12328302636445328,
+      "epoch": 0.04380242311276794,
+      "grad_norm": 3.6073521421449497,
+      "learning_rate": 2.919254658385093e-06,
+      "loss": 3.3391,
+      "step": 47
+    },
+    {
+      "cosine_similarity": 0.18483818453964024,
+      "epoch": 0.04473438956197577,
+      "grad_norm": 3.2404718187424866,
+      "learning_rate": 2.9813664596273295e-06,
+      "loss": 3.7476,
+      "step": 48
+    },
+    {
+      "cosine_similarity": 0.15193119498009877,
+      "epoch": 0.045666356011183594,
+      "grad_norm": 3.82221062431233,
+      "learning_rate": 3.043478260869566e-06,
+      "loss": 3.4744,
+      "step": 49
+    },
+    {
+      "cosine_similarity": 0.20963567345451825,
+      "epoch": 0.046598322460391424,
+      "grad_norm": 3.88209847981197,
+      "learning_rate": 3.1055900621118013e-06,
+      "loss": 3.3236,
+      "step": 50
+    },
+    {
+      "cosine_similarity": 0.3261588835530013,
+      "epoch": 0.047530288909599254,
+      "grad_norm": 5.507639468985589,
+      "learning_rate": 3.1677018633540376e-06,
+      "loss": 3.3221,
+      "step": 51
+    },
+    {
+      "cosine_similarity": 0.1454631457839041,
+      "epoch": 0.048462255358807084,
+      "grad_norm": 4.715007702531043,
+      "learning_rate": 3.2298136645962735e-06,
+      "loss": 3.232,
+      "step": 52
+    },
+    {
+      "cosine_similarity": -0.0011571119532082824,
+      "epoch": 0.049394221808014914,
+      "grad_norm": 3.2380148383779943,
+      "learning_rate": 3.29192546583851e-06,
+      "loss": 3.5702,
+      "step": 53
+    },
+    {
+      "cosine_similarity": 0.25766475204899847,
+      "epoch": 0.05032618825722274,
+      "grad_norm": 4.466518961608811,
+      "learning_rate": 3.3540372670807457e-06,
+      "loss": 3.4877,
+      "step": 54
+    },
+    {
+      "cosine_similarity": 0.22011978253212744,
+      "epoch": 0.05125815470643057,
+      "grad_norm": 3.7617185823598147,
+      "learning_rate": 3.4161490683229816e-06,
+      "loss": 3.602,
+      "step": 55
+    },
+    {
+      "cosine_similarity": 0.11883796290153766,
+      "epoch": 0.0521901211556384,
+      "grad_norm": 2.993355268740408,
+      "learning_rate": 3.4782608695652175e-06,
+      "loss": 3.1671,
+      "step": 56
+    },
+    {
+      "cosine_similarity": 0.0913175398618713,
+      "epoch": 0.05312208760484623,
+      "grad_norm": 7.279082549581904,
+      "learning_rate": 3.540372670807454e-06,
+      "loss": 3.6167,
+      "step": 57
+    },
+    {
+      "cosine_similarity": 0.04323035945730606,
+      "epoch": 0.05405405405405406,
+      "grad_norm": 3.063265789448478,
+      "learning_rate": 3.6024844720496897e-06,
+      "loss": 3.5314,
+      "step": 58
+    },
+    {
+      "cosine_similarity": 0.020764274554297777,
+      "epoch": 0.05498602050326188,
+      "grad_norm": 3.2603697681957553,
+      "learning_rate": 3.664596273291926e-06,
+      "loss": 3.6319,
+      "step": 59
+    },
+    {
+      "cosine_similarity": 0.05006690163334345,
+      "epoch": 0.05591798695246971,
+      "grad_norm": 3.289196955865107,
+      "learning_rate": 3.7267080745341615e-06,
+      "loss": 3.5556,
+      "step": 60
+    },
+    {
+      "cosine_similarity": -0.11310769067906375,
+      "epoch": 0.05684995340167754,
+      "grad_norm": 3.7131385456269803,
+      "learning_rate": 3.788819875776398e-06,
+      "loss": 3.061,
+      "step": 61
+    },
+    {
+      "cosine_similarity": 0.16882161751373778,
+      "epoch": 0.05778191985088537,
+      "grad_norm": 3.7043734224249647,
+      "learning_rate": 3.850931677018634e-06,
+      "loss": 3.3152,
+      "step": 62
+    },
+    {
+      "cosine_similarity": 0.02069913662820599,
+      "epoch": 0.05871388630009319,
+      "grad_norm": 2.7576749054271197,
+      "learning_rate": 3.91304347826087e-06,
+      "loss": 2.7973,
+      "step": 63
+    },
+    {
+      "cosine_similarity": 0.02743213758381044,
+      "epoch": 0.05964585274930102,
+      "grad_norm": 3.305812727081138,
+      "learning_rate": 3.975155279503106e-06,
+      "loss": 3.264,
+      "step": 64
+    },
+    {
+      "cosine_similarity": 0.14837094926584057,
+      "epoch": 0.06057781919850885,
+      "grad_norm": 2.858638916665973,
+      "learning_rate": 4.037267080745342e-06,
+      "loss": 3.4028,
+      "step": 65
+    },
+    {
+      "cosine_similarity": 0.1756202336674398,
+      "epoch": 0.06150978564771668,
+      "grad_norm": 2.5148214396554995,
+      "learning_rate": 4.099378881987578e-06,
+      "loss": 3.0922,
+      "step": 66
+    },
+    {
+      "cosine_similarity": 0.08992767711758687,
+      "epoch": 0.06244175209692451,
+      "grad_norm": 3.0093653873791086,
+      "learning_rate": 4.1614906832298145e-06,
+      "loss": 3.1175,
+      "step": 67
+    },
+    {
+      "cosine_similarity": 0.07266539111497981,
+      "epoch": 0.06337371854613234,
+      "grad_norm": 3.1730753225740056,
+      "learning_rate": 4.22360248447205e-06,
+      "loss": 3.0173,
+      "step": 68
+    },
+    {
+      "cosine_similarity": 0.18691765404360278,
+      "epoch": 0.06430568499534017,
+      "grad_norm": 3.380559560280933,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 3.5192,
+      "step": 69
+    },
+    {
+      "cosine_similarity": 0.11796335563166692,
+      "epoch": 0.06523765144454799,
+      "grad_norm": 2.9077104204861013,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 3.3629,
+      "step": 70
+    },
+    {
+      "cosine_similarity": 0.07765571343818317,
+      "epoch": 0.06616961789375582,
+      "grad_norm": 2.81712273827463,
+      "learning_rate": 4.409937888198758e-06,
+      "loss": 3.4043,
+      "step": 71
+    },
+    {
+      "cosine_similarity": -0.005905371568084928,
+      "epoch": 0.06710158434296365,
+      "grad_norm": 2.8308671698650603,
+      "learning_rate": 4.472049689440994e-06,
+      "loss": 3.2782,
+      "step": 72
+    },
+    {
+      "cosine_similarity": -0.028849168785963768,
+      "epoch": 0.06803355079217148,
+      "grad_norm": 3.518102128947131,
+      "learning_rate": 4.534161490683231e-06,
+      "loss": 3.4272,
+      "step": 73
+    },
+    {
+      "cosine_similarity": 0.13081321860671472,
+      "epoch": 0.06896551724137931,
+      "grad_norm": 2.6865474511762684,
+      "learning_rate": 4.596273291925466e-06,
+      "loss": 3.0313,
+      "step": 74
+    },
+    {
+      "cosine_similarity": 0.01857335565448794,
+      "epoch": 0.06989748369058714,
+      "grad_norm": 4.237893268910791,
+      "learning_rate": 4.6583850931677025e-06,
+      "loss": 3.2184,
+      "step": 75
+    },
+    {
+      "cosine_similarity": -0.061990325988919986,
+      "epoch": 0.07082945013979497,
+      "grad_norm": 2.8560462101055193,
+      "learning_rate": 4.7204968944099384e-06,
+      "loss": 3.0826,
+      "step": 76
+    },
+    {
+      "cosine_similarity": -0.07769631064425384,
+      "epoch": 0.0717614165890028,
+      "grad_norm": 2.697660870633607,
+      "learning_rate": 4.782608695652174e-06,
+      "loss": 2.6425,
+      "step": 77
+    },
+    {
+      "cosine_similarity": 0.03379178485325301,
+      "epoch": 0.07269338303821063,
+      "grad_norm": 3.461750561598614,
+      "learning_rate": 4.84472049689441e-06,
+      "loss": 3.1374,
+      "step": 78
+    },
+    {
+      "cosine_similarity": 0.08790697162845619,
+      "epoch": 0.07362534948741846,
+      "grad_norm": 2.9553106531061504,
+      "learning_rate": 4.906832298136646e-06,
+      "loss": 2.9852,
+      "step": 79
+    },
+    {
+      "cosine_similarity": 0.01551796419618813,
+      "epoch": 0.07455731593662628,
+      "grad_norm": 2.579611560131107,
+      "learning_rate": 4.968944099378882e-06,
+      "loss": 2.9431,
+      "step": 80
+    },
+    {
+      "cosine_similarity": -0.05199207151302162,
+      "epoch": 0.0754892823858341,
+      "grad_norm": 3.344913061353071,
+      "learning_rate": 5.031055900621118e-06,
+      "loss": 3.1528,
+      "step": 81
+    },
+    {
+      "cosine_similarity": -0.09804524151604324,
+      "epoch": 0.07642124883504194,
+      "grad_norm": 2.9477693101289213,
+      "learning_rate": 5.093167701863354e-06,
+      "loss": 3.2263,
+      "step": 82
+    },
+    {
+      "cosine_similarity": 0.04493768019680253,
+      "epoch": 0.07735321528424977,
+      "grad_norm": 3.0159497145898087,
+      "learning_rate": 5.155279503105591e-06,
+      "loss": 2.8373,
+      "step": 83
+    },
+    {
+      "cosine_similarity": -0.02560940006986939,
+      "epoch": 0.0782851817334576,
+      "grad_norm": 1.9346018979718442,
+      "learning_rate": 5.2173913043478265e-06,
+      "loss": 2.8216,
+      "step": 84
+    },
+    {
+      "cosine_similarity": 0.0047267089493431115,
+      "epoch": 0.07921714818266543,
+      "grad_norm": 3.5115423295838655,
+      "learning_rate": 5.279503105590062e-06,
+      "loss": 2.8983,
+      "step": 85
+    },
+    {
+      "cosine_similarity": 0.04777551274715519,
+      "epoch": 0.08014911463187326,
+      "grad_norm": 2.3285273256130687,
+      "learning_rate": 5.341614906832298e-06,
+      "loss": 3.392,
+      "step": 86
+    },
+    {
+      "cosine_similarity": 0.15837026972135484,
+      "epoch": 0.08108108108108109,
+      "grad_norm": 2.6548254048514397,
+      "learning_rate": 5.403726708074535e-06,
+      "loss": 3.6959,
+      "step": 87
+    },
+    {
+      "cosine_similarity": 0.12936241686509342,
+      "epoch": 0.08201304753028892,
+      "grad_norm": 3.001230907208606,
+      "learning_rate": 5.465838509316771e-06,
+      "loss": 3.2694,
+      "step": 88
+    },
+    {
+      "cosine_similarity": -0.040637434905574914,
+      "epoch": 0.08294501397949673,
+      "grad_norm": 2.473864082356364,
+      "learning_rate": 5.527950310559007e-06,
+      "loss": 3.1506,
+      "step": 89
+    },
+    {
+      "cosine_similarity": 0.090851368582997,
+      "epoch": 0.08387698042870456,
+      "grad_norm": 2.5085551262143415,
+      "learning_rate": 5.590062111801242e-06,
+      "loss": 2.7281,
+      "step": 90
+    },
+    {
+      "cosine_similarity": 0.00410117628562569,
+      "epoch": 0.08480894687791239,
+      "grad_norm": 2.2437454530760803,
+      "learning_rate": 5.652173913043479e-06,
+      "loss": 2.8306,
+      "step": 91
+    },
+    {
+      "cosine_similarity": 0.12056030317065973,
+      "epoch": 0.08574091332712022,
+      "grad_norm": 2.3592448424989265,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 2.7541,
+      "step": 92
+    },
+    {
+      "cosine_similarity": 0.05655905131466834,
+      "epoch": 0.08667287977632805,
+      "grad_norm": 3.2728637678380648,
+      "learning_rate": 5.77639751552795e-06,
+      "loss": 3.0322,
+      "step": 93
+    },
+    {
+      "cosine_similarity": -0.009127957087249611,
+      "epoch": 0.08760484622553588,
+      "grad_norm": 2.605975016060143,
+      "learning_rate": 5.838509316770186e-06,
+      "loss": 2.9535,
+      "step": 94
+    },
+    {
+      "cosine_similarity": 0.16142528664542521,
+      "epoch": 0.08853681267474371,
+      "grad_norm": 2.7112046610568092,
+      "learning_rate": 5.900621118012423e-06,
+      "loss": 3.0488,
+      "step": 95
+    },
+    {
+      "cosine_similarity": 0.04000625724413939,
+      "epoch": 0.08946877912395154,
+      "grad_norm": 2.8380246824471946,
+      "learning_rate": 5.962732919254659e-06,
+      "loss": 3.1753,
+      "step": 96
+    },
+    {
+      "cosine_similarity": 0.2228138647888587,
+      "epoch": 0.09040074557315937,
+      "grad_norm": 2.2788847114568735,
+      "learning_rate": 6.024844720496895e-06,
+      "loss": 2.799,
+      "step": 97
+    },
+    {
+      "cosine_similarity": 0.006548131520681677,
+      "epoch": 0.09133271202236719,
+      "grad_norm": 2.625569695177881,
+      "learning_rate": 6.086956521739132e-06,
+      "loss": 2.524,
+      "step": 98
+    },
+    {
+      "cosine_similarity": -0.03157155082147631,
+      "epoch": 0.09226467847157502,
+      "grad_norm": 2.9528242829284177,
+      "learning_rate": 6.1490683229813675e-06,
+      "loss": 3.0968,
+      "step": 99
+    },
+    {
+      "cosine_similarity": 0.09034029043682348,
+      "epoch": 0.09319664492078285,
+      "grad_norm": 3.0235277803812135,
+      "learning_rate": 6.2111801242236025e-06,
+      "loss": 2.8286,
+      "step": 100
+    },
+    {
+      "cosine_similarity": 0.07436478731257104,
+      "epoch": 0.09412861136999068,
+      "grad_norm": 2.560906124636941,
+      "learning_rate": 6.2732919254658384e-06,
+      "loss": 2.8044,
+      "step": 101
+    },
+    {
+      "cosine_similarity": -0.15583003144735535,
+      "epoch": 0.09506057781919851,
+      "grad_norm": 2.7500777886057732,
+      "learning_rate": 6.335403726708075e-06,
+      "loss": 2.8882,
+      "step": 102
+    },
+    {
+      "cosine_similarity": 0.11359092846724232,
+      "epoch": 0.09599254426840634,
+      "grad_norm": 2.960950385643161,
+      "learning_rate": 6.397515527950311e-06,
+      "loss": 2.9331,
+      "step": 103
+    },
+    {
+      "cosine_similarity": 0.03869583285612722,
+      "epoch": 0.09692451071761417,
+      "grad_norm": 2.1952847589503652,
+      "learning_rate": 6.459627329192547e-06,
+      "loss": 2.7317,
+      "step": 104
+    },
+    {
+      "cosine_similarity": 0.08285973218841164,
+      "epoch": 0.097856477166822,
+      "grad_norm": 3.078025265359008,
+      "learning_rate": 6.521739130434783e-06,
+      "loss": 3.0993,
+      "step": 105
+    },
+    {
+      "cosine_similarity": 0.32440586210515576,
+      "epoch": 0.09878844361602983,
+      "grad_norm": 2.5262742652220376,
+      "learning_rate": 6.58385093167702e-06,
+      "loss": 3.4761,
+      "step": 106
+    },
+    {
+      "cosine_similarity": 0.26636346435421665,
+      "epoch": 0.09972041006523766,
+      "grad_norm": 2.6136688539723263,
+      "learning_rate": 6.6459627329192555e-06,
+      "loss": 2.7231,
+      "step": 107
+    },
+    {
+      "cosine_similarity": -0.007477734832091979,
+      "epoch": 0.10065237651444547,
+      "grad_norm": 3.440305021081347,
+      "learning_rate": 6.7080745341614914e-06,
+      "loss": 2.9946,
+      "step": 108
+    },
+    {
+      "cosine_similarity": -0.06720966546137398,
+      "epoch": 0.1015843429636533,
+      "grad_norm": 2.1198318784940846,
+      "learning_rate": 6.7701863354037265e-06,
+      "loss": 2.5291,
+      "step": 109
+    },
+    {
+      "cosine_similarity": 0.008711696210686713,
+      "epoch": 0.10251630941286113,
+      "grad_norm": 2.5856692006608766,
+      "learning_rate": 6.832298136645963e-06,
+      "loss": 2.6883,
+      "step": 110
+    },
+    {
+      "cosine_similarity": -0.291879422913748,
+      "epoch": 0.10344827586206896,
+      "grad_norm": 2.653244852451326,
+      "learning_rate": 6.894409937888199e-06,
+      "loss": 2.9116,
+      "step": 111
+    },
+    {
+      "cosine_similarity": 0.011479439503935276,
+      "epoch": 0.1043802423112768,
+      "grad_norm": 2.662120308419673,
+      "learning_rate": 6.956521739130435e-06,
+      "loss": 2.8957,
+      "step": 112
+    },
+    {
+      "cosine_similarity": 0.03543276944342394,
+      "epoch": 0.10531220876048462,
+      "grad_norm": 3.3842285125616587,
+      "learning_rate": 7.018633540372671e-06,
+      "loss": 3.3667,
+      "step": 113
+    },
+    {
+      "cosine_similarity": 0.030381151767984928,
+      "epoch": 0.10624417520969245,
+      "grad_norm": 2.7123135762630515,
+      "learning_rate": 7.080745341614908e-06,
+      "loss": 2.8034,
+      "step": 114
+    },
+    {
+      "cosine_similarity": -0.18067040422213398,
+      "epoch": 0.10717614165890028,
+      "grad_norm": 2.4944016304468,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 3.7675,
+      "step": 115
+    },
+    {
+      "cosine_similarity": 0.0036799698224721416,
+      "epoch": 0.10810810810810811,
+      "grad_norm": 2.729924987870328,
+      "learning_rate": 7.2049689440993795e-06,
+      "loss": 2.7291,
+      "step": 116
+    },
+    {
+      "cosine_similarity": -0.10854812192669502,
+      "epoch": 0.10904007455731593,
+      "grad_norm": 2.3447202198440666,
+      "learning_rate": 7.267080745341616e-06,
+      "loss": 3.1276,
+      "step": 117
+    },
+    {
+      "cosine_similarity": 0.0021824262948271556,
+      "epoch": 0.10997204100652376,
+      "grad_norm": 2.4042209144860602,
+      "learning_rate": 7.329192546583852e-06,
+      "loss": 2.9314,
+      "step": 118
+    },
+    {
+      "cosine_similarity": 0.0806582118792736,
+      "epoch": 0.11090400745573159,
+      "grad_norm": 3.587962421062773,
+      "learning_rate": 7.391304347826087e-06,
+      "loss": 3.0224,
+      "step": 119
+    },
+    {
+      "cosine_similarity": 0.06309933086537466,
+      "epoch": 0.11183597390493942,
+      "grad_norm": 2.0581969626680636,
+      "learning_rate": 7.453416149068323e-06,
+      "loss": 2.6937,
+      "step": 120
+    },
+    {
+      "cosine_similarity": -0.09918415482616469,
+      "epoch": 0.11276794035414725,
+      "grad_norm": 2.369760515586048,
+      "learning_rate": 7.515527950310559e-06,
+      "loss": 3.0983,
+      "step": 121
+    },
+    {
+      "cosine_similarity": 0.08954945355743162,
+      "epoch": 0.11369990680335508,
+      "grad_norm": 4.331523988258371,
+      "learning_rate": 7.577639751552796e-06,
+      "loss": 2.6219,
+      "step": 122
+    },
+    {
+      "cosine_similarity": 0.14091683657832252,
+      "epoch": 0.11463187325256291,
+      "grad_norm": 2.4059614988826135,
+      "learning_rate": 7.639751552795032e-06,
+      "loss": 2.9888,
+      "step": 123
+    },
+    {
+      "cosine_similarity": -0.2149231096435358,
+      "epoch": 0.11556383970177074,
+      "grad_norm": 2.8276043790397587,
+      "learning_rate": 7.701863354037268e-06,
+      "loss": 2.7276,
+      "step": 124
+    },
+    {
+      "cosine_similarity": 0.1824376204259737,
+      "epoch": 0.11649580615097857,
+      "grad_norm": 2.963712562258993,
+      "learning_rate": 7.763975155279503e-06,
+      "loss": 3.0399,
+      "step": 125
+    },
+    {
+      "cosine_similarity": 0.11051339722481493,
+      "epoch": 0.11742777260018639,
+      "grad_norm": 3.1611846477029637,
+      "learning_rate": 7.82608695652174e-06,
+      "loss": 3.1904,
+      "step": 126
+    },
+    {
+      "cosine_similarity": 0.028118418832456183,
+      "epoch": 0.11835973904939422,
+      "grad_norm": 2.734872354057879,
+      "learning_rate": 7.888198757763977e-06,
+      "loss": 2.988,
+      "step": 127
+    },
+    {
+      "cosine_similarity": -0.12282250165331576,
+      "epoch": 0.11929170549860205,
+      "grad_norm": 2.6983168319904234,
+      "learning_rate": 7.950310559006212e-06,
+      "loss": 2.8769,
+      "step": 128
+    },
+    {
+      "cosine_similarity": -0.11162265454284301,
+      "epoch": 0.12022367194780988,
+      "grad_norm": 2.6586873395768302,
+      "learning_rate": 8.012422360248447e-06,
+      "loss": 3.0888,
+      "step": 129
+    },
+    {
+      "cosine_similarity": 0.18898509322818216,
+      "epoch": 0.1211556383970177,
+      "grad_norm": 3.169833091320079,
+      "learning_rate": 8.074534161490684e-06,
+      "loss": 3.2677,
+      "step": 130
+    },
+    {
+      "cosine_similarity": -0.07678078383474732,
+      "epoch": 0.12208760484622554,
+      "grad_norm": 2.8386676190686284,
+      "learning_rate": 8.13664596273292e-06,
+      "loss": 2.9494,
+      "step": 131
+    },
+    {
+      "cosine_similarity": -0.1801307567298009,
+      "epoch": 0.12301957129543337,
+      "grad_norm": 2.3888365995278953,
+      "learning_rate": 8.198757763975156e-06,
+      "loss": 3.0078,
+      "step": 132
+    },
+    {
+      "cosine_similarity": 0.29758999234962086,
+      "epoch": 0.1239515377446412,
+      "grad_norm": 2.0965314883140582,
+      "learning_rate": 8.260869565217392e-06,
+      "loss": 2.639,
+      "step": 133
+    },
+    {
+      "cosine_similarity": -0.0442814248015238,
+      "epoch": 0.12488350419384903,
+      "grad_norm": 3.2082610908542497,
+      "learning_rate": 8.322981366459629e-06,
+      "loss": 3.2078,
+      "step": 134
+    },
+    {
+      "cosine_similarity": 0.31098731290368115,
+      "epoch": 0.12581547064305684,
+      "grad_norm": 3.6659669078679467,
+      "learning_rate": 8.385093167701864e-06,
+      "loss": 3.0556,
+      "step": 135
+    },
+    {
+      "cosine_similarity": 0.020290723216259826,
+      "epoch": 0.1267474370922647,
+      "grad_norm": 2.153756213463185,
+      "learning_rate": 8.4472049689441e-06,
+      "loss": 2.5411,
+      "step": 136
+    },
+    {
+      "cosine_similarity": 0.18305040171094997,
+      "epoch": 0.1276794035414725,
+      "grad_norm": 2.5999956693904718,
+      "learning_rate": 8.509316770186336e-06,
+      "loss": 2.6705,
+      "step": 137
+    },
+    {
+      "cosine_similarity": 0.06995263678900498,
+      "epoch": 0.12861136999068035,
+      "grad_norm": 2.8414522369079345,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 3.1234,
+      "step": 138
+    },
+    {
+      "cosine_similarity": 0.0014792464350454975,
+      "epoch": 0.12954333643988816,
+      "grad_norm": 2.3118940381922215,
+      "learning_rate": 8.633540372670808e-06,
+      "loss": 2.8353,
+      "step": 139
+    },
+    {
+      "cosine_similarity": -0.14836137813961112,
+      "epoch": 0.13047530288909598,
+      "grad_norm": 2.3660401516106906,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 2.8235,
+      "step": 140
+    },
+    {
+      "cosine_similarity": 0.22405302039407188,
+      "epoch": 0.13140726933830382,
+      "grad_norm": 2.1403813544548673,
+      "learning_rate": 8.75776397515528e-06,
+      "loss": 3.0769,
+      "step": 141
+    },
+    {
+      "cosine_similarity": 0.2632948817179481,
+      "epoch": 0.13233923578751164,
+      "grad_norm": 2.5054438943516297,
+      "learning_rate": 8.819875776397516e-06,
+      "loss": 2.8645,
+      "step": 142
+    },
+    {
+      "cosine_similarity": 0.15416115087634974,
+      "epoch": 0.13327120223671948,
+      "grad_norm": 2.9362410496402016,
+      "learning_rate": 8.881987577639753e-06,
+      "loss": 2.7488,
+      "step": 143
+    },
+    {
+      "cosine_similarity": -0.04032506139979276,
+      "epoch": 0.1342031686859273,
+      "grad_norm": 2.9672113674833644,
+      "learning_rate": 8.944099378881988e-06,
+      "loss": 2.2696,
+      "step": 144
+    },
+    {
+      "cosine_similarity": 0.16206528994373745,
+      "epoch": 0.13513513513513514,
+      "grad_norm": 2.179707741019099,
+      "learning_rate": 9.006211180124225e-06,
+      "loss": 2.7081,
+      "step": 145
+    },
+    {
+      "cosine_similarity": 0.07434734829870723,
+      "epoch": 0.13606710158434296,
+      "grad_norm": 1.9398457542323295,
+      "learning_rate": 9.068322981366461e-06,
+      "loss": 3.0446,
+      "step": 146
+    },
+    {
+      "cosine_similarity": 0.3017774153389423,
+      "epoch": 0.1369990680335508,
+      "grad_norm": 2.1313777244104606,
+      "learning_rate": 9.130434782608697e-06,
+      "loss": 2.9255,
+      "step": 147
+    },
+    {
+      "cosine_similarity": 0.3362156207144012,
+      "epoch": 0.13793103448275862,
+      "grad_norm": 2.7074945227675844,
+      "learning_rate": 9.192546583850932e-06,
+      "loss": 3.2607,
+      "step": 148
+    },
+    {
+      "cosine_similarity": 0.1359323515573761,
+      "epoch": 0.13886300093196646,
+      "grad_norm": 2.389777710607772,
+      "learning_rate": 9.254658385093168e-06,
+      "loss": 3.1288,
+      "step": 149
+    },
+    {
+      "cosine_similarity": 0.12290134890612232,
+      "epoch": 0.13979496738117428,
+      "grad_norm": 1.8034042801175294,
+      "learning_rate": 9.316770186335405e-06,
+      "loss": 2.3919,
+      "step": 150
+    },
+    {
+      "cosine_similarity": 0.13332025958049878,
+      "epoch": 0.1407269338303821,
+      "grad_norm": 3.136965395734737,
+      "learning_rate": 9.37888198757764e-06,
+      "loss": 2.7018,
+      "step": 151
+    },
+    {
+      "cosine_similarity": 0.039105024594871994,
+      "epoch": 0.14165890027958994,
+      "grad_norm": 2.1770621106391412,
+      "learning_rate": 9.440993788819877e-06,
+      "loss": 2.7761,
+      "step": 152
+    },
+    {
+      "cosine_similarity": 0.19132732326714216,
+      "epoch": 0.14259086672879775,
+      "grad_norm": 2.7775793687585453,
+      "learning_rate": 9.503105590062112e-06,
+      "loss": 2.7123,
+      "step": 153
+    },
+    {
+      "cosine_similarity": 0.24455928498113183,
+      "epoch": 0.1435228331780056,
+      "grad_norm": 3.055134098446427,
+      "learning_rate": 9.565217391304349e-06,
+      "loss": 2.8785,
+      "step": 154
+    },
+    {
+      "cosine_similarity": -0.14775398395737255,
+      "epoch": 0.14445479962721341,
+      "grad_norm": 4.581176526310496,
+      "learning_rate": 9.627329192546585e-06,
+      "loss": 2.9351,
+      "step": 155
+    },
+    {
+      "cosine_similarity": -0.015343036088559804,
+      "epoch": 0.14538676607642126,
+      "grad_norm": 2.701183798350378,
+      "learning_rate": 9.68944099378882e-06,
+      "loss": 2.8927,
+      "step": 156
+    },
+    {
+      "cosine_similarity": 0.07606577251052844,
+      "epoch": 0.14631873252562907,
+      "grad_norm": 2.5439851810714775,
+      "learning_rate": 9.751552795031056e-06,
+      "loss": 2.6526,
+      "step": 157
+    },
+    {
+      "cosine_similarity": -0.1768876996541186,
+      "epoch": 0.14725069897483692,
+      "grad_norm": 2.1553041453965514,
+      "learning_rate": 9.813664596273292e-06,
+      "loss": 2.5836,
+      "step": 158
+    },
+    {
+      "cosine_similarity": 0.22803528509802187,
+      "epoch": 0.14818266542404473,
+      "grad_norm": 2.712513712586945,
+      "learning_rate": 9.875776397515529e-06,
+      "loss": 2.8446,
+      "step": 159
+    },
+    {
+      "cosine_similarity": 0.012041863655901694,
+      "epoch": 0.14911463187325255,
+      "grad_norm": 2.724172275762252,
+      "learning_rate": 9.937888198757764e-06,
+      "loss": 3.1572,
+      "step": 160
+    },
+    {
+      "cosine_similarity": 0.1853591054623505,
+      "epoch": 0.1500465983224604,
+      "grad_norm": 2.509083519859802,
+      "learning_rate": 1e-05,
+      "loss": 2.8052,
+      "step": 161
+    },
+    {
+      "cosine_similarity": 0.17565755199624122,
+      "epoch": 0.1509785647716682,
+      "grad_norm": 2.270342994472416,
+      "learning_rate": 1.0062111801242236e-05,
+      "loss": 2.815,
+      "step": 162
+    },
+    {
+      "cosine_similarity": -0.21496059209600674,
+      "epoch": 0.15191053122087605,
+      "grad_norm": 3.8740490947574484,
+      "learning_rate": 1.0124223602484473e-05,
+      "loss": 3.0495,
+      "step": 163
+    },
+    {
+      "cosine_similarity": 0.09993191234585677,
+      "epoch": 0.15284249767008387,
+      "grad_norm": 2.483879787335979,
+      "learning_rate": 1.0186335403726708e-05,
+      "loss": 3.0112,
+      "step": 164
+    },
+    {
+      "cosine_similarity": 0.11467387666046279,
+      "epoch": 0.15377446411929171,
+      "grad_norm": 2.140393591920432,
+      "learning_rate": 1.0248447204968946e-05,
+      "loss": 2.5961,
+      "step": 165
+    },
+    {
+      "cosine_similarity": 0.13532590657679036,
+      "epoch": 0.15470643056849953,
+      "grad_norm": 3.0874581005134862,
+      "learning_rate": 1.0310559006211181e-05,
+      "loss": 2.8443,
+      "step": 166
+    },
+    {
+      "cosine_similarity": 0.03257888056196846,
+      "epoch": 0.15563839701770738,
+      "grad_norm": 2.3791438217733027,
+      "learning_rate": 1.0372670807453418e-05,
+      "loss": 2.8704,
+      "step": 167
+    },
+    {
+      "cosine_similarity": -0.061069505690213674,
+      "epoch": 0.1565703634669152,
+      "grad_norm": 2.8544500088446236,
+      "learning_rate": 1.0434782608695653e-05,
+      "loss": 2.7843,
+      "step": 168
+    },
+    {
+      "cosine_similarity": 0.010056129192511018,
+      "epoch": 0.157502329916123,
+      "grad_norm": 2.489264783914348,
+      "learning_rate": 1.049689440993789e-05,
+      "loss": 2.8756,
+      "step": 169
+    },
+    {
+      "cosine_similarity": 0.12160047801918382,
+      "epoch": 0.15843429636533085,
+      "grad_norm": 1.8275171008040243,
+      "learning_rate": 1.0559006211180125e-05,
+      "loss": 2.1851,
+      "step": 170
+    },
+    {
+      "cosine_similarity": 0.1341347142662194,
+      "epoch": 0.15936626281453867,
+      "grad_norm": 2.122331657536347,
+      "learning_rate": 1.062111801242236e-05,
+      "loss": 2.6161,
+      "step": 171
+    },
+    {
+      "cosine_similarity": 0.013479897879095173,
+      "epoch": 0.1602982292637465,
+      "grad_norm": 2.458215496475715,
+      "learning_rate": 1.0683229813664597e-05,
+      "loss": 3.0425,
+      "step": 172
+    },
+    {
+      "cosine_similarity": 0.1353669213721682,
+      "epoch": 0.16123019571295433,
+      "grad_norm": 2.3512048954994307,
+      "learning_rate": 1.0745341614906832e-05,
+      "loss": 2.6497,
+      "step": 173
+    },
+    {
+      "cosine_similarity": -0.02408179377230293,
+      "epoch": 0.16216216216216217,
+      "grad_norm": 2.152482675721413,
+      "learning_rate": 1.080745341614907e-05,
+      "loss": 2.3519,
+      "step": 174
+    },
+    {
+      "cosine_similarity": 0.04890694150640415,
+      "epoch": 0.16309412861137,
+      "grad_norm": 1.9910256291190547,
+      "learning_rate": 1.0869565217391305e-05,
+      "loss": 2.6838,
+      "step": 175
+    },
+    {
+      "cosine_similarity": 0.22387222550753502,
+      "epoch": 0.16402609506057783,
+      "grad_norm": 2.80228567418662,
+      "learning_rate": 1.0931677018633542e-05,
+      "loss": 2.9426,
+      "step": 176
+    },
+    {
+      "cosine_similarity": 0.036687169818855676,
+      "epoch": 0.16495806150978565,
+      "grad_norm": 2.0549149693295776,
+      "learning_rate": 1.0993788819875777e-05,
+      "loss": 3.0609,
+      "step": 177
+    },
+    {
+      "cosine_similarity": 0.03536997793019986,
+      "epoch": 0.16589002795899346,
+      "grad_norm": 2.724943820333908,
+      "learning_rate": 1.1055900621118014e-05,
+      "loss": 2.6444,
+      "step": 178
+    },
+    {
+      "cosine_similarity": 0.0023947641860323572,
+      "epoch": 0.1668219944082013,
+      "grad_norm": 1.8758740578222666,
+      "learning_rate": 1.1118012422360249e-05,
+      "loss": 2.7462,
+      "step": 179
+    },
+    {
+      "cosine_similarity": -0.10369119016807216,
+      "epoch": 0.16775396085740912,
+      "grad_norm": 1.9748959679205955,
+      "learning_rate": 1.1180124223602484e-05,
+      "loss": 2.4115,
+      "step": 180
+    },
+    {
+      "cosine_similarity": 0.028563939582027523,
+      "epoch": 0.16868592730661697,
+      "grad_norm": 2.336912598799085,
+      "learning_rate": 1.1242236024844722e-05,
+      "loss": 2.927,
+      "step": 181
+    },
+    {
+      "cosine_similarity": 0.13696798484803271,
+      "epoch": 0.16961789375582478,
+      "grad_norm": 2.3530752962135164,
+      "learning_rate": 1.1304347826086957e-05,
+      "loss": 2.8188,
+      "step": 182
+    },
+    {
+      "cosine_similarity": 0.4404692578545227,
+      "epoch": 0.17054986020503263,
+      "grad_norm": 2.685863383620477,
+      "learning_rate": 1.1366459627329194e-05,
+      "loss": 2.8385,
+      "step": 183
+    },
+    {
+      "cosine_similarity": 0.06296351445447497,
+      "epoch": 0.17148182665424044,
+      "grad_norm": 2.8302560839042665,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 2.6896,
+      "step": 184
+    },
+    {
+      "cosine_similarity": -0.08619020806947263,
+      "epoch": 0.1724137931034483,
+      "grad_norm": 2.2251874490640433,
+      "learning_rate": 1.1490683229813666e-05,
+      "loss": 3.014,
+      "step": 185
+    },
+    {
+      "cosine_similarity": -0.12486713915689324,
+      "epoch": 0.1733457595526561,
+      "grad_norm": 2.716715931864157,
+      "learning_rate": 1.15527950310559e-05,
+      "loss": 2.182,
+      "step": 186
+    },
+    {
+      "cosine_similarity": 0.2822339650893711,
+      "epoch": 0.17427772600186392,
+      "grad_norm": 2.280852053603259,
+      "learning_rate": 1.161490683229814e-05,
+      "loss": 2.7578,
+      "step": 187
+    },
+    {
+      "cosine_similarity": 0.14469434648831617,
+      "epoch": 0.17520969245107176,
+      "grad_norm": 1.9615797369352297,
+      "learning_rate": 1.1677018633540373e-05,
+      "loss": 2.7254,
+      "step": 188
+    },
+    {
+      "cosine_similarity": 0.20716612603669346,
+      "epoch": 0.17614165890027958,
+      "grad_norm": 1.8258168969463555,
+      "learning_rate": 1.1739130434782611e-05,
+      "loss": 2.5094,
+      "step": 189
+    },
+    {
+      "cosine_similarity": 0.08601419715549019,
+      "epoch": 0.17707362534948742,
+      "grad_norm": 1.7510106774685688,
+      "learning_rate": 1.1801242236024846e-05,
+      "loss": 2.4882,
+      "step": 190
+    },
+    {
+      "cosine_similarity": 0.19385520781762347,
+      "epoch": 0.17800559179869524,
+      "grad_norm": 2.0170855296601156,
+      "learning_rate": 1.1863354037267081e-05,
+      "loss": 2.9033,
+      "step": 191
+    },
+    {
+      "cosine_similarity": 0.11654424360726759,
+      "epoch": 0.17893755824790308,
+      "grad_norm": 1.7745537085715386,
+      "learning_rate": 1.1925465838509318e-05,
+      "loss": 2.4641,
+      "step": 192
+    },
+    {
+      "cosine_similarity": 0.2128066464594931,
+      "epoch": 0.1798695246971109,
+      "grad_norm": 2.3442616542490367,
+      "learning_rate": 1.1987577639751553e-05,
+      "loss": 2.5621,
+      "step": 193
+    },
+    {
+      "cosine_similarity": 0.03967416544500937,
+      "epoch": 0.18080149114631874,
+      "grad_norm": 2.2794807522428004,
+      "learning_rate": 1.204968944099379e-05,
+      "loss": 2.598,
+      "step": 194
+    },
+    {
+      "cosine_similarity": 0.038243980349782776,
+      "epoch": 0.18173345759552656,
+      "grad_norm": 2.449840456512398,
+      "learning_rate": 1.2111801242236025e-05,
+      "loss": 2.9628,
+      "step": 195
+    },
+    {
+      "cosine_similarity": -0.16791453185324634,
+      "epoch": 0.18266542404473438,
+      "grad_norm": 2.070005192248288,
+      "learning_rate": 1.2173913043478263e-05,
+      "loss": 2.5119,
+      "step": 196
+    },
+    {
+      "cosine_similarity": 0.38663443152877,
+      "epoch": 0.18359739049394222,
+      "grad_norm": 2.111108012186067,
+      "learning_rate": 1.2236024844720498e-05,
+      "loss": 2.5142,
+      "step": 197
+    },
+    {
+      "cosine_similarity": 0.6399396684429696,
+      "epoch": 0.18452935694315004,
+      "grad_norm": 2.199376278977748,
+      "learning_rate": 1.2298136645962735e-05,
+      "loss": 2.8899,
+      "step": 198
+    },
+    {
+      "cosine_similarity": 0.088362105729341,
+      "epoch": 0.18546132339235788,
+      "grad_norm": 2.4828595343074946,
+      "learning_rate": 1.236024844720497e-05,
+      "loss": 2.8001,
+      "step": 199
+    },
+    {
+      "cosine_similarity": 0.31488344723997136,
+      "epoch": 0.1863932898415657,
+      "grad_norm": 2.6129176531698928,
+      "learning_rate": 1.2422360248447205e-05,
+      "loss": 3.0784,
+      "step": 200
+    },
+    {
+      "cosine_similarity": 0.0012520795904411865,
+      "epoch": 0.18732525629077354,
+      "grad_norm": 2.0706695249403166,
+      "learning_rate": 1.2484472049689442e-05,
+      "loss": 2.747,
+      "step": 201
+    },
+    {
+      "cosine_similarity": 0.037913399741076846,
+      "epoch": 0.18825722273998136,
+      "grad_norm": 1.8731949105463293,
+      "learning_rate": 1.2546583850931677e-05,
+      "loss": 2.5872,
+      "step": 202
+    },
+    {
+      "cosine_similarity": 0.14664540862724212,
+      "epoch": 0.1891891891891892,
+      "grad_norm": 2.5054797260912505,
+      "learning_rate": 1.2608695652173915e-05,
+      "loss": 2.7675,
+      "step": 203
+    },
+    {
+      "cosine_similarity": 0.1498963370459379,
+      "epoch": 0.19012115563839702,
+      "grad_norm": 2.5780680213277876,
+      "learning_rate": 1.267080745341615e-05,
+      "loss": 2.9624,
+      "step": 204
+    },
+    {
+      "cosine_similarity": 0.17411695891507986,
+      "epoch": 0.19105312208760486,
+      "grad_norm": 2.559312674948975,
+      "learning_rate": 1.2732919254658387e-05,
+      "loss": 2.6223,
+      "step": 205
+    },
+    {
+      "cosine_similarity": 0.12071525876545683,
+      "epoch": 0.19198508853681268,
+      "grad_norm": 2.202494772991211,
+      "learning_rate": 1.2795031055900622e-05,
+      "loss": 2.8792,
+      "step": 206
+    },
+    {
+      "cosine_similarity": 0.22532877452473185,
+      "epoch": 0.1929170549860205,
+      "grad_norm": 2.6284701932741896,
+      "learning_rate": 1.2857142857142859e-05,
+      "loss": 2.7574,
+      "step": 207
+    },
+    {
+      "cosine_similarity": 0.02430991667340344,
+      "epoch": 0.19384902143522834,
+      "grad_norm": 2.062718346160552,
+      "learning_rate": 1.2919254658385094e-05,
+      "loss": 2.8885,
+      "step": 208
+    },
+    {
+      "cosine_similarity": -0.17404002150896736,
+      "epoch": 0.19478098788443615,
+      "grad_norm": 2.0334880440096614,
+      "learning_rate": 1.2981366459627329e-05,
+      "loss": 1.9556,
+      "step": 209
+    },
+    {
+      "cosine_similarity": 0.015181321543412646,
+      "epoch": 0.195712954333644,
+      "grad_norm": 2.335952962310549,
+      "learning_rate": 1.3043478260869566e-05,
+      "loss": 2.6842,
+      "step": 210
+    },
+    {
+      "cosine_similarity": 0.21867115389739197,
+      "epoch": 0.1966449207828518,
+      "grad_norm": 1.5940025506558826,
+      "learning_rate": 1.31055900621118e-05,
+      "loss": 2.2245,
+      "step": 211
+    },
+    {
+      "cosine_similarity": 0.051446786675514014,
+      "epoch": 0.19757688723205966,
+      "grad_norm": 1.9749046091291527,
+      "learning_rate": 1.316770186335404e-05,
+      "loss": 2.7446,
+      "step": 212
+    },
+    {
+      "cosine_similarity": 0.018613770323745116,
+      "epoch": 0.19850885368126747,
+      "grad_norm": 2.436950556336193,
+      "learning_rate": 1.3229813664596274e-05,
+      "loss": 2.5934,
+      "step": 213
+    },
+    {
+      "cosine_similarity": -0.04067587273837224,
+      "epoch": 0.19944082013047532,
+      "grad_norm": 2.9673942202175807,
+      "learning_rate": 1.3291925465838511e-05,
+      "loss": 3.1013,
+      "step": 214
+    },
+    {
+      "cosine_similarity": 0.03975035869463719,
+      "epoch": 0.20037278657968313,
+      "grad_norm": 2.1093387481719903,
+      "learning_rate": 1.3354037267080746e-05,
+      "loss": 2.3759,
+      "step": 215
+    },
+    {
+      "cosine_similarity": 0.14691179554807585,
+      "epoch": 0.20130475302889095,
+      "grad_norm": 2.2697280450387596,
+      "learning_rate": 1.3416149068322983e-05,
+      "loss": 2.4131,
+      "step": 216
+    },
+    {
+      "cosine_similarity": 0.1074552021965198,
+      "epoch": 0.2022367194780988,
+      "grad_norm": 2.3191647534251434,
+      "learning_rate": 1.3478260869565218e-05,
+      "loss": 2.6252,
+      "step": 217
+    },
+    {
+      "cosine_similarity": 0.21156684752211738,
+      "epoch": 0.2031686859273066,
+      "grad_norm": 1.5255014285727475,
+      "learning_rate": 1.3540372670807453e-05,
+      "loss": 2.5975,
+      "step": 218
+    },
+    {
+      "cosine_similarity": 0.49937794158611915,
+      "epoch": 0.20410065237651445,
+      "grad_norm": 2.020111168004662,
+      "learning_rate": 1.3602484472049691e-05,
+      "loss": 2.7941,
+      "step": 219
+    },
+    {
+      "cosine_similarity": 0.2071877378182251,
+      "epoch": 0.20503261882572227,
+      "grad_norm": 2.823028751092971,
+      "learning_rate": 1.3664596273291926e-05,
+      "loss": 2.7259,
+      "step": 220
+    },
+    {
+      "cosine_similarity": 0.2779968408062057,
+      "epoch": 0.2059645852749301,
+      "grad_norm": 2.0304479716392554,
+      "learning_rate": 1.3726708074534163e-05,
+      "loss": 2.7469,
+      "step": 221
+    },
+    {
+      "cosine_similarity": 0.5941860586326967,
+      "epoch": 0.20689655172413793,
+      "grad_norm": 2.116037990695354,
+      "learning_rate": 1.3788819875776398e-05,
+      "loss": 2.6503,
+      "step": 222
+    },
+    {
+      "cosine_similarity": 0.25841198399149756,
+      "epoch": 0.20782851817334577,
+      "grad_norm": 2.273776729631011,
+      "learning_rate": 1.3850931677018635e-05,
+      "loss": 2.7533,
+      "step": 223
+    },
+    {
+      "cosine_similarity": 0.20982608770011313,
+      "epoch": 0.2087604846225536,
+      "grad_norm": 2.390319456477384,
+      "learning_rate": 1.391304347826087e-05,
+      "loss": 2.8115,
+      "step": 224
+    },
+    {
+      "cosine_similarity": 0.2145804243483251,
+      "epoch": 0.2096924510717614,
+      "grad_norm": 3.3733200213557715,
+      "learning_rate": 1.3975155279503107e-05,
+      "loss": 2.8709,
+      "step": 225
+    },
+    {
+      "cosine_similarity": 0.24600629161068266,
+      "epoch": 0.21062441752096925,
+      "grad_norm": 2.696125813736635,
+      "learning_rate": 1.4037267080745342e-05,
+      "loss": 2.72,
+      "step": 226
+    },
+    {
+      "cosine_similarity": 0.28551383531850166,
+      "epoch": 0.21155638397017706,
+      "grad_norm": 2.6621902040772354,
+      "learning_rate": 1.409937888198758e-05,
+      "loss": 2.6979,
+      "step": 227
+    },
+    {
+      "cosine_similarity": -0.08323564701941198,
+      "epoch": 0.2124883504193849,
+      "grad_norm": 2.2099457665492483,
+      "learning_rate": 1.4161490683229815e-05,
+      "loss": 2.8068,
+      "step": 228
+    },
+    {
+      "cosine_similarity": 0.3890730056485353,
+      "epoch": 0.21342031686859272,
+      "grad_norm": 2.3533898043319903,
+      "learning_rate": 1.422360248447205e-05,
+      "loss": 2.9563,
+      "step": 229
+    },
+    {
+      "cosine_similarity": 0.40501076921113527,
+      "epoch": 0.21435228331780057,
+      "grad_norm": 1.8861851289904854,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 2.7762,
+      "step": 230
+    },
+    {
+      "cosine_similarity": -0.11319918518459651,
+      "epoch": 0.21528424976700838,
+      "grad_norm": 2.09327191898752,
+      "learning_rate": 1.4347826086956522e-05,
+      "loss": 2.2258,
+      "step": 231
+    },
+    {
+      "cosine_similarity": 0.08671217733348321,
+      "epoch": 0.21621621621621623,
+      "grad_norm": 2.656457552356113,
+      "learning_rate": 1.4409937888198759e-05,
+      "loss": 2.5548,
+      "step": 232
+    },
+    {
+      "cosine_similarity": 0.08540243515502648,
+      "epoch": 0.21714818266542404,
+      "grad_norm": 2.011571227324112,
+      "learning_rate": 1.4472049689440994e-05,
+      "loss": 3.0256,
+      "step": 233
+    },
+    {
+      "cosine_similarity": -0.03849159771093411,
+      "epoch": 0.21808014911463186,
+      "grad_norm": 2.0493211576071615,
+      "learning_rate": 1.4534161490683232e-05,
+      "loss": 2.4543,
+      "step": 234
+    },
+    {
+      "cosine_similarity": 0.22544276971722893,
+      "epoch": 0.2190121155638397,
+      "grad_norm": 2.0381515746590324,
+      "learning_rate": 1.4596273291925467e-05,
+      "loss": 2.2933,
+      "step": 235
+    },
+    {
+      "cosine_similarity": 0.09040047386834384,
+      "epoch": 0.21994408201304752,
+      "grad_norm": 1.7860262236134328,
+      "learning_rate": 1.4658385093167704e-05,
+      "loss": 2.1988,
+      "step": 236
+    },
+    {
+      "cosine_similarity": 0.031180938501780595,
+      "epoch": 0.22087604846225536,
+      "grad_norm": 2.28137581241705,
+      "learning_rate": 1.472049689440994e-05,
+      "loss": 2.3206,
+      "step": 237
+    },
+    {
+      "cosine_similarity": 0.15278021105203837,
+      "epoch": 0.22180801491146318,
+      "grad_norm": 2.1587424577506913,
+      "learning_rate": 1.4782608695652174e-05,
+      "loss": 2.3977,
+      "step": 238
+    },
+    {
+      "cosine_similarity": 0.42392370211874025,
+      "epoch": 0.22273998136067102,
+      "grad_norm": 2.2779744676471285,
+      "learning_rate": 1.4844720496894411e-05,
+      "loss": 2.449,
+      "step": 239
+    },
+    {
+      "cosine_similarity": 0.3392490407964442,
+      "epoch": 0.22367194780987884,
+      "grad_norm": 2.33021711295839,
+      "learning_rate": 1.4906832298136646e-05,
+      "loss": 2.7955,
+      "step": 240
+    },
+    {
+      "cosine_similarity": -0.22844515514212568,
+      "epoch": 0.22460391425908668,
+      "grad_norm": 2.4630043898862386,
+      "learning_rate": 1.4968944099378885e-05,
+      "loss": 2.8513,
+      "step": 241
+    },
+    {
+      "cosine_similarity": 0.016818424574561414,
+      "epoch": 0.2255358807082945,
+      "grad_norm": 1.9043142043674224,
+      "learning_rate": 1.5031055900621118e-05,
+      "loss": 2.7238,
+      "step": 242
+    },
+    {
+      "cosine_similarity": -0.09051335689740284,
+      "epoch": 0.22646784715750232,
+      "grad_norm": 2.3642859969960552,
+      "learning_rate": 1.5093167701863356e-05,
+      "loss": 2.5866,
+      "step": 243
+    },
+    {
+      "cosine_similarity": 0.3825300988709888,
+      "epoch": 0.22739981360671016,
+      "grad_norm": 1.996326640030743,
+      "learning_rate": 1.5155279503105591e-05,
+      "loss": 2.6992,
+      "step": 244
+    },
+    {
+      "cosine_similarity": 0.07632216736108406,
+      "epoch": 0.22833178005591798,
+      "grad_norm": 2.382509256189736,
+      "learning_rate": 1.5217391304347828e-05,
+      "loss": 2.7696,
+      "step": 245
+    },
+    {
+      "cosine_similarity": -0.1981477755955585,
+      "epoch": 0.22926374650512582,
+      "grad_norm": 1.6501186080097598,
+      "learning_rate": 1.5279503105590063e-05,
+      "loss": 2.8425,
+      "step": 246
+    },
+    {
+      "cosine_similarity": 0.6134837459005472,
+      "epoch": 0.23019571295433364,
+      "grad_norm": 3.152668414183632,
+      "learning_rate": 1.5341614906832298e-05,
+      "loss": 2.9685,
+      "step": 247
+    },
+    {
+      "cosine_similarity": -0.01351020213123913,
+      "epoch": 0.23112767940354148,
+      "grad_norm": 2.1470955093726363,
+      "learning_rate": 1.5403726708074537e-05,
+      "loss": 2.5723,
+      "step": 248
+    },
+    {
+      "cosine_similarity": 0.09194007449851382,
+      "epoch": 0.2320596458527493,
+      "grad_norm": 1.7174075526109014,
+      "learning_rate": 1.5465838509316772e-05,
+      "loss": 2.5353,
+      "step": 249
+    },
+    {
+      "cosine_similarity": 0.29599865262609537,
+      "epoch": 0.23299161230195714,
+      "grad_norm": 1.9203124581929933,
+      "learning_rate": 1.5527950310559007e-05,
+      "loss": 2.418,
+      "step": 250
+    },
+    {
+      "cosine_similarity": 0.17348183367287143,
+      "epoch": 0.23392357875116496,
+      "grad_norm": 1.9470406230123074,
+      "learning_rate": 1.5590062111801242e-05,
+      "loss": 2.7415,
+      "step": 251
+    },
+    {
+      "cosine_similarity": 0.169113443821796,
+      "epoch": 0.23485554520037277,
+      "grad_norm": 1.8419840585875962,
+      "learning_rate": 1.565217391304348e-05,
+      "loss": 2.6448,
+      "step": 252
+    },
+    {
+      "cosine_similarity": 0.2337204495654513,
+      "epoch": 0.23578751164958062,
+      "grad_norm": 2.3594943534029524,
+      "learning_rate": 1.5714285714285715e-05,
+      "loss": 2.6217,
+      "step": 253
+    },
+    {
+      "cosine_similarity": 0.06315499042526225,
+      "epoch": 0.23671947809878843,
+      "grad_norm": 2.031913468164309,
+      "learning_rate": 1.5776397515527954e-05,
+      "loss": 2.7201,
+      "step": 254
+    },
+    {
+      "cosine_similarity": 0.027601404703635757,
+      "epoch": 0.23765144454799628,
+      "grad_norm": 1.738898101207263,
+      "learning_rate": 1.583850931677019e-05,
+      "loss": 2.3544,
+      "step": 255
+    },
+    {
+      "cosine_similarity": 0.0109328266243132,
+      "epoch": 0.2385834109972041,
+      "grad_norm": 2.1344829402081995,
+      "learning_rate": 1.5900621118012424e-05,
+      "loss": 2.1579,
+      "step": 256
+    },
+    {
+      "cosine_similarity": 0.084127370013018,
+      "epoch": 0.23951537744641194,
+      "grad_norm": 2.4916879666208245,
+      "learning_rate": 1.596273291925466e-05,
+      "loss": 2.8568,
+      "step": 257
+    },
+    {
+      "cosine_similarity": 0.21105779773407396,
+      "epoch": 0.24044734389561975,
+      "grad_norm": 1.9059781749319604,
+      "learning_rate": 1.6024844720496894e-05,
+      "loss": 2.6175,
+      "step": 258
+    },
+    {
+      "cosine_similarity": 0.04536774586563962,
+      "epoch": 0.2413793103448276,
+      "grad_norm": 1.627657695934529,
+      "learning_rate": 1.6086956521739132e-05,
+      "loss": 2.5245,
+      "step": 259
+    },
+    {
+      "cosine_similarity": 0.01953979623763381,
+      "epoch": 0.2423112767940354,
+      "grad_norm": 1.9129882445772612,
+      "learning_rate": 1.6149068322981367e-05,
+      "loss": 2.5974,
+      "step": 260
+    },
+    {
+      "cosine_similarity": 0.001947343959375657,
+      "epoch": 0.24324324324324326,
+      "grad_norm": 2.4746766923971566,
+      "learning_rate": 1.6211180124223606e-05,
+      "loss": 2.5244,
+      "step": 261
+    },
+    {
+      "cosine_similarity": 0.13466792863019336,
+      "epoch": 0.24417520969245107,
+      "grad_norm": 2.0171004299573165,
+      "learning_rate": 1.627329192546584e-05,
+      "loss": 2.3888,
+      "step": 262
+    },
+    {
+      "cosine_similarity": 0.10065912934223041,
+      "epoch": 0.2451071761416589,
+      "grad_norm": 1.9860427403611296,
+      "learning_rate": 1.6335403726708076e-05,
+      "loss": 2.5442,
+      "step": 263
+    },
+    {
+      "cosine_similarity": 0.015214211201356048,
+      "epoch": 0.24603914259086673,
+      "grad_norm": 2.11244793570608,
+      "learning_rate": 1.639751552795031e-05,
+      "loss": 2.577,
+      "step": 264
+    },
+    {
+      "cosine_similarity": 0.3522623525946097,
+      "epoch": 0.24697110904007455,
+      "grad_norm": 2.394599956081536,
+      "learning_rate": 1.645962732919255e-05,
+      "loss": 2.5416,
+      "step": 265
+    },
+    {
+      "cosine_similarity": 0.20137661482280134,
+      "epoch": 0.2479030754892824,
+      "grad_norm": 1.922797009059207,
+      "learning_rate": 1.6521739130434785e-05,
+      "loss": 2.4818,
+      "step": 266
+    },
+    {
+      "cosine_similarity": 0.3221207615807883,
+      "epoch": 0.2488350419384902,
+      "grad_norm": 1.7190205737426048,
+      "learning_rate": 1.658385093167702e-05,
+      "loss": 2.8559,
+      "step": 267
+    },
+    {
+      "cosine_similarity": 0.04890367139366374,
+      "epoch": 0.24976700838769805,
+      "grad_norm": 1.3822262131105125,
+      "learning_rate": 1.6645962732919258e-05,
+      "loss": 2.0072,
+      "step": 268
+    },
+    {
+      "cosine_similarity": 0.1372221898019281,
+      "epoch": 0.2506989748369059,
+      "grad_norm": 2.21447155154385,
+      "learning_rate": 1.670807453416149e-05,
+      "loss": 2.7052,
+      "step": 269
+    },
+    {
+      "cosine_similarity": 0.4239350049122637,
+      "epoch": 0.2516309412861137,
+      "grad_norm": 2.9046241638738226,
+      "learning_rate": 1.6770186335403728e-05,
+      "loss": 2.613,
+      "step": 270
+    },
+    {
+      "cosine_similarity": 0.03270492450357759,
+      "epoch": 0.25256290773532153,
+      "grad_norm": 2.0301808536151134,
+      "learning_rate": 1.6832298136645963e-05,
+      "loss": 2.4054,
+      "step": 271
+    },
+    {
+      "cosine_similarity": 0.41455504748465233,
+      "epoch": 0.2534948741845294,
+      "grad_norm": 2.1517200943406296,
+      "learning_rate": 1.68944099378882e-05,
+      "loss": 2.6595,
+      "step": 272
+    },
+    {
+      "cosine_similarity": -0.02047411548800589,
+      "epoch": 0.25442684063373716,
+      "grad_norm": 1.973528742376363,
+      "learning_rate": 1.6956521739130437e-05,
+      "loss": 2.1408,
+      "step": 273
+    },
+    {
+      "cosine_similarity": 0.042610120239126886,
+      "epoch": 0.255358807082945,
+      "grad_norm": 1.7927961888029353,
+      "learning_rate": 1.7018633540372672e-05,
+      "loss": 2.6807,
+      "step": 274
+    },
+    {
+      "cosine_similarity": 0.35251235759813426,
+      "epoch": 0.25629077353215285,
+      "grad_norm": 1.7610498937881829,
+      "learning_rate": 1.7080745341614907e-05,
+      "loss": 2.8623,
+      "step": 275
+    },
+    {
+      "cosine_similarity": 0.19072247886690888,
+      "epoch": 0.2572227399813607,
+      "grad_norm": 2.1050587574514332,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 2.6658,
+      "step": 276
+    },
+    {
+      "cosine_similarity": 0.282453522463849,
+      "epoch": 0.2581547064305685,
+      "grad_norm": 2.7147250314049547,
+      "learning_rate": 1.720496894409938e-05,
+      "loss": 2.2757,
+      "step": 277
+    },
+    {
+      "cosine_similarity": 0.4000379850641914,
+      "epoch": 0.2590866728797763,
+      "grad_norm": 1.8804233179242107,
+      "learning_rate": 1.7267080745341615e-05,
+      "loss": 2.5789,
+      "step": 278
+    },
+    {
+      "cosine_similarity": 0.14153282591216737,
+      "epoch": 0.26001863932898417,
+      "grad_norm": 2.0422956879183944,
+      "learning_rate": 1.7329192546583854e-05,
+      "loss": 2.2138,
+      "step": 279
+    },
+    {
+      "cosine_similarity": 0.17811668372057307,
+      "epoch": 0.26095060577819196,
+      "grad_norm": 2.0596233632274155,
+      "learning_rate": 1.739130434782609e-05,
+      "loss": 2.8271,
+      "step": 280
+    },
+    {
+      "cosine_similarity": 0.8688052668859707,
+      "epoch": 0.2618825722273998,
+      "grad_norm": 1.96474726527618,
+      "learning_rate": 1.7453416149068324e-05,
+      "loss": 2.6715,
+      "step": 281
+    },
+    {
+      "cosine_similarity": 0.22747519017671683,
+      "epoch": 0.26281453867660765,
+      "grad_norm": 1.6541364017900806,
+      "learning_rate": 1.751552795031056e-05,
+      "loss": 2.0105,
+      "step": 282
+    },
+    {
+      "cosine_similarity": 0.04492486273303857,
+      "epoch": 0.2637465051258155,
+      "grad_norm": 2.0785141182426328,
+      "learning_rate": 1.7577639751552797e-05,
+      "loss": 2.3451,
+      "step": 283
+    },
+    {
+      "cosine_similarity": -0.024259653124682683,
+      "epoch": 0.2646784715750233,
+      "grad_norm": 1.8830982420214546,
+      "learning_rate": 1.7639751552795032e-05,
+      "loss": 2.1352,
+      "step": 284
+    },
+    {
+      "cosine_similarity": 0.39020347735281724,
+      "epoch": 0.2656104380242311,
+      "grad_norm": 1.785639749347082,
+      "learning_rate": 1.7701863354037267e-05,
+      "loss": 2.7539,
+      "step": 285
+    },
+    {
+      "cosine_similarity": 0.1497875728513161,
+      "epoch": 0.26654240447343897,
+      "grad_norm": 1.9313989537832612,
+      "learning_rate": 1.7763975155279506e-05,
+      "loss": 2.6341,
+      "step": 286
+    },
+    {
+      "cosine_similarity": -0.25174185664324567,
+      "epoch": 0.2674743709226468,
+      "grad_norm": 2.076987158449556,
+      "learning_rate": 1.782608695652174e-05,
+      "loss": 2.8272,
+      "step": 287
+    },
+    {
+      "cosine_similarity": 0.3891480342107464,
+      "epoch": 0.2684063373718546,
+      "grad_norm": 1.9770227357657058,
+      "learning_rate": 1.7888198757763976e-05,
+      "loss": 2.7009,
+      "step": 288
+    },
+    {
+      "cosine_similarity": 0.24184230588363703,
+      "epoch": 0.26933830382106244,
+      "grad_norm": 2.001602409773861,
+      "learning_rate": 1.795031055900621e-05,
+      "loss": 2.9797,
+      "step": 289
+    },
+    {
+      "cosine_similarity": 0.3277327222369322,
+      "epoch": 0.2702702702702703,
+      "grad_norm": 1.831316600481878,
+      "learning_rate": 1.801242236024845e-05,
+      "loss": 2.5377,
+      "step": 290
+    },
+    {
+      "cosine_similarity": 0.036599841745852006,
+      "epoch": 0.2712022367194781,
+      "grad_norm": 2.312997234211653,
+      "learning_rate": 1.8074534161490685e-05,
+      "loss": 1.9708,
+      "step": 291
+    },
+    {
+      "cosine_similarity": 0.141478094074968,
+      "epoch": 0.2721342031686859,
+      "grad_norm": 1.8919801794156792,
+      "learning_rate": 1.8136645962732923e-05,
+      "loss": 2.3505,
+      "step": 292
+    },
+    {
+      "cosine_similarity": 0.6108015094287753,
+      "epoch": 0.27306616961789376,
+      "grad_norm": 1.7564756062085785,
+      "learning_rate": 1.8198757763975158e-05,
+      "loss": 2.7965,
+      "step": 293
+    },
+    {
+      "cosine_similarity": 0.1826706534095156,
+      "epoch": 0.2739981360671016,
+      "grad_norm": 1.8302241887267887,
+      "learning_rate": 1.8260869565217393e-05,
+      "loss": 2.6505,
+      "step": 294
+    },
+    {
+      "cosine_similarity": 0.21112402783605042,
+      "epoch": 0.2749301025163094,
+      "grad_norm": 2.0406829102965753,
+      "learning_rate": 1.8322981366459628e-05,
+      "loss": 2.757,
+      "step": 295
+    },
+    {
+      "cosine_similarity": 0.10827938437901753,
+      "epoch": 0.27586206896551724,
+      "grad_norm": 1.8260796044663556,
+      "learning_rate": 1.8385093167701863e-05,
+      "loss": 2.338,
+      "step": 296
+    },
+    {
+      "cosine_similarity": 0.37640013270223766,
+      "epoch": 0.2767940354147251,
+      "grad_norm": 1.6780080495896175,
+      "learning_rate": 1.84472049689441e-05,
+      "loss": 2.2468,
+      "step": 297
+    },
+    {
+      "cosine_similarity": 0.08174150428001573,
+      "epoch": 0.2777260018639329,
+      "grad_norm": 1.958745575145591,
+      "learning_rate": 1.8509316770186337e-05,
+      "loss": 2.6921,
+      "step": 298
+    },
+    {
+      "cosine_similarity": -0.033161509249935696,
+      "epoch": 0.2786579683131407,
+      "grad_norm": 2.1892536506662377,
+      "learning_rate": 1.8571428571428575e-05,
+      "loss": 2.5336,
+      "step": 299
+    },
+    {
+      "cosine_similarity": 0.03472493806291892,
+      "epoch": 0.27958993476234856,
+      "grad_norm": 2.0574362760942013,
+      "learning_rate": 1.863354037267081e-05,
+      "loss": 2.7969,
+      "step": 300
+    },
+    {
+      "cosine_similarity": 0.03471615274100366,
+      "epoch": 0.2805219012115564,
+      "grad_norm": 1.6548474684408008,
+      "learning_rate": 1.8695652173913045e-05,
+      "loss": 1.8272,
+      "step": 301
+    },
+    {
+      "cosine_similarity": 0.09884599355281357,
+      "epoch": 0.2814538676607642,
+      "grad_norm": 1.6152157706382257,
+      "learning_rate": 1.875776397515528e-05,
+      "loss": 2.1151,
+      "step": 302
+    },
+    {
+      "cosine_similarity": 0.5104094379303181,
+      "epoch": 0.28238583410997203,
+      "grad_norm": 3.1845029438849504,
+      "learning_rate": 1.881987577639752e-05,
+      "loss": 2.6596,
+      "step": 303
+    },
+    {
+      "cosine_similarity": 0.7187136396199281,
+      "epoch": 0.2833178005591799,
+      "grad_norm": 2.3168754243048206,
+      "learning_rate": 1.8881987577639754e-05,
+      "loss": 2.5943,
+      "step": 304
+    },
+    {
+      "cosine_similarity": 0.10766034430370382,
+      "epoch": 0.2842497670083877,
+      "grad_norm": 2.174051992234828,
+      "learning_rate": 1.894409937888199e-05,
+      "loss": 2.5151,
+      "step": 305
+    },
+    {
+      "cosine_similarity": 0.2575948453866291,
+      "epoch": 0.2851817334575955,
+      "grad_norm": 2.517360166165975,
+      "learning_rate": 1.9006211180124224e-05,
+      "loss": 2.7756,
+      "step": 306
+    },
+    {
+      "cosine_similarity": 0.037108518191977295,
+      "epoch": 0.28611369990680335,
+      "grad_norm": 1.8585964859647544,
+      "learning_rate": 1.906832298136646e-05,
+      "loss": 2.5313,
+      "step": 307
+    },
+    {
+      "cosine_similarity": 0.551073616270824,
+      "epoch": 0.2870456663560112,
+      "grad_norm": 1.9537820594945614,
+      "learning_rate": 1.9130434782608697e-05,
+      "loss": 2.6658,
+      "step": 308
+    },
+    {
+      "cosine_similarity": 0.06321052541929961,
+      "epoch": 0.287977632805219,
+      "grad_norm": 1.9327856454433918,
+      "learning_rate": 1.9192546583850932e-05,
+      "loss": 2.4954,
+      "step": 309
+    },
+    {
+      "cosine_similarity": 0.12305976876006555,
+      "epoch": 0.28890959925442683,
+      "grad_norm": 1.5328328135053981,
+      "learning_rate": 1.925465838509317e-05,
+      "loss": 2.0428,
+      "step": 310
+    },
+    {
+      "cosine_similarity": 0.3401513451515945,
+      "epoch": 0.2898415657036347,
+      "grad_norm": 2.818826598349484,
+      "learning_rate": 1.9316770186335406e-05,
+      "loss": 2.7862,
+      "step": 311
+    },
+    {
+      "cosine_similarity": 0.010869098173710436,
+      "epoch": 0.2907735321528425,
+      "grad_norm": 2.569724230085298,
+      "learning_rate": 1.937888198757764e-05,
+      "loss": 2.4741,
+      "step": 312
+    },
+    {
+      "cosine_similarity": 0.07336670917390764,
+      "epoch": 0.2917054986020503,
+      "grad_norm": 1.412704177810487,
+      "learning_rate": 1.9440993788819876e-05,
+      "loss": 2.0735,
+      "step": 313
+    },
+    {
+      "cosine_similarity": 0.09135173063792407,
+      "epoch": 0.29263746505125815,
+      "grad_norm": 2.4001300897352564,
+      "learning_rate": 1.950310559006211e-05,
+      "loss": 2.6318,
+      "step": 314
+    },
+    {
+      "cosine_similarity": 0.10342584941086222,
+      "epoch": 0.293569431500466,
+      "grad_norm": 1.918648680925207,
+      "learning_rate": 1.956521739130435e-05,
+      "loss": 2.713,
+      "step": 315
+    },
+    {
+      "cosine_similarity": 0.5741095090667167,
+      "epoch": 0.29450139794967384,
+      "grad_norm": 1.6109667233040341,
+      "learning_rate": 1.9627329192546585e-05,
+      "loss": 2.4806,
+      "step": 316
+    },
+    {
+      "cosine_similarity": 0.08495209792272432,
+      "epoch": 0.2954333643988816,
+      "grad_norm": 1.8950571678120118,
+      "learning_rate": 1.9689440993788823e-05,
+      "loss": 2.4934,
+      "step": 317
+    },
+    {
+      "cosine_similarity": 0.03454801601258479,
+      "epoch": 0.29636533084808947,
+      "grad_norm": 1.8731339604660653,
+      "learning_rate": 1.9751552795031058e-05,
+      "loss": 2.4396,
+      "step": 318
+    },
+    {
+      "cosine_similarity": 0.22396272890505514,
+      "epoch": 0.2972972972972973,
+      "grad_norm": 1.687323198944775,
+      "learning_rate": 1.9813664596273293e-05,
+      "loss": 2.6104,
+      "step": 319
+    },
+    {
+      "cosine_similarity": -0.011127045216001236,
+      "epoch": 0.2982292637465051,
+      "grad_norm": 2.2873982295833204,
+      "learning_rate": 1.9875776397515528e-05,
+      "loss": 2.5479,
+      "step": 320
+    },
+    {
+      "cosine_similarity": 0.1188929932359796,
+      "epoch": 0.29916123019571295,
+      "grad_norm": 1.8043863889603848,
+      "learning_rate": 1.9937888198757767e-05,
+      "loss": 2.5565,
+      "step": 321
+    },
+    {
+      "cosine_similarity": 0.25272161775255275,
+      "epoch": 0.3000931966449208,
+      "grad_norm": 1.4396666058982401,
+      "learning_rate": 2e-05,
+      "loss": 2.2043,
+      "step": 322
+    },
+    {
+      "cosine_similarity": 0.17909302925582807,
+      "epoch": 0.30102516309412863,
+      "grad_norm": 2.2790793187459073,
+      "learning_rate": 1.999309630652399e-05,
+      "loss": 2.7947,
+      "step": 323
+    },
+    {
+      "cosine_similarity": 0.7257057651949927,
+      "epoch": 0.3019571295433364,
+      "grad_norm": 1.7327188502236992,
+      "learning_rate": 1.9986192613047983e-05,
+      "loss": 1.7898,
+      "step": 324
+    },
+    {
+      "cosine_similarity": 0.6029475250912587,
+      "epoch": 0.30288909599254427,
+      "grad_norm": 1.8323191371649268,
+      "learning_rate": 1.9979288919571972e-05,
+      "loss": 2.6497,
+      "step": 325
+    },
+    {
+      "cosine_similarity": 0.0026485781919862303,
+      "epoch": 0.3038210624417521,
+      "grad_norm": 1.7797221944209,
+      "learning_rate": 1.997238522609596e-05,
+      "loss": 2.2815,
+      "step": 326
+    },
+    {
+      "cosine_similarity": 0.6683629807459791,
+      "epoch": 0.3047530288909599,
+      "grad_norm": 1.7912118586652257,
+      "learning_rate": 1.9965481532619954e-05,
+      "loss": 2.3027,
+      "step": 327
+    },
+    {
+      "cosine_similarity": 0.4579788462627103,
+      "epoch": 0.30568499534016774,
+      "grad_norm": 1.7122337711528663,
+      "learning_rate": 1.9958577839143946e-05,
+      "loss": 2.5434,
+      "step": 328
+    },
+    {
+      "cosine_similarity": -0.01745498354414477,
+      "epoch": 0.3066169617893756,
+      "grad_norm": 2.7448032324450713,
+      "learning_rate": 1.9951674145667935e-05,
+      "loss": 2.5497,
+      "step": 329
+    },
+    {
+      "cosine_similarity": 0.1728722016826905,
+      "epoch": 0.30754892823858343,
+      "grad_norm": 1.6345186948639916,
+      "learning_rate": 1.9944770452191924e-05,
+      "loss": 2.4184,
+      "step": 330
+    },
+    {
+      "cosine_similarity": 0.1957164634226352,
+      "epoch": 0.3084808946877912,
+      "grad_norm": 1.7450112522902077,
+      "learning_rate": 1.9937866758715913e-05,
+      "loss": 2.151,
+      "step": 331
+    },
+    {
+      "cosine_similarity": 0.2725352010021395,
+      "epoch": 0.30941286113699906,
+      "grad_norm": 1.5776336027514666,
+      "learning_rate": 1.9930963065239906e-05,
+      "loss": 1.9193,
+      "step": 332
+    },
+    {
+      "cosine_similarity": 0.20239288391639307,
+      "epoch": 0.3103448275862069,
+      "grad_norm": 1.853741106583004,
+      "learning_rate": 1.9924059371763895e-05,
+      "loss": 2.2188,
+      "step": 333
+    },
+    {
+      "cosine_similarity": 0.3192733730510847,
+      "epoch": 0.31127679403541475,
+      "grad_norm": 2.242643937137718,
+      "learning_rate": 1.9917155678287884e-05,
+      "loss": 2.5411,
+      "step": 334
+    },
+    {
+      "cosine_similarity": 0.2522909088704467,
+      "epoch": 0.31220876048462254,
+      "grad_norm": 1.7711508231783126,
+      "learning_rate": 1.9910251984811876e-05,
+      "loss": 2.142,
+      "step": 335
+    },
+    {
+      "cosine_similarity": 0.24370202099538593,
+      "epoch": 0.3131407269338304,
+      "grad_norm": 1.8745677267962981,
+      "learning_rate": 1.9903348291335865e-05,
+      "loss": 2.5642,
+      "step": 336
+    },
+    {
+      "cosine_similarity": 0.2749455414031907,
+      "epoch": 0.3140726933830382,
+      "grad_norm": 1.9137358011131067,
+      "learning_rate": 1.9896444597859858e-05,
+      "loss": 2.3313,
+      "step": 337
+    },
+    {
+      "cosine_similarity": 0.17569333036015797,
+      "epoch": 0.315004659832246,
+      "grad_norm": 1.806524222244171,
+      "learning_rate": 1.9889540904383847e-05,
+      "loss": 2.1867,
+      "step": 338
+    },
+    {
+      "cosine_similarity": 0.13353273973658197,
+      "epoch": 0.31593662628145386,
+      "grad_norm": 2.430530562545359,
+      "learning_rate": 1.988263721090784e-05,
+      "loss": 2.4495,
+      "step": 339
+    },
+    {
+      "cosine_similarity": 0.44734888016912716,
+      "epoch": 0.3168685927306617,
+      "grad_norm": 1.7715630436661114,
+      "learning_rate": 1.9875733517431828e-05,
+      "loss": 2.6872,
+      "step": 340
+    },
+    {
+      "cosine_similarity": 0.5148077269089032,
+      "epoch": 0.31780055917986955,
+      "grad_norm": 1.822199919864696,
+      "learning_rate": 1.9868829823955817e-05,
+      "loss": 2.5645,
+      "step": 341
+    },
+    {
+      "cosine_similarity": 0.2602128588835826,
+      "epoch": 0.31873252562907733,
+      "grad_norm": 2.436483165125212,
+      "learning_rate": 1.9861926130479806e-05,
+      "loss": 2.3682,
+      "step": 342
+    },
+    {
+      "cosine_similarity": 0.9905707701416825,
+      "epoch": 0.3196644920782852,
+      "grad_norm": 1.7980008424629639,
+      "learning_rate": 1.98550224370038e-05,
+      "loss": 2.3925,
+      "step": 343
+    },
+    {
+      "cosine_similarity": 0.5308331687878489,
+      "epoch": 0.320596458527493,
+      "grad_norm": 2.002683622948421,
+      "learning_rate": 1.9848118743527788e-05,
+      "loss": 2.5469,
+      "step": 344
+    },
+    {
+      "cosine_similarity": 0.5339997746117608,
+      "epoch": 0.32152842497670087,
+      "grad_norm": 2.309398197947059,
+      "learning_rate": 1.984121505005178e-05,
+      "loss": 2.7302,
+      "step": 345
+    },
+    {
+      "cosine_similarity": 0.5159616652553183,
+      "epoch": 0.32246039142590865,
+      "grad_norm": 1.6248930946161664,
+      "learning_rate": 1.983431135657577e-05,
+      "loss": 2.1001,
+      "step": 346
+    },
+    {
+      "cosine_similarity": 0.3734294046480442,
+      "epoch": 0.3233923578751165,
+      "grad_norm": 1.7670915331095598,
+      "learning_rate": 1.982740766309976e-05,
+      "loss": 2.1903,
+      "step": 347
+    },
+    {
+      "cosine_similarity": 0.6478431361335384,
+      "epoch": 0.32432432432432434,
+      "grad_norm": 1.4933615147444528,
+      "learning_rate": 1.982050396962375e-05,
+      "loss": 2.1585,
+      "step": 348
+    },
+    {
+      "cosine_similarity": 0.6601761002538271,
+      "epoch": 0.32525629077353213,
+      "grad_norm": 2.2115195286907485,
+      "learning_rate": 1.981360027614774e-05,
+      "loss": 2.4346,
+      "step": 349
+    },
+    {
+      "cosine_similarity": 0.033551776340731004,
+      "epoch": 0.32618825722274,
+      "grad_norm": 1.673421919527606,
+      "learning_rate": 1.9806696582671732e-05,
+      "loss": 2.6585,
+      "step": 350
+    },
+    {
+      "cosine_similarity": 0.017133806336675893,
+      "epoch": 0.3271202236719478,
+      "grad_norm": 1.777653726545176,
+      "learning_rate": 1.979979288919572e-05,
+      "loss": 2.3724,
+      "step": 351
+    },
+    {
+      "cosine_similarity": 0.6129822313105453,
+      "epoch": 0.32805219012115566,
+      "grad_norm": 1.741376389576883,
+      "learning_rate": 1.979288919571971e-05,
+      "loss": 2.5298,
+      "step": 352
+    },
+    {
+      "cosine_similarity": 0.28871869493750874,
+      "epoch": 0.32898415657036345,
+      "grad_norm": 2.0957077822393595,
+      "learning_rate": 1.9785985502243702e-05,
+      "loss": 2.8433,
+      "step": 353
+    },
+    {
+      "cosine_similarity": 0.22163185845103409,
+      "epoch": 0.3299161230195713,
+      "grad_norm": 1.9386352371277886,
+      "learning_rate": 1.977908180876769e-05,
+      "loss": 2.4426,
+      "step": 354
+    },
+    {
+      "cosine_similarity": 0.3005620524879735,
+      "epoch": 0.33084808946877914,
+      "grad_norm": 1.714062633822939,
+      "learning_rate": 1.9772178115291684e-05,
+      "loss": 2.4457,
+      "step": 355
+    },
+    {
+      "cosine_similarity": -0.00810541180329141,
+      "epoch": 0.3317800559179869,
+      "grad_norm": 2.123780046878567,
+      "learning_rate": 1.9765274421815673e-05,
+      "loss": 2.7253,
+      "step": 356
+    },
+    {
+      "cosine_similarity": 0.04414604079280311,
+      "epoch": 0.33271202236719477,
+      "grad_norm": 1.7510546206677564,
+      "learning_rate": 1.9758370728339665e-05,
+      "loss": 1.9673,
+      "step": 357
+    },
+    {
+      "cosine_similarity": 0.4900772028014338,
+      "epoch": 0.3336439888164026,
+      "grad_norm": 2.0274579598375126,
+      "learning_rate": 1.9751467034863654e-05,
+      "loss": 2.4127,
+      "step": 358
+    },
+    {
+      "cosine_similarity": -0.004989649384033018,
+      "epoch": 0.33457595526561046,
+      "grad_norm": 1.7290653760282848,
+      "learning_rate": 1.9744563341387643e-05,
+      "loss": 2.4318,
+      "step": 359
+    },
+    {
+      "cosine_similarity": -0.0030430554043300957,
+      "epoch": 0.33550792171481825,
+      "grad_norm": 1.606450984303285,
+      "learning_rate": 1.9737659647911633e-05,
+      "loss": 2.7513,
+      "step": 360
+    },
+    {
+      "cosine_similarity": 0.3009578434382676,
+      "epoch": 0.3364398881640261,
+      "grad_norm": 1.9954439352132862,
+      "learning_rate": 1.9730755954435625e-05,
+      "loss": 2.5123,
+      "step": 361
+    },
+    {
+      "cosine_similarity": 0.24922737083403448,
+      "epoch": 0.33737185461323393,
+      "grad_norm": 1.818538468131917,
+      "learning_rate": 1.9723852260959614e-05,
+      "loss": 1.8516,
+      "step": 362
+    },
+    {
+      "cosine_similarity": 0.5565853667925567,
+      "epoch": 0.3383038210624418,
+      "grad_norm": 1.8564243050729992,
+      "learning_rate": 1.9716948567483606e-05,
+      "loss": 2.4098,
+      "step": 363
+    },
+    {
+      "cosine_similarity": 0.06795742238133258,
+      "epoch": 0.33923578751164957,
+      "grad_norm": 1.9778605534612486,
+      "learning_rate": 1.9710044874007595e-05,
+      "loss": 2.7604,
+      "step": 364
+    },
+    {
+      "cosine_similarity": 0.22718077833631167,
+      "epoch": 0.3401677539608574,
+      "grad_norm": 1.7113017935922632,
+      "learning_rate": 1.9703141180531588e-05,
+      "loss": 2.2686,
+      "step": 365
+    },
+    {
+      "cosine_similarity": 0.6466314218129638,
+      "epoch": 0.34109972041006525,
+      "grad_norm": 2.031993197910752,
+      "learning_rate": 1.9696237487055577e-05,
+      "loss": 2.7322,
+      "step": 366
+    },
+    {
+      "cosine_similarity": 0.19428583621191572,
+      "epoch": 0.34203168685927304,
+      "grad_norm": 1.6584264960538244,
+      "learning_rate": 1.9689333793579566e-05,
+      "loss": 2.7181,
+      "step": 367
+    },
+    {
+      "cosine_similarity": 0.01384767918573874,
+      "epoch": 0.3429636533084809,
+      "grad_norm": 1.5817875097619023,
+      "learning_rate": 1.968243010010356e-05,
+      "loss": 2.0305,
+      "step": 368
+    },
+    {
+      "cosine_similarity": 0.4165502408209592,
+      "epoch": 0.34389561975768873,
+      "grad_norm": 2.172867667928234,
+      "learning_rate": 1.9675526406627547e-05,
+      "loss": 2.3591,
+      "step": 369
+    },
+    {
+      "cosine_similarity": 0.47646039668773305,
+      "epoch": 0.3448275862068966,
+      "grad_norm": 2.707390099593222,
+      "learning_rate": 1.9668622713151536e-05,
+      "loss": 2.6585,
+      "step": 370
+    },
+    {
+      "cosine_similarity": 0.467156343317729,
+      "epoch": 0.34575955265610436,
+      "grad_norm": 2.0282900108803554,
+      "learning_rate": 1.9661719019675526e-05,
+      "loss": 2.5167,
+      "step": 371
+    },
+    {
+      "cosine_similarity": -0.14265967273051883,
+      "epoch": 0.3466915191053122,
+      "grad_norm": 1.8093158575049355,
+      "learning_rate": 1.9654815326199518e-05,
+      "loss": 2.5221,
+      "step": 372
+    },
+    {
+      "cosine_similarity": 0.2334942859511812,
+      "epoch": 0.34762348555452005,
+      "grad_norm": 1.5848283540245596,
+      "learning_rate": 1.9647911632723507e-05,
+      "loss": 2.1208,
+      "step": 373
+    },
+    {
+      "cosine_similarity": 0.3479185111958745,
+      "epoch": 0.34855545200372784,
+      "grad_norm": 2.0502487199482835,
+      "learning_rate": 1.96410079392475e-05,
+      "loss": 2.3723,
+      "step": 374
+    },
+    {
+      "cosine_similarity": 0.3269020907654852,
+      "epoch": 0.3494874184529357,
+      "grad_norm": 1.7555514089362232,
+      "learning_rate": 1.9634104245771492e-05,
+      "loss": 2.5739,
+      "step": 375
+    },
+    {
+      "cosine_similarity": 0.2902242766425429,
+      "epoch": 0.3504193849021435,
+      "grad_norm": 2.109190286793207,
+      "learning_rate": 1.962720055229548e-05,
+      "loss": 2.5491,
+      "step": 376
+    },
+    {
+      "cosine_similarity": 0.1730811767777797,
+      "epoch": 0.35135135135135137,
+      "grad_norm": 1.4935407934625649,
+      "learning_rate": 1.962029685881947e-05,
+      "loss": 2.0314,
+      "step": 377
+    },
+    {
+      "cosine_similarity": 0.1841138953777689,
+      "epoch": 0.35228331780055916,
+      "grad_norm": 1.98704040177181,
+      "learning_rate": 1.961339316534346e-05,
+      "loss": 2.4391,
+      "step": 378
+    },
+    {
+      "cosine_similarity": 0.18104282162241722,
+      "epoch": 0.353215284249767,
+      "grad_norm": 1.7499501291681245,
+      "learning_rate": 1.960648947186745e-05,
+      "loss": 2.2631,
+      "step": 379
+    },
+    {
+      "cosine_similarity": 0.03956318601606164,
+      "epoch": 0.35414725069897485,
+      "grad_norm": 1.5808959029988983,
+      "learning_rate": 1.959958577839144e-05,
+      "loss": 2.5912,
+      "step": 380
+    },
+    {
+      "cosine_similarity": 0.15479963332200633,
+      "epoch": 0.3550792171481827,
+      "grad_norm": 1.9512486237081657,
+      "learning_rate": 1.959268208491543e-05,
+      "loss": 2.2606,
+      "step": 381
+    },
+    {
+      "cosine_similarity": 0.48653097692726954,
+      "epoch": 0.3560111835973905,
+      "grad_norm": 1.673746332699039,
+      "learning_rate": 1.9585778391439422e-05,
+      "loss": 2.2468,
+      "step": 382
+    },
+    {
+      "cosine_similarity": 0.44761708925624677,
+      "epoch": 0.3569431500465983,
+      "grad_norm": 1.840453395716475,
+      "learning_rate": 1.957887469796341e-05,
+      "loss": 2.6003,
+      "step": 383
+    },
+    {
+      "cosine_similarity": -0.05754760497552247,
+      "epoch": 0.35787511649580617,
+      "grad_norm": 1.2692712115733702,
+      "learning_rate": 1.9571971004487403e-05,
+      "loss": 2.0253,
+      "step": 384
+    },
+    {
+      "cosine_similarity": 0.3319997441627461,
+      "epoch": 0.35880708294501396,
+      "grad_norm": 1.8989907646382862,
+      "learning_rate": 1.9565067311011392e-05,
+      "loss": 2.5905,
+      "step": 385
+    },
+    {
+      "cosine_similarity": 0.5884646934863367,
+      "epoch": 0.3597390493942218,
+      "grad_norm": 1.7290166842001895,
+      "learning_rate": 1.9558163617535385e-05,
+      "loss": 2.4701,
+      "step": 386
+    },
+    {
+      "cosine_similarity": 0.406662425138792,
+      "epoch": 0.36067101584342964,
+      "grad_norm": 1.6539708520958532,
+      "learning_rate": 1.9551259924059374e-05,
+      "loss": 2.6101,
+      "step": 387
+    },
+    {
+      "cosine_similarity": 0.3659154956171942,
+      "epoch": 0.3616029822926375,
+      "grad_norm": 1.7711590358762308,
+      "learning_rate": 1.9544356230583363e-05,
+      "loss": 2.6861,
+      "step": 388
+    },
+    {
+      "cosine_similarity": 0.18496376628667893,
+      "epoch": 0.3625349487418453,
+      "grad_norm": 1.3367841723765757,
+      "learning_rate": 1.9537452537107352e-05,
+      "loss": 2.3974,
+      "step": 389
+    },
+    {
+      "cosine_similarity": 0.023089564687308592,
+      "epoch": 0.3634669151910531,
+      "grad_norm": 1.508727319393286,
+      "learning_rate": 1.9530548843631344e-05,
+      "loss": 2.0327,
+      "step": 390
+    },
+    {
+      "cosine_similarity": 0.4261458803796201,
+      "epoch": 0.36439888164026096,
+      "grad_norm": 1.9466522204515864,
+      "learning_rate": 1.9523645150155333e-05,
+      "loss": 2.4462,
+      "step": 391
+    },
+    {
+      "cosine_similarity": -0.3509288040727624,
+      "epoch": 0.36533084808946875,
+      "grad_norm": 3.766333922961592,
+      "learning_rate": 1.9516741456679326e-05,
+      "loss": 3.0689,
+      "step": 392
+    },
+    {
+      "cosine_similarity": -0.07190658023950408,
+      "epoch": 0.3662628145386766,
+      "grad_norm": 1.833682502009459,
+      "learning_rate": 1.9509837763203315e-05,
+      "loss": 2.4614,
+      "step": 393
+    },
+    {
+      "cosine_similarity": 0.10157331856498894,
+      "epoch": 0.36719478098788444,
+      "grad_norm": 1.7014312921901942,
+      "learning_rate": 1.9502934069727307e-05,
+      "loss": 2.3424,
+      "step": 394
+    },
+    {
+      "cosine_similarity": 0.35577205791090855,
+      "epoch": 0.3681267474370923,
+      "grad_norm": 2.1351619479291046,
+      "learning_rate": 1.9496030376251296e-05,
+      "loss": 2.1201,
+      "step": 395
+    },
+    {
+      "cosine_similarity": -0.033085959643376225,
+      "epoch": 0.36905871388630007,
+      "grad_norm": 1.910527318702723,
+      "learning_rate": 1.9489126682775285e-05,
+      "loss": 2.7485,
+      "step": 396
+    },
+    {
+      "cosine_similarity": 0.12175432447531684,
+      "epoch": 0.3699906803355079,
+      "grad_norm": 1.8998342862440556,
+      "learning_rate": 1.9482222989299278e-05,
+      "loss": 2.0682,
+      "step": 397
+    },
+    {
+      "cosine_similarity": 0.031051282621675065,
+      "epoch": 0.37092264678471576,
+      "grad_norm": 1.7495644524206744,
+      "learning_rate": 1.9475319295823267e-05,
+      "loss": 2.351,
+      "step": 398
+    },
+    {
+      "cosine_similarity": 0.33107420431383083,
+      "epoch": 0.3718546132339236,
+      "grad_norm": 1.4268339925170084,
+      "learning_rate": 1.9468415602347256e-05,
+      "loss": 2.5046,
+      "step": 399
+    },
+    {
+      "cosine_similarity": 0.04074459267316289,
+      "epoch": 0.3727865796831314,
+      "grad_norm": 2.5221601383984016,
+      "learning_rate": 1.9461511908871248e-05,
+      "loss": 2.3392,
+      "step": 400
+    },
+    {
+      "cosine_similarity": 0.26405734589415636,
+      "epoch": 0.37371854613233924,
+      "grad_norm": 2.266464738635341,
+      "learning_rate": 1.9454608215395237e-05,
+      "loss": 2.5821,
+      "step": 401
+    },
+    {
+      "cosine_similarity": 0.13042726688765854,
+      "epoch": 0.3746505125815471,
+      "grad_norm": 2.189418558617591,
+      "learning_rate": 1.944770452191923e-05,
+      "loss": 2.5639,
+      "step": 402
+    },
+    {
+      "cosine_similarity": 0.0782212577290033,
+      "epoch": 0.37558247903075487,
+      "grad_norm": 1.5806066905661282,
+      "learning_rate": 1.944080082844322e-05,
+      "loss": 2.5556,
+      "step": 403
+    },
+    {
+      "cosine_similarity": 0.21374880525655335,
+      "epoch": 0.3765144454799627,
+      "grad_norm": 1.3387605894817636,
+      "learning_rate": 1.943389713496721e-05,
+      "loss": 2.3404,
+      "step": 404
+    },
+    {
+      "cosine_similarity": 0.15238502817011543,
+      "epoch": 0.37744641192917056,
+      "grad_norm": 1.8483032684355467,
+      "learning_rate": 1.94269934414912e-05,
+      "loss": 2.1114,
+      "step": 405
+    },
+    {
+      "cosine_similarity": 0.3501452678381618,
+      "epoch": 0.3783783783783784,
+      "grad_norm": 1.8276272177173762,
+      "learning_rate": 1.942008974801519e-05,
+      "loss": 2.2458,
+      "step": 406
+    },
+    {
+      "cosine_similarity": 0.1307753850781423,
+      "epoch": 0.3793103448275862,
+      "grad_norm": 1.7909614178498445,
+      "learning_rate": 1.941318605453918e-05,
+      "loss": 2.3103,
+      "step": 407
+    },
+    {
+      "cosine_similarity": 0.21268210947867577,
+      "epoch": 0.38024231127679403,
+      "grad_norm": 1.662877533366739,
+      "learning_rate": 1.940628236106317e-05,
+      "loss": 2.1261,
+      "step": 408
+    },
+    {
+      "cosine_similarity": 0.10997665683270502,
+      "epoch": 0.3811742777260019,
+      "grad_norm": 2.3244534685777327,
+      "learning_rate": 1.939937866758716e-05,
+      "loss": 2.0986,
+      "step": 409
+    },
+    {
+      "cosine_similarity": 0.14467483187111438,
+      "epoch": 0.3821062441752097,
+      "grad_norm": 2.1360438402079236,
+      "learning_rate": 1.9392474974111152e-05,
+      "loss": 2.4894,
+      "step": 410
+    },
+    {
+      "cosine_similarity": 0.12280600806029718,
+      "epoch": 0.3830382106244175,
+      "grad_norm": 1.725275631635974,
+      "learning_rate": 1.938557128063514e-05,
+      "loss": 2.212,
+      "step": 411
+    },
+    {
+      "cosine_similarity": 0.25952246002217105,
+      "epoch": 0.38397017707362535,
+      "grad_norm": 1.9028669775493474,
+      "learning_rate": 1.9378667587159134e-05,
+      "loss": 2.8009,
+      "step": 412
+    },
+    {
+      "cosine_similarity": 0.21299963428637458,
+      "epoch": 0.3849021435228332,
+      "grad_norm": 2.3261042576961035,
+      "learning_rate": 1.9371763893683123e-05,
+      "loss": 2.6453,
+      "step": 413
+    },
+    {
+      "cosine_similarity": 1.010834055025008,
+      "epoch": 0.385834109972041,
+      "grad_norm": 1.4130355325468476,
+      "learning_rate": 1.9364860200207112e-05,
+      "loss": 2.0033,
+      "step": 414
+    },
+    {
+      "cosine_similarity": 0.21040586901086136,
+      "epoch": 0.38676607642124883,
+      "grad_norm": 2.158196375778959,
+      "learning_rate": 1.9357956506731104e-05,
+      "loss": 2.1832,
+      "step": 415
+    },
+    {
+      "cosine_similarity": 0.19076368215239578,
+      "epoch": 0.38769804287045667,
+      "grad_norm": 1.6105795614499274,
+      "learning_rate": 1.9351052813255093e-05,
+      "loss": 2.2325,
+      "step": 416
+    },
+    {
+      "cosine_similarity": 0.3767539574744773,
+      "epoch": 0.3886300093196645,
+      "grad_norm": 1.5623961428733137,
+      "learning_rate": 1.9344149119779082e-05,
+      "loss": 2.3042,
+      "step": 417
+    },
+    {
+      "cosine_similarity": 0.1507077009138864,
+      "epoch": 0.3895619757688723,
+      "grad_norm": 1.607823703696593,
+      "learning_rate": 1.933724542630307e-05,
+      "loss": 2.3993,
+      "step": 418
+    },
+    {
+      "cosine_similarity": 0.3018622863851018,
+      "epoch": 0.39049394221808015,
+      "grad_norm": 1.8129286999257817,
+      "learning_rate": 1.9330341732827064e-05,
+      "loss": 2.4895,
+      "step": 419
+    },
+    {
+      "cosine_similarity": 0.23820298893051037,
+      "epoch": 0.391425908667288,
+      "grad_norm": 1.4085893011929052,
+      "learning_rate": 1.9323438039351056e-05,
+      "loss": 2.5387,
+      "step": 420
+    },
+    {
+      "cosine_similarity": 0.12232854907037384,
+      "epoch": 0.3923578751164958,
+      "grad_norm": 1.5702077983719323,
+      "learning_rate": 1.9316534345875045e-05,
+      "loss": 2.3855,
+      "step": 421
+    },
+    {
+      "cosine_similarity": 0.013022837860411607,
+      "epoch": 0.3932898415657036,
+      "grad_norm": 1.6374321893811856,
+      "learning_rate": 1.9309630652399034e-05,
+      "loss": 2.3269,
+      "step": 422
+    },
+    {
+      "cosine_similarity": 0.10142901016423446,
+      "epoch": 0.39422180801491147,
+      "grad_norm": 2.0045920745517356,
+      "learning_rate": 1.9302726958923027e-05,
+      "loss": 2.1937,
+      "step": 423
+    },
+    {
+      "cosine_similarity": 0.3241714708523656,
+      "epoch": 0.3951537744641193,
+      "grad_norm": 2.232667143250206,
+      "learning_rate": 1.9295823265447016e-05,
+      "loss": 2.673,
+      "step": 424
+    },
+    {
+      "cosine_similarity": 0.3225132970222915,
+      "epoch": 0.3960857409133271,
+      "grad_norm": 1.3829934952936336,
+      "learning_rate": 1.9288919571971005e-05,
+      "loss": 1.9911,
+      "step": 425
+    },
+    {
+      "cosine_similarity": 0.050688256083479766,
+      "epoch": 0.39701770736253494,
+      "grad_norm": 1.4146822032778665,
+      "learning_rate": 1.9282015878494997e-05,
+      "loss": 1.8457,
+      "step": 426
+    },
+    {
+      "cosine_similarity": 0.3554852848058485,
+      "epoch": 0.3979496738117428,
+      "grad_norm": 1.4878390663884808,
+      "learning_rate": 1.9275112185018986e-05,
+      "loss": 2.2361,
+      "step": 427
+    },
+    {
+      "cosine_similarity": 0.2649086918044299,
+      "epoch": 0.39888164026095063,
+      "grad_norm": 1.8684821293374514,
+      "learning_rate": 1.9268208491542975e-05,
+      "loss": 2.3383,
+      "step": 428
+    },
+    {
+      "cosine_similarity": 0.1305956700113134,
+      "epoch": 0.3998136067101584,
+      "grad_norm": 1.6070342186473343,
+      "learning_rate": 1.9261304798066968e-05,
+      "loss": 2.4468,
+      "step": 429
+    },
+    {
+      "cosine_similarity": 0.3120460828956451,
+      "epoch": 0.40074557315936626,
+      "grad_norm": 1.5279075423994986,
+      "learning_rate": 1.925440110459096e-05,
+      "loss": 2.3281,
+      "step": 430
+    },
+    {
+      "cosine_similarity": 0.1774579586542875,
+      "epoch": 0.4016775396085741,
+      "grad_norm": 1.6950756371083593,
+      "learning_rate": 1.924749741111495e-05,
+      "loss": 2.674,
+      "step": 431
+    },
+    {
+      "cosine_similarity": 0.21714882357869095,
+      "epoch": 0.4026095060577819,
+      "grad_norm": 1.4555809882645372,
+      "learning_rate": 1.9240593717638938e-05,
+      "loss": 1.978,
+      "step": 432
+    },
+    {
+      "cosine_similarity": 0.21195598687144185,
+      "epoch": 0.40354147250698974,
+      "grad_norm": 1.965390105667436,
+      "learning_rate": 1.9233690024162927e-05,
+      "loss": 2.7993,
+      "step": 433
+    },
+    {
+      "cosine_similarity": 0.39987813054152915,
+      "epoch": 0.4044734389561976,
+      "grad_norm": 1.7183987649904287,
+      "learning_rate": 1.922678633068692e-05,
+      "loss": 2.1919,
+      "step": 434
+    },
+    {
+      "cosine_similarity": 0.3618191877125198,
+      "epoch": 0.40540540540540543,
+      "grad_norm": 1.6432967434124623,
+      "learning_rate": 1.921988263721091e-05,
+      "loss": 2.135,
+      "step": 435
+    },
+    {
+      "cosine_similarity": -0.07300024851840854,
+      "epoch": 0.4063373718546132,
+      "grad_norm": 1.8465243170909025,
+      "learning_rate": 1.9212978943734898e-05,
+      "loss": 2.6408,
+      "step": 436
+    },
+    {
+      "cosine_similarity": 0.19416823403900102,
+      "epoch": 0.40726933830382106,
+      "grad_norm": 2.149749526196831,
+      "learning_rate": 1.920607525025889e-05,
+      "loss": 2.4511,
+      "step": 437
+    },
+    {
+      "cosine_similarity": 0.40900567252015746,
+      "epoch": 0.4082013047530289,
+      "grad_norm": 2.1059697798467507,
+      "learning_rate": 1.919917155678288e-05,
+      "loss": 2.3944,
+      "step": 438
+    },
+    {
+      "cosine_similarity": 0.24667635017010459,
+      "epoch": 0.4091332712022367,
+      "grad_norm": 1.5564693818101423,
+      "learning_rate": 1.919226786330687e-05,
+      "loss": 2.1918,
+      "step": 439
+    },
+    {
+      "cosine_similarity": 0.0777647744583153,
+      "epoch": 0.41006523765144454,
+      "grad_norm": 2.788444329036235,
+      "learning_rate": 1.918536416983086e-05,
+      "loss": 2.3804,
+      "step": 440
+    },
+    {
+      "cosine_similarity": 0.19327906067956208,
+      "epoch": 0.4109972041006524,
+      "grad_norm": 1.5719915153420057,
+      "learning_rate": 1.9178460476354853e-05,
+      "loss": 2.4029,
+      "step": 441
+    },
+    {
+      "cosine_similarity": 0.17653376444130997,
+      "epoch": 0.4119291705498602,
+      "grad_norm": 1.9813318505510218,
+      "learning_rate": 1.9171556782878842e-05,
+      "loss": 2.3671,
+      "step": 442
+    },
+    {
+      "cosine_similarity": 0.08637658781101987,
+      "epoch": 0.412861136999068,
+      "grad_norm": 1.7001206984505435,
+      "learning_rate": 1.916465308940283e-05,
+      "loss": 2.247,
+      "step": 443
+    },
+    {
+      "cosine_similarity": 0.05224229702192771,
+      "epoch": 0.41379310344827586,
+      "grad_norm": 1.8684475168205008,
+      "learning_rate": 1.9157749395926824e-05,
+      "loss": 2.504,
+      "step": 444
+    },
+    {
+      "cosine_similarity": 0.5087067294430269,
+      "epoch": 0.4147250698974837,
+      "grad_norm": 2.0186161682140247,
+      "learning_rate": 1.9150845702450813e-05,
+      "loss": 2.4394,
+      "step": 445
+    },
+    {
+      "cosine_similarity": 0.1611446925324606,
+      "epoch": 0.41565703634669154,
+      "grad_norm": 1.631412323040377,
+      "learning_rate": 1.91439420089748e-05,
+      "loss": 2.5374,
+      "step": 446
+    },
+    {
+      "cosine_similarity": 0.07341612608562313,
+      "epoch": 0.41658900279589933,
+      "grad_norm": 2.2628088041019674,
+      "learning_rate": 1.9137038315498794e-05,
+      "loss": 2.3697,
+      "step": 447
+    },
+    {
+      "cosine_similarity": 0.019825162947612524,
+      "epoch": 0.4175209692451072,
+      "grad_norm": 1.3583015799787534,
+      "learning_rate": 1.9130134622022783e-05,
+      "loss": 2.053,
+      "step": 448
+    },
+    {
+      "cosine_similarity": 0.17313794261629104,
+      "epoch": 0.418452935694315,
+      "grad_norm": 1.68333008390759,
+      "learning_rate": 1.9123230928546776e-05,
+      "loss": 2.6936,
+      "step": 449
+    },
+    {
+      "cosine_similarity": 0.258902690363509,
+      "epoch": 0.4193849021435228,
+      "grad_norm": 2.237307646290344,
+      "learning_rate": 1.9116327235070765e-05,
+      "loss": 2.4418,
+      "step": 450
+    },
+    {
+      "cosine_similarity": 0.16004350253032112,
+      "epoch": 0.42031686859273065,
+      "grad_norm": 1.6234374701771574,
+      "learning_rate": 1.9109423541594754e-05,
+      "loss": 2.3924,
+      "step": 451
+    },
+    {
+      "cosine_similarity": 0.001909942669556958,
+      "epoch": 0.4212488350419385,
+      "grad_norm": 1.7772754512044233,
+      "learning_rate": 1.9102519848118746e-05,
+      "loss": 2.4705,
+      "step": 452
+    },
+    {
+      "cosine_similarity": 0.013553709892954159,
+      "epoch": 0.42218080149114634,
+      "grad_norm": 1.6707632757599227,
+      "learning_rate": 1.9095616154642735e-05,
+      "loss": 2.6611,
+      "step": 453
+    },
+    {
+      "cosine_similarity": -0.09610195559544102,
+      "epoch": 0.42311276794035413,
+      "grad_norm": 1.3590531171249627,
+      "learning_rate": 1.9088712461166724e-05,
+      "loss": 2.1749,
+      "step": 454
+    },
+    {
+      "cosine_similarity": 0.23384684642911177,
+      "epoch": 0.424044734389562,
+      "grad_norm": 1.4784268609127824,
+      "learning_rate": 1.9081808767690717e-05,
+      "loss": 2.3538,
+      "step": 455
+    },
+    {
+      "cosine_similarity": 0.34835989823532915,
+      "epoch": 0.4249767008387698,
+      "grad_norm": 1.5177834511486503,
+      "learning_rate": 1.9074905074214706e-05,
+      "loss": 2.384,
+      "step": 456
+    },
+    {
+      "cosine_similarity": 0.05580890486189941,
+      "epoch": 0.42590866728797766,
+      "grad_norm": 1.6705169348457243,
+      "learning_rate": 1.9068001380738698e-05,
+      "loss": 2.4117,
+      "step": 457
+    },
+    {
+      "cosine_similarity": 0.17903429281637862,
+      "epoch": 0.42684063373718545,
+      "grad_norm": 1.8220381631180265,
+      "learning_rate": 1.9061097687262687e-05,
+      "loss": 2.2104,
+      "step": 458
+    },
+    {
+      "cosine_similarity": 0.3351878279268973,
+      "epoch": 0.4277726001863933,
+      "grad_norm": 2.5043709117031288,
+      "learning_rate": 1.905419399378668e-05,
+      "loss": 2.462,
+      "step": 459
+    },
+    {
+      "cosine_similarity": 0.2885152877121994,
+      "epoch": 0.42870456663560114,
+      "grad_norm": 2.0953472810656235,
+      "learning_rate": 1.904729030031067e-05,
+      "loss": 2.3995,
+      "step": 460
+    },
+    {
+      "cosine_similarity": 0.09504641973421787,
+      "epoch": 0.4296365330848089,
+      "grad_norm": 1.7261897420849432,
+      "learning_rate": 1.9040386606834658e-05,
+      "loss": 2.2554,
+      "step": 461
+    },
+    {
+      "cosine_similarity": -0.01351431220933538,
+      "epoch": 0.43056849953401677,
+      "grad_norm": 1.8062946977104863,
+      "learning_rate": 1.9033482913358647e-05,
+      "loss": 2.3769,
+      "step": 462
+    },
+    {
+      "cosine_similarity": -0.032278629243062865,
+      "epoch": 0.4315004659832246,
+      "grad_norm": 1.2870700129665027,
+      "learning_rate": 1.902657921988264e-05,
+      "loss": 1.958,
+      "step": 463
+    },
+    {
+      "cosine_similarity": 0.5105897483335852,
+      "epoch": 0.43243243243243246,
+      "grad_norm": 1.7503175709799492,
+      "learning_rate": 1.9019675526406628e-05,
+      "loss": 2.4377,
+      "step": 464
+    },
+    {
+      "cosine_similarity": 0.7053411070071873,
+      "epoch": 0.43336439888164024,
+      "grad_norm": 1.5786095455669757,
+      "learning_rate": 1.901277183293062e-05,
+      "loss": 2.613,
+      "step": 465
+    },
+    {
+      "cosine_similarity": 0.2503245685658635,
+      "epoch": 0.4342963653308481,
+      "grad_norm": 1.6597548745282962,
+      "learning_rate": 1.900586813945461e-05,
+      "loss": 2.155,
+      "step": 466
+    },
+    {
+      "cosine_similarity": -0.12271201173172894,
+      "epoch": 0.43522833178005593,
+      "grad_norm": 2.164754805370105,
+      "learning_rate": 1.8998964445978602e-05,
+      "loss": 2.2143,
+      "step": 467
+    },
+    {
+      "cosine_similarity": 0.3188094706645858,
+      "epoch": 0.4361602982292637,
+      "grad_norm": 3.0976910324671776,
+      "learning_rate": 1.899206075250259e-05,
+      "loss": 2.5596,
+      "step": 468
+    },
+    {
+      "cosine_similarity": 0.4545027205761624,
+      "epoch": 0.43709226467847156,
+      "grad_norm": 1.4422371180939322,
+      "learning_rate": 1.898515705902658e-05,
+      "loss": 2.2682,
+      "step": 469
+    },
+    {
+      "cosine_similarity": 0.17021360701432267,
+      "epoch": 0.4380242311276794,
+      "grad_norm": 1.6193556455179252,
+      "learning_rate": 1.8978253365550572e-05,
+      "loss": 2.4523,
+      "step": 470
+    },
+    {
+      "cosine_similarity": 0.44261642367353243,
+      "epoch": 0.43895619757688725,
+      "grad_norm": 1.967999304157631,
+      "learning_rate": 1.897134967207456e-05,
+      "loss": 2.6164,
+      "step": 471
+    },
+    {
+      "cosine_similarity": 0.4349607679275104,
+      "epoch": 0.43988816402609504,
+      "grad_norm": 1.5414800599935854,
+      "learning_rate": 1.896444597859855e-05,
+      "loss": 2.4529,
+      "step": 472
+    },
+    {
+      "cosine_similarity": 0.08535641893987479,
+      "epoch": 0.4408201304753029,
+      "grad_norm": 1.5443693422850655,
+      "learning_rate": 1.895754228512254e-05,
+      "loss": 2.659,
+      "step": 473
+    },
+    {
+      "cosine_similarity": -0.09702549047691951,
+      "epoch": 0.44175209692451073,
+      "grad_norm": 1.3522774328210292,
+      "learning_rate": 1.8950638591646532e-05,
+      "loss": 1.8775,
+      "step": 474
+    },
+    {
+      "cosine_similarity": 0.23354702638989633,
+      "epoch": 0.4426840633737186,
+      "grad_norm": 1.7793866281721817,
+      "learning_rate": 1.894373489817052e-05,
+      "loss": 2.6487,
+      "step": 475
+    },
+    {
+      "cosine_similarity": 0.24682544224790012,
+      "epoch": 0.44361602982292636,
+      "grad_norm": 1.5266684819469696,
+      "learning_rate": 1.8936831204694513e-05,
+      "loss": 1.9687,
+      "step": 476
+    },
+    {
+      "cosine_similarity": 0.15858732359013855,
+      "epoch": 0.4445479962721342,
+      "grad_norm": 1.3631581517750824,
+      "learning_rate": 1.8929927511218506e-05,
+      "loss": 2.2625,
+      "step": 477
+    },
+    {
+      "cosine_similarity": 0.0698920733645452,
+      "epoch": 0.44547996272134205,
+      "grad_norm": 2.307309439372145,
+      "learning_rate": 1.8923023817742495e-05,
+      "loss": 2.3195,
+      "step": 478
+    },
+    {
+      "cosine_similarity": 0.2964822007266134,
+      "epoch": 0.44641192917054984,
+      "grad_norm": 1.537303807812485,
+      "learning_rate": 1.8916120124266484e-05,
+      "loss": 2.2077,
+      "step": 479
+    },
+    {
+      "cosine_similarity": 0.1969617306866071,
+      "epoch": 0.4473438956197577,
+      "grad_norm": 1.9108759254612193,
+      "learning_rate": 1.8909216430790473e-05,
+      "loss": 2.0776,
+      "step": 480
+    },
+    {
+      "cosine_similarity": 0.06160409266288757,
+      "epoch": 0.4482758620689655,
+      "grad_norm": 1.363641338015681,
+      "learning_rate": 1.8902312737314465e-05,
+      "loss": 2.039,
+      "step": 481
+    },
+    {
+      "cosine_similarity": 0.02397466473611598,
+      "epoch": 0.44920782851817337,
+      "grad_norm": 1.6566176723247086,
+      "learning_rate": 1.8895409043838454e-05,
+      "loss": 2.0478,
+      "step": 482
+    },
+    {
+      "cosine_similarity": 0.013230848682469237,
+      "epoch": 0.45013979496738116,
+      "grad_norm": 1.7377481410157034,
+      "learning_rate": 1.8888505350362443e-05,
+      "loss": 2.7105,
+      "step": 483
+    },
+    {
+      "cosine_similarity": -0.0943785504938457,
+      "epoch": 0.451071761416589,
+      "grad_norm": 1.7904258920222162,
+      "learning_rate": 1.8881601656886436e-05,
+      "loss": 2.1859,
+      "step": 484
+    },
+    {
+      "cosine_similarity": 0.11253577994508548,
+      "epoch": 0.45200372786579684,
+      "grad_norm": 1.6109675859693493,
+      "learning_rate": 1.8874697963410425e-05,
+      "loss": 2.252,
+      "step": 485
+    },
+    {
+      "cosine_similarity": 0.11321716590504212,
+      "epoch": 0.45293569431500463,
+      "grad_norm": 1.5096856007692818,
+      "learning_rate": 1.8867794269934417e-05,
+      "loss": 2.3237,
+      "step": 486
+    },
+    {
+      "cosine_similarity": 0.47456381247147367,
+      "epoch": 0.4538676607642125,
+      "grad_norm": 1.6096952395292248,
+      "learning_rate": 1.8860890576458406e-05,
+      "loss": 2.3883,
+      "step": 487
+    },
+    {
+      "cosine_similarity": 0.004957588130464206,
+      "epoch": 0.4547996272134203,
+      "grad_norm": 1.4646953096800839,
+      "learning_rate": 1.88539868829824e-05,
+      "loss": 2.3162,
+      "step": 488
+    },
+    {
+      "cosine_similarity": 0.24544432663029317,
+      "epoch": 0.45573159366262816,
+      "grad_norm": 1.5042014758089763,
+      "learning_rate": 1.8847083189506388e-05,
+      "loss": 1.7876,
+      "step": 489
+    },
+    {
+      "cosine_similarity": 0.06790727627490112,
+      "epoch": 0.45666356011183595,
+      "grad_norm": 1.9028084494703996,
+      "learning_rate": 1.8840179496030377e-05,
+      "loss": 2.3371,
+      "step": 490
+    },
+    {
+      "cosine_similarity": -0.1580613149445917,
+      "epoch": 0.4575955265610438,
+      "grad_norm": 1.5943644570878184,
+      "learning_rate": 1.8833275802554366e-05,
+      "loss": 2.204,
+      "step": 491
+    },
+    {
+      "cosine_similarity": 0.33394896449034644,
+      "epoch": 0.45852749301025164,
+      "grad_norm": 1.8545406371961963,
+      "learning_rate": 1.882637210907836e-05,
+      "loss": 2.5318,
+      "step": 492
+    },
+    {
+      "cosine_similarity": 0.4273283369062749,
+      "epoch": 0.4594594594594595,
+      "grad_norm": 1.6451496716654457,
+      "learning_rate": 1.8819468415602347e-05,
+      "loss": 2.5685,
+      "step": 493
+    },
+    {
+      "cosine_similarity": 0.1317603489669688,
+      "epoch": 0.4603914259086673,
+      "grad_norm": 1.561877384848377,
+      "learning_rate": 1.881256472212634e-05,
+      "loss": 2.5266,
+      "step": 494
+    },
+    {
+      "cosine_similarity": 0.20133326989662825,
+      "epoch": 0.4613233923578751,
+      "grad_norm": 1.6170298785184387,
+      "learning_rate": 1.880566102865033e-05,
+      "loss": 2.4136,
+      "step": 495
+    },
+    {
+      "cosine_similarity": -0.1523711509717022,
+      "epoch": 0.46225535880708296,
+      "grad_norm": 1.6918265873987866,
+      "learning_rate": 1.879875733517432e-05,
+      "loss": 2.3492,
+      "step": 496
+    },
+    {
+      "cosine_similarity": 0.43706284804430146,
+      "epoch": 0.46318732525629075,
+      "grad_norm": 1.739613068573366,
+      "learning_rate": 1.879185364169831e-05,
+      "loss": 2.482,
+      "step": 497
+    },
+    {
+      "cosine_similarity": 0.39865517757459307,
+      "epoch": 0.4641192917054986,
+      "grad_norm": 1.5263722465853993,
+      "learning_rate": 1.87849499482223e-05,
+      "loss": 1.672,
+      "step": 498
+    },
+    {
+      "cosine_similarity": 0.4702668844514776,
+      "epoch": 0.46505125815470644,
+      "grad_norm": 1.4749849865121285,
+      "learning_rate": 1.8778046254746292e-05,
+      "loss": 2.3267,
+      "step": 499
+    },
+    {
+      "cosine_similarity": 0.02815709783263527,
+      "epoch": 0.4659832246039143,
+      "grad_norm": 1.9629201340149562,
+      "learning_rate": 1.877114256127028e-05,
+      "loss": 2.3557,
+      "step": 500
+    },
+    {
+      "cosine_similarity": 0.3837201285873125,
+      "epoch": 0.46691519105312207,
+      "grad_norm": 1.9199550428331142,
+      "learning_rate": 1.876423886779427e-05,
+      "loss": 2.4766,
+      "step": 501
+    },
+    {
+      "cosine_similarity": 0.4012428049294842,
+      "epoch": 0.4678471575023299,
+      "grad_norm": 1.5974828389291793,
+      "learning_rate": 1.8757335174318262e-05,
+      "loss": 2.5615,
+      "step": 502
+    },
+    {
+      "cosine_similarity": 0.5027174716098477,
+      "epoch": 0.46877912395153776,
+      "grad_norm": 2.490802432181201,
+      "learning_rate": 1.875043148084225e-05,
+      "loss": 2.2232,
+      "step": 503
+    },
+    {
+      "cosine_similarity": 0.2685332116216082,
+      "epoch": 0.46971109040074555,
+      "grad_norm": 1.5169666739584526,
+      "learning_rate": 1.8743527787366244e-05,
+      "loss": 2.2072,
+      "step": 504
+    },
+    {
+      "cosine_similarity": 0.22058828500718913,
+      "epoch": 0.4706430568499534,
+      "grad_norm": 1.5847917002114478,
+      "learning_rate": 1.8736624093890233e-05,
+      "loss": 2.0824,
+      "step": 505
+    },
+    {
+      "cosine_similarity": 0.26361656638550535,
+      "epoch": 0.47157502329916123,
+      "grad_norm": 1.8294547131958747,
+      "learning_rate": 1.8729720400414225e-05,
+      "loss": 2.0247,
+      "step": 506
+    },
+    {
+      "cosine_similarity": 0.31903303656324217,
+      "epoch": 0.4725069897483691,
+      "grad_norm": 2.025677546749489,
+      "learning_rate": 1.8722816706938214e-05,
+      "loss": 2.5117,
+      "step": 507
+    },
+    {
+      "cosine_similarity": 0.04553384541625001,
+      "epoch": 0.47343895619757687,
+      "grad_norm": 1.4859082352932536,
+      "learning_rate": 1.8715913013462203e-05,
+      "loss": 2.3674,
+      "step": 508
+    },
+    {
+      "cosine_similarity": 0.12051140452687495,
+      "epoch": 0.4743709226467847,
+      "grad_norm": 1.4922738402095508,
+      "learning_rate": 1.8709009319986192e-05,
+      "loss": 2.2562,
+      "step": 509
+    },
+    {
+      "cosine_similarity": 0.28928624434374267,
+      "epoch": 0.47530288909599255,
+      "grad_norm": 1.4275229588865208,
+      "learning_rate": 1.8702105626510185e-05,
+      "loss": 2.0588,
+      "step": 510
+    },
+    {
+      "cosine_similarity": 0.36863084875274593,
+      "epoch": 0.4762348555452004,
+      "grad_norm": 1.6517586902101333,
+      "learning_rate": 1.8695201933034174e-05,
+      "loss": 2.0749,
+      "step": 511
+    },
+    {
+      "cosine_similarity": 0.008017876949513274,
+      "epoch": 0.4771668219944082,
+      "grad_norm": 1.5061039645493342,
+      "learning_rate": 1.8688298239558166e-05,
+      "loss": 2.1058,
+      "step": 512
+    },
+    {
+      "cosine_similarity": 0.30873661512198025,
+      "epoch": 0.47809878844361603,
+      "grad_norm": 1.5130569548395707,
+      "learning_rate": 1.8681394546082155e-05,
+      "loss": 2.0733,
+      "step": 513
+    },
+    {
+      "cosine_similarity": 0.17960605537385002,
+      "epoch": 0.4790307548928239,
+      "grad_norm": 1.397598240426557,
+      "learning_rate": 1.8674490852606148e-05,
+      "loss": 2.2269,
+      "step": 514
+    },
+    {
+      "cosine_similarity": -0.07195443822087576,
+      "epoch": 0.47996272134203166,
+      "grad_norm": 1.5699848800155156,
+      "learning_rate": 1.8667587159130137e-05,
+      "loss": 2.2551,
+      "step": 515
+    },
+    {
+      "cosine_similarity": 0.04273801644742807,
+      "epoch": 0.4808946877912395,
+      "grad_norm": 1.5136946448211104,
+      "learning_rate": 1.8660683465654126e-05,
+      "loss": 2.1906,
+      "step": 516
+    },
+    {
+      "cosine_similarity": 0.06766738874776812,
+      "epoch": 0.48182665424044735,
+      "grad_norm": 1.7024420797090165,
+      "learning_rate": 1.8653779772178118e-05,
+      "loss": 2.275,
+      "step": 517
+    },
+    {
+      "cosine_similarity": 0.028576495165730798,
+      "epoch": 0.4827586206896552,
+      "grad_norm": 1.6325595729575977,
+      "learning_rate": 1.8646876078702107e-05,
+      "loss": 2.1584,
+      "step": 518
+    },
+    {
+      "cosine_similarity": 0.06606098452260592,
+      "epoch": 0.483690587138863,
+      "grad_norm": 1.4527779115768524,
+      "learning_rate": 1.8639972385226096e-05,
+      "loss": 2.3565,
+      "step": 519
+    },
+    {
+      "cosine_similarity": -0.1274937283154779,
+      "epoch": 0.4846225535880708,
+      "grad_norm": 2.8963839328898118,
+      "learning_rate": 1.8633068691750085e-05,
+      "loss": 2.3905,
+      "step": 520
+    },
+    {
+      "cosine_similarity": 0.525719894514184,
+      "epoch": 0.48555452003727867,
+      "grad_norm": 1.5091547550707998,
+      "learning_rate": 1.8626164998274078e-05,
+      "loss": 2.1079,
+      "step": 521
+    },
+    {
+      "cosine_similarity": 0.3668568908830949,
+      "epoch": 0.4864864864864865,
+      "grad_norm": 1.3640380240066856,
+      "learning_rate": 1.861926130479807e-05,
+      "loss": 2.4602,
+      "step": 522
+    },
+    {
+      "cosine_similarity": 0.09876032511362984,
+      "epoch": 0.4874184529356943,
+      "grad_norm": 1.5530542153919484,
+      "learning_rate": 1.861235761132206e-05,
+      "loss": 2.2886,
+      "step": 523
+    },
+    {
+      "cosine_similarity": 0.34211720823875463,
+      "epoch": 0.48835041938490215,
+      "grad_norm": 1.547848206613281,
+      "learning_rate": 1.860545391784605e-05,
+      "loss": 2.2394,
+      "step": 524
+    },
+    {
+      "cosine_similarity": 0.44998707573762925,
+      "epoch": 0.48928238583411,
+      "grad_norm": 1.678973366163009,
+      "learning_rate": 1.859855022437004e-05,
+      "loss": 2.3646,
+      "step": 525
+    },
+    {
+      "cosine_similarity": 0.04153809722610209,
+      "epoch": 0.4902143522833178,
+      "grad_norm": 2.4099749994380586,
+      "learning_rate": 1.859164653089403e-05,
+      "loss": 2.2191,
+      "step": 526
+    },
+    {
+      "cosine_similarity": 0.03938221400488619,
+      "epoch": 0.4911463187325256,
+      "grad_norm": 1.5856175644542074,
+      "learning_rate": 1.858474283741802e-05,
+      "loss": 2.5646,
+      "step": 527
+    },
+    {
+      "cosine_similarity": 0.10680148827947189,
+      "epoch": 0.49207828518173347,
+      "grad_norm": 1.795080550025169,
+      "learning_rate": 1.857783914394201e-05,
+      "loss": 2.0756,
+      "step": 528
+    },
+    {
+      "cosine_similarity": 0.173411030318332,
+      "epoch": 0.4930102516309413,
+      "grad_norm": 1.7602580266256918,
+      "learning_rate": 1.8570935450466e-05,
+      "loss": 2.5383,
+      "step": 529
+    },
+    {
+      "cosine_similarity": -0.01809473715843265,
+      "epoch": 0.4939422180801491,
+      "grad_norm": 1.777431096324314,
+      "learning_rate": 1.856403175698999e-05,
+      "loss": 2.5549,
+      "step": 530
+    },
+    {
+      "cosine_similarity": 0.0027290580415205035,
+      "epoch": 0.49487418452935694,
+      "grad_norm": 2.1117631181689998,
+      "learning_rate": 1.8557128063513982e-05,
+      "loss": 2.4267,
+      "step": 531
+    },
+    {
+      "cosine_similarity": 0.10285721856424829,
+      "epoch": 0.4958061509785648,
+      "grad_norm": 1.492471008790975,
+      "learning_rate": 1.8550224370037974e-05,
+      "loss": 2.064,
+      "step": 532
+    },
+    {
+      "cosine_similarity": 0.34276199237830146,
+      "epoch": 0.4967381174277726,
+      "grad_norm": 1.726933391701041,
+      "learning_rate": 1.8543320676561963e-05,
+      "loss": 2.1291,
+      "step": 533
+    },
+    {
+      "cosine_similarity": 0.019433729854791165,
+      "epoch": 0.4976700838769804,
+      "grad_norm": 1.3592985920603635,
+      "learning_rate": 1.8536416983085952e-05,
+      "loss": 2.5504,
+      "step": 534
+    },
+    {
+      "cosine_similarity": 0.25700490376204344,
+      "epoch": 0.49860205032618826,
+      "grad_norm": 2.490826463065261,
+      "learning_rate": 1.8529513289609945e-05,
+      "loss": 1.9071,
+      "step": 535
+    },
+    {
+      "cosine_similarity": 0.023011333758089295,
+      "epoch": 0.4995340167753961,
+      "grad_norm": 1.6141786624000816,
+      "learning_rate": 1.8522609596133934e-05,
+      "loss": 2.1539,
+      "step": 536
+    },
+    {
+      "cosine_similarity": -0.01518612530792854,
+      "epoch": 0.5004659832246039,
+      "grad_norm": 1.4497051505170755,
+      "learning_rate": 1.8515705902657923e-05,
+      "loss": 2.0919,
+      "step": 537
+    },
+    {
+      "cosine_similarity": 0.378581125365567,
+      "epoch": 0.5013979496738118,
+      "grad_norm": 1.7677221300657004,
+      "learning_rate": 1.8508802209181912e-05,
+      "loss": 1.9936,
+      "step": 538
+    },
+    {
+      "cosine_similarity": 0.2652552628583175,
+      "epoch": 0.5023299161230196,
+      "grad_norm": 1.573312165452935,
+      "learning_rate": 1.8501898515705904e-05,
+      "loss": 2.2755,
+      "step": 539
+    },
+    {
+      "cosine_similarity": 0.47374752299504325,
+      "epoch": 0.5032618825722274,
+      "grad_norm": 1.5743706086089233,
+      "learning_rate": 1.8494994822229893e-05,
+      "loss": 2.4809,
+      "step": 540
+    },
+    {
+      "cosine_similarity": -0.40825858595869585,
+      "epoch": 0.5041938490214353,
+      "grad_norm": 1.7158700412506644,
+      "learning_rate": 1.8488091128753886e-05,
+      "loss": 2.4032,
+      "step": 541
+    },
+    {
+      "cosine_similarity": 0.3992474501601676,
+      "epoch": 0.5051258154706431,
+      "grad_norm": 1.8643727752694237,
+      "learning_rate": 1.8481187435277875e-05,
+      "loss": 2.3948,
+      "step": 542
+    },
+    {
+      "cosine_similarity": 0.2479413935906308,
+      "epoch": 0.5060577819198508,
+      "grad_norm": 2.1103368706614534,
+      "learning_rate": 1.8474283741801867e-05,
+      "loss": 2.3405,
+      "step": 543
+    },
+    {
+      "cosine_similarity": 0.2119012663397894,
+      "epoch": 0.5069897483690587,
+      "grad_norm": 1.4366094937184466,
+      "learning_rate": 1.8467380048325856e-05,
+      "loss": 2.0952,
+      "step": 544
+    },
+    {
+      "cosine_similarity": 0.08174949719919729,
+      "epoch": 0.5079217148182665,
+      "grad_norm": 1.791158853557985,
+      "learning_rate": 1.8460476354849845e-05,
+      "loss": 2.3088,
+      "step": 545
+    },
+    {
+      "cosine_similarity": 0.14199470234303457,
+      "epoch": 0.5088536812674743,
+      "grad_norm": 2.0070239512247223,
+      "learning_rate": 1.8453572661373838e-05,
+      "loss": 2.8008,
+      "step": 546
+    },
+    {
+      "cosine_similarity": 0.3614149827698578,
+      "epoch": 0.5097856477166822,
+      "grad_norm": 1.6432582565848415,
+      "learning_rate": 1.8446668967897827e-05,
+      "loss": 2.3867,
+      "step": 547
+    },
+    {
+      "cosine_similarity": 0.027950724348131936,
+      "epoch": 0.51071761416589,
+      "grad_norm": 1.391519115184979,
+      "learning_rate": 1.8439765274421816e-05,
+      "loss": 2.1763,
+      "step": 548
+    },
+    {
+      "cosine_similarity": 0.06757023231925204,
+      "epoch": 0.5116495806150979,
+      "grad_norm": 1.589525656858041,
+      "learning_rate": 1.8432861580945808e-05,
+      "loss": 2.7758,
+      "step": 549
+    },
+    {
+      "cosine_similarity": -0.19578959555591038,
+      "epoch": 0.5125815470643057,
+      "grad_norm": 1.7738012591951968,
+      "learning_rate": 1.8425957887469797e-05,
+      "loss": 2.416,
+      "step": 550
+    },
+    {
+      "cosine_similarity": 0.14931776122761103,
+      "epoch": 0.5135135135135135,
+      "grad_norm": 1.6639052297116894,
+      "learning_rate": 1.841905419399379e-05,
+      "loss": 2.3235,
+      "step": 551
+    },
+    {
+      "cosine_similarity": 0.30535626751797873,
+      "epoch": 0.5144454799627214,
+      "grad_norm": 2.274312665924869,
+      "learning_rate": 1.841215050051778e-05,
+      "loss": 2.3838,
+      "step": 552
+    },
+    {
+      "cosine_similarity": 0.37714324107843783,
+      "epoch": 0.5153774464119292,
+      "grad_norm": 1.896624868170278,
+      "learning_rate": 1.8405246807041768e-05,
+      "loss": 2.2392,
+      "step": 553
+    },
+    {
+      "cosine_similarity": 0.09957689045747879,
+      "epoch": 0.516309412861137,
+      "grad_norm": 1.8139338532853564,
+      "learning_rate": 1.839834311356576e-05,
+      "loss": 2.2661,
+      "step": 554
+    },
+    {
+      "cosine_similarity": 0.009439126016240725,
+      "epoch": 0.5172413793103449,
+      "grad_norm": 1.866444956406939,
+      "learning_rate": 1.839143942008975e-05,
+      "loss": 2.4796,
+      "step": 555
+    },
+    {
+      "cosine_similarity": -0.18994832410296608,
+      "epoch": 0.5181733457595527,
+      "grad_norm": 1.6034415479033473,
+      "learning_rate": 1.8384535726613738e-05,
+      "loss": 2.5022,
+      "step": 556
+    },
+    {
+      "cosine_similarity": 0.062000748575051334,
+      "epoch": 0.5191053122087604,
+      "grad_norm": 1.4943670510540883,
+      "learning_rate": 1.837763203313773e-05,
+      "loss": 2.3487,
+      "step": 557
+    },
+    {
+      "cosine_similarity": -0.06370998666371543,
+      "epoch": 0.5200372786579683,
+      "grad_norm": 1.711825609474896,
+      "learning_rate": 1.837072833966172e-05,
+      "loss": 2.3896,
+      "step": 558
+    },
+    {
+      "cosine_similarity": -0.17117902684319675,
+      "epoch": 0.5209692451071761,
+      "grad_norm": 1.5966339829287137,
+      "learning_rate": 1.8363824646185712e-05,
+      "loss": 2.0117,
+      "step": 559
+    },
+    {
+      "cosine_similarity": 0.041389761696114115,
+      "epoch": 0.5219012115563839,
+      "grad_norm": 1.487525784632725,
+      "learning_rate": 1.83569209527097e-05,
+      "loss": 2.0456,
+      "step": 560
+    },
+    {
+      "cosine_similarity": 0.4443462864545553,
+      "epoch": 0.5228331780055918,
+      "grad_norm": 1.5025438584203867,
+      "learning_rate": 1.8350017259233694e-05,
+      "loss": 2.1849,
+      "step": 561
+    },
+    {
+      "cosine_similarity": 0.028829052205361316,
+      "epoch": 0.5237651444547996,
+      "grad_norm": 1.5177542228875873,
+      "learning_rate": 1.8343113565757683e-05,
+      "loss": 1.9915,
+      "step": 562
+    },
+    {
+      "cosine_similarity": 0.4152223464972061,
+      "epoch": 0.5246971109040075,
+      "grad_norm": 1.4270002199937513,
+      "learning_rate": 1.833620987228167e-05,
+      "loss": 2.2566,
+      "step": 563
+    },
+    {
+      "cosine_similarity": 0.13264880072799823,
+      "epoch": 0.5256290773532153,
+      "grad_norm": 1.3267601789935253,
+      "learning_rate": 1.832930617880566e-05,
+      "loss": 1.9133,
+      "step": 564
+    },
+    {
+      "cosine_similarity": 0.30694944490433507,
+      "epoch": 0.5265610438024231,
+      "grad_norm": 1.4185651922192466,
+      "learning_rate": 1.8322402485329653e-05,
+      "loss": 2.0661,
+      "step": 565
+    },
+    {
+      "cosine_similarity": 0.13919445088573199,
+      "epoch": 0.527493010251631,
+      "grad_norm": 1.9899836825535,
+      "learning_rate": 1.8315498791853642e-05,
+      "loss": 2.3225,
+      "step": 566
+    },
+    {
+      "cosine_similarity": 0.09489556041072426,
+      "epoch": 0.5284249767008388,
+      "grad_norm": 1.671840655968411,
+      "learning_rate": 1.830859509837763e-05,
+      "loss": 2.2044,
+      "step": 567
+    },
+    {
+      "cosine_similarity": 0.36524623594327216,
+      "epoch": 0.5293569431500466,
+      "grad_norm": 1.435017181793532,
+      "learning_rate": 1.8301691404901624e-05,
+      "loss": 2.1471,
+      "step": 568
+    },
+    {
+      "cosine_similarity": 0.6379538168333889,
+      "epoch": 0.5302889095992545,
+      "grad_norm": 1.8672470375020935,
+      "learning_rate": 1.8294787711425616e-05,
+      "loss": 2.5014,
+      "step": 569
+    },
+    {
+      "cosine_similarity": 0.17888926678048178,
+      "epoch": 0.5312208760484622,
+      "grad_norm": 1.901936621476807,
+      "learning_rate": 1.8287884017949605e-05,
+      "loss": 2.4217,
+      "step": 570
+    },
+    {
+      "cosine_similarity": 0.39279999049428227,
+      "epoch": 0.53215284249767,
+      "grad_norm": 1.4363933126677908,
+      "learning_rate": 1.8280980324473594e-05,
+      "loss": 1.9325,
+      "step": 571
+    },
+    {
+      "cosine_similarity": 0.6066764733003902,
+      "epoch": 0.5330848089468779,
+      "grad_norm": 1.5824722527044108,
+      "learning_rate": 1.8274076630997587e-05,
+      "loss": 2.3326,
+      "step": 572
+    },
+    {
+      "cosine_similarity": 0.081222990977912,
+      "epoch": 0.5340167753960857,
+      "grad_norm": 1.4506474174701471,
+      "learning_rate": 1.8267172937521576e-05,
+      "loss": 2.5247,
+      "step": 573
+    },
+    {
+      "cosine_similarity": 0.05664170043314926,
+      "epoch": 0.5349487418452936,
+      "grad_norm": 1.4215350495189158,
+      "learning_rate": 1.8260269244045565e-05,
+      "loss": 2.1342,
+      "step": 574
+    },
+    {
+      "cosine_similarity": 0.03284329976992502,
+      "epoch": 0.5358807082945014,
+      "grad_norm": 1.558454647786123,
+      "learning_rate": 1.8253365550569557e-05,
+      "loss": 2.3477,
+      "step": 575
+    },
+    {
+      "cosine_similarity": 0.2747536421785615,
+      "epoch": 0.5368126747437092,
+      "grad_norm": 1.5394176621887679,
+      "learning_rate": 1.8246461857093546e-05,
+      "loss": 2.1294,
+      "step": 576
+    },
+    {
+      "cosine_similarity": 0.49644074865340315,
+      "epoch": 0.5377446411929171,
+      "grad_norm": 1.4391893478083546,
+      "learning_rate": 1.8239558163617535e-05,
+      "loss": 2.2757,
+      "step": 577
+    },
+    {
+      "cosine_similarity": 0.1349390503483804,
+      "epoch": 0.5386766076421249,
+      "grad_norm": 1.7043520393873124,
+      "learning_rate": 1.8232654470141528e-05,
+      "loss": 2.4026,
+      "step": 578
+    },
+    {
+      "cosine_similarity": 0.1918094416925076,
+      "epoch": 0.5396085740913327,
+      "grad_norm": 1.522058294600809,
+      "learning_rate": 1.822575077666552e-05,
+      "loss": 2.3024,
+      "step": 579
+    },
+    {
+      "cosine_similarity": 0.6459425547844191,
+      "epoch": 0.5405405405405406,
+      "grad_norm": 1.75157785351565,
+      "learning_rate": 1.821884708318951e-05,
+      "loss": 2.469,
+      "step": 580
+    },
+    {
+      "cosine_similarity": 0.38265251959359914,
+      "epoch": 0.5414725069897484,
+      "grad_norm": 1.3381394653880025,
+      "learning_rate": 1.8211943389713498e-05,
+      "loss": 2.4717,
+      "step": 581
+    },
+    {
+      "cosine_similarity": -0.015752819299796404,
+      "epoch": 0.5424044734389561,
+      "grad_norm": 1.284395512969127,
+      "learning_rate": 1.8205039696237487e-05,
+      "loss": 2.1725,
+      "step": 582
+    },
+    {
+      "cosine_similarity": 0.1081615745691915,
+      "epoch": 0.543336439888164,
+      "grad_norm": 1.3838362667242134,
+      "learning_rate": 1.819813600276148e-05,
+      "loss": 2.0011,
+      "step": 583
+    },
+    {
+      "cosine_similarity": 0.10668030384837492,
+      "epoch": 0.5442684063373718,
+      "grad_norm": 1.3278170360883623,
+      "learning_rate": 1.819123230928547e-05,
+      "loss": 2.1972,
+      "step": 584
+    },
+    {
+      "cosine_similarity": 0.24000169695139834,
+      "epoch": 0.5452003727865797,
+      "grad_norm": 1.5581123333373765,
+      "learning_rate": 1.8184328615809458e-05,
+      "loss": 2.1833,
+      "step": 585
+    },
+    {
+      "cosine_similarity": 0.23787668585414343,
+      "epoch": 0.5461323392357875,
+      "grad_norm": 1.60403937350136,
+      "learning_rate": 1.817742492233345e-05,
+      "loss": 1.9001,
+      "step": 586
+    },
+    {
+      "cosine_similarity": 0.06275289110967515,
+      "epoch": 0.5470643056849953,
+      "grad_norm": 1.6843520906270397,
+      "learning_rate": 1.817052122885744e-05,
+      "loss": 2.2176,
+      "step": 587
+    },
+    {
+      "cosine_similarity": 0.10222727136546114,
+      "epoch": 0.5479962721342032,
+      "grad_norm": 9.29960975656335,
+      "learning_rate": 1.816361753538143e-05,
+      "loss": 2.5448,
+      "step": 588
+    },
+    {
+      "cosine_similarity": 0.5078540850648455,
+      "epoch": 0.548928238583411,
+      "grad_norm": 1.3744874532258324,
+      "learning_rate": 1.815671384190542e-05,
+      "loss": 2.3523,
+      "step": 589
+    },
+    {
+      "cosine_similarity": 0.17067892896441106,
+      "epoch": 0.5498602050326188,
+      "grad_norm": 1.385513107301942,
+      "learning_rate": 1.8149810148429413e-05,
+      "loss": 2.2291,
+      "step": 590
+    },
+    {
+      "cosine_similarity": 0.42740083868969564,
+      "epoch": 0.5507921714818267,
+      "grad_norm": 1.4684399444724072,
+      "learning_rate": 1.8142906454953402e-05,
+      "loss": 2.5281,
+      "step": 591
+    },
+    {
+      "cosine_similarity": 0.6419959189426638,
+      "epoch": 0.5517241379310345,
+      "grad_norm": 1.3253038523397653,
+      "learning_rate": 1.813600276147739e-05,
+      "loss": 2.0732,
+      "step": 592
+    },
+    {
+      "cosine_similarity": 0.2080536280410808,
+      "epoch": 0.5526561043802423,
+      "grad_norm": 1.6937639910980802,
+      "learning_rate": 1.812909906800138e-05,
+      "loss": 2.2679,
+      "step": 593
+    },
+    {
+      "cosine_similarity": 0.2638215159531632,
+      "epoch": 0.5535880708294502,
+      "grad_norm": 1.5832580815323487,
+      "learning_rate": 1.8122195374525372e-05,
+      "loss": 2.5314,
+      "step": 594
+    },
+    {
+      "cosine_similarity": 0.5204055135038093,
+      "epoch": 0.554520037278658,
+      "grad_norm": 1.4952211628421477,
+      "learning_rate": 1.811529168104936e-05,
+      "loss": 2.1864,
+      "step": 595
+    },
+    {
+      "cosine_similarity": 0.3167140985818974,
+      "epoch": 0.5554520037278659,
+      "grad_norm": 1.6766846515537877,
+      "learning_rate": 1.8108387987573354e-05,
+      "loss": 2.23,
+      "step": 596
+    },
+    {
+      "cosine_similarity": 0.5869245637363544,
+      "epoch": 0.5563839701770736,
+      "grad_norm": 1.7061405272620023,
+      "learning_rate": 1.8101484294097343e-05,
+      "loss": 2.4056,
+      "step": 597
+    },
+    {
+      "cosine_similarity": 0.14668786116679752,
+      "epoch": 0.5573159366262814,
+      "grad_norm": 1.421500985773056,
+      "learning_rate": 1.8094580600621335e-05,
+      "loss": 2.1626,
+      "step": 598
+    },
+    {
+      "cosine_similarity": -0.04401393217295715,
+      "epoch": 0.5582479030754893,
+      "grad_norm": 1.5487149468277344,
+      "learning_rate": 1.8087676907145324e-05,
+      "loss": 2.5311,
+      "step": 599
+    },
+    {
+      "cosine_similarity": 0.1182013688362847,
+      "epoch": 0.5591798695246971,
+      "grad_norm": 1.4215262244942672,
+      "learning_rate": 1.8080773213669313e-05,
+      "loss": 2.3267,
+      "step": 600
+    },
+    {
+      "cosine_similarity": 0.23121409457278708,
+      "epoch": 0.5601118359739049,
+      "grad_norm": 1.9793208018823212,
+      "learning_rate": 1.8073869520193306e-05,
+      "loss": 2.1786,
+      "step": 601
+    },
+    {
+      "cosine_similarity": 0.04241595490945055,
+      "epoch": 0.5610438024231128,
+      "grad_norm": 1.119877667125998,
+      "learning_rate": 1.8066965826717295e-05,
+      "loss": 1.9447,
+      "step": 602
+    },
+    {
+      "cosine_similarity": 0.25093958951314105,
+      "epoch": 0.5619757688723206,
+      "grad_norm": 1.6916315783829228,
+      "learning_rate": 1.8060062133241284e-05,
+      "loss": 2.1343,
+      "step": 603
+    },
+    {
+      "cosine_similarity": 0.19475487371639394,
+      "epoch": 0.5629077353215284,
+      "grad_norm": 2.291761568614523,
+      "learning_rate": 1.8053158439765276e-05,
+      "loss": 2.5114,
+      "step": 604
+    },
+    {
+      "cosine_similarity": 0.3350525019470858,
+      "epoch": 0.5638397017707363,
+      "grad_norm": 1.527905060132524,
+      "learning_rate": 1.8046254746289265e-05,
+      "loss": 2.3623,
+      "step": 605
+    },
+    {
+      "cosine_similarity": 0.3433063823438943,
+      "epoch": 0.5647716682199441,
+      "grad_norm": 1.6249397429042538,
+      "learning_rate": 1.8039351052813258e-05,
+      "loss": 2.1812,
+      "step": 606
+    },
+    {
+      "cosine_similarity": 0.46586317292750073,
+      "epoch": 0.5657036346691519,
+      "grad_norm": 1.5142587136764594,
+      "learning_rate": 1.8032447359337247e-05,
+      "loss": 2.3985,
+      "step": 607
+    },
+    {
+      "cosine_similarity": 0.050655732532233946,
+      "epoch": 0.5666356011183598,
+      "grad_norm": 1.7310949452308582,
+      "learning_rate": 1.802554366586124e-05,
+      "loss": 2.1117,
+      "step": 608
+    },
+    {
+      "cosine_similarity": 0.2908794288638867,
+      "epoch": 0.5675675675675675,
+      "grad_norm": 1.757557524235986,
+      "learning_rate": 1.801863997238523e-05,
+      "loss": 2.2638,
+      "step": 609
+    },
+    {
+      "cosine_similarity": 0.30834215762539124,
+      "epoch": 0.5684995340167754,
+      "grad_norm": 2.0795080639967227,
+      "learning_rate": 1.8011736278909217e-05,
+      "loss": 2.1512,
+      "step": 610
+    },
+    {
+      "cosine_similarity": 0.16086225744209348,
+      "epoch": 0.5694315004659832,
+      "grad_norm": 1.5722081745938274,
+      "learning_rate": 1.8004832585433206e-05,
+      "loss": 2.3759,
+      "step": 611
+    },
+    {
+      "cosine_similarity": 0.46546663566720275,
+      "epoch": 0.570363466915191,
+      "grad_norm": 1.6970784679750548,
+      "learning_rate": 1.79979288919572e-05,
+      "loss": 2.2583,
+      "step": 612
+    },
+    {
+      "cosine_similarity": 0.2928617169139129,
+      "epoch": 0.5712954333643989,
+      "grad_norm": 1.7770649570435417,
+      "learning_rate": 1.7991025198481188e-05,
+      "loss": 2.3304,
+      "step": 613
+    },
+    {
+      "cosine_similarity": 0.246595969767217,
+      "epoch": 0.5722273998136067,
+      "grad_norm": 1.508847417254712,
+      "learning_rate": 1.798412150500518e-05,
+      "loss": 2.4042,
+      "step": 614
+    },
+    {
+      "cosine_similarity": 0.21570896669308062,
+      "epoch": 0.5731593662628145,
+      "grad_norm": 1.3859085992132538,
+      "learning_rate": 1.797721781152917e-05,
+      "loss": 2.0054,
+      "step": 615
+    },
+    {
+      "cosine_similarity": 0.3371829008282903,
+      "epoch": 0.5740913327120224,
+      "grad_norm": 1.6475522877880227,
+      "learning_rate": 1.7970314118053162e-05,
+      "loss": 2.4938,
+      "step": 616
+    },
+    {
+      "cosine_similarity": 0.6206190332694745,
+      "epoch": 0.5750232991612302,
+      "grad_norm": 1.5496408295884303,
+      "learning_rate": 1.796341042457715e-05,
+      "loss": 2.4543,
+      "step": 617
+    },
+    {
+      "cosine_similarity": 0.43967271440797073,
+      "epoch": 0.575955265610438,
+      "grad_norm": 1.436328610112255,
+      "learning_rate": 1.795650673110114e-05,
+      "loss": 2.3754,
+      "step": 618
+    },
+    {
+      "cosine_similarity": 0.4758728329542979,
+      "epoch": 0.5768872320596459,
+      "grad_norm": 1.5721452869337906,
+      "learning_rate": 1.7949603037625132e-05,
+      "loss": 2.1273,
+      "step": 619
+    },
+    {
+      "cosine_similarity": 0.005853049746501301,
+      "epoch": 0.5778191985088537,
+      "grad_norm": 1.9836331381186756,
+      "learning_rate": 1.794269934414912e-05,
+      "loss": 2.0923,
+      "step": 620
+    },
+    {
+      "cosine_similarity": 0.2683336993020741,
+      "epoch": 0.5787511649580616,
+      "grad_norm": 1.1946577220734367,
+      "learning_rate": 1.793579565067311e-05,
+      "loss": 1.9509,
+      "step": 621
+    },
+    {
+      "cosine_similarity": 0.4960244736532969,
+      "epoch": 0.5796831314072693,
+      "grad_norm": 1.306986177627536,
+      "learning_rate": 1.79288919571971e-05,
+      "loss": 2.363,
+      "step": 622
+    },
+    {
+      "cosine_similarity": 0.10928669950997154,
+      "epoch": 0.5806150978564771,
+      "grad_norm": 1.4807917081500288,
+      "learning_rate": 1.7921988263721092e-05,
+      "loss": 1.9138,
+      "step": 623
+    },
+    {
+      "cosine_similarity": -0.030127913976738925,
+      "epoch": 0.581547064305685,
+      "grad_norm": 2.3301268520500185,
+      "learning_rate": 1.7915084570245084e-05,
+      "loss": 2.353,
+      "step": 624
+    },
+    {
+      "cosine_similarity": 0.1984751437284382,
+      "epoch": 0.5824790307548928,
+      "grad_norm": 1.4106249901529944,
+      "learning_rate": 1.7908180876769073e-05,
+      "loss": 2.4325,
+      "step": 625
+    },
+    {
+      "cosine_similarity": -0.0957274496245627,
+      "epoch": 0.5834109972041006,
+      "grad_norm": 1.8231651844783685,
+      "learning_rate": 1.7901277183293066e-05,
+      "loss": 2.5265,
+      "step": 626
+    },
+    {
+      "cosine_similarity": 0.3008922262749748,
+      "epoch": 0.5843429636533085,
+      "grad_norm": 1.498835807308005,
+      "learning_rate": 1.7894373489817055e-05,
+      "loss": 2.2265,
+      "step": 627
+    },
+    {
+      "cosine_similarity": 0.08037614295162522,
+      "epoch": 0.5852749301025163,
+      "grad_norm": 1.5450857715360873,
+      "learning_rate": 1.7887469796341044e-05,
+      "loss": 2.2961,
+      "step": 628
+    },
+    {
+      "cosine_similarity": 0.028179057513112536,
+      "epoch": 0.5862068965517241,
+      "grad_norm": 1.6816668788711702,
+      "learning_rate": 1.7880566102865033e-05,
+      "loss": 2.255,
+      "step": 629
+    },
+    {
+      "cosine_similarity": 0.34365842683092307,
+      "epoch": 0.587138863000932,
+      "grad_norm": 2.40464055786463,
+      "learning_rate": 1.7873662409389025e-05,
+      "loss": 2.3541,
+      "step": 630
+    },
+    {
+      "cosine_similarity": 0.26986296370171514,
+      "epoch": 0.5880708294501398,
+      "grad_norm": 1.5976342616825931,
+      "learning_rate": 1.7866758715913014e-05,
+      "loss": 2.3125,
+      "step": 631
+    },
+    {
+      "cosine_similarity": -0.08927579275560223,
+      "epoch": 0.5890027958993477,
+      "grad_norm": 1.3255268326014484,
+      "learning_rate": 1.7859855022437003e-05,
+      "loss": 2.1333,
+      "step": 632
+    },
+    {
+      "cosine_similarity": 0.04988816200268718,
+      "epoch": 0.5899347623485555,
+      "grad_norm": 1.9769022366681437,
+      "learning_rate": 1.7852951328960996e-05,
+      "loss": 2.3893,
+      "step": 633
+    },
+    {
+      "cosine_similarity": 0.27225560078918687,
+      "epoch": 0.5908667287977633,
+      "grad_norm": 1.3882996821806006,
+      "learning_rate": 1.7846047635484988e-05,
+      "loss": 2.2665,
+      "step": 634
+    },
+    {
+      "cosine_similarity": 0.33197153989471523,
+      "epoch": 0.5917986952469712,
+      "grad_norm": 1.4645061883980406,
+      "learning_rate": 1.7839143942008977e-05,
+      "loss": 2.3521,
+      "step": 635
+    },
+    {
+      "cosine_similarity": 0.25041618948454897,
+      "epoch": 0.5927306616961789,
+      "grad_norm": 1.5223042904387163,
+      "learning_rate": 1.7832240248532966e-05,
+      "loss": 2.1634,
+      "step": 636
+    },
+    {
+      "cosine_similarity": 0.13036792373733547,
+      "epoch": 0.5936626281453867,
+      "grad_norm": 1.4533767561603526,
+      "learning_rate": 1.782533655505696e-05,
+      "loss": 2.4528,
+      "step": 637
+    },
+    {
+      "cosine_similarity": 0.3215504664949043,
+      "epoch": 0.5945945945945946,
+      "grad_norm": 2.266246555305096,
+      "learning_rate": 1.7818432861580948e-05,
+      "loss": 2.3711,
+      "step": 638
+    },
+    {
+      "cosine_similarity": 0.1777019631019742,
+      "epoch": 0.5955265610438024,
+      "grad_norm": 1.5476156888298622,
+      "learning_rate": 1.7811529168104937e-05,
+      "loss": 2.3519,
+      "step": 639
+    },
+    {
+      "cosine_similarity": 0.3910851578204587,
+      "epoch": 0.5964585274930102,
+      "grad_norm": 2.2807612977123983,
+      "learning_rate": 1.7804625474628926e-05,
+      "loss": 2.1176,
+      "step": 640
+    },
+    {
+      "cosine_similarity": 0.29509476797475137,
+      "epoch": 0.5973904939422181,
+      "grad_norm": 1.66486481800934,
+      "learning_rate": 1.7797721781152918e-05,
+      "loss": 2.1555,
+      "step": 641
+    },
+    {
+      "cosine_similarity": 0.2554017200694273,
+      "epoch": 0.5983224603914259,
+      "grad_norm": 1.2421962636534154,
+      "learning_rate": 1.7790818087676907e-05,
+      "loss": 2.1629,
+      "step": 642
+    },
+    {
+      "cosine_similarity": 0.2590395490999218,
+      "epoch": 0.5992544268406338,
+      "grad_norm": 1.6137129915321247,
+      "learning_rate": 1.77839143942009e-05,
+      "loss": 2.2083,
+      "step": 643
+    },
+    {
+      "cosine_similarity": 0.20882713280053794,
+      "epoch": 0.6001863932898416,
+      "grad_norm": 1.4091819388479963,
+      "learning_rate": 1.777701070072489e-05,
+      "loss": 2.1765,
+      "step": 644
+    },
+    {
+      "cosine_similarity": 0.2499120544708068,
+      "epoch": 0.6011183597390494,
+      "grad_norm": 1.5725489637252026,
+      "learning_rate": 1.777010700724888e-05,
+      "loss": 2.2715,
+      "step": 645
+    },
+    {
+      "cosine_similarity": 0.3376138917931363,
+      "epoch": 0.6020503261882573,
+      "grad_norm": 1.4166580470087173,
+      "learning_rate": 1.776320331377287e-05,
+      "loss": 1.839,
+      "step": 646
+    },
+    {
+      "cosine_similarity": 0.2493270357944935,
+      "epoch": 0.6029822926374651,
+      "grad_norm": 1.8219549278891733,
+      "learning_rate": 1.775629962029686e-05,
+      "loss": 2.2043,
+      "step": 647
+    },
+    {
+      "cosine_similarity": 0.15026793436285332,
+      "epoch": 0.6039142590866728,
+      "grad_norm": 1.6276614738538706,
+      "learning_rate": 1.774939592682085e-05,
+      "loss": 2.0879,
+      "step": 648
+    },
+    {
+      "cosine_similarity": 0.16632815723700226,
+      "epoch": 0.6048462255358807,
+      "grad_norm": 1.7787995066102995,
+      "learning_rate": 1.774249223334484e-05,
+      "loss": 2.0023,
+      "step": 649
+    },
+    {
+      "cosine_similarity": 0.03108913007240282,
+      "epoch": 0.6057781919850885,
+      "grad_norm": 1.3219699842122543,
+      "learning_rate": 1.773558853986883e-05,
+      "loss": 1.9024,
+      "step": 650
+    },
+    {
+      "cosine_similarity": 0.0868199128245599,
+      "epoch": 0.6067101584342963,
+      "grad_norm": 1.6669858725869293,
+      "learning_rate": 1.7728684846392822e-05,
+      "loss": 2.1585,
+      "step": 651
+    },
+    {
+      "cosine_similarity": 0.13831246407967904,
+      "epoch": 0.6076421248835042,
+      "grad_norm": 1.510153298360638,
+      "learning_rate": 1.772178115291681e-05,
+      "loss": 2.0304,
+      "step": 652
+    },
+    {
+      "cosine_similarity": 0.2623996334119673,
+      "epoch": 0.608574091332712,
+      "grad_norm": 1.4260334089387494,
+      "learning_rate": 1.7714877459440804e-05,
+      "loss": 2.28,
+      "step": 653
+    },
+    {
+      "cosine_similarity": 0.3374928331479142,
+      "epoch": 0.6095060577819198,
+      "grad_norm": 1.5780825216286487,
+      "learning_rate": 1.7707973765964793e-05,
+      "loss": 2.295,
+      "step": 654
+    },
+    {
+      "cosine_similarity": 0.08515496391946785,
+      "epoch": 0.6104380242311277,
+      "grad_norm": 1.7339164141908738,
+      "learning_rate": 1.7701070072488782e-05,
+      "loss": 2.27,
+      "step": 655
+    },
+    {
+      "cosine_similarity": 0.36601491566862854,
+      "epoch": 0.6113699906803355,
+      "grad_norm": 1.923846695059125,
+      "learning_rate": 1.7694166379012774e-05,
+      "loss": 2.2041,
+      "step": 656
+    },
+    {
+      "cosine_similarity": 0.38600735883401804,
+      "epoch": 0.6123019571295434,
+      "grad_norm": 1.8874367385326638,
+      "learning_rate": 1.7687262685536763e-05,
+      "loss": 2.363,
+      "step": 657
+    },
+    {
+      "cosine_similarity": 0.008531682285142159,
+      "epoch": 0.6132339235787512,
+      "grad_norm": 1.432229377920772,
+      "learning_rate": 1.7680358992060752e-05,
+      "loss": 2.2258,
+      "step": 658
+    },
+    {
+      "cosine_similarity": 0.18239173895793526,
+      "epoch": 0.614165890027959,
+      "grad_norm": 1.2394665101059001,
+      "learning_rate": 1.7673455298584745e-05,
+      "loss": 1.9782,
+      "step": 659
+    },
+    {
+      "cosine_similarity": 0.14084360847690935,
+      "epoch": 0.6150978564771669,
+      "grad_norm": 1.7235704709576314,
+      "learning_rate": 1.7666551605108734e-05,
+      "loss": 1.9527,
+      "step": 660
+    },
+    {
+      "cosine_similarity": 0.4380390234705507,
+      "epoch": 0.6160298229263746,
+      "grad_norm": 1.673655014631557,
+      "learning_rate": 1.7659647911632726e-05,
+      "loss": 1.9759,
+      "step": 661
+    },
+    {
+      "cosine_similarity": 0.07716737455985016,
+      "epoch": 0.6169617893755824,
+      "grad_norm": 1.264604815570779,
+      "learning_rate": 1.7652744218156715e-05,
+      "loss": 1.8647,
+      "step": 662
+    },
+    {
+      "cosine_similarity": -0.0036292590838499614,
+      "epoch": 0.6178937558247903,
+      "grad_norm": 1.3592957729517812,
+      "learning_rate": 1.7645840524680708e-05,
+      "loss": 1.9935,
+      "step": 663
+    },
+    {
+      "cosine_similarity": 0.3867483083142738,
+      "epoch": 0.6188257222739981,
+      "grad_norm": 1.75664151274146,
+      "learning_rate": 1.7638936831204697e-05,
+      "loss": 2.1942,
+      "step": 664
+    },
+    {
+      "cosine_similarity": 0.1299859590538899,
+      "epoch": 0.6197576887232059,
+      "grad_norm": 1.2986061368709245,
+      "learning_rate": 1.7632033137728686e-05,
+      "loss": 2.3555,
+      "step": 665
+    },
+    {
+      "cosine_similarity": 0.2537039830936324,
+      "epoch": 0.6206896551724138,
+      "grad_norm": 1.5156797091518885,
+      "learning_rate": 1.7625129444252678e-05,
+      "loss": 2.2751,
+      "step": 666
+    },
+    {
+      "cosine_similarity": 0.35953872137364445,
+      "epoch": 0.6216216216216216,
+      "grad_norm": 1.313526734477539,
+      "learning_rate": 1.7618225750776667e-05,
+      "loss": 2.4836,
+      "step": 667
+    },
+    {
+      "cosine_similarity": 0.11924309941244184,
+      "epoch": 0.6225535880708295,
+      "grad_norm": 1.4241817347349992,
+      "learning_rate": 1.7611322057300656e-05,
+      "loss": 2.145,
+      "step": 668
+    },
+    {
+      "cosine_similarity": 0.13533919913389916,
+      "epoch": 0.6234855545200373,
+      "grad_norm": 1.8401649487876932,
+      "learning_rate": 1.7604418363824645e-05,
+      "loss": 2.3352,
+      "step": 669
+    },
+    {
+      "cosine_similarity": 0.5189776044951683,
+      "epoch": 0.6244175209692451,
+      "grad_norm": 1.9517085852841993,
+      "learning_rate": 1.7597514670348638e-05,
+      "loss": 2.246,
+      "step": 670
+    },
+    {
+      "cosine_similarity": 0.5181673845816088,
+      "epoch": 0.625349487418453,
+      "grad_norm": 1.349342855489934,
+      "learning_rate": 1.759061097687263e-05,
+      "loss": 2.1677,
+      "step": 671
+    },
+    {
+      "cosine_similarity": 0.2103524401699914,
+      "epoch": 0.6262814538676608,
+      "grad_norm": 1.4937719668536884,
+      "learning_rate": 1.758370728339662e-05,
+      "loss": 2.2508,
+      "step": 672
+    },
+    {
+      "cosine_similarity": 0.5335700724642364,
+      "epoch": 0.6272134203168686,
+      "grad_norm": 1.3018880242332718,
+      "learning_rate": 1.7576803589920608e-05,
+      "loss": 2.1958,
+      "step": 673
+    },
+    {
+      "cosine_similarity": 0.42192567322136415,
+      "epoch": 0.6281453867660765,
+      "grad_norm": 1.648170255060068,
+      "learning_rate": 1.75698998964446e-05,
+      "loss": 2.1596,
+      "step": 674
+    },
+    {
+      "cosine_similarity": -0.16297561054074403,
+      "epoch": 0.6290773532152842,
+      "grad_norm": 1.4379657711762703,
+      "learning_rate": 1.756299620296859e-05,
+      "loss": 1.9945,
+      "step": 675
+    },
+    {
+      "cosine_similarity": 0.5826260610915126,
+      "epoch": 0.630009319664492,
+      "grad_norm": 1.5410978297047615,
+      "learning_rate": 1.755609250949258e-05,
+      "loss": 2.5361,
+      "step": 676
+    },
+    {
+      "cosine_similarity": 0.42213174526993336,
+      "epoch": 0.6309412861136999,
+      "grad_norm": 1.7082183828830988,
+      "learning_rate": 1.754918881601657e-05,
+      "loss": 2.3925,
+      "step": 677
+    },
+    {
+      "cosine_similarity": 0.26712345282473815,
+      "epoch": 0.6318732525629077,
+      "grad_norm": 1.4783035453987425,
+      "learning_rate": 1.754228512254056e-05,
+      "loss": 2.5615,
+      "step": 678
+    },
+    {
+      "cosine_similarity": 0.45549641885234,
+      "epoch": 0.6328052190121156,
+      "grad_norm": 1.693395239921197,
+      "learning_rate": 1.753538142906455e-05,
+      "loss": 2.4547,
+      "step": 679
+    },
+    {
+      "cosine_similarity": 0.4150134735552354,
+      "epoch": 0.6337371854613234,
+      "grad_norm": 1.653769321973139,
+      "learning_rate": 1.752847773558854e-05,
+      "loss": 2.3763,
+      "step": 680
+    },
+    {
+      "cosine_similarity": 0.33464454701364843,
+      "epoch": 0.6346691519105312,
+      "grad_norm": 1.347836268414311,
+      "learning_rate": 1.7521574042112534e-05,
+      "loss": 2.3469,
+      "step": 681
+    },
+    {
+      "cosine_similarity": 0.20503313325071884,
+      "epoch": 0.6356011183597391,
+      "grad_norm": 1.407351497882235,
+      "learning_rate": 1.7514670348636523e-05,
+      "loss": 2.3068,
+      "step": 682
+    },
+    {
+      "cosine_similarity": 0.19018300499635404,
+      "epoch": 0.6365330848089469,
+      "grad_norm": 1.3449677524945904,
+      "learning_rate": 1.7507766655160512e-05,
+      "loss": 2.2387,
+      "step": 683
+    },
+    {
+      "cosine_similarity": 0.19769038336469721,
+      "epoch": 0.6374650512581547,
+      "grad_norm": 1.7258739584961038,
+      "learning_rate": 1.75008629616845e-05,
+      "loss": 2.1981,
+      "step": 684
+    },
+    {
+      "cosine_similarity": 0.01695213394759131,
+      "epoch": 0.6383970177073626,
+      "grad_norm": 1.3883511663324586,
+      "learning_rate": 1.7493959268208494e-05,
+      "loss": 2.3466,
+      "step": 685
+    },
+    {
+      "cosine_similarity": 0.07704133868292311,
+      "epoch": 0.6393289841565704,
+      "grad_norm": 1.545485821057894,
+      "learning_rate": 1.7487055574732483e-05,
+      "loss": 2.1089,
+      "step": 686
+    },
+    {
+      "cosine_similarity": 0.20279334927560197,
+      "epoch": 0.6402609506057781,
+      "grad_norm": 1.5776916117808166,
+      "learning_rate": 1.748015188125647e-05,
+      "loss": 2.091,
+      "step": 687
+    },
+    {
+      "cosine_similarity": 0.15164431929717537,
+      "epoch": 0.641192917054986,
+      "grad_norm": 1.6150159212796578,
+      "learning_rate": 1.7473248187780464e-05,
+      "loss": 2.1633,
+      "step": 688
+    },
+    {
+      "cosine_similarity": 0.4742200726029016,
+      "epoch": 0.6421248835041938,
+      "grad_norm": 1.3756658701976674,
+      "learning_rate": 1.7466344494304453e-05,
+      "loss": 2.3924,
+      "step": 689
+    },
+    {
+      "cosine_similarity": -0.06404728434740005,
+      "epoch": 0.6430568499534017,
+      "grad_norm": 1.5812940797555206,
+      "learning_rate": 1.7459440800828446e-05,
+      "loss": 2.1367,
+      "step": 690
+    },
+    {
+      "cosine_similarity": 0.3959252450511365,
+      "epoch": 0.6439888164026095,
+      "grad_norm": 1.781765718915774,
+      "learning_rate": 1.7452537107352435e-05,
+      "loss": 1.9554,
+      "step": 691
+    },
+    {
+      "cosine_similarity": 0.6152792936437798,
+      "epoch": 0.6449207828518173,
+      "grad_norm": 1.4222742574473686,
+      "learning_rate": 1.7445633413876427e-05,
+      "loss": 2.0153,
+      "step": 692
+    },
+    {
+      "cosine_similarity": 0.2487462831506036,
+      "epoch": 0.6458527493010252,
+      "grad_norm": 1.4912821468363817,
+      "learning_rate": 1.7438729720400416e-05,
+      "loss": 2.4503,
+      "step": 693
+    },
+    {
+      "cosine_similarity": 0.07553313841053107,
+      "epoch": 0.646784715750233,
+      "grad_norm": 1.341593178319673,
+      "learning_rate": 1.7431826026924405e-05,
+      "loss": 1.8842,
+      "step": 694
+    },
+    {
+      "cosine_similarity": 0.33426355734212637,
+      "epoch": 0.6477166821994408,
+      "grad_norm": 1.3428848198144696,
+      "learning_rate": 1.7424922333448394e-05,
+      "loss": 2.2258,
+      "step": 695
+    },
+    {
+      "cosine_similarity": 0.5744282146160679,
+      "epoch": 0.6486486486486487,
+      "grad_norm": 1.8142300713701809,
+      "learning_rate": 1.7418018639972387e-05,
+      "loss": 2.2129,
+      "step": 696
+    },
+    {
+      "cosine_similarity": 0.3709161663252145,
+      "epoch": 0.6495806150978565,
+      "grad_norm": 1.5345803110903347,
+      "learning_rate": 1.7411114946496376e-05,
+      "loss": 1.9698,
+      "step": 697
+    },
+    {
+      "cosine_similarity": -0.3837277073346395,
+      "epoch": 0.6505125815470643,
+      "grad_norm": 1.791852722383533,
+      "learning_rate": 1.7404211253020368e-05,
+      "loss": 2.4899,
+      "step": 698
+    },
+    {
+      "cosine_similarity": -0.05638468084974728,
+      "epoch": 0.6514445479962722,
+      "grad_norm": 2.6723854850624273,
+      "learning_rate": 1.7397307559544357e-05,
+      "loss": 2.4024,
+      "step": 699
+    },
+    {
+      "cosine_similarity": 0.47918447691842486,
+      "epoch": 0.65237651444548,
+      "grad_norm": 1.8210721944111286,
+      "learning_rate": 1.739040386606835e-05,
+      "loss": 2.252,
+      "step": 700
+    },
+    {
+      "cosine_similarity": 0.4805090332907061,
+      "epoch": 0.6533084808946877,
+      "grad_norm": 1.3827397820416836,
+      "learning_rate": 1.738350017259234e-05,
+      "loss": 2.241,
+      "step": 701
+    },
+    {
+      "cosine_similarity": 0.16454209629459401,
+      "epoch": 0.6542404473438956,
+      "grad_norm": 1.7963490915201135,
+      "learning_rate": 1.7376596479116328e-05,
+      "loss": 2.4615,
+      "step": 702
+    },
+    {
+      "cosine_similarity": 0.34825279806427184,
+      "epoch": 0.6551724137931034,
+      "grad_norm": 1.3655144475996015,
+      "learning_rate": 1.736969278564032e-05,
+      "loss": 2.442,
+      "step": 703
+    },
+    {
+      "cosine_similarity": 0.09415776325064694,
+      "epoch": 0.6561043802423113,
+      "grad_norm": 1.6339317657397607,
+      "learning_rate": 1.736278909216431e-05,
+      "loss": 2.313,
+      "step": 704
+    },
+    {
+      "cosine_similarity": 0.20737519376455074,
+      "epoch": 0.6570363466915191,
+      "grad_norm": 1.2317544935761169,
+      "learning_rate": 1.7355885398688298e-05,
+      "loss": 2.1548,
+      "step": 705
+    },
+    {
+      "cosine_similarity": 0.4180357231239121,
+      "epoch": 0.6579683131407269,
+      "grad_norm": 1.5839109317113107,
+      "learning_rate": 1.734898170521229e-05,
+      "loss": 2.1009,
+      "step": 706
+    },
+    {
+      "cosine_similarity": 0.09613192139184223,
+      "epoch": 0.6589002795899348,
+      "grad_norm": 1.5397853130070611,
+      "learning_rate": 1.734207801173628e-05,
+      "loss": 2.5301,
+      "step": 707
+    },
+    {
+      "cosine_similarity": 0.07834554579532917,
+      "epoch": 0.6598322460391426,
+      "grad_norm": 1.688873575399597,
+      "learning_rate": 1.7335174318260272e-05,
+      "loss": 1.9098,
+      "step": 708
+    },
+    {
+      "cosine_similarity": 0.7766640311760398,
+      "epoch": 0.6607642124883504,
+      "grad_norm": 1.4845099564643083,
+      "learning_rate": 1.732827062478426e-05,
+      "loss": 2.5454,
+      "step": 709
+    },
+    {
+      "cosine_similarity": 0.23290732060237418,
+      "epoch": 0.6616961789375583,
+      "grad_norm": 1.3858548100451757,
+      "learning_rate": 1.7321366931308253e-05,
+      "loss": 2.1709,
+      "step": 710
+    },
+    {
+      "cosine_similarity": -0.10840906102594784,
+      "epoch": 0.6626281453867661,
+      "grad_norm": 1.6707144871709017,
+      "learning_rate": 1.7314463237832242e-05,
+      "loss": 2.2321,
+      "step": 711
+    },
+    {
+      "cosine_similarity": 0.17281315099671685,
+      "epoch": 0.6635601118359739,
+      "grad_norm": 1.2457568365244849,
+      "learning_rate": 1.730755954435623e-05,
+      "loss": 2.4624,
+      "step": 712
+    },
+    {
+      "cosine_similarity": 0.14060993793942872,
+      "epoch": 0.6644920782851818,
+      "grad_norm": 1.6988527470409855,
+      "learning_rate": 1.730065585088022e-05,
+      "loss": 2.0266,
+      "step": 713
+    },
+    {
+      "cosine_similarity": 0.17029679957767838,
+      "epoch": 0.6654240447343895,
+      "grad_norm": 1.3254630292547525,
+      "learning_rate": 1.7293752157404213e-05,
+      "loss": 2.3442,
+      "step": 714
+    },
+    {
+      "cosine_similarity": 0.05238979478474368,
+      "epoch": 0.6663560111835974,
+      "grad_norm": 1.3961296472775093,
+      "learning_rate": 1.7286848463928202e-05,
+      "loss": 2.1571,
+      "step": 715
+    },
+    {
+      "cosine_similarity": 0.13639863623265805,
+      "epoch": 0.6672879776328052,
+      "grad_norm": 1.356166021208571,
+      "learning_rate": 1.7279944770452194e-05,
+      "loss": 1.9877,
+      "step": 716
+    },
+    {
+      "cosine_similarity": 0.1908063602720427,
+      "epoch": 0.668219944082013,
+      "grad_norm": 1.3210772603650225,
+      "learning_rate": 1.7273041076976183e-05,
+      "loss": 2.3595,
+      "step": 717
+    },
+    {
+      "cosine_similarity": 0.2997470790555872,
+      "epoch": 0.6691519105312209,
+      "grad_norm": 1.2612299109576681,
+      "learning_rate": 1.7266137383500176e-05,
+      "loss": 2.2055,
+      "step": 718
+    },
+    {
+      "cosine_similarity": 0.04870698764437837,
+      "epoch": 0.6700838769804287,
+      "grad_norm": 1.5191672047152178,
+      "learning_rate": 1.7259233690024165e-05,
+      "loss": 2.3114,
+      "step": 719
+    },
+    {
+      "cosine_similarity": 0.24899859276752828,
+      "epoch": 0.6710158434296365,
+      "grad_norm": 1.4919706726999207,
+      "learning_rate": 1.7252329996548154e-05,
+      "loss": 2.3099,
+      "step": 720
+    },
+    {
+      "cosine_similarity": 0.14649174174127005,
+      "epoch": 0.6719478098788444,
+      "grad_norm": 1.6553096590780922,
+      "learning_rate": 1.7245426303072146e-05,
+      "loss": 2.1004,
+      "step": 721
+    },
+    {
+      "cosine_similarity": 0.3195440185911667,
+      "epoch": 0.6728797763280522,
+      "grad_norm": 1.360923710736457,
+      "learning_rate": 1.7238522609596135e-05,
+      "loss": 2.3179,
+      "step": 722
+    },
+    {
+      "cosine_similarity": 0.19507928135165656,
+      "epoch": 0.67381174277726,
+      "grad_norm": 1.3717168036006826,
+      "learning_rate": 1.7231618916120124e-05,
+      "loss": 2.0828,
+      "step": 723
+    },
+    {
+      "cosine_similarity": 0.2615442469103937,
+      "epoch": 0.6747437092264679,
+      "grad_norm": 2.154962670288846,
+      "learning_rate": 1.7224715222644113e-05,
+      "loss": 2.3111,
+      "step": 724
+    },
+    {
+      "cosine_similarity": 0.2308000791417178,
+      "epoch": 0.6756756756756757,
+      "grad_norm": 1.4704082119463906,
+      "learning_rate": 1.7217811529168106e-05,
+      "loss": 2.5043,
+      "step": 725
+    },
+    {
+      "cosine_similarity": 0.18473852724213888,
+      "epoch": 0.6766076421248836,
+      "grad_norm": 1.8486737332551197,
+      "learning_rate": 1.72109078356921e-05,
+      "loss": 2.2452,
+      "step": 726
+    },
+    {
+      "cosine_similarity": 0.14800486623800427,
+      "epoch": 0.6775396085740913,
+      "grad_norm": 1.4583162975706356,
+      "learning_rate": 1.7204004142216087e-05,
+      "loss": 2.0382,
+      "step": 727
+    },
+    {
+      "cosine_similarity": 0.08119314232966178,
+      "epoch": 0.6784715750232991,
+      "grad_norm": 1.3405134984579061,
+      "learning_rate": 1.719710044874008e-05,
+      "loss": 2.1761,
+      "step": 728
+    },
+    {
+      "cosine_similarity": 0.08791905996628079,
+      "epoch": 0.679403541472507,
+      "grad_norm": 1.8135506766336844,
+      "learning_rate": 1.719019675526407e-05,
+      "loss": 2.0079,
+      "step": 729
+    },
+    {
+      "cosine_similarity": 0.37947590238049034,
+      "epoch": 0.6803355079217148,
+      "grad_norm": 1.5807178640875017,
+      "learning_rate": 1.7183293061788058e-05,
+      "loss": 2.0752,
+      "step": 730
+    },
+    {
+      "cosine_similarity": 0.03719084108922302,
+      "epoch": 0.6812674743709226,
+      "grad_norm": 1.2882858963213313,
+      "learning_rate": 1.7176389368312047e-05,
+      "loss": 2.1517,
+      "step": 731
+    },
+    {
+      "cosine_similarity": 0.3206953049626113,
+      "epoch": 0.6821994408201305,
+      "grad_norm": 1.3500165005674547,
+      "learning_rate": 1.716948567483604e-05,
+      "loss": 2.2089,
+      "step": 732
+    },
+    {
+      "cosine_similarity": 0.03966383319196566,
+      "epoch": 0.6831314072693383,
+      "grad_norm": 1.5462267751927081,
+      "learning_rate": 1.716258198136003e-05,
+      "loss": 2.1678,
+      "step": 733
+    },
+    {
+      "cosine_similarity": 0.15528817946964357,
+      "epoch": 0.6840633737185461,
+      "grad_norm": 1.3079102577338002,
+      "learning_rate": 1.7155678287884017e-05,
+      "loss": 2.3424,
+      "step": 734
+    },
+    {
+      "cosine_similarity": 0.4108571355419645,
+      "epoch": 0.684995340167754,
+      "grad_norm": 1.6552688218516096,
+      "learning_rate": 1.714877459440801e-05,
+      "loss": 2.0467,
+      "step": 735
+    },
+    {
+      "cosine_similarity": 0.5686287001589877,
+      "epoch": 0.6859273066169618,
+      "grad_norm": 1.3828913818986823,
+      "learning_rate": 1.7141870900932002e-05,
+      "loss": 2.1638,
+      "step": 736
+    },
+    {
+      "cosine_similarity": 0.10246299523546752,
+      "epoch": 0.6868592730661697,
+      "grad_norm": 2.1033937241329475,
+      "learning_rate": 1.713496720745599e-05,
+      "loss": 2.2129,
+      "step": 737
+    },
+    {
+      "cosine_similarity": -0.4118948474920426,
+      "epoch": 0.6877912395153775,
+      "grad_norm": 1.8754061026692626,
+      "learning_rate": 1.712806351397998e-05,
+      "loss": 2.2979,
+      "step": 738
+    },
+    {
+      "cosine_similarity": 0.0010621632671874192,
+      "epoch": 0.6887232059645852,
+      "grad_norm": 1.4187547227120985,
+      "learning_rate": 1.7121159820503973e-05,
+      "loss": 2.3253,
+      "step": 739
+    },
+    {
+      "cosine_similarity": 0.21099935239956658,
+      "epoch": 0.6896551724137931,
+      "grad_norm": 1.2382379139723347,
+      "learning_rate": 1.7114256127027962e-05,
+      "loss": 1.931,
+      "step": 740
+    },
+    {
+      "cosine_similarity": 0.18777392233518228,
+      "epoch": 0.6905871388630009,
+      "grad_norm": 1.2043750408025158,
+      "learning_rate": 1.710735243355195e-05,
+      "loss": 2.0181,
+      "step": 741
+    },
+    {
+      "cosine_similarity": -0.06073761870865879,
+      "epoch": 0.6915191053122087,
+      "grad_norm": 1.4132033547714014,
+      "learning_rate": 1.710044874007594e-05,
+      "loss": 1.9972,
+      "step": 742
+    },
+    {
+      "cosine_similarity": -0.16956872155677657,
+      "epoch": 0.6924510717614166,
+      "grad_norm": 1.3668351288129459,
+      "learning_rate": 1.7093545046599932e-05,
+      "loss": 2.2089,
+      "step": 743
+    },
+    {
+      "cosine_similarity": -0.26371706654193333,
+      "epoch": 0.6933830382106244,
+      "grad_norm": 1.6969452054593235,
+      "learning_rate": 1.708664135312392e-05,
+      "loss": 2.1095,
+      "step": 744
+    },
+    {
+      "cosine_similarity": 0.17802015118175585,
+      "epoch": 0.6943150046598322,
+      "grad_norm": 1.4619674904690259,
+      "learning_rate": 1.7079737659647914e-05,
+      "loss": 2.3362,
+      "step": 745
+    },
+    {
+      "cosine_similarity": 0.26864299149891924,
+      "epoch": 0.6952469711090401,
+      "grad_norm": 1.4833885367376598,
+      "learning_rate": 1.7072833966171903e-05,
+      "loss": 2.1875,
+      "step": 746
+    },
+    {
+      "cosine_similarity": 0.5098131095355585,
+      "epoch": 0.6961789375582479,
+      "grad_norm": 1.4127405308716516,
+      "learning_rate": 1.7065930272695895e-05,
+      "loss": 2.0366,
+      "step": 747
+    },
+    {
+      "cosine_similarity": 0.0805994355018322,
+      "epoch": 0.6971109040074557,
+      "grad_norm": 1.3622932038836773,
+      "learning_rate": 1.7059026579219884e-05,
+      "loss": 2.1347,
+      "step": 748
+    },
+    {
+      "cosine_similarity": 0.4426900060422734,
+      "epoch": 0.6980428704566636,
+      "grad_norm": 1.3644334922597743,
+      "learning_rate": 1.7052122885743873e-05,
+      "loss": 2.1915,
+      "step": 749
+    },
+    {
+      "cosine_similarity": 0.0022344778040265553,
+      "epoch": 0.6989748369058714,
+      "grad_norm": 1.7702169358417938,
+      "learning_rate": 1.7045219192267866e-05,
+      "loss": 1.9601,
+      "step": 750
+    },
+    {
+      "cosine_similarity": 0.17893472475364505,
+      "epoch": 0.6999068033550793,
+      "grad_norm": 1.2953596140741745,
+      "learning_rate": 1.7038315498791855e-05,
+      "loss": 2.0127,
+      "step": 751
+    },
+    {
+      "cosine_similarity": 0.3833159609518247,
+      "epoch": 0.700838769804287,
+      "grad_norm": 2.0816342837747324,
+      "learning_rate": 1.7031411805315844e-05,
+      "loss": 2.1138,
+      "step": 752
+    },
+    {
+      "cosine_similarity": 0.1107973588825389,
+      "epoch": 0.7017707362534948,
+      "grad_norm": 1.4082249295475158,
+      "learning_rate": 1.7024508111839836e-05,
+      "loss": 2.4246,
+      "step": 753
+    },
+    {
+      "cosine_similarity": -0.0895376481790825,
+      "epoch": 0.7027027027027027,
+      "grad_norm": 1.4485712520853549,
+      "learning_rate": 1.7017604418363825e-05,
+      "loss": 2.374,
+      "step": 754
+    },
+    {
+      "cosine_similarity": 0.659165527545006,
+      "epoch": 0.7036346691519105,
+      "grad_norm": 1.2927784943869396,
+      "learning_rate": 1.7010700724887818e-05,
+      "loss": 2.3257,
+      "step": 755
+    },
+    {
+      "cosine_similarity": 0.2677224455259327,
+      "epoch": 0.7045666356011183,
+      "grad_norm": 1.4898252483260226,
+      "learning_rate": 1.7003797031411807e-05,
+      "loss": 1.96,
+      "step": 756
+    },
+    {
+      "cosine_similarity": 0.014408810211580378,
+      "epoch": 0.7054986020503262,
+      "grad_norm": 1.5874047978397055,
+      "learning_rate": 1.69968933379358e-05,
+      "loss": 1.933,
+      "step": 757
+    },
+    {
+      "cosine_similarity": 0.10361462304014546,
+      "epoch": 0.706430568499534,
+      "grad_norm": 1.7169545889147082,
+      "learning_rate": 1.6989989644459788e-05,
+      "loss": 2.3426,
+      "step": 758
+    },
+    {
+      "cosine_similarity": 0.8429669801991377,
+      "epoch": 0.7073625349487418,
+      "grad_norm": 1.3734707917687008,
+      "learning_rate": 1.6983085950983777e-05,
+      "loss": 2.2759,
+      "step": 759
+    },
+    {
+      "cosine_similarity": 0.28926672952450083,
+      "epoch": 0.7082945013979497,
+      "grad_norm": 1.446327816101935,
+      "learning_rate": 1.6976182257507766e-05,
+      "loss": 2.041,
+      "step": 760
+    },
+    {
+      "cosine_similarity": 0.5786801500241564,
+      "epoch": 0.7092264678471575,
+      "grad_norm": 1.5347125173571818,
+      "learning_rate": 1.696927856403176e-05,
+      "loss": 2.0175,
+      "step": 761
+    },
+    {
+      "cosine_similarity": 0.4836560793471901,
+      "epoch": 0.7101584342963654,
+      "grad_norm": 1.4425072290201555,
+      "learning_rate": 1.6962374870555748e-05,
+      "loss": 2.8634,
+      "step": 762
+    },
+    {
+      "cosine_similarity": 0.10137585582117084,
+      "epoch": 0.7110904007455732,
+      "grad_norm": 1.4892229745016707,
+      "learning_rate": 1.695547117707974e-05,
+      "loss": 2.3236,
+      "step": 763
+    },
+    {
+      "cosine_similarity": 0.38647254834692046,
+      "epoch": 0.712022367194781,
+      "grad_norm": 1.5666249294190633,
+      "learning_rate": 1.694856748360373e-05,
+      "loss": 2.2746,
+      "step": 764
+    },
+    {
+      "cosine_similarity": 0.35048235513912485,
+      "epoch": 0.7129543336439889,
+      "grad_norm": 1.3083473519541076,
+      "learning_rate": 1.694166379012772e-05,
+      "loss": 2.0945,
+      "step": 765
+    },
+    {
+      "cosine_similarity": 0.19832660521314383,
+      "epoch": 0.7138863000931966,
+      "grad_norm": 2.9060691911489487,
+      "learning_rate": 1.693476009665171e-05,
+      "loss": 2.3111,
+      "step": 766
+    },
+    {
+      "cosine_similarity": 0.4942236118695565,
+      "epoch": 0.7148182665424044,
+      "grad_norm": 1.7656166252817864,
+      "learning_rate": 1.69278564031757e-05,
+      "loss": 2.1217,
+      "step": 767
+    },
+    {
+      "cosine_similarity": 0.48857092509464045,
+      "epoch": 0.7157502329916123,
+      "grad_norm": 1.2686110309695684,
+      "learning_rate": 1.6920952709699692e-05,
+      "loss": 2.1964,
+      "step": 768
+    },
+    {
+      "cosine_similarity": 0.27787511746096033,
+      "epoch": 0.7166821994408201,
+      "grad_norm": 1.16952885737967,
+      "learning_rate": 1.691404901622368e-05,
+      "loss": 2.0784,
+      "step": 769
+    },
+    {
+      "cosine_similarity": 0.21972981793755492,
+      "epoch": 0.7176141658900279,
+      "grad_norm": 1.5420237154590706,
+      "learning_rate": 1.690714532274767e-05,
+      "loss": 1.9872,
+      "step": 770
+    },
+    {
+      "cosine_similarity": 0.7928378590483918,
+      "epoch": 0.7185461323392358,
+      "grad_norm": 1.820921756106068,
+      "learning_rate": 1.690024162927166e-05,
+      "loss": 2.0491,
+      "step": 771
+    },
+    {
+      "cosine_similarity": 0.5821221088335734,
+      "epoch": 0.7194780987884436,
+      "grad_norm": 1.4013846486683317,
+      "learning_rate": 1.689333793579565e-05,
+      "loss": 2.2539,
+      "step": 772
+    },
+    {
+      "cosine_similarity": 0.6607554653308144,
+      "epoch": 0.7204100652376515,
+      "grad_norm": 1.3030320379516342,
+      "learning_rate": 1.6886434242319644e-05,
+      "loss": 2.0757,
+      "step": 773
+    },
+    {
+      "cosine_similarity": 0.408166388497929,
+      "epoch": 0.7213420316868593,
+      "grad_norm": 1.2761224868360794,
+      "learning_rate": 1.6879530548843633e-05,
+      "loss": 2.0559,
+      "step": 774
+    },
+    {
+      "cosine_similarity": 0.21458876668526572,
+      "epoch": 0.7222739981360671,
+      "grad_norm": 1.4606104570794478,
+      "learning_rate": 1.6872626855367622e-05,
+      "loss": 2.5228,
+      "step": 775
+    },
+    {
+      "cosine_similarity": 0.129389525810783,
+      "epoch": 0.723205964585275,
+      "grad_norm": 1.5137828920426981,
+      "learning_rate": 1.6865723161891615e-05,
+      "loss": 2.1278,
+      "step": 776
+    },
+    {
+      "cosine_similarity": 0.3253364675638196,
+      "epoch": 0.7241379310344828,
+      "grad_norm": 1.4009079358026033,
+      "learning_rate": 1.6858819468415604e-05,
+      "loss": 1.8376,
+      "step": 777
+    },
+    {
+      "cosine_similarity": 0.6905118400835146,
+      "epoch": 0.7250698974836906,
+      "grad_norm": 1.719663614376492,
+      "learning_rate": 1.6851915774939593e-05,
+      "loss": 1.9378,
+      "step": 778
+    },
+    {
+      "cosine_similarity": 0.21090253795844205,
+      "epoch": 0.7260018639328985,
+      "grad_norm": 2.27724516984217,
+      "learning_rate": 1.6845012081463585e-05,
+      "loss": 2.3529,
+      "step": 779
+    },
+    {
+      "cosine_similarity": -0.03864112192599968,
+      "epoch": 0.7269338303821062,
+      "grad_norm": 1.4676334147030825,
+      "learning_rate": 1.6838108387987574e-05,
+      "loss": 1.8202,
+      "step": 780
+    },
+    {
+      "cosine_similarity": 0.4704509215169893,
+      "epoch": 0.727865796831314,
+      "grad_norm": 1.3453183836135063,
+      "learning_rate": 1.6831204694511563e-05,
+      "loss": 2.2876,
+      "step": 781
+    },
+    {
+      "cosine_similarity": -0.047672786143606365,
+      "epoch": 0.7287977632805219,
+      "grad_norm": 1.3196141674852604,
+      "learning_rate": 1.6824301001035556e-05,
+      "loss": 2.1415,
+      "step": 782
+    },
+    {
+      "cosine_similarity": 0.45145026973137026,
+      "epoch": 0.7297297297297297,
+      "grad_norm": 1.2503938229259295,
+      "learning_rate": 1.6817397307559548e-05,
+      "loss": 1.7689,
+      "step": 783
+    },
+    {
+      "cosine_similarity": 0.06241632942797138,
+      "epoch": 0.7306616961789375,
+      "grad_norm": 1.798325775966069,
+      "learning_rate": 1.6810493614083537e-05,
+      "loss": 2.0161,
+      "step": 784
+    },
+    {
+      "cosine_similarity": 0.17545520318075736,
+      "epoch": 0.7315936626281454,
+      "grad_norm": 2.1153504020825253,
+      "learning_rate": 1.6803589920607526e-05,
+      "loss": 2.5741,
+      "step": 785
+    },
+    {
+      "cosine_similarity": 0.43394495432550906,
+      "epoch": 0.7325256290773532,
+      "grad_norm": 1.8694171896522163,
+      "learning_rate": 1.6796686227131515e-05,
+      "loss": 2.4856,
+      "step": 786
+    },
+    {
+      "cosine_similarity": 0.18400497123332707,
+      "epoch": 0.7334575955265611,
+      "grad_norm": 1.6387775341931108,
+      "learning_rate": 1.6789782533655508e-05,
+      "loss": 2.0938,
+      "step": 787
+    },
+    {
+      "cosine_similarity": -0.29314958946220826,
+      "epoch": 0.7343895619757689,
+      "grad_norm": 1.8947051886139157,
+      "learning_rate": 1.6782878840179497e-05,
+      "loss": 2.3258,
+      "step": 788
+    },
+    {
+      "cosine_similarity": 0.03564832327698337,
+      "epoch": 0.7353215284249767,
+      "grad_norm": 2.1930883664824816,
+      "learning_rate": 1.6775975146703486e-05,
+      "loss": 2.1662,
+      "step": 789
+    },
+    {
+      "cosine_similarity": 0.6010400246448155,
+      "epoch": 0.7362534948741846,
+      "grad_norm": 1.5316697218036917,
+      "learning_rate": 1.6769071453227478e-05,
+      "loss": 2.3847,
+      "step": 790
+    },
+    {
+      "cosine_similarity": 0.42061012607353565,
+      "epoch": 0.7371854613233924,
+      "grad_norm": 1.3065044056859134,
+      "learning_rate": 1.6762167759751467e-05,
+      "loss": 1.9106,
+      "step": 791
+    },
+    {
+      "cosine_similarity": 0.46507063495719114,
+      "epoch": 0.7381174277726001,
+      "grad_norm": 1.3638821687944265,
+      "learning_rate": 1.675526406627546e-05,
+      "loss": 2.0212,
+      "step": 792
+    },
+    {
+      "cosine_similarity": 0.18693361882725024,
+      "epoch": 0.739049394221808,
+      "grad_norm": 1.5985161155545224,
+      "learning_rate": 1.674836037279945e-05,
+      "loss": 2.2749,
+      "step": 793
+    },
+    {
+      "cosine_similarity": -0.10453531956300227,
+      "epoch": 0.7399813606710158,
+      "grad_norm": 1.4756213205868058,
+      "learning_rate": 1.674145667932344e-05,
+      "loss": 2.0069,
+      "step": 794
+    },
+    {
+      "cosine_similarity": 0.2763105387895728,
+      "epoch": 0.7409133271202236,
+      "grad_norm": 1.480997693000602,
+      "learning_rate": 1.673455298584743e-05,
+      "loss": 2.1562,
+      "step": 795
+    },
+    {
+      "cosine_similarity": 0.5235206939209689,
+      "epoch": 0.7418452935694315,
+      "grad_norm": 1.2454757607293239,
+      "learning_rate": 1.672764929237142e-05,
+      "loss": 1.7992,
+      "step": 796
+    },
+    {
+      "cosine_similarity": 0.14402023930515162,
+      "epoch": 0.7427772600186393,
+      "grad_norm": 1.3831435906788003,
+      "learning_rate": 1.672074559889541e-05,
+      "loss": 2.4108,
+      "step": 797
+    },
+    {
+      "cosine_similarity": 0.2639980713457272,
+      "epoch": 0.7437092264678472,
+      "grad_norm": 1.552305757558759,
+      "learning_rate": 1.67138419054194e-05,
+      "loss": 2.0854,
+      "step": 798
+    },
+    {
+      "cosine_similarity": 0.09200569363688607,
+      "epoch": 0.744641192917055,
+      "grad_norm": 2.397246200450407,
+      "learning_rate": 1.670693821194339e-05,
+      "loss": 2.1416,
+      "step": 799
+    },
+    {
+      "cosine_similarity": 0.4708733673080242,
+      "epoch": 0.7455731593662628,
+      "grad_norm": 1.7352785305765037,
+      "learning_rate": 1.6700034518467382e-05,
+      "loss": 2.0815,
+      "step": 800
+    },
+    {
+      "cosine_similarity": 0.10570396474147721,
+      "epoch": 0.7465051258154707,
+      "grad_norm": 1.4956973746205988,
+      "learning_rate": 1.669313082499137e-05,
+      "loss": 2.0761,
+      "step": 801
+    },
+    {
+      "cosine_similarity": 0.2383015917672328,
+      "epoch": 0.7474370922646785,
+      "grad_norm": 1.2665137170679859,
+      "learning_rate": 1.6686227131515363e-05,
+      "loss": 2.0544,
+      "step": 802
+    },
+    {
+      "cosine_similarity": -0.09086735093866392,
+      "epoch": 0.7483690587138863,
+      "grad_norm": 1.3829729471043029,
+      "learning_rate": 1.6679323438039353e-05,
+      "loss": 1.9121,
+      "step": 803
+    },
+    {
+      "cosine_similarity": 0.36985461786082086,
+      "epoch": 0.7493010251630942,
+      "grad_norm": 1.5472939261477696,
+      "learning_rate": 1.667241974456334e-05,
+      "loss": 1.9328,
+      "step": 804
+    },
+    {
+      "cosine_similarity": 0.5828921137899554,
+      "epoch": 0.750232991612302,
+      "grad_norm": 1.3052323632816571,
+      "learning_rate": 1.6665516051087334e-05,
+      "loss": 2.0891,
+      "step": 805
+    },
+    {
+      "cosine_similarity": 0.4495715799360949,
+      "epoch": 0.7511649580615097,
+      "grad_norm": 1.8564858929420038,
+      "learning_rate": 1.6658612357611323e-05,
+      "loss": 2.0714,
+      "step": 806
+    },
+    {
+      "cosine_similarity": 0.042457640246512544,
+      "epoch": 0.7520969245107176,
+      "grad_norm": 1.5779542920335738,
+      "learning_rate": 1.6651708664135312e-05,
+      "loss": 2.4188,
+      "step": 807
+    },
+    {
+      "cosine_similarity": 0.20959685505504963,
+      "epoch": 0.7530288909599254,
+      "grad_norm": 1.441604389940688,
+      "learning_rate": 1.6644804970659305e-05,
+      "loss": 2.502,
+      "step": 808
+    },
+    {
+      "cosine_similarity": 0.10176069772019865,
+      "epoch": 0.7539608574091333,
+      "grad_norm": 1.3935042507950215,
+      "learning_rate": 1.6637901277183294e-05,
+      "loss": 2.3354,
+      "step": 809
+    },
+    {
+      "cosine_similarity": 0.11563787599566736,
+      "epoch": 0.7548928238583411,
+      "grad_norm": 1.3650107426103366,
+      "learning_rate": 1.6630997583707286e-05,
+      "loss": 2.0096,
+      "step": 810
+    },
+    {
+      "cosine_similarity": 0.15972223945846667,
+      "epoch": 0.7558247903075489,
+      "grad_norm": 1.2795345803322344,
+      "learning_rate": 1.6624093890231275e-05,
+      "loss": 2.1731,
+      "step": 811
+    },
+    {
+      "cosine_similarity": 0.21791257505499825,
+      "epoch": 0.7567567567567568,
+      "grad_norm": 1.7078422438065641,
+      "learning_rate": 1.6617190196755267e-05,
+      "loss": 2.2594,
+      "step": 812
+    },
+    {
+      "cosine_similarity": 0.27089045980565196,
+      "epoch": 0.7576887232059646,
+      "grad_norm": 1.4155063316899799,
+      "learning_rate": 1.6610286503279256e-05,
+      "loss": 2.2671,
+      "step": 813
+    },
+    {
+      "cosine_similarity": -0.07809140821364682,
+      "epoch": 0.7586206896551724,
+      "grad_norm": 1.3187080948279781,
+      "learning_rate": 1.6603382809803246e-05,
+      "loss": 2.3231,
+      "step": 814
+    },
+    {
+      "cosine_similarity": 0.5077544024360523,
+      "epoch": 0.7595526561043803,
+      "grad_norm": 1.2309448000594245,
+      "learning_rate": 1.6596479116327235e-05,
+      "loss": 1.8347,
+      "step": 815
+    },
+    {
+      "cosine_similarity": 0.305998489213082,
+      "epoch": 0.7604846225535881,
+      "grad_norm": 1.2639568228736908,
+      "learning_rate": 1.6589575422851227e-05,
+      "loss": 1.663,
+      "step": 816
+    },
+    {
+      "cosine_similarity": 0.20366019620389889,
+      "epoch": 0.7614165890027959,
+      "grad_norm": 1.5075888854049195,
+      "learning_rate": 1.6582671729375216e-05,
+      "loss": 2.2405,
+      "step": 817
+    },
+    {
+      "cosine_similarity": 0.30188361714813977,
+      "epoch": 0.7623485554520038,
+      "grad_norm": 1.1164218626980655,
+      "learning_rate": 1.657576803589921e-05,
+      "loss": 1.7883,
+      "step": 818
+    },
+    {
+      "cosine_similarity": 0.5430489844791261,
+      "epoch": 0.7632805219012115,
+      "grad_norm": 1.3068802153747885,
+      "learning_rate": 1.6568864342423197e-05,
+      "loss": 1.9692,
+      "step": 819
+    },
+    {
+      "cosine_similarity": 0.08842494751307206,
+      "epoch": 0.7642124883504194,
+      "grad_norm": 1.2513766089618228,
+      "learning_rate": 1.656196064894719e-05,
+      "loss": 1.9189,
+      "step": 820
+    },
+    {
+      "cosine_similarity": 0.430263846255103,
+      "epoch": 0.7651444547996272,
+      "grad_norm": 1.3430278579494774,
+      "learning_rate": 1.655505695547118e-05,
+      "loss": 2.0985,
+      "step": 821
+    },
+    {
+      "cosine_similarity": 0.3340099287891539,
+      "epoch": 0.766076421248835,
+      "grad_norm": 1.8465958630443766,
+      "learning_rate": 1.6548153261995168e-05,
+      "loss": 2.12,
+      "step": 822
+    },
+    {
+      "cosine_similarity": -0.08681685519980213,
+      "epoch": 0.7670083876980429,
+      "grad_norm": 1.6666585851548426,
+      "learning_rate": 1.654124956851916e-05,
+      "loss": 2.2903,
+      "step": 823
+    },
+    {
+      "cosine_similarity": 0.06113244162572767,
+      "epoch": 0.7679403541472507,
+      "grad_norm": 1.5765492019847922,
+      "learning_rate": 1.653434587504315e-05,
+      "loss": 2.5968,
+      "step": 824
+    },
+    {
+      "cosine_similarity": -0.023234633832773886,
+      "epoch": 0.7688723205964585,
+      "grad_norm": 1.9809444359166968,
+      "learning_rate": 1.652744218156714e-05,
+      "loss": 2.3812,
+      "step": 825
+    },
+    {
+      "cosine_similarity": -0.020493760916813937,
+      "epoch": 0.7698042870456664,
+      "grad_norm": 1.1234798888335866,
+      "learning_rate": 1.6520538488091128e-05,
+      "loss": 1.9185,
+      "step": 826
+    },
+    {
+      "cosine_similarity": 0.16365276036931867,
+      "epoch": 0.7707362534948742,
+      "grad_norm": 2.5893118731458578,
+      "learning_rate": 1.651363479461512e-05,
+      "loss": 2.2645,
+      "step": 827
+    },
+    {
+      "cosine_similarity": 0.2681550138976699,
+      "epoch": 0.771668219944082,
+      "grad_norm": 1.3084781083371513,
+      "learning_rate": 1.6506731101139112e-05,
+      "loss": 2.1862,
+      "step": 828
+    },
+    {
+      "cosine_similarity": 0.2593114036939322,
+      "epoch": 0.7726001863932899,
+      "grad_norm": 2.22364221582845,
+      "learning_rate": 1.64998274076631e-05,
+      "loss": 1.8956,
+      "step": 829
+    },
+    {
+      "cosine_similarity": 0.576582284695463,
+      "epoch": 0.7735321528424977,
+      "grad_norm": 1.6062021223731229,
+      "learning_rate": 1.6492923714187094e-05,
+      "loss": 2.4267,
+      "step": 830
+    },
+    {
+      "cosine_similarity": 0.7186475674638538,
+      "epoch": 0.7744641192917054,
+      "grad_norm": 1.1946558388698483,
+      "learning_rate": 1.6486020020711083e-05,
+      "loss": 1.9747,
+      "step": 831
+    },
+    {
+      "cosine_similarity": 0.03315254769782236,
+      "epoch": 0.7753960857409133,
+      "grad_norm": 1.3396227854807465,
+      "learning_rate": 1.6479116327235072e-05,
+      "loss": 1.9413,
+      "step": 832
+    },
+    {
+      "cosine_similarity": 0.5334904789772058,
+      "epoch": 0.7763280521901211,
+      "grad_norm": 1.3091237643843323,
+      "learning_rate": 1.647221263375906e-05,
+      "loss": 2.2201,
+      "step": 833
+    },
+    {
+      "cosine_similarity": 0.1263890761154602,
+      "epoch": 0.777260018639329,
+      "grad_norm": 1.1546421072060815,
+      "learning_rate": 1.6465308940283053e-05,
+      "loss": 2.1112,
+      "step": 834
+    },
+    {
+      "cosine_similarity": 0.5846491933810507,
+      "epoch": 0.7781919850885368,
+      "grad_norm": 1.754385364777399,
+      "learning_rate": 1.6458405246807042e-05,
+      "loss": 2.2058,
+      "step": 835
+    },
+    {
+      "cosine_similarity": 0.10981157936471259,
+      "epoch": 0.7791239515377446,
+      "grad_norm": 2.2213948505531738,
+      "learning_rate": 1.645150155333103e-05,
+      "loss": 2.1559,
+      "step": 836
+    },
+    {
+      "cosine_similarity": 0.6218145914829942,
+      "epoch": 0.7800559179869525,
+      "grad_norm": 1.2852911571851666,
+      "learning_rate": 1.6444597859855024e-05,
+      "loss": 2.3255,
+      "step": 837
+    },
+    {
+      "cosine_similarity": 0.04856623412028995,
+      "epoch": 0.7809878844361603,
+      "grad_norm": 1.6419583211745767,
+      "learning_rate": 1.6437694166379013e-05,
+      "loss": 2.024,
+      "step": 838
+    },
+    {
+      "cosine_similarity": 0.3862351854558424,
+      "epoch": 0.7819198508853681,
+      "grad_norm": 1.4666020815811667,
+      "learning_rate": 1.6430790472903005e-05,
+      "loss": 2.3661,
+      "step": 839
+    },
+    {
+      "cosine_similarity": 0.45002071506526936,
+      "epoch": 0.782851817334576,
+      "grad_norm": 1.4062835691571336,
+      "learning_rate": 1.6423886779426994e-05,
+      "loss": 2.2297,
+      "step": 840
+    },
+    {
+      "cosine_similarity": 0.23711774734464636,
+      "epoch": 0.7837837837837838,
+      "grad_norm": 1.5919951914655948,
+      "learning_rate": 1.6416983085950987e-05,
+      "loss": 2.2127,
+      "step": 841
+    },
+    {
+      "cosine_similarity": 0.3222378225092349,
+      "epoch": 0.7847157502329916,
+      "grad_norm": 1.3337201015558455,
+      "learning_rate": 1.6410079392474976e-05,
+      "loss": 1.9071,
+      "step": 842
+    },
+    {
+      "cosine_similarity": 0.19239010290391143,
+      "epoch": 0.7856477166821995,
+      "grad_norm": 1.586938885698843,
+      "learning_rate": 1.6403175698998965e-05,
+      "loss": 2.2326,
+      "step": 843
+    },
+    {
+      "cosine_similarity": 0.1375715877255559,
+      "epoch": 0.7865796831314072,
+      "grad_norm": 1.388293176527509,
+      "learning_rate": 1.6396272005522954e-05,
+      "loss": 1.8949,
+      "step": 844
+    },
+    {
+      "cosine_similarity": -0.23307315983824814,
+      "epoch": 0.7875116495806151,
+      "grad_norm": 1.6616684540560074,
+      "learning_rate": 1.6389368312046946e-05,
+      "loss": 2.1413,
+      "step": 845
+    },
+    {
+      "cosine_similarity": 0.08269379887140411,
+      "epoch": 0.7884436160298229,
+      "grad_norm": 1.5117715363046869,
+      "learning_rate": 1.6382464618570935e-05,
+      "loss": 2.3194,
+      "step": 846
+    },
+    {
+      "cosine_similarity": 0.11069404851079649,
+      "epoch": 0.7893755824790307,
+      "grad_norm": 1.4122073348892812,
+      "learning_rate": 1.6375560925094928e-05,
+      "loss": 1.8398,
+      "step": 847
+    },
+    {
+      "cosine_similarity": 0.4256384739383273,
+      "epoch": 0.7903075489282386,
+      "grad_norm": 1.2808806860055273,
+      "learning_rate": 1.6368657231618917e-05,
+      "loss": 1.9182,
+      "step": 848
+    },
+    {
+      "cosine_similarity": 0.23302215063884196,
+      "epoch": 0.7912395153774464,
+      "grad_norm": 1.3668757897784194,
+      "learning_rate": 1.636175353814291e-05,
+      "loss": 1.8185,
+      "step": 849
+    },
+    {
+      "cosine_similarity": 0.01624136820746965,
+      "epoch": 0.7921714818266542,
+      "grad_norm": 1.319396049183175,
+      "learning_rate": 1.63548498446669e-05,
+      "loss": 2.1217,
+      "step": 850
+    },
+    {
+      "cosine_similarity": 0.09709587578742133,
+      "epoch": 0.7931034482758621,
+      "grad_norm": 1.2923996452164963,
+      "learning_rate": 1.6347946151190887e-05,
+      "loss": 2.4457,
+      "step": 851
+    },
+    {
+      "cosine_similarity": 0.03751642980274204,
+      "epoch": 0.7940354147250699,
+      "grad_norm": 2.3420803633276077,
+      "learning_rate": 1.634104245771488e-05,
+      "loss": 2.6317,
+      "step": 852
+    },
+    {
+      "cosine_similarity": 0.2090588499996113,
+      "epoch": 0.7949673811742777,
+      "grad_norm": 1.639967476848662,
+      "learning_rate": 1.633413876423887e-05,
+      "loss": 2.1502,
+      "step": 853
+    },
+    {
+      "cosine_similarity": 0.341409219188151,
+      "epoch": 0.7958993476234856,
+      "grad_norm": 1.5246027970656424,
+      "learning_rate": 1.6327235070762858e-05,
+      "loss": 2.5248,
+      "step": 854
+    },
+    {
+      "cosine_similarity": 0.1688687939862678,
+      "epoch": 0.7968313140726934,
+      "grad_norm": 2.228118622585711,
+      "learning_rate": 1.632033137728685e-05,
+      "loss": 2.168,
+      "step": 855
+    },
+    {
+      "cosine_similarity": 0.026292941017532143,
+      "epoch": 0.7977632805219013,
+      "grad_norm": 1.7230532071065878,
+      "learning_rate": 1.631342768381084e-05,
+      "loss": 2.4399,
+      "step": 856
+    },
+    {
+      "cosine_similarity": 0.01164136786275396,
+      "epoch": 0.798695246971109,
+      "grad_norm": 1.0148454249804013,
+      "learning_rate": 1.6306523990334832e-05,
+      "loss": 1.6779,
+      "step": 857
+    },
+    {
+      "cosine_similarity": 0.41944917013207106,
+      "epoch": 0.7996272134203168,
+      "grad_norm": 1.1409489895992901,
+      "learning_rate": 1.629962029685882e-05,
+      "loss": 1.838,
+      "step": 858
+    },
+    {
+      "cosine_similarity": 0.2508821040227736,
+      "epoch": 0.8005591798695247,
+      "grad_norm": 2.073030774393582,
+      "learning_rate": 1.6292716603382813e-05,
+      "loss": 2.5023,
+      "step": 859
+    },
+    {
+      "cosine_similarity": 0.17134578635083023,
+      "epoch": 0.8014911463187325,
+      "grad_norm": 1.450423799450184,
+      "learning_rate": 1.6285812909906802e-05,
+      "loss": 2.0599,
+      "step": 860
+    },
+    {
+      "cosine_similarity": 0.1929464768336525,
+      "epoch": 0.8024231127679403,
+      "grad_norm": 1.3118292876971445,
+      "learning_rate": 1.627890921643079e-05,
+      "loss": 2.1429,
+      "step": 861
+    },
+    {
+      "cosine_similarity": 0.05267723228791421,
+      "epoch": 0.8033550792171482,
+      "grad_norm": 1.0853183772183956,
+      "learning_rate": 1.627200552295478e-05,
+      "loss": 1.8924,
+      "step": 862
+    },
+    {
+      "cosine_similarity": 0.1539584993869459,
+      "epoch": 0.804287045666356,
+      "grad_norm": 1.369383738258699,
+      "learning_rate": 1.6265101829478773e-05,
+      "loss": 1.9977,
+      "step": 863
+    },
+    {
+      "cosine_similarity": 0.09877195802466851,
+      "epoch": 0.8052190121155638,
+      "grad_norm": 1.5520013339543346,
+      "learning_rate": 1.6258198136002762e-05,
+      "loss": 2.1485,
+      "step": 864
+    },
+    {
+      "cosine_similarity": 0.1805458114469288,
+      "epoch": 0.8061509785647717,
+      "grad_norm": 2.29981386481859,
+      "learning_rate": 1.6251294442526754e-05,
+      "loss": 2.2166,
+      "step": 865
+    },
+    {
+      "cosine_similarity": 0.11484820733044943,
+      "epoch": 0.8070829450139795,
+      "grad_norm": 1.4196682523075745,
+      "learning_rate": 1.6244390749050743e-05,
+      "loss": 2.3335,
+      "step": 866
+    },
+    {
+      "cosine_similarity": 0.11631824309902969,
+      "epoch": 0.8080149114631874,
+      "grad_norm": 1.2379874611949857,
+      "learning_rate": 1.6237487055574736e-05,
+      "loss": 2.0792,
+      "step": 867
+    },
+    {
+      "cosine_similarity": 0.13545174082900094,
+      "epoch": 0.8089468779123952,
+      "grad_norm": 1.5860198059057342,
+      "learning_rate": 1.6230583362098725e-05,
+      "loss": 2.3907,
+      "step": 868
+    },
+    {
+      "cosine_similarity": 0.1580620899926608,
+      "epoch": 0.809878844361603,
+      "grad_norm": 1.7386521448688146,
+      "learning_rate": 1.6223679668622714e-05,
+      "loss": 2.3433,
+      "step": 869
+    },
+    {
+      "cosine_similarity": 0.043838268957800144,
+      "epoch": 0.8108108108108109,
+      "grad_norm": 1.4734573525589478,
+      "learning_rate": 1.6216775975146706e-05,
+      "loss": 2.244,
+      "step": 870
+    },
+    {
+      "cosine_similarity": 0.34923825756704263,
+      "epoch": 0.8117427772600186,
+      "grad_norm": 1.5140456329322676,
+      "learning_rate": 1.6209872281670695e-05,
+      "loss": 1.9543,
+      "step": 871
+    },
+    {
+      "cosine_similarity": 0.1512672228072688,
+      "epoch": 0.8126747437092264,
+      "grad_norm": 1.3594062714131716,
+      "learning_rate": 1.6202968588194684e-05,
+      "loss": 2.1319,
+      "step": 872
+    },
+    {
+      "cosine_similarity": -0.16650843741998125,
+      "epoch": 0.8136067101584343,
+      "grad_norm": 1.7440837721973441,
+      "learning_rate": 1.6196064894718673e-05,
+      "loss": 2.2743,
+      "step": 873
+    },
+    {
+      "cosine_similarity": -0.1266511950921545,
+      "epoch": 0.8145386766076421,
+      "grad_norm": 1.1938561822797988,
+      "learning_rate": 1.6189161201242666e-05,
+      "loss": 1.8413,
+      "step": 874
+    },
+    {
+      "cosine_similarity": -0.35349189721198626,
+      "epoch": 0.8154706430568499,
+      "grad_norm": 1.3953971575538793,
+      "learning_rate": 1.6182257507766658e-05,
+      "loss": 2.2704,
+      "step": 875
+    },
+    {
+      "cosine_similarity": 0.020642772840171287,
+      "epoch": 0.8164026095060578,
+      "grad_norm": 1.3891637571657591,
+      "learning_rate": 1.6175353814290647e-05,
+      "loss": 2.1173,
+      "step": 876
+    },
+    {
+      "cosine_similarity": 0.22396387747252214,
+      "epoch": 0.8173345759552656,
+      "grad_norm": 1.41491237264784,
+      "learning_rate": 1.6168450120814636e-05,
+      "loss": 1.9165,
+      "step": 877
+    },
+    {
+      "cosine_similarity": -0.23941060746984397,
+      "epoch": 0.8182665424044734,
+      "grad_norm": 1.5554032876954806,
+      "learning_rate": 1.616154642733863e-05,
+      "loss": 2.0328,
+      "step": 878
+    },
+    {
+      "cosine_similarity": 0.23281051422583557,
+      "epoch": 0.8191985088536813,
+      "grad_norm": 1.1531284428259956,
+      "learning_rate": 1.6154642733862618e-05,
+      "loss": 1.9265,
+      "step": 879
+    },
+    {
+      "cosine_similarity": 0.0933258707575431,
+      "epoch": 0.8201304753028891,
+      "grad_norm": 1.202062180569725,
+      "learning_rate": 1.6147739040386607e-05,
+      "loss": 1.8522,
+      "step": 880
+    },
+    {
+      "cosine_similarity": 0.26898817306905043,
+      "epoch": 0.821062441752097,
+      "grad_norm": 1.6244286989195984,
+      "learning_rate": 1.61408353469106e-05,
+      "loss": 2.2275,
+      "step": 881
+    },
+    {
+      "cosine_similarity": 0.2788604775516811,
+      "epoch": 0.8219944082013048,
+      "grad_norm": 1.4557095681471557,
+      "learning_rate": 1.6133931653434588e-05,
+      "loss": 2.1883,
+      "step": 882
+    },
+    {
+      "cosine_similarity": 0.46738557508554096,
+      "epoch": 0.8229263746505125,
+      "grad_norm": 1.253384355830844,
+      "learning_rate": 1.6127027959958577e-05,
+      "loss": 2.0447,
+      "step": 883
+    },
+    {
+      "cosine_similarity": 0.36366450609864914,
+      "epoch": 0.8238583410997204,
+      "grad_norm": 1.2155142305584654,
+      "learning_rate": 1.612012426648257e-05,
+      "loss": 2.0001,
+      "step": 884
+    },
+    {
+      "cosine_similarity": 0.48302312433641414,
+      "epoch": 0.8247903075489282,
+      "grad_norm": 1.4040395425311873,
+      "learning_rate": 1.6113220573006562e-05,
+      "loss": 2.2178,
+      "step": 885
+    },
+    {
+      "cosine_similarity": 0.35894373522201034,
+      "epoch": 0.825722273998136,
+      "grad_norm": 1.4837476970727894,
+      "learning_rate": 1.610631687953055e-05,
+      "loss": 1.9525,
+      "step": 886
+    },
+    {
+      "cosine_similarity": 0.4965540690745698,
+      "epoch": 0.8266542404473439,
+      "grad_norm": 1.433585125969006,
+      "learning_rate": 1.609941318605454e-05,
+      "loss": 1.9027,
+      "step": 887
+    },
+    {
+      "cosine_similarity": 0.799203755049935,
+      "epoch": 0.8275862068965517,
+      "grad_norm": 1.813791378138503,
+      "learning_rate": 1.6092509492578533e-05,
+      "loss": 2.0574,
+      "step": 888
+    },
+    {
+      "cosine_similarity": 0.16864862636690162,
+      "epoch": 0.8285181733457595,
+      "grad_norm": 1.2733056182649907,
+      "learning_rate": 1.608560579910252e-05,
+      "loss": 2.0183,
+      "step": 889
+    },
+    {
+      "cosine_similarity": 0.2795438924828913,
+      "epoch": 0.8294501397949674,
+      "grad_norm": 1.54030218975004,
+      "learning_rate": 1.607870210562651e-05,
+      "loss": 2.1217,
+      "step": 890
+    },
+    {
+      "cosine_similarity": 0.06755616884065635,
+      "epoch": 0.8303821062441752,
+      "grad_norm": 1.4438851142985103,
+      "learning_rate": 1.60717984121505e-05,
+      "loss": 1.98,
+      "step": 891
+    },
+    {
+      "cosine_similarity": 0.20511267882073766,
+      "epoch": 0.8313140726933831,
+      "grad_norm": 1.4094433187867166,
+      "learning_rate": 1.6064894718674492e-05,
+      "loss": 1.7462,
+      "step": 892
+    },
+    {
+      "cosine_similarity": 0.16264060429141708,
+      "epoch": 0.8322460391425909,
+      "grad_norm": 1.1453286757312462,
+      "learning_rate": 1.605799102519848e-05,
+      "loss": 1.9452,
+      "step": 893
+    },
+    {
+      "cosine_similarity": 0.3527465225729456,
+      "epoch": 0.8331780055917987,
+      "grad_norm": 1.227012618822391,
+      "learning_rate": 1.6051087331722474e-05,
+      "loss": 2.04,
+      "step": 894
+    },
+    {
+      "cosine_similarity": 0.4008654771245587,
+      "epoch": 0.8341099720410066,
+      "grad_norm": 1.4474407235815152,
+      "learning_rate": 1.6044183638246463e-05,
+      "loss": 2.1987,
+      "step": 895
+    },
+    {
+      "cosine_similarity": 0.2000039775727819,
+      "epoch": 0.8350419384902144,
+      "grad_norm": 1.5181460774649096,
+      "learning_rate": 1.6037279944770455e-05,
+      "loss": 2.4344,
+      "step": 896
+    },
+    {
+      "cosine_similarity": 0.023439630294889695,
+      "epoch": 0.8359739049394221,
+      "grad_norm": 1.4280584472230928,
+      "learning_rate": 1.6030376251294444e-05,
+      "loss": 1.8123,
+      "step": 897
+    },
+    {
+      "cosine_similarity": 0.22218738116867137,
+      "epoch": 0.83690587138863,
+      "grad_norm": 1.191942641771474,
+      "learning_rate": 1.6023472557818433e-05,
+      "loss": 1.729,
+      "step": 898
+    },
+    {
+      "cosine_similarity": -0.010216304813396703,
+      "epoch": 0.8378378378378378,
+      "grad_norm": 1.3965004298398374,
+      "learning_rate": 1.6016568864342426e-05,
+      "loss": 2.2878,
+      "step": 899
+    },
+    {
+      "cosine_similarity": 0.31155736668379114,
+      "epoch": 0.8387698042870456,
+      "grad_norm": 1.7305578744370342,
+      "learning_rate": 1.6009665170866415e-05,
+      "loss": 2.3222,
+      "step": 900
+    },
+    {
+      "cosine_similarity": 0.25001742872698557,
+      "epoch": 0.8397017707362535,
+      "grad_norm": 1.5880569101659838,
+      "learning_rate": 1.6002761477390404e-05,
+      "loss": 2.2557,
+      "step": 901
+    },
+    {
+      "cosine_similarity": 0.3848655347552907,
+      "epoch": 0.8406337371854613,
+      "grad_norm": 1.3416625508700657,
+      "learning_rate": 1.5995857783914396e-05,
+      "loss": 2.0053,
+      "step": 902
+    },
+    {
+      "cosine_similarity": 0.37632088773113725,
+      "epoch": 0.8415657036346692,
+      "grad_norm": 1.2262332601385397,
+      "learning_rate": 1.5988954090438385e-05,
+      "loss": 1.7862,
+      "step": 903
+    },
+    {
+      "cosine_similarity": 0.10530713883824558,
+      "epoch": 0.842497670083877,
+      "grad_norm": 1.2125126732231806,
+      "learning_rate": 1.5982050396962378e-05,
+      "loss": 2.0917,
+      "step": 904
+    },
+    {
+      "cosine_similarity": 0.06694926939499121,
+      "epoch": 0.8434296365330848,
+      "grad_norm": 1.4106331703211281,
+      "learning_rate": 1.5975146703486367e-05,
+      "loss": 2.0446,
+      "step": 905
+    },
+    {
+      "cosine_similarity": 0.09370496881036838,
+      "epoch": 0.8443616029822927,
+      "grad_norm": 1.457629302820333,
+      "learning_rate": 1.5968243010010356e-05,
+      "loss": 2.163,
+      "step": 906
+    },
+    {
+      "cosine_similarity": 0.10932660991190403,
+      "epoch": 0.8452935694315005,
+      "grad_norm": 1.5277493053503193,
+      "learning_rate": 1.5961339316534348e-05,
+      "loss": 2.386,
+      "step": 907
+    },
+    {
+      "cosine_similarity": 0.43163275180538996,
+      "epoch": 0.8462255358807083,
+      "grad_norm": 1.2765380335480436,
+      "learning_rate": 1.5954435623058337e-05,
+      "loss": 2.1812,
+      "step": 908
+    },
+    {
+      "cosine_similarity": -0.36925542973370756,
+      "epoch": 0.8471575023299162,
+      "grad_norm": 1.3091669606418423,
+      "learning_rate": 1.5947531929582326e-05,
+      "loss": 1.8268,
+      "step": 909
+    },
+    {
+      "cosine_similarity": 0.4749515121820803,
+      "epoch": 0.848089468779124,
+      "grad_norm": 3.4149496384353104,
+      "learning_rate": 1.594062823610632e-05,
+      "loss": 1.9482,
+      "step": 910
+    },
+    {
+      "cosine_similarity": 0.47045085526463354,
+      "epoch": 0.8490214352283317,
+      "grad_norm": 1.2259893864368065,
+      "learning_rate": 1.5933724542630308e-05,
+      "loss": 2.0999,
+      "step": 911
+    },
+    {
+      "cosine_similarity": 0.4145617372897093,
+      "epoch": 0.8499534016775396,
+      "grad_norm": 1.2304109401583578,
+      "learning_rate": 1.59268208491543e-05,
+      "loss": 2.1695,
+      "step": 912
+    },
+    {
+      "cosine_similarity": 0.01574269311950758,
+      "epoch": 0.8508853681267474,
+      "grad_norm": 1.2592302682254115,
+      "learning_rate": 1.591991715567829e-05,
+      "loss": 2.152,
+      "step": 913
+    },
+    {
+      "cosine_similarity": 0.010570331221126238,
+      "epoch": 0.8518173345759553,
+      "grad_norm": 1.4020829318140295,
+      "learning_rate": 1.591301346220228e-05,
+      "loss": 1.9962,
+      "step": 914
+    },
+    {
+      "cosine_similarity": 0.0922146553689938,
+      "epoch": 0.8527493010251631,
+      "grad_norm": 4.0130584478413285,
+      "learning_rate": 1.590610976872627e-05,
+      "loss": 2.0853,
+      "step": 915
+    },
+    {
+      "cosine_similarity": 0.34463068726944035,
+      "epoch": 0.8536812674743709,
+      "grad_norm": 1.3341971056141935,
+      "learning_rate": 1.589920607525026e-05,
+      "loss": 2.451,
+      "step": 916
+    },
+    {
+      "cosine_similarity": 0.4935944704175949,
+      "epoch": 0.8546132339235788,
+      "grad_norm": 1.5059780196418247,
+      "learning_rate": 1.589230238177425e-05,
+      "loss": 2.3015,
+      "step": 917
+    },
+    {
+      "cosine_similarity": -0.05198707338629173,
+      "epoch": 0.8555452003727866,
+      "grad_norm": 1.2187456207835066,
+      "learning_rate": 1.588539868829824e-05,
+      "loss": 1.8644,
+      "step": 918
+    },
+    {
+      "cosine_similarity": -0.5017126656340588,
+      "epoch": 0.8564771668219944,
+      "grad_norm": 1.3805675173221028,
+      "learning_rate": 1.587849499482223e-05,
+      "loss": 2.3656,
+      "step": 919
+    },
+    {
+      "cosine_similarity": -0.0479676400461762,
+      "epoch": 0.8574091332712023,
+      "grad_norm": 1.2273534442233733,
+      "learning_rate": 1.5871591301346222e-05,
+      "loss": 2.0096,
+      "step": 920
+    },
+    {
+      "cosine_similarity": 0.15943776646471336,
+      "epoch": 0.8583410997204101,
+      "grad_norm": 1.2378919396363712,
+      "learning_rate": 1.586468760787021e-05,
+      "loss": 1.8654,
+      "step": 921
+    },
+    {
+      "cosine_similarity": 0.7400270050406339,
+      "epoch": 0.8592730661696178,
+      "grad_norm": 1.3731195551509825,
+      "learning_rate": 1.5857783914394204e-05,
+      "loss": 2.3446,
+      "step": 922
+    },
+    {
+      "cosine_similarity": 0.10999190810966566,
+      "epoch": 0.8602050326188257,
+      "grad_norm": 1.52022371208532,
+      "learning_rate": 1.5850880220918193e-05,
+      "loss": 2.0456,
+      "step": 923
+    },
+    {
+      "cosine_similarity": -0.015335150947161062,
+      "epoch": 0.8611369990680335,
+      "grad_norm": 1.1781233770542006,
+      "learning_rate": 1.5843976527442182e-05,
+      "loss": 1.9706,
+      "step": 924
+    },
+    {
+      "cosine_similarity": 0.20287671872554638,
+      "epoch": 0.8620689655172413,
+      "grad_norm": 1.8368136747332728,
+      "learning_rate": 1.5837072833966174e-05,
+      "loss": 2.137,
+      "step": 925
+    },
+    {
+      "cosine_similarity": 0.39950826771785974,
+      "epoch": 0.8630009319664492,
+      "grad_norm": 1.2534312262952725,
+      "learning_rate": 1.5830169140490164e-05,
+      "loss": 2.2011,
+      "step": 926
+    },
+    {
+      "cosine_similarity": -0.027470802416286497,
+      "epoch": 0.863932898415657,
+      "grad_norm": 1.1400928915047106,
+      "learning_rate": 1.5823265447014153e-05,
+      "loss": 1.8757,
+      "step": 927
+    },
+    {
+      "cosine_similarity": -0.07017712167145307,
+      "epoch": 0.8648648648648649,
+      "grad_norm": 1.4330322231668846,
+      "learning_rate": 1.5816361753538145e-05,
+      "loss": 2.3327,
+      "step": 928
+    },
+    {
+      "cosine_similarity": 0.5289497026879353,
+      "epoch": 0.8657968313140727,
+      "grad_norm": 1.7138101777087764,
+      "learning_rate": 1.5809458060062134e-05,
+      "loss": 2.0572,
+      "step": 929
+    },
+    {
+      "cosine_similarity": 0.263261363039919,
+      "epoch": 0.8667287977632805,
+      "grad_norm": 1.4728724898258605,
+      "learning_rate": 1.5802554366586126e-05,
+      "loss": 1.9729,
+      "step": 930
+    },
+    {
+      "cosine_similarity": 0.1281543578777674,
+      "epoch": 0.8676607642124884,
+      "grad_norm": 1.278240384470866,
+      "learning_rate": 1.5795650673110115e-05,
+      "loss": 2.3315,
+      "step": 931
+    },
+    {
+      "cosine_similarity": 0.258749919698254,
+      "epoch": 0.8685927306616962,
+      "grad_norm": 3.9747116926879444,
+      "learning_rate": 1.5788746979634108e-05,
+      "loss": 2.1907,
+      "step": 932
+    },
+    {
+      "cosine_similarity": 0.16496003866049314,
+      "epoch": 0.869524697110904,
+      "grad_norm": 1.2085069607189074,
+      "learning_rate": 1.5781843286158097e-05,
+      "loss": 1.9187,
+      "step": 933
+    },
+    {
+      "cosine_similarity": 0.2672883026980527,
+      "epoch": 0.8704566635601119,
+      "grad_norm": 1.3017655521489113,
+      "learning_rate": 1.5774939592682086e-05,
+      "loss": 2.0771,
+      "step": 934
+    },
+    {
+      "cosine_similarity": 0.3689129959543753,
+      "epoch": 0.8713886300093197,
+      "grad_norm": 1.1526337961925919,
+      "learning_rate": 1.5768035899206075e-05,
+      "loss": 2.0359,
+      "step": 935
+    },
+    {
+      "cosine_similarity": 0.3916730697598186,
+      "epoch": 0.8723205964585274,
+      "grad_norm": 1.7179831573170172,
+      "learning_rate": 1.5761132205730067e-05,
+      "loss": 2.3032,
+      "step": 936
+    },
+    {
+      "cosine_similarity": -0.002280184744569954,
+      "epoch": 0.8732525629077353,
+      "grad_norm": 1.4501692250348002,
+      "learning_rate": 1.5754228512254056e-05,
+      "loss": 2.1611,
+      "step": 937
+    },
+    {
+      "cosine_similarity": 0.20749414132977845,
+      "epoch": 0.8741845293569431,
+      "grad_norm": 1.3695833601375023,
+      "learning_rate": 1.5747324818778046e-05,
+      "loss": 1.798,
+      "step": 938
+    },
+    {
+      "cosine_similarity": 0.39672754271461047,
+      "epoch": 0.875116495806151,
+      "grad_norm": 1.263971498208846,
+      "learning_rate": 1.5740421125302038e-05,
+      "loss": 2.1095,
+      "step": 939
+    },
+    {
+      "cosine_similarity": 0.32712718520185835,
+      "epoch": 0.8760484622553588,
+      "grad_norm": 1.4486764761680664,
+      "learning_rate": 1.5733517431826027e-05,
+      "loss": 2.2822,
+      "step": 940
+    },
+    {
+      "cosine_similarity": 0.4434852793821396,
+      "epoch": 0.8769804287045666,
+      "grad_norm": 1.7164056709213014,
+      "learning_rate": 1.572661373835002e-05,
+      "loss": 2.2771,
+      "step": 941
+    },
+    {
+      "cosine_similarity": 0.35522780741574705,
+      "epoch": 0.8779123951537745,
+      "grad_norm": 1.2557970129097427,
+      "learning_rate": 1.571971004487401e-05,
+      "loss": 2.0784,
+      "step": 942
+    },
+    {
+      "cosine_similarity": 0.19273670242871738,
+      "epoch": 0.8788443616029823,
+      "grad_norm": 1.2607418403505704,
+      "learning_rate": 1.5712806351398e-05,
+      "loss": 2.1141,
+      "step": 943
+    },
+    {
+      "cosine_similarity": 0.07086938731881223,
+      "epoch": 0.8797763280521901,
+      "grad_norm": 1.4707760732075879,
+      "learning_rate": 1.570590265792199e-05,
+      "loss": 2.1383,
+      "step": 944
+    },
+    {
+      "cosine_similarity": 0.41046656871728476,
+      "epoch": 0.880708294501398,
+      "grad_norm": 1.4004981738868982,
+      "learning_rate": 1.569899896444598e-05,
+      "loss": 1.7198,
+      "step": 945
+    },
+    {
+      "cosine_similarity": 0.6614427359645152,
+      "epoch": 0.8816402609506058,
+      "grad_norm": 1.604197986384341,
+      "learning_rate": 1.5692095270969968e-05,
+      "loss": 2.1607,
+      "step": 946
+    },
+    {
+      "cosine_similarity": 0.2339834316996676,
+      "epoch": 0.8825722273998136,
+      "grad_norm": 1.1532419706465582,
+      "learning_rate": 1.568519157749396e-05,
+      "loss": 2.1792,
+      "step": 947
+    },
+    {
+      "cosine_similarity": -0.10710775896165622,
+      "epoch": 0.8835041938490215,
+      "grad_norm": 2.233977343371551,
+      "learning_rate": 1.567828788401795e-05,
+      "loss": 2.4003,
+      "step": 948
+    },
+    {
+      "cosine_similarity": 0.26625428343464164,
+      "epoch": 0.8844361602982292,
+      "grad_norm": 1.1186085190116408,
+      "learning_rate": 1.5671384190541942e-05,
+      "loss": 1.9305,
+      "step": 949
+    },
+    {
+      "cosine_similarity": 0.2640265878830857,
+      "epoch": 0.8853681267474371,
+      "grad_norm": 1.526496646940089,
+      "learning_rate": 1.566448049706593e-05,
+      "loss": 2.3089,
+      "step": 950
+    },
+    {
+      "cosine_similarity": 0.3176813320739606,
+      "epoch": 0.8863000931966449,
+      "grad_norm": 1.499300139086924,
+      "learning_rate": 1.5657576803589923e-05,
+      "loss": 2.0112,
+      "step": 951
+    },
+    {
+      "cosine_similarity": 0.33870304022858183,
+      "epoch": 0.8872320596458527,
+      "grad_norm": 1.7153597241915572,
+      "learning_rate": 1.5650673110113912e-05,
+      "loss": 2.2162,
+      "step": 952
+    },
+    {
+      "cosine_similarity": 0.08405015595166841,
+      "epoch": 0.8881640260950606,
+      "grad_norm": 1.2325862770470175,
+      "learning_rate": 1.56437694166379e-05,
+      "loss": 2.0964,
+      "step": 953
+    },
+    {
+      "cosine_similarity": 0.21030216539707852,
+      "epoch": 0.8890959925442684,
+      "grad_norm": 1.4343845723170632,
+      "learning_rate": 1.5636865723161894e-05,
+      "loss": 2.0572,
+      "step": 954
+    },
+    {
+      "cosine_similarity": 0.023526752623093262,
+      "epoch": 0.8900279589934762,
+      "grad_norm": 1.3488229182596545,
+      "learning_rate": 1.5629962029685883e-05,
+      "loss": 2.0449,
+      "step": 955
+    },
+    {
+      "cosine_similarity": 0.2082300039532623,
+      "epoch": 0.8909599254426841,
+      "grad_norm": 1.4227316079089132,
+      "learning_rate": 1.5623058336209872e-05,
+      "loss": 2.058,
+      "step": 956
+    },
+    {
+      "cosine_similarity": -0.006528106006805303,
+      "epoch": 0.8918918918918919,
+      "grad_norm": 1.4673600610786017,
+      "learning_rate": 1.5616154642733864e-05,
+      "loss": 2.2987,
+      "step": 957
+    },
+    {
+      "cosine_similarity": 0.2884994145088987,
+      "epoch": 0.8928238583410997,
+      "grad_norm": 1.311602115388731,
+      "learning_rate": 1.5609250949257853e-05,
+      "loss": 2.316,
+      "step": 958
+    },
+    {
+      "cosine_similarity": 0.16884966828812203,
+      "epoch": 0.8937558247903076,
+      "grad_norm": 1.1467394949413703,
+      "learning_rate": 1.5602347255781846e-05,
+      "loss": 1.6958,
+      "step": 959
+    },
+    {
+      "cosine_similarity": 0.08054344713333157,
+      "epoch": 0.8946877912395154,
+      "grad_norm": 1.2900192734726486,
+      "learning_rate": 1.5595443562305835e-05,
+      "loss": 2.4569,
+      "step": 960
+    },
+    {
+      "cosine_similarity": -0.017750096622361294,
+      "epoch": 0.8956197576887233,
+      "grad_norm": 1.7812466460368521,
+      "learning_rate": 1.5588539868829827e-05,
+      "loss": 2.2927,
+      "step": 961
+    },
+    {
+      "cosine_similarity": 0.03958706379941647,
+      "epoch": 0.896551724137931,
+      "grad_norm": 1.3072627432565784,
+      "learning_rate": 1.5581636175353816e-05,
+      "loss": 2.1663,
+      "step": 962
+    },
+    {
+      "cosine_similarity": -0.03832512302205757,
+      "epoch": 0.8974836905871388,
+      "grad_norm": 1.5218564642803227,
+      "learning_rate": 1.5574732481877805e-05,
+      "loss": 1.9296,
+      "step": 963
+    },
+    {
+      "cosine_similarity": 0.11165169038482241,
+      "epoch": 0.8984156570363467,
+      "grad_norm": 1.4403005417739658,
+      "learning_rate": 1.5567828788401794e-05,
+      "loss": 1.7115,
+      "step": 964
+    },
+    {
+      "cosine_similarity": 0.22854207295972742,
+      "epoch": 0.8993476234855545,
+      "grad_norm": 1.5408946014868454,
+      "learning_rate": 1.5560925094925787e-05,
+      "loss": 2.0445,
+      "step": 965
+    },
+    {
+      "cosine_similarity": 0.044685529299604156,
+      "epoch": 0.9002795899347623,
+      "grad_norm": 1.1506213667331866,
+      "learning_rate": 1.5554021401449776e-05,
+      "loss": 1.8034,
+      "step": 966
+    },
+    {
+      "cosine_similarity": 0.02023867044564871,
+      "epoch": 0.9012115563839702,
+      "grad_norm": 1.149750436874357,
+      "learning_rate": 1.5547117707973768e-05,
+      "loss": 2.133,
+      "step": 967
+    },
+    {
+      "cosine_similarity": -0.12043354029455199,
+      "epoch": 0.902143522833178,
+      "grad_norm": 1.5151808591999492,
+      "learning_rate": 1.5540214014497757e-05,
+      "loss": 2.0671,
+      "step": 968
+    },
+    {
+      "cosine_similarity": -0.0080207864545536,
+      "epoch": 0.9030754892823858,
+      "grad_norm": 1.6293272759515234,
+      "learning_rate": 1.553331032102175e-05,
+      "loss": 2.3009,
+      "step": 969
+    },
+    {
+      "cosine_similarity": 0.39585228219259017,
+      "epoch": 0.9040074557315937,
+      "grad_norm": 1.5085735920610301,
+      "learning_rate": 1.552640662754574e-05,
+      "loss": 2.2593,
+      "step": 970
+    },
+    {
+      "cosine_similarity": 0.1531588164570011,
+      "epoch": 0.9049394221808015,
+      "grad_norm": 1.9006640421249312,
+      "learning_rate": 1.5519502934069728e-05,
+      "loss": 2.0689,
+      "step": 971
+    },
+    {
+      "cosine_similarity": 0.21244030042676493,
+      "epoch": 0.9058713886300093,
+      "grad_norm": 1.5325297061619638,
+      "learning_rate": 1.551259924059372e-05,
+      "loss": 2.1917,
+      "step": 972
+    },
+    {
+      "cosine_similarity": -0.16392196847491575,
+      "epoch": 0.9068033550792172,
+      "grad_norm": 1.4378110072108554,
+      "learning_rate": 1.550569554711771e-05,
+      "loss": 1.8231,
+      "step": 973
+    },
+    {
+      "cosine_similarity": 0.7351953603239533,
+      "epoch": 0.907735321528425,
+      "grad_norm": 1.6067521744354043,
+      "learning_rate": 1.54987918536417e-05,
+      "loss": 1.9521,
+      "step": 974
+    },
+    {
+      "cosine_similarity": -0.16116742754088395,
+      "epoch": 0.9086672879776329,
+      "grad_norm": 1.807337617739289,
+      "learning_rate": 1.5491888160165687e-05,
+      "loss": 2.3258,
+      "step": 975
+    },
+    {
+      "cosine_similarity": 0.3626839139313336,
+      "epoch": 0.9095992544268406,
+      "grad_norm": 1.357088093731775,
+      "learning_rate": 1.548498446668968e-05,
+      "loss": 2.2351,
+      "step": 976
+    },
+    {
+      "cosine_similarity": 0.3526947565364157,
+      "epoch": 0.9105312208760484,
+      "grad_norm": 1.1009104215950547,
+      "learning_rate": 1.5478080773213672e-05,
+      "loss": 2.038,
+      "step": 977
+    },
+    {
+      "cosine_similarity": 0.11796063918623183,
+      "epoch": 0.9114631873252563,
+      "grad_norm": 1.6703366928668235,
+      "learning_rate": 1.547117707973766e-05,
+      "loss": 1.4994,
+      "step": 978
+    },
+    {
+      "cosine_similarity": 0.10951535482769233,
+      "epoch": 0.9123951537744641,
+      "grad_norm": 1.4921104012424855,
+      "learning_rate": 1.5464273386261654e-05,
+      "loss": 1.8274,
+      "step": 979
+    },
+    {
+      "cosine_similarity": 0.45776704249135597,
+      "epoch": 0.9133271202236719,
+      "grad_norm": 1.4435113612600219,
+      "learning_rate": 1.5457369692785643e-05,
+      "loss": 2.0379,
+      "step": 980
+    },
+    {
+      "cosine_similarity": 0.13201208089932265,
+      "epoch": 0.9142590866728798,
+      "grad_norm": 1.4217330433223012,
+      "learning_rate": 1.5450465999309632e-05,
+      "loss": 2.1895,
+      "step": 981
+    },
+    {
+      "cosine_similarity": 0.047953885488511046,
+      "epoch": 0.9151910531220876,
+      "grad_norm": 1.2583927282382459,
+      "learning_rate": 1.544356230583362e-05,
+      "loss": 2.3245,
+      "step": 982
+    },
+    {
+      "cosine_similarity": 0.23669009132344,
+      "epoch": 0.9161230195712954,
+      "grad_norm": 1.7259565902451957,
+      "learning_rate": 1.5436658612357613e-05,
+      "loss": 2.0423,
+      "step": 983
+    },
+    {
+      "cosine_similarity": 0.504278420811588,
+      "epoch": 0.9170549860205033,
+      "grad_norm": 1.2659885422191661,
+      "learning_rate": 1.5429754918881602e-05,
+      "loss": 2.022,
+      "step": 984
+    },
+    {
+      "cosine_similarity": 0.11095003159548293,
+      "epoch": 0.9179869524697111,
+      "grad_norm": 1.4279649835455934,
+      "learning_rate": 1.542285122540559e-05,
+      "loss": 2.0724,
+      "step": 985
+    },
+    {
+      "cosine_similarity": 0.19157042047927392,
+      "epoch": 0.918918918918919,
+      "grad_norm": 1.7228827289138928,
+      "learning_rate": 1.5415947531929584e-05,
+      "loss": 2.4961,
+      "step": 986
+    },
+    {
+      "cosine_similarity": 0.45073121473411815,
+      "epoch": 0.9198508853681268,
+      "grad_norm": 1.2483786461548951,
+      "learning_rate": 1.5409043838453576e-05,
+      "loss": 1.8313,
+      "step": 987
+    },
+    {
+      "cosine_similarity": 0.3767487742603566,
+      "epoch": 0.9207828518173345,
+      "grad_norm": 1.1102643685200004,
+      "learning_rate": 1.5402140144977565e-05,
+      "loss": 2.2281,
+      "step": 988
+    },
+    {
+      "cosine_similarity": 0.10663302383174238,
+      "epoch": 0.9217148182665424,
+      "grad_norm": 1.393722358429717,
+      "learning_rate": 1.5395236451501554e-05,
+      "loss": 2.126,
+      "step": 989
+    },
+    {
+      "cosine_similarity": 0.30598579094621287,
+      "epoch": 0.9226467847157502,
+      "grad_norm": 1.2486681452053576,
+      "learning_rate": 1.5388332758025547e-05,
+      "loss": 1.9196,
+      "step": 990
+    },
+    {
+      "cosine_similarity": 0.43450930640744156,
+      "epoch": 0.923578751164958,
+      "grad_norm": 1.4310056458692497,
+      "learning_rate": 1.5381429064549536e-05,
+      "loss": 2.29,
+      "step": 991
+    },
+    {
+      "cosine_similarity": 0.14548334535528432,
+      "epoch": 0.9245107176141659,
+      "grad_norm": 1.3904504513307347,
+      "learning_rate": 1.5374525371073525e-05,
+      "loss": 2.2432,
+      "step": 992
+    },
+    {
+      "cosine_similarity": 0.475716573778351,
+      "epoch": 0.9254426840633737,
+      "grad_norm": 1.9293057792571173,
+      "learning_rate": 1.5367621677597514e-05,
+      "loss": 2.3344,
+      "step": 993
+    },
+    {
+      "cosine_similarity": 0.4410959436015716,
+      "epoch": 0.9263746505125815,
+      "grad_norm": 1.2640593553776058,
+      "learning_rate": 1.5360717984121506e-05,
+      "loss": 2.2224,
+      "step": 994
+    },
+    {
+      "cosine_similarity": 0.28150992310427014,
+      "epoch": 0.9273066169617894,
+      "grad_norm": 1.7123013201772173,
+      "learning_rate": 1.5353814290645495e-05,
+      "loss": 2.0748,
+      "step": 995
+    },
+    {
+      "cosine_similarity": 0.22679138473220403,
+      "epoch": 0.9282385834109972,
+      "grad_norm": 1.5034715831940557,
+      "learning_rate": 1.5346910597169488e-05,
+      "loss": 2.137,
+      "step": 996
+    },
+    {
+      "cosine_similarity": 0.19974429634201185,
+      "epoch": 0.9291705498602051,
+      "grad_norm": 1.9949429738954663,
+      "learning_rate": 1.5340006903693477e-05,
+      "loss": 2.1561,
+      "step": 997
+    },
+    {
+      "cosine_similarity": 0.23666091963641264,
+      "epoch": 0.9301025163094129,
+      "grad_norm": 1.6657349868455449,
+      "learning_rate": 1.533310321021747e-05,
+      "loss": 2.1507,
+      "step": 998
+    },
+    {
+      "cosine_similarity": 0.024547194861113352,
+      "epoch": 0.9310344827586207,
+      "grad_norm": 1.40199393276661,
+      "learning_rate": 1.5326199516741458e-05,
+      "loss": 1.859,
+      "step": 999
+    },
+    {
+      "cosine_similarity": -0.020844241402873994,
+      "epoch": 0.9319664492078286,
+      "grad_norm": 1.540836798236812,
+      "learning_rate": 1.5319295823265447e-05,
+      "loss": 2.3447,
+      "step": 1000
+    },
+    {
+      "cosine_similarity": 0.26515947277894386,
+      "epoch": 0.9328984156570364,
+      "grad_norm": 1.3008341748472811,
+      "learning_rate": 1.531239212978944e-05,
+      "loss": 2.0752,
+      "step": 1001
+    },
+    {
+      "cosine_similarity": 0.2761568660616188,
+      "epoch": 0.9338303821062441,
+      "grad_norm": 1.3657945543015357,
+      "learning_rate": 1.530548843631343e-05,
+      "loss": 2.1326,
+      "step": 1002
+    },
+    {
+      "cosine_similarity": 0.15822633131601777,
+      "epoch": 0.934762348555452,
+      "grad_norm": 1.1529237973583473,
+      "learning_rate": 1.5298584742837418e-05,
+      "loss": 1.9756,
+      "step": 1003
+    },
+    {
+      "cosine_similarity": 0.4099494606781677,
+      "epoch": 0.9356943150046598,
+      "grad_norm": 1.2721489399336918,
+      "learning_rate": 1.529168104936141e-05,
+      "loss": 2.3852,
+      "step": 1004
+    },
+    {
+      "cosine_similarity": -0.08159895916200811,
+      "epoch": 0.9366262814538676,
+      "grad_norm": 1.4294942627529115,
+      "learning_rate": 1.52847773558854e-05,
+      "loss": 2.1792,
+      "step": 1005
+    },
+    {
+      "cosine_similarity": 0.1796434708373194,
+      "epoch": 0.9375582479030755,
+      "grad_norm": 2.370336240422075,
+      "learning_rate": 1.527787366240939e-05,
+      "loss": 2.2547,
+      "step": 1006
+    },
+    {
+      "cosine_similarity": 0.33435725918650366,
+      "epoch": 0.9384902143522833,
+      "grad_norm": 1.3034666264118728,
+      "learning_rate": 1.527096996893338e-05,
+      "loss": 2.1768,
+      "step": 1007
+    },
+    {
+      "cosine_similarity": 0.532453163337686,
+      "epoch": 0.9394221808014911,
+      "grad_norm": 1.2005448745826834,
+      "learning_rate": 1.526406627545737e-05,
+      "loss": 2.4203,
+      "step": 1008
+    },
+    {
+      "cosine_similarity": 0.1117233306475784,
+      "epoch": 0.940354147250699,
+      "grad_norm": 1.1718925214715077,
+      "learning_rate": 1.5257162581981362e-05,
+      "loss": 1.7646,
+      "step": 1009
+    },
+    {
+      "cosine_similarity": 0.00012727363785703278,
+      "epoch": 0.9412861136999068,
+      "grad_norm": 1.0966724295012191,
+      "learning_rate": 1.5250258888505351e-05,
+      "loss": 1.7183,
+      "step": 1010
+    },
+    {
+      "cosine_similarity": 0.20700001816233868,
+      "epoch": 0.9422180801491147,
+      "grad_norm": 1.358899891501129,
+      "learning_rate": 1.5243355195029342e-05,
+      "loss": 1.6995,
+      "step": 1011
+    },
+    {
+      "cosine_similarity": 0.6800364137511705,
+      "epoch": 0.9431500465983225,
+      "grad_norm": 1.1987478608317146,
+      "learning_rate": 1.5236451501553333e-05,
+      "loss": 2.6013,
+      "step": 1012
+    },
+    {
+      "cosine_similarity": 0.08511546634225985,
+      "epoch": 0.9440820130475303,
+      "grad_norm": 1.0318352896054572,
+      "learning_rate": 1.5229547808077323e-05,
+      "loss": 1.6667,
+      "step": 1013
+    },
+    {
+      "cosine_similarity": 0.26098414221956057,
+      "epoch": 0.9450139794967382,
+      "grad_norm": 1.6941285185065362,
+      "learning_rate": 1.5222644114601312e-05,
+      "loss": 2.15,
+      "step": 1014
+    },
+    {
+      "cosine_similarity": 0.38155380880027184,
+      "epoch": 0.9459459459459459,
+      "grad_norm": 1.3472935086140394,
+      "learning_rate": 1.5215740421125303e-05,
+      "loss": 2.2054,
+      "step": 1015
+    },
+    {
+      "cosine_similarity": 0.5514421086112882,
+      "epoch": 0.9468779123951537,
+      "grad_norm": 1.390307615705129,
+      "learning_rate": 1.5208836727649294e-05,
+      "loss": 1.9998,
+      "step": 1016
+    },
+    {
+      "cosine_similarity": 0.16243316267322008,
+      "epoch": 0.9478098788443616,
+      "grad_norm": 1.439184063657177,
+      "learning_rate": 1.5201933034173285e-05,
+      "loss": 2.0105,
+      "step": 1017
+    },
+    {
+      "cosine_similarity": 0.13741556283607398,
+      "epoch": 0.9487418452935694,
+      "grad_norm": 1.2948604821230842,
+      "learning_rate": 1.5195029340697274e-05,
+      "loss": 1.9852,
+      "step": 1018
+    },
+    {
+      "cosine_similarity": 0.06803636291217216,
+      "epoch": 0.9496738117427772,
+      "grad_norm": 1.6094915577510391,
+      "learning_rate": 1.5188125647221266e-05,
+      "loss": 2.1481,
+      "step": 1019
+    },
+    {
+      "cosine_similarity": 0.47226928998255274,
+      "epoch": 0.9506057781919851,
+      "grad_norm": 1.4867540862328323,
+      "learning_rate": 1.5181221953745255e-05,
+      "loss": 2.1168,
+      "step": 1020
+    },
+    {
+      "cosine_similarity": 0.07809476405815864,
+      "epoch": 0.9515377446411929,
+      "grad_norm": 1.2490852190440302,
+      "learning_rate": 1.5174318260269246e-05,
+      "loss": 2.1854,
+      "step": 1021
+    },
+    {
+      "cosine_similarity": -0.1307070927839768,
+      "epoch": 0.9524697110904008,
+      "grad_norm": 1.4090359319951997,
+      "learning_rate": 1.5167414566793235e-05,
+      "loss": 2.1732,
+      "step": 1022
+    },
+    {
+      "cosine_similarity": 0.06638039468834948,
+      "epoch": 0.9534016775396086,
+      "grad_norm": 1.328641597069533,
+      "learning_rate": 1.5160510873317227e-05,
+      "loss": 2.2125,
+      "step": 1023
+    },
+    {
+      "cosine_similarity": 0.19361345235788968,
+      "epoch": 0.9543336439888164,
+      "grad_norm": 1.5432117409760144,
+      "learning_rate": 1.5153607179841216e-05,
+      "loss": 2.135,
+      "step": 1024
+    },
+    {
+      "cosine_similarity": 0.21758196158576668,
+      "epoch": 0.9552656104380243,
+      "grad_norm": 1.2702922669137466,
+      "learning_rate": 1.5146703486365207e-05,
+      "loss": 2.2508,
+      "step": 1025
+    },
+    {
+      "cosine_similarity": 0.17232256361004006,
+      "epoch": 0.9561975768872321,
+      "grad_norm": 1.1208554377178244,
+      "learning_rate": 1.5139799792889196e-05,
+      "loss": 2.08,
+      "step": 1026
+    },
+    {
+      "cosine_similarity": -0.06598656892728635,
+      "epoch": 0.9571295433364398,
+      "grad_norm": 1.3445064182408677,
+      "learning_rate": 1.5132896099413189e-05,
+      "loss": 2.2001,
+      "step": 1027
+    },
+    {
+      "cosine_similarity": 0.19032629587049016,
+      "epoch": 0.9580615097856477,
+      "grad_norm": 1.4064859243420158,
+      "learning_rate": 1.5125992405937178e-05,
+      "loss": 2.1859,
+      "step": 1028
+    },
+    {
+      "cosine_similarity": -0.09004582357658053,
+      "epoch": 0.9589934762348555,
+      "grad_norm": 1.1517093094757542,
+      "learning_rate": 1.5119088712461167e-05,
+      "loss": 1.9517,
+      "step": 1029
+    },
+    {
+      "cosine_similarity": 0.18783428493713367,
+      "epoch": 0.9599254426840633,
+      "grad_norm": 1.4189644229952443,
+      "learning_rate": 1.5112185018985159e-05,
+      "loss": 2.0366,
+      "step": 1030
+    },
+    {
+      "cosine_similarity": 0.031081570945288007,
+      "epoch": 0.9608574091332712,
+      "grad_norm": 1.291240981403698,
+      "learning_rate": 1.510528132550915e-05,
+      "loss": 2.1385,
+      "step": 1031
+    },
+    {
+      "cosine_similarity": 0.0203616656188463,
+      "epoch": 0.961789375582479,
+      "grad_norm": 1.3450715852339998,
+      "learning_rate": 1.5098377632033139e-05,
+      "loss": 2.1987,
+      "step": 1032
+    },
+    {
+      "cosine_similarity": 0.28370517035720466,
+      "epoch": 0.9627213420316869,
+      "grad_norm": 1.4888821997050758,
+      "learning_rate": 1.5091473938557128e-05,
+      "loss": 2.075,
+      "step": 1033
+    },
+    {
+      "cosine_similarity": 0.2497445718479643,
+      "epoch": 0.9636533084808947,
+      "grad_norm": 2.0190258826477705,
+      "learning_rate": 1.508457024508112e-05,
+      "loss": 2.0631,
+      "step": 1034
+    },
+    {
+      "cosine_similarity": 0.21395325553460917,
+      "epoch": 0.9645852749301025,
+      "grad_norm": 1.4844295075817002,
+      "learning_rate": 1.5077666551605111e-05,
+      "loss": 2.3477,
+      "step": 1035
+    },
+    {
+      "cosine_similarity": 0.4822750282542871,
+      "epoch": 0.9655172413793104,
+      "grad_norm": 3.4702090711376474,
+      "learning_rate": 1.50707628581291e-05,
+      "loss": 2.1245,
+      "step": 1036
+    },
+    {
+      "cosine_similarity": 0.1888264405603915,
+      "epoch": 0.9664492078285182,
+      "grad_norm": 1.3083047199907454,
+      "learning_rate": 1.5063859164653089e-05,
+      "loss": 1.7439,
+      "step": 1037
+    },
+    {
+      "cosine_similarity": -0.02591202993666662,
+      "epoch": 0.967381174277726,
+      "grad_norm": 1.0384047623650035,
+      "learning_rate": 1.5056955471177081e-05,
+      "loss": 1.8417,
+      "step": 1038
+    },
+    {
+      "cosine_similarity": -0.02280993084485188,
+      "epoch": 0.9683131407269339,
+      "grad_norm": 1.5613288870708972,
+      "learning_rate": 1.505005177770107e-05,
+      "loss": 2.1609,
+      "step": 1039
+    },
+    {
+      "cosine_similarity": -0.011039291757467495,
+      "epoch": 0.9692451071761417,
+      "grad_norm": 1.9966463342581167,
+      "learning_rate": 1.5043148084225061e-05,
+      "loss": 1.7426,
+      "step": 1040
+    },
+    {
+      "cosine_similarity": 0.1281726357435528,
+      "epoch": 0.9701770736253494,
+      "grad_norm": 1.2995996945117685,
+      "learning_rate": 1.5036244390749054e-05,
+      "loss": 1.9734,
+      "step": 1041
+    },
+    {
+      "cosine_similarity": 0.17271340132382698,
+      "epoch": 0.9711090400745573,
+      "grad_norm": 1.808080482318741,
+      "learning_rate": 1.5029340697273043e-05,
+      "loss": 2.2947,
+      "step": 1042
+    },
+    {
+      "cosine_similarity": -0.022452882232605997,
+      "epoch": 0.9720410065237651,
+      "grad_norm": 1.7096685917380159,
+      "learning_rate": 1.5022437003797032e-05,
+      "loss": 2.2236,
+      "step": 1043
+    },
+    {
+      "cosine_similarity": 0.22836577577137226,
+      "epoch": 0.972972972972973,
+      "grad_norm": 1.6022860717333487,
+      "learning_rate": 1.5015533310321023e-05,
+      "loss": 1.9525,
+      "step": 1044
+    },
+    {
+      "cosine_similarity": 0.07704452050539021,
+      "epoch": 0.9739049394221808,
+      "grad_norm": 1.2937264324007687,
+      "learning_rate": 1.5008629616845013e-05,
+      "loss": 2.2589,
+      "step": 1045
+    },
+    {
+      "cosine_similarity": 0.08124701035323081,
+      "epoch": 0.9748369058713886,
+      "grad_norm": 1.3388450618728054,
+      "learning_rate": 1.5001725923369004e-05,
+      "loss": 2.1924,
+      "step": 1046
+    },
+    {
+      "cosine_similarity": -0.07779909580456801,
+      "epoch": 0.9757688723205965,
+      "grad_norm": 1.703440576032822,
+      "learning_rate": 1.4994822229892993e-05,
+      "loss": 2.6841,
+      "step": 1047
+    },
+    {
+      "cosine_similarity": 0.20357902586200455,
+      "epoch": 0.9767008387698043,
+      "grad_norm": 1.2290993455019885,
+      "learning_rate": 1.4987918536416984e-05,
+      "loss": 2.003,
+      "step": 1048
+    },
+    {
+      "cosine_similarity": 0.10634444812442158,
+      "epoch": 0.9776328052190121,
+      "grad_norm": 1.5080214457431809,
+      "learning_rate": 1.4981014842940974e-05,
+      "loss": 2.3933,
+      "step": 1049
+    },
+    {
+      "cosine_similarity": 0.0646212952791913,
+      "epoch": 0.97856477166822,
+      "grad_norm": 1.6745678058414704,
+      "learning_rate": 1.4974111149464965e-05,
+      "loss": 2.0809,
+      "step": 1050
+    },
+    {
+      "cosine_similarity": -0.17135341584118274,
+      "epoch": 0.9794967381174278,
+      "grad_norm": 1.631618712497268,
+      "learning_rate": 1.4967207455988954e-05,
+      "loss": 2.041,
+      "step": 1051
+    },
+    {
+      "cosine_similarity": 0.19354104683899312,
+      "epoch": 0.9804287045666356,
+      "grad_norm": 1.3948999895809806,
+      "learning_rate": 1.4960303762512947e-05,
+      "loss": 1.9146,
+      "step": 1052
+    },
+    {
+      "cosine_similarity": 0.2566514223172982,
+      "epoch": 0.9813606710158435,
+      "grad_norm": 1.458687169774932,
+      "learning_rate": 1.4953400069036936e-05,
+      "loss": 2.2098,
+      "step": 1053
+    },
+    {
+      "cosine_similarity": 0.01089511470300713,
+      "epoch": 0.9822926374650512,
+      "grad_norm": 1.239051033317669,
+      "learning_rate": 1.4946496375560926e-05,
+      "loss": 2.1534,
+      "step": 1054
+    },
+    {
+      "cosine_similarity": -0.05075415905057637,
+      "epoch": 0.983224603914259,
+      "grad_norm": 1.1318060336123361,
+      "learning_rate": 1.4939592682084915e-05,
+      "loss": 2.295,
+      "step": 1055
+    },
+    {
+      "cosine_similarity": 0.18875631976332405,
+      "epoch": 0.9841565703634669,
+      "grad_norm": 1.0717975572009704,
+      "learning_rate": 1.4932688988608908e-05,
+      "loss": 1.8584,
+      "step": 1056
+    },
+    {
+      "cosine_similarity": 0.43939499115362907,
+      "epoch": 0.9850885368126747,
+      "grad_norm": 1.5525470635493885,
+      "learning_rate": 1.4925785295132897e-05,
+      "loss": 2.455,
+      "step": 1057
+    },
+    {
+      "cosine_similarity": 0.35755342429100373,
+      "epoch": 0.9860205032618826,
+      "grad_norm": 1.4892712229577392,
+      "learning_rate": 1.4918881601656888e-05,
+      "loss": 2.0982,
+      "step": 1058
+    },
+    {
+      "cosine_similarity": 0.09650883480344401,
+      "epoch": 0.9869524697110904,
+      "grad_norm": 1.1512067932659096,
+      "learning_rate": 1.4911977908180878e-05,
+      "loss": 2.4642,
+      "step": 1059
+    },
+    {
+      "cosine_similarity": 0.19244295076365037,
+      "epoch": 0.9878844361602982,
+      "grad_norm": 1.6252442366208697,
+      "learning_rate": 1.4905074214704869e-05,
+      "loss": 1.9725,
+      "step": 1060
+    },
+    {
+      "cosine_similarity": 0.23461164920897795,
+      "epoch": 0.9888164026095061,
+      "grad_norm": 1.2442273894804083,
+      "learning_rate": 1.4898170521228858e-05,
+      "loss": 2.1364,
+      "step": 1061
+    },
+    {
+      "cosine_similarity": 0.4246035013608187,
+      "epoch": 0.9897483690587139,
+      "grad_norm": 1.6276026783563797,
+      "learning_rate": 1.4891266827752849e-05,
+      "loss": 2.0837,
+      "step": 1062
+    },
+    {
+      "cosine_similarity": 0.3298847330711108,
+      "epoch": 0.9906803355079217,
+      "grad_norm": 1.1619620170908622,
+      "learning_rate": 1.488436313427684e-05,
+      "loss": 1.8184,
+      "step": 1063
+    },
+    {
+      "cosine_similarity": 0.05037092135489437,
+      "epoch": 0.9916123019571296,
+      "grad_norm": 1.4069101034583071,
+      "learning_rate": 1.487745944080083e-05,
+      "loss": 2.2549,
+      "step": 1064
+    },
+    {
+      "cosine_similarity": 0.17294446988466272,
+      "epoch": 0.9925442684063374,
+      "grad_norm": 1.0897605578724396,
+      "learning_rate": 1.487055574732482e-05,
+      "loss": 1.6499,
+      "step": 1065
+    },
+    {
+      "cosine_similarity": 0.23246590760030097,
+      "epoch": 0.9934762348555451,
+      "grad_norm": 1.2669725641476643,
+      "learning_rate": 1.486365205384881e-05,
+      "loss": 2.0854,
+      "step": 1066
+    },
+    {
+      "cosine_similarity": 0.05849097650157266,
+      "epoch": 0.994408201304753,
+      "grad_norm": 1.3172327058156597,
+      "learning_rate": 1.4856748360372801e-05,
+      "loss": 1.834,
+      "step": 1067
+    },
+    {
+      "cosine_similarity": 0.17495233123874596,
+      "epoch": 0.9953401677539608,
+      "grad_norm": 1.4306923589330696,
+      "learning_rate": 1.4849844666896792e-05,
+      "loss": 2.2323,
+      "step": 1068
+    },
+    {
+      "cosine_similarity": 0.04289576812140693,
+      "epoch": 0.9962721342031687,
+      "grad_norm": 1.3179980894355625,
+      "learning_rate": 1.484294097342078e-05,
+      "loss": 2.3579,
+      "step": 1069
+    },
+    {
+      "cosine_similarity": 0.37038287932337444,
+      "epoch": 0.9972041006523765,
+      "grad_norm": 1.182899670699838,
+      "learning_rate": 1.4836037279944773e-05,
+      "loss": 1.988,
+      "step": 1070
+    },
+    {
+      "cosine_similarity": 0.45731200541556266,
+      "epoch": 0.9981360671015843,
+      "grad_norm": 1.2754317260936419,
+      "learning_rate": 1.4829133586468762e-05,
+      "loss": 1.9228,
+      "step": 1071
+    },
+    {
+      "cosine_similarity": 0.10395217453166762,
+      "epoch": 0.9990680335507922,
+      "grad_norm": 1.4609470583979731,
+      "learning_rate": 1.4822229892992753e-05,
+      "loss": 2.4198,
+      "step": 1072
+    },
+    {
+      "cosine_similarity": 0.16234145093665803,
+      "epoch": 1.0,
+      "grad_norm": 1.110401072992952,
+      "learning_rate": 1.4815326199516742e-05,
+      "loss": 2.0412,
+      "step": 1073
+    },
+    {
+      "cosine_similarity": 0.2529346874194412,
+      "epoch": 1.0009319664492078,
+      "grad_norm": 1.3617292978013524,
+      "learning_rate": 1.4808422506040734e-05,
+      "loss": 1.9489,
+      "step": 1074
+    },
+    {
+      "cosine_similarity": 0.12505382806399865,
+      "epoch": 1.0018639328984156,
+      "grad_norm": 1.4588943414354731,
+      "learning_rate": 1.4801518812564723e-05,
+      "loss": 2.0745,
+      "step": 1075
+    },
+    {
+      "cosine_similarity": 0.6939809963419343,
+      "epoch": 1.0027958993476236,
+      "grad_norm": 1.072041623189185,
+      "learning_rate": 1.4794615119088714e-05,
+      "loss": 1.6922,
+      "step": 1076
+    },
+    {
+      "cosine_similarity": 0.17122290196026077,
+      "epoch": 1.0037278657968314,
+      "grad_norm": 1.2247248706507248,
+      "learning_rate": 1.4787711425612703e-05,
+      "loss": 1.705,
+      "step": 1077
+    },
+    {
+      "cosine_similarity": -0.21629725254285204,
+      "epoch": 1.0046598322460392,
+      "grad_norm": 1.0791012092856853,
+      "learning_rate": 1.4780807732136696e-05,
+      "loss": 1.9911,
+      "step": 1078
+    },
+    {
+      "cosine_similarity": 0.2775679514864521,
+      "epoch": 1.005591798695247,
+      "grad_norm": 1.0798626250212844,
+      "learning_rate": 1.4773904038660685e-05,
+      "loss": 1.6312,
+      "step": 1079
+    },
+    {
+      "cosine_similarity": 0.09360194451795631,
+      "epoch": 1.0065237651444547,
+      "grad_norm": 1.0661963150756615,
+      "learning_rate": 1.4767000345184674e-05,
+      "loss": 1.4921,
+      "step": 1080
+    },
+    {
+      "cosine_similarity": 0.24079247102875317,
+      "epoch": 1.0074557315936625,
+      "grad_norm": 1.1398521097105654,
+      "learning_rate": 1.4760096651708666e-05,
+      "loss": 2.0307,
+      "step": 1081
+    },
+    {
+      "cosine_similarity": 0.13518781160186,
+      "epoch": 1.0083876980428705,
+      "grad_norm": 1.1576578821122472,
+      "learning_rate": 1.4753192958232657e-05,
+      "loss": 1.7429,
+      "step": 1082
+    },
+    {
+      "cosine_similarity": -0.2545791509798496,
+      "epoch": 1.0093196644920783,
+      "grad_norm": 1.379385332297065,
+      "learning_rate": 1.4746289264756646e-05,
+      "loss": 1.8179,
+      "step": 1083
+    },
+    {
+      "cosine_similarity": 0.16599558270277578,
+      "epoch": 1.0102516309412861,
+      "grad_norm": 1.280370180356584,
+      "learning_rate": 1.4739385571280635e-05,
+      "loss": 1.6286,
+      "step": 1084
+    },
+    {
+      "cosine_similarity": 0.2296921439281264,
+      "epoch": 1.011183597390494,
+      "grad_norm": 1.0981003833326275,
+      "learning_rate": 1.4732481877804627e-05,
+      "loss": 1.8926,
+      "step": 1085
+    },
+    {
+      "cosine_similarity": 0.0992210621702367,
+      "epoch": 1.0121155638397017,
+      "grad_norm": 1.1258991381253616,
+      "learning_rate": 1.4725578184328618e-05,
+      "loss": 1.7227,
+      "step": 1086
+    },
+    {
+      "cosine_similarity": 0.001106874976138019,
+      "epoch": 1.0130475302889097,
+      "grad_norm": 0.9354344955306583,
+      "learning_rate": 1.4718674490852607e-05,
+      "loss": 1.6936,
+      "step": 1087
+    },
+    {
+      "cosine_similarity": -0.12160694533863554,
+      "epoch": 1.0139794967381175,
+      "grad_norm": 1.1780997938604063,
+      "learning_rate": 1.4711770797376596e-05,
+      "loss": 1.7146,
+      "step": 1088
+    },
+    {
+      "cosine_similarity": 0.0803660870588389,
+      "epoch": 1.0149114631873253,
+      "grad_norm": 1.7433608489567716,
+      "learning_rate": 1.4704867103900589e-05,
+      "loss": 1.8965,
+      "step": 1089
+    },
+    {
+      "cosine_similarity": 0.33777778731482394,
+      "epoch": 1.015843429636533,
+      "grad_norm": 1.314249399842105,
+      "learning_rate": 1.4697963410424578e-05,
+      "loss": 1.8454,
+      "step": 1090
+    },
+    {
+      "cosine_similarity": 0.3549736839185438,
+      "epoch": 1.0167753960857409,
+      "grad_norm": 1.0864663471291827,
+      "learning_rate": 1.4691059716948568e-05,
+      "loss": 1.7389,
+      "step": 1091
+    },
+    {
+      "cosine_similarity": 0.22522805876604884,
+      "epoch": 1.0177073625349486,
+      "grad_norm": 1.2626168252258414,
+      "learning_rate": 1.468415602347256e-05,
+      "loss": 1.9366,
+      "step": 1092
+    },
+    {
+      "cosine_similarity": -0.14223574455152616,
+      "epoch": 1.0186393289841567,
+      "grad_norm": 1.1612981599488108,
+      "learning_rate": 1.467725232999655e-05,
+      "loss": 2.0743,
+      "step": 1093
+    },
+    {
+      "cosine_similarity": 0.2955552292456157,
+      "epoch": 1.0195712954333644,
+      "grad_norm": 1.5457365132569814,
+      "learning_rate": 1.4670348636520539e-05,
+      "loss": 1.8654,
+      "step": 1094
+    },
+    {
+      "cosine_similarity": -0.3299086231492028,
+      "epoch": 1.0205032618825722,
+      "grad_norm": 1.1115742047645485,
+      "learning_rate": 1.466344494304453e-05,
+      "loss": 1.5546,
+      "step": 1095
+    },
+    {
+      "cosine_similarity": 0.16537231693853383,
+      "epoch": 1.02143522833178,
+      "grad_norm": 1.0390593208031087,
+      "learning_rate": 1.465654124956852e-05,
+      "loss": 1.6541,
+      "step": 1096
+    },
+    {
+      "cosine_similarity": 0.5077772054743692,
+      "epoch": 1.0223671947809878,
+      "grad_norm": 1.159237841690778,
+      "learning_rate": 1.4649637556092511e-05,
+      "loss": 1.6826,
+      "step": 1097
+    },
+    {
+      "cosine_similarity": 0.32851599433246975,
+      "epoch": 1.0232991612301958,
+      "grad_norm": 1.2133893871373183,
+      "learning_rate": 1.46427338626165e-05,
+      "loss": 1.4998,
+      "step": 1098
+    },
+    {
+      "cosine_similarity": -0.05128227909484984,
+      "epoch": 1.0242311276794036,
+      "grad_norm": 1.510704683667745,
+      "learning_rate": 1.463583016914049e-05,
+      "loss": 2.0258,
+      "step": 1099
+    },
+    {
+      "cosine_similarity": 0.10857124829435968,
+      "epoch": 1.0251630941286114,
+      "grad_norm": 1.2463370548284476,
+      "learning_rate": 1.4628926475664481e-05,
+      "loss": 1.7312,
+      "step": 1100
+    },
+    {
+      "cosine_similarity": 0.1674608797767576,
+      "epoch": 1.0260950605778192,
+      "grad_norm": 1.1160122944498199,
+      "learning_rate": 1.4622022782188472e-05,
+      "loss": 1.6393,
+      "step": 1101
+    },
+    {
+      "cosine_similarity": 0.2119730258633674,
+      "epoch": 1.027027027027027,
+      "grad_norm": 1.5155571992629024,
+      "learning_rate": 1.4615119088712461e-05,
+      "loss": 1.9385,
+      "step": 1102
+    },
+    {
+      "cosine_similarity": 0.24933627301255554,
+      "epoch": 1.0279589934762348,
+      "grad_norm": 1.4665554689961096,
+      "learning_rate": 1.4608215395236454e-05,
+      "loss": 1.7221,
+      "step": 1103
+    },
+    {
+      "cosine_similarity": -0.17997303575585125,
+      "epoch": 1.0288909599254428,
+      "grad_norm": 1.046525726286338,
+      "learning_rate": 1.4601311701760443e-05,
+      "loss": 1.8879,
+      "step": 1104
+    },
+    {
+      "cosine_similarity": 0.08175177554473247,
+      "epoch": 1.0298229263746506,
+      "grad_norm": 1.4476010358504714,
+      "learning_rate": 1.4594408008284433e-05,
+      "loss": 1.8039,
+      "step": 1105
+    },
+    {
+      "cosine_similarity": 0.013816853618349478,
+      "epoch": 1.0307548928238583,
+      "grad_norm": 1.2444606380243677,
+      "learning_rate": 1.4587504314808423e-05,
+      "loss": 1.688,
+      "step": 1106
+    },
+    {
+      "cosine_similarity": -0.10910910747317412,
+      "epoch": 1.0316868592730661,
+      "grad_norm": 1.3510601766285246,
+      "learning_rate": 1.4580600621332415e-05,
+      "loss": 1.7189,
+      "step": 1107
+    },
+    {
+      "cosine_similarity": 0.25992373594815676,
+      "epoch": 1.032618825722274,
+      "grad_norm": 1.2922022910336397,
+      "learning_rate": 1.4573696927856404e-05,
+      "loss": 1.95,
+      "step": 1108
+    },
+    {
+      "cosine_similarity": 0.0748506962852779,
+      "epoch": 1.0335507921714817,
+      "grad_norm": 1.3090711802218793,
+      "learning_rate": 1.4566793234380395e-05,
+      "loss": 1.6682,
+      "step": 1109
+    },
+    {
+      "cosine_similarity": 0.08641999675645493,
+      "epoch": 1.0344827586206897,
+      "grad_norm": 1.2038686559865803,
+      "learning_rate": 1.4559889540904385e-05,
+      "loss": 1.7591,
+      "step": 1110
+    },
+    {
+      "cosine_similarity": -0.014767699195481428,
+      "epoch": 1.0354147250698975,
+      "grad_norm": 1.3588492966171186,
+      "learning_rate": 1.4552985847428376e-05,
+      "loss": 1.572,
+      "step": 1111
+    },
+    {
+      "cosine_similarity": 0.21807629407566423,
+      "epoch": 1.0363466915191053,
+      "grad_norm": 1.2438944882899685,
+      "learning_rate": 1.4546082153952365e-05,
+      "loss": 1.8076,
+      "step": 1112
+    },
+    {
+      "cosine_similarity": 0.4818181159400519,
+      "epoch": 1.037278657968313,
+      "grad_norm": 1.2518726459172205,
+      "learning_rate": 1.4539178460476356e-05,
+      "loss": 1.7175,
+      "step": 1113
+    },
+    {
+      "cosine_similarity": 0.2845454573224555,
+      "epoch": 1.0382106244175209,
+      "grad_norm": 1.284314797110271,
+      "learning_rate": 1.4532274767000347e-05,
+      "loss": 1.9365,
+      "step": 1114
+    },
+    {
+      "cosine_similarity": 0.45988360773770454,
+      "epoch": 1.0391425908667289,
+      "grad_norm": 1.308613123154501,
+      "learning_rate": 1.4525371073524337e-05,
+      "loss": 1.956,
+      "step": 1115
+    },
+    {
+      "cosine_similarity": 0.0992433557165712,
+      "epoch": 1.0400745573159367,
+      "grad_norm": 1.2022736610156528,
+      "learning_rate": 1.4518467380048326e-05,
+      "loss": 1.597,
+      "step": 1116
+    },
+    {
+      "cosine_similarity": 0.43863394827871843,
+      "epoch": 1.0410065237651445,
+      "grad_norm": 1.2458772079641987,
+      "learning_rate": 1.4511563686572317e-05,
+      "loss": 1.8865,
+      "step": 1117
+    },
+    {
+      "cosine_similarity": 0.16302690469221556,
+      "epoch": 1.0419384902143523,
+      "grad_norm": 1.4360668893335387,
+      "learning_rate": 1.4504659993096308e-05,
+      "loss": 2.0251,
+      "step": 1118
+    },
+    {
+      "cosine_similarity": 0.0892991810542551,
+      "epoch": 1.04287045666356,
+      "grad_norm": 1.304966445898737,
+      "learning_rate": 1.4497756299620299e-05,
+      "loss": 2.2204,
+      "step": 1119
+    },
+    {
+      "cosine_similarity": 0.19080567705750084,
+      "epoch": 1.0438024231127678,
+      "grad_norm": 1.1721339344486374,
+      "learning_rate": 1.4490852606144288e-05,
+      "loss": 1.9584,
+      "step": 1120
+    },
+    {
+      "cosine_similarity": 0.035551125267700844,
+      "epoch": 1.0447343895619758,
+      "grad_norm": 1.121646990833946,
+      "learning_rate": 1.448394891266828e-05,
+      "loss": 1.6443,
+      "step": 1121
+    },
+    {
+      "cosine_similarity": 0.021693727410633162,
+      "epoch": 1.0456663560111836,
+      "grad_norm": 1.138522342163322,
+      "learning_rate": 1.4477045219192269e-05,
+      "loss": 1.7132,
+      "step": 1122
+    },
+    {
+      "cosine_similarity": 0.12543636281545129,
+      "epoch": 1.0465983224603914,
+      "grad_norm": 1.2924304806075655,
+      "learning_rate": 1.447014152571626e-05,
+      "loss": 2.0823,
+      "step": 1123
+    },
+    {
+      "cosine_similarity": 0.6115519766726314,
+      "epoch": 1.0475302889095992,
+      "grad_norm": 1.1485756130825457,
+      "learning_rate": 1.4463237832240249e-05,
+      "loss": 2.0029,
+      "step": 1124
+    },
+    {
+      "cosine_similarity": 0.40230857605683096,
+      "epoch": 1.048462255358807,
+      "grad_norm": 1.1471098552370453,
+      "learning_rate": 1.4456334138764241e-05,
+      "loss": 1.9096,
+      "step": 1125
+    },
+    {
+      "cosine_similarity": 0.27952964020976817,
+      "epoch": 1.049394221808015,
+      "grad_norm": 1.297870624779003,
+      "learning_rate": 1.444943044528823e-05,
+      "loss": 2.011,
+      "step": 1126
+    },
+    {
+      "cosine_similarity": 0.2443269790160249,
+      "epoch": 1.0503261882572228,
+      "grad_norm": 1.0508554738463078,
+      "learning_rate": 1.4442526751812221e-05,
+      "loss": 1.6311,
+      "step": 1127
+    },
+    {
+      "cosine_similarity": -0.09179355221279609,
+      "epoch": 1.0512581547064306,
+      "grad_norm": 1.1522277041120523,
+      "learning_rate": 1.443562305833621e-05,
+      "loss": 1.6904,
+      "step": 1128
+    },
+    {
+      "cosine_similarity": 0.03182056514984807,
+      "epoch": 1.0521901211556384,
+      "grad_norm": 1.2000002345037215,
+      "learning_rate": 1.4428719364860203e-05,
+      "loss": 1.9921,
+      "step": 1129
+    },
+    {
+      "cosine_similarity": -0.1658448228405271,
+      "epoch": 1.0531220876048462,
+      "grad_norm": 1.4847642273707433,
+      "learning_rate": 1.4421815671384192e-05,
+      "loss": 1.7826,
+      "step": 1130
+    },
+    {
+      "cosine_similarity": 0.051080030747223115,
+      "epoch": 1.054054054054054,
+      "grad_norm": 1.5133469120778413,
+      "learning_rate": 1.441491197790818e-05,
+      "loss": 2.109,
+      "step": 1131
+    },
+    {
+      "cosine_similarity": -0.12506904932053423,
+      "epoch": 1.054986020503262,
+      "grad_norm": 1.249962935768059,
+      "learning_rate": 1.4408008284432173e-05,
+      "loss": 1.4441,
+      "step": 1132
+    },
+    {
+      "cosine_similarity": -0.06313598584689964,
+      "epoch": 1.0559179869524697,
+      "grad_norm": 1.0989749579935741,
+      "learning_rate": 1.4401104590956164e-05,
+      "loss": 1.8988,
+      "step": 1133
+    },
+    {
+      "cosine_similarity": 0.13953362826902593,
+      "epoch": 1.0568499534016775,
+      "grad_norm": 1.1356400725305877,
+      "learning_rate": 1.4394200897480153e-05,
+      "loss": 1.906,
+      "step": 1134
+    },
+    {
+      "cosine_similarity": 0.1801894464315253,
+      "epoch": 1.0577819198508853,
+      "grad_norm": 1.2433493092054353,
+      "learning_rate": 1.4387297204004142e-05,
+      "loss": 1.875,
+      "step": 1135
+    },
+    {
+      "cosine_similarity": -0.014659194796595595,
+      "epoch": 1.058713886300093,
+      "grad_norm": 1.1726783617751757,
+      "learning_rate": 1.4380393510528134e-05,
+      "loss": 1.7492,
+      "step": 1136
+    },
+    {
+      "cosine_similarity": -0.14643125509130253,
+      "epoch": 1.0596458527493011,
+      "grad_norm": 1.0586581996320994,
+      "learning_rate": 1.4373489817052125e-05,
+      "loss": 1.9589,
+      "step": 1137
+    },
+    {
+      "cosine_similarity": -0.15769365248499123,
+      "epoch": 1.060577819198509,
+      "grad_norm": 1.3026320573336487,
+      "learning_rate": 1.4366586123576114e-05,
+      "loss": 1.901,
+      "step": 1138
+    },
+    {
+      "cosine_similarity": 0.49058268780852976,
+      "epoch": 1.0615097856477167,
+      "grad_norm": 1.4372011821082495,
+      "learning_rate": 1.4359682430100103e-05,
+      "loss": 1.7901,
+      "step": 1139
+    },
+    {
+      "cosine_similarity": 0.0035195421821929826,
+      "epoch": 1.0624417520969245,
+      "grad_norm": 1.7490732779932037,
+      "learning_rate": 1.4352778736624096e-05,
+      "loss": 2.2865,
+      "step": 1140
+    },
+    {
+      "cosine_similarity": 0.10719188640712826,
+      "epoch": 1.0633737185461323,
+      "grad_norm": 1.154519538921614,
+      "learning_rate": 1.4345875043148085e-05,
+      "loss": 1.8247,
+      "step": 1141
+    },
+    {
+      "cosine_similarity": -0.07717265248912088,
+      "epoch": 1.06430568499534,
+      "grad_norm": 0.9867257392089064,
+      "learning_rate": 1.4338971349672075e-05,
+      "loss": 1.3635,
+      "step": 1142
+    },
+    {
+      "cosine_similarity": -0.28860282376197577,
+      "epoch": 1.065237651444548,
+      "grad_norm": 1.0386584301030013,
+      "learning_rate": 1.4332067656196068e-05,
+      "loss": 1.7767,
+      "step": 1143
+    },
+    {
+      "cosine_similarity": -0.12117281616078136,
+      "epoch": 1.0661696178937559,
+      "grad_norm": 1.2508144438968467,
+      "learning_rate": 1.4325163962720057e-05,
+      "loss": 1.9135,
+      "step": 1144
+    },
+    {
+      "cosine_similarity": 0.14880862986322396,
+      "epoch": 1.0671015843429636,
+      "grad_norm": 1.254751673955984,
+      "learning_rate": 1.4318260269244046e-05,
+      "loss": 1.7296,
+      "step": 1145
+    },
+    {
+      "cosine_similarity": 0.13336916207304436,
+      "epoch": 1.0680335507921714,
+      "grad_norm": 1.254245950472177,
+      "learning_rate": 1.4311356575768037e-05,
+      "loss": 1.8949,
+      "step": 1146
+    },
+    {
+      "cosine_similarity": 0.0647566864982512,
+      "epoch": 1.0689655172413792,
+      "grad_norm": 1.1714603840131765,
+      "learning_rate": 1.4304452882292027e-05,
+      "loss": 1.9557,
+      "step": 1147
+    },
+    {
+      "cosine_similarity": 0.26038353855969304,
+      "epoch": 1.0698974836905872,
+      "grad_norm": 1.107058107201564,
+      "learning_rate": 1.4297549188816018e-05,
+      "loss": 1.7805,
+      "step": 1148
+    },
+    {
+      "cosine_similarity": 0.3092361509925611,
+      "epoch": 1.070829450139795,
+      "grad_norm": 1.5844356466006753,
+      "learning_rate": 1.4290645495340007e-05,
+      "loss": 1.9913,
+      "step": 1149
+    },
+    {
+      "cosine_similarity": -0.025836474839227835,
+      "epoch": 1.0717614165890028,
+      "grad_norm": 1.1838918411101036,
+      "learning_rate": 1.4283741801864e-05,
+      "loss": 1.7494,
+      "step": 1150
+    },
+    {
+      "cosine_similarity": 0.18798169269453047,
+      "epoch": 1.0726933830382106,
+      "grad_norm": 1.5769899178181435,
+      "learning_rate": 1.4276838108387989e-05,
+      "loss": 2.1024,
+      "step": 1151
+    },
+    {
+      "cosine_similarity": -0.06163564648796004,
+      "epoch": 1.0736253494874184,
+      "grad_norm": 1.5046414148546292,
+      "learning_rate": 1.426993441491198e-05,
+      "loss": 1.8461,
+      "step": 1152
+    },
+    {
+      "cosine_similarity": 0.1398342086865286,
+      "epoch": 1.0745573159366262,
+      "grad_norm": 1.2890025098790083,
+      "learning_rate": 1.4263030721435968e-05,
+      "loss": 1.8445,
+      "step": 1153
+    },
+    {
+      "cosine_similarity": -0.16075193175113578,
+      "epoch": 1.0754892823858342,
+      "grad_norm": 1.112946098861442,
+      "learning_rate": 1.425612702795996e-05,
+      "loss": 1.7358,
+      "step": 1154
+    },
+    {
+      "cosine_similarity": 0.2028696280860701,
+      "epoch": 1.076421248835042,
+      "grad_norm": 1.3093039071658568,
+      "learning_rate": 1.424922333448395e-05,
+      "loss": 2.1783,
+      "step": 1155
+    },
+    {
+      "cosine_similarity": 0.2241639210284079,
+      "epoch": 1.0773532152842498,
+      "grad_norm": 1.1406646066036454,
+      "learning_rate": 1.424231964100794e-05,
+      "loss": 1.9752,
+      "step": 1156
+    },
+    {
+      "cosine_similarity": 0.0630209202288915,
+      "epoch": 1.0782851817334576,
+      "grad_norm": 1.1243990283665775,
+      "learning_rate": 1.423541594753193e-05,
+      "loss": 2.0986,
+      "step": 1157
+    },
+    {
+      "cosine_similarity": 0.3457683912997958,
+      "epoch": 1.0792171481826653,
+      "grad_norm": 1.249758425960956,
+      "learning_rate": 1.4228512254055922e-05,
+      "loss": 2.1414,
+      "step": 1158
+    },
+    {
+      "cosine_similarity": 0.2915769019810277,
+      "epoch": 1.0801491146318734,
+      "grad_norm": 1.1959437455847182,
+      "learning_rate": 1.4221608560579911e-05,
+      "loss": 1.5645,
+      "step": 1159
+    },
+    {
+      "cosine_similarity": 0.19319332898499209,
+      "epoch": 1.0810810810810811,
+      "grad_norm": 1.278548981864792,
+      "learning_rate": 1.4214704867103902e-05,
+      "loss": 1.6706,
+      "step": 1160
+    },
+    {
+      "cosine_similarity": 0.0924752173250572,
+      "epoch": 1.082013047530289,
+      "grad_norm": 1.430457025716276,
+      "learning_rate": 1.4207801173627892e-05,
+      "loss": 1.4691,
+      "step": 1161
+    },
+    {
+      "cosine_similarity": 0.36264224296123965,
+      "epoch": 1.0829450139794967,
+      "grad_norm": 1.0740773422140608,
+      "learning_rate": 1.4200897480151883e-05,
+      "loss": 1.6117,
+      "step": 1162
+    },
+    {
+      "cosine_similarity": 0.4296808393625959,
+      "epoch": 1.0838769804287045,
+      "grad_norm": 1.418366702545325,
+      "learning_rate": 1.4193993786675872e-05,
+      "loss": 2.0781,
+      "step": 1163
+    },
+    {
+      "cosine_similarity": 0.29150979506600017,
+      "epoch": 1.0848089468779123,
+      "grad_norm": 1.1172555342610189,
+      "learning_rate": 1.4187090093199863e-05,
+      "loss": 1.7843,
+      "step": 1164
+    },
+    {
+      "cosine_similarity": 0.27975366130639195,
+      "epoch": 1.0857409133271203,
+      "grad_norm": 1.274654222936517,
+      "learning_rate": 1.4180186399723854e-05,
+      "loss": 1.8219,
+      "step": 1165
+    },
+    {
+      "cosine_similarity": 0.023058040490803143,
+      "epoch": 1.086672879776328,
+      "grad_norm": 1.0868471481493422,
+      "learning_rate": 1.4173282706247844e-05,
+      "loss": 1.6408,
+      "step": 1166
+    },
+    {
+      "cosine_similarity": 0.7156414739168062,
+      "epoch": 1.0876048462255359,
+      "grad_norm": 1.4593952373614507,
+      "learning_rate": 1.4166379012771833e-05,
+      "loss": 1.9581,
+      "step": 1167
+    },
+    {
+      "cosine_similarity": 0.3057617169937069,
+      "epoch": 1.0885368126747437,
+      "grad_norm": 1.2173442937392973,
+      "learning_rate": 1.4159475319295824e-05,
+      "loss": 1.8636,
+      "step": 1168
+    },
+    {
+      "cosine_similarity": 0.18059729735312263,
+      "epoch": 1.0894687791239515,
+      "grad_norm": 1.3105194503418656,
+      "learning_rate": 1.4152571625819815e-05,
+      "loss": 2.0886,
+      "step": 1169
+    },
+    {
+      "cosine_similarity": 0.25303599524344483,
+      "epoch": 1.0904007455731595,
+      "grad_norm": 1.0084104023398257,
+      "learning_rate": 1.4145667932343806e-05,
+      "loss": 1.8017,
+      "step": 1170
+    },
+    {
+      "cosine_similarity": 0.180120544896426,
+      "epoch": 1.0913327120223673,
+      "grad_norm": 1.29778138942217,
+      "learning_rate": 1.4138764238867795e-05,
+      "loss": 1.7146,
+      "step": 1171
+    },
+    {
+      "cosine_similarity": 0.10734535615489398,
+      "epoch": 1.092264678471575,
+      "grad_norm": 1.154697225766186,
+      "learning_rate": 1.4131860545391787e-05,
+      "loss": 1.6983,
+      "step": 1172
+    },
+    {
+      "cosine_similarity": 0.2235011958732202,
+      "epoch": 1.0931966449207828,
+      "grad_norm": 1.6350993642542326,
+      "learning_rate": 1.4124956851915776e-05,
+      "loss": 1.5134,
+      "step": 1173
+    },
+    {
+      "cosine_similarity": 0.43971460119152694,
+      "epoch": 1.0941286113699906,
+      "grad_norm": 1.252344246782257,
+      "learning_rate": 1.4118053158439767e-05,
+      "loss": 1.5702,
+      "step": 1174
+    },
+    {
+      "cosine_similarity": 0.4529720944353802,
+      "epoch": 1.0950605778191984,
+      "grad_norm": 1.208874695819535,
+      "learning_rate": 1.4111149464963756e-05,
+      "loss": 1.9655,
+      "step": 1175
+    },
+    {
+      "cosine_similarity": 0.1795578775044819,
+      "epoch": 1.0959925442684064,
+      "grad_norm": 1.2858997731117519,
+      "learning_rate": 1.4104245771487748e-05,
+      "loss": 1.8548,
+      "step": 1176
+    },
+    {
+      "cosine_similarity": 0.37975427827312097,
+      "epoch": 1.0969245107176142,
+      "grad_norm": 1.2566740362092266,
+      "learning_rate": 1.4097342078011737e-05,
+      "loss": 2.13,
+      "step": 1177
+    },
+    {
+      "cosine_similarity": 0.43065899352144793,
+      "epoch": 1.097856477166822,
+      "grad_norm": 1.2560823350054364,
+      "learning_rate": 1.4090438384535728e-05,
+      "loss": 1.5751,
+      "step": 1178
+    },
+    {
+      "cosine_similarity": 0.16113905353898547,
+      "epoch": 1.0987884436160298,
+      "grad_norm": 1.145452610318921,
+      "learning_rate": 1.4083534691059717e-05,
+      "loss": 1.9612,
+      "step": 1179
+    },
+    {
+      "cosine_similarity": 0.25548119961341903,
+      "epoch": 1.0997204100652376,
+      "grad_norm": 1.1695760236154962,
+      "learning_rate": 1.407663099758371e-05,
+      "loss": 1.6045,
+      "step": 1180
+    },
+    {
+      "cosine_similarity": 0.13651814412817884,
+      "epoch": 1.1006523765144456,
+      "grad_norm": 1.3856352412565192,
+      "learning_rate": 1.4069727304107699e-05,
+      "loss": 1.7047,
+      "step": 1181
+    },
+    {
+      "cosine_similarity": 0.34526990299304505,
+      "epoch": 1.1015843429636534,
+      "grad_norm": 1.0647165803900291,
+      "learning_rate": 1.4062823610631688e-05,
+      "loss": 1.5285,
+      "step": 1182
+    },
+    {
+      "cosine_similarity": 0.20689516380602174,
+      "epoch": 1.1025163094128612,
+      "grad_norm": 1.2188504510081544,
+      "learning_rate": 1.405591991715568e-05,
+      "loss": 1.7027,
+      "step": 1183
+    },
+    {
+      "cosine_similarity": 0.29578530603162156,
+      "epoch": 1.103448275862069,
+      "grad_norm": 1.213996939844059,
+      "learning_rate": 1.404901622367967e-05,
+      "loss": 2.0108,
+      "step": 1184
+    },
+    {
+      "cosine_similarity": 0.5567784482099395,
+      "epoch": 1.1043802423112767,
+      "grad_norm": 1.3542667227996446,
+      "learning_rate": 1.404211253020366e-05,
+      "loss": 1.7093,
+      "step": 1185
+    },
+    {
+      "cosine_similarity": 0.4080762635987535,
+      "epoch": 1.1053122087604845,
+      "grad_norm": 1.2350884934954112,
+      "learning_rate": 1.4035208836727649e-05,
+      "loss": 1.8568,
+      "step": 1186
+    },
+    {
+      "cosine_similarity": 0.15678801857599262,
+      "epoch": 1.1062441752096925,
+      "grad_norm": 1.2565296796967884,
+      "learning_rate": 1.4028305143251641e-05,
+      "loss": 1.931,
+      "step": 1187
+    },
+    {
+      "cosine_similarity": 0.006608660888693946,
+      "epoch": 1.1071761416589003,
+      "grad_norm": 1.3171907641869995,
+      "learning_rate": 1.402140144977563e-05,
+      "loss": 1.9252,
+      "step": 1188
+    },
+    {
+      "cosine_similarity": 0.41489501514692834,
+      "epoch": 1.1081081081081081,
+      "grad_norm": 1.1224283246517597,
+      "learning_rate": 1.4014497756299621e-05,
+      "loss": 1.8865,
+      "step": 1189
+    },
+    {
+      "cosine_similarity": 0.23342354958091077,
+      "epoch": 1.109040074557316,
+      "grad_norm": 1.2257141491214307,
+      "learning_rate": 1.400759406282361e-05,
+      "loss": 1.561,
+      "step": 1190
+    },
+    {
+      "cosine_similarity": 0.13102201763839075,
+      "epoch": 1.1099720410065237,
+      "grad_norm": 1.1024395117624823,
+      "learning_rate": 1.4000690369347603e-05,
+      "loss": 1.8646,
+      "step": 1191
+    },
+    {
+      "cosine_similarity": -0.015712415895278206,
+      "epoch": 1.1109040074557317,
+      "grad_norm": 1.2247368511518664,
+      "learning_rate": 1.3993786675871592e-05,
+      "loss": 1.7602,
+      "step": 1192
+    },
+    {
+      "cosine_similarity": 0.24483525640838283,
+      "epoch": 1.1118359739049395,
+      "grad_norm": 1.1586760432948737,
+      "learning_rate": 1.3986882982395582e-05,
+      "loss": 1.878,
+      "step": 1193
+    },
+    {
+      "cosine_similarity": 0.4572024834967652,
+      "epoch": 1.1127679403541473,
+      "grad_norm": 1.2937088868920412,
+      "learning_rate": 1.3979979288919575e-05,
+      "loss": 1.9166,
+      "step": 1194
+    },
+    {
+      "cosine_similarity": 0.2912023376469478,
+      "epoch": 1.113699906803355,
+      "grad_norm": 1.1964574828099606,
+      "learning_rate": 1.3973075595443564e-05,
+      "loss": 1.7294,
+      "step": 1195
+    },
+    {
+      "cosine_similarity": 0.21286126276341635,
+      "epoch": 1.1146318732525629,
+      "grad_norm": 1.4791992047385127,
+      "learning_rate": 1.3966171901967553e-05,
+      "loss": 1.477,
+      "step": 1196
+    },
+    {
+      "cosine_similarity": 0.2744076507216444,
+      "epoch": 1.1155638397017706,
+      "grad_norm": 1.086116327189491,
+      "learning_rate": 1.3959268208491544e-05,
+      "loss": 2.1449,
+      "step": 1197
+    },
+    {
+      "cosine_similarity": 0.23244020849648236,
+      "epoch": 1.1164958061509787,
+      "grad_norm": 1.2004186949918338,
+      "learning_rate": 1.3952364515015534e-05,
+      "loss": 2.2223,
+      "step": 1198
+    },
+    {
+      "cosine_similarity": -0.10630333230111076,
+      "epoch": 1.1174277726001864,
+      "grad_norm": 1.4159260703642604,
+      "learning_rate": 1.3945460821539525e-05,
+      "loss": 1.9545,
+      "step": 1199
+    },
+    {
+      "cosine_similarity": 0.16747908331172445,
+      "epoch": 1.1183597390493942,
+      "grad_norm": 1.5852640957333297,
+      "learning_rate": 1.3938557128063514e-05,
+      "loss": 1.6396,
+      "step": 1200
+    },
+    {
+      "cosine_similarity": 0.5407438284715911,
+      "epoch": 1.119291705498602,
+      "grad_norm": 1.4686940453824462,
+      "learning_rate": 1.3931653434587507e-05,
+      "loss": 1.9678,
+      "step": 1201
+    },
+    {
+      "cosine_similarity": 0.15527354664855988,
+      "epoch": 1.1202236719478098,
+      "grad_norm": 1.134037626656289,
+      "learning_rate": 1.3924749741111496e-05,
+      "loss": 2.3005,
+      "step": 1202
+    },
+    {
+      "cosine_similarity": 0.24353187676500454,
+      "epoch": 1.1211556383970178,
+      "grad_norm": 1.465526230313108,
+      "learning_rate": 1.3917846047635486e-05,
+      "loss": 1.8744,
+      "step": 1203
+    },
+    {
+      "cosine_similarity": 0.0503685706007102,
+      "epoch": 1.1220876048462256,
+      "grad_norm": 1.2971987849343976,
+      "learning_rate": 1.3910942354159475e-05,
+      "loss": 1.7256,
+      "step": 1204
+    },
+    {
+      "cosine_similarity": 0.23171499016893526,
+      "epoch": 1.1230195712954334,
+      "grad_norm": 1.4230712427825654,
+      "learning_rate": 1.3904038660683468e-05,
+      "loss": 1.8763,
+      "step": 1205
+    },
+    {
+      "cosine_similarity": -0.0714719506722462,
+      "epoch": 1.1239515377446412,
+      "grad_norm": 1.2297361758387266,
+      "learning_rate": 1.3897134967207457e-05,
+      "loss": 1.9658,
+      "step": 1206
+    },
+    {
+      "cosine_similarity": 0.008703631417887443,
+      "epoch": 1.124883504193849,
+      "grad_norm": 1.2328533323229458,
+      "learning_rate": 1.3890231273731448e-05,
+      "loss": 1.7896,
+      "step": 1207
+    },
+    {
+      "cosine_similarity": 0.3597989481931174,
+      "epoch": 1.1258154706430568,
+      "grad_norm": 1.3431838528688704,
+      "learning_rate": 1.3883327580255437e-05,
+      "loss": 1.9221,
+      "step": 1208
+    },
+    {
+      "cosine_similarity": -0.09231570330321794,
+      "epoch": 1.1267474370922648,
+      "grad_norm": 1.3290144029280808,
+      "learning_rate": 1.3876423886779429e-05,
+      "loss": 2.0112,
+      "step": 1209
+    },
+    {
+      "cosine_similarity": 0.022393872836714013,
+      "epoch": 1.1276794035414726,
+      "grad_norm": 2.1355610484329026,
+      "learning_rate": 1.3869520193303418e-05,
+      "loss": 1.885,
+      "step": 1210
+    },
+    {
+      "cosine_similarity": 0.17087704486067984,
+      "epoch": 1.1286113699906803,
+      "grad_norm": 1.2607714768565705,
+      "learning_rate": 1.3862616499827409e-05,
+      "loss": 1.8951,
+      "step": 1211
+    },
+    {
+      "cosine_similarity": 0.18138604208787065,
+      "epoch": 1.1295433364398881,
+      "grad_norm": 0.9247752280784443,
+      "learning_rate": 1.38557128063514e-05,
+      "loss": 1.3892,
+      "step": 1212
+    },
+    {
+      "cosine_similarity": 0.386411811386714,
+      "epoch": 1.130475302889096,
+      "grad_norm": 1.4194787149593049,
+      "learning_rate": 1.384880911287539e-05,
+      "loss": 2.2945,
+      "step": 1213
+    },
+    {
+      "cosine_similarity": -0.0012223340566554098,
+      "epoch": 1.131407269338304,
+      "grad_norm": 1.2881222495090459,
+      "learning_rate": 1.384190541939938e-05,
+      "loss": 1.9419,
+      "step": 1214
+    },
+    {
+      "cosine_similarity": 0.20268975952196702,
+      "epoch": 1.1323392357875117,
+      "grad_norm": 1.0732035477579893,
+      "learning_rate": 1.383500172592337e-05,
+      "loss": 1.7969,
+      "step": 1215
+    },
+    {
+      "cosine_similarity": 0.13466430992473633,
+      "epoch": 1.1332712022367195,
+      "grad_norm": 1.373086486484411,
+      "learning_rate": 1.382809803244736e-05,
+      "loss": 2.129,
+      "step": 1216
+    },
+    {
+      "cosine_similarity": 0.001999202912223561,
+      "epoch": 1.1342031686859273,
+      "grad_norm": 1.4336964684758742,
+      "learning_rate": 1.3821194338971351e-05,
+      "loss": 1.6834,
+      "step": 1217
+    },
+    {
+      "cosine_similarity": 0.19019552475337403,
+      "epoch": 1.135135135135135,
+      "grad_norm": 1.1697796459908991,
+      "learning_rate": 1.381429064549534e-05,
+      "loss": 2.01,
+      "step": 1218
+    },
+    {
+      "cosine_similarity": 0.13991921255057999,
+      "epoch": 1.1360671015843429,
+      "grad_norm": 0.9515090344568532,
+      "learning_rate": 1.3807386952019331e-05,
+      "loss": 1.7612,
+      "step": 1219
+    },
+    {
+      "cosine_similarity": 0.14739749863732385,
+      "epoch": 1.1369990680335509,
+      "grad_norm": 1.3214479210755046,
+      "learning_rate": 1.3800483258543322e-05,
+      "loss": 2.0084,
+      "step": 1220
+    },
+    {
+      "cosine_similarity": -0.005706568764932192,
+      "epoch": 1.1379310344827587,
+      "grad_norm": 1.2405630710659916,
+      "learning_rate": 1.3793579565067313e-05,
+      "loss": 1.7497,
+      "step": 1221
+    },
+    {
+      "cosine_similarity": 0.02404684214250351,
+      "epoch": 1.1388630009319665,
+      "grad_norm": 1.1859000715109518,
+      "learning_rate": 1.3786675871591302e-05,
+      "loss": 1.6382,
+      "step": 1222
+    },
+    {
+      "cosine_similarity": 0.4847179695855016,
+      "epoch": 1.1397949673811743,
+      "grad_norm": 1.1853121400770483,
+      "learning_rate": 1.3779772178115294e-05,
+      "loss": 1.9374,
+      "step": 1223
+    },
+    {
+      "cosine_similarity": 0.20867768449636998,
+      "epoch": 1.140726933830382,
+      "grad_norm": 1.4181306275123626,
+      "learning_rate": 1.3772868484639283e-05,
+      "loss": 2.0473,
+      "step": 1224
+    },
+    {
+      "cosine_similarity": 0.33267106737953683,
+      "epoch": 1.14165890027959,
+      "grad_norm": 1.091314646377402,
+      "learning_rate": 1.3765964791163274e-05,
+      "loss": 1.7633,
+      "step": 1225
+    },
+    {
+      "cosine_similarity": 0.06054437176216545,
+      "epoch": 1.1425908667287978,
+      "grad_norm": 1.024314294845468,
+      "learning_rate": 1.3759061097687263e-05,
+      "loss": 1.6078,
+      "step": 1226
+    },
+    {
+      "cosine_similarity": 0.18403126961505822,
+      "epoch": 1.1435228331780056,
+      "grad_norm": 1.1614674175237152,
+      "learning_rate": 1.3752157404211255e-05,
+      "loss": 1.4234,
+      "step": 1227
+    },
+    {
+      "cosine_similarity": 0.14113621860701298,
+      "epoch": 1.1444547996272134,
+      "grad_norm": 2.1762275229063546,
+      "learning_rate": 1.3745253710735244e-05,
+      "loss": 2.4592,
+      "step": 1228
+    },
+    {
+      "cosine_similarity": 0.20737840298101964,
+      "epoch": 1.1453867660764212,
+      "grad_norm": 1.403012723445906,
+      "learning_rate": 1.3738350017259235e-05,
+      "loss": 1.4602,
+      "step": 1229
+    },
+    {
+      "cosine_similarity": -0.23461544576132737,
+      "epoch": 1.146318732525629,
+      "grad_norm": 1.3803127402143696,
+      "learning_rate": 1.3731446323783224e-05,
+      "loss": 1.9526,
+      "step": 1230
+    },
+    {
+      "cosine_similarity": 0.016892230743915635,
+      "epoch": 1.147250698974837,
+      "grad_norm": 1.2048630806772194,
+      "learning_rate": 1.3724542630307217e-05,
+      "loss": 1.7646,
+      "step": 1231
+    },
+    {
+      "cosine_similarity": 0.10142910246808284,
+      "epoch": 1.1481826654240448,
+      "grad_norm": 1.2450783193347053,
+      "learning_rate": 1.3717638936831206e-05,
+      "loss": 1.9568,
+      "step": 1232
+    },
+    {
+      "cosine_similarity": -0.2668144526029428,
+      "epoch": 1.1491146318732526,
+      "grad_norm": 1.1991038502400309,
+      "learning_rate": 1.3710735243355195e-05,
+      "loss": 1.483,
+      "step": 1233
+    },
+    {
+      "cosine_similarity": 0.09032053787233629,
+      "epoch": 1.1500465983224604,
+      "grad_norm": 1.321535204746285,
+      "learning_rate": 1.3703831549879187e-05,
+      "loss": 1.6976,
+      "step": 1234
+    },
+    {
+      "cosine_similarity": 0.2958490916973732,
+      "epoch": 1.1509785647716682,
+      "grad_norm": 1.3861350498688063,
+      "learning_rate": 1.3696927856403178e-05,
+      "loss": 1.9191,
+      "step": 1235
+    },
+    {
+      "cosine_similarity": -0.021752811333204017,
+      "epoch": 1.1519105312208762,
+      "grad_norm": 1.0808629102904213,
+      "learning_rate": 1.3690024162927167e-05,
+      "loss": 1.9995,
+      "step": 1236
+    },
+    {
+      "cosine_similarity": 0.1386973061648661,
+      "epoch": 1.152842497670084,
+      "grad_norm": 1.190012476588744,
+      "learning_rate": 1.3683120469451156e-05,
+      "loss": 1.9531,
+      "step": 1237
+    },
+    {
+      "cosine_similarity": 0.12968122782660077,
+      "epoch": 1.1537744641192917,
+      "grad_norm": 1.140014466632025,
+      "learning_rate": 1.3676216775975148e-05,
+      "loss": 1.8895,
+      "step": 1238
+    },
+    {
+      "cosine_similarity": -0.11018975034450969,
+      "epoch": 1.1547064305684995,
+      "grad_norm": 1.262284490825013,
+      "learning_rate": 1.3669313082499137e-05,
+      "loss": 1.777,
+      "step": 1239
+    },
+    {
+      "cosine_similarity": 0.2741627406373057,
+      "epoch": 1.1556383970177073,
+      "grad_norm": 1.4804773530313093,
+      "learning_rate": 1.3662409389023128e-05,
+      "loss": 1.7863,
+      "step": 1240
+    },
+    {
+      "cosine_similarity": 0.001672322940945594,
+      "epoch": 1.156570363466915,
+      "grad_norm": 1.2100442906262658,
+      "learning_rate": 1.365550569554712e-05,
+      "loss": 1.7823,
+      "step": 1241
+    },
+    {
+      "cosine_similarity": 0.015004057391056319,
+      "epoch": 1.157502329916123,
+      "grad_norm": 1.3433708787727539,
+      "learning_rate": 1.364860200207111e-05,
+      "loss": 1.9991,
+      "step": 1242
+    },
+    {
+      "cosine_similarity": 0.008609213458730721,
+      "epoch": 1.158434296365331,
+      "grad_norm": 1.5723832363193704,
+      "learning_rate": 1.3641698308595099e-05,
+      "loss": 1.7247,
+      "step": 1243
+    },
+    {
+      "cosine_similarity": 0.06414055930659962,
+      "epoch": 1.1593662628145387,
+      "grad_norm": 1.3491790681133176,
+      "learning_rate": 1.363479461511909e-05,
+      "loss": 2.0424,
+      "step": 1244
+    },
+    {
+      "cosine_similarity": -0.031082478746911266,
+      "epoch": 1.1602982292637465,
+      "grad_norm": 1.1784312982947371,
+      "learning_rate": 1.3627890921643082e-05,
+      "loss": 1.8356,
+      "step": 1245
+    },
+    {
+      "cosine_similarity": 0.09089327914640358,
+      "epoch": 1.1612301957129543,
+      "grad_norm": 1.0140015364019501,
+      "learning_rate": 1.362098722816707e-05,
+      "loss": 1.7015,
+      "step": 1246
+    },
+    {
+      "cosine_similarity": 0.39858741751590815,
+      "epoch": 1.1621621621621623,
+      "grad_norm": 1.115162456951588,
+      "learning_rate": 1.361408353469106e-05,
+      "loss": 1.8333,
+      "step": 1247
+    },
+    {
+      "cosine_similarity": -0.21958602213209977,
+      "epoch": 1.16309412861137,
+      "grad_norm": 1.682323063057243,
+      "learning_rate": 1.360717984121505e-05,
+      "loss": 1.6504,
+      "step": 1248
+    },
+    {
+      "cosine_similarity": -0.12113496621387689,
+      "epoch": 1.1640260950605779,
+      "grad_norm": 2.021181727579927,
+      "learning_rate": 1.3600276147739041e-05,
+      "loss": 2.351,
+      "step": 1249
+    },
+    {
+      "cosine_similarity": 0.13660873433252105,
+      "epoch": 1.1649580615097856,
+      "grad_norm": 1.1575820113650808,
+      "learning_rate": 1.3593372454263032e-05,
+      "loss": 1.8238,
+      "step": 1250
+    },
+    {
+      "cosine_similarity": 0.40335619840936027,
+      "epoch": 1.1658900279589934,
+      "grad_norm": 1.164825243925656,
+      "learning_rate": 1.3586468760787021e-05,
+      "loss": 1.736,
+      "step": 1251
+    },
+    {
+      "cosine_similarity": -0.021213702255943948,
+      "epoch": 1.1668219944082012,
+      "grad_norm": 1.1457045875105774,
+      "learning_rate": 1.3579565067311014e-05,
+      "loss": 1.5296,
+      "step": 1252
+    },
+    {
+      "cosine_similarity": -0.0048091510062141735,
+      "epoch": 1.167753960857409,
+      "grad_norm": 1.3048042617265034,
+      "learning_rate": 1.3572661373835003e-05,
+      "loss": 1.7927,
+      "step": 1253
+    },
+    {
+      "cosine_similarity": 0.013981247728614855,
+      "epoch": 1.168685927306617,
+      "grad_norm": 1.1858436629747842,
+      "learning_rate": 1.3565757680358993e-05,
+      "loss": 1.892,
+      "step": 1254
+    },
+    {
+      "cosine_similarity": 0.14841467847869572,
+      "epoch": 1.1696178937558248,
+      "grad_norm": 1.3450246679307583,
+      "learning_rate": 1.3558853986882982e-05,
+      "loss": 1.9083,
+      "step": 1255
+    },
+    {
+      "cosine_similarity": 0.15108143527926926,
+      "epoch": 1.1705498602050326,
+      "grad_norm": 1.4650054813375852,
+      "learning_rate": 1.3551950293406975e-05,
+      "loss": 1.9947,
+      "step": 1256
+    },
+    {
+      "cosine_similarity": 0.2515487263819254,
+      "epoch": 1.1714818266542404,
+      "grad_norm": 1.0997087864959447,
+      "learning_rate": 1.3545046599930964e-05,
+      "loss": 2.0166,
+      "step": 1257
+    },
+    {
+      "cosine_similarity": 0.10848037008649515,
+      "epoch": 1.1724137931034484,
+      "grad_norm": 1.1101498713796447,
+      "learning_rate": 1.3538142906454955e-05,
+      "loss": 1.5183,
+      "step": 1258
+    },
+    {
+      "cosine_similarity": 0.19649397922373601,
+      "epoch": 1.1733457595526562,
+      "grad_norm": 1.0368019498843224,
+      "learning_rate": 1.3531239212978944e-05,
+      "loss": 1.7064,
+      "step": 1259
+    },
+    {
+      "cosine_similarity": 0.2640260808495064,
+      "epoch": 1.174277726001864,
+      "grad_norm": 1.0830412627933457,
+      "learning_rate": 1.3524335519502936e-05,
+      "loss": 1.781,
+      "step": 1260
+    },
+    {
+      "cosine_similarity": 0.3330544707515436,
+      "epoch": 1.1752096924510718,
+      "grad_norm": 1.4732607290066215,
+      "learning_rate": 1.3517431826026925e-05,
+      "loss": 1.6936,
+      "step": 1261
+    },
+    {
+      "cosine_similarity": 0.3119327726475695,
+      "epoch": 1.1761416589002796,
+      "grad_norm": 1.3120396885231187,
+      "learning_rate": 1.3510528132550916e-05,
+      "loss": 1.7405,
+      "step": 1262
+    },
+    {
+      "cosine_similarity": 0.3751614192506316,
+      "epoch": 1.1770736253494873,
+      "grad_norm": 1.0670069134944014,
+      "learning_rate": 1.3503624439074907e-05,
+      "loss": 1.7869,
+      "step": 1263
+    },
+    {
+      "cosine_similarity": 0.2996339482070278,
+      "epoch": 1.1780055917986951,
+      "grad_norm": 1.1699282186681366,
+      "learning_rate": 1.3496720745598897e-05,
+      "loss": 1.8349,
+      "step": 1264
+    },
+    {
+      "cosine_similarity": 0.024537782498732293,
+      "epoch": 1.1789375582479031,
+      "grad_norm": 1.1025220863264156,
+      "learning_rate": 1.3489817052122886e-05,
+      "loss": 1.8644,
+      "step": 1265
+    },
+    {
+      "cosine_similarity": 0.0940379484070052,
+      "epoch": 1.179869524697111,
+      "grad_norm": 1.161412867094022,
+      "learning_rate": 1.3482913358646877e-05,
+      "loss": 1.7081,
+      "step": 1266
+    },
+    {
+      "cosine_similarity": 0.4399312333424312,
+      "epoch": 1.1808014911463187,
+      "grad_norm": 1.2147606052503348,
+      "learning_rate": 1.3476009665170868e-05,
+      "loss": 1.8983,
+      "step": 1267
+    },
+    {
+      "cosine_similarity": 0.20151884054014838,
+      "epoch": 1.1817334575955265,
+      "grad_norm": 1.3807545153040144,
+      "learning_rate": 1.3469105971694858e-05,
+      "loss": 1.8768,
+      "step": 1268
+    },
+    {
+      "cosine_similarity": 0.018794152706518626,
+      "epoch": 1.1826654240447343,
+      "grad_norm": 1.073904402244953,
+      "learning_rate": 1.3462202278218848e-05,
+      "loss": 1.8191,
+      "step": 1269
+    },
+    {
+      "cosine_similarity": 0.09086785266905795,
+      "epoch": 1.1835973904939423,
+      "grad_norm": 1.1638692943188689,
+      "learning_rate": 1.3455298584742838e-05,
+      "loss": 1.8523,
+      "step": 1270
+    },
+    {
+      "cosine_similarity": 0.25139478233291923,
+      "epoch": 1.18452935694315,
+      "grad_norm": 1.0311211302579175,
+      "learning_rate": 1.3448394891266829e-05,
+      "loss": 1.5458,
+      "step": 1271
+    },
+    {
+      "cosine_similarity": 0.23005359763791244,
+      "epoch": 1.1854613233923579,
+      "grad_norm": 1.0833738671574638,
+      "learning_rate": 1.344149119779082e-05,
+      "loss": 1.5627,
+      "step": 1272
+    },
+    {
+      "cosine_similarity": 0.23768721934057407,
+      "epoch": 1.1863932898415657,
+      "grad_norm": 1.2126350295600588,
+      "learning_rate": 1.3434587504314809e-05,
+      "loss": 1.6985,
+      "step": 1273
+    },
+    {
+      "cosine_similarity": 0.13298891624712844,
+      "epoch": 1.1873252562907735,
+      "grad_norm": 1.2861788009826234,
+      "learning_rate": 1.3427683810838801e-05,
+      "loss": 1.7109,
+      "step": 1274
+    },
+    {
+      "cosine_similarity": 0.4363285729993991,
+      "epoch": 1.1882572227399812,
+      "grad_norm": 1.419289008587863,
+      "learning_rate": 1.342078011736279e-05,
+      "loss": 1.9217,
+      "step": 1275
+    },
+    {
+      "cosine_similarity": -0.1736073221744842,
+      "epoch": 1.1891891891891893,
+      "grad_norm": 1.1556085273943761,
+      "learning_rate": 1.3413876423886781e-05,
+      "loss": 1.568,
+      "step": 1276
+    },
+    {
+      "cosine_similarity": 0.18762314312155295,
+      "epoch": 1.190121155638397,
+      "grad_norm": 1.2713186512299584,
+      "learning_rate": 1.340697273041077e-05,
+      "loss": 1.8632,
+      "step": 1277
+    },
+    {
+      "cosine_similarity": 0.35293797572290275,
+      "epoch": 1.1910531220876048,
+      "grad_norm": 1.0719519724094952,
+      "learning_rate": 1.3400069036934762e-05,
+      "loss": 1.8083,
+      "step": 1278
+    },
+    {
+      "cosine_similarity": 0.12299358410592605,
+      "epoch": 1.1919850885368126,
+      "grad_norm": 1.269199584418817,
+      "learning_rate": 1.3393165343458751e-05,
+      "loss": 1.8338,
+      "step": 1279
+    },
+    {
+      "cosine_similarity": 0.022061813183002984,
+      "epoch": 1.1929170549860204,
+      "grad_norm": 1.140209662212806,
+      "learning_rate": 1.3386261649982742e-05,
+      "loss": 1.486,
+      "step": 1280
+    },
+    {
+      "cosine_similarity": 0.16433142517267138,
+      "epoch": 1.1938490214352284,
+      "grad_norm": 1.1765647277772948,
+      "learning_rate": 1.3379357956506733e-05,
+      "loss": 1.9563,
+      "step": 1281
+    },
+    {
+      "cosine_similarity": 0.6631103691479069,
+      "epoch": 1.1947809878844362,
+      "grad_norm": 1.2270044194574175,
+      "learning_rate": 1.3372454263030724e-05,
+      "loss": 1.8358,
+      "step": 1282
+    },
+    {
+      "cosine_similarity": 0.23722251116404472,
+      "epoch": 1.195712954333644,
+      "grad_norm": 1.0304390881337036,
+      "learning_rate": 1.3365550569554713e-05,
+      "loss": 1.5631,
+      "step": 1283
+    },
+    {
+      "cosine_similarity": 0.07670419552250872,
+      "epoch": 1.1966449207828518,
+      "grad_norm": 1.1693815149908946,
+      "learning_rate": 1.3358646876078702e-05,
+      "loss": 1.8932,
+      "step": 1284
+    },
+    {
+      "cosine_similarity": 0.25406635348190054,
+      "epoch": 1.1975768872320596,
+      "grad_norm": 1.2909839194217774,
+      "learning_rate": 1.3351743182602694e-05,
+      "loss": 1.9873,
+      "step": 1285
+    },
+    {
+      "cosine_similarity": 0.15023519368911953,
+      "epoch": 1.1985088536812674,
+      "grad_norm": 0.9989694625314747,
+      "learning_rate": 1.3344839489126685e-05,
+      "loss": 1.6468,
+      "step": 1286
+    },
+    {
+      "cosine_similarity": -0.007357864012191874,
+      "epoch": 1.1994408201304754,
+      "grad_norm": 1.2658501922345704,
+      "learning_rate": 1.3337935795650674e-05,
+      "loss": 1.8554,
+      "step": 1287
+    },
+    {
+      "cosine_similarity": 0.004653494163979659,
+      "epoch": 1.2003727865796832,
+      "grad_norm": 1.077964719541428,
+      "learning_rate": 1.3331032102174663e-05,
+      "loss": 1.6131,
+      "step": 1288
+    },
+    {
+      "cosine_similarity": 0.3820790709886342,
+      "epoch": 1.201304753028891,
+      "grad_norm": 0.9706225755463557,
+      "learning_rate": 1.3324128408698655e-05,
+      "loss": 1.1597,
+      "step": 1289
+    },
+    {
+      "cosine_similarity": 0.7122271786941493,
+      "epoch": 1.2022367194780987,
+      "grad_norm": 1.3999563814895966,
+      "learning_rate": 1.3317224715222644e-05,
+      "loss": 1.8615,
+      "step": 1290
+    },
+    {
+      "cosine_similarity": -0.17822535297619282,
+      "epoch": 1.2031686859273065,
+      "grad_norm": 1.600749476014216,
+      "learning_rate": 1.3310321021746635e-05,
+      "loss": 1.6229,
+      "step": 1291
+    },
+    {
+      "cosine_similarity": 0.15304221228863138,
+      "epoch": 1.2041006523765145,
+      "grad_norm": 2.1520336457819993,
+      "learning_rate": 1.3303417328270628e-05,
+      "loss": 1.7702,
+      "step": 1292
+    },
+    {
+      "cosine_similarity": -0.0481511122941759,
+      "epoch": 1.2050326188257223,
+      "grad_norm": 1.4265946387493382,
+      "learning_rate": 1.3296513634794617e-05,
+      "loss": 1.8836,
+      "step": 1293
+    },
+    {
+      "cosine_similarity": 0.4301958505787666,
+      "epoch": 1.2059645852749301,
+      "grad_norm": 1.3563813705615997,
+      "learning_rate": 1.3289609941318606e-05,
+      "loss": 2.2906,
+      "step": 1294
+    },
+    {
+      "cosine_similarity": 0.23541118872053166,
+      "epoch": 1.206896551724138,
+      "grad_norm": 1.2410938383170411,
+      "learning_rate": 1.3282706247842596e-05,
+      "loss": 1.6195,
+      "step": 1295
+    },
+    {
+      "cosine_similarity": 0.25862099648397324,
+      "epoch": 1.2078285181733457,
+      "grad_norm": 1.5898963673483646,
+      "learning_rate": 1.3275802554366589e-05,
+      "loss": 1.7495,
+      "step": 1296
+    },
+    {
+      "cosine_similarity": 0.15232600003715924,
+      "epoch": 1.2087604846225535,
+      "grad_norm": 1.2575086767154686,
+      "learning_rate": 1.3268898860890578e-05,
+      "loss": 1.7418,
+      "step": 1297
+    },
+    {
+      "cosine_similarity": 0.2057080403569417,
+      "epoch": 1.2096924510717615,
+      "grad_norm": 1.1710916833779106,
+      "learning_rate": 1.3261995167414567e-05,
+      "loss": 1.8115,
+      "step": 1298
+    },
+    {
+      "cosine_similarity": 0.20169066536476288,
+      "epoch": 1.2106244175209693,
+      "grad_norm": 1.0630925819959889,
+      "learning_rate": 1.3255091473938558e-05,
+      "loss": 1.7193,
+      "step": 1299
+    },
+    {
+      "cosine_similarity": 0.26806020113758305,
+      "epoch": 1.211556383970177,
+      "grad_norm": 0.9825244542540802,
+      "learning_rate": 1.3248187780462548e-05,
+      "loss": 1.6058,
+      "step": 1300
+    },
+    {
+      "cosine_similarity": 0.0652048866496896,
+      "epoch": 1.2124883504193849,
+      "grad_norm": 1.1640637520394275,
+      "learning_rate": 1.3241284086986539e-05,
+      "loss": 2.0198,
+      "step": 1301
+    },
+    {
+      "cosine_similarity": 0.13022162240535554,
+      "epoch": 1.2134203168685926,
+      "grad_norm": 3.208468899478075,
+      "learning_rate": 1.3234380393510528e-05,
+      "loss": 1.9208,
+      "step": 1302
+    },
+    {
+      "cosine_similarity": 0.021904894644120015,
+      "epoch": 1.2143522833178007,
+      "grad_norm": 1.1203125960597746,
+      "learning_rate": 1.322747670003452e-05,
+      "loss": 1.6375,
+      "step": 1303
+    },
+    {
+      "cosine_similarity": 0.12036452170525042,
+      "epoch": 1.2152842497670084,
+      "grad_norm": 1.1580970213969182,
+      "learning_rate": 1.322057300655851e-05,
+      "loss": 1.9087,
+      "step": 1304
+    },
+    {
+      "cosine_similarity": 0.14914393756694097,
+      "epoch": 1.2162162162162162,
+      "grad_norm": 1.383082918680524,
+      "learning_rate": 1.32136693130825e-05,
+      "loss": 1.9587,
+      "step": 1305
+    },
+    {
+      "cosine_similarity": 0.331206710637301,
+      "epoch": 1.217148182665424,
+      "grad_norm": 1.5093197910623477,
+      "learning_rate": 1.320676561960649e-05,
+      "loss": 1.7722,
+      "step": 1306
+    },
+    {
+      "cosine_similarity": 0.06395461976699653,
+      "epoch": 1.2180801491146318,
+      "grad_norm": 1.3479789095041694,
+      "learning_rate": 1.3199861926130482e-05,
+      "loss": 1.9846,
+      "step": 1307
+    },
+    {
+      "cosine_similarity": 0.015963525346401067,
+      "epoch": 1.2190121155638396,
+      "grad_norm": 1.44210509737739,
+      "learning_rate": 1.3192958232654471e-05,
+      "loss": 1.9792,
+      "step": 1308
+    },
+    {
+      "cosine_similarity": -0.13493834463466373,
+      "epoch": 1.2199440820130476,
+      "grad_norm": 1.1439311122221962,
+      "learning_rate": 1.3186054539178462e-05,
+      "loss": 1.5731,
+      "step": 1309
+    },
+    {
+      "cosine_similarity": 0.4052615983802866,
+      "epoch": 1.2208760484622554,
+      "grad_norm": 1.5045601803004056,
+      "learning_rate": 1.317915084570245e-05,
+      "loss": 1.6575,
+      "step": 1310
+    },
+    {
+      "cosine_similarity": 0.12758645996333484,
+      "epoch": 1.2218080149114632,
+      "grad_norm": 1.312557352249991,
+      "learning_rate": 1.3172247152226443e-05,
+      "loss": 1.823,
+      "step": 1311
+    },
+    {
+      "cosine_similarity": 0.24466915534823883,
+      "epoch": 1.222739981360671,
+      "grad_norm": 1.1312950138213729,
+      "learning_rate": 1.3165343458750432e-05,
+      "loss": 1.9603,
+      "step": 1312
+    },
+    {
+      "cosine_similarity": 0.17258225526905574,
+      "epoch": 1.2236719478098788,
+      "grad_norm": 1.1716809581650152,
+      "learning_rate": 1.3158439765274423e-05,
+      "loss": 1.5977,
+      "step": 1313
+    },
+    {
+      "cosine_similarity": 0.37492299776790466,
+      "epoch": 1.2246039142590868,
+      "grad_norm": 1.1278839407091696,
+      "learning_rate": 1.3151536071798414e-05,
+      "loss": 1.7431,
+      "step": 1314
+    },
+    {
+      "cosine_similarity": 0.1800551416798403,
+      "epoch": 1.2255358807082946,
+      "grad_norm": 1.1773452033869147,
+      "learning_rate": 1.3144632378322404e-05,
+      "loss": 1.7517,
+      "step": 1315
+    },
+    {
+      "cosine_similarity": 0.14290074158804106,
+      "epoch": 1.2264678471575023,
+      "grad_norm": 1.2571684765324032,
+      "learning_rate": 1.3137728684846393e-05,
+      "loss": 1.7742,
+      "step": 1316
+    },
+    {
+      "cosine_similarity": 0.12949879007308643,
+      "epoch": 1.2273998136067101,
+      "grad_norm": 1.2033563069310347,
+      "learning_rate": 1.3130824991370384e-05,
+      "loss": 1.7797,
+      "step": 1317
+    },
+    {
+      "cosine_similarity": 0.19587338579642805,
+      "epoch": 1.228331780055918,
+      "grad_norm": 1.1023370608109604,
+      "learning_rate": 1.3123921297894375e-05,
+      "loss": 1.7495,
+      "step": 1318
+    },
+    {
+      "cosine_similarity": -0.010158148288970157,
+      "epoch": 1.2292637465051257,
+      "grad_norm": 1.7335637384162246,
+      "learning_rate": 1.3117017604418366e-05,
+      "loss": 1.9674,
+      "step": 1319
+    },
+    {
+      "cosine_similarity": 0.17636343512616778,
+      "epoch": 1.2301957129543337,
+      "grad_norm": 1.4558858454301138,
+      "learning_rate": 1.3110113910942355e-05,
+      "loss": 1.9106,
+      "step": 1320
+    },
+    {
+      "cosine_similarity": 0.2864037704151485,
+      "epoch": 1.2311276794035415,
+      "grad_norm": 1.2218864755463403,
+      "learning_rate": 1.3103210217466345e-05,
+      "loss": 1.6922,
+      "step": 1321
+    },
+    {
+      "cosine_similarity": -0.10185619732061371,
+      "epoch": 1.2320596458527493,
+      "grad_norm": 1.1065629386424738,
+      "learning_rate": 1.3096306523990336e-05,
+      "loss": 1.5327,
+      "step": 1322
+    },
+    {
+      "cosine_similarity": -0.1541996861673849,
+      "epoch": 1.232991612301957,
+      "grad_norm": 1.2571635797015557,
+      "learning_rate": 1.3089402830514327e-05,
+      "loss": 1.8631,
+      "step": 1323
+    },
+    {
+      "cosine_similarity": 0.10015322403074947,
+      "epoch": 1.2339235787511649,
+      "grad_norm": 1.169044430768509,
+      "learning_rate": 1.3082499137038316e-05,
+      "loss": 1.796,
+      "step": 1324
+    },
+    {
+      "cosine_similarity": 0.29477512283627616,
+      "epoch": 1.2348555452003729,
+      "grad_norm": 1.2343322403328816,
+      "learning_rate": 1.3075595443562308e-05,
+      "loss": 1.6094,
+      "step": 1325
+    },
+    {
+      "cosine_similarity": 0.05220290082126813,
+      "epoch": 1.2357875116495807,
+      "grad_norm": 1.299978532646063,
+      "learning_rate": 1.3068691750086297e-05,
+      "loss": 1.7309,
+      "step": 1326
+    },
+    {
+      "cosine_similarity": 0.2568244874318337,
+      "epoch": 1.2367194780987885,
+      "grad_norm": 1.130085295436261,
+      "learning_rate": 1.3061788056610288e-05,
+      "loss": 2.2196,
+      "step": 1327
+    },
+    {
+      "cosine_similarity": 0.030861737252864282,
+      "epoch": 1.2376514445479962,
+      "grad_norm": 1.0424217135173868,
+      "learning_rate": 1.3054884363134277e-05,
+      "loss": 1.5053,
+      "step": 1328
+    },
+    {
+      "cosine_similarity": 0.04946981449279249,
+      "epoch": 1.238583410997204,
+      "grad_norm": 1.227932649540319,
+      "learning_rate": 1.304798066965827e-05,
+      "loss": 1.5727,
+      "step": 1329
+    },
+    {
+      "cosine_similarity": 0.15476704821115075,
+      "epoch": 1.2395153774464118,
+      "grad_norm": 1.324857175666478,
+      "learning_rate": 1.3041076976182258e-05,
+      "loss": 1.6357,
+      "step": 1330
+    },
+    {
+      "cosine_similarity": 0.05792850482297169,
+      "epoch": 1.2404473438956198,
+      "grad_norm": 1.2072610259293393,
+      "learning_rate": 1.3034173282706248e-05,
+      "loss": 2.0541,
+      "step": 1331
+    },
+    {
+      "cosine_similarity": 0.025005714808112145,
+      "epoch": 1.2413793103448276,
+      "grad_norm": 1.177672987104257,
+      "learning_rate": 1.302726958923024e-05,
+      "loss": 1.8939,
+      "step": 1332
+    },
+    {
+      "cosine_similarity": 0.15252264236885826,
+      "epoch": 1.2423112767940354,
+      "grad_norm": 1.2664550779366515,
+      "learning_rate": 1.302036589575423e-05,
+      "loss": 1.602,
+      "step": 1333
+    },
+    {
+      "cosine_similarity": 0.09325578186542449,
+      "epoch": 1.2432432432432432,
+      "grad_norm": 1.2208579500376098,
+      "learning_rate": 1.301346220227822e-05,
+      "loss": 1.8034,
+      "step": 1334
+    },
+    {
+      "cosine_similarity": -0.07459105035769889,
+      "epoch": 1.244175209692451,
+      "grad_norm": 1.3140566088407062,
+      "learning_rate": 1.3006558508802209e-05,
+      "loss": 1.9173,
+      "step": 1335
+    },
+    {
+      "cosine_similarity": 0.3388498155527339,
+      "epoch": 1.245107176141659,
+      "grad_norm": 1.2515794570772594,
+      "learning_rate": 1.2999654815326201e-05,
+      "loss": 2.089,
+      "step": 1336
+    },
+    {
+      "cosine_similarity": -0.13358393891834014,
+      "epoch": 1.2460391425908668,
+      "grad_norm": 1.1133486074063792,
+      "learning_rate": 1.2992751121850192e-05,
+      "loss": 1.7132,
+      "step": 1337
+    },
+    {
+      "cosine_similarity": -0.1509013655255534,
+      "epoch": 1.2469711090400746,
+      "grad_norm": 1.1172066322839114,
+      "learning_rate": 1.2985847428374181e-05,
+      "loss": 1.9719,
+      "step": 1338
+    },
+    {
+      "cosine_similarity": -0.09834379338586438,
+      "epoch": 1.2479030754892824,
+      "grad_norm": 1.2272462540322988,
+      "learning_rate": 1.297894373489817e-05,
+      "loss": 1.7247,
+      "step": 1339
+    },
+    {
+      "cosine_similarity": 0.23890448314653526,
+      "epoch": 1.2488350419384902,
+      "grad_norm": 1.3805747358336073,
+      "learning_rate": 1.2972040041422162e-05,
+      "loss": 2.0412,
+      "step": 1340
+    },
+    {
+      "cosine_similarity": 0.20804826269175547,
+      "epoch": 1.249767008387698,
+      "grad_norm": 1.3073144218926147,
+      "learning_rate": 1.2965136347946151e-05,
+      "loss": 1.9272,
+      "step": 1341
+    },
+    {
+      "cosine_similarity": -0.1819614433620234,
+      "epoch": 1.250698974836906,
+      "grad_norm": 1.6157075878043377,
+      "learning_rate": 1.2958232654470142e-05,
+      "loss": 1.9214,
+      "step": 1342
+    },
+    {
+      "cosine_similarity": -0.13102084175151488,
+      "epoch": 1.2516309412861137,
+      "grad_norm": 1.416665423379183,
+      "learning_rate": 1.2951328960994135e-05,
+      "loss": 1.5898,
+      "step": 1343
+    },
+    {
+      "cosine_similarity": 0.023552104582332024,
+      "epoch": 1.2525629077353215,
+      "grad_norm": 1.576533284260416,
+      "learning_rate": 1.2944425267518124e-05,
+      "loss": 1.9383,
+      "step": 1344
+    },
+    {
+      "cosine_similarity": -0.10549481136720751,
+      "epoch": 1.2534948741845293,
+      "grad_norm": 1.195190215027469,
+      "learning_rate": 1.2937521574042113e-05,
+      "loss": 1.7074,
+      "step": 1345
+    },
+    {
+      "cosine_similarity": 0.00236959946601518,
+      "epoch": 1.254426840633737,
+      "grad_norm": 1.1083915436959966,
+      "learning_rate": 1.2930617880566103e-05,
+      "loss": 1.9121,
+      "step": 1346
+    },
+    {
+      "cosine_similarity": 0.08358268975637544,
+      "epoch": 1.2553588070829451,
+      "grad_norm": 1.1849282941250325,
+      "learning_rate": 1.2923714187090096e-05,
+      "loss": 2.047,
+      "step": 1347
+    },
+    {
+      "cosine_similarity": 0.49085870061271253,
+      "epoch": 1.256290773532153,
+      "grad_norm": 1.2509638121924367,
+      "learning_rate": 1.2916810493614085e-05,
+      "loss": 1.8409,
+      "step": 1348
+    },
+    {
+      "cosine_similarity": 0.41523686473007154,
+      "epoch": 1.2572227399813607,
+      "grad_norm": 1.048679790802513,
+      "learning_rate": 1.2909906800138074e-05,
+      "loss": 1.8252,
+      "step": 1349
+    },
+    {
+      "cosine_similarity": 0.025452648990387076,
+      "epoch": 1.2581547064305685,
+      "grad_norm": 1.0804322235736719,
+      "learning_rate": 1.2903003106662065e-05,
+      "loss": 1.5345,
+      "step": 1350
+    },
+    {
+      "cosine_similarity": 0.5167551710468028,
+      "epoch": 1.2590866728797763,
+      "grad_norm": 1.1054328889859268,
+      "learning_rate": 1.2896099413186055e-05,
+      "loss": 1.4774,
+      "step": 1351
+    },
+    {
+      "cosine_similarity": 0.41169027907166644,
+      "epoch": 1.260018639328984,
+      "grad_norm": 1.3166901543484197,
+      "learning_rate": 1.2889195719710046e-05,
+      "loss": 2.049,
+      "step": 1352
+    },
+    {
+      "cosine_similarity": 0.13934065825330713,
+      "epoch": 1.2609506057781918,
+      "grad_norm": 1.2563157219574745,
+      "learning_rate": 1.2882292026234035e-05,
+      "loss": 1.8601,
+      "step": 1353
+    },
+    {
+      "cosine_similarity": -0.06259731058557179,
+      "epoch": 1.2618825722273999,
+      "grad_norm": 1.174150167916975,
+      "learning_rate": 1.2875388332758028e-05,
+      "loss": 2.068,
+      "step": 1354
+    },
+    {
+      "cosine_similarity": 0.2666076900170404,
+      "epoch": 1.2628145386766076,
+      "grad_norm": 1.3402051924644076,
+      "learning_rate": 1.2868484639282017e-05,
+      "loss": 1.7679,
+      "step": 1355
+    },
+    {
+      "cosine_similarity": 0.26420421152646173,
+      "epoch": 1.2637465051258154,
+      "grad_norm": 1.2261493548766167,
+      "learning_rate": 1.2861580945806007e-05,
+      "loss": 1.841,
+      "step": 1356
+    },
+    {
+      "cosine_similarity": 0.1505648496053853,
+      "epoch": 1.2646784715750232,
+      "grad_norm": 1.12673079462477,
+      "learning_rate": 1.2854677252329996e-05,
+      "loss": 1.9126,
+      "step": 1357
+    },
+    {
+      "cosine_similarity": 0.31174631663043695,
+      "epoch": 1.2656104380242312,
+      "grad_norm": 1.1449183172417146,
+      "learning_rate": 1.2847773558853989e-05,
+      "loss": 1.6115,
+      "step": 1358
+    },
+    {
+      "cosine_similarity": 0.10683687388662802,
+      "epoch": 1.266542404473439,
+      "grad_norm": 1.1046826339082028,
+      "learning_rate": 1.2840869865377978e-05,
+      "loss": 1.4207,
+      "step": 1359
+    },
+    {
+      "cosine_similarity": 0.10470574173091644,
+      "epoch": 1.2674743709226468,
+      "grad_norm": 1.0339311400334132,
+      "learning_rate": 1.2833966171901969e-05,
+      "loss": 1.7397,
+      "step": 1360
+    },
+    {
+      "cosine_similarity": 0.1916186376865284,
+      "epoch": 1.2684063373718546,
+      "grad_norm": 1.0987112019486207,
+      "learning_rate": 1.2827062478425958e-05,
+      "loss": 1.8583,
+      "step": 1361
+    },
+    {
+      "cosine_similarity": 0.316180036241573,
+      "epoch": 1.2693383038210624,
+      "grad_norm": 1.2643051298065897,
+      "learning_rate": 1.282015878494995e-05,
+      "loss": 1.8468,
+      "step": 1362
+    },
+    {
+      "cosine_similarity": 0.12398619410240275,
+      "epoch": 1.2702702702702702,
+      "grad_norm": 1.041716136830765,
+      "learning_rate": 1.2813255091473939e-05,
+      "loss": 1.4967,
+      "step": 1363
+    },
+    {
+      "cosine_similarity": 0.17535116217029498,
+      "epoch": 1.271202236719478,
+      "grad_norm": 1.212419781169882,
+      "learning_rate": 1.280635139799793e-05,
+      "loss": 2.0035,
+      "step": 1364
+    },
+    {
+      "cosine_similarity": 0.11403137022199714,
+      "epoch": 1.272134203168686,
+      "grad_norm": 1.2621806738597316,
+      "learning_rate": 1.279944770452192e-05,
+      "loss": 1.5883,
+      "step": 1365
+    },
+    {
+      "cosine_similarity": 0.21184399104695228,
+      "epoch": 1.2730661696178938,
+      "grad_norm": 1.401542264404816,
+      "learning_rate": 1.2792544011045911e-05,
+      "loss": 1.7655,
+      "step": 1366
+    },
+    {
+      "cosine_similarity": 0.37160230665328814,
+      "epoch": 1.2739981360671015,
+      "grad_norm": 1.303668358635084,
+      "learning_rate": 1.27856403175699e-05,
+      "loss": 2.1736,
+      "step": 1367
+    },
+    {
+      "cosine_similarity": 0.09919565025414556,
+      "epoch": 1.2749301025163093,
+      "grad_norm": 1.2351365115868898,
+      "learning_rate": 1.2778736624093891e-05,
+      "loss": 1.9546,
+      "step": 1368
+    },
+    {
+      "cosine_similarity": -0.17880520000719802,
+      "epoch": 1.2758620689655173,
+      "grad_norm": 1.1607836686594462,
+      "learning_rate": 1.2771832930617882e-05,
+      "loss": 2.0214,
+      "step": 1369
+    },
+    {
+      "cosine_similarity": -0.004579280458732797,
+      "epoch": 1.2767940354147251,
+      "grad_norm": 1.2211356937098046,
+      "learning_rate": 1.2764929237141873e-05,
+      "loss": 1.7403,
+      "step": 1370
+    },
+    {
+      "cosine_similarity": 0.4475741770684633,
+      "epoch": 1.277726001863933,
+      "grad_norm": 1.2555160265240397,
+      "learning_rate": 1.2758025543665862e-05,
+      "loss": 1.828,
+      "step": 1371
+    },
+    {
+      "cosine_similarity": 0.5368207550876661,
+      "epoch": 1.2786579683131407,
+      "grad_norm": 1.0323851363315157,
+      "learning_rate": 1.2751121850189854e-05,
+      "loss": 2.0302,
+      "step": 1372
+    },
+    {
+      "cosine_similarity": -0.08726106625681616,
+      "epoch": 1.2795899347623485,
+      "grad_norm": 1.1995386557168295,
+      "learning_rate": 1.2744218156713843e-05,
+      "loss": 2.0844,
+      "step": 1373
+    },
+    {
+      "cosine_similarity": 0.3220048372992778,
+      "epoch": 1.2805219012115563,
+      "grad_norm": 1.0431789948235197,
+      "learning_rate": 1.2737314463237834e-05,
+      "loss": 1.643,
+      "step": 1374
+    },
+    {
+      "cosine_similarity": 0.19226946576565243,
+      "epoch": 1.281453867660764,
+      "grad_norm": 0.9585334235710281,
+      "learning_rate": 1.2730410769761823e-05,
+      "loss": 1.9041,
+      "step": 1375
+    },
+    {
+      "cosine_similarity": -0.05106523523636362,
+      "epoch": 1.282385834109972,
+      "grad_norm": 1.2445288731545616,
+      "learning_rate": 1.2723507076285815e-05,
+      "loss": 1.6385,
+      "step": 1376
+    },
+    {
+      "cosine_similarity": 0.39877819547714566,
+      "epoch": 1.2833178005591799,
+      "grad_norm": 1.296309887470932,
+      "learning_rate": 1.2716603382809804e-05,
+      "loss": 1.7856,
+      "step": 1377
+    },
+    {
+      "cosine_similarity": 0.34767951382948076,
+      "epoch": 1.2842497670083877,
+      "grad_norm": 1.2572836996884362,
+      "learning_rate": 1.2709699689333795e-05,
+      "loss": 1.6488,
+      "step": 1378
+    },
+    {
+      "cosine_similarity": 0.3006832347124235,
+      "epoch": 1.2851817334575955,
+      "grad_norm": 1.491487335853907,
+      "learning_rate": 1.2702795995857784e-05,
+      "loss": 1.7359,
+      "step": 1379
+    },
+    {
+      "cosine_similarity": 0.13637176202409929,
+      "epoch": 1.2861136999068035,
+      "grad_norm": 1.319666284566579,
+      "learning_rate": 1.2695892302381776e-05,
+      "loss": 1.841,
+      "step": 1380
+    },
+    {
+      "cosine_similarity": 0.05974988488145149,
+      "epoch": 1.2870456663560113,
+      "grad_norm": 1.1951573013333494,
+      "learning_rate": 1.2688988608905766e-05,
+      "loss": 1.9982,
+      "step": 1381
+    },
+    {
+      "cosine_similarity": 0.5650355532979081,
+      "epoch": 1.287977632805219,
+      "grad_norm": 1.078692246377579,
+      "learning_rate": 1.2682084915429755e-05,
+      "loss": 1.5916,
+      "step": 1382
+    },
+    {
+      "cosine_similarity": 0.2019679027126029,
+      "epoch": 1.2889095992544268,
+      "grad_norm": 1.2603495601368877,
+      "learning_rate": 1.2675181221953747e-05,
+      "loss": 1.8179,
+      "step": 1383
+    },
+    {
+      "cosine_similarity": -0.00431327002233403,
+      "epoch": 1.2898415657036346,
+      "grad_norm": 1.298809588796041,
+      "learning_rate": 1.2668277528477738e-05,
+      "loss": 1.8606,
+      "step": 1384
+    },
+    {
+      "cosine_similarity": 0.1330816228873664,
+      "epoch": 1.2907735321528424,
+      "grad_norm": 1.274604038651491,
+      "learning_rate": 1.2661373835001727e-05,
+      "loss": 1.7963,
+      "step": 1385
+    },
+    {
+      "cosine_similarity": 0.23687119170205664,
+      "epoch": 1.2917054986020502,
+      "grad_norm": 1.2016781875195248,
+      "learning_rate": 1.2654470141525716e-05,
+      "loss": 1.9524,
+      "step": 1386
+    },
+    {
+      "cosine_similarity": -0.023168213355911904,
+      "epoch": 1.2926374650512582,
+      "grad_norm": 1.2903259055770968,
+      "learning_rate": 1.2647566448049708e-05,
+      "loss": 1.5727,
+      "step": 1387
+    },
+    {
+      "cosine_similarity": 0.29888965972224874,
+      "epoch": 1.293569431500466,
+      "grad_norm": 1.1352539365967182,
+      "learning_rate": 1.2640662754573699e-05,
+      "loss": 1.5235,
+      "step": 1388
+    },
+    {
+      "cosine_similarity": 0.4581690371669237,
+      "epoch": 1.2945013979496738,
+      "grad_norm": 1.2942414033535488,
+      "learning_rate": 1.2633759061097688e-05,
+      "loss": 1.645,
+      "step": 1389
+    },
+    {
+      "cosine_similarity": -0.05827750945258534,
+      "epoch": 1.2954333643988816,
+      "grad_norm": 1.1394526587479132,
+      "learning_rate": 1.2626855367621677e-05,
+      "loss": 1.7344,
+      "step": 1390
+    },
+    {
+      "cosine_similarity": 0.30587142202909084,
+      "epoch": 1.2963653308480896,
+      "grad_norm": 1.4961321216185834,
+      "learning_rate": 1.261995167414567e-05,
+      "loss": 1.7661,
+      "step": 1391
+    },
+    {
+      "cosine_similarity": 0.3584491675251394,
+      "epoch": 1.2972972972972974,
+      "grad_norm": 1.261751390716187,
+      "learning_rate": 1.2613047980669658e-05,
+      "loss": 2.0092,
+      "step": 1392
+    },
+    {
+      "cosine_similarity": 0.3239068110635054,
+      "epoch": 1.2982292637465052,
+      "grad_norm": 1.3853594969816136,
+      "learning_rate": 1.260614428719365e-05,
+      "loss": 1.725,
+      "step": 1393
+    },
+    {
+      "cosine_similarity": 0.003351687757704881,
+      "epoch": 1.299161230195713,
+      "grad_norm": 1.151239124847404,
+      "learning_rate": 1.2599240593717642e-05,
+      "loss": 1.7583,
+      "step": 1394
+    },
+    {
+      "cosine_similarity": -0.011519417093918786,
+      "epoch": 1.3000931966449207,
+      "grad_norm": 1.1595630862612996,
+      "learning_rate": 1.259233690024163e-05,
+      "loss": 1.759,
+      "step": 1395
+    },
+    {
+      "cosine_similarity": 0.05975666196943286,
+      "epoch": 1.3010251630941285,
+      "grad_norm": 1.288807090005187,
+      "learning_rate": 1.258543320676562e-05,
+      "loss": 1.9912,
+      "step": 1396
+    },
+    {
+      "cosine_similarity": 0.23998708921645856,
+      "epoch": 1.3019571295433363,
+      "grad_norm": 1.1013482203585336,
+      "learning_rate": 1.257852951328961e-05,
+      "loss": 1.6435,
+      "step": 1397
+    },
+    {
+      "cosine_similarity": 0.35398656867769396,
+      "epoch": 1.3028890959925443,
+      "grad_norm": 1.087421242692138,
+      "learning_rate": 1.2571625819813603e-05,
+      "loss": 1.2912,
+      "step": 1398
+    },
+    {
+      "cosine_similarity": -0.005620387188545184,
+      "epoch": 1.303821062441752,
+      "grad_norm": 1.2551572532609085,
+      "learning_rate": 1.2564722126337592e-05,
+      "loss": 1.755,
+      "step": 1399
+    },
+    {
+      "cosine_similarity": -0.10313324061736488,
+      "epoch": 1.30475302889096,
+      "grad_norm": 1.0019363039620366,
+      "learning_rate": 1.2557818432861581e-05,
+      "loss": 1.6003,
+      "step": 1400
+    },
+    {
+      "cosine_similarity": -0.168775346051508,
+      "epoch": 1.3056849953401677,
+      "grad_norm": 1.291007489223207,
+      "learning_rate": 1.2550914739385572e-05,
+      "loss": 1.5717,
+      "step": 1401
+    },
+    {
+      "cosine_similarity": 0.17055148068293016,
+      "epoch": 1.3066169617893757,
+      "grad_norm": 1.1087737636427952,
+      "learning_rate": 1.2544011045909562e-05,
+      "loss": 1.6679,
+      "step": 1402
+    },
+    {
+      "cosine_similarity": 0.2937229082580676,
+      "epoch": 1.3075489282385835,
+      "grad_norm": 1.249746446103539,
+      "learning_rate": 1.2537107352433553e-05,
+      "loss": 1.7689,
+      "step": 1403
+    },
+    {
+      "cosine_similarity": 0.2772264865616702,
+      "epoch": 1.3084808946877913,
+      "grad_norm": 1.1983717833529273,
+      "learning_rate": 1.2530203658957542e-05,
+      "loss": 1.9173,
+      "step": 1404
+    },
+    {
+      "cosine_similarity": 0.1979311056331807,
+      "epoch": 1.309412861136999,
+      "grad_norm": 1.4965018978459477,
+      "learning_rate": 1.2523299965481535e-05,
+      "loss": 1.7782,
+      "step": 1405
+    },
+    {
+      "cosine_similarity": 0.30645029782986244,
+      "epoch": 1.3103448275862069,
+      "grad_norm": 1.2805184949568655,
+      "learning_rate": 1.2516396272005524e-05,
+      "loss": 2.0246,
+      "step": 1406
+    },
+    {
+      "cosine_similarity": 0.03376377402003057,
+      "epoch": 1.3112767940354146,
+      "grad_norm": 1.347100165603231,
+      "learning_rate": 1.2509492578529514e-05,
+      "loss": 1.9525,
+      "step": 1407
+    },
+    {
+      "cosine_similarity": 0.10585858969841222,
+      "epoch": 1.3122087604846224,
+      "grad_norm": 1.2669653261446574,
+      "learning_rate": 1.2502588885053503e-05,
+      "loss": 2.0771,
+      "step": 1408
+    },
+    {
+      "cosine_similarity": -0.18444152305531725,
+      "epoch": 1.3131407269338304,
+      "grad_norm": 1.4797992069041324,
+      "learning_rate": 1.2495685191577496e-05,
+      "loss": 1.772,
+      "step": 1409
+    },
+    {
+      "cosine_similarity": 0.18421038727961841,
+      "epoch": 1.3140726933830382,
+      "grad_norm": 1.3355027881763761,
+      "learning_rate": 1.2488781498101485e-05,
+      "loss": 1.8994,
+      "step": 1410
+    },
+    {
+      "cosine_similarity": 0.4967118376236295,
+      "epoch": 1.315004659832246,
+      "grad_norm": 1.406921666034876,
+      "learning_rate": 1.2481877804625476e-05,
+      "loss": 1.8021,
+      "step": 1411
+    },
+    {
+      "cosine_similarity": 0.5407327072932685,
+      "epoch": 1.3159366262814538,
+      "grad_norm": 1.4075340963492868,
+      "learning_rate": 1.2474974111149465e-05,
+      "loss": 1.5891,
+      "step": 1412
+    },
+    {
+      "cosine_similarity": -0.1220101096007825,
+      "epoch": 1.3168685927306618,
+      "grad_norm": 1.332468055408898,
+      "learning_rate": 1.2468070417673457e-05,
+      "loss": 1.9405,
+      "step": 1413
+    },
+    {
+      "cosine_similarity": 0.06850002907147626,
+      "epoch": 1.3178005591798696,
+      "grad_norm": 1.6365180289819605,
+      "learning_rate": 1.2461166724197446e-05,
+      "loss": 1.7636,
+      "step": 1414
+    },
+    {
+      "cosine_similarity": 0.7249960814869663,
+      "epoch": 1.3187325256290774,
+      "grad_norm": 1.136412358130612,
+      "learning_rate": 1.2454263030721437e-05,
+      "loss": 2.0425,
+      "step": 1415
+    },
+    {
+      "cosine_similarity": 0.09590187150313219,
+      "epoch": 1.3196644920782852,
+      "grad_norm": 1.1362622492916723,
+      "learning_rate": 1.2447359337245428e-05,
+      "loss": 1.7774,
+      "step": 1416
+    },
+    {
+      "cosine_similarity": -0.09856407529701881,
+      "epoch": 1.320596458527493,
+      "grad_norm": 1.0024088319275506,
+      "learning_rate": 1.2440455643769418e-05,
+      "loss": 1.7334,
+      "step": 1417
+    },
+    {
+      "cosine_similarity": 0.03993695374224958,
+      "epoch": 1.3215284249767008,
+      "grad_norm": 1.1735555323405558,
+      "learning_rate": 1.2433551950293407e-05,
+      "loss": 1.7684,
+      "step": 1418
+    },
+    {
+      "cosine_similarity": -0.005753466093593761,
+      "epoch": 1.3224603914259085,
+      "grad_norm": 1.1569053847445596,
+      "learning_rate": 1.2426648256817398e-05,
+      "loss": 1.8979,
+      "step": 1419
+    },
+    {
+      "cosine_similarity": 0.5155085593288434,
+      "epoch": 1.3233923578751166,
+      "grad_norm": 1.2405785208085551,
+      "learning_rate": 1.2419744563341389e-05,
+      "loss": 1.8891,
+      "step": 1420
+    },
+    {
+      "cosine_similarity": 0.22851133442654686,
+      "epoch": 1.3243243243243243,
+      "grad_norm": 1.4014646805970041,
+      "learning_rate": 1.241284086986538e-05,
+      "loss": 2.0086,
+      "step": 1421
+    },
+    {
+      "cosine_similarity": 0.06508362878053035,
+      "epoch": 1.3252562907735321,
+      "grad_norm": 1.0662280252076952,
+      "learning_rate": 1.2405937176389369e-05,
+      "loss": 1.5274,
+      "step": 1422
+    },
+    {
+      "cosine_similarity": 0.33307358378860036,
+      "epoch": 1.32618825722274,
+      "grad_norm": 1.2251324877185115,
+      "learning_rate": 1.2399033482913361e-05,
+      "loss": 1.9009,
+      "step": 1423
+    },
+    {
+      "cosine_similarity": 0.24273013266205076,
+      "epoch": 1.327120223671948,
+      "grad_norm": 1.2097292044170138,
+      "learning_rate": 1.239212978943735e-05,
+      "loss": 1.7317,
+      "step": 1424
+    },
+    {
+      "cosine_similarity": -0.08610919296827033,
+      "epoch": 1.3280521901211557,
+      "grad_norm": 0.9842348108889346,
+      "learning_rate": 1.238522609596134e-05,
+      "loss": 1.5909,
+      "step": 1425
+    },
+    {
+      "cosine_similarity": 0.06558809949475103,
+      "epoch": 1.3289841565703635,
+      "grad_norm": 1.264805883329856,
+      "learning_rate": 1.237832240248533e-05,
+      "loss": 1.9068,
+      "step": 1426
+    },
+    {
+      "cosine_similarity": 0.1931846058855837,
+      "epoch": 1.3299161230195713,
+      "grad_norm": 1.1803963045793122,
+      "learning_rate": 1.2371418709009322e-05,
+      "loss": 1.784,
+      "step": 1427
+    },
+    {
+      "cosine_similarity": 0.36465150042319244,
+      "epoch": 1.330848089468779,
+      "grad_norm": 1.093956431275058,
+      "learning_rate": 1.2364515015533311e-05,
+      "loss": 1.8564,
+      "step": 1428
+    },
+    {
+      "cosine_similarity": 0.258321778317455,
+      "epoch": 1.3317800559179869,
+      "grad_norm": 1.1343678583402443,
+      "learning_rate": 1.2357611322057302e-05,
+      "loss": 1.9484,
+      "step": 1429
+    },
+    {
+      "cosine_similarity": 0.18494805554294516,
+      "epoch": 1.3327120223671947,
+      "grad_norm": 1.3541649894452275,
+      "learning_rate": 1.2350707628581291e-05,
+      "loss": 1.8858,
+      "step": 1430
+    },
+    {
+      "cosine_similarity": 0.27835353021231785,
+      "epoch": 1.3336439888164027,
+      "grad_norm": 1.2173008716991547,
+      "learning_rate": 1.2343803935105284e-05,
+      "loss": 1.8346,
+      "step": 1431
+    },
+    {
+      "cosine_similarity": 0.24356668809473983,
+      "epoch": 1.3345759552656105,
+      "grad_norm": 1.1531816608844483,
+      "learning_rate": 1.2336900241629273e-05,
+      "loss": 1.985,
+      "step": 1432
+    },
+    {
+      "cosine_similarity": 0.01221131970011076,
+      "epoch": 1.3355079217148182,
+      "grad_norm": 1.2182930849595255,
+      "learning_rate": 1.2329996548153262e-05,
+      "loss": 1.5868,
+      "step": 1433
+    },
+    {
+      "cosine_similarity": 0.23050410645975192,
+      "epoch": 1.336439888164026,
+      "grad_norm": 1.454268928206258,
+      "learning_rate": 1.2323092854677254e-05,
+      "loss": 1.4012,
+      "step": 1434
+    },
+    {
+      "cosine_similarity": -0.1875197823120269,
+      "epoch": 1.337371854613234,
+      "grad_norm": 1.338721407639912,
+      "learning_rate": 1.2316189161201245e-05,
+      "loss": 1.7605,
+      "step": 1435
+    },
+    {
+      "cosine_similarity": 0.3793059459695217,
+      "epoch": 1.3383038210624418,
+      "grad_norm": 1.1243346258825213,
+      "learning_rate": 1.2309285467725234e-05,
+      "loss": 1.7925,
+      "step": 1436
+    },
+    {
+      "cosine_similarity": 0.16188595844677447,
+      "epoch": 1.3392357875116496,
+      "grad_norm": 1.6800822977472898,
+      "learning_rate": 1.2302381774249223e-05,
+      "loss": 1.7408,
+      "step": 1437
+    },
+    {
+      "cosine_similarity": 0.3325018097226398,
+      "epoch": 1.3401677539608574,
+      "grad_norm": 1.1554148098914743,
+      "learning_rate": 1.2295478080773215e-05,
+      "loss": 1.7152,
+      "step": 1438
+    },
+    {
+      "cosine_similarity": 0.1289948070938897,
+      "epoch": 1.3410997204100652,
+      "grad_norm": 1.1940497621931176,
+      "learning_rate": 1.2288574387297206e-05,
+      "loss": 2.0834,
+      "step": 1439
+    },
+    {
+      "cosine_similarity": 0.05146046661460739,
+      "epoch": 1.342031686859273,
+      "grad_norm": 1.1011638154726344,
+      "learning_rate": 1.2281670693821195e-05,
+      "loss": 2.146,
+      "step": 1440
+    },
+    {
+      "cosine_similarity": 0.03299708664952559,
+      "epoch": 1.3429636533084808,
+      "grad_norm": 2.189362902156097,
+      "learning_rate": 1.2274767000345184e-05,
+      "loss": 2.0171,
+      "step": 1441
+    },
+    {
+      "cosine_similarity": 0.009558766323511679,
+      "epoch": 1.3438956197576888,
+      "grad_norm": 1.2480987009830808,
+      "learning_rate": 1.2267863306869176e-05,
+      "loss": 1.8394,
+      "step": 1442
+    },
+    {
+      "cosine_similarity": 0.05553360916145056,
+      "epoch": 1.3448275862068966,
+      "grad_norm": 1.0169199907014514,
+      "learning_rate": 1.2260959613393166e-05,
+      "loss": 1.5908,
+      "step": 1443
+    },
+    {
+      "cosine_similarity": 0.11855107474086968,
+      "epoch": 1.3457595526561044,
+      "grad_norm": 1.463201518236932,
+      "learning_rate": 1.2254055919917156e-05,
+      "loss": 1.8734,
+      "step": 1444
+    },
+    {
+      "cosine_similarity": -0.15434047115889707,
+      "epoch": 1.3466915191053122,
+      "grad_norm": 1.1069682578928637,
+      "learning_rate": 1.2247152226441149e-05,
+      "loss": 1.9683,
+      "step": 1445
+    },
+    {
+      "cosine_similarity": 0.12100041736049422,
+      "epoch": 1.3476234855545202,
+      "grad_norm": 1.1935897843759413,
+      "learning_rate": 1.2240248532965138e-05,
+      "loss": 1.9507,
+      "step": 1446
+    },
+    {
+      "cosine_similarity": 0.1459788918576497,
+      "epoch": 1.348555452003728,
+      "grad_norm": 1.2777931350468648,
+      "learning_rate": 1.2233344839489127e-05,
+      "loss": 1.9631,
+      "step": 1447
+    },
+    {
+      "cosine_similarity": 0.2712405520948209,
+      "epoch": 1.3494874184529357,
+      "grad_norm": 1.314732495228994,
+      "learning_rate": 1.2226441146013117e-05,
+      "loss": 1.6035,
+      "step": 1448
+    },
+    {
+      "cosine_similarity": 0.258778745193998,
+      "epoch": 1.3504193849021435,
+      "grad_norm": 1.0680654520804906,
+      "learning_rate": 1.221953745253711e-05,
+      "loss": 1.699,
+      "step": 1449
+    },
+    {
+      "cosine_similarity": 0.06287008388014327,
+      "epoch": 1.3513513513513513,
+      "grad_norm": 1.113565399502801,
+      "learning_rate": 1.2212633759061099e-05,
+      "loss": 1.6116,
+      "step": 1450
+    },
+    {
+      "cosine_similarity": 0.3043271295157761,
+      "epoch": 1.352283317800559,
+      "grad_norm": 1.361766247660109,
+      "learning_rate": 1.2205730065585088e-05,
+      "loss": 1.786,
+      "step": 1451
+    },
+    {
+      "cosine_similarity": -0.006221746019856339,
+      "epoch": 1.353215284249767,
+      "grad_norm": 1.4751388004550376,
+      "learning_rate": 1.2198826372109079e-05,
+      "loss": 1.6884,
+      "step": 1452
+    },
+    {
+      "cosine_similarity": 0.05260843846117343,
+      "epoch": 1.354147250698975,
+      "grad_norm": 1.0250845953184522,
+      "learning_rate": 1.219192267863307e-05,
+      "loss": 1.3349,
+      "step": 1453
+    },
+    {
+      "cosine_similarity": 0.16354271589247177,
+      "epoch": 1.3550792171481827,
+      "grad_norm": 0.9166629101661402,
+      "learning_rate": 1.218501898515706e-05,
+      "loss": 1.4515,
+      "step": 1454
+    },
+    {
+      "cosine_similarity": 0.336595891396137,
+      "epoch": 1.3560111835973905,
+      "grad_norm": 1.3029901617953847,
+      "learning_rate": 1.217811529168105e-05,
+      "loss": 1.6183,
+      "step": 1455
+    },
+    {
+      "cosine_similarity": 0.24794396807035138,
+      "epoch": 1.3569431500465983,
+      "grad_norm": 1.18079485837695,
+      "learning_rate": 1.2171211598205042e-05,
+      "loss": 1.6479,
+      "step": 1456
+    },
+    {
+      "cosine_similarity": 0.1746965139595073,
+      "epoch": 1.3578751164958063,
+      "grad_norm": 1.4889132596655048,
+      "learning_rate": 1.216430790472903e-05,
+      "loss": 1.7863,
+      "step": 1457
+    },
+    {
+      "cosine_similarity": 0.03457707159133615,
+      "epoch": 1.358807082945014,
+      "grad_norm": 1.2713901366351075,
+      "learning_rate": 1.2157404211253021e-05,
+      "loss": 1.7013,
+      "step": 1458
+    },
+    {
+      "cosine_similarity": 0.3642826713604233,
+      "epoch": 1.3597390493942219,
+      "grad_norm": 1.417236830135734,
+      "learning_rate": 1.215050051777701e-05,
+      "loss": 2.1976,
+      "step": 1459
+    },
+    {
+      "cosine_similarity": 0.21353788888389685,
+      "epoch": 1.3606710158434296,
+      "grad_norm": 1.2006740559664284,
+      "learning_rate": 1.2143596824301003e-05,
+      "loss": 1.9215,
+      "step": 1460
+    },
+    {
+      "cosine_similarity": 0.19111166558371928,
+      "epoch": 1.3616029822926374,
+      "grad_norm": 1.0264557105697065,
+      "learning_rate": 1.2136693130824992e-05,
+      "loss": 1.5601,
+      "step": 1461
+    },
+    {
+      "cosine_similarity": 0.06562445723350731,
+      "epoch": 1.3625349487418452,
+      "grad_norm": 1.3462651209158227,
+      "learning_rate": 1.2129789437348983e-05,
+      "loss": 1.7873,
+      "step": 1462
+    },
+    {
+      "cosine_similarity": -0.15042099422009855,
+      "epoch": 1.363466915191053,
+      "grad_norm": 1.2380936950160597,
+      "learning_rate": 1.2122885743872973e-05,
+      "loss": 1.9882,
+      "step": 1463
+    },
+    {
+      "cosine_similarity": 0.14989861063490112,
+      "epoch": 1.364398881640261,
+      "grad_norm": 1.0297831632198906,
+      "learning_rate": 1.2115982050396964e-05,
+      "loss": 1.561,
+      "step": 1464
+    },
+    {
+      "cosine_similarity": 0.29375546719555945,
+      "epoch": 1.3653308480894688,
+      "grad_norm": 1.2725820468340738,
+      "learning_rate": 1.2109078356920953e-05,
+      "loss": 1.7096,
+      "step": 1465
+    },
+    {
+      "cosine_similarity": 0.3961707260744218,
+      "epoch": 1.3662628145386766,
+      "grad_norm": 1.1830897120456088,
+      "learning_rate": 1.2102174663444944e-05,
+      "loss": 2.0283,
+      "step": 1466
+    },
+    {
+      "cosine_similarity": 0.0740570957336863,
+      "epoch": 1.3671947809878844,
+      "grad_norm": 1.3650070583281284,
+      "learning_rate": 1.2095270969968935e-05,
+      "loss": 1.6443,
+      "step": 1467
+    },
+    {
+      "cosine_similarity": 0.2341269762983164,
+      "epoch": 1.3681267474370924,
+      "grad_norm": 1.3076068579695812,
+      "learning_rate": 1.2088367276492925e-05,
+      "loss": 2.0621,
+      "step": 1468
+    },
+    {
+      "cosine_similarity": 0.023083611995263573,
+      "epoch": 1.3690587138863002,
+      "grad_norm": 1.3071158370109928,
+      "learning_rate": 1.2081463583016914e-05,
+      "loss": 1.7743,
+      "step": 1469
+    },
+    {
+      "cosine_similarity": 0.23729223212618586,
+      "epoch": 1.369990680335508,
+      "grad_norm": 1.2338981320014082,
+      "learning_rate": 1.2074559889540905e-05,
+      "loss": 2.1976,
+      "step": 1470
+    },
+    {
+      "cosine_similarity": 0.02161494903368805,
+      "epoch": 1.3709226467847158,
+      "grad_norm": 1.3147482344366057,
+      "learning_rate": 1.2067656196064896e-05,
+      "loss": 1.6653,
+      "step": 1471
+    },
+    {
+      "cosine_similarity": 0.26359961445008556,
+      "epoch": 1.3718546132339235,
+      "grad_norm": 1.1525266682096904,
+      "learning_rate": 1.2060752502588887e-05,
+      "loss": 2.0734,
+      "step": 1472
+    },
+    {
+      "cosine_similarity": 0.11367245797026594,
+      "epoch": 1.3727865796831313,
+      "grad_norm": 1.0728233900903046,
+      "learning_rate": 1.2053848809112876e-05,
+      "loss": 1.8363,
+      "step": 1473
+    },
+    {
+      "cosine_similarity": 0.5406908744847091,
+      "epoch": 1.3737185461323391,
+      "grad_norm": 1.0857804282995491,
+      "learning_rate": 1.2046945115636868e-05,
+      "loss": 1.7218,
+      "step": 1474
+    },
+    {
+      "cosine_similarity": 0.3141112587950942,
+      "epoch": 1.3746505125815471,
+      "grad_norm": 1.2941484479689513,
+      "learning_rate": 1.2040041422160857e-05,
+      "loss": 1.7645,
+      "step": 1475
+    },
+    {
+      "cosine_similarity": 0.17652957937559094,
+      "epoch": 1.375582479030755,
+      "grad_norm": 1.068747793654174,
+      "learning_rate": 1.2033137728684848e-05,
+      "loss": 1.7132,
+      "step": 1476
+    },
+    {
+      "cosine_similarity": 0.016901735483970514,
+      "epoch": 1.3765144454799627,
+      "grad_norm": 1.0283989685222696,
+      "learning_rate": 1.2026234035208837e-05,
+      "loss": 1.5462,
+      "step": 1477
+    },
+    {
+      "cosine_similarity": 0.15522560786819983,
+      "epoch": 1.3774464119291705,
+      "grad_norm": 1.092149594609746,
+      "learning_rate": 1.201933034173283e-05,
+      "loss": 1.8898,
+      "step": 1478
+    },
+    {
+      "cosine_similarity": 0.32871065482662776,
+      "epoch": 1.3783783783783785,
+      "grad_norm": 1.110127270595407,
+      "learning_rate": 1.2012426648256818e-05,
+      "loss": 1.426,
+      "step": 1479
+    },
+    {
+      "cosine_similarity": 0.02810653586714183,
+      "epoch": 1.3793103448275863,
+      "grad_norm": 1.0366934074718506,
+      "learning_rate": 1.2005522954780809e-05,
+      "loss": 1.9109,
+      "step": 1480
+    },
+    {
+      "cosine_similarity": -0.024726264024574066,
+      "epoch": 1.380242311276794,
+      "grad_norm": 1.4368960949081813,
+      "learning_rate": 1.1998619261304798e-05,
+      "loss": 1.7932,
+      "step": 1481
+    },
+    {
+      "cosine_similarity": 0.20621676854749652,
+      "epoch": 1.3811742777260019,
+      "grad_norm": 1.2755568122210519,
+      "learning_rate": 1.199171556782879e-05,
+      "loss": 1.8678,
+      "step": 1482
+    },
+    {
+      "cosine_similarity": 0.20277633176163695,
+      "epoch": 1.3821062441752097,
+      "grad_norm": 1.16821490129891,
+      "learning_rate": 1.198481187435278e-05,
+      "loss": 1.7787,
+      "step": 1483
+    },
+    {
+      "cosine_similarity": -0.03121746198632975,
+      "epoch": 1.3830382106244175,
+      "grad_norm": 1.167727031974801,
+      "learning_rate": 1.1977908180876769e-05,
+      "loss": 1.9833,
+      "step": 1484
+    },
+    {
+      "cosine_similarity": 0.22803000211534946,
+      "epoch": 1.3839701770736252,
+      "grad_norm": 1.40370886177378,
+      "learning_rate": 1.1971004487400761e-05,
+      "loss": 1.926,
+      "step": 1485
+    },
+    {
+      "cosine_similarity": 0.16559175132754742,
+      "epoch": 1.3849021435228333,
+      "grad_norm": 1.2105689870807477,
+      "learning_rate": 1.1964100793924752e-05,
+      "loss": 1.8156,
+      "step": 1486
+    },
+    {
+      "cosine_similarity": 0.10208788736274604,
+      "epoch": 1.385834109972041,
+      "grad_norm": 1.164111523889717,
+      "learning_rate": 1.195719710044874e-05,
+      "loss": 1.4058,
+      "step": 1487
+    },
+    {
+      "cosine_similarity": 0.09451749896636837,
+      "epoch": 1.3867660764212488,
+      "grad_norm": 0.9941065340906555,
+      "learning_rate": 1.195029340697273e-05,
+      "loss": 1.5565,
+      "step": 1488
+    },
+    {
+      "cosine_similarity": 0.3210079191260303,
+      "epoch": 1.3876980428704566,
+      "grad_norm": 1.2626503492633108,
+      "learning_rate": 1.1943389713496722e-05,
+      "loss": 1.6943,
+      "step": 1489
+    },
+    {
+      "cosine_similarity": 0.6013960324602535,
+      "epoch": 1.3886300093196646,
+      "grad_norm": 1.1401190034864375,
+      "learning_rate": 1.1936486020020713e-05,
+      "loss": 1.9213,
+      "step": 1490
+    },
+    {
+      "cosine_similarity": -0.03691077672759331,
+      "epoch": 1.3895619757688724,
+      "grad_norm": 1.123210977729181,
+      "learning_rate": 1.1929582326544702e-05,
+      "loss": 1.6101,
+      "step": 1491
+    },
+    {
+      "cosine_similarity": 0.3490258948908722,
+      "epoch": 1.3904939422180802,
+      "grad_norm": 1.262388718737804,
+      "learning_rate": 1.1922678633068691e-05,
+      "loss": 1.8468,
+      "step": 1492
+    },
+    {
+      "cosine_similarity": 0.45233819409768555,
+      "epoch": 1.391425908667288,
+      "grad_norm": 1.296692701113406,
+      "learning_rate": 1.1915774939592684e-05,
+      "loss": 1.8711,
+      "step": 1493
+    },
+    {
+      "cosine_similarity": 0.21750953058384534,
+      "epoch": 1.3923578751164958,
+      "grad_norm": 1.2524519384097237,
+      "learning_rate": 1.1908871246116673e-05,
+      "loss": 1.6915,
+      "step": 1494
+    },
+    {
+      "cosine_similarity": 0.2318699022340053,
+      "epoch": 1.3932898415657036,
+      "grad_norm": 1.2347290725661217,
+      "learning_rate": 1.1901967552640663e-05,
+      "loss": 1.7399,
+      "step": 1495
+    },
+    {
+      "cosine_similarity": 0.39250710141787964,
+      "epoch": 1.3942218080149114,
+      "grad_norm": 1.3310214574045574,
+      "learning_rate": 1.1895063859164656e-05,
+      "loss": 1.8261,
+      "step": 1496
+    },
+    {
+      "cosine_similarity": 0.14235888957204465,
+      "epoch": 1.3951537744641194,
+      "grad_norm": 1.3150256926828296,
+      "learning_rate": 1.1888160165688645e-05,
+      "loss": 1.8436,
+      "step": 1497
+    },
+    {
+      "cosine_similarity": 0.16727170662108698,
+      "epoch": 1.3960857409133272,
+      "grad_norm": 1.1536053357162963,
+      "learning_rate": 1.1881256472212634e-05,
+      "loss": 1.5759,
+      "step": 1498
+    },
+    {
+      "cosine_similarity": 0.13035898464242923,
+      "epoch": 1.397017707362535,
+      "grad_norm": 1.1429945137831605,
+      "learning_rate": 1.1874352778736625e-05,
+      "loss": 1.7122,
+      "step": 1499
+    },
+    {
+      "cosine_similarity": 0.3321658791112968,
+      "epoch": 1.3979496738117427,
+      "grad_norm": 1.3956703602142162,
+      "learning_rate": 1.1867449085260617e-05,
+      "loss": 1.742,
+      "step": 1500
+    },
+    {
+      "cosine_similarity": 0.4098520271941178,
+      "epoch": 1.3988816402609507,
+      "grad_norm": 1.3985075171858699,
+      "learning_rate": 1.1860545391784606e-05,
+      "loss": 1.6817,
+      "step": 1501
+    },
+    {
+      "cosine_similarity": 0.05192152796156245,
+      "epoch": 1.3998136067101585,
+      "grad_norm": 1.3090350317788346,
+      "learning_rate": 1.1853641698308595e-05,
+      "loss": 1.7515,
+      "step": 1502
+    },
+    {
+      "cosine_similarity": 0.17258297630030248,
+      "epoch": 1.4007455731593663,
+      "grad_norm": 1.1023817797645485,
+      "learning_rate": 1.1846738004832586e-05,
+      "loss": 1.7409,
+      "step": 1503
+    },
+    {
+      "cosine_similarity": -0.15328025065439962,
+      "epoch": 1.401677539608574,
+      "grad_norm": 1.2455586198254853,
+      "learning_rate": 1.1839834311356576e-05,
+      "loss": 1.6715,
+      "step": 1504
+    },
+    {
+      "cosine_similarity": 0.22665326762805466,
+      "epoch": 1.402609506057782,
+      "grad_norm": 1.374921303156203,
+      "learning_rate": 1.1832930617880567e-05,
+      "loss": 2.0522,
+      "step": 1505
+    },
+    {
+      "cosine_similarity": 0.07309039712113655,
+      "epoch": 1.4035414725069897,
+      "grad_norm": 1.1425741039298067,
+      "learning_rate": 1.1826026924404556e-05,
+      "loss": 1.9628,
+      "step": 1506
+    },
+    {
+      "cosine_similarity": 0.3159705960116702,
+      "epoch": 1.4044734389561975,
+      "grad_norm": 1.4201207440403463,
+      "learning_rate": 1.1819123230928549e-05,
+      "loss": 2.0434,
+      "step": 1507
+    },
+    {
+      "cosine_similarity": 0.02310376911325343,
+      "epoch": 1.4054054054054055,
+      "grad_norm": 1.2985753705691117,
+      "learning_rate": 1.1812219537452538e-05,
+      "loss": 1.5583,
+      "step": 1508
+    },
+    {
+      "cosine_similarity": 0.1631063789388168,
+      "epoch": 1.4063373718546133,
+      "grad_norm": 1.5966736204159928,
+      "learning_rate": 1.1805315843976528e-05,
+      "loss": 2.3052,
+      "step": 1509
+    },
+    {
+      "cosine_similarity": 0.058597522382748325,
+      "epoch": 1.407269338303821,
+      "grad_norm": 1.1889846870131207,
+      "learning_rate": 1.1798412150500517e-05,
+      "loss": 1.6338,
+      "step": 1510
+    },
+    {
+      "cosine_similarity": 0.24708463209025788,
+      "epoch": 1.4082013047530288,
+      "grad_norm": 1.4355260419296796,
+      "learning_rate": 1.179150845702451e-05,
+      "loss": 1.7196,
+      "step": 1511
+    },
+    {
+      "cosine_similarity": 0.1193638916241147,
+      "epoch": 1.4091332712022366,
+      "grad_norm": 1.006931510367032,
+      "learning_rate": 1.1784604763548499e-05,
+      "loss": 1.3936,
+      "step": 1512
+    },
+    {
+      "cosine_similarity": 0.13935933484397534,
+      "epoch": 1.4100652376514446,
+      "grad_norm": 1.321438307676119,
+      "learning_rate": 1.177770107007249e-05,
+      "loss": 1.8891,
+      "step": 1513
+    },
+    {
+      "cosine_similarity": -0.007374439062927547,
+      "epoch": 1.4109972041006524,
+      "grad_norm": 1.4242209614646215,
+      "learning_rate": 1.177079737659648e-05,
+      "loss": 1.9768,
+      "step": 1514
+    },
+    {
+      "cosine_similarity": 0.16971613869128344,
+      "epoch": 1.4119291705498602,
+      "grad_norm": 1.3000067438628489,
+      "learning_rate": 1.1763893683120471e-05,
+      "loss": 1.5309,
+      "step": 1515
+    },
+    {
+      "cosine_similarity": -0.02412783990094373,
+      "epoch": 1.412861136999068,
+      "grad_norm": 1.4149467700019496,
+      "learning_rate": 1.175698998964446e-05,
+      "loss": 1.7313,
+      "step": 1516
+    },
+    {
+      "cosine_similarity": 0.5362378737342758,
+      "epoch": 1.4137931034482758,
+      "grad_norm": 0.9760949950057142,
+      "learning_rate": 1.1750086296168451e-05,
+      "loss": 1.8111,
+      "step": 1517
+    },
+    {
+      "cosine_similarity": 0.4233183950573839,
+      "epoch": 1.4147250698974836,
+      "grad_norm": 1.091946625011952,
+      "learning_rate": 1.1743182602692442e-05,
+      "loss": 1.7604,
+      "step": 1518
+    },
+    {
+      "cosine_similarity": -0.029011324261660244,
+      "epoch": 1.4156570363466916,
+      "grad_norm": 1.609450952302318,
+      "learning_rate": 1.1736278909216432e-05,
+      "loss": 1.5715,
+      "step": 1519
+    },
+    {
+      "cosine_similarity": 0.23183126154588607,
+      "epoch": 1.4165890027958994,
+      "grad_norm": 1.1322558850583313,
+      "learning_rate": 1.1729375215740421e-05,
+      "loss": 1.6644,
+      "step": 1520
+    },
+    {
+      "cosine_similarity": 0.0807729809290148,
+      "epoch": 1.4175209692451072,
+      "grad_norm": 1.019838849912351,
+      "learning_rate": 1.1722471522264412e-05,
+      "loss": 1.8671,
+      "step": 1521
+    },
+    {
+      "cosine_similarity": -0.006769548289185504,
+      "epoch": 1.418452935694315,
+      "grad_norm": 1.8940512873876225,
+      "learning_rate": 1.1715567828788403e-05,
+      "loss": 2.1097,
+      "step": 1522
+    },
+    {
+      "cosine_similarity": 0.44539266321885496,
+      "epoch": 1.4193849021435228,
+      "grad_norm": 1.1498573384725639,
+      "learning_rate": 1.1708664135312394e-05,
+      "loss": 1.6929,
+      "step": 1523
+    },
+    {
+      "cosine_similarity": 0.2610219752313373,
+      "epoch": 1.4203168685927308,
+      "grad_norm": 1.9469246063789565,
+      "learning_rate": 1.1701760441836383e-05,
+      "loss": 1.7218,
+      "step": 1524
+    },
+    {
+      "cosine_similarity": 0.0846997947527241,
+      "epoch": 1.4212488350419386,
+      "grad_norm": 1.3813487077988436,
+      "learning_rate": 1.1694856748360375e-05,
+      "loss": 1.8308,
+      "step": 1525
+    },
+    {
+      "cosine_similarity": -0.05287537554183834,
+      "epoch": 1.4221808014911463,
+      "grad_norm": 1.4707740643207792,
+      "learning_rate": 1.1687953054884364e-05,
+      "loss": 2.0971,
+      "step": 1526
+    },
+    {
+      "cosine_similarity": 0.10196135601184039,
+      "epoch": 1.4231127679403541,
+      "grad_norm": 1.2842891926142497,
+      "learning_rate": 1.1681049361408355e-05,
+      "loss": 1.8314,
+      "step": 1527
+    },
+    {
+      "cosine_similarity": -0.0743443174087208,
+      "epoch": 1.424044734389562,
+      "grad_norm": 1.3519513881014105,
+      "learning_rate": 1.1674145667932344e-05,
+      "loss": 1.7225,
+      "step": 1528
+    },
+    {
+      "cosine_similarity": 0.2131184183421666,
+      "epoch": 1.4249767008387697,
+      "grad_norm": 1.218979743982011,
+      "learning_rate": 1.1667241974456336e-05,
+      "loss": 1.5101,
+      "step": 1529
+    },
+    {
+      "cosine_similarity": 0.2787573512721672,
+      "epoch": 1.4259086672879777,
+      "grad_norm": 1.590491242933055,
+      "learning_rate": 1.1660338280980325e-05,
+      "loss": 1.8358,
+      "step": 1530
+    },
+    {
+      "cosine_similarity": 0.31866769287195323,
+      "epoch": 1.4268406337371855,
+      "grad_norm": 1.176810680351995,
+      "learning_rate": 1.1653434587504316e-05,
+      "loss": 1.6001,
+      "step": 1531
+    },
+    {
+      "cosine_similarity": 0.27355771778446525,
+      "epoch": 1.4277726001863933,
+      "grad_norm": 1.4541338253168004,
+      "learning_rate": 1.1646530894028305e-05,
+      "loss": 2.0649,
+      "step": 1532
+    },
+    {
+      "cosine_similarity": 0.0015866827452528226,
+      "epoch": 1.428704566635601,
+      "grad_norm": 1.3948242238169875,
+      "learning_rate": 1.1639627200552298e-05,
+      "loss": 1.8304,
+      "step": 1533
+    },
+    {
+      "cosine_similarity": 0.39592540982887464,
+      "epoch": 1.4296365330848089,
+      "grad_norm": 1.2133307897027559,
+      "learning_rate": 1.1632723507076287e-05,
+      "loss": 1.8157,
+      "step": 1534
+    },
+    {
+      "cosine_similarity": 0.6440754871378195,
+      "epoch": 1.4305684995340169,
+      "grad_norm": 1.1397713086752186,
+      "learning_rate": 1.1625819813600276e-05,
+      "loss": 1.7456,
+      "step": 1535
+    },
+    {
+      "cosine_similarity": 0.06728402106175323,
+      "epoch": 1.4315004659832247,
+      "grad_norm": 1.3266580887363932,
+      "learning_rate": 1.1618916120124268e-05,
+      "loss": 1.4858,
+      "step": 1536
+    },
+    {
+      "cosine_similarity": 0.32213615200044127,
+      "epoch": 1.4324324324324325,
+      "grad_norm": 1.204314030791446,
+      "learning_rate": 1.1612012426648259e-05,
+      "loss": 1.9759,
+      "step": 1537
+    },
+    {
+      "cosine_similarity": 0.20850364301902125,
+      "epoch": 1.4333643988816402,
+      "grad_norm": 1.0501948513484993,
+      "learning_rate": 1.1605108733172248e-05,
+      "loss": 1.74,
+      "step": 1538
+    },
+    {
+      "cosine_similarity": 0.22304390885043232,
+      "epoch": 1.434296365330848,
+      "grad_norm": 1.2412826111140063,
+      "learning_rate": 1.1598205039696237e-05,
+      "loss": 1.8684,
+      "step": 1539
+    },
+    {
+      "cosine_similarity": -0.2597670335351652,
+      "epoch": 1.4352283317800558,
+      "grad_norm": 1.068657949429026,
+      "learning_rate": 1.159130134622023e-05,
+      "loss": 1.7486,
+      "step": 1540
+    },
+    {
+      "cosine_similarity": 0.4863679834359496,
+      "epoch": 1.4361602982292636,
+      "grad_norm": 1.0631334047543437,
+      "learning_rate": 1.158439765274422e-05,
+      "loss": 1.6811,
+      "step": 1541
+    },
+    {
+      "cosine_similarity": 0.05157346269832577,
+      "epoch": 1.4370922646784716,
+      "grad_norm": 1.4258845444462018,
+      "learning_rate": 1.1577493959268209e-05,
+      "loss": 1.8688,
+      "step": 1542
+    },
+    {
+      "cosine_similarity": 0.30740990657095607,
+      "epoch": 1.4380242311276794,
+      "grad_norm": 2.257664144391525,
+      "learning_rate": 1.1570590265792198e-05,
+      "loss": 1.987,
+      "step": 1543
+    },
+    {
+      "cosine_similarity": 0.21624282023224115,
+      "epoch": 1.4389561975768872,
+      "grad_norm": 1.151474155163488,
+      "learning_rate": 1.156368657231619e-05,
+      "loss": 1.7183,
+      "step": 1544
+    },
+    {
+      "cosine_similarity": 0.2596795733728236,
+      "epoch": 1.439888164026095,
+      "grad_norm": 1.2652903603168795,
+      "learning_rate": 1.155678287884018e-05,
+      "loss": 1.6213,
+      "step": 1545
+    },
+    {
+      "cosine_similarity": 0.10270367270719893,
+      "epoch": 1.440820130475303,
+      "grad_norm": 1.0933571561420956,
+      "learning_rate": 1.154987918536417e-05,
+      "loss": 1.7856,
+      "step": 1546
+    },
+    {
+      "cosine_similarity": -0.05023177005674341,
+      "epoch": 1.4417520969245108,
+      "grad_norm": 0.9474738297543831,
+      "learning_rate": 1.1542975491888163e-05,
+      "loss": 1.5921,
+      "step": 1547
+    },
+    {
+      "cosine_similarity": 0.19866845091252014,
+      "epoch": 1.4426840633737186,
+      "grad_norm": 1.529116981426749,
+      "learning_rate": 1.1536071798412152e-05,
+      "loss": 1.6958,
+      "step": 1548
+    },
+    {
+      "cosine_similarity": 0.640251439359715,
+      "epoch": 1.4436160298229264,
+      "grad_norm": 1.2076067807568052,
+      "learning_rate": 1.152916810493614e-05,
+      "loss": 1.865,
+      "step": 1549
+    },
+    {
+      "cosine_similarity": 0.1341249891256666,
+      "epoch": 1.4445479962721341,
+      "grad_norm": 1.204505013242542,
+      "learning_rate": 1.1522264411460132e-05,
+      "loss": 1.8523,
+      "step": 1550
+    },
+    {
+      "cosine_similarity": 0.11533031886060982,
+      "epoch": 1.445479962721342,
+      "grad_norm": 1.4748521555000087,
+      "learning_rate": 1.1515360717984124e-05,
+      "loss": 1.8198,
+      "step": 1551
+    },
+    {
+      "cosine_similarity": 0.43365847046298006,
+      "epoch": 1.4464119291705497,
+      "grad_norm": 1.1264148351473087,
+      "learning_rate": 1.1508457024508113e-05,
+      "loss": 1.6895,
+      "step": 1552
+    },
+    {
+      "cosine_similarity": 0.3447927948699419,
+      "epoch": 1.4473438956197577,
+      "grad_norm": 1.0758964237777784,
+      "learning_rate": 1.1501553331032102e-05,
+      "loss": 1.5213,
+      "step": 1553
+    },
+    {
+      "cosine_similarity": 0.044902330510150315,
+      "epoch": 1.4482758620689655,
+      "grad_norm": 1.0840276646186835,
+      "learning_rate": 1.1494649637556094e-05,
+      "loss": 1.9522,
+      "step": 1554
+    },
+    {
+      "cosine_similarity": 0.20243559852148546,
+      "epoch": 1.4492078285181733,
+      "grad_norm": 1.2358838221294102,
+      "learning_rate": 1.1487745944080084e-05,
+      "loss": 1.6324,
+      "step": 1555
+    },
+    {
+      "cosine_similarity": 0.30636150856489486,
+      "epoch": 1.450139794967381,
+      "grad_norm": 1.0759312015919749,
+      "learning_rate": 1.1480842250604074e-05,
+      "loss": 1.9135,
+      "step": 1556
+    },
+    {
+      "cosine_similarity": 0.21187743214388552,
+      "epoch": 1.4510717614165891,
+      "grad_norm": 0.9893974300314241,
+      "learning_rate": 1.1473938557128063e-05,
+      "loss": 1.932,
+      "step": 1557
+    },
+    {
+      "cosine_similarity": 0.24948935581712864,
+      "epoch": 1.452003727865797,
+      "grad_norm": 1.137145700798436,
+      "learning_rate": 1.1467034863652056e-05,
+      "loss": 1.8547,
+      "step": 1558
+    },
+    {
+      "cosine_similarity": 0.1636982030383784,
+      "epoch": 1.4529356943150047,
+      "grad_norm": 1.371520250308096,
+      "learning_rate": 1.1460131170176045e-05,
+      "loss": 1.7153,
+      "step": 1559
+    },
+    {
+      "cosine_similarity": 0.10798225521693551,
+      "epoch": 1.4538676607642125,
+      "grad_norm": 1.2476223012762084,
+      "learning_rate": 1.1453227476700035e-05,
+      "loss": 1.5649,
+      "step": 1560
+    },
+    {
+      "cosine_similarity": 0.11450244044102657,
+      "epoch": 1.4547996272134203,
+      "grad_norm": 1.644019836245965,
+      "learning_rate": 1.1446323783224025e-05,
+      "loss": 1.7789,
+      "step": 1561
+    },
+    {
+      "cosine_similarity": 0.0652898118972232,
+      "epoch": 1.455731593662628,
+      "grad_norm": 1.147569805895284,
+      "learning_rate": 1.1439420089748017e-05,
+      "loss": 1.8304,
+      "step": 1562
+    },
+    {
+      "cosine_similarity": -0.10188593904779925,
+      "epoch": 1.4566635601118358,
+      "grad_norm": 1.6573623343184476,
+      "learning_rate": 1.1432516396272006e-05,
+      "loss": 1.7644,
+      "step": 1563
+    },
+    {
+      "cosine_similarity": 0.1523019721170746,
+      "epoch": 1.4575955265610439,
+      "grad_norm": 1.3632337558477379,
+      "learning_rate": 1.1425612702795997e-05,
+      "loss": 2.1599,
+      "step": 1564
+    },
+    {
+      "cosine_similarity": 0.22589403861834664,
+      "epoch": 1.4585274930102516,
+      "grad_norm": 1.1497607252150246,
+      "learning_rate": 1.1418709009319987e-05,
+      "loss": 1.6502,
+      "step": 1565
+    },
+    {
+      "cosine_similarity": -0.033262388895503404,
+      "epoch": 1.4594594594594594,
+      "grad_norm": 1.1588815998546507,
+      "learning_rate": 1.1411805315843978e-05,
+      "loss": 1.855,
+      "step": 1566
+    },
+    {
+      "cosine_similarity": -0.05266178091788692,
+      "epoch": 1.4603914259086672,
+      "grad_norm": 1.4794173274813245,
+      "learning_rate": 1.1404901622367967e-05,
+      "loss": 1.6194,
+      "step": 1567
+    },
+    {
+      "cosine_similarity": -0.05195688467460765,
+      "epoch": 1.4613233923578752,
+      "grad_norm": 1.3291504883141214,
+      "learning_rate": 1.1397997928891958e-05,
+      "loss": 1.946,
+      "step": 1568
+    },
+    {
+      "cosine_similarity": 0.05505533216199474,
+      "epoch": 1.462255358807083,
+      "grad_norm": 1.360937297144579,
+      "learning_rate": 1.1391094235415949e-05,
+      "loss": 2.0336,
+      "step": 1569
+    },
+    {
+      "cosine_similarity": 0.085969780206034,
+      "epoch": 1.4631873252562908,
+      "grad_norm": 1.3315835775812122,
+      "learning_rate": 1.138419054193994e-05,
+      "loss": 1.9545,
+      "step": 1570
+    },
+    {
+      "cosine_similarity": 0.10582720605252052,
+      "epoch": 1.4641192917054986,
+      "grad_norm": 1.1121142635688215,
+      "learning_rate": 1.1377286848463928e-05,
+      "loss": 1.6338,
+      "step": 1571
+    },
+    {
+      "cosine_similarity": 0.15814469315982074,
+      "epoch": 1.4650512581547064,
+      "grad_norm": 1.0802300550134074,
+      "learning_rate": 1.137038315498792e-05,
+      "loss": 1.7881,
+      "step": 1572
+    },
+    {
+      "cosine_similarity": 0.008603515877482478,
+      "epoch": 1.4659832246039142,
+      "grad_norm": 1.0884263231828137,
+      "learning_rate": 1.136347946151191e-05,
+      "loss": 1.6484,
+      "step": 1573
+    },
+    {
+      "cosine_similarity": 0.17191572056665153,
+      "epoch": 1.466915191053122,
+      "grad_norm": 1.1856924082388378,
+      "learning_rate": 1.13565757680359e-05,
+      "loss": 1.7913,
+      "step": 1574
+    },
+    {
+      "cosine_similarity": 0.1339896554474356,
+      "epoch": 1.46784715750233,
+      "grad_norm": 1.2767401533256046,
+      "learning_rate": 1.134967207455989e-05,
+      "loss": 1.8097,
+      "step": 1575
+    },
+    {
+      "cosine_similarity": 0.08381731888322398,
+      "epoch": 1.4687791239515378,
+      "grad_norm": 1.547290321272811,
+      "learning_rate": 1.1342768381083882e-05,
+      "loss": 1.6921,
+      "step": 1576
+    },
+    {
+      "cosine_similarity": 0.4484140431417436,
+      "epoch": 1.4697110904007455,
+      "grad_norm": 1.4183504363777373,
+      "learning_rate": 1.1335864687607871e-05,
+      "loss": 1.8299,
+      "step": 1577
+    },
+    {
+      "cosine_similarity": 0.18274863245839063,
+      "epoch": 1.4706430568499533,
+      "grad_norm": 1.1056194092779719,
+      "learning_rate": 1.1328960994131862e-05,
+      "loss": 1.9007,
+      "step": 1578
+    },
+    {
+      "cosine_similarity": 0.19704960354615686,
+      "epoch": 1.4715750232991613,
+      "grad_norm": 1.5431910866929668,
+      "learning_rate": 1.1322057300655851e-05,
+      "loss": 2.0706,
+      "step": 1579
+    },
+    {
+      "cosine_similarity": 0.08541354111131988,
+      "epoch": 1.4725069897483691,
+      "grad_norm": 1.2315343352074197,
+      "learning_rate": 1.1315153607179843e-05,
+      "loss": 1.7628,
+      "step": 1580
+    },
+    {
+      "cosine_similarity": 0.1225605705178567,
+      "epoch": 1.473438956197577,
+      "grad_norm": 2.974444681509899,
+      "learning_rate": 1.1308249913703832e-05,
+      "loss": 1.5585,
+      "step": 1581
+    },
+    {
+      "cosine_similarity": 0.429113673065914,
+      "epoch": 1.4743709226467847,
+      "grad_norm": 1.6191626014618217,
+      "learning_rate": 1.1301346220227823e-05,
+      "loss": 1.6762,
+      "step": 1582
+    },
+    {
+      "cosine_similarity": 0.13614474267913196,
+      "epoch": 1.4753028890959925,
+      "grad_norm": 1.1918718206791326,
+      "learning_rate": 1.1294442526751812e-05,
+      "loss": 1.6612,
+      "step": 1583
+    },
+    {
+      "cosine_similarity": 0.14586529436986959,
+      "epoch": 1.4762348555452003,
+      "grad_norm": 1.297934313966871,
+      "learning_rate": 1.1287538833275805e-05,
+      "loss": 2.1827,
+      "step": 1584
+    },
+    {
+      "cosine_similarity": 0.12502012862741946,
+      "epoch": 1.477166821994408,
+      "grad_norm": 1.137487522180814,
+      "learning_rate": 1.1280635139799794e-05,
+      "loss": 1.7741,
+      "step": 1585
+    },
+    {
+      "cosine_similarity": 0.2312425289449637,
+      "epoch": 1.478098788443616,
+      "grad_norm": 1.2242489038176434,
+      "learning_rate": 1.1273731446323783e-05,
+      "loss": 1.8783,
+      "step": 1586
+    },
+    {
+      "cosine_similarity": 0.05023847310261108,
+      "epoch": 1.4790307548928239,
+      "grad_norm": 1.0226056477326346,
+      "learning_rate": 1.1266827752847775e-05,
+      "loss": 1.5947,
+      "step": 1587
+    },
+    {
+      "cosine_similarity": 0.1493093092769437,
+      "epoch": 1.4799627213420317,
+      "grad_norm": 1.165502072416851,
+      "learning_rate": 1.1259924059371766e-05,
+      "loss": 1.949,
+      "step": 1588
+    },
+    {
+      "cosine_similarity": 0.4353452903391891,
+      "epoch": 1.4808946877912395,
+      "grad_norm": 1.1705189458215346,
+      "learning_rate": 1.1253020365895755e-05,
+      "loss": 1.7275,
+      "step": 1589
+    },
+    {
+      "cosine_similarity": 0.09112677652226812,
+      "epoch": 1.4818266542404475,
+      "grad_norm": 1.1067729598646263,
+      "learning_rate": 1.1246116672419744e-05,
+      "loss": 1.9462,
+      "step": 1590
+    },
+    {
+      "cosine_similarity": 0.14877072569450875,
+      "epoch": 1.4827586206896552,
+      "grad_norm": 1.3496365641686643,
+      "learning_rate": 1.1239212978943736e-05,
+      "loss": 1.8303,
+      "step": 1591
+    },
+    {
+      "cosine_similarity": 0.06771576792963004,
+      "epoch": 1.483690587138863,
+      "grad_norm": 1.2085709635743629,
+      "learning_rate": 1.1232309285467727e-05,
+      "loss": 2.1436,
+      "step": 1592
+    },
+    {
+      "cosine_similarity": 0.15412339110425147,
+      "epoch": 1.4846225535880708,
+      "grad_norm": 1.0715182233576859,
+      "learning_rate": 1.1225405591991716e-05,
+      "loss": 1.4764,
+      "step": 1593
+    },
+    {
+      "cosine_similarity": 0.4450054332597659,
+      "epoch": 1.4855545200372786,
+      "grad_norm": 1.0945387326630494,
+      "learning_rate": 1.1218501898515709e-05,
+      "loss": 2.0354,
+      "step": 1594
+    },
+    {
+      "cosine_similarity": 0.35825148660464745,
+      "epoch": 1.4864864864864864,
+      "grad_norm": 1.3439883197969966,
+      "learning_rate": 1.1211598205039698e-05,
+      "loss": 1.757,
+      "step": 1595
+    },
+    {
+      "cosine_similarity": 0.4529444795197679,
+      "epoch": 1.4874184529356942,
+      "grad_norm": 1.2006454693233126,
+      "learning_rate": 1.1204694511563687e-05,
+      "loss": 1.8125,
+      "step": 1596
+    },
+    {
+      "cosine_similarity": 0.3957998122867132,
+      "epoch": 1.4883504193849022,
+      "grad_norm": 1.21582806423762,
+      "learning_rate": 1.1197790818087677e-05,
+      "loss": 1.7742,
+      "step": 1597
+    },
+    {
+      "cosine_similarity": 0.08360395511690945,
+      "epoch": 1.48928238583411,
+      "grad_norm": 1.2837042176315432,
+      "learning_rate": 1.119088712461167e-05,
+      "loss": 1.6494,
+      "step": 1598
+    },
+    {
+      "cosine_similarity": 0.010572303680564927,
+      "epoch": 1.4902143522833178,
+      "grad_norm": 1.4121085436968446,
+      "learning_rate": 1.1183983431135659e-05,
+      "loss": 1.7795,
+      "step": 1599
+    },
+    {
+      "cosine_similarity": 0.3025661434924069,
+      "epoch": 1.4911463187325256,
+      "grad_norm": 1.244641742039753,
+      "learning_rate": 1.1177079737659648e-05,
+      "loss": 1.5185,
+      "step": 1600
+    },
+    {
+      "cosine_similarity": 0.6063774594668032,
+      "epoch": 1.4920782851817336,
+      "grad_norm": 1.131468339920731,
+      "learning_rate": 1.1170176044183639e-05,
+      "loss": 2.0931,
+      "step": 1601
+    },
+    {
+      "cosine_similarity": 0.1057308182349369,
+      "epoch": 1.4930102516309414,
+      "grad_norm": 1.1152530474105224,
+      "learning_rate": 1.1163272350707631e-05,
+      "loss": 1.8327,
+      "step": 1602
+    },
+    {
+      "cosine_similarity": 0.029201557198500788,
+      "epoch": 1.4939422180801492,
+      "grad_norm": 1.078740076742949,
+      "learning_rate": 1.115636865723162e-05,
+      "loss": 1.6668,
+      "step": 1603
+    },
+    {
+      "cosine_similarity": 0.020132934543301718,
+      "epoch": 1.494874184529357,
+      "grad_norm": 1.2617749690227125,
+      "learning_rate": 1.1149464963755609e-05,
+      "loss": 1.736,
+      "step": 1604
+    },
+    {
+      "cosine_similarity": 0.4616640758937417,
+      "epoch": 1.4958061509785647,
+      "grad_norm": 1.1664111038722131,
+      "learning_rate": 1.1142561270279602e-05,
+      "loss": 2.171,
+      "step": 1605
+    },
+    {
+      "cosine_similarity": 0.03540473847756257,
+      "epoch": 1.4967381174277725,
+      "grad_norm": 1.4003252913337803,
+      "learning_rate": 1.113565757680359e-05,
+      "loss": 2.2114,
+      "step": 1606
+    },
+    {
+      "cosine_similarity": 0.0747446903436925,
+      "epoch": 1.4976700838769803,
+      "grad_norm": 1.1683338855966716,
+      "learning_rate": 1.1128753883327581e-05,
+      "loss": 1.7711,
+      "step": 1607
+    },
+    {
+      "cosine_similarity": -0.07965097322522381,
+      "epoch": 1.4986020503261883,
+      "grad_norm": 1.198950721336685,
+      "learning_rate": 1.112185018985157e-05,
+      "loss": 2.0059,
+      "step": 1608
+    },
+    {
+      "cosine_similarity": -0.02879516489536783,
+      "epoch": 1.499534016775396,
+      "grad_norm": 0.9756792281340697,
+      "learning_rate": 1.1114946496375563e-05,
+      "loss": 1.445,
+      "step": 1609
+    },
+    {
+      "cosine_similarity": 0.46860413383009153,
+      "epoch": 1.500465983224604,
+      "grad_norm": 1.4821727952238735,
+      "learning_rate": 1.1108042802899552e-05,
+      "loss": 1.9093,
+      "step": 1610
+    },
+    {
+      "cosine_similarity": 0.19668919109355779,
+      "epoch": 1.501397949673812,
+      "grad_norm": 1.1634414868927774,
+      "learning_rate": 1.1101139109423543e-05,
+      "loss": 1.9468,
+      "step": 1611
+    },
+    {
+      "cosine_similarity": 0.21512391635944822,
+      "epoch": 1.5023299161230197,
+      "grad_norm": 1.3525202514261405,
+      "learning_rate": 1.1094235415947532e-05,
+      "loss": 1.8918,
+      "step": 1612
+    },
+    {
+      "cosine_similarity": 0.10364588142083095,
+      "epoch": 1.5032618825722275,
+      "grad_norm": 1.2141221297869202,
+      "learning_rate": 1.1087331722471524e-05,
+      "loss": 1.7461,
+      "step": 1613
+    },
+    {
+      "cosine_similarity": 0.07276535331087762,
+      "epoch": 1.5041938490214353,
+      "grad_norm": 1.3020528015839303,
+      "learning_rate": 1.1080428028995513e-05,
+      "loss": 1.8669,
+      "step": 1614
+    },
+    {
+      "cosine_similarity": -0.10166615832574326,
+      "epoch": 1.505125815470643,
+      "grad_norm": 1.169204613802049,
+      "learning_rate": 1.1073524335519504e-05,
+      "loss": 1.4854,
+      "step": 1615
+    },
+    {
+      "cosine_similarity": 0.3464662428091075,
+      "epoch": 1.5060577819198508,
+      "grad_norm": 1.2353628032608732,
+      "learning_rate": 1.1066620642043494e-05,
+      "loss": 1.7309,
+      "step": 1616
+    },
+    {
+      "cosine_similarity": 0.07730827502023507,
+      "epoch": 1.5069897483690586,
+      "grad_norm": 1.0830756988415184,
+      "learning_rate": 1.1059716948567485e-05,
+      "loss": 1.4556,
+      "step": 1617
+    },
+    {
+      "cosine_similarity": 0.27120272408121676,
+      "epoch": 1.5079217148182664,
+      "grad_norm": 1.5145114446871637,
+      "learning_rate": 1.1052813255091474e-05,
+      "loss": 1.7101,
+      "step": 1618
+    },
+    {
+      "cosine_similarity": 0.08162766911830244,
+      "epoch": 1.5088536812674742,
+      "grad_norm": 1.0369242040645028,
+      "learning_rate": 1.1045909561615465e-05,
+      "loss": 1.8108,
+      "step": 1619
+    },
+    {
+      "cosine_similarity": -0.15782448943481578,
+      "epoch": 1.5097856477166822,
+      "grad_norm": 1.1361799210130796,
+      "learning_rate": 1.1039005868139456e-05,
+      "loss": 1.8781,
+      "step": 1620
+    },
+    {
+      "cosine_similarity": 0.010055257688623247,
+      "epoch": 1.51071761416589,
+      "grad_norm": 1.0643407172485253,
+      "learning_rate": 1.1032102174663446e-05,
+      "loss": 1.8708,
+      "step": 1621
+    },
+    {
+      "cosine_similarity": 0.41080830897540405,
+      "epoch": 1.511649580615098,
+      "grad_norm": 1.1162890694686025,
+      "learning_rate": 1.1025198481187435e-05,
+      "loss": 1.7067,
+      "step": 1622
+    },
+    {
+      "cosine_similarity": 0.5493286574391498,
+      "epoch": 1.5125815470643058,
+      "grad_norm": 1.3928010084336102,
+      "learning_rate": 1.1018294787711426e-05,
+      "loss": 1.8241,
+      "step": 1623
+    },
+    {
+      "cosine_similarity": 0.20729122761418028,
+      "epoch": 1.5135135135135136,
+      "grad_norm": 1.293278512766048,
+      "learning_rate": 1.1011391094235417e-05,
+      "loss": 1.4952,
+      "step": 1624
+    },
+    {
+      "cosine_similarity": 0.15064432294993843,
+      "epoch": 1.5144454799627214,
+      "grad_norm": 1.3773929651536834,
+      "learning_rate": 1.1004487400759408e-05,
+      "loss": 1.5829,
+      "step": 1625
+    },
+    {
+      "cosine_similarity": 0.45246456700887117,
+      "epoch": 1.5153774464119292,
+      "grad_norm": 1.3524937184117463,
+      "learning_rate": 1.0997583707283397e-05,
+      "loss": 1.7217,
+      "step": 1626
+    },
+    {
+      "cosine_similarity": 0.5018581165343886,
+      "epoch": 1.516309412861137,
+      "grad_norm": 1.0267854348162087,
+      "learning_rate": 1.0990680013807389e-05,
+      "loss": 1.4,
+      "step": 1627
+    },
+    {
+      "cosine_similarity": 0.11537615862509035,
+      "epoch": 1.5172413793103448,
+      "grad_norm": 1.0723088774286276,
+      "learning_rate": 1.0983776320331378e-05,
+      "loss": 1.5342,
+      "step": 1628
+    },
+    {
+      "cosine_similarity": 0.3294780059113493,
+      "epoch": 1.5181733457595525,
+      "grad_norm": 1.1240660724577671,
+      "learning_rate": 1.0976872626855369e-05,
+      "loss": 1.6319,
+      "step": 1629
+    },
+    {
+      "cosine_similarity": 0.06187974184317936,
+      "epoch": 1.5191053122087603,
+      "grad_norm": 1.3121383853570565,
+      "learning_rate": 1.0969968933379358e-05,
+      "loss": 1.6762,
+      "step": 1630
+    },
+    {
+      "cosine_similarity": 0.583868447044204,
+      "epoch": 1.5200372786579683,
+      "grad_norm": 1.1534693771589806,
+      "learning_rate": 1.096306523990335e-05,
+      "loss": 2.8051,
+      "step": 1631
+    },
+    {
+      "cosine_similarity": 0.8396547280091321,
+      "epoch": 1.5209692451071761,
+      "grad_norm": 1.226503320375056,
+      "learning_rate": 1.095616154642734e-05,
+      "loss": 1.7566,
+      "step": 1632
+    },
+    {
+      "cosine_similarity": -0.09148641960466578,
+      "epoch": 1.521901211556384,
+      "grad_norm": 1.2528867256231386,
+      "learning_rate": 1.094925785295133e-05,
+      "loss": 1.8054,
+      "step": 1633
+    },
+    {
+      "cosine_similarity": 0.7065648680954368,
+      "epoch": 1.522833178005592,
+      "grad_norm": 1.3258647809364366,
+      "learning_rate": 1.094235415947532e-05,
+      "loss": 1.658,
+      "step": 1634
+    },
+    {
+      "cosine_similarity": 0.9243662889322254,
+      "epoch": 1.5237651444547997,
+      "grad_norm": 1.0440519315429113,
+      "learning_rate": 1.0935450465999312e-05,
+      "loss": 1.7121,
+      "step": 1635
+    },
+    {
+      "cosine_similarity": 0.4626480345640239,
+      "epoch": 1.5246971109040075,
+      "grad_norm": 1.134551898208798,
+      "learning_rate": 1.09285467725233e-05,
+      "loss": 1.6233,
+      "step": 1636
+    },
+    {
+      "cosine_similarity": 0.13887765157654053,
+      "epoch": 1.5256290773532153,
+      "grad_norm": 1.2296393753732853,
+      "learning_rate": 1.092164307904729e-05,
+      "loss": 1.6392,
+      "step": 1637
+    },
+    {
+      "cosine_similarity": 0.11222317023230492,
+      "epoch": 1.526561043802423,
+      "grad_norm": 1.151990755163757,
+      "learning_rate": 1.0914739385571282e-05,
+      "loss": 1.4589,
+      "step": 1638
+    },
+    {
+      "cosine_similarity": 0.014941320529103849,
+      "epoch": 1.5274930102516309,
+      "grad_norm": 1.364251960179271,
+      "learning_rate": 1.0907835692095273e-05,
+      "loss": 1.8775,
+      "step": 1639
+    },
+    {
+      "cosine_similarity": 0.09595939479501854,
+      "epoch": 1.5284249767008387,
+      "grad_norm": 1.2398866351941806,
+      "learning_rate": 1.0900931998619262e-05,
+      "loss": 1.6345,
+      "step": 1640
+    },
+    {
+      "cosine_similarity": 0.10839211028184098,
+      "epoch": 1.5293569431500464,
+      "grad_norm": 1.2609407205346579,
+      "learning_rate": 1.0894028305143251e-05,
+      "loss": 1.8201,
+      "step": 1641
+    },
+    {
+      "cosine_similarity": -0.07705718649492628,
+      "epoch": 1.5302889095992545,
+      "grad_norm": 1.0387489796048253,
+      "learning_rate": 1.0887124611667243e-05,
+      "loss": 1.7734,
+      "step": 1642
+    },
+    {
+      "cosine_similarity": 0.2940382737873345,
+      "epoch": 1.5312208760484622,
+      "grad_norm": 1.2037977445485293,
+      "learning_rate": 1.0880220918191234e-05,
+      "loss": 1.7879,
+      "step": 1643
+    },
+    {
+      "cosine_similarity": 0.27438238291828265,
+      "epoch": 1.53215284249767,
+      "grad_norm": 1.2552962297204484,
+      "learning_rate": 1.0873317224715223e-05,
+      "loss": 1.7764,
+      "step": 1644
+    },
+    {
+      "cosine_similarity": 0.13388516880548348,
+      "epoch": 1.533084808946878,
+      "grad_norm": 1.1155554624866921,
+      "learning_rate": 1.0866413531239216e-05,
+      "loss": 1.7952,
+      "step": 1645
+    },
+    {
+      "cosine_similarity": 0.3979838644914757,
+      "epoch": 1.5340167753960858,
+      "grad_norm": 1.0845490675156026,
+      "learning_rate": 1.0859509837763205e-05,
+      "loss": 1.6502,
+      "step": 1646
+    },
+    {
+      "cosine_similarity": 0.23813361053884785,
+      "epoch": 1.5349487418452936,
+      "grad_norm": 1.3175431836123397,
+      "learning_rate": 1.0852606144287194e-05,
+      "loss": 1.8509,
+      "step": 1647
+    },
+    {
+      "cosine_similarity": 0.057763454849218,
+      "epoch": 1.5358807082945014,
+      "grad_norm": 1.1980783559688404,
+      "learning_rate": 1.0845702450811184e-05,
+      "loss": 1.9087,
+      "step": 1648
+    },
+    {
+      "cosine_similarity": 0.01707753324342289,
+      "epoch": 1.5368126747437092,
+      "grad_norm": 1.1696160692811175,
+      "learning_rate": 1.0838798757335177e-05,
+      "loss": 1.8401,
+      "step": 1649
+    },
+    {
+      "cosine_similarity": 0.12304627100472641,
+      "epoch": 1.537744641192917,
+      "grad_norm": 1.150541285502057,
+      "learning_rate": 1.0831895063859166e-05,
+      "loss": 1.5202,
+      "step": 1650
+    },
+    {
+      "cosine_similarity": 0.5043949710148455,
+      "epoch": 1.5386766076421248,
+      "grad_norm": 1.4453690083577075,
+      "learning_rate": 1.0824991370383155e-05,
+      "loss": 1.8987,
+      "step": 1651
+    },
+    {
+      "cosine_similarity": 0.12177694541964283,
+      "epoch": 1.5396085740913326,
+      "grad_norm": 1.2317110215705744,
+      "learning_rate": 1.0818087676907146e-05,
+      "loss": 1.9277,
+      "step": 1652
+    },
+    {
+      "cosine_similarity": 0.17703344626445358,
+      "epoch": 1.5405405405405406,
+      "grad_norm": 1.3145518834222232,
+      "learning_rate": 1.0811183983431136e-05,
+      "loss": 1.8701,
+      "step": 1653
+    },
+    {
+      "cosine_similarity": 0.11112851446653911,
+      "epoch": 1.5414725069897484,
+      "grad_norm": 1.3318953445624493,
+      "learning_rate": 1.0804280289955127e-05,
+      "loss": 1.6549,
+      "step": 1654
+    },
+    {
+      "cosine_similarity": 0.05921254371600335,
+      "epoch": 1.5424044734389561,
+      "grad_norm": 2.54403625654924,
+      "learning_rate": 1.0797376596479116e-05,
+      "loss": 1.7867,
+      "step": 1655
+    },
+    {
+      "cosine_similarity": 0.04574426488507089,
+      "epoch": 1.5433364398881642,
+      "grad_norm": 1.3063570807620881,
+      "learning_rate": 1.0790472903003109e-05,
+      "loss": 1.6538,
+      "step": 1656
+    },
+    {
+      "cosine_similarity": 0.032647437920194505,
+      "epoch": 1.544268406337372,
+      "grad_norm": 1.3832016876774909,
+      "learning_rate": 1.0783569209527098e-05,
+      "loss": 2.1104,
+      "step": 1657
+    },
+    {
+      "cosine_similarity": 0.03566816691070663,
+      "epoch": 1.5452003727865797,
+      "grad_norm": 1.412070906328608,
+      "learning_rate": 1.0776665516051088e-05,
+      "loss": 2.0278,
+      "step": 1658
+    },
+    {
+      "cosine_similarity": 0.20966018339956163,
+      "epoch": 1.5461323392357875,
+      "grad_norm": 1.1488840080454041,
+      "learning_rate": 1.0769761822575077e-05,
+      "loss": 1.6276,
+      "step": 1659
+    },
+    {
+      "cosine_similarity": 0.1606305208231818,
+      "epoch": 1.5470643056849953,
+      "grad_norm": 1.1285720865099171,
+      "learning_rate": 1.076285812909907e-05,
+      "loss": 1.633,
+      "step": 1660
+    },
+    {
+      "cosine_similarity": 0.15168500444225175,
+      "epoch": 1.547996272134203,
+      "grad_norm": 1.3134039860611515,
+      "learning_rate": 1.0755954435623059e-05,
+      "loss": 1.8262,
+      "step": 1661
+    },
+    {
+      "cosine_similarity": 0.15902860318718276,
+      "epoch": 1.5489282385834109,
+      "grad_norm": 1.4583078112775343,
+      "learning_rate": 1.074905074214705e-05,
+      "loss": 1.6934,
+      "step": 1662
+    },
+    {
+      "cosine_similarity": 0.19675713974816283,
+      "epoch": 1.5498602050326187,
+      "grad_norm": 1.1879928157520698,
+      "learning_rate": 1.0742147048671039e-05,
+      "loss": 1.8024,
+      "step": 1663
+    },
+    {
+      "cosine_similarity": 0.06745227740275064,
+      "epoch": 1.5507921714818267,
+      "grad_norm": 1.303205303194602,
+      "learning_rate": 1.0735243355195031e-05,
+      "loss": 1.7942,
+      "step": 1664
+    },
+    {
+      "cosine_similarity": -0.03655231882117249,
+      "epoch": 1.5517241379310345,
+      "grad_norm": 1.2037539138787243,
+      "learning_rate": 1.072833966171902e-05,
+      "loss": 1.7481,
+      "step": 1665
+    },
+    {
+      "cosine_similarity": 0.060029063053202734,
+      "epoch": 1.5526561043802423,
+      "grad_norm": 1.1051894289203454,
+      "learning_rate": 1.072143596824301e-05,
+      "loss": 1.6281,
+      "step": 1666
+    },
+    {
+      "cosine_similarity": 0.26014093509808894,
+      "epoch": 1.5535880708294503,
+      "grad_norm": 1.1821146361985235,
+      "learning_rate": 1.0714532274767002e-05,
+      "loss": 1.777,
+      "step": 1667
+    },
+    {
+      "cosine_similarity": 0.06096183145819955,
+      "epoch": 1.554520037278658,
+      "grad_norm": 1.285742666752431,
+      "learning_rate": 1.0707628581290992e-05,
+      "loss": 1.5762,
+      "step": 1668
+    },
+    {
+      "cosine_similarity": 0.18077472608662862,
+      "epoch": 1.5554520037278659,
+      "grad_norm": 0.9600412255465472,
+      "learning_rate": 1.0700724887814981e-05,
+      "loss": 1.6336,
+      "step": 1669
+    },
+    {
+      "cosine_similarity": 0.3057543851066333,
+      "epoch": 1.5563839701770736,
+      "grad_norm": 1.2928927947975668,
+      "learning_rate": 1.0693821194338972e-05,
+      "loss": 1.5726,
+      "step": 1670
+    },
+    {
+      "cosine_similarity": 0.23347031661812076,
+      "epoch": 1.5573159366262814,
+      "grad_norm": 1.1423655331165752,
+      "learning_rate": 1.0686917500862963e-05,
+      "loss": 1.7458,
+      "step": 1671
+    },
+    {
+      "cosine_similarity": -0.11344538184545336,
+      "epoch": 1.5582479030754892,
+      "grad_norm": 1.289716466944667,
+      "learning_rate": 1.0680013807386953e-05,
+      "loss": 1.8835,
+      "step": 1672
+    },
+    {
+      "cosine_similarity": 0.3209668431833733,
+      "epoch": 1.559179869524697,
+      "grad_norm": 1.0778581592288414,
+      "learning_rate": 1.0673110113910943e-05,
+      "loss": 1.496,
+      "step": 1673
+    },
+    {
+      "cosine_similarity": 0.09157588461426236,
+      "epoch": 1.5601118359739048,
+      "grad_norm": 1.22393111343103,
+      "learning_rate": 1.0666206420434933e-05,
+      "loss": 1.5585,
+      "step": 1674
+    },
+    {
+      "cosine_similarity": -0.0830344599306271,
+      "epoch": 1.5610438024231128,
+      "grad_norm": 1.4433637355549618,
+      "learning_rate": 1.0659302726958924e-05,
+      "loss": 1.7821,
+      "step": 1675
+    },
+    {
+      "cosine_similarity": 0.10313950997304852,
+      "epoch": 1.5619757688723206,
+      "grad_norm": 1.369229800655836,
+      "learning_rate": 1.0652399033482915e-05,
+      "loss": 1.7446,
+      "step": 1676
+    },
+    {
+      "cosine_similarity": 0.07180080297964603,
+      "epoch": 1.5629077353215284,
+      "grad_norm": 1.3070627177309304,
+      "learning_rate": 1.0645495340006904e-05,
+      "loss": 1.4949,
+      "step": 1677
+    },
+    {
+      "cosine_similarity": -0.23550182066784073,
+      "epoch": 1.5638397017707364,
+      "grad_norm": 1.0467315962444017,
+      "learning_rate": 1.0638591646530896e-05,
+      "loss": 1.5306,
+      "step": 1678
+    },
+    {
+      "cosine_similarity": 0.5695778087213639,
+      "epoch": 1.5647716682199442,
+      "grad_norm": 1.1917858401473844,
+      "learning_rate": 1.0631687953054885e-05,
+      "loss": 1.5943,
+      "step": 1679
+    },
+    {
+      "cosine_similarity": 0.29365178519020646,
+      "epoch": 1.565703634669152,
+      "grad_norm": 1.149227153560679,
+      "learning_rate": 1.0624784259578876e-05,
+      "loss": 2.0206,
+      "step": 1680
+    },
+    {
+      "cosine_similarity": 0.21227096341486348,
+      "epoch": 1.5666356011183598,
+      "grad_norm": 0.9860735085389059,
+      "learning_rate": 1.0617880566102865e-05,
+      "loss": 1.7283,
+      "step": 1681
+    },
+    {
+      "cosine_similarity": 0.05071271248560576,
+      "epoch": 1.5675675675675675,
+      "grad_norm": 1.0932266996963977,
+      "learning_rate": 1.0610976872626857e-05,
+      "loss": 1.5426,
+      "step": 1682
+    },
+    {
+      "cosine_similarity": 0.5808109529517542,
+      "epoch": 1.5684995340167753,
+      "grad_norm": 1.1158825369863292,
+      "learning_rate": 1.0604073179150846e-05,
+      "loss": 1.8507,
+      "step": 1683
+    },
+    {
+      "cosine_similarity": 0.48188873529715587,
+      "epoch": 1.5694315004659831,
+      "grad_norm": 1.0886942641931114,
+      "learning_rate": 1.0597169485674837e-05,
+      "loss": 2.0198,
+      "step": 1684
+    },
+    {
+      "cosine_similarity": 0.14977062844684377,
+      "epoch": 1.570363466915191,
+      "grad_norm": 1.1865770551735209,
+      "learning_rate": 1.0590265792198828e-05,
+      "loss": 2.1478,
+      "step": 1685
+    },
+    {
+      "cosine_similarity": 0.045341445412619544,
+      "epoch": 1.571295433364399,
+      "grad_norm": 1.1208084829450862,
+      "learning_rate": 1.0583362098722819e-05,
+      "loss": 1.7218,
+      "step": 1686
+    },
+    {
+      "cosine_similarity": 0.2580508953061591,
+      "epoch": 1.5722273998136067,
+      "grad_norm": 1.2504247070869061,
+      "learning_rate": 1.0576458405246808e-05,
+      "loss": 1.9269,
+      "step": 1687
+    },
+    {
+      "cosine_similarity": 0.09631442503104545,
+      "epoch": 1.5731593662628145,
+      "grad_norm": 1.1366967082565096,
+      "learning_rate": 1.0569554711770797e-05,
+      "loss": 1.657,
+      "step": 1688
+    },
+    {
+      "cosine_similarity": 0.17061911856359732,
+      "epoch": 1.5740913327120225,
+      "grad_norm": 1.4903960075613862,
+      "learning_rate": 1.056265101829479e-05,
+      "loss": 1.8251,
+      "step": 1689
+    },
+    {
+      "cosine_similarity": 0.0684367935177907,
+      "epoch": 1.5750232991612303,
+      "grad_norm": 1.0694150158039262,
+      "learning_rate": 1.055574732481878e-05,
+      "loss": 1.6955,
+      "step": 1690
+    },
+    {
+      "cosine_similarity": 0.403640662494887,
+      "epoch": 1.575955265610438,
+      "grad_norm": 1.3706471022493105,
+      "learning_rate": 1.0548843631342769e-05,
+      "loss": 1.9829,
+      "step": 1691
+    },
+    {
+      "cosine_similarity": 0.07283105327196593,
+      "epoch": 1.5768872320596459,
+      "grad_norm": 1.108265725209936,
+      "learning_rate": 1.0541939937866758e-05,
+      "loss": 1.3458,
+      "step": 1692
+    },
+    {
+      "cosine_similarity": 0.19367689755205877,
+      "epoch": 1.5778191985088537,
+      "grad_norm": 1.0393439970114908,
+      "learning_rate": 1.053503624439075e-05,
+      "loss": 1.9086,
+      "step": 1693
+    },
+    {
+      "cosine_similarity": 0.0313750642264905,
+      "epoch": 1.5787511649580614,
+      "grad_norm": 1.0348591664876614,
+      "learning_rate": 1.0528132550914741e-05,
+      "loss": 1.6938,
+      "step": 1694
+    },
+    {
+      "cosine_similarity": 0.291094382714992,
+      "epoch": 1.5796831314072692,
+      "grad_norm": 1.4345265722396028,
+      "learning_rate": 1.052122885743873e-05,
+      "loss": 1.7309,
+      "step": 1695
+    },
+    {
+      "cosine_similarity": 0.333342879021879,
+      "epoch": 1.580615097856477,
+      "grad_norm": 1.3081227695785484,
+      "learning_rate": 1.0514325163962723e-05,
+      "loss": 1.8254,
+      "step": 1696
+    },
+    {
+      "cosine_similarity": 0.0064931929150656735,
+      "epoch": 1.581547064305685,
+      "grad_norm": 1.0729118854907025,
+      "learning_rate": 1.0507421470486712e-05,
+      "loss": 1.5967,
+      "step": 1697
+    },
+    {
+      "cosine_similarity": 0.19887751103063278,
+      "epoch": 1.5824790307548928,
+      "grad_norm": 1.164613328954487,
+      "learning_rate": 1.05005177770107e-05,
+      "loss": 1.7833,
+      "step": 1698
+    },
+    {
+      "cosine_similarity": 0.03292166955158032,
+      "epoch": 1.5834109972041006,
+      "grad_norm": 1.3687769543651225,
+      "learning_rate": 1.0493614083534691e-05,
+      "loss": 1.8204,
+      "step": 1699
+    },
+    {
+      "cosine_similarity": 0.05411965466468989,
+      "epoch": 1.5843429636533086,
+      "grad_norm": 1.1230218372925338,
+      "learning_rate": 1.0486710390058684e-05,
+      "loss": 1.8594,
+      "step": 1700
+    },
+    {
+      "cosine_similarity": -0.1087264550020906,
+      "epoch": 1.5852749301025164,
+      "grad_norm": 1.0205659571813226,
+      "learning_rate": 1.0479806696582673e-05,
+      "loss": 1.5532,
+      "step": 1701
+    },
+    {
+      "cosine_similarity": -0.0023713584045163562,
+      "epoch": 1.5862068965517242,
+      "grad_norm": 1.3806094757061815,
+      "learning_rate": 1.0472903003106662e-05,
+      "loss": 1.5819,
+      "step": 1702
+    },
+    {
+      "cosine_similarity": 0.059944495957127655,
+      "epoch": 1.587138863000932,
+      "grad_norm": 1.1812070932702696,
+      "learning_rate": 1.0465999309630653e-05,
+      "loss": 1.6686,
+      "step": 1703
+    },
+    {
+      "cosine_similarity": 0.18648135992217205,
+      "epoch": 1.5880708294501398,
+      "grad_norm": 1.102113374761161,
+      "learning_rate": 1.0459095616154643e-05,
+      "loss": 1.7458,
+      "step": 1704
+    },
+    {
+      "cosine_similarity": 0.19657907727170265,
+      "epoch": 1.5890027958993476,
+      "grad_norm": 1.3741159687565676,
+      "learning_rate": 1.0452191922678634e-05,
+      "loss": 1.6876,
+      "step": 1705
+    },
+    {
+      "cosine_similarity": -0.20994567921346666,
+      "epoch": 1.5899347623485554,
+      "grad_norm": 1.1142862830356637,
+      "learning_rate": 1.0445288229202623e-05,
+      "loss": 1.5302,
+      "step": 1706
+    },
+    {
+      "cosine_similarity": 0.047604850410245834,
+      "epoch": 1.5908667287977631,
+      "grad_norm": 1.2387781525976707,
+      "learning_rate": 1.0438384535726616e-05,
+      "loss": 1.7999,
+      "step": 1707
+    },
+    {
+      "cosine_similarity": 0.24498929800800312,
+      "epoch": 1.5917986952469712,
+      "grad_norm": 1.5667456304423575,
+      "learning_rate": 1.0431480842250605e-05,
+      "loss": 1.8933,
+      "step": 1708
+    },
+    {
+      "cosine_similarity": -0.01725851117952045,
+      "epoch": 1.592730661696179,
+      "grad_norm": 1.5721041049154267,
+      "learning_rate": 1.0424577148774595e-05,
+      "loss": 1.7375,
+      "step": 1709
+    },
+    {
+      "cosine_similarity": 0.15464082682340116,
+      "epoch": 1.5936626281453867,
+      "grad_norm": 1.2776326206404405,
+      "learning_rate": 1.0417673455298584e-05,
+      "loss": 1.882,
+      "step": 1710
+    },
+    {
+      "cosine_similarity": 0.35755908068115966,
+      "epoch": 1.5945945945945947,
+      "grad_norm": 1.1610361392899493,
+      "learning_rate": 1.0410769761822577e-05,
+      "loss": 1.5924,
+      "step": 1711
+    },
+    {
+      "cosine_similarity": 0.4566471965685942,
+      "epoch": 1.5955265610438025,
+      "grad_norm": 1.0645954756089393,
+      "learning_rate": 1.0403866068346566e-05,
+      "loss": 1.6577,
+      "step": 1712
+    },
+    {
+      "cosine_similarity": 0.4540180485549546,
+      "epoch": 1.5964585274930103,
+      "grad_norm": 1.2431491614613523,
+      "learning_rate": 1.0396962374870557e-05,
+      "loss": 1.5973,
+      "step": 1713
+    },
+    {
+      "cosine_similarity": -0.38697723306044063,
+      "epoch": 1.597390493942218,
+      "grad_norm": 1.4907798932142697,
+      "learning_rate": 1.0390058681394546e-05,
+      "loss": 1.8612,
+      "step": 1714
+    },
+    {
+      "cosine_similarity": 0.2187445246999306,
+      "epoch": 1.598322460391426,
+      "grad_norm": 1.1279776919728275,
+      "learning_rate": 1.0383154987918538e-05,
+      "loss": 1.8183,
+      "step": 1715
+    },
+    {
+      "cosine_similarity": 0.3037076365926407,
+      "epoch": 1.5992544268406337,
+      "grad_norm": 1.0672610625200396,
+      "learning_rate": 1.0376251294442527e-05,
+      "loss": 1.5736,
+      "step": 1716
+    },
+    {
+      "cosine_similarity": 0.3532760169302846,
+      "epoch": 1.6001863932898415,
+      "grad_norm": 1.0717612137877661,
+      "learning_rate": 1.0369347600966518e-05,
+      "loss": 1.7916,
+      "step": 1717
+    },
+    {
+      "cosine_similarity": 0.11046130142559296,
+      "epoch": 1.6011183597390493,
+      "grad_norm": 1.2717531974023413,
+      "learning_rate": 1.0362443907490509e-05,
+      "loss": 1.8479,
+      "step": 1718
+    },
+    {
+      "cosine_similarity": 0.193468552713398,
+      "epoch": 1.6020503261882573,
+      "grad_norm": 1.1392525803803275,
+      "learning_rate": 1.03555402140145e-05,
+      "loss": 1.8615,
+      "step": 1719
+    },
+    {
+      "cosine_similarity": -0.04418235970125103,
+      "epoch": 1.602982292637465,
+      "grad_norm": 1.2810489029568677,
+      "learning_rate": 1.0348636520538488e-05,
+      "loss": 1.7655,
+      "step": 1720
+    },
+    {
+      "cosine_similarity": 0.33004283252707384,
+      "epoch": 1.6039142590866728,
+      "grad_norm": 1.0625404621219348,
+      "learning_rate": 1.0341732827062479e-05,
+      "loss": 1.9351,
+      "step": 1721
+    },
+    {
+      "cosine_similarity": -0.0021977673295014723,
+      "epoch": 1.6048462255358809,
+      "grad_norm": 0.955629209639426,
+      "learning_rate": 1.033482913358647e-05,
+      "loss": 1.3537,
+      "step": 1722
+    },
+    {
+      "cosine_similarity": 0.04749310553682023,
+      "epoch": 1.6057781919850886,
+      "grad_norm": 1.0884801698331326,
+      "learning_rate": 1.032792544011046e-05,
+      "loss": 1.4393,
+      "step": 1723
+    },
+    {
+      "cosine_similarity": 0.17145471460862596,
+      "epoch": 1.6067101584342964,
+      "grad_norm": 1.209853914654163,
+      "learning_rate": 1.032102174663445e-05,
+      "loss": 1.9104,
+      "step": 1724
+    },
+    {
+      "cosine_similarity": -0.12212931159164382,
+      "epoch": 1.6076421248835042,
+      "grad_norm": 1.166901338166021,
+      "learning_rate": 1.031411805315844e-05,
+      "loss": 1.8276,
+      "step": 1725
+    },
+    {
+      "cosine_similarity": 0.2883773433583493,
+      "epoch": 1.608574091332712,
+      "grad_norm": 1.2242558415397697,
+      "learning_rate": 1.0307214359682431e-05,
+      "loss": 1.7056,
+      "step": 1726
+    },
+    {
+      "cosine_similarity": 0.03355109685651949,
+      "epoch": 1.6095060577819198,
+      "grad_norm": 1.108046794623555,
+      "learning_rate": 1.0300310666206422e-05,
+      "loss": 1.7544,
+      "step": 1727
+    },
+    {
+      "cosine_similarity": -0.21104566760034926,
+      "epoch": 1.6104380242311276,
+      "grad_norm": 1.484754123981437,
+      "learning_rate": 1.029340697273041e-05,
+      "loss": 1.724,
+      "step": 1728
+    },
+    {
+      "cosine_similarity": -0.054085952705755856,
+      "epoch": 1.6113699906803354,
+      "grad_norm": 1.1238639938937551,
+      "learning_rate": 1.0286503279254403e-05,
+      "loss": 1.532,
+      "step": 1729
+    },
+    {
+      "cosine_similarity": 0.10230554331723375,
+      "epoch": 1.6123019571295434,
+      "grad_norm": 1.1025017386941434,
+      "learning_rate": 1.0279599585778392e-05,
+      "loss": 1.7596,
+      "step": 1730
+    },
+    {
+      "cosine_similarity": -0.16564740144423074,
+      "epoch": 1.6132339235787512,
+      "grad_norm": 1.172117366012158,
+      "learning_rate": 1.0272695892302383e-05,
+      "loss": 1.7667,
+      "step": 1731
+    },
+    {
+      "cosine_similarity": 0.5793550406915497,
+      "epoch": 1.614165890027959,
+      "grad_norm": 1.1210625173334223,
+      "learning_rate": 1.0265792198826372e-05,
+      "loss": 1.7264,
+      "step": 1732
+    },
+    {
+      "cosine_similarity": 0.3650320951900019,
+      "epoch": 1.615097856477167,
+      "grad_norm": 1.115925662669345,
+      "learning_rate": 1.0258888505350364e-05,
+      "loss": 1.5991,
+      "step": 1733
+    },
+    {
+      "cosine_similarity": 0.12470335844336469,
+      "epoch": 1.6160298229263748,
+      "grad_norm": 0.9356113156717938,
+      "learning_rate": 1.0251984811874353e-05,
+      "loss": 1.399,
+      "step": 1734
+    },
+    {
+      "cosine_similarity": 0.16095196912539414,
+      "epoch": 1.6169617893755825,
+      "grad_norm": 1.3296845725576438,
+      "learning_rate": 1.0245081118398344e-05,
+      "loss": 1.708,
+      "step": 1735
+    },
+    {
+      "cosine_similarity": -0.07511950067476982,
+      "epoch": 1.6178937558247903,
+      "grad_norm": 1.307655241973434,
+      "learning_rate": 1.0238177424922335e-05,
+      "loss": 1.738,
+      "step": 1736
+    },
+    {
+      "cosine_similarity": 0.404676808771281,
+      "epoch": 1.6188257222739981,
+      "grad_norm": 1.3083520718989554,
+      "learning_rate": 1.0231273731446326e-05,
+      "loss": 1.6146,
+      "step": 1737
+    },
+    {
+      "cosine_similarity": 0.04436225974778861,
+      "epoch": 1.619757688723206,
+      "grad_norm": 0.9488680398237176,
+      "learning_rate": 1.0224370037970315e-05,
+      "loss": 1.5411,
+      "step": 1738
+    },
+    {
+      "cosine_similarity": -0.04009473828587093,
+      "epoch": 1.6206896551724137,
+      "grad_norm": 1.1317938998404509,
+      "learning_rate": 1.0217466344494304e-05,
+      "loss": 1.5085,
+      "step": 1739
+    },
+    {
+      "cosine_similarity": 0.11232295322704,
+      "epoch": 1.6216216216216215,
+      "grad_norm": 1.2934191283132288,
+      "learning_rate": 1.0210562651018296e-05,
+      "loss": 1.5297,
+      "step": 1740
+    },
+    {
+      "cosine_similarity": 0.16861663404299423,
+      "epoch": 1.6225535880708295,
+      "grad_norm": 1.1520492305873087,
+      "learning_rate": 1.0203658957542287e-05,
+      "loss": 1.7633,
+      "step": 1741
+    },
+    {
+      "cosine_similarity": 0.14420061432821396,
+      "epoch": 1.6234855545200373,
+      "grad_norm": 1.358256208263621,
+      "learning_rate": 1.0196755264066276e-05,
+      "loss": 1.8731,
+      "step": 1742
+    },
+    {
+      "cosine_similarity": 0.2055993025411818,
+      "epoch": 1.624417520969245,
+      "grad_norm": 1.2304033107616463,
+      "learning_rate": 1.0189851570590265e-05,
+      "loss": 1.847,
+      "step": 1743
+    },
+    {
+      "cosine_similarity": 0.1419486442099629,
+      "epoch": 1.625349487418453,
+      "grad_norm": 1.2476551234891253,
+      "learning_rate": 1.0182947877114257e-05,
+      "loss": 1.7464,
+      "step": 1744
+    },
+    {
+      "cosine_similarity": 0.07789818382296612,
+      "epoch": 1.6262814538676609,
+      "grad_norm": 1.2973647493436835,
+      "learning_rate": 1.0176044183638248e-05,
+      "loss": 1.758,
+      "step": 1745
+    },
+    {
+      "cosine_similarity": -0.21659948646903857,
+      "epoch": 1.6272134203168687,
+      "grad_norm": 1.23992066266897,
+      "learning_rate": 1.0169140490162237e-05,
+      "loss": 1.7192,
+      "step": 1746
+    },
+    {
+      "cosine_similarity": 0.008837562198504121,
+      "epoch": 1.6281453867660765,
+      "grad_norm": 1.540035080508733,
+      "learning_rate": 1.016223679668623e-05,
+      "loss": 2.0073,
+      "step": 1747
+    },
+    {
+      "cosine_similarity": 0.006856065753621176,
+      "epoch": 1.6290773532152842,
+      "grad_norm": 1.1583007470331235,
+      "learning_rate": 1.0155333103210219e-05,
+      "loss": 1.4963,
+      "step": 1748
+    },
+    {
+      "cosine_similarity": 0.2924760608382533,
+      "epoch": 1.630009319664492,
+      "grad_norm": 1.3238109845442267,
+      "learning_rate": 1.0148429409734208e-05,
+      "loss": 2.1663,
+      "step": 1749
+    },
+    {
+      "cosine_similarity": 0.0031829333698723226,
+      "epoch": 1.6309412861136998,
+      "grad_norm": 1.1921622426164429,
+      "learning_rate": 1.0141525716258198e-05,
+      "loss": 1.7121,
+      "step": 1750
+    },
+    {
+      "cosine_similarity": 0.011664194830225834,
+      "epoch": 1.6318732525629076,
+      "grad_norm": 1.2249528397131004,
+      "learning_rate": 1.0134622022782191e-05,
+      "loss": 1.9524,
+      "step": 1751
+    },
+    {
+      "cosine_similarity": -0.05206065185977983,
+      "epoch": 1.6328052190121156,
+      "grad_norm": 1.1060930425257443,
+      "learning_rate": 1.012771832930618e-05,
+      "loss": 1.7302,
+      "step": 1752
+    },
+    {
+      "cosine_similarity": -0.07522396214409958,
+      "epoch": 1.6337371854613234,
+      "grad_norm": 1.0395377736670575,
+      "learning_rate": 1.0120814635830169e-05,
+      "loss": 1.5517,
+      "step": 1753
+    },
+    {
+      "cosine_similarity": 0.25875430087415924,
+      "epoch": 1.6346691519105312,
+      "grad_norm": 1.169382547991344,
+      "learning_rate": 1.011391094235416e-05,
+      "loss": 1.6755,
+      "step": 1754
+    },
+    {
+      "cosine_similarity": 0.11189424555645879,
+      "epoch": 1.6356011183597392,
+      "grad_norm": 1.1426728487578277,
+      "learning_rate": 1.010700724887815e-05,
+      "loss": 1.5053,
+      "step": 1755
+    },
+    {
+      "cosine_similarity": 0.14343175573821834,
+      "epoch": 1.636533084808947,
+      "grad_norm": 1.2244931367401641,
+      "learning_rate": 1.0100103555402141e-05,
+      "loss": 1.4026,
+      "step": 1756
+    },
+    {
+      "cosine_similarity": 0.08879933030893859,
+      "epoch": 1.6374650512581548,
+      "grad_norm": 1.1925309307663432,
+      "learning_rate": 1.009319986192613e-05,
+      "loss": 1.7271,
+      "step": 1757
+    },
+    {
+      "cosine_similarity": 0.15827381038558405,
+      "epoch": 1.6383970177073626,
+      "grad_norm": 1.0961557440754899,
+      "learning_rate": 1.0086296168450123e-05,
+      "loss": 1.6745,
+      "step": 1758
+    },
+    {
+      "cosine_similarity": 0.23619421504245136,
+      "epoch": 1.6393289841565704,
+      "grad_norm": 1.3460456936873801,
+      "learning_rate": 1.0079392474974112e-05,
+      "loss": 1.9907,
+      "step": 1759
+    },
+    {
+      "cosine_similarity": 0.13535639053458878,
+      "epoch": 1.6402609506057781,
+      "grad_norm": 1.178934478534969,
+      "learning_rate": 1.0072488781498102e-05,
+      "loss": 2.23,
+      "step": 1760
+    },
+    {
+      "cosine_similarity": 0.05009341354431523,
+      "epoch": 1.641192917054986,
+      "grad_norm": 1.5036809353993876,
+      "learning_rate": 1.0065585088022091e-05,
+      "loss": 1.5242,
+      "step": 1761
+    },
+    {
+      "cosine_similarity": 0.006282949664797946,
+      "epoch": 1.6421248835041937,
+      "grad_norm": 1.2018518469482358,
+      "learning_rate": 1.0058681394546084e-05,
+      "loss": 1.6945,
+      "step": 1762
+    },
+    {
+      "cosine_similarity": 0.11372230956713907,
+      "epoch": 1.6430568499534017,
+      "grad_norm": 1.148765209226536,
+      "learning_rate": 1.0051777701070073e-05,
+      "loss": 2.0087,
+      "step": 1763
+    },
+    {
+      "cosine_similarity": 0.040049309396376125,
+      "epoch": 1.6439888164026095,
+      "grad_norm": 1.1460325353031917,
+      "learning_rate": 1.0044874007594064e-05,
+      "loss": 1.6061,
+      "step": 1764
+    },
+    {
+      "cosine_similarity": 0.06494893644044078,
+      "epoch": 1.6449207828518173,
+      "grad_norm": 1.1232272693981444,
+      "learning_rate": 1.0037970314118053e-05,
+      "loss": 1.6341,
+      "step": 1765
+    },
+    {
+      "cosine_similarity": 0.22988068267337528,
+      "epoch": 1.6458527493010253,
+      "grad_norm": 1.326416107027173,
+      "learning_rate": 1.0031066620642045e-05,
+      "loss": 1.9691,
+      "step": 1766
+    },
+    {
+      "cosine_similarity": 0.2569510349759423,
+      "epoch": 1.646784715750233,
+      "grad_norm": 1.2458156680526924,
+      "learning_rate": 1.0024162927166034e-05,
+      "loss": 1.8402,
+      "step": 1767
+    },
+    {
+      "cosine_similarity": 0.2851090984911518,
+      "epoch": 1.647716682199441,
+      "grad_norm": 0.9493309676329716,
+      "learning_rate": 1.0017259233690025e-05,
+      "loss": 1.7973,
+      "step": 1768
+    },
+    {
+      "cosine_similarity": 0.11970767298966664,
+      "epoch": 1.6486486486486487,
+      "grad_norm": 1.3269660804184056,
+      "learning_rate": 1.0010355540214016e-05,
+      "loss": 1.7968,
+      "step": 1769
+    },
+    {
+      "cosine_similarity": 0.19751875005932054,
+      "epoch": 1.6495806150978565,
+      "grad_norm": 1.075703685201617,
+      "learning_rate": 1.0003451846738006e-05,
+      "loss": 1.7566,
+      "step": 1770
+    },
+    {
+      "cosine_similarity": 0.29852654324544897,
+      "epoch": 1.6505125815470643,
+      "grad_norm": 1.1139596861172822,
+      "learning_rate": 9.996548153261995e-06,
+      "loss": 1.7063,
+      "step": 1771
+    },
+    {
+      "cosine_similarity": -0.07215671036869517,
+      "epoch": 1.651444547996272,
+      "grad_norm": 1.1709335148098265,
+      "learning_rate": 9.989644459785986e-06,
+      "loss": 2.0346,
+      "step": 1772
+    },
+    {
+      "cosine_similarity": 0.07471877236559,
+      "epoch": 1.6523765144454798,
+      "grad_norm": 1.068560730318174,
+      "learning_rate": 9.982740766309977e-06,
+      "loss": 1.4819,
+      "step": 1773
+    },
+    {
+      "cosine_similarity": 0.28770542117931236,
+      "epoch": 1.6533084808946876,
+      "grad_norm": 1.115572504058935,
+      "learning_rate": 9.975837072833968e-06,
+      "loss": 1.6451,
+      "step": 1774
+    },
+    {
+      "cosine_similarity": 0.15446144530082725,
+      "epoch": 1.6542404473438956,
+      "grad_norm": 1.152335366389935,
+      "learning_rate": 9.968933379357957e-06,
+      "loss": 1.8949,
+      "step": 1775
+    },
+    {
+      "cosine_similarity": -0.03802068117637605,
+      "epoch": 1.6551724137931034,
+      "grad_norm": 1.2640501770739763,
+      "learning_rate": 9.962029685881947e-06,
+      "loss": 1.511,
+      "step": 1776
+    },
+    {
+      "cosine_similarity": 0.21124967440662462,
+      "epoch": 1.6561043802423114,
+      "grad_norm": 1.2319713512306198,
+      "learning_rate": 9.955125992405938e-06,
+      "loss": 1.5546,
+      "step": 1777
+    },
+    {
+      "cosine_similarity": 0.32086327068165477,
+      "epoch": 1.6570363466915192,
+      "grad_norm": 1.5855896119857333,
+      "learning_rate": 9.948222298929929e-06,
+      "loss": 1.7863,
+      "step": 1778
+    },
+    {
+      "cosine_similarity": 0.14542097159223563,
+      "epoch": 1.657968313140727,
+      "grad_norm": 1.302866862664669,
+      "learning_rate": 9.94131860545392e-06,
+      "loss": 1.7547,
+      "step": 1779
+    },
+    {
+      "cosine_similarity": 0.09947598263803957,
+      "epoch": 1.6589002795899348,
+      "grad_norm": 1.086549216236332,
+      "learning_rate": 9.934414911977909e-06,
+      "loss": 1.6963,
+      "step": 1780
+    },
+    {
+      "cosine_similarity": 0.059248341850770006,
+      "epoch": 1.6598322460391426,
+      "grad_norm": 0.9871610777033947,
+      "learning_rate": 9.9275112185019e-06,
+      "loss": 1.6076,
+      "step": 1781
+    },
+    {
+      "cosine_similarity": -0.014864432838954793,
+      "epoch": 1.6607642124883504,
+      "grad_norm": 1.1146721000267765,
+      "learning_rate": 9.92060752502589e-06,
+      "loss": 1.7867,
+      "step": 1782
+    },
+    {
+      "cosine_similarity": 0.22488591782062192,
+      "epoch": 1.6616961789375582,
+      "grad_norm": 1.3651847448936245,
+      "learning_rate": 9.91370383154988e-06,
+      "loss": 1.7695,
+      "step": 1783
+    },
+    {
+      "cosine_similarity": 0.16800727539145788,
+      "epoch": 1.662628145386766,
+      "grad_norm": 1.2015623395236465,
+      "learning_rate": 9.90680013807387e-06,
+      "loss": 1.5983,
+      "step": 1784
+    },
+    {
+      "cosine_similarity": 0.07894180485173952,
+      "epoch": 1.6635601118359737,
+      "grad_norm": 1.1454387387220466,
+      "learning_rate": 9.89989644459786e-06,
+      "loss": 2.0868,
+      "step": 1785
+    },
+    {
+      "cosine_similarity": 0.011345992685336969,
+      "epoch": 1.6644920782851818,
+      "grad_norm": 1.2747823333824475,
+      "learning_rate": 9.892992751121851e-06,
+      "loss": 2.2219,
+      "step": 1786
+    },
+    {
+      "cosine_similarity": 0.29172830529623667,
+      "epoch": 1.6654240447343895,
+      "grad_norm": 1.3567229513407617,
+      "learning_rate": 9.886089057645842e-06,
+      "loss": 1.6351,
+      "step": 1787
+    },
+    {
+      "cosine_similarity": 0.3082610165795005,
+      "epoch": 1.6663560111835976,
+      "grad_norm": 1.077329686358408,
+      "learning_rate": 9.879185364169833e-06,
+      "loss": 1.8455,
+      "step": 1788
+    },
+    {
+      "cosine_similarity": 0.33808263340814904,
+      "epoch": 1.6672879776328053,
+      "grad_norm": 1.065080745123489,
+      "learning_rate": 9.872281670693822e-06,
+      "loss": 1.9244,
+      "step": 1789
+    },
+    {
+      "cosine_similarity": 0.02932314412045086,
+      "epoch": 1.6682199440820131,
+      "grad_norm": 1.411433933010272,
+      "learning_rate": 9.865377977217812e-06,
+      "loss": 1.7174,
+      "step": 1790
+    },
+    {
+      "cosine_similarity": 0.09361502628861958,
+      "epoch": 1.669151910531221,
+      "grad_norm": 1.0664785866925777,
+      "learning_rate": 9.858474283741803e-06,
+      "loss": 1.6936,
+      "step": 1791
+    },
+    {
+      "cosine_similarity": 0.33875814016130257,
+      "epoch": 1.6700838769804287,
+      "grad_norm": 1.2476783965515554,
+      "learning_rate": 9.851570590265794e-06,
+      "loss": 1.7258,
+      "step": 1792
+    },
+    {
+      "cosine_similarity": -0.11082369853433256,
+      "epoch": 1.6710158434296365,
+      "grad_norm": 1.2956530826334276,
+      "learning_rate": 9.844666896789783e-06,
+      "loss": 2.1242,
+      "step": 1793
+    },
+    {
+      "cosine_similarity": 0.06036201557249193,
+      "epoch": 1.6719478098788443,
+      "grad_norm": 1.3557600412318418,
+      "learning_rate": 9.837763203313774e-06,
+      "loss": 1.8869,
+      "step": 1794
+    },
+    {
+      "cosine_similarity": -0.011561052424039382,
+      "epoch": 1.672879776328052,
+      "grad_norm": 1.3271785788756736,
+      "learning_rate": 9.830859509837763e-06,
+      "loss": 1.8717,
+      "step": 1795
+    },
+    {
+      "cosine_similarity": 0.6404358277116855,
+      "epoch": 1.6738117427772599,
+      "grad_norm": 1.2407877995054537,
+      "learning_rate": 9.823955816361753e-06,
+      "loss": 1.8747,
+      "step": 1796
+    },
+    {
+      "cosine_similarity": 0.22881779694330068,
+      "epoch": 1.6747437092264679,
+      "grad_norm": 1.1777364893321598,
+      "learning_rate": 9.817052122885746e-06,
+      "loss": 1.5349,
+      "step": 1797
+    },
+    {
+      "cosine_similarity": 0.407886652524228,
+      "epoch": 1.6756756756756757,
+      "grad_norm": 1.1594665831065338,
+      "learning_rate": 9.810148429409735e-06,
+      "loss": 1.8786,
+      "step": 1798
+    },
+    {
+      "cosine_similarity": 0.22942445518735824,
+      "epoch": 1.6766076421248837,
+      "grad_norm": 1.2152361408591403,
+      "learning_rate": 9.803244735933726e-06,
+      "loss": 1.5522,
+      "step": 1799
+    },
+    {
+      "cosine_similarity": 0.03674616371955342,
+      "epoch": 1.6775396085740915,
+      "grad_norm": 1.45356202253188,
+      "learning_rate": 9.796341042457715e-06,
+      "loss": 1.8586,
+      "step": 1800
+    },
+    {
+      "cosine_similarity": 0.33915150597085436,
+      "epoch": 1.6784715750232992,
+      "grad_norm": 1.3892373434992955,
+      "learning_rate": 9.789437348981705e-06,
+      "loss": 1.9234,
+      "step": 1801
+    },
+    {
+      "cosine_similarity": 0.003361778196952766,
+      "epoch": 1.679403541472507,
+      "grad_norm": 1.1449849251191169,
+      "learning_rate": 9.782533655505696e-06,
+      "loss": 1.6346,
+      "step": 1802
+    },
+    {
+      "cosine_similarity": 0.1802472491767407,
+      "epoch": 1.6803355079217148,
+      "grad_norm": 1.1176285133334067,
+      "learning_rate": 9.775629962029687e-06,
+      "loss": 1.7218,
+      "step": 1803
+    },
+    {
+      "cosine_similarity": 0.018996372406768464,
+      "epoch": 1.6812674743709226,
+      "grad_norm": 1.1092037916818769,
+      "learning_rate": 9.768726268553676e-06,
+      "loss": 1.7589,
+      "step": 1804
+    },
+    {
+      "cosine_similarity": 0.3783764849339527,
+      "epoch": 1.6821994408201304,
+      "grad_norm": 1.172423695611715,
+      "learning_rate": 9.761822575077667e-06,
+      "loss": 1.8191,
+      "step": 1805
+    },
+    {
+      "cosine_similarity": 0.22111449776639094,
+      "epoch": 1.6831314072693382,
+      "grad_norm": 1.2166663177908899,
+      "learning_rate": 9.754918881601657e-06,
+      "loss": 1.5967,
+      "step": 1806
+    },
+    {
+      "cosine_similarity": 0.08044324782498191,
+      "epoch": 1.684063373718546,
+      "grad_norm": 1.1945961144414612,
+      "learning_rate": 9.748015188125648e-06,
+      "loss": 1.6655,
+      "step": 1807
+    },
+    {
+      "cosine_similarity": 0.3632402354252731,
+      "epoch": 1.684995340167754,
+      "grad_norm": 1.1686666069412803,
+      "learning_rate": 9.741111494649639e-06,
+      "loss": 2.0752,
+      "step": 1808
+    },
+    {
+      "cosine_similarity": 0.11269857714781265,
+      "epoch": 1.6859273066169618,
+      "grad_norm": 1.0900709234693515,
+      "learning_rate": 9.734207801173628e-06,
+      "loss": 1.7076,
+      "step": 1809
+    },
+    {
+      "cosine_similarity": 0.2867421044000691,
+      "epoch": 1.6868592730661698,
+      "grad_norm": 1.2668865998097456,
+      "learning_rate": 9.727304107697619e-06,
+      "loss": 1.7044,
+      "step": 1810
+    },
+    {
+      "cosine_similarity": 0.4430334390584754,
+      "epoch": 1.6877912395153776,
+      "grad_norm": 1.293304380153078,
+      "learning_rate": 9.72040041422161e-06,
+      "loss": 1.9446,
+      "step": 1811
+    },
+    {
+      "cosine_similarity": 0.13417546295248886,
+      "epoch": 1.6887232059645854,
+      "grad_norm": 1.154402399575446,
+      "learning_rate": 9.7134967207456e-06,
+      "loss": 1.5522,
+      "step": 1812
+    },
+    {
+      "cosine_similarity": -0.15788498486879649,
+      "epoch": 1.6896551724137931,
+      "grad_norm": 1.5115135817088083,
+      "learning_rate": 9.70659302726959e-06,
+      "loss": 1.6785,
+      "step": 1813
+    },
+    {
+      "cosine_similarity": 0.14630575226889506,
+      "epoch": 1.690587138863001,
+      "grad_norm": 1.2030381168702267,
+      "learning_rate": 9.69968933379358e-06,
+      "loss": 1.5293,
+      "step": 1814
+    },
+    {
+      "cosine_similarity": 0.3861873681184119,
+      "epoch": 1.6915191053122087,
+      "grad_norm": 1.4793051697845805,
+      "learning_rate": 9.69278564031757e-06,
+      "loss": 2.2109,
+      "step": 1815
+    },
+    {
+      "cosine_similarity": 0.3834773139151327,
+      "epoch": 1.6924510717614165,
+      "grad_norm": 1.0910833983929873,
+      "learning_rate": 9.685881946841561e-06,
+      "loss": 1.8641,
+      "step": 1816
+    },
+    {
+      "cosine_similarity": 0.006401128033754425,
+      "epoch": 1.6933830382106243,
+      "grad_norm": 1.1896776583858546,
+      "learning_rate": 9.678978253365552e-06,
+      "loss": 1.6539,
+      "step": 1817
+    },
+    {
+      "cosine_similarity": -0.021339266662954486,
+      "epoch": 1.694315004659832,
+      "grad_norm": 0.8958080972651747,
+      "learning_rate": 9.672074559889541e-06,
+      "loss": 1.4358,
+      "step": 1818
+    },
+    {
+      "cosine_similarity": 0.5172327817448896,
+      "epoch": 1.69524697110904,
+      "grad_norm": 1.4091254005302138,
+      "learning_rate": 9.665170866413532e-06,
+      "loss": 1.7136,
+      "step": 1819
+    },
+    {
+      "cosine_similarity": -0.01566874291489338,
+      "epoch": 1.696178937558248,
+      "grad_norm": 1.247498015699088,
+      "learning_rate": 9.658267172937523e-06,
+      "loss": 1.4721,
+      "step": 1820
+    },
+    {
+      "cosine_similarity": 0.19662626864525298,
+      "epoch": 1.6971109040074557,
+      "grad_norm": 1.0936278439883103,
+      "learning_rate": 9.651363479461513e-06,
+      "loss": 1.6829,
+      "step": 1821
+    },
+    {
+      "cosine_similarity": 0.18989087537940988,
+      "epoch": 1.6980428704566637,
+      "grad_norm": 1.4137241616139042,
+      "learning_rate": 9.644459785985502e-06,
+      "loss": 1.5899,
+      "step": 1822
+    },
+    {
+      "cosine_similarity": 0.20645550624227244,
+      "epoch": 1.6989748369058715,
+      "grad_norm": 1.2183798619593835,
+      "learning_rate": 9.637556092509493e-06,
+      "loss": 1.6693,
+      "step": 1823
+    },
+    {
+      "cosine_similarity": 0.6397117004977263,
+      "epoch": 1.6999068033550793,
+      "grad_norm": 1.4140625356695944,
+      "learning_rate": 9.630652399033484e-06,
+      "loss": 1.8909,
+      "step": 1824
+    },
+    {
+      "cosine_similarity": -0.017209945897660773,
+      "epoch": 1.700838769804287,
+      "grad_norm": 1.1833296112404956,
+      "learning_rate": 9.623748705557475e-06,
+      "loss": 1.8571,
+      "step": 1825
+    },
+    {
+      "cosine_similarity": 0.1607183381508325,
+      "epoch": 1.7017707362534948,
+      "grad_norm": 1.4467275362624605,
+      "learning_rate": 9.616845012081464e-06,
+      "loss": 2.0847,
+      "step": 1826
+    },
+    {
+      "cosine_similarity": 0.2668751670445859,
+      "epoch": 1.7027027027027026,
+      "grad_norm": 1.124988855759916,
+      "learning_rate": 9.609941318605454e-06,
+      "loss": 1.7676,
+      "step": 1827
+    },
+    {
+      "cosine_similarity": 0.22817549734036877,
+      "epoch": 1.7036346691519104,
+      "grad_norm": 1.430595701933953,
+      "learning_rate": 9.603037625129445e-06,
+      "loss": 1.9621,
+      "step": 1828
+    },
+    {
+      "cosine_similarity": 0.09915843687543313,
+      "epoch": 1.7045666356011182,
+      "grad_norm": 1.4105226280983119,
+      "learning_rate": 9.596133931653436e-06,
+      "loss": 2.0505,
+      "step": 1829
+    },
+    {
+      "cosine_similarity": 0.003465386804078875,
+      "epoch": 1.7054986020503262,
+      "grad_norm": 1.2390180733761964,
+      "learning_rate": 9.589230238177427e-06,
+      "loss": 1.9557,
+      "step": 1830
+    },
+    {
+      "cosine_similarity": 0.2767454528637983,
+      "epoch": 1.706430568499534,
+      "grad_norm": 1.1125978738165898,
+      "learning_rate": 9.582326544701416e-06,
+      "loss": 1.8638,
+      "step": 1831
+    },
+    {
+      "cosine_similarity": 0.4679803542865848,
+      "epoch": 1.7073625349487418,
+      "grad_norm": 1.1259635832974124,
+      "learning_rate": 9.575422851225406e-06,
+      "loss": 1.7134,
+      "step": 1832
+    },
+    {
+      "cosine_similarity": -0.04320483596130231,
+      "epoch": 1.7082945013979498,
+      "grad_norm": 1.259788540988728,
+      "learning_rate": 9.568519157749397e-06,
+      "loss": 1.5896,
+      "step": 1833
+    },
+    {
+      "cosine_similarity": 0.054842516074907074,
+      "epoch": 1.7092264678471576,
+      "grad_norm": 1.070670443240842,
+      "learning_rate": 9.561615464273388e-06,
+      "loss": 1.6696,
+      "step": 1834
+    },
+    {
+      "cosine_similarity": -0.12147417096278633,
+      "epoch": 1.7101584342963654,
+      "grad_norm": 1.0791253685020015,
+      "learning_rate": 9.554711770797377e-06,
+      "loss": 1.3869,
+      "step": 1835
+    },
+    {
+      "cosine_similarity": 0.15720401785913782,
+      "epoch": 1.7110904007455732,
+      "grad_norm": 1.212429740582367,
+      "learning_rate": 9.547808077321368e-06,
+      "loss": 2.052,
+      "step": 1836
+    },
+    {
+      "cosine_similarity": -0.07281131965440596,
+      "epoch": 1.712022367194781,
+      "grad_norm": 1.240269844740701,
+      "learning_rate": 9.540904383845358e-06,
+      "loss": 1.6579,
+      "step": 1837
+    },
+    {
+      "cosine_similarity": 0.19557760755012255,
+      "epoch": 1.7129543336439887,
+      "grad_norm": 1.129757504579066,
+      "learning_rate": 9.534000690369349e-06,
+      "loss": 1.6457,
+      "step": 1838
+    },
+    {
+      "cosine_similarity": 0.16871098255578806,
+      "epoch": 1.7138863000931965,
+      "grad_norm": 1.3818762237346673,
+      "learning_rate": 9.52709699689334e-06,
+      "loss": 1.8282,
+      "step": 1839
+    },
+    {
+      "cosine_similarity": 0.1355243024977273,
+      "epoch": 1.7148182665424043,
+      "grad_norm": 1.360889238973069,
+      "learning_rate": 9.520193303417329e-06,
+      "loss": 1.4992,
+      "step": 1840
+    },
+    {
+      "cosine_similarity": -0.013418337797978032,
+      "epoch": 1.7157502329916123,
+      "grad_norm": 1.3380482576863706,
+      "learning_rate": 9.51328960994132e-06,
+      "loss": 1.9011,
+      "step": 1841
+    },
+    {
+      "cosine_similarity": -0.04404898821103065,
+      "epoch": 1.7166821994408201,
+      "grad_norm": 1.372986302011595,
+      "learning_rate": 9.50638591646531e-06,
+      "loss": 1.8203,
+      "step": 1842
+    },
+    {
+      "cosine_similarity": -0.007771493984933043,
+      "epoch": 1.717614165890028,
+      "grad_norm": 1.2988566322070947,
+      "learning_rate": 9.499482222989301e-06,
+      "loss": 1.8172,
+      "step": 1843
+    },
+    {
+      "cosine_similarity": 0.1149762286349186,
+      "epoch": 1.718546132339236,
+      "grad_norm": 0.9044591565645814,
+      "learning_rate": 9.49257852951329e-06,
+      "loss": 1.549,
+      "step": 1844
+    },
+    {
+      "cosine_similarity": -0.020586360939312306,
+      "epoch": 1.7194780987884437,
+      "grad_norm": 1.4452318323470272,
+      "learning_rate": 9.48567483603728e-06,
+      "loss": 1.8743,
+      "step": 1845
+    },
+    {
+      "cosine_similarity": 0.088847685742657,
+      "epoch": 1.7204100652376515,
+      "grad_norm": 1.2051449023440075,
+      "learning_rate": 9.47877114256127e-06,
+      "loss": 1.7977,
+      "step": 1846
+    },
+    {
+      "cosine_similarity": 0.14360098014786024,
+      "epoch": 1.7213420316868593,
+      "grad_norm": 1.0945218387620304,
+      "learning_rate": 9.47186744908526e-06,
+      "loss": 1.5731,
+      "step": 1847
+    },
+    {
+      "cosine_similarity": 0.19287100373595856,
+      "epoch": 1.722273998136067,
+      "grad_norm": 1.2977223172779984,
+      "learning_rate": 9.464963755609253e-06,
+      "loss": 1.7217,
+      "step": 1848
+    },
+    {
+      "cosine_similarity": 0.29749186174170494,
+      "epoch": 1.7232059645852749,
+      "grad_norm": 1.176635555997596,
+      "learning_rate": 9.458060062133242e-06,
+      "loss": 1.4199,
+      "step": 1849
+    },
+    {
+      "cosine_similarity": 0.04703330970820499,
+      "epoch": 1.7241379310344827,
+      "grad_norm": 1.2709676254405229,
+      "learning_rate": 9.451156368657233e-06,
+      "loss": 1.5819,
+      "step": 1850
+    },
+    {
+      "cosine_similarity": 0.2795254456013907,
+      "epoch": 1.7250698974836904,
+      "grad_norm": 1.2229656352100062,
+      "learning_rate": 9.444252675181222e-06,
+      "loss": 1.8787,
+      "step": 1851
+    },
+    {
+      "cosine_similarity": 0.031935866768045856,
+      "epoch": 1.7260018639328985,
+      "grad_norm": 1.0913253156359093,
+      "learning_rate": 9.437348981705212e-06,
+      "loss": 1.6995,
+      "step": 1852
+    },
+    {
+      "cosine_similarity": 0.2108123249070698,
+      "epoch": 1.7269338303821062,
+      "grad_norm": 1.3012587648399572,
+      "learning_rate": 9.430445288229203e-06,
+      "loss": 1.7144,
+      "step": 1853
+    },
+    {
+      "cosine_similarity": -0.004129746006769018,
+      "epoch": 1.727865796831314,
+      "grad_norm": 0.9983918574754842,
+      "learning_rate": 9.423541594753194e-06,
+      "loss": 1.6091,
+      "step": 1854
+    },
+    {
+      "cosine_similarity": 0.4952046393123229,
+      "epoch": 1.728797763280522,
+      "grad_norm": 1.3434100759588816,
+      "learning_rate": 9.416637901277183e-06,
+      "loss": 1.4951,
+      "step": 1855
+    },
+    {
+      "cosine_similarity": -0.05475874673854047,
+      "epoch": 1.7297297297297298,
+      "grad_norm": 1.061762405243965,
+      "learning_rate": 9.409734207801174e-06,
+      "loss": 1.5263,
+      "step": 1856
+    },
+    {
+      "cosine_similarity": -0.0063481718973317225,
+      "epoch": 1.7306616961789376,
+      "grad_norm": 1.104650516350256,
+      "learning_rate": 9.402830514325164e-06,
+      "loss": 1.6107,
+      "step": 1857
+    },
+    {
+      "cosine_similarity": 0.11213768431514445,
+      "epoch": 1.7315936626281454,
+      "grad_norm": 1.1465366067463436,
+      "learning_rate": 9.395926820849155e-06,
+      "loss": 1.8286,
+      "step": 1858
+    },
+    {
+      "cosine_similarity": 0.30077670007049323,
+      "epoch": 1.7325256290773532,
+      "grad_norm": 1.1143738380589867,
+      "learning_rate": 9.389023127373146e-06,
+      "loss": 1.7576,
+      "step": 1859
+    },
+    {
+      "cosine_similarity": 0.33735485070462695,
+      "epoch": 1.733457595526561,
+      "grad_norm": 1.2235131987600836,
+      "learning_rate": 9.382119433897135e-06,
+      "loss": 1.9618,
+      "step": 1860
+    },
+    {
+      "cosine_similarity": 0.14592303547961166,
+      "epoch": 1.7343895619757688,
+      "grad_norm": 1.2381332856254421,
+      "learning_rate": 9.375215740421126e-06,
+      "loss": 1.7516,
+      "step": 1861
+    },
+    {
+      "cosine_similarity": 0.05973867992021711,
+      "epoch": 1.7353215284249766,
+      "grad_norm": 1.0906220970486205,
+      "learning_rate": 9.368312046945116e-06,
+      "loss": 1.6168,
+      "step": 1862
+    },
+    {
+      "cosine_similarity": 0.0038927719491917948,
+      "epoch": 1.7362534948741846,
+      "grad_norm": 1.1471803126185314,
+      "learning_rate": 9.361408353469107e-06,
+      "loss": 1.4919,
+      "step": 1863
+    },
+    {
+      "cosine_similarity": 0.1738167180917232,
+      "epoch": 1.7371854613233924,
+      "grad_norm": 1.484033588848656,
+      "learning_rate": 9.354504659993096e-06,
+      "loss": 1.7164,
+      "step": 1864
+    },
+    {
+      "cosine_similarity": 0.13296313372556082,
+      "epoch": 1.7381174277726001,
+      "grad_norm": 1.1821135083264003,
+      "learning_rate": 9.347600966517087e-06,
+      "loss": 1.5346,
+      "step": 1865
+    },
+    {
+      "cosine_similarity": 0.14396212872677383,
+      "epoch": 1.7390493942218082,
+      "grad_norm": 1.4233715833741791,
+      "learning_rate": 9.340697273041078e-06,
+      "loss": 1.9239,
+      "step": 1866
+    },
+    {
+      "cosine_similarity": 0.05857173753373368,
+      "epoch": 1.739981360671016,
+      "grad_norm": 1.0989194380827074,
+      "learning_rate": 9.333793579565068e-06,
+      "loss": 1.8997,
+      "step": 1867
+    },
+    {
+      "cosine_similarity": 0.28820909833771285,
+      "epoch": 1.7409133271202237,
+      "grad_norm": 1.329414358966972,
+      "learning_rate": 9.326889886089059e-06,
+      "loss": 1.6634,
+      "step": 1868
+    },
+    {
+      "cosine_similarity": 0.30140646450418096,
+      "epoch": 1.7418452935694315,
+      "grad_norm": 1.0073018393412119,
+      "learning_rate": 9.319986192613048e-06,
+      "loss": 1.6455,
+      "step": 1869
+    },
+    {
+      "cosine_similarity": -0.007261650373019727,
+      "epoch": 1.7427772600186393,
+      "grad_norm": 1.710062672388617,
+      "learning_rate": 9.313082499137039e-06,
+      "loss": 1.9576,
+      "step": 1870
+    },
+    {
+      "cosine_similarity": 0.32926439207916136,
+      "epoch": 1.743709226467847,
+      "grad_norm": 1.2313244501471252,
+      "learning_rate": 9.30617880566103e-06,
+      "loss": 1.9592,
+      "step": 1871
+    },
+    {
+      "cosine_similarity": 0.03983106853423816,
+      "epoch": 1.7446411929170549,
+      "grad_norm": 1.213300330065946,
+      "learning_rate": 9.29927511218502e-06,
+      "loss": 1.691,
+      "step": 1872
+    },
+    {
+      "cosine_similarity": 0.2995962621370624,
+      "epoch": 1.7455731593662627,
+      "grad_norm": 1.3325574843080734,
+      "learning_rate": 9.29237141870901e-06,
+      "loss": 2.1088,
+      "step": 1873
+    },
+    {
+      "cosine_similarity": 0.05403796175882507,
+      "epoch": 1.7465051258154707,
+      "grad_norm": 1.5483334008246366,
+      "learning_rate": 9.285467725233e-06,
+      "loss": 1.7731,
+      "step": 1874
+    },
+    {
+      "cosine_similarity": 0.3498784674635732,
+      "epoch": 1.7474370922646785,
+      "grad_norm": 1.3440422759012287,
+      "learning_rate": 9.278564031756991e-06,
+      "loss": 2.5788,
+      "step": 1875
+    },
+    {
+      "cosine_similarity": 0.3811793422842541,
+      "epoch": 1.7483690587138863,
+      "grad_norm": 1.171193033471971,
+      "learning_rate": 9.271660338280982e-06,
+      "loss": 1.738,
+      "step": 1876
+    },
+    {
+      "cosine_similarity": 0.3707758413951055,
+      "epoch": 1.7493010251630943,
+      "grad_norm": 1.2737987431435631,
+      "learning_rate": 9.264756644804972e-06,
+      "loss": 1.5781,
+      "step": 1877
+    },
+    {
+      "cosine_similarity": 0.08996297989496709,
+      "epoch": 1.750232991612302,
+      "grad_norm": 1.3543801323580202,
+      "learning_rate": 9.257852951328961e-06,
+      "loss": 1.9882,
+      "step": 1878
+    },
+    {
+      "cosine_similarity": 0.2285563282407523,
+      "epoch": 1.7511649580615098,
+      "grad_norm": 1.238861765387152,
+      "learning_rate": 9.250949257852952e-06,
+      "loss": 1.5787,
+      "step": 1879
+    },
+    {
+      "cosine_similarity": 0.0641083094408615,
+      "epoch": 1.7520969245107176,
+      "grad_norm": 1.1861689893875154,
+      "learning_rate": 9.244045564376943e-06,
+      "loss": 1.4134,
+      "step": 1880
+    },
+    {
+      "cosine_similarity": 0.053003492016036664,
+      "epoch": 1.7530288909599254,
+      "grad_norm": 1.0438416103189732,
+      "learning_rate": 9.237141870900934e-06,
+      "loss": 1.7895,
+      "step": 1881
+    },
+    {
+      "cosine_similarity": 0.18303927529953928,
+      "epoch": 1.7539608574091332,
+      "grad_norm": 1.1432469108611474,
+      "learning_rate": 9.230238177424923e-06,
+      "loss": 1.8353,
+      "step": 1882
+    },
+    {
+      "cosine_similarity": 0.26640935608060656,
+      "epoch": 1.754892823858341,
+      "grad_norm": 1.208251396709718,
+      "learning_rate": 9.223334483948913e-06,
+      "loss": 1.437,
+      "step": 1883
+    },
+    {
+      "cosine_similarity": 0.0278898249251788,
+      "epoch": 1.7558247903075488,
+      "grad_norm": 1.108768869395917,
+      "learning_rate": 9.216430790472904e-06,
+      "loss": 1.9278,
+      "step": 1884
+    },
+    {
+      "cosine_similarity": -0.042516409904588846,
+      "epoch": 1.7567567567567568,
+      "grad_norm": 2.114524332957321,
+      "learning_rate": 9.209527096996895e-06,
+      "loss": 1.5585,
+      "step": 1885
+    },
+    {
+      "cosine_similarity": 0.13942109546193873,
+      "epoch": 1.7576887232059646,
+      "grad_norm": 1.2207872583611294,
+      "learning_rate": 9.202623403520884e-06,
+      "loss": 1.9836,
+      "step": 1886
+    },
+    {
+      "cosine_similarity": 0.02103184137177447,
+      "epoch": 1.7586206896551724,
+      "grad_norm": 1.0334154465706546,
+      "learning_rate": 9.195719710044875e-06,
+      "loss": 1.9855,
+      "step": 1887
+    },
+    {
+      "cosine_similarity": 0.08080240019507902,
+      "epoch": 1.7595526561043804,
+      "grad_norm": 1.3594621362273658,
+      "learning_rate": 9.188816016568865e-06,
+      "loss": 2.0291,
+      "step": 1888
+    },
+    {
+      "cosine_similarity": 0.12196906541901707,
+      "epoch": 1.7604846225535882,
+      "grad_norm": 1.3192330116870568,
+      "learning_rate": 9.181912323092856e-06,
+      "loss": 1.8036,
+      "step": 1889
+    },
+    {
+      "cosine_similarity": 0.20044628850879742,
+      "epoch": 1.761416589002796,
+      "grad_norm": 1.1589334052917588,
+      "learning_rate": 9.175008629616847e-06,
+      "loss": 1.64,
+      "step": 1890
+    },
+    {
+      "cosine_similarity": 0.06635928535813694,
+      "epoch": 1.7623485554520038,
+      "grad_norm": 1.6797990768612396,
+      "learning_rate": 9.168104936140836e-06,
+      "loss": 2.1354,
+      "step": 1891
+    },
+    {
+      "cosine_similarity": 0.4869580397499899,
+      "epoch": 1.7632805219012115,
+      "grad_norm": 1.2063463127570273,
+      "learning_rate": 9.161201242664827e-06,
+      "loss": 1.9866,
+      "step": 1892
+    },
+    {
+      "cosine_similarity": 0.731138537443921,
+      "epoch": 1.7642124883504193,
+      "grad_norm": 1.3607175241657539,
+      "learning_rate": 9.154297549188816e-06,
+      "loss": 1.5502,
+      "step": 1893
+    },
+    {
+      "cosine_similarity": 0.408840253616137,
+      "epoch": 1.7651444547996271,
+      "grad_norm": 1.2900206180265856,
+      "learning_rate": 9.147393855712808e-06,
+      "loss": 1.9971,
+      "step": 1894
+    },
+    {
+      "cosine_similarity": -0.18947734629092203,
+      "epoch": 1.766076421248835,
+      "grad_norm": 1.2793182567423245,
+      "learning_rate": 9.140490162236797e-06,
+      "loss": 1.9515,
+      "step": 1895
+    },
+    {
+      "cosine_similarity": -0.2520126490651911,
+      "epoch": 1.767008387698043,
+      "grad_norm": 1.51513290303227,
+      "learning_rate": 9.133586468760788e-06,
+      "loss": 1.7816,
+      "step": 1896
+    },
+    {
+      "cosine_similarity": 0.4427113154405146,
+      "epoch": 1.7679403541472507,
+      "grad_norm": 1.3618498563185155,
+      "learning_rate": 9.126682775284779e-06,
+      "loss": 1.8922,
+      "step": 1897
+    },
+    {
+      "cosine_similarity": 0.8752261966501979,
+      "epoch": 1.7688723205964585,
+      "grad_norm": 1.2829796929958417,
+      "learning_rate": 9.119779081808768e-06,
+      "loss": 2.1419,
+      "step": 1898
+    },
+    {
+      "cosine_similarity": 0.21961292324439344,
+      "epoch": 1.7698042870456665,
+      "grad_norm": 1.1410676647808433,
+      "learning_rate": 9.11287538833276e-06,
+      "loss": 1.7545,
+      "step": 1899
+    },
+    {
+      "cosine_similarity": 0.21622048030378546,
+      "epoch": 1.7707362534948743,
+      "grad_norm": 1.3689928843005104,
+      "learning_rate": 9.105971694856749e-06,
+      "loss": 1.6737,
+      "step": 1900
+    },
+    {
+      "cosine_similarity": 0.3566264412804844,
+      "epoch": 1.771668219944082,
+      "grad_norm": 1.0852303248280175,
+      "learning_rate": 9.09906800138074e-06,
+      "loss": 1.6306,
+      "step": 1901
+    },
+    {
+      "cosine_similarity": -0.04310814208976749,
+      "epoch": 1.7726001863932899,
+      "grad_norm": 0.9574902154093584,
+      "learning_rate": 9.092164307904729e-06,
+      "loss": 1.1813,
+      "step": 1902
+    },
+    {
+      "cosine_similarity": 0.41200184091284925,
+      "epoch": 1.7735321528424977,
+      "grad_norm": 1.172893128410929,
+      "learning_rate": 9.08526061442872e-06,
+      "loss": 1.7649,
+      "step": 1903
+    },
+    {
+      "cosine_similarity": 0.007839145527361781,
+      "epoch": 1.7744641192917054,
+      "grad_norm": 1.1903871964065569,
+      "learning_rate": 9.07835692095271e-06,
+      "loss": 1.6636,
+      "step": 1904
+    },
+    {
+      "cosine_similarity": 0.06010803152461836,
+      "epoch": 1.7753960857409132,
+      "grad_norm": 1.1702137413129765,
+      "learning_rate": 9.071453227476701e-06,
+      "loss": 1.9912,
+      "step": 1905
+    },
+    {
+      "cosine_similarity": 0.3030992787409179,
+      "epoch": 1.776328052190121,
+      "grad_norm": 1.1984180813406324,
+      "learning_rate": 9.06454953400069e-06,
+      "loss": 1.5585,
+      "step": 1906
+    },
+    {
+      "cosine_similarity": 0.2649410405588889,
+      "epoch": 1.777260018639329,
+      "grad_norm": 1.0669529474591926,
+      "learning_rate": 9.05764584052468e-06,
+      "loss": 1.7739,
+      "step": 1907
+    },
+    {
+      "cosine_similarity": 0.14515215180894367,
+      "epoch": 1.7781919850885368,
+      "grad_norm": 0.9986504243484332,
+      "learning_rate": 9.050742147048671e-06,
+      "loss": 1.7988,
+      "step": 1908
+    },
+    {
+      "cosine_similarity": -0.023100823105454447,
+      "epoch": 1.7791239515377446,
+      "grad_norm": 1.1752044371280945,
+      "learning_rate": 9.043838453572662e-06,
+      "loss": 1.9093,
+      "step": 1909
+    },
+    {
+      "cosine_similarity": 0.06361981872255502,
+      "epoch": 1.7800559179869526,
+      "grad_norm": 1.2886682403372336,
+      "learning_rate": 9.036934760096653e-06,
+      "loss": 1.8911,
+      "step": 1910
+    },
+    {
+      "cosine_similarity": 0.07860492894716484,
+      "epoch": 1.7809878844361604,
+      "grad_norm": 1.1638870832600765,
+      "learning_rate": 9.030031066620642e-06,
+      "loss": 1.9987,
+      "step": 1911
+    },
+    {
+      "cosine_similarity": 0.33719819234780746,
+      "epoch": 1.7819198508853682,
+      "grad_norm": 1.008492673967196,
+      "learning_rate": 9.023127373144633e-06,
+      "loss": 1.3139,
+      "step": 1912
+    },
+    {
+      "cosine_similarity": 0.13258194182908378,
+      "epoch": 1.782851817334576,
+      "grad_norm": 1.0178867416159152,
+      "learning_rate": 9.016223679668623e-06,
+      "loss": 1.543,
+      "step": 1913
+    },
+    {
+      "cosine_similarity": 0.09921047376164442,
+      "epoch": 1.7837837837837838,
+      "grad_norm": 1.2006148747349974,
+      "learning_rate": 9.009319986192614e-06,
+      "loss": 1.4636,
+      "step": 1914
+    },
+    {
+      "cosine_similarity": 0.1452957968858803,
+      "epoch": 1.7847157502329916,
+      "grad_norm": 1.321035329659599,
+      "learning_rate": 9.002416292716603e-06,
+      "loss": 1.6821,
+      "step": 1915
+    },
+    {
+      "cosine_similarity": 0.36721103478295214,
+      "epoch": 1.7856477166821993,
+      "grad_norm": 1.5238877970057747,
+      "learning_rate": 8.995512599240594e-06,
+      "loss": 1.7469,
+      "step": 1916
+    },
+    {
+      "cosine_similarity": 0.4892225184769114,
+      "epoch": 1.7865796831314071,
+      "grad_norm": 1.2357802115584662,
+      "learning_rate": 8.988608905764585e-06,
+      "loss": 1.6108,
+      "step": 1917
+    },
+    {
+      "cosine_similarity": 0.8572535313153034,
+      "epoch": 1.7875116495806151,
+      "grad_norm": 1.3960613446651913,
+      "learning_rate": 8.981705212288575e-06,
+      "loss": 1.9695,
+      "step": 1918
+    },
+    {
+      "cosine_similarity": 0.10245607567535735,
+      "epoch": 1.788443616029823,
+      "grad_norm": 1.2594846058667597,
+      "learning_rate": 8.974801518812566e-06,
+      "loss": 1.6669,
+      "step": 1919
+    },
+    {
+      "cosine_similarity": -0.1222966359361916,
+      "epoch": 1.7893755824790307,
+      "grad_norm": 1.079355255361701,
+      "learning_rate": 8.967897825336555e-06,
+      "loss": 1.5637,
+      "step": 1920
+    },
+    {
+      "cosine_similarity": 0.0820544795343138,
+      "epoch": 1.7903075489282387,
+      "grad_norm": 1.2390483194104476,
+      "learning_rate": 8.960994131860546e-06,
+      "loss": 1.5205,
+      "step": 1921
+    },
+    {
+      "cosine_similarity": 0.4181390576621248,
+      "epoch": 1.7912395153774465,
+      "grad_norm": 1.7724960789686026,
+      "learning_rate": 8.954090438384537e-06,
+      "loss": 1.6215,
+      "step": 1922
+    },
+    {
+      "cosine_similarity": 0.3216659013487731,
+      "epoch": 1.7921714818266543,
+      "grad_norm": 1.2123806531207757,
+      "learning_rate": 8.947186744908527e-06,
+      "loss": 1.7548,
+      "step": 1923
+    },
+    {
+      "cosine_similarity": 0.10083869768692329,
+      "epoch": 1.793103448275862,
+      "grad_norm": 1.135715977827818,
+      "learning_rate": 8.940283051432516e-06,
+      "loss": 1.5513,
+      "step": 1924
+    },
+    {
+      "cosine_similarity": 0.11115050660143788,
+      "epoch": 1.7940354147250699,
+      "grad_norm": 1.448045970824058,
+      "learning_rate": 8.933379357956507e-06,
+      "loss": 2.1199,
+      "step": 1925
+    },
+    {
+      "cosine_similarity": -0.11973071849039503,
+      "epoch": 1.7949673811742777,
+      "grad_norm": 1.0363339588288354,
+      "learning_rate": 8.926475664480498e-06,
+      "loss": 1.4613,
+      "step": 1926
+    },
+    {
+      "cosine_similarity": 0.31940435793320643,
+      "epoch": 1.7958993476234855,
+      "grad_norm": 1.1545910256790965,
+      "learning_rate": 8.919571971004489e-06,
+      "loss": 1.6644,
+      "step": 1927
+    },
+    {
+      "cosine_similarity": 0.3338694982466685,
+      "epoch": 1.7968313140726933,
+      "grad_norm": 1.0257883909536476,
+      "learning_rate": 8.91266827752848e-06,
+      "loss": 1.5985,
+      "step": 1928
+    },
+    {
+      "cosine_similarity": 0.18653344065503616,
+      "epoch": 1.7977632805219013,
+      "grad_norm": 1.1516548686710617,
+      "learning_rate": 8.905764584052468e-06,
+      "loss": 1.8365,
+      "step": 1929
+    },
+    {
+      "cosine_similarity": -0.024718552831992938,
+      "epoch": 1.798695246971109,
+      "grad_norm": 1.1073941673783785,
+      "learning_rate": 8.898860890576459e-06,
+      "loss": 1.8611,
+      "step": 1930
+    },
+    {
+      "cosine_similarity": 0.1302366710553835,
+      "epoch": 1.7996272134203168,
+      "grad_norm": 1.1128992383051124,
+      "learning_rate": 8.89195719710045e-06,
+      "loss": 1.6971,
+      "step": 1931
+    },
+    {
+      "cosine_similarity": 0.3122873800205302,
+      "epoch": 1.8005591798695249,
+      "grad_norm": 1.0337987793439887,
+      "learning_rate": 8.88505350362444e-06,
+      "loss": 1.6049,
+      "step": 1932
+    },
+    {
+      "cosine_similarity": 0.3065076297130843,
+      "epoch": 1.8014911463187326,
+      "grad_norm": 1.2606399879261387,
+      "learning_rate": 8.87814981014843e-06,
+      "loss": 1.7841,
+      "step": 1933
+    },
+    {
+      "cosine_similarity": 0.15195703983681097,
+      "epoch": 1.8024231127679404,
+      "grad_norm": 1.2012824494742502,
+      "learning_rate": 8.87124611667242e-06,
+      "loss": 1.7774,
+      "step": 1934
+    },
+    {
+      "cosine_similarity": 0.17526571972467989,
+      "epoch": 1.8033550792171482,
+      "grad_norm": 1.4559529545690695,
+      "learning_rate": 8.864342423196411e-06,
+      "loss": 1.5838,
+      "step": 1935
+    },
+    {
+      "cosine_similarity": 0.09938822528730153,
+      "epoch": 1.804287045666356,
+      "grad_norm": 1.1163235691914637,
+      "learning_rate": 8.857438729720402e-06,
+      "loss": 1.7744,
+      "step": 1936
+    },
+    {
+      "cosine_similarity": 0.37924070117065084,
+      "epoch": 1.8052190121155638,
+      "grad_norm": 1.270658555153114,
+      "learning_rate": 8.850535036244391e-06,
+      "loss": 1.7472,
+      "step": 1937
+    },
+    {
+      "cosine_similarity": -0.017212633512031675,
+      "epoch": 1.8061509785647716,
+      "grad_norm": 1.0819687971547538,
+      "learning_rate": 8.843631342768382e-06,
+      "loss": 1.4568,
+      "step": 1938
+    },
+    {
+      "cosine_similarity": -0.13211226353546787,
+      "epoch": 1.8070829450139794,
+      "grad_norm": 1.2338902227473658,
+      "learning_rate": 8.836727649292372e-06,
+      "loss": 1.3992,
+      "step": 1939
+    },
+    {
+      "cosine_similarity": 0.32366055312871084,
+      "epoch": 1.8080149114631874,
+      "grad_norm": 1.3082831692943946,
+      "learning_rate": 8.829823955816363e-06,
+      "loss": 2.0307,
+      "step": 1940
+    },
+    {
+      "cosine_similarity": 0.0551298877812234,
+      "epoch": 1.8089468779123952,
+      "grad_norm": 1.491326166516012,
+      "learning_rate": 8.822920262340354e-06,
+      "loss": 1.8941,
+      "step": 1941
+    },
+    {
+      "cosine_similarity": -0.014650489088072661,
+      "epoch": 1.809878844361603,
+      "grad_norm": 1.0551052770389207,
+      "learning_rate": 8.816016568864343e-06,
+      "loss": 1.6745,
+      "step": 1942
+    },
+    {
+      "cosine_similarity": -0.06885748455600377,
+      "epoch": 1.810810810810811,
+      "grad_norm": 1.1692530010698272,
+      "learning_rate": 8.809112875388334e-06,
+      "loss": 1.5441,
+      "step": 1943
+    },
+    {
+      "cosine_similarity": 0.0703489763642999,
+      "epoch": 1.8117427772600188,
+      "grad_norm": 1.0313707801635026,
+      "learning_rate": 8.802209181912323e-06,
+      "loss": 1.7197,
+      "step": 1944
+    },
+    {
+      "cosine_similarity": 0.0008498133673025532,
+      "epoch": 1.8126747437092265,
+      "grad_norm": 1.309514636669631,
+      "learning_rate": 8.795305488436315e-06,
+      "loss": 1.9652,
+      "step": 1945
+    },
+    {
+      "cosine_similarity": 0.27357329558469373,
+      "epoch": 1.8136067101584343,
+      "grad_norm": 1.1237444211025729,
+      "learning_rate": 8.788401794960304e-06,
+      "loss": 1.561,
+      "step": 1946
+    },
+    {
+      "cosine_similarity": 0.318172384165475,
+      "epoch": 1.8145386766076421,
+      "grad_norm": 1.2462098783341196,
+      "learning_rate": 8.781498101484295e-06,
+      "loss": 1.7611,
+      "step": 1947
+    },
+    {
+      "cosine_similarity": 0.12540252930015497,
+      "epoch": 1.81547064305685,
+      "grad_norm": 1.7945713123633584,
+      "learning_rate": 8.774594408008286e-06,
+      "loss": 1.8411,
+      "step": 1948
+    },
+    {
+      "cosine_similarity": 0.07112662033872849,
+      "epoch": 1.8164026095060577,
+      "grad_norm": 1.035849866063896,
+      "learning_rate": 8.767690714532275e-06,
+      "loss": 1.5385,
+      "step": 1949
+    },
+    {
+      "cosine_similarity": 0.21544099581759116,
+      "epoch": 1.8173345759552655,
+      "grad_norm": 1.0833009561993758,
+      "learning_rate": 8.760787021056267e-06,
+      "loss": 1.4838,
+      "step": 1950
+    },
+    {
+      "cosine_similarity": 0.22467345889618365,
+      "epoch": 1.8182665424044733,
+      "grad_norm": 1.1609271296443908,
+      "learning_rate": 8.753883327580256e-06,
+      "loss": 1.5291,
+      "step": 1951
+    },
+    {
+      "cosine_similarity": 0.0472873727057794,
+      "epoch": 1.8191985088536813,
+      "grad_norm": 1.2033240850696447,
+      "learning_rate": 8.746979634104247e-06,
+      "loss": 1.5369,
+      "step": 1952
+    },
+    {
+      "cosine_similarity": 0.13345945208734578,
+      "epoch": 1.820130475302889,
+      "grad_norm": 1.4934513821794275,
+      "learning_rate": 8.740075940628236e-06,
+      "loss": 1.6118,
+      "step": 1953
+    },
+    {
+      "cosine_similarity": 0.1775832798964298,
+      "epoch": 1.821062441752097,
+      "grad_norm": 1.4324318924591681,
+      "learning_rate": 8.733172247152227e-06,
+      "loss": 1.8303,
+      "step": 1954
+    },
+    {
+      "cosine_similarity": 0.22542968097517238,
+      "epoch": 1.8219944082013049,
+      "grad_norm": 1.1095214585036652,
+      "learning_rate": 8.726268553676217e-06,
+      "loss": 1.4799,
+      "step": 1955
+    },
+    {
+      "cosine_similarity": 0.1087432021141328,
+      "epoch": 1.8229263746505127,
+      "grad_norm": 1.3935180217776928,
+      "learning_rate": 8.719364860200208e-06,
+      "loss": 1.7424,
+      "step": 1956
+    },
+    {
+      "cosine_similarity": 0.11505377150169872,
+      "epoch": 1.8238583410997204,
+      "grad_norm": 1.20389592892656,
+      "learning_rate": 8.712461166724197e-06,
+      "loss": 2.0447,
+      "step": 1957
+    },
+    {
+      "cosine_similarity": 0.16478269563553433,
+      "epoch": 1.8247903075489282,
+      "grad_norm": 1.5094976683818297,
+      "learning_rate": 8.705557473248188e-06,
+      "loss": 1.7411,
+      "step": 1958
+    },
+    {
+      "cosine_similarity": 0.12102620708233862,
+      "epoch": 1.825722273998136,
+      "grad_norm": 1.2690269734519286,
+      "learning_rate": 8.698653779772179e-06,
+      "loss": 2.0348,
+      "step": 1959
+    },
+    {
+      "cosine_similarity": 0.05049789650619809,
+      "epoch": 1.8266542404473438,
+      "grad_norm": 1.2419959137669156,
+      "learning_rate": 8.69175008629617e-06,
+      "loss": 2.0451,
+      "step": 1960
+    },
+    {
+      "cosine_similarity": -0.07370901645583434,
+      "epoch": 1.8275862068965516,
+      "grad_norm": 1.00807961064055,
+      "learning_rate": 8.68484639282016e-06,
+      "loss": 1.6365,
+      "step": 1961
+    },
+    {
+      "cosine_similarity": -0.04765309648262065,
+      "epoch": 1.8285181733457594,
+      "grad_norm": 1.2218994519766138,
+      "learning_rate": 8.677942699344149e-06,
+      "loss": 2.0099,
+      "step": 1962
+    },
+    {
+      "cosine_similarity": 0.24849316623091977,
+      "epoch": 1.8294501397949674,
+      "grad_norm": 1.3211863620192268,
+      "learning_rate": 8.67103900586814e-06,
+      "loss": 2.0481,
+      "step": 1963
+    },
+    {
+      "cosine_similarity": 0.18664214909156132,
+      "epoch": 1.8303821062441752,
+      "grad_norm": 1.1338355341153457,
+      "learning_rate": 8.66413531239213e-06,
+      "loss": 1.5402,
+      "step": 1964
+    },
+    {
+      "cosine_similarity": 0.06959557284309983,
+      "epoch": 1.8313140726933832,
+      "grad_norm": 1.0706168897795032,
+      "learning_rate": 8.657231618916121e-06,
+      "loss": 1.8223,
+      "step": 1965
+    },
+    {
+      "cosine_similarity": 0.08293071517345232,
+      "epoch": 1.832246039142591,
+      "grad_norm": 1.2098031118077799,
+      "learning_rate": 8.65032792544011e-06,
+      "loss": 1.4244,
+      "step": 1966
+    },
+    {
+      "cosine_similarity": 0.32313643060998254,
+      "epoch": 1.8331780055917988,
+      "grad_norm": 1.1302904801537657,
+      "learning_rate": 8.643424231964101e-06,
+      "loss": 1.8421,
+      "step": 1967
+    },
+    {
+      "cosine_similarity": 0.43284624140532646,
+      "epoch": 1.8341099720410066,
+      "grad_norm": 1.346013337862704,
+      "learning_rate": 8.636520538488092e-06,
+      "loss": 1.83,
+      "step": 1968
+    },
+    {
+      "cosine_similarity": 0.13917508844940749,
+      "epoch": 1.8350419384902144,
+      "grad_norm": 1.1660453292618655,
+      "learning_rate": 8.629616845012082e-06,
+      "loss": 1.9839,
+      "step": 1969
+    },
+    {
+      "cosine_similarity": 0.14224596297870332,
+      "epoch": 1.8359739049394221,
+      "grad_norm": 1.146061010771569,
+      "learning_rate": 8.622713151536073e-06,
+      "loss": 1.6066,
+      "step": 1970
+    },
+    {
+      "cosine_similarity": -0.004308632998650107,
+      "epoch": 1.83690587138863,
+      "grad_norm": 1.0955505609536262,
+      "learning_rate": 8.615809458060062e-06,
+      "loss": 1.6377,
+      "step": 1971
+    },
+    {
+      "cosine_similarity": 0.4130242159899954,
+      "epoch": 1.8378378378378377,
+      "grad_norm": 1.3030581114297928,
+      "learning_rate": 8.608905764584053e-06,
+      "loss": 1.6535,
+      "step": 1972
+    },
+    {
+      "cosine_similarity": 0.2765790418817425,
+      "epoch": 1.8387698042870455,
+      "grad_norm": 1.1919370831596632,
+      "learning_rate": 8.602002071108044e-06,
+      "loss": 1.7339,
+      "step": 1973
+    },
+    {
+      "cosine_similarity": -0.06927237772330957,
+      "epoch": 1.8397017707362535,
+      "grad_norm": 1.3966510721067895,
+      "learning_rate": 8.595098377632034e-06,
+      "loss": 1.8386,
+      "step": 1974
+    },
+    {
+      "cosine_similarity": 0.21615812265512716,
+      "epoch": 1.8406337371854613,
+      "grad_norm": 1.3025536504686956,
+      "learning_rate": 8.588194684156023e-06,
+      "loss": 1.6587,
+      "step": 1975
+    },
+    {
+      "cosine_similarity": 0.07790340480084515,
+      "epoch": 1.8415657036346693,
+      "grad_norm": 1.134139607674466,
+      "learning_rate": 8.581290990680014e-06,
+      "loss": 1.4033,
+      "step": 1976
+    },
+    {
+      "cosine_similarity": 0.45173510936279787,
+      "epoch": 1.842497670083877,
+      "grad_norm": 1.1491984902838532,
+      "learning_rate": 8.574387297204005e-06,
+      "loss": 1.6511,
+      "step": 1977
+    },
+    {
+      "cosine_similarity": 0.290533206792949,
+      "epoch": 1.843429636533085,
+      "grad_norm": 1.1478922204391815,
+      "learning_rate": 8.567483603727996e-06,
+      "loss": 1.6465,
+      "step": 1978
+    },
+    {
+      "cosine_similarity": -0.1583793540913016,
+      "epoch": 1.8443616029822927,
+      "grad_norm": 1.4456288879683257,
+      "learning_rate": 8.560579910251986e-06,
+      "loss": 1.7675,
+      "step": 1979
+    },
+    {
+      "cosine_similarity": 0.09672992860217436,
+      "epoch": 1.8452935694315005,
+      "grad_norm": 1.1050957119159668,
+      "learning_rate": 8.553676216775975e-06,
+      "loss": 1.6643,
+      "step": 1980
+    },
+    {
+      "cosine_similarity": 0.02260302645442038,
+      "epoch": 1.8462255358807083,
+      "grad_norm": 1.2390221604291103,
+      "learning_rate": 8.546772523299966e-06,
+      "loss": 1.818,
+      "step": 1981
+    },
+    {
+      "cosine_similarity": 0.1107625899041805,
+      "epoch": 1.847157502329916,
+      "grad_norm": 1.061665145696232,
+      "learning_rate": 8.539868829823957e-06,
+      "loss": 1.6271,
+      "step": 1982
+    },
+    {
+      "cosine_similarity": 0.22781299998274096,
+      "epoch": 1.8480894687791238,
+      "grad_norm": 1.229162577708579,
+      "learning_rate": 8.532965136347948e-06,
+      "loss": 1.697,
+      "step": 1983
+    },
+    {
+      "cosine_similarity": 0.279539289161191,
+      "epoch": 1.8490214352283316,
+      "grad_norm": 1.2941821880750322,
+      "learning_rate": 8.526061442871937e-06,
+      "loss": 1.6024,
+      "step": 1984
+    },
+    {
+      "cosine_similarity": 0.26251738863651475,
+      "epoch": 1.8499534016775396,
+      "grad_norm": 1.2524582126908386,
+      "learning_rate": 8.519157749395927e-06,
+      "loss": 1.4501,
+      "step": 1985
+    },
+    {
+      "cosine_similarity": 0.2966970515354647,
+      "epoch": 1.8508853681267474,
+      "grad_norm": 1.59971621613397,
+      "learning_rate": 8.512254055919918e-06,
+      "loss": 1.7256,
+      "step": 1986
+    },
+    {
+      "cosine_similarity": -0.03308518334414638,
+      "epoch": 1.8518173345759554,
+      "grad_norm": 1.0799984061783958,
+      "learning_rate": 8.505350362443909e-06,
+      "loss": 1.5847,
+      "step": 1987
+    },
+    {
+      "cosine_similarity": -0.08495129750253076,
+      "epoch": 1.8527493010251632,
+      "grad_norm": 1.1853182501798045,
+      "learning_rate": 8.4984466689679e-06,
+      "loss": 1.5096,
+      "step": 1988
+    },
+    {
+      "cosine_similarity": 0.24167795895235591,
+      "epoch": 1.853681267474371,
+      "grad_norm": 1.1159998988215072,
+      "learning_rate": 8.491542975491889e-06,
+      "loss": 1.7796,
+      "step": 1989
+    },
+    {
+      "cosine_similarity": 0.4983875182237957,
+      "epoch": 1.8546132339235788,
+      "grad_norm": 1.329997052446646,
+      "learning_rate": 8.48463928201588e-06,
+      "loss": 1.8383,
+      "step": 1990
+    },
+    {
+      "cosine_similarity": 0.048739234734712145,
+      "epoch": 1.8555452003727866,
+      "grad_norm": 1.0893181202033222,
+      "learning_rate": 8.47773558853987e-06,
+      "loss": 1.7128,
+      "step": 1991
+    },
+    {
+      "cosine_similarity": 0.17526280398194688,
+      "epoch": 1.8564771668219944,
+      "grad_norm": 1.2159710176255847,
+      "learning_rate": 8.47083189506386e-06,
+      "loss": 2.0877,
+      "step": 1992
+    },
+    {
+      "cosine_similarity": 0.2908502157302147,
+      "epoch": 1.8574091332712022,
+      "grad_norm": 1.2783104438663688,
+      "learning_rate": 8.46392820158785e-06,
+      "loss": 1.974,
+      "step": 1993
+    },
+    {
+      "cosine_similarity": 0.27349247362182366,
+      "epoch": 1.85834109972041,
+      "grad_norm": 1.2276869922005993,
+      "learning_rate": 8.45702450811184e-06,
+      "loss": 1.779,
+      "step": 1994
+    },
+    {
+      "cosine_similarity": 0.27263143324611644,
+      "epoch": 1.8592730661696177,
+      "grad_norm": 1.2564676840866564,
+      "learning_rate": 8.45012081463583e-06,
+      "loss": 1.8014,
+      "step": 1995
+    },
+    {
+      "cosine_similarity": -0.15787975883043529,
+      "epoch": 1.8602050326188257,
+      "grad_norm": 1.1465009957307408,
+      "learning_rate": 8.443217121159822e-06,
+      "loss": 1.7246,
+      "step": 1996
+    },
+    {
+      "cosine_similarity": 0.29335105209308004,
+      "epoch": 1.8611369990680335,
+      "grad_norm": 1.186639583462888,
+      "learning_rate": 8.436313427683811e-06,
+      "loss": 1.5008,
+      "step": 1997
+    },
+    {
+      "cosine_similarity": 0.33458667848642243,
+      "epoch": 1.8620689655172413,
+      "grad_norm": 1.227947236934869,
+      "learning_rate": 8.429409734207802e-06,
+      "loss": 1.6288,
+      "step": 1998
+    },
+    {
+      "cosine_similarity": 0.5188514730634283,
+      "epoch": 1.8630009319664493,
+      "grad_norm": 1.1792379190236089,
+      "learning_rate": 8.422506040731793e-06,
+      "loss": 1.9102,
+      "step": 1999
+    },
+    {
+      "cosine_similarity": -0.018183487502381284,
+      "epoch": 1.8639328984156571,
+      "grad_norm": 1.3606358107671939,
+      "learning_rate": 8.415602347255782e-06,
+      "loss": 2.3035,
+      "step": 2000
+    },
+    {
+      "cosine_similarity": 0.15076274809747564,
+      "epoch": 1.864864864864865,
+      "grad_norm": 1.3600581927820248,
+      "learning_rate": 8.408698653779774e-06,
+      "loss": 1.6357,
+      "step": 2001
+    },
+    {
+      "cosine_similarity": 0.3107817085183365,
+      "epoch": 1.8657968313140727,
+      "grad_norm": 1.1948416539852187,
+      "learning_rate": 8.401794960303763e-06,
+      "loss": 1.8389,
+      "step": 2002
+    },
+    {
+      "cosine_similarity": 0.15529352734616209,
+      "epoch": 1.8667287977632805,
+      "grad_norm": 1.359653478011412,
+      "learning_rate": 8.394891266827754e-06,
+      "loss": 1.7425,
+      "step": 2003
+    },
+    {
+      "cosine_similarity": -0.0875050796100778,
+      "epoch": 1.8676607642124883,
+      "grad_norm": 1.5113293150166962,
+      "learning_rate": 8.387987573351743e-06,
+      "loss": 2.0325,
+      "step": 2004
+    },
+    {
+      "cosine_similarity": -0.021824158638160795,
+      "epoch": 1.868592730661696,
+      "grad_norm": 0.9079571333068369,
+      "learning_rate": 8.381083879875734e-06,
+      "loss": 1.4522,
+      "step": 2005
+    },
+    {
+      "cosine_similarity": 0.3047202011552034,
+      "epoch": 1.8695246971109039,
+      "grad_norm": 1.4153102613576338,
+      "learning_rate": 8.374180186399724e-06,
+      "loss": 1.9869,
+      "step": 2006
+    },
+    {
+      "cosine_similarity": 0.22016989679642207,
+      "epoch": 1.8704566635601119,
+      "grad_norm": 1.1975372214160858,
+      "learning_rate": 8.367276492923715e-06,
+      "loss": 1.7474,
+      "step": 2007
+    },
+    {
+      "cosine_similarity": 0.032513220846575416,
+      "epoch": 1.8713886300093197,
+      "grad_norm": 1.3768693709428497,
+      "learning_rate": 8.360372799447706e-06,
+      "loss": 1.9267,
+      "step": 2008
+    },
+    {
+      "cosine_similarity": 0.22698309416766613,
+      "epoch": 1.8723205964585274,
+      "grad_norm": 1.3678418870987925,
+      "learning_rate": 8.353469105971695e-06,
+      "loss": 1.9265,
+      "step": 2009
+    },
+    {
+      "cosine_similarity": 0.25677206428337107,
+      "epoch": 1.8732525629077355,
+      "grad_norm": 1.2172201702130974,
+      "learning_rate": 8.346565412495686e-06,
+      "loss": 1.6029,
+      "step": 2010
+    },
+    {
+      "cosine_similarity": -0.08339508024489416,
+      "epoch": 1.8741845293569432,
+      "grad_norm": 1.2688583874002057,
+      "learning_rate": 8.339661719019676e-06,
+      "loss": 1.6448,
+      "step": 2011
+    },
+    {
+      "cosine_similarity": 0.00592439659142771,
+      "epoch": 1.875116495806151,
+      "grad_norm": 1.3151534586392342,
+      "learning_rate": 8.332758025543667e-06,
+      "loss": 1.7343,
+      "step": 2012
+    },
+    {
+      "cosine_similarity": -0.030544560681917726,
+      "epoch": 1.8760484622553588,
+      "grad_norm": 0.9975227366615849,
+      "learning_rate": 8.325854332067656e-06,
+      "loss": 1.4845,
+      "step": 2013
+    },
+    {
+      "cosine_similarity": -0.09540304350925527,
+      "epoch": 1.8769804287045666,
+      "grad_norm": 1.2854464650791082,
+      "learning_rate": 8.318950638591647e-06,
+      "loss": 1.5805,
+      "step": 2014
+    },
+    {
+      "cosine_similarity": 0.2758353628094084,
+      "epoch": 1.8779123951537744,
+      "grad_norm": 1.0989903785995365,
+      "learning_rate": 8.312046945115638e-06,
+      "loss": 1.5023,
+      "step": 2015
+    },
+    {
+      "cosine_similarity": -0.1768059048903213,
+      "epoch": 1.8788443616029822,
+      "grad_norm": 1.014303367716466,
+      "learning_rate": 8.305143251639628e-06,
+      "loss": 1.4322,
+      "step": 2016
+    },
+    {
+      "cosine_similarity": 0.15559669349460709,
+      "epoch": 1.87977632805219,
+      "grad_norm": 1.4017555926428062,
+      "learning_rate": 8.298239558163617e-06,
+      "loss": 1.7924,
+      "step": 2017
+    },
+    {
+      "cosine_similarity": 0.3277356581239844,
+      "epoch": 1.880708294501398,
+      "grad_norm": 1.1227917458325087,
+      "learning_rate": 8.291335864687608e-06,
+      "loss": 1.4744,
+      "step": 2018
+    },
+    {
+      "cosine_similarity": 0.15445341855139447,
+      "epoch": 1.8816402609506058,
+      "grad_norm": 1.337199654486138,
+      "learning_rate": 8.284432171211599e-06,
+      "loss": 1.949,
+      "step": 2019
+    },
+    {
+      "cosine_similarity": -0.10758767547977795,
+      "epoch": 1.8825722273998136,
+      "grad_norm": 1.1317650958623153,
+      "learning_rate": 8.27752847773559e-06,
+      "loss": 1.6682,
+      "step": 2020
+    },
+    {
+      "cosine_similarity": 0.11713041241078427,
+      "epoch": 1.8835041938490216,
+      "grad_norm": 1.2271909603559847,
+      "learning_rate": 8.27062478425958e-06,
+      "loss": 1.8021,
+      "step": 2021
+    },
+    {
+      "cosine_similarity": 0.1989845123209029,
+      "epoch": 1.8844361602982294,
+      "grad_norm": 1.4595797443874436,
+      "learning_rate": 8.26372109078357e-06,
+      "loss": 1.7932,
+      "step": 2022
+    },
+    {
+      "cosine_similarity": -0.13863032893548932,
+      "epoch": 1.8853681267474371,
+      "grad_norm": 1.4079839844507371,
+      "learning_rate": 8.25681739730756e-06,
+      "loss": 1.8516,
+      "step": 2023
+    },
+    {
+      "cosine_similarity": 0.1587939663304333,
+      "epoch": 1.886300093196645,
+      "grad_norm": 1.686398870744004,
+      "learning_rate": 8.24991370383155e-06,
+      "loss": 1.8174,
+      "step": 2024
+    },
+    {
+      "cosine_similarity": 0.26759289197310726,
+      "epoch": 1.8872320596458527,
+      "grad_norm": 1.0989691866838853,
+      "learning_rate": 8.243010010355541e-06,
+      "loss": 1.5421,
+      "step": 2025
+    },
+    {
+      "cosine_similarity": 0.03546380009689646,
+      "epoch": 1.8881640260950605,
+      "grad_norm": 1.1667356989608013,
+      "learning_rate": 8.23610631687953e-06,
+      "loss": 1.6907,
+      "step": 2026
+    },
+    {
+      "cosine_similarity": -0.03828350495375238,
+      "epoch": 1.8890959925442683,
+      "grad_norm": 1.2587814404977806,
+      "learning_rate": 8.229202623403521e-06,
+      "loss": 1.5019,
+      "step": 2027
+    },
+    {
+      "cosine_similarity": 0.6994476723439341,
+      "epoch": 1.890027958993476,
+      "grad_norm": 1.321158827743466,
+      "learning_rate": 8.222298929927512e-06,
+      "loss": 1.8962,
+      "step": 2028
+    },
+    {
+      "cosine_similarity": 0.3838334768417961,
+      "epoch": 1.890959925442684,
+      "grad_norm": 1.0835263709897343,
+      "learning_rate": 8.215395236451503e-06,
+      "loss": 1.7327,
+      "step": 2029
+    },
+    {
+      "cosine_similarity": 0.04472034228525716,
+      "epoch": 1.8918918918918919,
+      "grad_norm": 1.1330200382733557,
+      "learning_rate": 8.208491542975493e-06,
+      "loss": 1.6001,
+      "step": 2030
+    },
+    {
+      "cosine_similarity": 0.3826916130040443,
+      "epoch": 1.8928238583410997,
+      "grad_norm": 1.1389505527527817,
+      "learning_rate": 8.201587849499482e-06,
+      "loss": 1.6614,
+      "step": 2031
+    },
+    {
+      "cosine_similarity": 0.0682390502922901,
+      "epoch": 1.8937558247903077,
+      "grad_norm": 1.2149857486223183,
+      "learning_rate": 8.194684156023473e-06,
+      "loss": 1.6943,
+      "step": 2032
+    },
+    {
+      "cosine_similarity": 0.20472155223941416,
+      "epoch": 1.8946877912395155,
+      "grad_norm": 0.9861772759090407,
+      "learning_rate": 8.187780462547464e-06,
+      "loss": 1.2525,
+      "step": 2033
+    },
+    {
+      "cosine_similarity": 0.04757125502710412,
+      "epoch": 1.8956197576887233,
+      "grad_norm": 1.1856978822113542,
+      "learning_rate": 8.180876769071455e-06,
+      "loss": 1.4933,
+      "step": 2034
+    },
+    {
+      "cosine_similarity": 0.2228182294709562,
+      "epoch": 1.896551724137931,
+      "grad_norm": 1.1963949446840352,
+      "learning_rate": 8.173973075595444e-06,
+      "loss": 1.9349,
+      "step": 2035
+    },
+    {
+      "cosine_similarity": 0.06460316505109881,
+      "epoch": 1.8974836905871388,
+      "grad_norm": 1.336796389168374,
+      "learning_rate": 8.167069382119434e-06,
+      "loss": 1.7337,
+      "step": 2036
+    },
+    {
+      "cosine_similarity": -0.059843220534454754,
+      "epoch": 1.8984156570363466,
+      "grad_norm": 1.0066983596348411,
+      "learning_rate": 8.160165688643425e-06,
+      "loss": 1.3793,
+      "step": 2037
+    },
+    {
+      "cosine_similarity": 0.15557315850312212,
+      "epoch": 1.8993476234855544,
+      "grad_norm": 1.2217829467900518,
+      "learning_rate": 8.153261995167416e-06,
+      "loss": 2.1121,
+      "step": 2038
+    },
+    {
+      "cosine_similarity": 0.1489385618362159,
+      "epoch": 1.9002795899347622,
+      "grad_norm": 1.2871325786600212,
+      "learning_rate": 8.146358301691407e-06,
+      "loss": 1.9362,
+      "step": 2039
+    },
+    {
+      "cosine_similarity": 0.12749065034945165,
+      "epoch": 1.9012115563839702,
+      "grad_norm": 1.1591955439770185,
+      "learning_rate": 8.139454608215396e-06,
+      "loss": 1.4478,
+      "step": 2040
+    },
+    {
+      "cosine_similarity": 0.16708446041443567,
+      "epoch": 1.902143522833178,
+      "grad_norm": 1.264091527285695,
+      "learning_rate": 8.132550914739386e-06,
+      "loss": 1.8534,
+      "step": 2041
+    },
+    {
+      "cosine_similarity": 0.10417749146457049,
+      "epoch": 1.9030754892823858,
+      "grad_norm": 1.1741375647090202,
+      "learning_rate": 8.125647221263377e-06,
+      "loss": 1.8548,
+      "step": 2042
+    },
+    {
+      "cosine_similarity": 0.11750028651333003,
+      "epoch": 1.9040074557315938,
+      "grad_norm": 1.0489704842562593,
+      "learning_rate": 8.118743527787368e-06,
+      "loss": 1.4905,
+      "step": 2043
+    },
+    {
+      "cosine_similarity": 0.0507101554171141,
+      "epoch": 1.9049394221808016,
+      "grad_norm": 1.219054376147004,
+      "learning_rate": 8.111839834311357e-06,
+      "loss": 1.9877,
+      "step": 2044
+    },
+    {
+      "cosine_similarity": 0.1090890724245039,
+      "epoch": 1.9058713886300094,
+      "grad_norm": 1.386972297955794,
+      "learning_rate": 8.104936140835348e-06,
+      "loss": 1.7893,
+      "step": 2045
+    },
+    {
+      "cosine_similarity": 0.0604824285199088,
+      "epoch": 1.9068033550792172,
+      "grad_norm": 1.1990085234778554,
+      "learning_rate": 8.098032447359337e-06,
+      "loss": 1.7253,
+      "step": 2046
+    },
+    {
+      "cosine_similarity": 0.07714061599556468,
+      "epoch": 1.907735321528425,
+      "grad_norm": 1.1253853560853426,
+      "learning_rate": 8.091128753883329e-06,
+      "loss": 1.5828,
+      "step": 2047
+    },
+    {
+      "cosine_similarity": -0.09378239195499985,
+      "epoch": 1.9086672879776327,
+      "grad_norm": 1.4485265217989411,
+      "learning_rate": 8.084225060407318e-06,
+      "loss": 1.8674,
+      "step": 2048
+    },
+    {
+      "cosine_similarity": 0.09533948534944775,
+      "epoch": 1.9095992544268405,
+      "grad_norm": 1.1350783769991653,
+      "learning_rate": 8.077321366931309e-06,
+      "loss": 1.8692,
+      "step": 2049
+    },
+    {
+      "cosine_similarity": 0.27587965119244223,
+      "epoch": 1.9105312208760483,
+      "grad_norm": 1.2427900312311388,
+      "learning_rate": 8.0704176734553e-06,
+      "loss": 1.6673,
+      "step": 2050
+    },
+    {
+      "cosine_similarity": 0.0655879494665749,
+      "epoch": 1.9114631873252563,
+      "grad_norm": 1.2208413836895762,
+      "learning_rate": 8.063513979979289e-06,
+      "loss": 1.8442,
+      "step": 2051
+    },
+    {
+      "cosine_similarity": 0.04671951457474431,
+      "epoch": 1.9123951537744641,
+      "grad_norm": 1.3259707262349842,
+      "learning_rate": 8.056610286503281e-06,
+      "loss": 1.6948,
+      "step": 2052
+    },
+    {
+      "cosine_similarity": 0.2916053870519502,
+      "epoch": 1.913327120223672,
+      "grad_norm": 1.3383915759868144,
+      "learning_rate": 8.04970659302727e-06,
+      "loss": 1.6453,
+      "step": 2053
+    },
+    {
+      "cosine_similarity": 0.13851416855970883,
+      "epoch": 1.91425908667288,
+      "grad_norm": 1.1971570539000127,
+      "learning_rate": 8.04280289955126e-06,
+      "loss": 1.6445,
+      "step": 2054
+    },
+    {
+      "cosine_similarity": -0.041645425889921504,
+      "epoch": 1.9151910531220877,
+      "grad_norm": 1.1531113974767788,
+      "learning_rate": 8.03589920607525e-06,
+      "loss": 1.8706,
+      "step": 2055
+    },
+    {
+      "cosine_similarity": 0.05553364436108187,
+      "epoch": 1.9161230195712955,
+      "grad_norm": 0.9550344708392858,
+      "learning_rate": 8.02899551259924e-06,
+      "loss": 1.5673,
+      "step": 2056
+    },
+    {
+      "cosine_similarity": 0.10019332670849512,
+      "epoch": 1.9170549860205033,
+      "grad_norm": 1.1999629008010755,
+      "learning_rate": 8.022091819123231e-06,
+      "loss": 1.9693,
+      "step": 2057
+    },
+    {
+      "cosine_similarity": 0.025141418989946977,
+      "epoch": 1.917986952469711,
+      "grad_norm": 1.1622819136529525,
+      "learning_rate": 8.015188125647222e-06,
+      "loss": 1.9919,
+      "step": 2058
+    },
+    {
+      "cosine_similarity": 0.08185134324446959,
+      "epoch": 1.9189189189189189,
+      "grad_norm": 1.2133100632801537,
+      "learning_rate": 8.008284432171213e-06,
+      "loss": 1.8086,
+      "step": 2059
+    },
+    {
+      "cosine_similarity": 0.061821593697635825,
+      "epoch": 1.9198508853681266,
+      "grad_norm": 1.4785700877952201,
+      "learning_rate": 8.001380738695202e-06,
+      "loss": 1.6421,
+      "step": 2060
+    },
+    {
+      "cosine_similarity": 0.01700299793459288,
+      "epoch": 1.9207828518173344,
+      "grad_norm": 1.4109195788176765,
+      "learning_rate": 7.994477045219193e-06,
+      "loss": 1.87,
+      "step": 2061
+    },
+    {
+      "cosine_similarity": 0.14510169495642974,
+      "epoch": 1.9217148182665424,
+      "grad_norm": 0.9987345233418284,
+      "learning_rate": 7.987573351743183e-06,
+      "loss": 1.7129,
+      "step": 2062
+    },
+    {
+      "cosine_similarity": 0.23219028875767697,
+      "epoch": 1.9226467847157502,
+      "grad_norm": 1.2673082229455028,
+      "learning_rate": 7.980669658267174e-06,
+      "loss": 1.8255,
+      "step": 2063
+    },
+    {
+      "cosine_similarity": 0.25264030316148983,
+      "epoch": 1.923578751164958,
+      "grad_norm": 1.0385465023988363,
+      "learning_rate": 7.973765964791163e-06,
+      "loss": 1.6449,
+      "step": 2064
+    },
+    {
+      "cosine_similarity": 0.01077370582853015,
+      "epoch": 1.924510717614166,
+      "grad_norm": 1.2153355764129674,
+      "learning_rate": 7.966862271315154e-06,
+      "loss": 1.72,
+      "step": 2065
+    },
+    {
+      "cosine_similarity": 0.02313248112537588,
+      "epoch": 1.9254426840633738,
+      "grad_norm": 1.0563876888554777,
+      "learning_rate": 7.959958577839145e-06,
+      "loss": 1.3366,
+      "step": 2066
+    },
+    {
+      "cosine_similarity": 0.137647559057031,
+      "epoch": 1.9263746505125816,
+      "grad_norm": 1.3299549799980477,
+      "learning_rate": 7.953054884363135e-06,
+      "loss": 1.651,
+      "step": 2067
+    },
+    {
+      "cosine_similarity": 0.17680482332249675,
+      "epoch": 1.9273066169617894,
+      "grad_norm": 1.1924923745575782,
+      "learning_rate": 7.946151190887124e-06,
+      "loss": 1.7852,
+      "step": 2068
+    },
+    {
+      "cosine_similarity": 0.24586876819836626,
+      "epoch": 1.9282385834109972,
+      "grad_norm": 1.019981039575043,
+      "learning_rate": 7.939247497411115e-06,
+      "loss": 1.5131,
+      "step": 2069
+    },
+    {
+      "cosine_similarity": -0.1868710588862423,
+      "epoch": 1.929170549860205,
+      "grad_norm": 1.0916567273701352,
+      "learning_rate": 7.932343803935106e-06,
+      "loss": 1.7899,
+      "step": 2070
+    },
+    {
+      "cosine_similarity": -0.2355774548322509,
+      "epoch": 1.9301025163094128,
+      "grad_norm": 1.1371763252059903,
+      "learning_rate": 7.925440110459096e-06,
+      "loss": 1.761,
+      "step": 2071
+    },
+    {
+      "cosine_similarity": 0.046022957349669644,
+      "epoch": 1.9310344827586206,
+      "grad_norm": 1.2297679499977552,
+      "learning_rate": 7.918536416983087e-06,
+      "loss": 1.7448,
+      "step": 2072
+    },
+    {
+      "cosine_similarity": 0.6132927919554544,
+      "epoch": 1.9319664492078286,
+      "grad_norm": 1.351041493253116,
+      "learning_rate": 7.911632723507076e-06,
+      "loss": 1.6347,
+      "step": 2073
+    },
+    {
+      "cosine_similarity": 0.33758497343999805,
+      "epoch": 1.9328984156570364,
+      "grad_norm": 1.296584546049266,
+      "learning_rate": 7.904729030031067e-06,
+      "loss": 1.5464,
+      "step": 2074
+    },
+    {
+      "cosine_similarity": 0.18578826511187568,
+      "epoch": 1.9338303821062441,
+      "grad_norm": 1.137157551538958,
+      "learning_rate": 7.897825336555058e-06,
+      "loss": 1.7257,
+      "step": 2075
+    },
+    {
+      "cosine_similarity": 0.06680003157738013,
+      "epoch": 1.9347623485554521,
+      "grad_norm": 1.1861493431175667,
+      "learning_rate": 7.890921643079048e-06,
+      "loss": 1.5577,
+      "step": 2076
+    },
+    {
+      "cosine_similarity": -0.13455661164520105,
+      "epoch": 1.93569431500466,
+      "grad_norm": 0.9909226712920376,
+      "learning_rate": 7.884017949603038e-06,
+      "loss": 1.4428,
+      "step": 2077
+    },
+    {
+      "cosine_similarity": 0.05247877979242345,
+      "epoch": 1.9366262814538677,
+      "grad_norm": 1.0949121989570185,
+      "learning_rate": 7.877114256127028e-06,
+      "loss": 1.8106,
+      "step": 2078
+    },
+    {
+      "cosine_similarity": 0.10244967564496832,
+      "epoch": 1.9375582479030755,
+      "grad_norm": 1.2408684605325646,
+      "learning_rate": 7.870210562651019e-06,
+      "loss": 1.7261,
+      "step": 2079
+    },
+    {
+      "cosine_similarity": 0.048682452364403654,
+      "epoch": 1.9384902143522833,
+      "grad_norm": 1.1722715837117823,
+      "learning_rate": 7.86330686917501e-06,
+      "loss": 1.7243,
+      "step": 2080
+    },
+    {
+      "cosine_similarity": -0.07760833286140566,
+      "epoch": 1.939422180801491,
+      "grad_norm": 1.7028407416061038,
+      "learning_rate": 7.856403175699e-06,
+      "loss": 1.704,
+      "step": 2081
+    },
+    {
+      "cosine_similarity": 0.3339029199477617,
+      "epoch": 1.9403541472506989,
+      "grad_norm": 1.0606804350104126,
+      "learning_rate": 7.84949948222299e-06,
+      "loss": 1.7256,
+      "step": 2082
+    },
+    {
+      "cosine_similarity": 0.1634411221732838,
+      "epoch": 1.9412861136999067,
+      "grad_norm": 1.201334636149946,
+      "learning_rate": 7.84259578874698e-06,
+      "loss": 1.828,
+      "step": 2083
+    },
+    {
+      "cosine_similarity": 0.16487523600738688,
+      "epoch": 1.9422180801491147,
+      "grad_norm": 1.106835625461138,
+      "learning_rate": 7.835692095270971e-06,
+      "loss": 1.4159,
+      "step": 2084
+    },
+    {
+      "cosine_similarity": 0.09986269952514173,
+      "epoch": 1.9431500465983225,
+      "grad_norm": 1.087014615010923,
+      "learning_rate": 7.828788401794962e-06,
+      "loss": 1.8072,
+      "step": 2085
+    },
+    {
+      "cosine_similarity": 0.0652338716089996,
+      "epoch": 1.9440820130475303,
+      "grad_norm": 1.1833413925365917,
+      "learning_rate": 7.82188470831895e-06,
+      "loss": 1.5688,
+      "step": 2086
+    },
+    {
+      "cosine_similarity": 0.13351126057691126,
+      "epoch": 1.9450139794967383,
+      "grad_norm": 1.5165757571500962,
+      "learning_rate": 7.814981014842941e-06,
+      "loss": 2.189,
+      "step": 2087
+    },
+    {
+      "cosine_similarity": 0.08427882533846663,
+      "epoch": 1.945945945945946,
+      "grad_norm": 1.1462323990355336,
+      "learning_rate": 7.808077321366932e-06,
+      "loss": 1.8342,
+      "step": 2088
+    },
+    {
+      "cosine_similarity": 0.03821674217815995,
+      "epoch": 1.9468779123951538,
+      "grad_norm": 1.2324434161116442,
+      "learning_rate": 7.801173627890923e-06,
+      "loss": 1.782,
+      "step": 2089
+    },
+    {
+      "cosine_similarity": 0.41237095229746673,
+      "epoch": 1.9478098788443616,
+      "grad_norm": 1.2820599453462547,
+      "learning_rate": 7.794269934414914e-06,
+      "loss": 1.649,
+      "step": 2090
+    },
+    {
+      "cosine_similarity": 0.259916972696634,
+      "epoch": 1.9487418452935694,
+      "grad_norm": 0.990961082014108,
+      "learning_rate": 7.787366240938903e-06,
+      "loss": 1.671,
+      "step": 2091
+    },
+    {
+      "cosine_similarity": -0.03979251016659043,
+      "epoch": 1.9496738117427772,
+      "grad_norm": 1.0900146963479682,
+      "learning_rate": 7.780462547462893e-06,
+      "loss": 1.5256,
+      "step": 2092
+    },
+    {
+      "cosine_similarity": -0.033852969776801696,
+      "epoch": 1.950605778191985,
+      "grad_norm": 1.3056358154331391,
+      "learning_rate": 7.773558853986884e-06,
+      "loss": 1.9472,
+      "step": 2093
+    },
+    {
+      "cosine_similarity": 0.4009840888285964,
+      "epoch": 1.9515377446411928,
+      "grad_norm": 1.051412721003859,
+      "learning_rate": 7.766655160510875e-06,
+      "loss": 1.6831,
+      "step": 2094
+    },
+    {
+      "cosine_similarity": 0.2024245701528789,
+      "epoch": 1.9524697110904008,
+      "grad_norm": 1.1998790163420676,
+      "learning_rate": 7.759751467034864e-06,
+      "loss": 1.5645,
+      "step": 2095
+    },
+    {
+      "cosine_similarity": 0.05681129732756744,
+      "epoch": 1.9534016775396086,
+      "grad_norm": 1.1425954562897638,
+      "learning_rate": 7.752847773558855e-06,
+      "loss": 1.7221,
+      "step": 2096
+    },
+    {
+      "cosine_similarity": 0.1299979170844386,
+      "epoch": 1.9543336439888164,
+      "grad_norm": 0.9877584861407736,
+      "learning_rate": 7.745944080082844e-06,
+      "loss": 1.6362,
+      "step": 2097
+    },
+    {
+      "cosine_similarity": 0.09530600189030809,
+      "epoch": 1.9552656104380244,
+      "grad_norm": 1.207138125650716,
+      "learning_rate": 7.739040386606836e-06,
+      "loss": 1.7862,
+      "step": 2098
+    },
+    {
+      "cosine_similarity": 0.08068478815377038,
+      "epoch": 1.9561975768872322,
+      "grad_norm": 1.0546053660423083,
+      "learning_rate": 7.732136693130827e-06,
+      "loss": 1.5435,
+      "step": 2099
+    },
+    {
+      "cosine_similarity": -0.026969148210197635,
+      "epoch": 1.95712954333644,
+      "grad_norm": 1.162358100985284,
+      "learning_rate": 7.725232999654816e-06,
+      "loss": 1.8268,
+      "step": 2100
+    },
+    {
+      "cosine_similarity": 0.04264017591812327,
+      "epoch": 1.9580615097856477,
+      "grad_norm": 1.158611532076786,
+      "learning_rate": 7.718329306178807e-06,
+      "loss": 1.8294,
+      "step": 2101
+    },
+    {
+      "cosine_similarity": 0.1965165217846064,
+      "epoch": 1.9589934762348555,
+      "grad_norm": 1.3187467570595102,
+      "learning_rate": 7.711425612702796e-06,
+      "loss": 1.4685,
+      "step": 2102
+    },
+    {
+      "cosine_similarity": 0.15443059019985964,
+      "epoch": 1.9599254426840633,
+      "grad_norm": 1.08256253202545,
+      "learning_rate": 7.704521919226788e-06,
+      "loss": 1.9092,
+      "step": 2103
+    },
+    {
+      "cosine_similarity": 0.17816718841600623,
+      "epoch": 1.9608574091332711,
+      "grad_norm": 1.1441682543079759,
+      "learning_rate": 7.697618225750777e-06,
+      "loss": 1.9077,
+      "step": 2104
+    },
+    {
+      "cosine_similarity": 0.0792493288139221,
+      "epoch": 1.961789375582479,
+      "grad_norm": 1.2479024930852207,
+      "learning_rate": 7.690714532274768e-06,
+      "loss": 1.93,
+      "step": 2105
+    },
+    {
+      "cosine_similarity": 0.257115592177867,
+      "epoch": 1.962721342031687,
+      "grad_norm": 1.1460624172332272,
+      "learning_rate": 7.683810838798757e-06,
+      "loss": 1.6149,
+      "step": 2106
+    },
+    {
+      "cosine_similarity": 0.14338478401846477,
+      "epoch": 1.9636533084808947,
+      "grad_norm": 1.1411209946763758,
+      "learning_rate": 7.676907145322748e-06,
+      "loss": 1.6319,
+      "step": 2107
+    },
+    {
+      "cosine_similarity": 0.07920319143018915,
+      "epoch": 1.9645852749301025,
+      "grad_norm": 1.1312917150890371,
+      "learning_rate": 7.670003451846738e-06,
+      "loss": 2.0526,
+      "step": 2108
+    },
+    {
+      "cosine_similarity": 0.09605301080380396,
+      "epoch": 1.9655172413793105,
+      "grad_norm": 1.0961978947907527,
+      "learning_rate": 7.663099758370729e-06,
+      "loss": 1.5648,
+      "step": 2109
+    },
+    {
+      "cosine_similarity": 0.07367673595167594,
+      "epoch": 1.9664492078285183,
+      "grad_norm": 1.4603410988779288,
+      "learning_rate": 7.65619606489472e-06,
+      "loss": 1.6378,
+      "step": 2110
+    },
+    {
+      "cosine_similarity": 0.5717661026690888,
+      "epoch": 1.967381174277726,
+      "grad_norm": 1.1825265687034325,
+      "learning_rate": 7.649292371418709e-06,
+      "loss": 1.7192,
+      "step": 2111
+    },
+    {
+      "cosine_similarity": 0.03902460896143627,
+      "epoch": 1.9683131407269339,
+      "grad_norm": 1.2840126024016443,
+      "learning_rate": 7.6423886779427e-06,
+      "loss": 1.9031,
+      "step": 2112
+    },
+    {
+      "cosine_similarity": 0.17580772507293013,
+      "epoch": 1.9692451071761417,
+      "grad_norm": 1.3549801989601864,
+      "learning_rate": 7.63548498446669e-06,
+      "loss": 1.8065,
+      "step": 2113
+    },
+    {
+      "cosine_similarity": -0.0055084868526384,
+      "epoch": 1.9701770736253494,
+      "grad_norm": 1.167647331393277,
+      "learning_rate": 7.628581290990681e-06,
+      "loss": 1.5786,
+      "step": 2114
+    },
+    {
+      "cosine_similarity": -0.10207665301712708,
+      "epoch": 1.9711090400745572,
+      "grad_norm": 1.0350727619817703,
+      "learning_rate": 7.621677597514671e-06,
+      "loss": 1.5923,
+      "step": 2115
+    },
+    {
+      "cosine_similarity": 0.33271702577709106,
+      "epoch": 1.972041006523765,
+      "grad_norm": 1.2085507829324191,
+      "learning_rate": 7.614773904038662e-06,
+      "loss": 1.7873,
+      "step": 2116
+    },
+    {
+      "cosine_similarity": -0.1700351814652679,
+      "epoch": 1.972972972972973,
+      "grad_norm": 1.1819803193459688,
+      "learning_rate": 7.6078702105626516e-06,
+      "loss": 1.4775,
+      "step": 2117
+    },
+    {
+      "cosine_similarity": 0.0652625527216822,
+      "epoch": 1.9739049394221808,
+      "grad_norm": 1.2471331947863977,
+      "learning_rate": 7.600966517086642e-06,
+      "loss": 1.56,
+      "step": 2118
+    },
+    {
+      "cosine_similarity": 0.14370040968914802,
+      "epoch": 1.9748369058713886,
+      "grad_norm": 1.2820551144088108,
+      "learning_rate": 7.594062823610633e-06,
+      "loss": 1.8962,
+      "step": 2119
+    },
+    {
+      "cosine_similarity": 0.03505454266283616,
+      "epoch": 1.9757688723205966,
+      "grad_norm": 1.195520047138079,
+      "learning_rate": 7.587159130134623e-06,
+      "loss": 1.7318,
+      "step": 2120
+    },
+    {
+      "cosine_similarity": -0.00799871888748129,
+      "epoch": 1.9767008387698044,
+      "grad_norm": 1.471707019614902,
+      "learning_rate": 7.580255436658614e-06,
+      "loss": 1.7453,
+      "step": 2121
+    },
+    {
+      "cosine_similarity": 0.30817612797162125,
+      "epoch": 1.9776328052190122,
+      "grad_norm": 1.373181353506156,
+      "learning_rate": 7.5733517431826035e-06,
+      "loss": 1.7899,
+      "step": 2122
+    },
+    {
+      "cosine_similarity": 0.08106377866937986,
+      "epoch": 1.97856477166822,
+      "grad_norm": 1.1987062179277355,
+      "learning_rate": 7.566448049706594e-06,
+      "loss": 1.9073,
+      "step": 2123
+    },
+    {
+      "cosine_similarity": -0.09785617604417858,
+      "epoch": 1.9794967381174278,
+      "grad_norm": 1.3224391947589986,
+      "learning_rate": 7.559544356230583e-06,
+      "loss": 1.7794,
+      "step": 2124
+    },
+    {
+      "cosine_similarity": -0.10677385712369568,
+      "epoch": 1.9804287045666356,
+      "grad_norm": 1.2700980238291752,
+      "learning_rate": 7.552640662754575e-06,
+      "loss": 1.965,
+      "step": 2125
+    },
+    {
+      "cosine_similarity": 0.3630660278522065,
+      "epoch": 1.9813606710158433,
+      "grad_norm": 1.1326779808284666,
+      "learning_rate": 7.545736969278564e-06,
+      "loss": 1.7886,
+      "step": 2126
+    },
+    {
+      "cosine_similarity": 0.14936041522020205,
+      "epoch": 1.9822926374650511,
+      "grad_norm": 1.368182481034118,
+      "learning_rate": 7.5388332758025555e-06,
+      "loss": 1.7787,
+      "step": 2127
+    },
+    {
+      "cosine_similarity": 0.08302351358324385,
+      "epoch": 1.983224603914259,
+      "grad_norm": 1.0638111598821416,
+      "learning_rate": 7.5319295823265445e-06,
+      "loss": 2.0079,
+      "step": 2128
+    },
+    {
+      "cosine_similarity": -0.0754131131453911,
+      "epoch": 1.984156570363467,
+      "grad_norm": 1.0484023335021238,
+      "learning_rate": 7.525025888850535e-06,
+      "loss": 1.7858,
+      "step": 2129
+    },
+    {
+      "cosine_similarity": -0.03626363885410668,
+      "epoch": 1.9850885368126747,
+      "grad_norm": 1.0460878724199563,
+      "learning_rate": 7.518122195374527e-06,
+      "loss": 1.7607,
+      "step": 2130
+    },
+    {
+      "cosine_similarity": 0.12453670250521798,
+      "epoch": 1.9860205032618827,
+      "grad_norm": 1.0289714467810296,
+      "learning_rate": 7.511218501898516e-06,
+      "loss": 1.4967,
+      "step": 2131
+    },
+    {
+      "cosine_similarity": 0.1436710292706813,
+      "epoch": 1.9869524697110905,
+      "grad_norm": 1.1604920661506328,
+      "learning_rate": 7.504314808422507e-06,
+      "loss": 1.7029,
+      "step": 2132
+    },
+    {
+      "cosine_similarity": 0.020996443085865745,
+      "epoch": 1.9878844361602983,
+      "grad_norm": 1.0781943241362826,
+      "learning_rate": 7.4974111149464965e-06,
+      "loss": 1.5953,
+      "step": 2133
+    },
+    {
+      "cosine_similarity": 0.005053190468316761,
+      "epoch": 1.988816402609506,
+      "grad_norm": 1.1061262842723447,
+      "learning_rate": 7.490507421470487e-06,
+      "loss": 1.6089,
+      "step": 2134
+    },
+    {
+      "cosine_similarity": 0.07675084112191378,
+      "epoch": 1.9897483690587139,
+      "grad_norm": 1.0833698853921265,
+      "learning_rate": 7.483603727994477e-06,
+      "loss": 1.5896,
+      "step": 2135
+    },
+    {
+      "cosine_similarity": 0.018252427498844245,
+      "epoch": 1.9906803355079217,
+      "grad_norm": 1.1320501133543164,
+      "learning_rate": 7.476700034518468e-06,
+      "loss": 1.7248,
+      "step": 2136
+    },
+    {
+      "cosine_similarity": -0.28779993168396106,
+      "epoch": 1.9916123019571295,
+      "grad_norm": 1.6063556729274904,
+      "learning_rate": 7.469796341042458e-06,
+      "loss": 1.616,
+      "step": 2137
+    },
+    {
+      "cosine_similarity": 0.16584510202689343,
+      "epoch": 1.9925442684063372,
+      "grad_norm": 0.985905572106495,
+      "learning_rate": 7.4628926475664485e-06,
+      "loss": 1.642,
+      "step": 2138
+    },
+    {
+      "cosine_similarity": 0.1718402899250609,
+      "epoch": 1.993476234855545,
+      "grad_norm": 1.093365657869586,
+      "learning_rate": 7.455988954090439e-06,
+      "loss": 1.6873,
+      "step": 2139
+    },
+    {
+      "cosine_similarity": -0.029576062178959337,
+      "epoch": 1.994408201304753,
+      "grad_norm": 1.032145943975548,
+      "learning_rate": 7.449085260614429e-06,
+      "loss": 1.6142,
+      "step": 2140
+    },
+    {
+      "cosine_similarity": 0.13713399296265083,
+      "epoch": 1.9953401677539608,
+      "grad_norm": 1.175821378612112,
+      "learning_rate": 7.44218156713842e-06,
+      "loss": 1.684,
+      "step": 2141
+    },
+    {
+      "cosine_similarity": -0.0923807336070711,
+      "epoch": 1.9962721342031688,
+      "grad_norm": 1.0443269638407784,
+      "learning_rate": 7.43527787366241e-06,
+      "loss": 1.644,
+      "step": 2142
+    },
+    {
+      "cosine_similarity": 0.04315918716549145,
+      "epoch": 1.9972041006523766,
+      "grad_norm": 1.2718854392106378,
+      "learning_rate": 7.4283741801864004e-06,
+      "loss": 1.6779,
+      "step": 2143
+    },
+    {
+      "cosine_similarity": 0.12473145432512177,
+      "epoch": 1.9981360671015844,
+      "grad_norm": 1.0840786750768499,
+      "learning_rate": 7.42147048671039e-06,
+      "loss": 1.3887,
+      "step": 2144
+    },
+    {
+      "cosine_similarity": 0.28127007913129587,
+      "epoch": 1.9990680335507922,
+      "grad_norm": 1.1508185277459004,
+      "learning_rate": 7.414566793234381e-06,
+      "loss": 1.4449,
+      "step": 2145
+    },
+    {
+      "cosine_similarity": 0.012989812963368503,
+      "epoch": 2.0,
+      "grad_norm": 1.1436957504110825,
+      "learning_rate": 7.407663099758371e-06,
+      "loss": 1.4584,
+      "step": 2146
+    },
+    {
+      "cosine_similarity": 0.3223960851231341,
+      "epoch": 2.000931966449208,
+      "grad_norm": 1.0307138173571466,
+      "learning_rate": 7.400759406282362e-06,
+      "loss": 1.8527,
+      "step": 2147
+    },
+    {
+      "cosine_similarity": 0.19273607207881338,
+      "epoch": 2.0018639328984156,
+      "grad_norm": 1.2840480536104246,
+      "learning_rate": 7.3938557128063516e-06,
+      "loss": 1.4986,
+      "step": 2148
+    },
+    {
+      "cosine_similarity": -0.038226261822738805,
+      "epoch": 2.0027958993476234,
+      "grad_norm": 1.0976690801160378,
+      "learning_rate": 7.386952019330342e-06,
+      "loss": 1.5647,
+      "step": 2149
+    },
+    {
+      "cosine_similarity": -0.060118513869865206,
+      "epoch": 2.003727865796831,
+      "grad_norm": 0.8817466948605353,
+      "learning_rate": 7.380048325854333e-06,
+      "loss": 1.4473,
+      "step": 2150
+    },
+    {
+      "cosine_similarity": -0.1444841678264523,
+      "epoch": 2.004659832246039,
+      "grad_norm": 1.0931358460169815,
+      "learning_rate": 7.373144632378323e-06,
+      "loss": 1.4896,
+      "step": 2151
+    },
+    {
+      "cosine_similarity": -0.022210070866622515,
+      "epoch": 2.005591798695247,
+      "grad_norm": 1.124810838229824,
+      "learning_rate": 7.366240938902314e-06,
+      "loss": 1.246,
+      "step": 2152
+    },
+    {
+      "cosine_similarity": 0.039749881997162236,
+      "epoch": 2.006523765144455,
+      "grad_norm": 1.0313076256411702,
+      "learning_rate": 7.3593372454263035e-06,
+      "loss": 1.3539,
+      "step": 2153
+    },
+    {
+      "cosine_similarity": 0.09523275702365049,
+      "epoch": 2.0074557315936628,
+      "grad_norm": 0.8925428950130772,
+      "learning_rate": 7.352433551950294e-06,
+      "loss": 1.2259,
+      "step": 2154
+    },
+    {
+      "cosine_similarity": 0.1583320529648728,
+      "epoch": 2.0083876980428705,
+      "grad_norm": 1.0387623765417786,
+      "learning_rate": 7.345529858474284e-06,
+      "loss": 1.45,
+      "step": 2155
+    },
+    {
+      "cosine_similarity": 0.29488446078533537,
+      "epoch": 2.0093196644920783,
+      "grad_norm": 1.4489971936592168,
+      "learning_rate": 7.338626164998275e-06,
+      "loss": 1.5628,
+      "step": 2156
+    },
+    {
+      "cosine_similarity": 0.01745565335002933,
+      "epoch": 2.010251630941286,
+      "grad_norm": 1.2965555515846905,
+      "learning_rate": 7.331722471522265e-06,
+      "loss": 1.4594,
+      "step": 2157
+    },
+    {
+      "cosine_similarity": 0.06376894815290556,
+      "epoch": 2.011183597390494,
+      "grad_norm": 1.0755681409414193,
+      "learning_rate": 7.3248187780462555e-06,
+      "loss": 1.4386,
+      "step": 2158
+    },
+    {
+      "cosine_similarity": 0.15821139517300364,
+      "epoch": 2.0121155638397017,
+      "grad_norm": 1.210476592657748,
+      "learning_rate": 7.317915084570245e-06,
+      "loss": 1.4435,
+      "step": 2159
+    },
+    {
+      "cosine_similarity": -0.0047552836672117745,
+      "epoch": 2.0130475302889095,
+      "grad_norm": 1.1859321490209827,
+      "learning_rate": 7.311011391094236e-06,
+      "loss": 1.3252,
+      "step": 2160
+    },
+    {
+      "cosine_similarity": 0.25154638156002207,
+      "epoch": 2.0139794967381173,
+      "grad_norm": 1.1135538224331532,
+      "learning_rate": 7.304107697618227e-06,
+      "loss": 1.6279,
+      "step": 2161
+    },
+    {
+      "cosine_similarity": 0.22273825504073783,
+      "epoch": 2.014911463187325,
+      "grad_norm": 1.2811591001161682,
+      "learning_rate": 7.297204004142217e-06,
+      "loss": 1.1481,
+      "step": 2162
+    },
+    {
+      "cosine_similarity": 0.006810736056699481,
+      "epoch": 2.0158434296365333,
+      "grad_norm": 1.0811845667961149,
+      "learning_rate": 7.2903003106662075e-06,
+      "loss": 1.6483,
+      "step": 2163
+    },
+    {
+      "cosine_similarity": -0.03130640440760994,
+      "epoch": 2.016775396085741,
+      "grad_norm": 1.2619110183673536,
+      "learning_rate": 7.283396617190197e-06,
+      "loss": 1.7333,
+      "step": 2164
+    },
+    {
+      "cosine_similarity": 0.16427585634793646,
+      "epoch": 2.017707362534949,
+      "grad_norm": 1.0333298630092695,
+      "learning_rate": 7.276492923714188e-06,
+      "loss": 1.362,
+      "step": 2165
+    },
+    {
+      "cosine_similarity": 0.07198869964798663,
+      "epoch": 2.0186393289841567,
+      "grad_norm": 1.1420758097821069,
+      "learning_rate": 7.269589230238178e-06,
+      "loss": 1.3586,
+      "step": 2166
+    },
+    {
+      "cosine_similarity": 0.30531493438445806,
+      "epoch": 2.0195712954333644,
+      "grad_norm": 1.1057975644657185,
+      "learning_rate": 7.262685536762169e-06,
+      "loss": 1.9273,
+      "step": 2167
+    },
+    {
+      "cosine_similarity": -0.1235161412274785,
+      "epoch": 2.0205032618825722,
+      "grad_norm": 1.0811350687836065,
+      "learning_rate": 7.255781843286159e-06,
+      "loss": 1.4306,
+      "step": 2168
+    },
+    {
+      "cosine_similarity": 0.1555109004271584,
+      "epoch": 2.02143522833178,
+      "grad_norm": 1.0859257054876141,
+      "learning_rate": 7.248878149810149e-06,
+      "loss": 1.4211,
+      "step": 2169
+    },
+    {
+      "cosine_similarity": 0.14320062063556566,
+      "epoch": 2.022367194780988,
+      "grad_norm": 1.4123815104185136,
+      "learning_rate": 7.24197445633414e-06,
+      "loss": 1.7051,
+      "step": 2170
+    },
+    {
+      "cosine_similarity": 0.161324685300927,
+      "epoch": 2.0232991612301956,
+      "grad_norm": 1.1384686903478474,
+      "learning_rate": 7.23507076285813e-06,
+      "loss": 1.4378,
+      "step": 2171
+    },
+    {
+      "cosine_similarity": 0.1412052605554306,
+      "epoch": 2.0242311276794034,
+      "grad_norm": 1.0561967058029826,
+      "learning_rate": 7.228167069382121e-06,
+      "loss": 1.447,
+      "step": 2172
+    },
+    {
+      "cosine_similarity": 0.12796438106616098,
+      "epoch": 2.025163094128611,
+      "grad_norm": 0.8916545468141477,
+      "learning_rate": 7.2212633759061106e-06,
+      "loss": 1.3776,
+      "step": 2173
+    },
+    {
+      "cosine_similarity": 0.30436072833801947,
+      "epoch": 2.0260950605778194,
+      "grad_norm": 1.1043404654518494,
+      "learning_rate": 7.214359682430101e-06,
+      "loss": 1.3658,
+      "step": 2174
+    },
+    {
+      "cosine_similarity": -0.0867103595534847,
+      "epoch": 2.027027027027027,
+      "grad_norm": 1.243466622933385,
+      "learning_rate": 7.20745598895409e-06,
+      "loss": 1.8291,
+      "step": 2175
+    },
+    {
+      "cosine_similarity": -0.02167963686869609,
+      "epoch": 2.027958993476235,
+      "grad_norm": 1.2073886654625614,
+      "learning_rate": 7.200552295478082e-06,
+      "loss": 1.4364,
+      "step": 2176
+    },
+    {
+      "cosine_similarity": -0.061677790192318664,
+      "epoch": 2.0288909599254428,
+      "grad_norm": 1.1899001034770742,
+      "learning_rate": 7.193648602002071e-06,
+      "loss": 1.6948,
+      "step": 2177
+    },
+    {
+      "cosine_similarity": -0.3413984610771727,
+      "epoch": 2.0298229263746506,
+      "grad_norm": 1.428755980293876,
+      "learning_rate": 7.1867449085260625e-06,
+      "loss": 1.4937,
+      "step": 2178
+    },
+    {
+      "cosine_similarity": -0.08988269664680544,
+      "epoch": 2.0307548928238583,
+      "grad_norm": 1.2654372979186375,
+      "learning_rate": 7.1798412150500516e-06,
+      "loss": 1.497,
+      "step": 2179
+    },
+    {
+      "cosine_similarity": 0.21556594460742678,
+      "epoch": 2.031686859273066,
+      "grad_norm": 0.917988257119976,
+      "learning_rate": 7.172937521574042e-06,
+      "loss": 1.2655,
+      "step": 2180
+    },
+    {
+      "cosine_similarity": 0.31167609417132225,
+      "epoch": 2.032618825722274,
+      "grad_norm": 1.090932338916807,
+      "learning_rate": 7.166033828098034e-06,
+      "loss": 1.4213,
+      "step": 2181
+    },
+    {
+      "cosine_similarity": 0.12042568990819853,
+      "epoch": 2.0335507921714817,
+      "grad_norm": 1.0260471980374855,
+      "learning_rate": 7.159130134622023e-06,
+      "loss": 1.1079,
+      "step": 2182
+    },
+    {
+      "cosine_similarity": -0.2776995924067827,
+      "epoch": 2.0344827586206895,
+      "grad_norm": 1.4321160019588635,
+      "learning_rate": 7.152226441146014e-06,
+      "loss": 1.7306,
+      "step": 2183
+    },
+    {
+      "cosine_similarity": 0.1692702487474962,
+      "epoch": 2.0354147250698973,
+      "grad_norm": 1.8247746788447912,
+      "learning_rate": 7.1453227476700035e-06,
+      "loss": 1.6478,
+      "step": 2184
+    },
+    {
+      "cosine_similarity": 0.2557933979003344,
+      "epoch": 2.0363466915191055,
+      "grad_norm": 1.2612060180253089,
+      "learning_rate": 7.138419054193994e-06,
+      "loss": 1.2788,
+      "step": 2185
+    },
+    {
+      "cosine_similarity": -0.04026678773597394,
+      "epoch": 2.0372786579683133,
+      "grad_norm": 1.261490721883179,
+      "learning_rate": 7.131515360717984e-06,
+      "loss": 1.3774,
+      "step": 2186
+    },
+    {
+      "cosine_similarity": 0.2629945980435487,
+      "epoch": 2.038210624417521,
+      "grad_norm": 1.108860649638962,
+      "learning_rate": 7.124611667241975e-06,
+      "loss": 1.6074,
+      "step": 2187
+    },
+    {
+      "cosine_similarity": 0.18685837652535173,
+      "epoch": 2.039142590866729,
+      "grad_norm": 1.2383175630036474,
+      "learning_rate": 7.117707973765965e-06,
+      "loss": 1.64,
+      "step": 2188
+    },
+    {
+      "cosine_similarity": 0.16689392151134058,
+      "epoch": 2.0400745573159367,
+      "grad_norm": 0.9299084542185169,
+      "learning_rate": 7.1108042802899555e-06,
+      "loss": 1.2438,
+      "step": 2189
+    },
+    {
+      "cosine_similarity": 0.20717806578098613,
+      "epoch": 2.0410065237651445,
+      "grad_norm": 1.4021646496386664,
+      "learning_rate": 7.103900586813946e-06,
+      "loss": 1.6063,
+      "step": 2190
+    },
+    {
+      "cosine_similarity": 0.3322473282919032,
+      "epoch": 2.0419384902143523,
+      "grad_norm": 1.7213239522004085,
+      "learning_rate": 7.096996893337936e-06,
+      "loss": 1.5406,
+      "step": 2191
+    },
+    {
+      "cosine_similarity": 0.3815861354234591,
+      "epoch": 2.04287045666356,
+      "grad_norm": 1.1278013121019694,
+      "learning_rate": 7.090093199861927e-06,
+      "loss": 1.7902,
+      "step": 2192
+    },
+    {
+      "cosine_similarity": 0.14332043768119196,
+      "epoch": 2.043802423112768,
+      "grad_norm": 1.1370960573796132,
+      "learning_rate": 7.083189506385917e-06,
+      "loss": 1.37,
+      "step": 2193
+    },
+    {
+      "cosine_similarity": 0.36499551420998133,
+      "epoch": 2.0447343895619756,
+      "grad_norm": 1.0228748223923994,
+      "learning_rate": 7.0762858129099075e-06,
+      "loss": 1.6752,
+      "step": 2194
+    },
+    {
+      "cosine_similarity": 0.153917599413449,
+      "epoch": 2.0456663560111834,
+      "grad_norm": 1.218135709045946,
+      "learning_rate": 7.069382119433897e-06,
+      "loss": 1.7175,
+      "step": 2195
+    },
+    {
+      "cosine_similarity": 0.11381960008215103,
+      "epoch": 2.0465983224603916,
+      "grad_norm": 1.0463360749029464,
+      "learning_rate": 7.062478425957888e-06,
+      "loss": 1.3601,
+      "step": 2196
+    },
+    {
+      "cosine_similarity": 0.19984989189277924,
+      "epoch": 2.0475302889095994,
+      "grad_norm": 1.1096402554064986,
+      "learning_rate": 7.055574732481878e-06,
+      "loss": 1.4562,
+      "step": 2197
+    },
+    {
+      "cosine_similarity": 0.1993558288381241,
+      "epoch": 2.048462255358807,
+      "grad_norm": 1.0738728585036046,
+      "learning_rate": 7.048671039005869e-06,
+      "loss": 1.1768,
+      "step": 2198
+    },
+    {
+      "cosine_similarity": 0.03242293934507698,
+      "epoch": 2.049394221808015,
+      "grad_norm": 1.0871258626644695,
+      "learning_rate": 7.041767345529859e-06,
+      "loss": 1.4774,
+      "step": 2199
+    },
+    {
+      "cosine_similarity": 0.02906720023019253,
+      "epoch": 2.050326188257223,
+      "grad_norm": 1.27552870008273,
+      "learning_rate": 7.034863652053849e-06,
+      "loss": 1.8274,
+      "step": 2200
+    },
+    {
+      "cosine_similarity": 0.0890952433669254,
+      "epoch": 2.0512581547064306,
+      "grad_norm": 1.083505838985043,
+      "learning_rate": 7.02795995857784e-06,
+      "loss": 1.1378,
+      "step": 2201
+    },
+    {
+      "cosine_similarity": -0.27186706896073637,
+      "epoch": 2.0521901211556384,
+      "grad_norm": 1.306432556164151,
+      "learning_rate": 7.02105626510183e-06,
+      "loss": 1.5848,
+      "step": 2202
+    },
+    {
+      "cosine_similarity": 0.16407730685108976,
+      "epoch": 2.053122087604846,
+      "grad_norm": 1.0173014702401304,
+      "learning_rate": 7.014152571625821e-06,
+      "loss": 1.4152,
+      "step": 2203
+    },
+    {
+      "cosine_similarity": -0.04821930956349521,
+      "epoch": 2.054054054054054,
+      "grad_norm": 1.0502401805863373,
+      "learning_rate": 7.0072488781498106e-06,
+      "loss": 1.4104,
+      "step": 2204
+    },
+    {
+      "cosine_similarity": 0.061559061276295846,
+      "epoch": 2.0549860205032617,
+      "grad_norm": 1.3282471594518561,
+      "learning_rate": 7.000345184673801e-06,
+      "loss": 1.379,
+      "step": 2205
+    },
+    {
+      "cosine_similarity": 0.239471547482031,
+      "epoch": 2.0559179869524695,
+      "grad_norm": 1.03199603903598,
+      "learning_rate": 6.993441491197791e-06,
+      "loss": 1.4567,
+      "step": 2206
+    },
+    {
+      "cosine_similarity": 0.041367956296834735,
+      "epoch": 2.0568499534016778,
+      "grad_norm": 1.2959725276158138,
+      "learning_rate": 6.986537797721782e-06,
+      "loss": 1.5454,
+      "step": 2207
+    },
+    {
+      "cosine_similarity": 0.028377125912748248,
+      "epoch": 2.0577819198508855,
+      "grad_norm": 1.5965527947047649,
+      "learning_rate": 6.979634104245772e-06,
+      "loss": 1.5407,
+      "step": 2208
+    },
+    {
+      "cosine_similarity": 0.06141508742980482,
+      "epoch": 2.0587138863000933,
+      "grad_norm": 0.9591556846882718,
+      "learning_rate": 6.9727304107697625e-06,
+      "loss": 1.4163,
+      "step": 2209
+    },
+    {
+      "cosine_similarity": 0.10960455475023839,
+      "epoch": 2.059645852749301,
+      "grad_norm": 0.9124996516266343,
+      "learning_rate": 6.965826717293753e-06,
+      "loss": 1.5107,
+      "step": 2210
+    },
+    {
+      "cosine_similarity": 0.04178004563300782,
+      "epoch": 2.060577819198509,
+      "grad_norm": 1.5672253689489561,
+      "learning_rate": 6.958923023817743e-06,
+      "loss": 1.5277,
+      "step": 2211
+    },
+    {
+      "cosine_similarity": 0.15205914235176451,
+      "epoch": 2.0615097856477167,
+      "grad_norm": 0.938326389384352,
+      "learning_rate": 6.952019330341734e-06,
+      "loss": 1.2719,
+      "step": 2212
+    },
+    {
+      "cosine_similarity": -0.23152756091270585,
+      "epoch": 2.0624417520969245,
+      "grad_norm": 1.296629899659066,
+      "learning_rate": 6.945115636865724e-06,
+      "loss": 1.3933,
+      "step": 2213
+    },
+    {
+      "cosine_similarity": -0.04622691438464182,
+      "epoch": 2.0633737185461323,
+      "grad_norm": 1.1754478636747383,
+      "learning_rate": 6.9382119433897145e-06,
+      "loss": 1.3544,
+      "step": 2214
+    },
+    {
+      "cosine_similarity": 0.007248362083098369,
+      "epoch": 2.06430568499534,
+      "grad_norm": 1.2099956483736054,
+      "learning_rate": 6.931308249913704e-06,
+      "loss": 1.4752,
+      "step": 2215
+    },
+    {
+      "cosine_similarity": -0.10145813387389024,
+      "epoch": 2.065237651444548,
+      "grad_norm": 1.147871788533159,
+      "learning_rate": 6.924404556437695e-06,
+      "loss": 1.2449,
+      "step": 2216
+    },
+    {
+      "cosine_similarity": -0.03582807652707785,
+      "epoch": 2.0661696178937556,
+      "grad_norm": 1.2277104619791186,
+      "learning_rate": 6.917500862961685e-06,
+      "loss": 1.3415,
+      "step": 2217
+    },
+    {
+      "cosine_similarity": 0.1356570505879576,
+      "epoch": 2.0671015843429634,
+      "grad_norm": 1.095325507430864,
+      "learning_rate": 6.910597169485676e-06,
+      "loss": 1.6356,
+      "step": 2218
+    },
+    {
+      "cosine_similarity": 0.06034625566258162,
+      "epoch": 2.0680335507921717,
+      "grad_norm": 1.5164902528280297,
+      "learning_rate": 6.903693476009666e-06,
+      "loss": 1.5583,
+      "step": 2219
+    },
+    {
+      "cosine_similarity": 0.1291538041472393,
+      "epoch": 2.0689655172413794,
+      "grad_norm": 1.2899518441764655,
+      "learning_rate": 6.896789782533656e-06,
+      "loss": 1.2596,
+      "step": 2220
+    },
+    {
+      "cosine_similarity": 0.19876262576276255,
+      "epoch": 2.0698974836905872,
+      "grad_norm": 1.143184898091243,
+      "learning_rate": 6.889886089057647e-06,
+      "loss": 1.5942,
+      "step": 2221
+    },
+    {
+      "cosine_similarity": 0.2325888412870003,
+      "epoch": 2.070829450139795,
+      "grad_norm": 1.3132906704296288,
+      "learning_rate": 6.882982395581637e-06,
+      "loss": 1.5391,
+      "step": 2222
+    },
+    {
+      "cosine_similarity": 0.06896641779022329,
+      "epoch": 2.071761416589003,
+      "grad_norm": 1.2271319342053877,
+      "learning_rate": 6.876078702105628e-06,
+      "loss": 1.4224,
+      "step": 2223
+    },
+    {
+      "cosine_similarity": -0.02035421546856814,
+      "epoch": 2.0726933830382106,
+      "grad_norm": 1.2973864443929712,
+      "learning_rate": 6.869175008629618e-06,
+      "loss": 1.708,
+      "step": 2224
+    },
+    {
+      "cosine_similarity": 0.04054137198374757,
+      "epoch": 2.0736253494874184,
+      "grad_norm": 1.183009162659481,
+      "learning_rate": 6.862271315153608e-06,
+      "loss": 1.4556,
+      "step": 2225
+    },
+    {
+      "cosine_similarity": 0.09500604582656118,
+      "epoch": 2.074557315936626,
+      "grad_norm": 1.1382481982049013,
+      "learning_rate": 6.855367621677597e-06,
+      "loss": 1.6248,
+      "step": 2226
+    },
+    {
+      "cosine_similarity": 0.04799612174186559,
+      "epoch": 2.075489282385834,
+      "grad_norm": 1.4131857778762325,
+      "learning_rate": 6.848463928201589e-06,
+      "loss": 1.5217,
+      "step": 2227
+    },
+    {
+      "cosine_similarity": 0.29570373128423477,
+      "epoch": 2.0764212488350418,
+      "grad_norm": 0.9611622708011329,
+      "learning_rate": 6.841560234725578e-06,
+      "loss": 1.5117,
+      "step": 2228
+    },
+    {
+      "cosine_similarity": 0.14938621935393234,
+      "epoch": 2.0773532152842495,
+      "grad_norm": 1.4984379323818426,
+      "learning_rate": 6.834656541249569e-06,
+      "loss": 1.4471,
+      "step": 2229
+    },
+    {
+      "cosine_similarity": 0.09836598227876078,
+      "epoch": 2.0782851817334578,
+      "grad_norm": 1.0856755409589331,
+      "learning_rate": 6.82775284777356e-06,
+      "loss": 1.2537,
+      "step": 2230
+    },
+    {
+      "cosine_similarity": 0.04566202944807287,
+      "epoch": 2.0792171481826656,
+      "grad_norm": 1.0207714684201565,
+      "learning_rate": 6.820849154297549e-06,
+      "loss": 1.3214,
+      "step": 2231
+    },
+    {
+      "cosine_similarity": -0.03241305176640825,
+      "epoch": 2.0801491146318734,
+      "grad_norm": 1.3493536578488485,
+      "learning_rate": 6.813945460821541e-06,
+      "loss": 1.516,
+      "step": 2232
+    },
+    {
+      "cosine_similarity": 0.08647039925550949,
+      "epoch": 2.081081081081081,
+      "grad_norm": 1.2541001116980683,
+      "learning_rate": 6.80704176734553e-06,
+      "loss": 1.1442,
+      "step": 2233
+    },
+    {
+      "cosine_similarity": 0.13293101992217649,
+      "epoch": 2.082013047530289,
+      "grad_norm": 1.1753357285091355,
+      "learning_rate": 6.800138073869521e-06,
+      "loss": 1.544,
+      "step": 2234
+    },
+    {
+      "cosine_similarity": -0.11786414829580164,
+      "epoch": 2.0829450139794967,
+      "grad_norm": 1.1814586718498254,
+      "learning_rate": 6.7932343803935106e-06,
+      "loss": 1.4741,
+      "step": 2235
+    },
+    {
+      "cosine_similarity": 0.14503272205100762,
+      "epoch": 2.0838769804287045,
+      "grad_norm": 1.2875792789502083,
+      "learning_rate": 6.786330686917501e-06,
+      "loss": 1.3368,
+      "step": 2236
+    },
+    {
+      "cosine_similarity": 0.20920783336813853,
+      "epoch": 2.0848089468779123,
+      "grad_norm": 1.1956004529989368,
+      "learning_rate": 6.779426993441491e-06,
+      "loss": 1.6156,
+      "step": 2237
+    },
+    {
+      "cosine_similarity": 0.2343235851037803,
+      "epoch": 2.08574091332712,
+      "grad_norm": 1.2005406016916118,
+      "learning_rate": 6.772523299965482e-06,
+      "loss": 1.6158,
+      "step": 2238
+    },
+    {
+      "cosine_similarity": 0.11535740621724988,
+      "epoch": 2.086672879776328,
+      "grad_norm": 1.05761215049746,
+      "learning_rate": 6.765619606489472e-06,
+      "loss": 1.4102,
+      "step": 2239
+    },
+    {
+      "cosine_similarity": 0.14140105339723247,
+      "epoch": 2.0876048462255357,
+      "grad_norm": 1.5465047682016824,
+      "learning_rate": 6.7587159130134625e-06,
+      "loss": 1.5874,
+      "step": 2240
+    },
+    {
+      "cosine_similarity": 0.011714830831493836,
+      "epoch": 2.088536812674744,
+      "grad_norm": 1.2736874163580296,
+      "learning_rate": 6.751812219537453e-06,
+      "loss": 1.617,
+      "step": 2241
+    },
+    {
+      "cosine_similarity": 0.17591217487850386,
+      "epoch": 2.0894687791239517,
+      "grad_norm": 1.2347667848241715,
+      "learning_rate": 6.744908526061443e-06,
+      "loss": 1.4225,
+      "step": 2242
+    },
+    {
+      "cosine_similarity": 0.01972333344618023,
+      "epoch": 2.0904007455731595,
+      "grad_norm": 1.0311788067859682,
+      "learning_rate": 6.738004832585434e-06,
+      "loss": 1.2771,
+      "step": 2243
+    },
+    {
+      "cosine_similarity": 0.08442919136288057,
+      "epoch": 2.0913327120223673,
+      "grad_norm": 1.2989944208947992,
+      "learning_rate": 6.731101139109424e-06,
+      "loss": 1.3617,
+      "step": 2244
+    },
+    {
+      "cosine_similarity": -0.09230233394079632,
+      "epoch": 2.092264678471575,
+      "grad_norm": 0.9926173158468751,
+      "learning_rate": 6.7241974456334145e-06,
+      "loss": 1.2923,
+      "step": 2245
+    },
+    {
+      "cosine_similarity": 0.02321952308733538,
+      "epoch": 2.093196644920783,
+      "grad_norm": 1.2943659559001317,
+      "learning_rate": 6.717293752157404e-06,
+      "loss": 1.4355,
+      "step": 2246
+    },
+    {
+      "cosine_similarity": 0.2965082428136477,
+      "epoch": 2.0941286113699906,
+      "grad_norm": 1.5100796493350028,
+      "learning_rate": 6.710390058681395e-06,
+      "loss": 1.5639,
+      "step": 2247
+    },
+    {
+      "cosine_similarity": 0.12176064463100993,
+      "epoch": 2.0950605778191984,
+      "grad_norm": 0.9926746527313939,
+      "learning_rate": 6.703486365205385e-06,
+      "loss": 1.189,
+      "step": 2248
+    },
+    {
+      "cosine_similarity": 0.033977950965654466,
+      "epoch": 2.095992544268406,
+      "grad_norm": 1.1019647949859974,
+      "learning_rate": 6.696582671729376e-06,
+      "loss": 1.3094,
+      "step": 2249
+    },
+    {
+      "cosine_similarity": 0.19626525923633337,
+      "epoch": 2.096924510717614,
+      "grad_norm": 1.4150967328894553,
+      "learning_rate": 6.6896789782533665e-06,
+      "loss": 1.5955,
+      "step": 2250
+    },
+    {
+      "cosine_similarity": 0.12103248035334573,
+      "epoch": 2.0978564771668218,
+      "grad_norm": 1.2820381061385766,
+      "learning_rate": 6.682775284777356e-06,
+      "loss": 1.4198,
+      "step": 2251
+    },
+    {
+      "cosine_similarity": -0.010190507811940423,
+      "epoch": 2.09878844361603,
+      "grad_norm": 1.1683789097052264,
+      "learning_rate": 6.675871591301347e-06,
+      "loss": 1.8632,
+      "step": 2252
+    },
+    {
+      "cosine_similarity": 0.01179468030370027,
+      "epoch": 2.099720410065238,
+      "grad_norm": 1.1783309357899916,
+      "learning_rate": 6.668967897825337e-06,
+      "loss": 1.418,
+      "step": 2253
+    },
+    {
+      "cosine_similarity": 0.11910952284797587,
+      "epoch": 2.1006523765144456,
+      "grad_norm": 1.5980244031687498,
+      "learning_rate": 6.662064204349328e-06,
+      "loss": 1.7296,
+      "step": 2254
+    },
+    {
+      "cosine_similarity": 0.08290942205556424,
+      "epoch": 2.1015843429636534,
+      "grad_norm": 0.9859805494615046,
+      "learning_rate": 6.655160510873318e-06,
+      "loss": 1.4065,
+      "step": 2255
+    },
+    {
+      "cosine_similarity": 0.17338005303655382,
+      "epoch": 2.102516309412861,
+      "grad_norm": 1.1822454572853889,
+      "learning_rate": 6.648256817397308e-06,
+      "loss": 1.9322,
+      "step": 2256
+    },
+    {
+      "cosine_similarity": 0.013378009144537927,
+      "epoch": 2.103448275862069,
+      "grad_norm": 1.17509088734795,
+      "learning_rate": 6.641353123921298e-06,
+      "loss": 1.8807,
+      "step": 2257
+    },
+    {
+      "cosine_similarity": -0.08332356429487585,
+      "epoch": 2.1043802423112767,
+      "grad_norm": 1.0305452836742224,
+      "learning_rate": 6.634449430445289e-06,
+      "loss": 1.3306,
+      "step": 2258
+    },
+    {
+      "cosine_similarity": 0.07102204730080676,
+      "epoch": 2.1053122087604845,
+      "grad_norm": 1.0034810396487375,
+      "learning_rate": 6.627545736969279e-06,
+      "loss": 1.6127,
+      "step": 2259
+    },
+    {
+      "cosine_similarity": -0.05066088375277593,
+      "epoch": 2.1062441752096923,
+      "grad_norm": 1.1201119143282723,
+      "learning_rate": 6.6206420434932696e-06,
+      "loss": 1.305,
+      "step": 2260
+    },
+    {
+      "cosine_similarity": 0.15356187387986284,
+      "epoch": 2.1071761416589,
+      "grad_norm": 1.161473544984363,
+      "learning_rate": 6.61373835001726e-06,
+      "loss": 1.5989,
+      "step": 2261
+    },
+    {
+      "cosine_similarity": 0.02773163154109181,
+      "epoch": 2.108108108108108,
+      "grad_norm": 1.2966445328446978,
+      "learning_rate": 6.60683465654125e-06,
+      "loss": 1.6964,
+      "step": 2262
+    },
+    {
+      "cosine_similarity": -0.01840743802169725,
+      "epoch": 2.109040074557316,
+      "grad_norm": 1.0697889566647678,
+      "learning_rate": 6.599930963065241e-06,
+      "loss": 1.4808,
+      "step": 2263
+    },
+    {
+      "cosine_similarity": 0.08916529178529572,
+      "epoch": 2.109972041006524,
+      "grad_norm": 1.1209033711315568,
+      "learning_rate": 6.593027269589231e-06,
+      "loss": 1.1848,
+      "step": 2264
+    },
+    {
+      "cosine_similarity": -0.061985472741196965,
+      "epoch": 2.1109040074557317,
+      "grad_norm": 1.0756412594374982,
+      "learning_rate": 6.5861235761132215e-06,
+      "loss": 1.3887,
+      "step": 2265
+    },
+    {
+      "cosine_similarity": 0.12950631662860518,
+      "epoch": 2.1118359739049395,
+      "grad_norm": 0.973455721622555,
+      "learning_rate": 6.579219882637211e-06,
+      "loss": 1.4144,
+      "step": 2266
+    },
+    {
+      "cosine_similarity": 0.15721363087527626,
+      "epoch": 2.1127679403541473,
+      "grad_norm": 1.3549364708620077,
+      "learning_rate": 6.572316189161202e-06,
+      "loss": 1.802,
+      "step": 2267
+    },
+    {
+      "cosine_similarity": 0.03296870317389809,
+      "epoch": 2.113699906803355,
+      "grad_norm": 1.1383871190337658,
+      "learning_rate": 6.565412495685192e-06,
+      "loss": 1.6187,
+      "step": 2268
+    },
+    {
+      "cosine_similarity": 0.13041988282693934,
+      "epoch": 2.114631873252563,
+      "grad_norm": 1.0107180613023174,
+      "learning_rate": 6.558508802209183e-06,
+      "loss": 1.4317,
+      "step": 2269
+    },
+    {
+      "cosine_similarity": 0.10801245248736836,
+      "epoch": 2.1155638397017706,
+      "grad_norm": 1.2569968965997307,
+      "learning_rate": 6.551605108733173e-06,
+      "loss": 1.804,
+      "step": 2270
+    },
+    {
+      "cosine_similarity": 0.004301736093193697,
+      "epoch": 2.1164958061509784,
+      "grad_norm": 1.156811666382713,
+      "learning_rate": 6.544701415257163e-06,
+      "loss": 1.4588,
+      "step": 2271
+    },
+    {
+      "cosine_similarity": -0.058075939188633956,
+      "epoch": 2.117427772600186,
+      "grad_norm": 1.095633133749295,
+      "learning_rate": 6.537797721781154e-06,
+      "loss": 1.612,
+      "step": 2272
+    },
+    {
+      "cosine_similarity": -0.06261819244504833,
+      "epoch": 2.118359739049394,
+      "grad_norm": 1.1191698355966548,
+      "learning_rate": 6.530894028305144e-06,
+      "loss": 1.3433,
+      "step": 2273
+    },
+    {
+      "cosine_similarity": 0.23318585051760454,
+      "epoch": 2.1192917054986022,
+      "grad_norm": 1.3658708399169945,
+      "learning_rate": 6.523990334829135e-06,
+      "loss": 1.532,
+      "step": 2274
+    },
+    {
+      "cosine_similarity": 0.1902853298619233,
+      "epoch": 2.12022367194781,
+      "grad_norm": 0.971509033761461,
+      "learning_rate": 6.517086641353124e-06,
+      "loss": 1.2547,
+      "step": 2275
+    },
+    {
+      "cosine_similarity": 0.1864602216342677,
+      "epoch": 2.121155638397018,
+      "grad_norm": 1.1622671242384337,
+      "learning_rate": 6.510182947877115e-06,
+      "loss": 1.5343,
+      "step": 2276
+    },
+    {
+      "cosine_similarity": 0.19472083970598775,
+      "epoch": 2.1220876048462256,
+      "grad_norm": 1.1918181129678855,
+      "learning_rate": 6.503279254401104e-06,
+      "loss": 1.4954,
+      "step": 2277
+    },
+    {
+      "cosine_similarity": 0.08435390308408587,
+      "epoch": 2.1230195712954334,
+      "grad_norm": 1.2827160033382268,
+      "learning_rate": 6.496375560925096e-06,
+      "loss": 1.6133,
+      "step": 2278
+    },
+    {
+      "cosine_similarity": 0.08667964515168311,
+      "epoch": 2.123951537744641,
+      "grad_norm": 1.3101978068120301,
+      "learning_rate": 6.489471867449085e-06,
+      "loss": 1.1938,
+      "step": 2279
+    },
+    {
+      "cosine_similarity": 0.1018517891238452,
+      "epoch": 2.124883504193849,
+      "grad_norm": 0.8879625415922378,
+      "learning_rate": 6.482568173973076e-06,
+      "loss": 1.1813,
+      "step": 2280
+    },
+    {
+      "cosine_similarity": 0.23560073934210138,
+      "epoch": 2.1258154706430568,
+      "grad_norm": 1.3300321996973066,
+      "learning_rate": 6.475664480497067e-06,
+      "loss": 1.2962,
+      "step": 2281
+    },
+    {
+      "cosine_similarity": 0.14167252189443622,
+      "epoch": 2.1267474370922645,
+      "grad_norm": 1.340209010417788,
+      "learning_rate": 6.468760787021056e-06,
+      "loss": 1.4262,
+      "step": 2282
+    },
+    {
+      "cosine_similarity": 0.1411279903101959,
+      "epoch": 2.1276794035414723,
+      "grad_norm": 1.2984610414844089,
+      "learning_rate": 6.461857093545048e-06,
+      "loss": 1.8933,
+      "step": 2283
+    },
+    {
+      "cosine_similarity": -0.18184957466395887,
+      "epoch": 2.12861136999068,
+      "grad_norm": 1.4240612601588736,
+      "learning_rate": 6.454953400069037e-06,
+      "loss": 1.5084,
+      "step": 2284
+    },
+    {
+      "cosine_similarity": -0.0008490977754388219,
+      "epoch": 2.1295433364398884,
+      "grad_norm": 1.075123002847833,
+      "learning_rate": 6.448049706593028e-06,
+      "loss": 1.2208,
+      "step": 2285
+    },
+    {
+      "cosine_similarity": 0.43309083684669436,
+      "epoch": 2.130475302889096,
+      "grad_norm": 1.3371031780919806,
+      "learning_rate": 6.441146013117018e-06,
+      "loss": 1.8664,
+      "step": 2286
+    },
+    {
+      "cosine_similarity": 0.1943871985181864,
+      "epoch": 2.131407269338304,
+      "grad_norm": 1.2161174926931881,
+      "learning_rate": 6.434242319641008e-06,
+      "loss": 1.4988,
+      "step": 2287
+    },
+    {
+      "cosine_similarity": 0.09946305752023865,
+      "epoch": 2.1323392357875117,
+      "grad_norm": 1.2246599768565403,
+      "learning_rate": 6.427338626164998e-06,
+      "loss": 1.3299,
+      "step": 2288
+    },
+    {
+      "cosine_similarity": 0.26119956784373116,
+      "epoch": 2.1332712022367195,
+      "grad_norm": 1.2040972028777432,
+      "learning_rate": 6.420434932688989e-06,
+      "loss": 1.6419,
+      "step": 2289
+    },
+    {
+      "cosine_similarity": 0.010543302027674114,
+      "epoch": 2.1342031686859273,
+      "grad_norm": 1.3453844599626434,
+      "learning_rate": 6.413531239212979e-06,
+      "loss": 1.6422,
+      "step": 2290
+    },
+    {
+      "cosine_similarity": 0.10294845829672986,
+      "epoch": 2.135135135135135,
+      "grad_norm": 1.1545168973796756,
+      "learning_rate": 6.4066275457369696e-06,
+      "loss": 1.4901,
+      "step": 2291
+    },
+    {
+      "cosine_similarity": 0.02060286867333543,
+      "epoch": 2.136067101584343,
+      "grad_norm": 1.0875492580582313,
+      "learning_rate": 6.39972385226096e-06,
+      "loss": 1.2521,
+      "step": 2292
+    },
+    {
+      "cosine_similarity": 0.043927635765535926,
+      "epoch": 2.1369990680335507,
+      "grad_norm": 1.1638611299443857,
+      "learning_rate": 6.39282015878495e-06,
+      "loss": 1.6216,
+      "step": 2293
+    },
+    {
+      "cosine_similarity": 0.0442858616884844,
+      "epoch": 2.1379310344827585,
+      "grad_norm": 1.131921533731666,
+      "learning_rate": 6.385916465308941e-06,
+      "loss": 1.4162,
+      "step": 2294
+    },
+    {
+      "cosine_similarity": 0.06990912883741955,
+      "epoch": 2.1388630009319662,
+      "grad_norm": 1.3094765896382061,
+      "learning_rate": 6.379012771832931e-06,
+      "loss": 1.7049,
+      "step": 2295
+    },
+    {
+      "cosine_similarity": 0.08796320768368462,
+      "epoch": 2.1397949673811745,
+      "grad_norm": 1.1730224192676528,
+      "learning_rate": 6.3721090783569215e-06,
+      "loss": 1.3682,
+      "step": 2296
+    },
+    {
+      "cosine_similarity": 0.05108141859365138,
+      "epoch": 2.1407269338303823,
+      "grad_norm": 1.53069114031854,
+      "learning_rate": 6.365205384880911e-06,
+      "loss": 1.6783,
+      "step": 2297
+    },
+    {
+      "cosine_similarity": 0.32342088038064964,
+      "epoch": 2.14165890027959,
+      "grad_norm": 1.1320835788881232,
+      "learning_rate": 6.358301691404902e-06,
+      "loss": 1.7306,
+      "step": 2298
+    },
+    {
+      "cosine_similarity": 0.19803310642857494,
+      "epoch": 2.142590866728798,
+      "grad_norm": 1.3165987774074177,
+      "learning_rate": 6.351397997928892e-06,
+      "loss": 1.4119,
+      "step": 2299
+    },
+    {
+      "cosine_similarity": 0.1044801160011276,
+      "epoch": 2.1435228331780056,
+      "grad_norm": 1.3906457515333663,
+      "learning_rate": 6.344494304452883e-06,
+      "loss": 1.4722,
+      "step": 2300
+    },
+    {
+      "cosine_similarity": 0.18650892356366666,
+      "epoch": 2.1444547996272134,
+      "grad_norm": 1.3009604315034706,
+      "learning_rate": 6.3375906109768735e-06,
+      "loss": 1.3456,
+      "step": 2301
+    },
+    {
+      "cosine_similarity": 0.11938815491289681,
+      "epoch": 2.145386766076421,
+      "grad_norm": 1.1111782984590985,
+      "learning_rate": 6.330686917500863e-06,
+      "loss": 1.3784,
+      "step": 2302
+    },
+    {
+      "cosine_similarity": -0.004444458156625772,
+      "epoch": 2.146318732525629,
+      "grad_norm": 1.0661991365876304,
+      "learning_rate": 6.323783224024854e-06,
+      "loss": 1.3729,
+      "step": 2303
+    },
+    {
+      "cosine_similarity": 0.017648568433062638,
+      "epoch": 2.147250698974837,
+      "grad_norm": 1.0122979032191142,
+      "learning_rate": 6.316879530548844e-06,
+      "loss": 1.4215,
+      "step": 2304
+    },
+    {
+      "cosine_similarity": 0.0594965106564988,
+      "epoch": 2.1481826654240446,
+      "grad_norm": 1.197281217404825,
+      "learning_rate": 6.309975837072835e-06,
+      "loss": 1.635,
+      "step": 2305
+    },
+    {
+      "cosine_similarity": -0.15091155418655955,
+      "epoch": 2.1491146318732524,
+      "grad_norm": 1.2123937837675076,
+      "learning_rate": 6.303072143596825e-06,
+      "loss": 1.3584,
+      "step": 2306
+    },
+    {
+      "cosine_similarity": 0.041332043274552026,
+      "epoch": 2.1500465983224606,
+      "grad_norm": 0.9822966272363558,
+      "learning_rate": 6.296168450120815e-06,
+      "loss": 1.3664,
+      "step": 2307
+    },
+    {
+      "cosine_similarity": 0.18202976051576006,
+      "epoch": 2.1509785647716684,
+      "grad_norm": 1.2376778421705807,
+      "learning_rate": 6.289264756644805e-06,
+      "loss": 1.5851,
+      "step": 2308
+    },
+    {
+      "cosine_similarity": 0.10539707096626684,
+      "epoch": 2.151910531220876,
+      "grad_norm": 1.1891551767120783,
+      "learning_rate": 6.282361063168796e-06,
+      "loss": 1.4261,
+      "step": 2309
+    },
+    {
+      "cosine_similarity": -0.015594345906758744,
+      "epoch": 2.152842497670084,
+      "grad_norm": 1.154480518127905,
+      "learning_rate": 6.275457369692786e-06,
+      "loss": 1.4044,
+      "step": 2310
+    },
+    {
+      "cosine_similarity": 0.16678593959251614,
+      "epoch": 2.1537744641192917,
+      "grad_norm": 1.2005049561857666,
+      "learning_rate": 6.268553676216777e-06,
+      "loss": 1.4656,
+      "step": 2311
+    },
+    {
+      "cosine_similarity": 0.024901548026671137,
+      "epoch": 2.1547064305684995,
+      "grad_norm": 1.3588062970898525,
+      "learning_rate": 6.261649982740767e-06,
+      "loss": 1.6411,
+      "step": 2312
+    },
+    {
+      "cosine_similarity": 0.028731690475463263,
+      "epoch": 2.1556383970177073,
+      "grad_norm": 1.2075542197585316,
+      "learning_rate": 6.254746289264757e-06,
+      "loss": 1.7453,
+      "step": 2313
+    },
+    {
+      "cosine_similarity": 0.17417977944293786,
+      "epoch": 2.156570363466915,
+      "grad_norm": 1.2389301892827864,
+      "learning_rate": 6.247842595788748e-06,
+      "loss": 1.2207,
+      "step": 2314
+    },
+    {
+      "cosine_similarity": 0.15784962112419892,
+      "epoch": 2.157502329916123,
+      "grad_norm": 1.1854089985596505,
+      "learning_rate": 6.240938902312738e-06,
+      "loss": 1.4301,
+      "step": 2315
+    },
+    {
+      "cosine_similarity": 0.22899530418640765,
+      "epoch": 2.1584342963653307,
+      "grad_norm": 1.2553598199651448,
+      "learning_rate": 6.2340352088367286e-06,
+      "loss": 1.7626,
+      "step": 2316
+    },
+    {
+      "cosine_similarity": 0.06853951662477369,
+      "epoch": 2.1593662628145385,
+      "grad_norm": 1.3101056730114105,
+      "learning_rate": 6.2271315153607184e-06,
+      "loss": 1.4332,
+      "step": 2317
+    },
+    {
+      "cosine_similarity": 0.17448392289260442,
+      "epoch": 2.1602982292637467,
+      "grad_norm": 1.2123050516280918,
+      "learning_rate": 6.220227821884709e-06,
+      "loss": 1.4663,
+      "step": 2318
+    },
+    {
+      "cosine_similarity": 0.02471808393453321,
+      "epoch": 2.1612301957129545,
+      "grad_norm": 1.1326452077050755,
+      "learning_rate": 6.213324128408699e-06,
+      "loss": 1.3045,
+      "step": 2319
+    },
+    {
+      "cosine_similarity": -0.06554499306719708,
+      "epoch": 2.1621621621621623,
+      "grad_norm": 1.1029010257551985,
+      "learning_rate": 6.20642043493269e-06,
+      "loss": 1.4907,
+      "step": 2320
+    },
+    {
+      "cosine_similarity": 0.3387139935268005,
+      "epoch": 2.16309412861137,
+      "grad_norm": 1.5401007239253661,
+      "learning_rate": 6.1995167414566805e-06,
+      "loss": 1.879,
+      "step": 2321
+    },
+    {
+      "cosine_similarity": 0.16919694696972076,
+      "epoch": 2.164026095060578,
+      "grad_norm": 1.210920596109071,
+      "learning_rate": 6.19261304798067e-06,
+      "loss": 1.547,
+      "step": 2322
+    },
+    {
+      "cosine_similarity": -0.08438833141245936,
+      "epoch": 2.1649580615097856,
+      "grad_norm": 1.2855796801272117,
+      "learning_rate": 6.185709354504661e-06,
+      "loss": 2.1619,
+      "step": 2323
+    },
+    {
+      "cosine_similarity": 0.2105415193781709,
+      "epoch": 2.1658900279589934,
+      "grad_norm": 1.0749504887662416,
+      "learning_rate": 6.178805661028651e-06,
+      "loss": 1.3142,
+      "step": 2324
+    },
+    {
+      "cosine_similarity": 0.3014412801997004,
+      "epoch": 2.1668219944082012,
+      "grad_norm": 1.0299390753840265,
+      "learning_rate": 6.171901967552642e-06,
+      "loss": 1.3343,
+      "step": 2325
+    },
+    {
+      "cosine_similarity": 0.19104946214739768,
+      "epoch": 2.167753960857409,
+      "grad_norm": 1.2814441167775084,
+      "learning_rate": 6.164998274076631e-06,
+      "loss": 1.5411,
+      "step": 2326
+    },
+    {
+      "cosine_similarity": 0.13191496433402186,
+      "epoch": 2.168685927306617,
+      "grad_norm": 1.1863262301318336,
+      "learning_rate": 6.158094580600622e-06,
+      "loss": 1.4029,
+      "step": 2327
+    },
+    {
+      "cosine_similarity": -0.04256987962080702,
+      "epoch": 2.1696178937558246,
+      "grad_norm": 1.337635127037114,
+      "learning_rate": 6.151190887124611e-06,
+      "loss": 1.3141,
+      "step": 2328
+    },
+    {
+      "cosine_similarity": -0.27250651861308145,
+      "epoch": 2.170549860205033,
+      "grad_norm": 1.1718820115884436,
+      "learning_rate": 6.144287193648603e-06,
+      "loss": 1.5923,
+      "step": 2329
+    },
+    {
+      "cosine_similarity": 0.12450933906496005,
+      "epoch": 2.1714818266542406,
+      "grad_norm": 0.9910987264960247,
+      "learning_rate": 6.137383500172592e-06,
+      "loss": 1.6117,
+      "step": 2330
+    },
+    {
+      "cosine_similarity": 0.04642030574414822,
+      "epoch": 2.1724137931034484,
+      "grad_norm": 1.1009305071401707,
+      "learning_rate": 6.130479806696583e-06,
+      "loss": 1.7542,
+      "step": 2331
+    },
+    {
+      "cosine_similarity": 0.30187687426162585,
+      "epoch": 2.173345759552656,
+      "grad_norm": 1.232606241555057,
+      "learning_rate": 6.123576113220574e-06,
+      "loss": 1.5808,
+      "step": 2332
+    },
+    {
+      "cosine_similarity": 0.13078795145247057,
+      "epoch": 2.174277726001864,
+      "grad_norm": 1.2105400746592259,
+      "learning_rate": 6.116672419744563e-06,
+      "loss": 1.4373,
+      "step": 2333
+    },
+    {
+      "cosine_similarity": -0.021610801997063372,
+      "epoch": 2.1752096924510718,
+      "grad_norm": 1.149682273576602,
+      "learning_rate": 6.109768726268555e-06,
+      "loss": 1.6185,
+      "step": 2334
+    },
+    {
+      "cosine_similarity": -0.08589832261499067,
+      "epoch": 2.1761416589002796,
+      "grad_norm": 1.2971916417709475,
+      "learning_rate": 6.102865032792544e-06,
+      "loss": 1.4257,
+      "step": 2335
+    },
+    {
+      "cosine_similarity": 0.03820428583451885,
+      "epoch": 2.1770736253494873,
+      "grad_norm": 1.2580318668384867,
+      "learning_rate": 6.095961339316535e-06,
+      "loss": 1.597,
+      "step": 2336
+    },
+    {
+      "cosine_similarity": -0.1275658607311519,
+      "epoch": 2.178005591798695,
+      "grad_norm": 1.1713171393335082,
+      "learning_rate": 6.089057645840525e-06,
+      "loss": 1.3457,
+      "step": 2337
+    },
+    {
+      "cosine_similarity": 0.12750065617544132,
+      "epoch": 2.178937558247903,
+      "grad_norm": 1.028876767780903,
+      "learning_rate": 6.082153952364515e-06,
+      "loss": 1.3161,
+      "step": 2338
+    },
+    {
+      "cosine_similarity": -0.061934828230773994,
+      "epoch": 2.1798695246971107,
+      "grad_norm": 1.4237236680720662,
+      "learning_rate": 6.075250258888505e-06,
+      "loss": 1.4709,
+      "step": 2339
+    },
+    {
+      "cosine_similarity": -0.13016391663035862,
+      "epoch": 2.180801491146319,
+      "grad_norm": 1.0363816122842122,
+      "learning_rate": 6.068346565412496e-06,
+      "loss": 1.4653,
+      "step": 2340
+    },
+    {
+      "cosine_similarity": -0.002517219090928886,
+      "epoch": 2.1817334575955267,
+      "grad_norm": 1.2548992435442645,
+      "learning_rate": 6.061442871936487e-06,
+      "loss": 1.4298,
+      "step": 2341
+    },
+    {
+      "cosine_similarity": -0.0034540599742463765,
+      "epoch": 2.1826654240447345,
+      "grad_norm": 1.2749583493968037,
+      "learning_rate": 6.054539178460477e-06,
+      "loss": 1.5999,
+      "step": 2342
+    },
+    {
+      "cosine_similarity": 0.10899274914372452,
+      "epoch": 2.1835973904939423,
+      "grad_norm": 1.0577267848526746,
+      "learning_rate": 6.047635484984467e-06,
+      "loss": 1.7169,
+      "step": 2343
+    },
+    {
+      "cosine_similarity": 0.21245707027260935,
+      "epoch": 2.18452935694315,
+      "grad_norm": 1.103300372941199,
+      "learning_rate": 6.040731791508457e-06,
+      "loss": 1.6314,
+      "step": 2344
+    },
+    {
+      "cosine_similarity": -0.15913137445215336,
+      "epoch": 2.185461323392358,
+      "grad_norm": 1.1605574327356831,
+      "learning_rate": 6.033828098032448e-06,
+      "loss": 1.576,
+      "step": 2345
+    },
+    {
+      "cosine_similarity": 0.27579169768969475,
+      "epoch": 2.1863932898415657,
+      "grad_norm": 1.1789528969923675,
+      "learning_rate": 6.026924404556438e-06,
+      "loss": 1.4894,
+      "step": 2346
+    },
+    {
+      "cosine_similarity": 0.07499460122752931,
+      "epoch": 2.1873252562907735,
+      "grad_norm": 1.2877585133511227,
+      "learning_rate": 6.0200207110804286e-06,
+      "loss": 1.5628,
+      "step": 2347
+    },
+    {
+      "cosine_similarity": 0.026171085697552714,
+      "epoch": 2.1882572227399812,
+      "grad_norm": 1.0386758204980036,
+      "learning_rate": 6.0131170176044184e-06,
+      "loss": 1.4485,
+      "step": 2348
+    },
+    {
+      "cosine_similarity": 0.15807983010889673,
+      "epoch": 2.189189189189189,
+      "grad_norm": 1.1230650790380672,
+      "learning_rate": 6.006213324128409e-06,
+      "loss": 1.5325,
+      "step": 2349
+    },
+    {
+      "cosine_similarity": 0.20021582969701562,
+      "epoch": 2.190121155638397,
+      "grad_norm": 1.2990574270271278,
+      "learning_rate": 5.999309630652399e-06,
+      "loss": 1.3655,
+      "step": 2350
+    },
+    {
+      "cosine_similarity": 0.030837047912556297,
+      "epoch": 2.191053122087605,
+      "grad_norm": 0.950217216682341,
+      "learning_rate": 5.99240593717639e-06,
+      "loss": 1.6549,
+      "step": 2351
+    },
+    {
+      "cosine_similarity": 0.2839239750880349,
+      "epoch": 2.191985088536813,
+      "grad_norm": 1.3288187532489024,
+      "learning_rate": 5.9855022437003805e-06,
+      "loss": 1.9885,
+      "step": 2352
+    },
+    {
+      "cosine_similarity": -0.025635775695117036,
+      "epoch": 2.1929170549860206,
+      "grad_norm": 1.400407760139919,
+      "learning_rate": 5.97859855022437e-06,
+      "loss": 1.2529,
+      "step": 2353
+    },
+    {
+      "cosine_similarity": -0.07388202092862642,
+      "epoch": 2.1938490214352284,
+      "grad_norm": 1.1458238021982352,
+      "learning_rate": 5.971694856748361e-06,
+      "loss": 1.6013,
+      "step": 2354
+    },
+    {
+      "cosine_similarity": 0.22062029031502087,
+      "epoch": 2.194780987884436,
+      "grad_norm": 1.0870916997058218,
+      "learning_rate": 5.964791163272351e-06,
+      "loss": 1.3364,
+      "step": 2355
+    },
+    {
+      "cosine_similarity": 0.07385382827510294,
+      "epoch": 2.195712954333644,
+      "grad_norm": 1.057090787829228,
+      "learning_rate": 5.957887469796342e-06,
+      "loss": 1.3718,
+      "step": 2356
+    },
+    {
+      "cosine_similarity": 0.06792023299800769,
+      "epoch": 2.196644920782852,
+      "grad_norm": 1.2064753719818964,
+      "learning_rate": 5.950983776320332e-06,
+      "loss": 1.8227,
+      "step": 2357
+    },
+    {
+      "cosine_similarity": 0.00023153791609373362,
+      "epoch": 2.1975768872320596,
+      "grad_norm": 1.0834936965345399,
+      "learning_rate": 5.944080082844322e-06,
+      "loss": 1.7063,
+      "step": 2358
+    },
+    {
+      "cosine_similarity": 0.018294223605253453,
+      "epoch": 2.1985088536812674,
+      "grad_norm": 1.1111726088408702,
+      "learning_rate": 5.937176389368312e-06,
+      "loss": 1.6398,
+      "step": 2359
+    },
+    {
+      "cosine_similarity": 0.29360441242329566,
+      "epoch": 2.199440820130475,
+      "grad_norm": 1.3056651936764478,
+      "learning_rate": 5.930272695892303e-06,
+      "loss": 1.6289,
+      "step": 2360
+    },
+    {
+      "cosine_similarity": -0.01585913792377906,
+      "epoch": 2.200372786579683,
+      "grad_norm": 1.0759906702724986,
+      "learning_rate": 5.923369002416293e-06,
+      "loss": 1.6306,
+      "step": 2361
+    },
+    {
+      "cosine_similarity": -0.039383492213823625,
+      "epoch": 2.201304753028891,
+      "grad_norm": 1.27684690096842,
+      "learning_rate": 5.916465308940284e-06,
+      "loss": 1.3377,
+      "step": 2362
+    },
+    {
+      "cosine_similarity": 0.17011357812055952,
+      "epoch": 2.202236719478099,
+      "grad_norm": 1.1670881695174928,
+      "learning_rate": 5.909561615464274e-06,
+      "loss": 1.6408,
+      "step": 2363
+    },
+    {
+      "cosine_similarity": 0.31377592096079693,
+      "epoch": 2.2031686859273067,
+      "grad_norm": 1.243283388881653,
+      "learning_rate": 5.902657921988264e-06,
+      "loss": 1.6075,
+      "step": 2364
+    },
+    {
+      "cosine_similarity": 0.22211328106090467,
+      "epoch": 2.2041006523765145,
+      "grad_norm": 1.2813685260226328,
+      "learning_rate": 5.895754228512255e-06,
+      "loss": 1.3654,
+      "step": 2365
+    },
+    {
+      "cosine_similarity": 0.11359098503822178,
+      "epoch": 2.2050326188257223,
+      "grad_norm": 1.1965910306592282,
+      "learning_rate": 5.888850535036245e-06,
+      "loss": 1.3222,
+      "step": 2366
+    },
+    {
+      "cosine_similarity": 0.12808166434144994,
+      "epoch": 2.20596458527493,
+      "grad_norm": 1.2188902665153645,
+      "learning_rate": 5.881946841560236e-06,
+      "loss": 1.4159,
+      "step": 2367
+    },
+    {
+      "cosine_similarity": 0.1617747946712549,
+      "epoch": 2.206896551724138,
+      "grad_norm": 1.390886417969745,
+      "learning_rate": 5.8750431480842255e-06,
+      "loss": 1.7987,
+      "step": 2368
+    },
+    {
+      "cosine_similarity": 0.03956140731105886,
+      "epoch": 2.2078285181733457,
+      "grad_norm": 1.1509354319759488,
+      "learning_rate": 5.868139454608216e-06,
+      "loss": 1.2335,
+      "step": 2369
+    },
+    {
+      "cosine_similarity": 0.19006911394270654,
+      "epoch": 2.2087604846225535,
+      "grad_norm": 1.2796236603003295,
+      "learning_rate": 5.861235761132206e-06,
+      "loss": 1.5779,
+      "step": 2370
+    },
+    {
+      "cosine_similarity": 0.18296014618601586,
+      "epoch": 2.2096924510717613,
+      "grad_norm": 1.1600945619081175,
+      "learning_rate": 5.854332067656197e-06,
+      "loss": 1.5064,
+      "step": 2371
+    },
+    {
+      "cosine_similarity": -0.09473525174452292,
+      "epoch": 2.210624417520969,
+      "grad_norm": 1.2111152008356765,
+      "learning_rate": 5.8474283741801875e-06,
+      "loss": 1.2841,
+      "step": 2372
+    },
+    {
+      "cosine_similarity": 0.009036004874040758,
+      "epoch": 2.2115563839701773,
+      "grad_norm": 1.1740593971516842,
+      "learning_rate": 5.8405246807041774e-06,
+      "loss": 1.6725,
+      "step": 2373
+    },
+    {
+      "cosine_similarity": 0.01972426354378221,
+      "epoch": 2.212488350419385,
+      "grad_norm": 1.1716865818719477,
+      "learning_rate": 5.833620987228168e-06,
+      "loss": 1.3499,
+      "step": 2374
+    },
+    {
+      "cosine_similarity": -0.11838915847958235,
+      "epoch": 2.213420316868593,
+      "grad_norm": 1.257450148186415,
+      "learning_rate": 5.826717293752158e-06,
+      "loss": 1.4279,
+      "step": 2375
+    },
+    {
+      "cosine_similarity": 0.022313584706347114,
+      "epoch": 2.2143522833178007,
+      "grad_norm": 1.6653522869085766,
+      "learning_rate": 5.819813600276149e-06,
+      "loss": 1.3699,
+      "step": 2376
+    },
+    {
+      "cosine_similarity": 0.0884680541546098,
+      "epoch": 2.2152842497670084,
+      "grad_norm": 1.036642926799536,
+      "learning_rate": 5.812909906800138e-06,
+      "loss": 1.5059,
+      "step": 2377
+    },
+    {
+      "cosine_similarity": 0.006253582715961041,
+      "epoch": 2.2162162162162162,
+      "grad_norm": 1.1095726558638466,
+      "learning_rate": 5.806006213324129e-06,
+      "loss": 1.3461,
+      "step": 2378
+    },
+    {
+      "cosine_similarity": 0.02734814245829689,
+      "epoch": 2.217148182665424,
+      "grad_norm": 1.1606700770137601,
+      "learning_rate": 5.7991025198481184e-06,
+      "loss": 1.4935,
+      "step": 2379
+    },
+    {
+      "cosine_similarity": 0.1052486398166746,
+      "epoch": 2.218080149114632,
+      "grad_norm": 1.5831601308653889,
+      "learning_rate": 5.79219882637211e-06,
+      "loss": 1.3741,
+      "step": 2380
+    },
+    {
+      "cosine_similarity": 0.2954157085517678,
+      "epoch": 2.2190121155638396,
+      "grad_norm": 1.1015000936683392,
+      "learning_rate": 5.785295132896099e-06,
+      "loss": 1.6546,
+      "step": 2381
+    },
+    {
+      "cosine_similarity": 0.4235864323958019,
+      "epoch": 2.2199440820130474,
+      "grad_norm": 1.2050584185145627,
+      "learning_rate": 5.77839143942009e-06,
+      "loss": 1.7233,
+      "step": 2382
+    },
+    {
+      "cosine_similarity": -0.05518590318976492,
+      "epoch": 2.220876048462255,
+      "grad_norm": 1.020243187454305,
+      "learning_rate": 5.771487745944081e-06,
+      "loss": 1.2298,
+      "step": 2383
+    },
+    {
+      "cosine_similarity": 0.17647495474259178,
+      "epoch": 2.2218080149114634,
+      "grad_norm": 1.3636977277646547,
+      "learning_rate": 5.76458405246807e-06,
+      "loss": 1.2047,
+      "step": 2384
+    },
+    {
+      "cosine_similarity": 0.1867280503113264,
+      "epoch": 2.222739981360671,
+      "grad_norm": 1.069585956360386,
+      "learning_rate": 5.757680358992062e-06,
+      "loss": 1.3046,
+      "step": 2385
+    },
+    {
+      "cosine_similarity": 0.1421619387472237,
+      "epoch": 2.223671947809879,
+      "grad_norm": 1.0844158534565083,
+      "learning_rate": 5.750776665516051e-06,
+      "loss": 1.6886,
+      "step": 2386
+    },
+    {
+      "cosine_similarity": 0.15739173502187306,
+      "epoch": 2.2246039142590868,
+      "grad_norm": 1.2935542458808893,
+      "learning_rate": 5.743872972040042e-06,
+      "loss": 1.577,
+      "step": 2387
+    },
+    {
+      "cosine_similarity": -0.019161723337490433,
+      "epoch": 2.2255358807082946,
+      "grad_norm": 1.1112242627603948,
+      "learning_rate": 5.736969278564032e-06,
+      "loss": 1.461,
+      "step": 2388
+    },
+    {
+      "cosine_similarity": 0.1014521382410846,
+      "epoch": 2.2264678471575023,
+      "grad_norm": 1.1204878914957208,
+      "learning_rate": 5.730065585088022e-06,
+      "loss": 1.575,
+      "step": 2389
+    },
+    {
+      "cosine_similarity": 0.04435247372765187,
+      "epoch": 2.22739981360671,
+      "grad_norm": 0.9994510174817347,
+      "learning_rate": 5.723161891612012e-06,
+      "loss": 1.2455,
+      "step": 2390
+    },
+    {
+      "cosine_similarity": 0.16317796545645347,
+      "epoch": 2.228331780055918,
+      "grad_norm": 1.0546476638139382,
+      "learning_rate": 5.716258198136003e-06,
+      "loss": 1.3201,
+      "step": 2391
+    },
+    {
+      "cosine_similarity": -0.10622717463349059,
+      "epoch": 2.2292637465051257,
+      "grad_norm": 1.2034261193455724,
+      "learning_rate": 5.709354504659994e-06,
+      "loss": 1.4417,
+      "step": 2392
+    },
+    {
+      "cosine_similarity": -0.008055967375279414,
+      "epoch": 2.2301957129543335,
+      "grad_norm": 1.281673755231543,
+      "learning_rate": 5.702450811183984e-06,
+      "loss": 1.5119,
+      "step": 2393
+    },
+    {
+      "cosine_similarity": 0.05181494009505516,
+      "epoch": 2.2311276794035413,
+      "grad_norm": 1.1921248018874848,
+      "learning_rate": 5.695547117707974e-06,
+      "loss": 1.5213,
+      "step": 2394
+    },
+    {
+      "cosine_similarity": -0.07990258488180219,
+      "epoch": 2.2320596458527495,
+      "grad_norm": 1.043072094276622,
+      "learning_rate": 5.688643424231964e-06,
+      "loss": 1.2131,
+      "step": 2395
+    },
+    {
+      "cosine_similarity": 0.11116443356517888,
+      "epoch": 2.2329916123019573,
+      "grad_norm": 1.4866167858429815,
+      "learning_rate": 5.681739730755955e-06,
+      "loss": 1.3857,
+      "step": 2396
+    },
+    {
+      "cosine_similarity": 0.09371175802851191,
+      "epoch": 2.233923578751165,
+      "grad_norm": 1.2182518176661885,
+      "learning_rate": 5.674836037279945e-06,
+      "loss": 1.4267,
+      "step": 2397
+    },
+    {
+      "cosine_similarity": 0.04147025347004539,
+      "epoch": 2.234855545200373,
+      "grad_norm": 1.0461710932252588,
+      "learning_rate": 5.667932343803936e-06,
+      "loss": 1.2696,
+      "step": 2398
+    },
+    {
+      "cosine_similarity": 0.12794646942969443,
+      "epoch": 2.2357875116495807,
+      "grad_norm": 1.2933750096650052,
+      "learning_rate": 5.6610286503279255e-06,
+      "loss": 1.5821,
+      "step": 2399
+    },
+    {
+      "cosine_similarity": 0.03264933753820703,
+      "epoch": 2.2367194780987885,
+      "grad_norm": 1.1880197035070228,
+      "learning_rate": 5.654124956851916e-06,
+      "loss": 1.7183,
+      "step": 2400
+    },
+    {
+      "cosine_similarity": 0.09498797785159611,
+      "epoch": 2.2376514445479962,
+      "grad_norm": 1.1327394452201287,
+      "learning_rate": 5.647221263375906e-06,
+      "loss": 1.7061,
+      "step": 2401
+    },
+    {
+      "cosine_similarity": 0.17602402991819308,
+      "epoch": 2.238583410997204,
+      "grad_norm": 1.01836329250445,
+      "learning_rate": 5.640317569899897e-06,
+      "loss": 1.4753,
+      "step": 2402
+    },
+    {
+      "cosine_similarity": -0.02773988124365714,
+      "epoch": 2.239515377446412,
+      "grad_norm": 1.2057901817802095,
+      "learning_rate": 5.6334138764238875e-06,
+      "loss": 1.5126,
+      "step": 2403
+    },
+    {
+      "cosine_similarity": 0.30186920393124983,
+      "epoch": 2.2404473438956196,
+      "grad_norm": 1.2728578280144853,
+      "learning_rate": 5.6265101829478774e-06,
+      "loss": 1.1376,
+      "step": 2404
+    },
+    {
+      "cosine_similarity": 0.047823971968098546,
+      "epoch": 2.2413793103448274,
+      "grad_norm": 1.2868948371115698,
+      "learning_rate": 5.619606489471868e-06,
+      "loss": 1.4175,
+      "step": 2405
+    },
+    {
+      "cosine_similarity": 0.06797338008009729,
+      "epoch": 2.2423112767940356,
+      "grad_norm": 1.1680751950806945,
+      "learning_rate": 5.612702795995858e-06,
+      "loss": 1.3238,
+      "step": 2406
+    },
+    {
+      "cosine_similarity": 0.3532456526452282,
+      "epoch": 2.2432432432432434,
+      "grad_norm": 1.1706489717089439,
+      "learning_rate": 5.605799102519849e-06,
+      "loss": 1.9061,
+      "step": 2407
+    },
+    {
+      "cosine_similarity": 0.09125884427402915,
+      "epoch": 2.244175209692451,
+      "grad_norm": 1.3716466413583217,
+      "learning_rate": 5.598895409043839e-06,
+      "loss": 1.3628,
+      "step": 2408
+    },
+    {
+      "cosine_similarity": 0.08874061289392889,
+      "epoch": 2.245107176141659,
+      "grad_norm": 1.629569022686214,
+      "learning_rate": 5.591991715567829e-06,
+      "loss": 1.4265,
+      "step": 2409
+    },
+    {
+      "cosine_similarity": 0.005160288147413453,
+      "epoch": 2.246039142590867,
+      "grad_norm": 1.3354836456207597,
+      "learning_rate": 5.585088022091819e-06,
+      "loss": 1.467,
+      "step": 2410
+    },
+    {
+      "cosine_similarity": 0.00013067423901251245,
+      "epoch": 2.2469711090400746,
+      "grad_norm": 1.5553971932275705,
+      "learning_rate": 5.57818432861581e-06,
+      "loss": 1.4899,
+      "step": 2411
+    },
+    {
+      "cosine_similarity": -0.11776395657951298,
+      "epoch": 2.2479030754892824,
+      "grad_norm": 1.0998269089859571,
+      "learning_rate": 5.571280635139801e-06,
+      "loss": 1.4526,
+      "step": 2412
+    },
+    {
+      "cosine_similarity": 0.016146576562431895,
+      "epoch": 2.24883504193849,
+      "grad_norm": 1.02727144602557,
+      "learning_rate": 5.564376941663791e-06,
+      "loss": 1.3864,
+      "step": 2413
+    },
+    {
+      "cosine_similarity": 0.15409788765591603,
+      "epoch": 2.249767008387698,
+      "grad_norm": 1.2498250304193927,
+      "learning_rate": 5.557473248187781e-06,
+      "loss": 1.4388,
+      "step": 2414
+    },
+    {
+      "cosine_similarity": 0.15969482355917342,
+      "epoch": 2.2506989748369057,
+      "grad_norm": 0.9633888620330026,
+      "learning_rate": 5.550569554711771e-06,
+      "loss": 1.4679,
+      "step": 2415
+    },
+    {
+      "cosine_similarity": 0.230954371744545,
+      "epoch": 2.2516309412861135,
+      "grad_norm": 1.6430461514578467,
+      "learning_rate": 5.543665861235762e-06,
+      "loss": 1.6196,
+      "step": 2416
+    },
+    {
+      "cosine_similarity": 0.1868414385218552,
+      "epoch": 2.2525629077353218,
+      "grad_norm": 1.2613787504465837,
+      "learning_rate": 5.536762167759752e-06,
+      "loss": 1.3911,
+      "step": 2417
+    },
+    {
+      "cosine_similarity": -0.0167139079944995,
+      "epoch": 2.2534948741845295,
+      "grad_norm": 1.3046154747862937,
+      "learning_rate": 5.529858474283743e-06,
+      "loss": 1.6291,
+      "step": 2418
+    },
+    {
+      "cosine_similarity": 0.09318791603726838,
+      "epoch": 2.2544268406337373,
+      "grad_norm": 1.0736478658140316,
+      "learning_rate": 5.5229547808077325e-06,
+      "loss": 1.4218,
+      "step": 2419
+    },
+    {
+      "cosine_similarity": 0.0624597987946558,
+      "epoch": 2.255358807082945,
+      "grad_norm": 1.2853915954959494,
+      "learning_rate": 5.516051087331723e-06,
+      "loss": 1.3516,
+      "step": 2420
+    },
+    {
+      "cosine_similarity": 0.024227500319276087,
+      "epoch": 2.256290773532153,
+      "grad_norm": 1.3401118439006796,
+      "learning_rate": 5.509147393855713e-06,
+      "loss": 1.265,
+      "step": 2421
+    },
+    {
+      "cosine_similarity": -0.16548533208283434,
+      "epoch": 2.2572227399813607,
+      "grad_norm": 1.044648742071634,
+      "learning_rate": 5.502243700379704e-06,
+      "loss": 1.3115,
+      "step": 2422
+    },
+    {
+      "cosine_similarity": 0.15869864904987652,
+      "epoch": 2.2581547064305685,
+      "grad_norm": 1.2978910526911653,
+      "learning_rate": 5.4953400069036946e-06,
+      "loss": 1.313,
+      "step": 2423
+    },
+    {
+      "cosine_similarity": 0.07162138697498868,
+      "epoch": 2.2590866728797763,
+      "grad_norm": 1.3108618012976396,
+      "learning_rate": 5.4884363134276845e-06,
+      "loss": 1.5214,
+      "step": 2424
+    },
+    {
+      "cosine_similarity": 0.056912858565185,
+      "epoch": 2.260018639328984,
+      "grad_norm": 1.0586486657754635,
+      "learning_rate": 5.481532619951675e-06,
+      "loss": 1.3996,
+      "step": 2425
+    },
+    {
+      "cosine_similarity": 0.1409808468811646,
+      "epoch": 2.260950605778192,
+      "grad_norm": 1.4633912077658482,
+      "learning_rate": 5.474628926475665e-06,
+      "loss": 1.6913,
+      "step": 2426
+    },
+    {
+      "cosine_similarity": 0.09415153323094498,
+      "epoch": 2.2618825722273996,
+      "grad_norm": 1.197610198851028,
+      "learning_rate": 5.467725232999656e-06,
+      "loss": 1.6623,
+      "step": 2427
+    },
+    {
+      "cosine_similarity": 0.052984616134021154,
+      "epoch": 2.262814538676608,
+      "grad_norm": 1.1195672088582778,
+      "learning_rate": 5.460821539523645e-06,
+      "loss": 1.4944,
+      "step": 2428
+    },
+    {
+      "cosine_similarity": 0.158676613736475,
+      "epoch": 2.2637465051258157,
+      "grad_norm": 1.1742280706719466,
+      "learning_rate": 5.4539178460476364e-06,
+      "loss": 1.4214,
+      "step": 2429
+    },
+    {
+      "cosine_similarity": -0.01913518924172565,
+      "epoch": 2.2646784715750234,
+      "grad_norm": 1.1084616603605193,
+      "learning_rate": 5.4470141525716255e-06,
+      "loss": 1.1667,
+      "step": 2430
+    },
+    {
+      "cosine_similarity": 0.20590906025934033,
+      "epoch": 2.2656104380242312,
+      "grad_norm": 1.0797208751093699,
+      "learning_rate": 5.440110459095617e-06,
+      "loss": 1.3738,
+      "step": 2431
+    },
+    {
+      "cosine_similarity": 0.09757710401561603,
+      "epoch": 2.266542404473439,
+      "grad_norm": 1.0424953218892208,
+      "learning_rate": 5.433206765619608e-06,
+      "loss": 1.5084,
+      "step": 2432
+    },
+    {
+      "cosine_similarity": 0.04100271404490905,
+      "epoch": 2.267474370922647,
+      "grad_norm": 1.1700397911064522,
+      "learning_rate": 5.426303072143597e-06,
+      "loss": 1.4111,
+      "step": 2433
+    },
+    {
+      "cosine_similarity": 0.05308092477330836,
+      "epoch": 2.2684063373718546,
+      "grad_norm": 1.1689502566069365,
+      "learning_rate": 5.419399378667588e-06,
+      "loss": 1.3784,
+      "step": 2434
+    },
+    {
+      "cosine_similarity": -0.05039611896666946,
+      "epoch": 2.2693383038210624,
+      "grad_norm": 1.2091708698140473,
+      "learning_rate": 5.4124956851915774e-06,
+      "loss": 1.5644,
+      "step": 2435
+    },
+    {
+      "cosine_similarity": 0.02639870610708524,
+      "epoch": 2.27027027027027,
+      "grad_norm": 1.2310201049195928,
+      "learning_rate": 5.405591991715568e-06,
+      "loss": 1.2492,
+      "step": 2436
+    },
+    {
+      "cosine_similarity": 0.09960548657796726,
+      "epoch": 2.271202236719478,
+      "grad_norm": 1.225900638092262,
+      "learning_rate": 5.398688298239558e-06,
+      "loss": 1.5903,
+      "step": 2437
+    },
+    {
+      "cosine_similarity": 0.09129142308929114,
+      "epoch": 2.2721342031686858,
+      "grad_norm": 1.0614778508691851,
+      "learning_rate": 5.391784604763549e-06,
+      "loss": 1.1982,
+      "step": 2438
+    },
+    {
+      "cosine_similarity": 0.07844055629534509,
+      "epoch": 2.273066169617894,
+      "grad_norm": 1.052280186331269,
+      "learning_rate": 5.384880911287539e-06,
+      "loss": 1.2975,
+      "step": 2439
+    },
+    {
+      "cosine_similarity": -0.06250736689132798,
+      "epoch": 2.2739981360671018,
+      "grad_norm": 1.1264583786742588,
+      "learning_rate": 5.377977217811529e-06,
+      "loss": 1.6153,
+      "step": 2440
+    },
+    {
+      "cosine_similarity": 0.1545409433407724,
+      "epoch": 2.2749301025163096,
+      "grad_norm": 1.1636358871006733,
+      "learning_rate": 5.371073524335519e-06,
+      "loss": 1.3341,
+      "step": 2441
+    },
+    {
+      "cosine_similarity": 0.046695531769306804,
+      "epoch": 2.2758620689655173,
+      "grad_norm": 1.0245930184175824,
+      "learning_rate": 5.36416983085951e-06,
+      "loss": 1.5408,
+      "step": 2442
+    },
+    {
+      "cosine_similarity": -0.026481558060446143,
+      "epoch": 2.276794035414725,
+      "grad_norm": 1.2317687081036357,
+      "learning_rate": 5.357266137383501e-06,
+      "loss": 1.3577,
+      "step": 2443
+    },
+    {
+      "cosine_similarity": 0.11898198868636516,
+      "epoch": 2.277726001863933,
+      "grad_norm": 1.0740425015620436,
+      "learning_rate": 5.350362443907491e-06,
+      "loss": 1.5823,
+      "step": 2444
+    },
+    {
+      "cosine_similarity": 0.18130769073420683,
+      "epoch": 2.2786579683131407,
+      "grad_norm": 1.1244257368282582,
+      "learning_rate": 5.343458750431481e-06,
+      "loss": 1.7473,
+      "step": 2445
+    },
+    {
+      "cosine_similarity": 0.016104657035050637,
+      "epoch": 2.2795899347623485,
+      "grad_norm": 1.1996979219213446,
+      "learning_rate": 5.336555056955471e-06,
+      "loss": 1.6334,
+      "step": 2446
+    },
+    {
+      "cosine_similarity": 0.04955709206934533,
+      "epoch": 2.2805219012115563,
+      "grad_norm": 1.2845201676065157,
+      "learning_rate": 5.329651363479462e-06,
+      "loss": 1.77,
+      "step": 2447
+    },
+    {
+      "cosine_similarity": 0.2217773704457782,
+      "epoch": 2.281453867660764,
+      "grad_norm": 1.2663140015859986,
+      "learning_rate": 5.322747670003452e-06,
+      "loss": 1.7409,
+      "step": 2448
+    },
+    {
+      "cosine_similarity": 0.2661500411156526,
+      "epoch": 2.282385834109972,
+      "grad_norm": 1.366535170761548,
+      "learning_rate": 5.315843976527443e-06,
+      "loss": 1.5142,
+      "step": 2449
+    },
+    {
+      "cosine_similarity": 0.10508706221652435,
+      "epoch": 2.28331780055918,
+      "grad_norm": 1.3325362072816562,
+      "learning_rate": 5.3089402830514325e-06,
+      "loss": 1.744,
+      "step": 2450
+    },
+    {
+      "cosine_similarity": 0.05311150298447941,
+      "epoch": 2.284249767008388,
+      "grad_norm": 1.120919068957598,
+      "learning_rate": 5.302036589575423e-06,
+      "loss": 1.5828,
+      "step": 2451
+    },
+    {
+      "cosine_similarity": 0.10589332441042466,
+      "epoch": 2.2851817334575957,
+      "grad_norm": 1.2250930038628791,
+      "learning_rate": 5.295132896099414e-06,
+      "loss": 1.4258,
+      "step": 2452
+    },
+    {
+      "cosine_similarity": 0.1111900103267423,
+      "epoch": 2.2861136999068035,
+      "grad_norm": 1.0145296602621143,
+      "learning_rate": 5.288229202623404e-06,
+      "loss": 1.3652,
+      "step": 2453
+    },
+    {
+      "cosine_similarity": 0.02420107373233089,
+      "epoch": 2.2870456663560113,
+      "grad_norm": 1.1614138159820797,
+      "learning_rate": 5.281325509147395e-06,
+      "loss": 1.6513,
+      "step": 2454
+    },
+    {
+      "cosine_similarity": -0.03566659492619705,
+      "epoch": 2.287977632805219,
+      "grad_norm": 1.271921141862331,
+      "learning_rate": 5.2744218156713845e-06,
+      "loss": 1.6881,
+      "step": 2455
+    },
+    {
+      "cosine_similarity": -0.01825007295934364,
+      "epoch": 2.288909599254427,
+      "grad_norm": 1.1083664918038818,
+      "learning_rate": 5.267518122195375e-06,
+      "loss": 1.5806,
+      "step": 2456
+    },
+    {
+      "cosine_similarity": -0.011229082657491684,
+      "epoch": 2.2898415657036346,
+      "grad_norm": 1.1989306993369662,
+      "learning_rate": 5.260614428719365e-06,
+      "loss": 1.6509,
+      "step": 2457
+    },
+    {
+      "cosine_similarity": 0.1771802350999973,
+      "epoch": 2.2907735321528424,
+      "grad_norm": 1.126117549783366,
+      "learning_rate": 5.253710735243356e-06,
+      "loss": 1.3515,
+      "step": 2458
+    },
+    {
+      "cosine_similarity": 0.2286188801573853,
+      "epoch": 2.29170549860205,
+      "grad_norm": 1.2881247053865805,
+      "learning_rate": 5.246807041767346e-06,
+      "loss": 1.3541,
+      "step": 2459
+    },
+    {
+      "cosine_similarity": 0.07207194154540127,
+      "epoch": 2.292637465051258,
+      "grad_norm": 1.2242062557141538,
+      "learning_rate": 5.2399033482913364e-06,
+      "loss": 1.3978,
+      "step": 2460
+    },
+    {
+      "cosine_similarity": 0.07622743400935507,
+      "epoch": 2.293569431500466,
+      "grad_norm": 1.0536728475895976,
+      "learning_rate": 5.232999654815326e-06,
+      "loss": 1.4763,
+      "step": 2461
+    },
+    {
+      "cosine_similarity": 0.12523551562758645,
+      "epoch": 2.294501397949674,
+      "grad_norm": 1.2622122294104328,
+      "learning_rate": 5.226095961339317e-06,
+      "loss": 1.3237,
+      "step": 2462
+    },
+    {
+      "cosine_similarity": 0.2441060564821483,
+      "epoch": 2.295433364398882,
+      "grad_norm": 1.3427390890760114,
+      "learning_rate": 5.219192267863308e-06,
+      "loss": 1.3532,
+      "step": 2463
+    },
+    {
+      "cosine_similarity": 0.25943394046728085,
+      "epoch": 2.2963653308480896,
+      "grad_norm": 1.070097865368194,
+      "learning_rate": 5.212288574387298e-06,
+      "loss": 1.4286,
+      "step": 2464
+    },
+    {
+      "cosine_similarity": 0.12420603584078317,
+      "epoch": 2.2972972972972974,
+      "grad_norm": 1.1944796640905524,
+      "learning_rate": 5.205384880911288e-06,
+      "loss": 1.7991,
+      "step": 2465
+    },
+    {
+      "cosine_similarity": 0.037869137159636405,
+      "epoch": 2.298229263746505,
+      "grad_norm": 1.0856849194427818,
+      "learning_rate": 5.198481187435278e-06,
+      "loss": 1.4378,
+      "step": 2466
+    },
+    {
+      "cosine_similarity": -0.04375099650975208,
+      "epoch": 2.299161230195713,
+      "grad_norm": 1.2809190108902409,
+      "learning_rate": 5.191577493959269e-06,
+      "loss": 1.3612,
+      "step": 2467
+    },
+    {
+      "cosine_similarity": 0.006210424094262228,
+      "epoch": 2.3000931966449207,
+      "grad_norm": 1.2600837850674338,
+      "learning_rate": 5.184673800483259e-06,
+      "loss": 1.617,
+      "step": 2468
+    },
+    {
+      "cosine_similarity": 0.12677176172369922,
+      "epoch": 2.3010251630941285,
+      "grad_norm": 1.0193053939938774,
+      "learning_rate": 5.17777010700725e-06,
+      "loss": 1.4331,
+      "step": 2469
+    },
+    {
+      "cosine_similarity": 0.1892818930061387,
+      "epoch": 2.3019571295433363,
+      "grad_norm": 1.2825452861839288,
+      "learning_rate": 5.1708664135312395e-06,
+      "loss": 1.5779,
+      "step": 2470
+    },
+    {
+      "cosine_similarity": 0.0670860378511708,
+      "epoch": 2.302889095992544,
+      "grad_norm": 1.1279342351302828,
+      "learning_rate": 5.16396272005523e-06,
+      "loss": 1.6083,
+      "step": 2471
+    },
+    {
+      "cosine_similarity": -0.08282639394377932,
+      "epoch": 2.3038210624417523,
+      "grad_norm": 1.1194063925431055,
+      "learning_rate": 5.15705902657922e-06,
+      "loss": 1.6041,
+      "step": 2472
+    },
+    {
+      "cosine_similarity": -0.015940588397877316,
+      "epoch": 2.3047530288909597,
+      "grad_norm": 1.066276145197786,
+      "learning_rate": 5.150155333103211e-06,
+      "loss": 1.4823,
+      "step": 2473
+    },
+    {
+      "cosine_similarity": 0.16571771656282658,
+      "epoch": 2.305684995340168,
+      "grad_norm": 1.2497905295427252,
+      "learning_rate": 5.143251639627202e-06,
+      "loss": 1.4288,
+      "step": 2474
+    },
+    {
+      "cosine_similarity": 0.15314932962612576,
+      "epoch": 2.3066169617893757,
+      "grad_norm": 1.053666129210027,
+      "learning_rate": 5.1363479461511915e-06,
+      "loss": 1.5423,
+      "step": 2475
+    },
+    {
+      "cosine_similarity": 0.1490389444366056,
+      "epoch": 2.3075489282385835,
+      "grad_norm": 0.990796191638096,
+      "learning_rate": 5.129444252675182e-06,
+      "loss": 1.5432,
+      "step": 2476
+    },
+    {
+      "cosine_similarity": 0.03487396653290785,
+      "epoch": 2.3084808946877913,
+      "grad_norm": 1.2310623324215435,
+      "learning_rate": 5.122540559199172e-06,
+      "loss": 1.4288,
+      "step": 2477
+    },
+    {
+      "cosine_similarity": 0.03103356454777282,
+      "epoch": 2.309412861136999,
+      "grad_norm": 1.2358934136196273,
+      "learning_rate": 5.115636865723163e-06,
+      "loss": 1.5154,
+      "step": 2478
+    },
+    {
+      "cosine_similarity": 0.07224830100785766,
+      "epoch": 2.310344827586207,
+      "grad_norm": 1.2533786294363947,
+      "learning_rate": 5.108733172247152e-06,
+      "loss": 1.4807,
+      "step": 2479
+    },
+    {
+      "cosine_similarity": 0.12251155905652365,
+      "epoch": 2.3112767940354146,
+      "grad_norm": 1.1392715259065131,
+      "learning_rate": 5.1018294787711435e-06,
+      "loss": 1.168,
+      "step": 2480
+    },
+    {
+      "cosine_similarity": 0.10132367224920642,
+      "epoch": 2.3122087604846224,
+      "grad_norm": 1.3232882102185568,
+      "learning_rate": 5.0949257852951325e-06,
+      "loss": 1.7522,
+      "step": 2481
+    },
+    {
+      "cosine_similarity": 0.15507763730208904,
+      "epoch": 2.31314072693383,
+      "grad_norm": 1.1168450299940325,
+      "learning_rate": 5.088022091819124e-06,
+      "loss": 1.3374,
+      "step": 2482
+    },
+    {
+      "cosine_similarity": -0.120302744886394,
+      "epoch": 2.3140726933830384,
+      "grad_norm": 1.2249995397816789,
+      "learning_rate": 5.081118398343115e-06,
+      "loss": 1.4225,
+      "step": 2483
+    },
+    {
+      "cosine_similarity": 0.011594454709094565,
+      "epoch": 2.315004659832246,
+      "grad_norm": 1.2212397481969242,
+      "learning_rate": 5.074214704867104e-06,
+      "loss": 1.4392,
+      "step": 2484
+    },
+    {
+      "cosine_similarity": 0.13817071391858426,
+      "epoch": 2.315936626281454,
+      "grad_norm": 1.1416054455452975,
+      "learning_rate": 5.0673110113910954e-06,
+      "loss": 1.5682,
+      "step": 2485
+    },
+    {
+      "cosine_similarity": -0.03322883362939465,
+      "epoch": 2.316868592730662,
+      "grad_norm": 1.2563584577489986,
+      "learning_rate": 5.0604073179150845e-06,
+      "loss": 1.4347,
+      "step": 2486
+    },
+    {
+      "cosine_similarity": 0.2200177137095743,
+      "epoch": 2.3178005591798696,
+      "grad_norm": 1.1151587929081712,
+      "learning_rate": 5.053503624439075e-06,
+      "loss": 1.585,
+      "step": 2487
+    },
+    {
+      "cosine_similarity": 0.11011197003641505,
+      "epoch": 2.3187325256290774,
+      "grad_norm": 1.425970518430547,
+      "learning_rate": 5.046599930963065e-06,
+      "loss": 1.5816,
+      "step": 2488
+    },
+    {
+      "cosine_similarity": 0.011756490666511939,
+      "epoch": 2.319664492078285,
+      "grad_norm": 1.3130014478028373,
+      "learning_rate": 5.039696237487056e-06,
+      "loss": 1.4726,
+      "step": 2489
+    },
+    {
+      "cosine_similarity": -0.030458891453618978,
+      "epoch": 2.320596458527493,
+      "grad_norm": 1.2229973901653648,
+      "learning_rate": 5.032792544011046e-06,
+      "loss": 1.6399,
+      "step": 2490
+    },
+    {
+      "cosine_similarity": 0.010643033938730749,
+      "epoch": 2.3215284249767008,
+      "grad_norm": 1.197164049630769,
+      "learning_rate": 5.0258888505350364e-06,
+      "loss": 1.9601,
+      "step": 2491
+    },
+    {
+      "cosine_similarity": 0.032150460207443894,
+      "epoch": 2.3224603914259085,
+      "grad_norm": 1.155894194317279,
+      "learning_rate": 5.018985157059026e-06,
+      "loss": 1.6669,
+      "step": 2492
+    },
+    {
+      "cosine_similarity": 0.0663985478536235,
+      "epoch": 2.3233923578751163,
+      "grad_norm": 1.2080226098848184,
+      "learning_rate": 5.012081463583017e-06,
+      "loss": 1.7119,
+      "step": 2493
+    },
+    {
+      "cosine_similarity": 0.07688608033115499,
+      "epoch": 2.3243243243243246,
+      "grad_norm": 1.224194854656665,
+      "learning_rate": 5.005177770107008e-06,
+      "loss": 1.6685,
+      "step": 2494
+    },
+    {
+      "cosine_similarity": -0.021849743721043095,
+      "epoch": 2.325256290773532,
+      "grad_norm": 1.0732309991855205,
+      "learning_rate": 4.998274076630998e-06,
+      "loss": 1.5626,
+      "step": 2495
+    },
+    {
+      "cosine_similarity": 0.02024243506921931,
+      "epoch": 2.32618825722274,
+      "grad_norm": 1.1784294229541907,
+      "learning_rate": 4.991370383154988e-06,
+      "loss": 1.4065,
+      "step": 2496
+    },
+    {
+      "cosine_similarity": -0.11651805445687917,
+      "epoch": 2.327120223671948,
+      "grad_norm": 1.1865841670442194,
+      "learning_rate": 4.984466689678978e-06,
+      "loss": 1.3332,
+      "step": 2497
+    },
+    {
+      "cosine_similarity": -0.14395612883010678,
+      "epoch": 2.3280521901211557,
+      "grad_norm": 1.3149371319654288,
+      "learning_rate": 4.977562996202969e-06,
+      "loss": 1.669,
+      "step": 2498
+    },
+    {
+      "cosine_similarity": 0.09206708896038074,
+      "epoch": 2.3289841565703635,
+      "grad_norm": 1.3810140758409781,
+      "learning_rate": 4.97065930272696e-06,
+      "loss": 1.4258,
+      "step": 2499
+    },
+    {
+      "cosine_similarity": 0.04110783802477544,
+      "epoch": 2.3299161230195713,
+      "grad_norm": 1.2759956956271983,
+      "learning_rate": 4.96375560925095e-06,
+      "loss": 1.4271,
+      "step": 2500
+    },
+    {
+      "cosine_similarity": 0.06332180719528349,
+      "epoch": 2.330848089468779,
+      "grad_norm": 1.402695930141495,
+      "learning_rate": 4.95685191577494e-06,
+      "loss": 1.4531,
+      "step": 2501
+    },
+    {
+      "cosine_similarity": 0.10730756336452243,
+      "epoch": 2.331780055917987,
+      "grad_norm": 1.3855439595845709,
+      "learning_rate": 4.94994822229893e-06,
+      "loss": 1.5043,
+      "step": 2502
+    },
+    {
+      "cosine_similarity": 0.09891015120136418,
+      "epoch": 2.3327120223671947,
+      "grad_norm": 1.2817197923398593,
+      "learning_rate": 4.943044528822921e-06,
+      "loss": 1.539,
+      "step": 2503
+    },
+    {
+      "cosine_similarity": -0.04411869539430975,
+      "epoch": 2.3336439888164024,
+      "grad_norm": 1.3987479228324393,
+      "learning_rate": 4.936140835346911e-06,
+      "loss": 1.3161,
+      "step": 2504
+    },
+    {
+      "cosine_similarity": 0.06472144987511783,
+      "epoch": 2.3345759552656107,
+      "grad_norm": 1.1096391184517607,
+      "learning_rate": 4.929237141870902e-06,
+      "loss": 1.4411,
+      "step": 2505
+    },
+    {
+      "cosine_similarity": 0.057236562527111394,
+      "epoch": 2.335507921714818,
+      "grad_norm": 1.2998581511912546,
+      "learning_rate": 4.9223334483948915e-06,
+      "loss": 1.375,
+      "step": 2506
+    },
+    {
+      "cosine_similarity": 0.02734060770597253,
+      "epoch": 2.3364398881640263,
+      "grad_norm": 1.2419874816630705,
+      "learning_rate": 4.915429754918881e-06,
+      "loss": 1.482,
+      "step": 2507
+    },
+    {
+      "cosine_similarity": 0.1421513184935593,
+      "epoch": 2.337371854613234,
+      "grad_norm": 1.3396138296808244,
+      "learning_rate": 4.908526061442873e-06,
+      "loss": 1.5378,
+      "step": 2508
+    },
+    {
+      "cosine_similarity": -0.02756057724494215,
+      "epoch": 2.338303821062442,
+      "grad_norm": 1.1419777729449971,
+      "learning_rate": 4.901622367966863e-06,
+      "loss": 1.7831,
+      "step": 2509
+    },
+    {
+      "cosine_similarity": -0.0481650750057585,
+      "epoch": 2.3392357875116496,
+      "grad_norm": 1.1480363217871474,
+      "learning_rate": 4.894718674490853e-06,
+      "loss": 1.2868,
+      "step": 2510
+    },
+    {
+      "cosine_similarity": 0.05887650786838049,
+      "epoch": 2.3401677539608574,
+      "grad_norm": 1.3020741047132411,
+      "learning_rate": 4.8878149810148435e-06,
+      "loss": 1.5739,
+      "step": 2511
+    },
+    {
+      "cosine_similarity": -0.11293918880074831,
+      "epoch": 2.341099720410065,
+      "grad_norm": 1.1909126187001278,
+      "learning_rate": 4.880911287538833e-06,
+      "loss": 1.3827,
+      "step": 2512
+    },
+    {
+      "cosine_similarity": 0.043578558703963384,
+      "epoch": 2.342031686859273,
+      "grad_norm": 1.1699863801827617,
+      "learning_rate": 4.874007594062824e-06,
+      "loss": 1.2963,
+      "step": 2513
+    },
+    {
+      "cosine_similarity": 0.0032575491795184867,
+      "epoch": 2.3429636533084808,
+      "grad_norm": 1.047827641343813,
+      "learning_rate": 4.867103900586814e-06,
+      "loss": 1.4542,
+      "step": 2514
+    },
+    {
+      "cosine_similarity": 0.11240127879334626,
+      "epoch": 2.3438956197576886,
+      "grad_norm": 1.2851130474168402,
+      "learning_rate": 4.860200207110805e-06,
+      "loss": 1.5183,
+      "step": 2515
+    },
+    {
+      "cosine_similarity": 0.017033059903132048,
+      "epoch": 2.344827586206897,
+      "grad_norm": 1.4373858561900932,
+      "learning_rate": 4.853296513634795e-06,
+      "loss": 1.4999,
+      "step": 2516
+    },
+    {
+      "cosine_similarity": 0.12546028187686808,
+      "epoch": 2.345759552656104,
+      "grad_norm": 1.0865717470659562,
+      "learning_rate": 4.846392820158785e-06,
+      "loss": 1.6996,
+      "step": 2517
+    },
+    {
+      "cosine_similarity": -0.0211379083678489,
+      "epoch": 2.3466915191053124,
+      "grad_norm": 1.1213347674780159,
+      "learning_rate": 4.839489126682776e-06,
+      "loss": 1.3605,
+      "step": 2518
+    },
+    {
+      "cosine_similarity": 0.03248660015728675,
+      "epoch": 2.34762348555452,
+      "grad_norm": 1.0403620325169005,
+      "learning_rate": 4.832585433206766e-06,
+      "loss": 1.385,
+      "step": 2519
+    },
+    {
+      "cosine_similarity": -0.04789090331491651,
+      "epoch": 2.348555452003728,
+      "grad_norm": 1.2176624243686736,
+      "learning_rate": 4.825681739730757e-06,
+      "loss": 1.3537,
+      "step": 2520
+    },
+    {
+      "cosine_similarity": 0.14958977628235942,
+      "epoch": 2.3494874184529357,
+      "grad_norm": 1.2952606951238055,
+      "learning_rate": 4.8187780462547465e-06,
+      "loss": 1.467,
+      "step": 2521
+    },
+    {
+      "cosine_similarity": 0.14816405719139886,
+      "epoch": 2.3504193849021435,
+      "grad_norm": 1.1130648834443755,
+      "learning_rate": 4.811874352778737e-06,
+      "loss": 1.0771,
+      "step": 2522
+    },
+    {
+      "cosine_similarity": 0.15711326644254764,
+      "epoch": 2.3513513513513513,
+      "grad_norm": 1.0360434501811393,
+      "learning_rate": 4.804970659302727e-06,
+      "loss": 1.8195,
+      "step": 2523
+    },
+    {
+      "cosine_similarity": 0.11037807334500968,
+      "epoch": 2.352283317800559,
+      "grad_norm": 1.2439184243126429,
+      "learning_rate": 4.798066965826718e-06,
+      "loss": 1.7721,
+      "step": 2524
+    },
+    {
+      "cosine_similarity": 0.17764095881587588,
+      "epoch": 2.353215284249767,
+      "grad_norm": 1.102637250035473,
+      "learning_rate": 4.791163272350708e-06,
+      "loss": 1.6562,
+      "step": 2525
+    },
+    {
+      "cosine_similarity": 0.07080071558239195,
+      "epoch": 2.3541472506989747,
+      "grad_norm": 1.1408407962300564,
+      "learning_rate": 4.7842595788746985e-06,
+      "loss": 1.3981,
+      "step": 2526
+    },
+    {
+      "cosine_similarity": -0.04946201146552058,
+      "epoch": 2.355079217148183,
+      "grad_norm": 1.150967573607901,
+      "learning_rate": 4.777355885398688e-06,
+      "loss": 1.5504,
+      "step": 2527
+    },
+    {
+      "cosine_similarity": 0.03602975345908873,
+      "epoch": 2.3560111835973903,
+      "grad_norm": 1.2653651797949548,
+      "learning_rate": 4.770452191922679e-06,
+      "loss": 1.5106,
+      "step": 2528
+    },
+    {
+      "cosine_similarity": -0.2359435771662035,
+      "epoch": 2.3569431500465985,
+      "grad_norm": 1.484254517689889,
+      "learning_rate": 4.76354849844667e-06,
+      "loss": 1.6553,
+      "step": 2529
+    },
+    {
+      "cosine_similarity": 0.025474045052547637,
+      "epoch": 2.3578751164958063,
+      "grad_norm": 1.4042239748503444,
+      "learning_rate": 4.75664480497066e-06,
+      "loss": 1.6496,
+      "step": 2530
+    },
+    {
+      "cosine_similarity": 0.13012600966030685,
+      "epoch": 2.358807082945014,
+      "grad_norm": 1.0363216557315584,
+      "learning_rate": 4.7497411114946505e-06,
+      "loss": 1.3356,
+      "step": 2531
+    },
+    {
+      "cosine_similarity": 0.024928019948621812,
+      "epoch": 2.359739049394222,
+      "grad_norm": 1.1830434849983573,
+      "learning_rate": 4.74283741801864e-06,
+      "loss": 1.4705,
+      "step": 2532
+    },
+    {
+      "cosine_similarity": 0.06964589282583405,
+      "epoch": 2.3606710158434296,
+      "grad_norm": 1.1713667992794476,
+      "learning_rate": 4.73593372454263e-06,
+      "loss": 1.4051,
+      "step": 2533
+    },
+    {
+      "cosine_similarity": 0.19745872732316008,
+      "epoch": 2.3616029822926374,
+      "grad_norm": 1.2672804103648123,
+      "learning_rate": 4.729030031066621e-06,
+      "loss": 1.5956,
+      "step": 2534
+    },
+    {
+      "cosine_similarity": -0.006457995248640105,
+      "epoch": 2.362534948741845,
+      "grad_norm": 1.3698645010950563,
+      "learning_rate": 4.722126337590611e-06,
+      "loss": 1.246,
+      "step": 2535
+    },
+    {
+      "cosine_similarity": 0.16207875681940065,
+      "epoch": 2.363466915191053,
+      "grad_norm": 1.1794321794848197,
+      "learning_rate": 4.715222644114602e-06,
+      "loss": 1.532,
+      "step": 2536
+    },
+    {
+      "cosine_similarity": -0.08629560493351407,
+      "epoch": 2.364398881640261,
+      "grad_norm": 1.2891580189011151,
+      "learning_rate": 4.7083189506385915e-06,
+      "loss": 1.3691,
+      "step": 2537
+    },
+    {
+      "cosine_similarity": 0.2391494577625243,
+      "epoch": 2.3653308480894686,
+      "grad_norm": 1.3896257846539173,
+      "learning_rate": 4.701415257162582e-06,
+      "loss": 1.4151,
+      "step": 2538
+    },
+    {
+      "cosine_similarity": 0.01775317655551694,
+      "epoch": 2.3662628145386764,
+      "grad_norm": 1.1540443721552494,
+      "learning_rate": 4.694511563686573e-06,
+      "loss": 1.537,
+      "step": 2539
+    },
+    {
+      "cosine_similarity": 0.07221422047135133,
+      "epoch": 2.3671947809878846,
+      "grad_norm": 1.0728287738186582,
+      "learning_rate": 4.687607870210563e-06,
+      "loss": 1.544,
+      "step": 2540
+    },
+    {
+      "cosine_similarity": 0.012119703814975425,
+      "epoch": 2.3681267474370924,
+      "grad_norm": 1.1347905409741226,
+      "learning_rate": 4.6807041767345536e-06,
+      "loss": 2.0713,
+      "step": 2541
+    },
+    {
+      "cosine_similarity": -0.07474387894235598,
+      "epoch": 2.3690587138863,
+      "grad_norm": 1.2080847447387262,
+      "learning_rate": 4.6738004832585435e-06,
+      "loss": 1.2422,
+      "step": 2542
+    },
+    {
+      "cosine_similarity": 0.0775076036507194,
+      "epoch": 2.369990680335508,
+      "grad_norm": 1.4145631260869538,
+      "learning_rate": 4.666896789782534e-06,
+      "loss": 1.6524,
+      "step": 2543
+    },
+    {
+      "cosine_similarity": -0.03518346291362403,
+      "epoch": 2.3709226467847158,
+      "grad_norm": 1.1294451167660529,
+      "learning_rate": 4.659993096306524e-06,
+      "loss": 1.511,
+      "step": 2544
+    },
+    {
+      "cosine_similarity": -0.09800907638493044,
+      "epoch": 2.3718546132339235,
+      "grad_norm": 1.2044578847460272,
+      "learning_rate": 4.653089402830515e-06,
+      "loss": 1.2688,
+      "step": 2545
+    },
+    {
+      "cosine_similarity": -0.007919346727008668,
+      "epoch": 2.3727865796831313,
+      "grad_norm": 1.2695456527126017,
+      "learning_rate": 4.646185709354505e-06,
+      "loss": 1.4775,
+      "step": 2546
+    },
+    {
+      "cosine_similarity": 0.33910861426600886,
+      "epoch": 2.373718546132339,
+      "grad_norm": 1.1336985808083013,
+      "learning_rate": 4.6392820158784954e-06,
+      "loss": 1.2978,
+      "step": 2547
+    },
+    {
+      "cosine_similarity": -0.018947437932067036,
+      "epoch": 2.374650512581547,
+      "grad_norm": 1.0386108470240625,
+      "learning_rate": 4.632378322402486e-06,
+      "loss": 1.2538,
+      "step": 2548
+    },
+    {
+      "cosine_similarity": 0.07066519206042342,
+      "epoch": 2.3755824790307547,
+      "grad_norm": 1.2708709957418292,
+      "learning_rate": 4.625474628926476e-06,
+      "loss": 1.5736,
+      "step": 2549
+    },
+    {
+      "cosine_similarity": 0.20213259304145254,
+      "epoch": 2.3765144454799625,
+      "grad_norm": 1.0800663071708976,
+      "learning_rate": 4.618570935450467e-06,
+      "loss": 1.2115,
+      "step": 2550
+    },
+    {
+      "cosine_similarity": 0.11342422469273533,
+      "epoch": 2.3774464119291707,
+      "grad_norm": 1.2219362750455496,
+      "learning_rate": 4.611667241974457e-06,
+      "loss": 1.4676,
+      "step": 2551
+    },
+    {
+      "cosine_similarity": 0.07762880726331058,
+      "epoch": 2.3783783783783785,
+      "grad_norm": 1.3625596889488227,
+      "learning_rate": 4.604763548498447e-06,
+      "loss": 1.333,
+      "step": 2552
+    },
+    {
+      "cosine_similarity": 0.06644584823580298,
+      "epoch": 2.3793103448275863,
+      "grad_norm": 1.0813738472915646,
+      "learning_rate": 4.597859855022437e-06,
+      "loss": 1.5054,
+      "step": 2553
+    },
+    {
+      "cosine_similarity": -0.10467743150821804,
+      "epoch": 2.380242311276794,
+      "grad_norm": 1.2025053624461113,
+      "learning_rate": 4.590956161546428e-06,
+      "loss": 1.5831,
+      "step": 2554
+    },
+    {
+      "cosine_similarity": 0.08863481535965209,
+      "epoch": 2.381174277726002,
+      "grad_norm": 1.2934980551424335,
+      "learning_rate": 4.584052468070418e-06,
+      "loss": 1.4854,
+      "step": 2555
+    },
+    {
+      "cosine_similarity": 0.036050152858014654,
+      "epoch": 2.3821062441752097,
+      "grad_norm": 1.1093978786649317,
+      "learning_rate": 4.577148774594408e-06,
+      "loss": 1.4446,
+      "step": 2556
+    },
+    {
+      "cosine_similarity": 0.060532123685742546,
+      "epoch": 2.3830382106244175,
+      "grad_norm": 1.2388355455325757,
+      "learning_rate": 4.5702450811183985e-06,
+      "loss": 1.7007,
+      "step": 2557
+    },
+    {
+      "cosine_similarity": 0.18244969688999887,
+      "epoch": 2.3839701770736252,
+      "grad_norm": 1.33706083756147,
+      "learning_rate": 4.563341387642389e-06,
+      "loss": 1.4132,
+      "step": 2558
+    },
+    {
+      "cosine_similarity": 0.1069352683987913,
+      "epoch": 2.384902143522833,
+      "grad_norm": 1.1979440433817898,
+      "learning_rate": 4.55643769416638e-06,
+      "loss": 1.3557,
+      "step": 2559
+    },
+    {
+      "cosine_similarity": 0.12989280869043351,
+      "epoch": 2.385834109972041,
+      "grad_norm": 1.3336597191730497,
+      "learning_rate": 4.54953400069037e-06,
+      "loss": 1.4015,
+      "step": 2560
+    },
+    {
+      "cosine_similarity": 0.15002121640586463,
+      "epoch": 2.3867660764212486,
+      "grad_norm": 1.3295200733561623,
+      "learning_rate": 4.54263030721436e-06,
+      "loss": 1.596,
+      "step": 2561
+    },
+    {
+      "cosine_similarity": 0.01893308436064819,
+      "epoch": 2.387698042870457,
+      "grad_norm": 1.059380380902673,
+      "learning_rate": 4.5357266137383505e-06,
+      "loss": 1.2263,
+      "step": 2562
+    },
+    {
+      "cosine_similarity": 0.19626268655825696,
+      "epoch": 2.3886300093196646,
+      "grad_norm": 1.1371654325462595,
+      "learning_rate": 4.52882292026234e-06,
+      "loss": 1.5062,
+      "step": 2563
+    },
+    {
+      "cosine_similarity": 0.018285342903296233,
+      "epoch": 2.3895619757688724,
+      "grad_norm": 1.3106862575055518,
+      "learning_rate": 4.521919226786331e-06,
+      "loss": 1.5746,
+      "step": 2564
+    },
+    {
+      "cosine_similarity": 0.12290468182115774,
+      "epoch": 2.39049394221808,
+      "grad_norm": 1.1597205636143506,
+      "learning_rate": 4.515015533310321e-06,
+      "loss": 1.4676,
+      "step": 2565
+    },
+    {
+      "cosine_similarity": 0.16273156074833703,
+      "epoch": 2.391425908667288,
+      "grad_norm": 1.1488527003130946,
+      "learning_rate": 4.508111839834312e-06,
+      "loss": 1.7156,
+      "step": 2566
+    },
+    {
+      "cosine_similarity": 0.09823126240418026,
+      "epoch": 2.392357875116496,
+      "grad_norm": 1.0448396846400383,
+      "learning_rate": 4.501208146358302e-06,
+      "loss": 1.2123,
+      "step": 2567
+    },
+    {
+      "cosine_similarity": 0.10099155472450567,
+      "epoch": 2.3932898415657036,
+      "grad_norm": 1.2525082046599705,
+      "learning_rate": 4.494304452882292e-06,
+      "loss": 1.5041,
+      "step": 2568
+    },
+    {
+      "cosine_similarity": 0.18419135508634446,
+      "epoch": 2.3942218080149114,
+      "grad_norm": 1.3016845965771027,
+      "learning_rate": 4.487400759406283e-06,
+      "loss": 1.649,
+      "step": 2569
+    },
+    {
+      "cosine_similarity": -0.07358822334662106,
+      "epoch": 2.395153774464119,
+      "grad_norm": 1.1647572747517518,
+      "learning_rate": 4.480497065930273e-06,
+      "loss": 1.3618,
+      "step": 2570
+    },
+    {
+      "cosine_similarity": -0.03186738401999288,
+      "epoch": 2.396085740913327,
+      "grad_norm": 1.1513852667562183,
+      "learning_rate": 4.473593372454264e-06,
+      "loss": 1.5518,
+      "step": 2571
+    },
+    {
+      "cosine_similarity": 0.063007149109284,
+      "epoch": 2.3970177073625347,
+      "grad_norm": 1.1461146532970943,
+      "learning_rate": 4.466689678978254e-06,
+      "loss": 1.6952,
+      "step": 2572
+    },
+    {
+      "cosine_similarity": -0.028734503686991646,
+      "epoch": 2.397949673811743,
+      "grad_norm": 1.390332349813879,
+      "learning_rate": 4.459785985502244e-06,
+      "loss": 1.2182,
+      "step": 2573
+    },
+    {
+      "cosine_similarity": 0.16909460467113893,
+      "epoch": 2.3988816402609507,
+      "grad_norm": 1.32508313567355,
+      "learning_rate": 4.452882292026234e-06,
+      "loss": 1.3864,
+      "step": 2574
+    },
+    {
+      "cosine_similarity": 0.12942882241362444,
+      "epoch": 2.3998136067101585,
+      "grad_norm": 1.266599938906296,
+      "learning_rate": 4.445978598550225e-06,
+      "loss": 1.5317,
+      "step": 2575
+    },
+    {
+      "cosine_similarity": -0.120308024779668,
+      "epoch": 2.4007455731593663,
+      "grad_norm": 1.102239730946869,
+      "learning_rate": 4.439074905074215e-06,
+      "loss": 1.4421,
+      "step": 2576
+    },
+    {
+      "cosine_similarity": 0.014639236035403336,
+      "epoch": 2.401677539608574,
+      "grad_norm": 1.1838259571373173,
+      "learning_rate": 4.4321712115982055e-06,
+      "loss": 1.8187,
+      "step": 2577
+    },
+    {
+      "cosine_similarity": -0.18097590581540596,
+      "epoch": 2.402609506057782,
+      "grad_norm": 1.2349581095139999,
+      "learning_rate": 4.4252675181221954e-06,
+      "loss": 1.4341,
+      "step": 2578
+    },
+    {
+      "cosine_similarity": -0.02899171718505072,
+      "epoch": 2.4035414725069897,
+      "grad_norm": 1.2607417605254245,
+      "learning_rate": 4.418363824646186e-06,
+      "loss": 1.4775,
+      "step": 2579
+    },
+    {
+      "cosine_similarity": 0.07158689281217455,
+      "epoch": 2.4044734389561975,
+      "grad_norm": 1.4592061993594396,
+      "learning_rate": 4.411460131170177e-06,
+      "loss": 1.6441,
+      "step": 2580
+    },
+    {
+      "cosine_similarity": -0.030620129897021557,
+      "epoch": 2.4054054054054053,
+      "grad_norm": 1.0240222943366155,
+      "learning_rate": 4.404556437694167e-06,
+      "loss": 1.1349,
+      "step": 2581
+    },
+    {
+      "cosine_similarity": -0.07828461948431467,
+      "epoch": 2.406337371854613,
+      "grad_norm": 1.1175397078243456,
+      "learning_rate": 4.3976527442181575e-06,
+      "loss": 1.3091,
+      "step": 2582
+    },
+    {
+      "cosine_similarity": 0.12317392209956156,
+      "epoch": 2.407269338303821,
+      "grad_norm": 1.2854855902102424,
+      "learning_rate": 4.390749050742147e-06,
+      "loss": 1.4739,
+      "step": 2583
+    },
+    {
+      "cosine_similarity": 0.08601567006739591,
+      "epoch": 2.408201304753029,
+      "grad_norm": 1.4115643391034336,
+      "learning_rate": 4.383845357266137e-06,
+      "loss": 1.4201,
+      "step": 2584
+    },
+    {
+      "cosine_similarity": -0.0676139032865179,
+      "epoch": 2.409133271202237,
+      "grad_norm": 1.2715565996541187,
+      "learning_rate": 4.376941663790128e-06,
+      "loss": 1.516,
+      "step": 2585
+    },
+    {
+      "cosine_similarity": 0.03580063945761315,
+      "epoch": 2.4100652376514446,
+      "grad_norm": 1.2253256646378639,
+      "learning_rate": 4.370037970314118e-06,
+      "loss": 1.4208,
+      "step": 2586
+    },
+    {
+      "cosine_similarity": -0.01977738278461475,
+      "epoch": 2.4109972041006524,
+      "grad_norm": 1.1038625928806296,
+      "learning_rate": 4.363134276838109e-06,
+      "loss": 1.572,
+      "step": 2587
+    },
+    {
+      "cosine_similarity": 0.01317068612139466,
+      "epoch": 2.4119291705498602,
+      "grad_norm": 1.108296254059261,
+      "learning_rate": 4.3562305833620985e-06,
+      "loss": 1.3301,
+      "step": 2588
+    },
+    {
+      "cosine_similarity": -0.009285801948887007,
+      "epoch": 2.412861136999068,
+      "grad_norm": 1.1094490531395762,
+      "learning_rate": 4.349326889886089e-06,
+      "loss": 1.4318,
+      "step": 2589
+    },
+    {
+      "cosine_similarity": -0.07704914187320298,
+      "epoch": 2.413793103448276,
+      "grad_norm": 1.2559791518492909,
+      "learning_rate": 4.34242319641008e-06,
+      "loss": 1.2698,
+      "step": 2590
+    },
+    {
+      "cosine_similarity": 0.08014772568675017,
+      "epoch": 2.4147250698974836,
+      "grad_norm": 1.3540074908928634,
+      "learning_rate": 4.33551950293407e-06,
+      "loss": 1.604,
+      "step": 2591
+    },
+    {
+      "cosine_similarity": 0.14240184079486506,
+      "epoch": 2.4156570363466914,
+      "grad_norm": 1.2401356472667997,
+      "learning_rate": 4.328615809458061e-06,
+      "loss": 1.3355,
+      "step": 2592
+    },
+    {
+      "cosine_similarity": 0.03354329853721655,
+      "epoch": 2.416589002795899,
+      "grad_norm": 1.0916987146213941,
+      "learning_rate": 4.3217121159820505e-06,
+      "loss": 1.6563,
+      "step": 2593
+    },
+    {
+      "cosine_similarity": 0.0004903961482281479,
+      "epoch": 2.417520969245107,
+      "grad_norm": 1.1017328042553731,
+      "learning_rate": 4.314808422506041e-06,
+      "loss": 1.3167,
+      "step": 2594
+    },
+    {
+      "cosine_similarity": 0.17455016488424002,
+      "epoch": 2.418452935694315,
+      "grad_norm": 1.2677364016349988,
+      "learning_rate": 4.307904729030031e-06,
+      "loss": 1.7991,
+      "step": 2595
+    },
+    {
+      "cosine_similarity": 0.016861049157964898,
+      "epoch": 2.419384902143523,
+      "grad_norm": 1.1128655842387907,
+      "learning_rate": 4.301001035554022e-06,
+      "loss": 1.265,
+      "step": 2596
+    },
+    {
+      "cosine_similarity": 0.04938993580950326,
+      "epoch": 2.4203168685927308,
+      "grad_norm": 1.0188662729515332,
+      "learning_rate": 4.294097342078012e-06,
+      "loss": 1.5389,
+      "step": 2597
+    },
+    {
+      "cosine_similarity": -0.0680604797485569,
+      "epoch": 2.4212488350419386,
+      "grad_norm": 1.3353145854479986,
+      "learning_rate": 4.2871936486020025e-06,
+      "loss": 1.6241,
+      "step": 2598
+    },
+    {
+      "cosine_similarity": 0.2315862243276701,
+      "epoch": 2.4221808014911463,
+      "grad_norm": 1.2603748625070297,
+      "learning_rate": 4.280289955125993e-06,
+      "loss": 1.4048,
+      "step": 2599
+    },
+    {
+      "cosine_similarity": 0.159358454447188,
+      "epoch": 2.423112767940354,
+      "grad_norm": 1.217394547702451,
+      "learning_rate": 4.273386261649983e-06,
+      "loss": 1.4993,
+      "step": 2600
+    },
+    {
+      "cosine_similarity": 0.09777348704880118,
+      "epoch": 2.424044734389562,
+      "grad_norm": 1.9554676980707864,
+      "learning_rate": 4.266482568173974e-06,
+      "loss": 1.7381,
+      "step": 2601
+    },
+    {
+      "cosine_similarity": 0.09765952034100807,
+      "epoch": 2.4249767008387697,
+      "grad_norm": 1.1919143432144321,
+      "learning_rate": 4.259578874697964e-06,
+      "loss": 1.4747,
+      "step": 2602
+    },
+    {
+      "cosine_similarity": -0.1408098096257469,
+      "epoch": 2.4259086672879775,
+      "grad_norm": 1.0693891852326036,
+      "learning_rate": 4.2526751812219544e-06,
+      "loss": 1.2804,
+      "step": 2603
+    },
+    {
+      "cosine_similarity": -0.11666829681035193,
+      "epoch": 2.4268406337371853,
+      "grad_norm": 1.153844584422594,
+      "learning_rate": 4.245771487745944e-06,
+      "loss": 1.3657,
+      "step": 2604
+    },
+    {
+      "cosine_similarity": 0.028138114571032928,
+      "epoch": 2.427772600186393,
+      "grad_norm": 1.1254042581876564,
+      "learning_rate": 4.238867794269935e-06,
+      "loss": 1.5409,
+      "step": 2605
+    },
+    {
+      "cosine_similarity": 0.1214336802560877,
+      "epoch": 2.4287045666356013,
+      "grad_norm": 1.2325239403412374,
+      "learning_rate": 4.231964100793925e-06,
+      "loss": 1.6482,
+      "step": 2606
+    },
+    {
+      "cosine_similarity": 0.12621906206626368,
+      "epoch": 2.429636533084809,
+      "grad_norm": 1.199048141850142,
+      "learning_rate": 4.225060407317915e-06,
+      "loss": 1.5027,
+      "step": 2607
+    },
+    {
+      "cosine_similarity": 0.0012218724678032154,
+      "epoch": 2.430568499534017,
+      "grad_norm": 1.4286079510681082,
+      "learning_rate": 4.2181567138419055e-06,
+      "loss": 1.6577,
+      "step": 2608
+    },
+    {
+      "cosine_similarity": 0.11417310493652329,
+      "epoch": 2.4315004659832247,
+      "grad_norm": 1.0670872889478584,
+      "learning_rate": 4.211253020365896e-06,
+      "loss": 1.3135,
+      "step": 2609
+    },
+    {
+      "cosine_similarity": -0.09188534661542276,
+      "epoch": 2.4324324324324325,
+      "grad_norm": 1.3231146140624495,
+      "learning_rate": 4.204349326889887e-06,
+      "loss": 1.624,
+      "step": 2610
+    },
+    {
+      "cosine_similarity": 0.12077514669333625,
+      "epoch": 2.4333643988816402,
+      "grad_norm": 1.184018049664963,
+      "learning_rate": 4.197445633413877e-06,
+      "loss": 1.6451,
+      "step": 2611
+    },
+    {
+      "cosine_similarity": 0.11071326417914579,
+      "epoch": 2.434296365330848,
+      "grad_norm": 1.1946715033457964,
+      "learning_rate": 4.190541939937867e-06,
+      "loss": 1.4873,
+      "step": 2612
+    },
+    {
+      "cosine_similarity": 0.029998594445813948,
+      "epoch": 2.435228331780056,
+      "grad_norm": 1.1065079168632697,
+      "learning_rate": 4.1836382464618575e-06,
+      "loss": 1.2291,
+      "step": 2613
+    },
+    {
+      "cosine_similarity": 0.014504684378075752,
+      "epoch": 2.4361602982292636,
+      "grad_norm": 1.1359731316785153,
+      "learning_rate": 4.176734552985847e-06,
+      "loss": 1.3263,
+      "step": 2614
+    },
+    {
+      "cosine_similarity": -0.12463031940320851,
+      "epoch": 2.4370922646784714,
+      "grad_norm": 1.236117028714913,
+      "learning_rate": 4.169830859509838e-06,
+      "loss": 1.4668,
+      "step": 2615
+    },
+    {
+      "cosine_similarity": 0.04209905735076754,
+      "epoch": 2.438024231127679,
+      "grad_norm": 1.1524100534629262,
+      "learning_rate": 4.162927166033828e-06,
+      "loss": 1.2005,
+      "step": 2616
+    },
+    {
+      "cosine_similarity": 0.04069087223329139,
+      "epoch": 2.4389561975768874,
+      "grad_norm": 1.3265596543138864,
+      "learning_rate": 4.156023472557819e-06,
+      "loss": 1.6394,
+      "step": 2617
+    },
+    {
+      "cosine_similarity": -0.026266564054618444,
+      "epoch": 2.439888164026095,
+      "grad_norm": 1.2291749547566506,
+      "learning_rate": 4.149119779081809e-06,
+      "loss": 1.6096,
+      "step": 2618
+    },
+    {
+      "cosine_similarity": 0.2300266528541254,
+      "epoch": 2.440820130475303,
+      "grad_norm": 1.2998311316236149,
+      "learning_rate": 4.142216085605799e-06,
+      "loss": 1.5783,
+      "step": 2619
+    },
+    {
+      "cosine_similarity": -0.1292329077002568,
+      "epoch": 2.441752096924511,
+      "grad_norm": 1.3027758477653495,
+      "learning_rate": 4.13531239212979e-06,
+      "loss": 1.4549,
+      "step": 2620
+    },
+    {
+      "cosine_similarity": 0.08602010589929451,
+      "epoch": 2.4426840633737186,
+      "grad_norm": 1.113424912217498,
+      "learning_rate": 4.12840869865378e-06,
+      "loss": 1.4547,
+      "step": 2621
+    },
+    {
+      "cosine_similarity": 0.047654105535727895,
+      "epoch": 2.4436160298229264,
+      "grad_norm": 1.1731202154010762,
+      "learning_rate": 4.121505005177771e-06,
+      "loss": 1.4271,
+      "step": 2622
+    },
+    {
+      "cosine_similarity": 0.04762978112449675,
+      "epoch": 2.444547996272134,
+      "grad_norm": 1.1764485344138855,
+      "learning_rate": 4.114601311701761e-06,
+      "loss": 1.4908,
+      "step": 2623
+    },
+    {
+      "cosine_similarity": -0.08079255226138668,
+      "epoch": 2.445479962721342,
+      "grad_norm": 1.0461140324996652,
+      "learning_rate": 4.107697618225751e-06,
+      "loss": 1.2264,
+      "step": 2624
+    },
+    {
+      "cosine_similarity": 0.027709335632464614,
+      "epoch": 2.4464119291705497,
+      "grad_norm": 1.2282674089923686,
+      "learning_rate": 4.100793924749741e-06,
+      "loss": 1.6416,
+      "step": 2625
+    },
+    {
+      "cosine_similarity": 0.04633486534295781,
+      "epoch": 2.4473438956197575,
+      "grad_norm": 1.086613312574748,
+      "learning_rate": 4.093890231273732e-06,
+      "loss": 1.4105,
+      "step": 2626
+    },
+    {
+      "cosine_similarity": -0.19310139486697991,
+      "epoch": 2.4482758620689653,
+      "grad_norm": 1.3278076083983552,
+      "learning_rate": 4.086986537797722e-06,
+      "loss": 1.5787,
+      "step": 2627
+    },
+    {
+      "cosine_similarity": -0.06243663914883341,
+      "epoch": 2.4492078285181735,
+      "grad_norm": 1.2290267190616686,
+      "learning_rate": 4.0800828443217126e-06,
+      "loss": 1.307,
+      "step": 2628
+    },
+    {
+      "cosine_similarity": 0.043395317602988204,
+      "epoch": 2.4501397949673813,
+      "grad_norm": 1.123889300934563,
+      "learning_rate": 4.073179150845703e-06,
+      "loss": 1.2832,
+      "step": 2629
+    },
+    {
+      "cosine_similarity": 0.024793179575927006,
+      "epoch": 2.451071761416589,
+      "grad_norm": 1.2681480769895772,
+      "learning_rate": 4.066275457369693e-06,
+      "loss": 1.6462,
+      "step": 2630
+    },
+    {
+      "cosine_similarity": 0.03242728919469126,
+      "epoch": 2.452003727865797,
+      "grad_norm": 1.1227149528986489,
+      "learning_rate": 4.059371763893684e-06,
+      "loss": 1.6345,
+      "step": 2631
+    },
+    {
+      "cosine_similarity": 0.029064322947811368,
+      "epoch": 2.4529356943150047,
+      "grad_norm": 1.107913873416979,
+      "learning_rate": 4.052468070417674e-06,
+      "loss": 1.3308,
+      "step": 2632
+    },
+    {
+      "cosine_similarity": 0.14140825978986277,
+      "epoch": 2.4538676607642125,
+      "grad_norm": 1.2387278448472443,
+      "learning_rate": 4.0455643769416645e-06,
+      "loss": 1.545,
+      "step": 2633
+    },
+    {
+      "cosine_similarity": 0.08774048730735096,
+      "epoch": 2.4547996272134203,
+      "grad_norm": 1.0188098495925013,
+      "learning_rate": 4.0386606834656544e-06,
+      "loss": 1.4345,
+      "step": 2634
+    },
+    {
+      "cosine_similarity": -0.06777250491489789,
+      "epoch": 2.455731593662628,
+      "grad_norm": 1.2294575813765514,
+      "learning_rate": 4.031756989989644e-06,
+      "loss": 1.4877,
+      "step": 2635
+    },
+    {
+      "cosine_similarity": 0.05975474684423293,
+      "epoch": 2.456663560111836,
+      "grad_norm": 1.2414499137418238,
+      "learning_rate": 4.024853296513635e-06,
+      "loss": 1.7333,
+      "step": 2636
+    },
+    {
+      "cosine_similarity": 0.1893723701730852,
+      "epoch": 2.4575955265610436,
+      "grad_norm": 1.3319766914028905,
+      "learning_rate": 4.017949603037625e-06,
+      "loss": 1.5285,
+      "step": 2637
+    },
+    {
+      "cosine_similarity": 0.13732411211292553,
+      "epoch": 2.4585274930102514,
+      "grad_norm": 1.6599869952322437,
+      "learning_rate": 4.011045909561616e-06,
+      "loss": 1.3195,
+      "step": 2638
+    },
+    {
+      "cosine_similarity": 0.0018471395236005048,
+      "epoch": 2.4594594594594597,
+      "grad_norm": 1.0196670073153398,
+      "learning_rate": 4.004142216085606e-06,
+      "loss": 1.4402,
+      "step": 2639
+    },
+    {
+      "cosine_similarity": 0.11610225045363307,
+      "epoch": 2.4603914259086674,
+      "grad_norm": 1.4081390682154118,
+      "learning_rate": 3.997238522609596e-06,
+      "loss": 1.4291,
+      "step": 2640
+    },
+    {
+      "cosine_similarity": 0.026943832113763753,
+      "epoch": 2.4613233923578752,
+      "grad_norm": 1.1467678797262444,
+      "learning_rate": 3.990334829133587e-06,
+      "loss": 1.5586,
+      "step": 2641
+    },
+    {
+      "cosine_similarity": 0.14272774167153834,
+      "epoch": 2.462255358807083,
+      "grad_norm": 1.1976773446466642,
+      "learning_rate": 3.983431135657577e-06,
+      "loss": 1.2929,
+      "step": 2642
+    },
+    {
+      "cosine_similarity": -0.0036398287678018728,
+      "epoch": 2.463187325256291,
+      "grad_norm": 1.0456688701621752,
+      "learning_rate": 3.976527442181568e-06,
+      "loss": 1.4696,
+      "step": 2643
+    },
+    {
+      "cosine_similarity": 0.0955135379827765,
+      "epoch": 2.4641192917054986,
+      "grad_norm": 1.087859334338723,
+      "learning_rate": 3.9696237487055575e-06,
+      "loss": 1.4696,
+      "step": 2644
+    },
+    {
+      "cosine_similarity": -0.08325564699231072,
+      "epoch": 2.4650512581547064,
+      "grad_norm": 1.32494580159811,
+      "learning_rate": 3.962720055229548e-06,
+      "loss": 1.4728,
+      "step": 2645
+    },
+    {
+      "cosine_similarity": -0.1483923248499018,
+      "epoch": 2.465983224603914,
+      "grad_norm": 1.2504959768971733,
+      "learning_rate": 3.955816361753538e-06,
+      "loss": 1.3896,
+      "step": 2646
+    },
+    {
+      "cosine_similarity": -0.0004813145342119689,
+      "epoch": 2.466915191053122,
+      "grad_norm": 1.0035719594226982,
+      "learning_rate": 3.948912668277529e-06,
+      "loss": 1.3671,
+      "step": 2647
+    },
+    {
+      "cosine_similarity": 0.17519895748398567,
+      "epoch": 2.4678471575023297,
+      "grad_norm": 1.3330289630061045,
+      "learning_rate": 3.942008974801519e-06,
+      "loss": 1.3109,
+      "step": 2648
+    },
+    {
+      "cosine_similarity": 0.13049986984193684,
+      "epoch": 2.4687791239515375,
+      "grad_norm": 1.099759510711388,
+      "learning_rate": 3.9351052813255095e-06,
+      "loss": 1.4237,
+      "step": 2649
+    },
+    {
+      "cosine_similarity": -0.10471115721246195,
+      "epoch": 2.4697110904007458,
+      "grad_norm": 1.0660444608702853,
+      "learning_rate": 3.9282015878495e-06,
+      "loss": 1.2463,
+      "step": 2650
+    },
+    {
+      "cosine_similarity": 0.1077901150857817,
+      "epoch": 2.4706430568499536,
+      "grad_norm": 1.3409939269687052,
+      "learning_rate": 3.92129789437349e-06,
+      "loss": 1.4654,
+      "step": 2651
+    },
+    {
+      "cosine_similarity": 0.03238489130137914,
+      "epoch": 2.4715750232991613,
+      "grad_norm": 1.125056904750956,
+      "learning_rate": 3.914394200897481e-06,
+      "loss": 1.5222,
+      "step": 2652
+    },
+    {
+      "cosine_similarity": 0.01564581477788187,
+      "epoch": 2.472506989748369,
+      "grad_norm": 1.1480710818026494,
+      "learning_rate": 3.907490507421471e-06,
+      "loss": 1.534,
+      "step": 2653
+    },
+    {
+      "cosine_similarity": 0.1473860667553795,
+      "epoch": 2.473438956197577,
+      "grad_norm": 1.2567732645820648,
+      "learning_rate": 3.9005868139454615e-06,
+      "loss": 1.4037,
+      "step": 2654
+    },
+    {
+      "cosine_similarity": 0.07155877987847704,
+      "epoch": 2.4743709226467847,
+      "grad_norm": 1.1139376037858706,
+      "learning_rate": 3.893683120469451e-06,
+      "loss": 1.7814,
+      "step": 2655
+    },
+    {
+      "cosine_similarity": 0.174746550679515,
+      "epoch": 2.4753028890959925,
+      "grad_norm": 1.2922997207244682,
+      "learning_rate": 3.886779426993442e-06,
+      "loss": 2.0633,
+      "step": 2656
+    },
+    {
+      "cosine_similarity": -0.06548707107188695,
+      "epoch": 2.4762348555452003,
+      "grad_norm": 1.3430593646826723,
+      "learning_rate": 3.879875733517432e-06,
+      "loss": 1.1832,
+      "step": 2657
+    },
+    {
+      "cosine_similarity": -0.20839401800794613,
+      "epoch": 2.477166821994408,
+      "grad_norm": 1.2971708706083542,
+      "learning_rate": 3.872972040041422e-06,
+      "loss": 1.5109,
+      "step": 2658
+    },
+    {
+      "cosine_similarity": -0.05575709476248112,
+      "epoch": 2.478098788443616,
+      "grad_norm": 1.2664267997038603,
+      "learning_rate": 3.866068346565413e-06,
+      "loss": 1.4569,
+      "step": 2659
+    },
+    {
+      "cosine_similarity": 0.196420516908976,
+      "epoch": 2.4790307548928237,
+      "grad_norm": 1.147251081817075,
+      "learning_rate": 3.859164653089403e-06,
+      "loss": 1.6319,
+      "step": 2660
+    },
+    {
+      "cosine_similarity": 0.24866911542148984,
+      "epoch": 2.479962721342032,
+      "grad_norm": 1.2561929012777568,
+      "learning_rate": 3.852260959613394e-06,
+      "loss": 1.4616,
+      "step": 2661
+    },
+    {
+      "cosine_similarity": -0.040217424973218494,
+      "epoch": 2.4808946877912397,
+      "grad_norm": 1.5984116600075298,
+      "learning_rate": 3.845357266137384e-06,
+      "loss": 1.5589,
+      "step": 2662
+    },
+    {
+      "cosine_similarity": 0.08064669169379102,
+      "epoch": 2.4818266542404475,
+      "grad_norm": 1.4096208355042,
+      "learning_rate": 3.838453572661374e-06,
+      "loss": 1.3958,
+      "step": 2663
+    },
+    {
+      "cosine_similarity": -0.035988086998185335,
+      "epoch": 2.4827586206896552,
+      "grad_norm": 1.1363400518544067,
+      "learning_rate": 3.8315498791853645e-06,
+      "loss": 1.5088,
+      "step": 2664
+    },
+    {
+      "cosine_similarity": 0.008727598440872371,
+      "epoch": 2.483690587138863,
+      "grad_norm": 1.028363711508309,
+      "learning_rate": 3.8246461857093544e-06,
+      "loss": 1.1893,
+      "step": 2665
+    },
+    {
+      "cosine_similarity": 0.3613464008211617,
+      "epoch": 2.484622553588071,
+      "grad_norm": 1.355598085465792,
+      "learning_rate": 3.817742492233345e-06,
+      "loss": 1.4655,
+      "step": 2666
+    },
+    {
+      "cosine_similarity": 0.11964610479221825,
+      "epoch": 2.4855545200372786,
+      "grad_norm": 1.2977090128045643,
+      "learning_rate": 3.8108387987573355e-06,
+      "loss": 1.4386,
+      "step": 2667
+    },
+    {
+      "cosine_similarity": 0.04757256360580174,
+      "epoch": 2.4864864864864864,
+      "grad_norm": 1.1873157823806602,
+      "learning_rate": 3.8039351052813258e-06,
+      "loss": 1.4425,
+      "step": 2668
+    },
+    {
+      "cosine_similarity": -0.1268056759105838,
+      "epoch": 2.487418452935694,
+      "grad_norm": 1.211513807592653,
+      "learning_rate": 3.7970314118053165e-06,
+      "loss": 1.6122,
+      "step": 2669
+    },
+    {
+      "cosine_similarity": 0.06783908167835119,
+      "epoch": 2.488350419384902,
+      "grad_norm": 1.0341287727634545,
+      "learning_rate": 3.790127718329307e-06,
+      "loss": 1.4038,
+      "step": 2670
+    },
+    {
+      "cosine_similarity": 0.03230554186488068,
+      "epoch": 2.4892823858341098,
+      "grad_norm": 1.1199887200910577,
+      "learning_rate": 3.783224024853297e-06,
+      "loss": 1.5883,
+      "step": 2671
+    },
+    {
+      "cosine_similarity": 0.08069514514262081,
+      "epoch": 2.490214352283318,
+      "grad_norm": 1.0342507650335997,
+      "learning_rate": 3.7763203313772874e-06,
+      "loss": 1.2397,
+      "step": 2672
+    },
+    {
+      "cosine_similarity": 0.09971890551261949,
+      "epoch": 2.491146318732526,
+      "grad_norm": 1.2200221368991206,
+      "learning_rate": 3.7694166379012777e-06,
+      "loss": 1.6656,
+      "step": 2673
+    },
+    {
+      "cosine_similarity": 0.04483243506547297,
+      "epoch": 2.4920782851817336,
+      "grad_norm": 1.0828168031378014,
+      "learning_rate": 3.7625129444252676e-06,
+      "loss": 1.4686,
+      "step": 2674
+    },
+    {
+      "cosine_similarity": 0.14868759570145626,
+      "epoch": 2.4930102516309414,
+      "grad_norm": 1.2800839800597634,
+      "learning_rate": 3.755609250949258e-06,
+      "loss": 1.522,
+      "step": 2675
+    },
+    {
+      "cosine_similarity": -0.06528520969753386,
+      "epoch": 2.493942218080149,
+      "grad_norm": 1.2960946266848137,
+      "learning_rate": 3.7487055574732483e-06,
+      "loss": 1.5989,
+      "step": 2676
+    },
+    {
+      "cosine_similarity": 0.055916894331302675,
+      "epoch": 2.494874184529357,
+      "grad_norm": 1.160758010620494,
+      "learning_rate": 3.7418018639972386e-06,
+      "loss": 1.5792,
+      "step": 2677
+    },
+    {
+      "cosine_similarity": 0.01060673440132384,
+      "epoch": 2.4958061509785647,
+      "grad_norm": 1.0833231571853152,
+      "learning_rate": 3.734898170521229e-06,
+      "loss": 1.1489,
+      "step": 2678
+    },
+    {
+      "cosine_similarity": 0.044759678621636184,
+      "epoch": 2.4967381174277725,
+      "grad_norm": 0.9899408318822777,
+      "learning_rate": 3.7279944770452196e-06,
+      "loss": 1.6342,
+      "step": 2679
+    },
+    {
+      "cosine_similarity": 0.12232918661180031,
+      "epoch": 2.4976700838769803,
+      "grad_norm": 1.1991142167257278,
+      "learning_rate": 3.72109078356921e-06,
+      "loss": 1.3923,
+      "step": 2680
+    },
+    {
+      "cosine_similarity": 0.08375300865026085,
+      "epoch": 2.498602050326188,
+      "grad_norm": 1.2375551431532916,
+      "learning_rate": 3.7141870900932002e-06,
+      "loss": 1.4347,
+      "step": 2681
+    },
+    {
+      "cosine_similarity": 0.11205471156041726,
+      "epoch": 2.499534016775396,
+      "grad_norm": 1.1637850423007914,
+      "learning_rate": 3.7072833966171905e-06,
+      "loss": 1.5833,
+      "step": 2682
+    },
+    {
+      "cosine_similarity": -0.05873188477460003,
+      "epoch": 2.500465983224604,
+      "grad_norm": 1.1887320406355573,
+      "learning_rate": 3.700379703141181e-06,
+      "loss": 1.6122,
+      "step": 2683
+    },
+    {
+      "cosine_similarity": 0.042675136910545815,
+      "epoch": 2.501397949673812,
+      "grad_norm": 1.2336515193498545,
+      "learning_rate": 3.693476009665171e-06,
+      "loss": 1.5418,
+      "step": 2684
+    },
+    {
+      "cosine_similarity": 0.08491255168871818,
+      "epoch": 2.5023299161230197,
+      "grad_norm": 1.077233650959817,
+      "learning_rate": 3.6865723161891615e-06,
+      "loss": 1.6203,
+      "step": 2685
+    },
+    {
+      "cosine_similarity": 0.11559143276743572,
+      "epoch": 2.5032618825722275,
+      "grad_norm": 1.0533476538167825,
+      "learning_rate": 3.6796686227131518e-06,
+      "loss": 1.4113,
+      "step": 2686
+    },
+    {
+      "cosine_similarity": 0.03071077454567781,
+      "epoch": 2.5041938490214353,
+      "grad_norm": 1.3314746174702454,
+      "learning_rate": 3.672764929237142e-06,
+      "loss": 1.6871,
+      "step": 2687
+    },
+    {
+      "cosine_similarity": 0.025292747197811976,
+      "epoch": 2.505125815470643,
+      "grad_norm": 1.065955841128825,
+      "learning_rate": 3.6658612357611324e-06,
+      "loss": 1.3644,
+      "step": 2688
+    },
+    {
+      "cosine_similarity": 0.08676665071547436,
+      "epoch": 2.506057781919851,
+      "grad_norm": 1.0737457697285318,
+      "learning_rate": 3.6589575422851227e-06,
+      "loss": 1.2924,
+      "step": 2689
+    },
+    {
+      "cosine_similarity": 0.010855778929297365,
+      "epoch": 2.5069897483690586,
+      "grad_norm": 1.2319007185941977,
+      "learning_rate": 3.6520538488091134e-06,
+      "loss": 1.5995,
+      "step": 2690
+    },
+    {
+      "cosine_similarity": -0.060364203064817365,
+      "epoch": 2.5079217148182664,
+      "grad_norm": 1.2022123332072254,
+      "learning_rate": 3.6451501553331037e-06,
+      "loss": 1.622,
+      "step": 2691
+    },
+    {
+      "cosine_similarity": -0.01600451477264818,
+      "epoch": 2.508853681267474,
+      "grad_norm": 1.3372595417479796,
+      "learning_rate": 3.638246461857094e-06,
+      "loss": 1.6311,
+      "step": 2692
+    },
+    {
+      "cosine_similarity": 0.1301454489436272,
+      "epoch": 2.509785647716682,
+      "grad_norm": 1.1791915809425644,
+      "learning_rate": 3.6313427683810844e-06,
+      "loss": 1.5284,
+      "step": 2693
+    },
+    {
+      "cosine_similarity": -0.052064675645304814,
+      "epoch": 2.5107176141658902,
+      "grad_norm": 1.1607806389101936,
+      "learning_rate": 3.6244390749050747e-06,
+      "loss": 1.4513,
+      "step": 2694
+    },
+    {
+      "cosine_similarity": -0.032375287177116636,
+      "epoch": 2.511649580615098,
+      "grad_norm": 1.1578023071375816,
+      "learning_rate": 3.617535381429065e-06,
+      "loss": 1.7545,
+      "step": 2695
+    },
+    {
+      "cosine_similarity": 0.10546742486929997,
+      "epoch": 2.512581547064306,
+      "grad_norm": 1.086249267636809,
+      "learning_rate": 3.6106316879530553e-06,
+      "loss": 1.5428,
+      "step": 2696
+    },
+    {
+      "cosine_similarity": -0.23611640739931128,
+      "epoch": 2.5135135135135136,
+      "grad_norm": 1.2707442821551596,
+      "learning_rate": 3.603727994477045e-06,
+      "loss": 1.1971,
+      "step": 2697
+    },
+    {
+      "cosine_similarity": 0.03183178219731563,
+      "epoch": 2.5144454799627214,
+      "grad_norm": 1.8296064404458356,
+      "learning_rate": 3.5968243010010355e-06,
+      "loss": 1.826,
+      "step": 2698
+    },
+    {
+      "cosine_similarity": -0.1429798049390814,
+      "epoch": 2.515377446411929,
+      "grad_norm": 1.1117035632329115,
+      "learning_rate": 3.5899206075250258e-06,
+      "loss": 1.2669,
+      "step": 2699
+    },
+    {
+      "cosine_similarity": 0.1452095794752861,
+      "epoch": 2.516309412861137,
+      "grad_norm": 1.5955193169274238,
+      "learning_rate": 3.583016914049017e-06,
+      "loss": 1.4297,
+      "step": 2700
+    },
+    {
+      "cosine_similarity": 0.035704246703812635,
+      "epoch": 2.5172413793103448,
+      "grad_norm": 1.3506834676219277,
+      "learning_rate": 3.576113220573007e-06,
+      "loss": 1.5788,
+      "step": 2701
+    },
+    {
+      "cosine_similarity": 0.09728655330671059,
+      "epoch": 2.5181733457595525,
+      "grad_norm": 1.088082812048027,
+      "learning_rate": 3.569209527096997e-06,
+      "loss": 1.4518,
+      "step": 2702
+    },
+    {
+      "cosine_similarity": 0.1197148670002469,
+      "epoch": 2.5191053122087603,
+      "grad_norm": 1.2495916183793345,
+      "learning_rate": 3.5623058336209874e-06,
+      "loss": 1.5391,
+      "step": 2703
+    },
+    {
+      "cosine_similarity": 0.21525800952493682,
+      "epoch": 2.520037278657968,
+      "grad_norm": 1.211573302743258,
+      "learning_rate": 3.5554021401449777e-06,
+      "loss": 1.3163,
+      "step": 2704
+    },
+    {
+      "cosine_similarity": 0.013308690155064527,
+      "epoch": 2.5209692451071763,
+      "grad_norm": 1.363091794014671,
+      "learning_rate": 3.548498446668968e-06,
+      "loss": 1.3911,
+      "step": 2705
+    },
+    {
+      "cosine_similarity": 0.13674946580265684,
+      "epoch": 2.5219012115563837,
+      "grad_norm": 1.354316181415006,
+      "learning_rate": 3.5415947531929584e-06,
+      "loss": 1.7774,
+      "step": 2706
+    },
+    {
+      "cosine_similarity": 0.10089727996625304,
+      "epoch": 2.522833178005592,
+      "grad_norm": 1.0699358114441115,
+      "learning_rate": 3.5346910597169487e-06,
+      "loss": 1.383,
+      "step": 2707
+    },
+    {
+      "cosine_similarity": -0.05255131403127064,
+      "epoch": 2.5237651444547997,
+      "grad_norm": 1.2069711740574396,
+      "learning_rate": 3.527787366240939e-06,
+      "loss": 1.4906,
+      "step": 2708
+    },
+    {
+      "cosine_similarity": 0.01504654850087514,
+      "epoch": 2.5246971109040075,
+      "grad_norm": 1.2598573377146531,
+      "learning_rate": 3.5208836727649293e-06,
+      "loss": 1.557,
+      "step": 2709
+    },
+    {
+      "cosine_similarity": -0.04126878205754991,
+      "epoch": 2.5256290773532153,
+      "grad_norm": 1.4228615840266332,
+      "learning_rate": 3.51397997928892e-06,
+      "loss": 1.6293,
+      "step": 2710
+    },
+    {
+      "cosine_similarity": 0.31901102460296815,
+      "epoch": 2.526561043802423,
+      "grad_norm": 1.7154306596015043,
+      "learning_rate": 3.5070762858129103e-06,
+      "loss": 1.7292,
+      "step": 2711
+    },
+    {
+      "cosine_similarity": 0.15344623885526235,
+      "epoch": 2.527493010251631,
+      "grad_norm": 1.2097466791113933,
+      "learning_rate": 3.5001725923369006e-06,
+      "loss": 1.3188,
+      "step": 2712
+    },
+    {
+      "cosine_similarity": 0.05333126596518861,
+      "epoch": 2.5284249767008387,
+      "grad_norm": 1.151326595519897,
+      "learning_rate": 3.493268898860891e-06,
+      "loss": 1.7136,
+      "step": 2713
+    },
+    {
+      "cosine_similarity": 0.08211866849409878,
+      "epoch": 2.5293569431500464,
+      "grad_norm": 1.2753313927127015,
+      "learning_rate": 3.4863652053848813e-06,
+      "loss": 1.7243,
+      "step": 2714
+    },
+    {
+      "cosine_similarity": 0.16190997625189324,
+      "epoch": 2.5302889095992542,
+      "grad_norm": 1.2402994909261922,
+      "learning_rate": 3.4794615119088716e-06,
+      "loss": 1.4966,
+      "step": 2715
+    },
+    {
+      "cosine_similarity": -0.08446167922868902,
+      "epoch": 2.5312208760484625,
+      "grad_norm": 1.133432964360887,
+      "learning_rate": 3.472557818432862e-06,
+      "loss": 1.6329,
+      "step": 2716
+    },
+    {
+      "cosine_similarity": 0.050036231104409995,
+      "epoch": 2.53215284249767,
+      "grad_norm": 1.2307160169803153,
+      "learning_rate": 3.465654124956852e-06,
+      "loss": 1.4339,
+      "step": 2717
+    },
+    {
+      "cosine_similarity": -0.04070676615642447,
+      "epoch": 2.533084808946878,
+      "grad_norm": 1.282379662689497,
+      "learning_rate": 3.4587504314808425e-06,
+      "loss": 1.4608,
+      "step": 2718
+    },
+    {
+      "cosine_similarity": -0.0001862844807517108,
+      "epoch": 2.534016775396086,
+      "grad_norm": 1.1580380392625726,
+      "learning_rate": 3.451846738004833e-06,
+      "loss": 1.3392,
+      "step": 2719
+    },
+    {
+      "cosine_similarity": 0.11217367280324606,
+      "epoch": 2.5349487418452936,
+      "grad_norm": 1.038681747538465,
+      "learning_rate": 3.4449430445288235e-06,
+      "loss": 1.4697,
+      "step": 2720
+    },
+    {
+      "cosine_similarity": -0.0005173204521471394,
+      "epoch": 2.5358807082945014,
+      "grad_norm": 1.5440432359925602,
+      "learning_rate": 3.438039351052814e-06,
+      "loss": 1.5092,
+      "step": 2721
+    },
+    {
+      "cosine_similarity": -0.018051191993814095,
+      "epoch": 2.536812674743709,
+      "grad_norm": 1.1833578228190615,
+      "learning_rate": 3.431135657576804e-06,
+      "loss": 1.2592,
+      "step": 2722
+    },
+    {
+      "cosine_similarity": 0.036949768701827666,
+      "epoch": 2.537744641192917,
+      "grad_norm": 1.1875624355445993,
+      "learning_rate": 3.4242319641007945e-06,
+      "loss": 1.5463,
+      "step": 2723
+    },
+    {
+      "cosine_similarity": 0.12015501678625118,
+      "epoch": 2.5386766076421248,
+      "grad_norm": 1.2079259957638477,
+      "learning_rate": 3.4173282706247844e-06,
+      "loss": 1.4254,
+      "step": 2724
+    },
+    {
+      "cosine_similarity": 0.10205793493236086,
+      "epoch": 2.5396085740913326,
+      "grad_norm": 1.1996238453102386,
+      "learning_rate": 3.4104245771487747e-06,
+      "loss": 1.2867,
+      "step": 2725
+    },
+    {
+      "cosine_similarity": 0.07653759650135888,
+      "epoch": 2.5405405405405403,
+      "grad_norm": 1.3814818215532123,
+      "learning_rate": 3.403520883672765e-06,
+      "loss": 1.6044,
+      "step": 2726
+    },
+    {
+      "cosine_similarity": 0.14473931199840862,
+      "epoch": 2.5414725069897486,
+      "grad_norm": 1.22284463729923,
+      "learning_rate": 3.3966171901967553e-06,
+      "loss": 1.6198,
+      "step": 2727
+    },
+    {
+      "cosine_similarity": 0.13436947762139928,
+      "epoch": 2.542404473438956,
+      "grad_norm": 1.2137359561836603,
+      "learning_rate": 3.3897134967207456e-06,
+      "loss": 1.8463,
+      "step": 2728
+    },
+    {
+      "cosine_similarity": 0.1975900021228663,
+      "epoch": 2.543336439888164,
+      "grad_norm": 1.4324250624595911,
+      "learning_rate": 3.382809803244736e-06,
+      "loss": 1.5034,
+      "step": 2729
+    },
+    {
+      "cosine_similarity": 0.4112426821018899,
+      "epoch": 2.544268406337372,
+      "grad_norm": 1.199102839028808,
+      "learning_rate": 3.3759061097687266e-06,
+      "loss": 1.5248,
+      "step": 2730
+    },
+    {
+      "cosine_similarity": 0.03935253462874996,
+      "epoch": 2.5452003727865797,
+      "grad_norm": 1.3697031488003293,
+      "learning_rate": 3.369002416292717e-06,
+      "loss": 1.7183,
+      "step": 2731
+    },
+    {
+      "cosine_similarity": -0.05240722031151399,
+      "epoch": 2.5461323392357875,
+      "grad_norm": 1.2031502692618647,
+      "learning_rate": 3.3620987228167072e-06,
+      "loss": 1.453,
+      "step": 2732
+    },
+    {
+      "cosine_similarity": 0.05903314568664495,
+      "epoch": 2.5470643056849953,
+      "grad_norm": 0.992694770102625,
+      "learning_rate": 3.3551950293406976e-06,
+      "loss": 1.3805,
+      "step": 2733
+    },
+    {
+      "cosine_similarity": 0.21862424693757862,
+      "epoch": 2.547996272134203,
+      "grad_norm": 1.3185543973923426,
+      "learning_rate": 3.348291335864688e-06,
+      "loss": 1.47,
+      "step": 2734
+    },
+    {
+      "cosine_similarity": 0.08228313467671433,
+      "epoch": 2.548928238583411,
+      "grad_norm": 1.1095702952249153,
+      "learning_rate": 3.341387642388678e-06,
+      "loss": 1.494,
+      "step": 2735
+    },
+    {
+      "cosine_similarity": 0.10836896544887624,
+      "epoch": 2.5498602050326187,
+      "grad_norm": 1.2519781882736947,
+      "learning_rate": 3.3344839489126685e-06,
+      "loss": 1.3737,
+      "step": 2736
+    },
+    {
+      "cosine_similarity": 0.11148381234048466,
+      "epoch": 2.5507921714818265,
+      "grad_norm": 1.3262089366940015,
+      "learning_rate": 3.327580255436659e-06,
+      "loss": 1.7372,
+      "step": 2737
+    },
+    {
+      "cosine_similarity": 0.007530171562922127,
+      "epoch": 2.5517241379310347,
+      "grad_norm": 0.9962071912962283,
+      "learning_rate": 3.320676561960649e-06,
+      "loss": 1.3381,
+      "step": 2738
+    },
+    {
+      "cosine_similarity": 0.037656260715014434,
+      "epoch": 2.552656104380242,
+      "grad_norm": 0.9893838395335439,
+      "learning_rate": 3.3137728684846394e-06,
+      "loss": 1.3808,
+      "step": 2739
+    },
+    {
+      "cosine_similarity": 0.02679224014265603,
+      "epoch": 2.5535880708294503,
+      "grad_norm": 1.403327575681813,
+      "learning_rate": 3.30686917500863e-06,
+      "loss": 1.5096,
+      "step": 2740
+    },
+    {
+      "cosine_similarity": -0.004747294008925276,
+      "epoch": 2.554520037278658,
+      "grad_norm": 1.0991490957501522,
+      "learning_rate": 3.2999654815326205e-06,
+      "loss": 1.5135,
+      "step": 2741
+    },
+    {
+      "cosine_similarity": -0.027333302302648747,
+      "epoch": 2.555452003727866,
+      "grad_norm": 1.1856216647215279,
+      "learning_rate": 3.2930617880566108e-06,
+      "loss": 1.354,
+      "step": 2742
+    },
+    {
+      "cosine_similarity": 0.016971025103326362,
+      "epoch": 2.5563839701770736,
+      "grad_norm": 1.141433613037645,
+      "learning_rate": 3.286158094580601e-06,
+      "loss": 1.3996,
+      "step": 2743
+    },
+    {
+      "cosine_similarity": 0.16645043914850496,
+      "epoch": 2.5573159366262814,
+      "grad_norm": 1.2097073407930037,
+      "learning_rate": 3.2792544011045914e-06,
+      "loss": 1.591,
+      "step": 2744
+    },
+    {
+      "cosine_similarity": -0.004983470820398046,
+      "epoch": 2.558247903075489,
+      "grad_norm": 1.1205812135807225,
+      "learning_rate": 3.2723507076285817e-06,
+      "loss": 1.2681,
+      "step": 2745
+    },
+    {
+      "cosine_similarity": 0.20254678347883354,
+      "epoch": 2.559179869524697,
+      "grad_norm": 1.3084281181752355,
+      "learning_rate": 3.265447014152572e-06,
+      "loss": 1.6856,
+      "step": 2746
+    },
+    {
+      "cosine_similarity": 0.14160511067845144,
+      "epoch": 2.560111835973905,
+      "grad_norm": 1.2600179928123083,
+      "learning_rate": 3.258543320676562e-06,
+      "loss": 1.6905,
+      "step": 2747
+    },
+    {
+      "cosine_similarity": -0.00745506011790102,
+      "epoch": 2.5610438024231126,
+      "grad_norm": 0.9847700448206255,
+      "learning_rate": 3.251639627200552e-06,
+      "loss": 1.4677,
+      "step": 2748
+    },
+    {
+      "cosine_similarity": -0.06635639240792668,
+      "epoch": 2.561975768872321,
+      "grad_norm": 1.2398822981251951,
+      "learning_rate": 3.2447359337245425e-06,
+      "loss": 1.5066,
+      "step": 2749
+    },
+    {
+      "cosine_similarity": 0.03102662668488175,
+      "epoch": 2.562907735321528,
+      "grad_norm": 1.1284168799224734,
+      "learning_rate": 3.2378322402485337e-06,
+      "loss": 1.4418,
+      "step": 2750
+    },
+    {
+      "cosine_similarity": 0.09855467420827711,
+      "epoch": 2.5638397017707364,
+      "grad_norm": 1.0376384433170647,
+      "learning_rate": 3.230928546772524e-06,
+      "loss": 1.3139,
+      "step": 2751
+    },
+    {
+      "cosine_similarity": -0.009799523671152737,
+      "epoch": 2.564771668219944,
+      "grad_norm": 1.2263907676672612,
+      "learning_rate": 3.224024853296514e-06,
+      "loss": 1.2716,
+      "step": 2752
+    },
+    {
+      "cosine_similarity": -0.08055528122577764,
+      "epoch": 2.565703634669152,
+      "grad_norm": 1.0636573197086991,
+      "learning_rate": 3.217121159820504e-06,
+      "loss": 1.485,
+      "step": 2753
+    },
+    {
+      "cosine_similarity": -0.07063575245127932,
+      "epoch": 2.5666356011183598,
+      "grad_norm": 1.0815053282730762,
+      "learning_rate": 3.2102174663444945e-06,
+      "loss": 1.4881,
+      "step": 2754
+    },
+    {
+      "cosine_similarity": 0.050530314346029906,
+      "epoch": 2.5675675675675675,
+      "grad_norm": 1.0914371432465426,
+      "learning_rate": 3.2033137728684848e-06,
+      "loss": 1.4111,
+      "step": 2755
+    },
+    {
+      "cosine_similarity": 0.14698574848298426,
+      "epoch": 2.5684995340167753,
+      "grad_norm": 1.1797269221422244,
+      "learning_rate": 3.196410079392475e-06,
+      "loss": 1.5764,
+      "step": 2756
+    },
+    {
+      "cosine_similarity": 0.11705480243485335,
+      "epoch": 2.569431500465983,
+      "grad_norm": 1.115909050743732,
+      "learning_rate": 3.1895063859164654e-06,
+      "loss": 1.2554,
+      "step": 2757
+    },
+    {
+      "cosine_similarity": -0.054217199970157004,
+      "epoch": 2.570363466915191,
+      "grad_norm": 1.3707275293699253,
+      "learning_rate": 3.1826026924404557e-06,
+      "loss": 1.416,
+      "step": 2758
+    },
+    {
+      "cosine_similarity": 0.006317808538668884,
+      "epoch": 2.5712954333643987,
+      "grad_norm": 1.0722799481193588,
+      "learning_rate": 3.175698998964446e-06,
+      "loss": 1.3397,
+      "step": 2759
+    },
+    {
+      "cosine_similarity": 0.06347690067369412,
+      "epoch": 2.572227399813607,
+      "grad_norm": 1.151667131663918,
+      "learning_rate": 3.1687953054884367e-06,
+      "loss": 1.2735,
+      "step": 2760
+    },
+    {
+      "cosine_similarity": -0.021203135912594442,
+      "epoch": 2.5731593662628143,
+      "grad_norm": 1.2746714845192821,
+      "learning_rate": 3.161891612012427e-06,
+      "loss": 1.3883,
+      "step": 2761
+    },
+    {
+      "cosine_similarity": -0.022865567274460666,
+      "epoch": 2.5740913327120225,
+      "grad_norm": 1.2422625567385686,
+      "learning_rate": 3.1549879185364174e-06,
+      "loss": 1.5406,
+      "step": 2762
+    },
+    {
+      "cosine_similarity": 0.08636950808908332,
+      "epoch": 2.5750232991612303,
+      "grad_norm": 1.1041932664353955,
+      "learning_rate": 3.1480842250604077e-06,
+      "loss": 1.3232,
+      "step": 2763
+    },
+    {
+      "cosine_similarity": -0.06128500508506854,
+      "epoch": 2.575955265610438,
+      "grad_norm": 1.1536794090336078,
+      "learning_rate": 3.141180531584398e-06,
+      "loss": 1.5808,
+      "step": 2764
+    },
+    {
+      "cosine_similarity": -0.08927761143005912,
+      "epoch": 2.576887232059646,
+      "grad_norm": 1.5728921357660908,
+      "learning_rate": 3.1342768381083883e-06,
+      "loss": 1.8429,
+      "step": 2765
+    },
+    {
+      "cosine_similarity": 0.09752502152077432,
+      "epoch": 2.5778191985088537,
+      "grad_norm": 1.1150919608923089,
+      "learning_rate": 3.1273731446323786e-06,
+      "loss": 1.1784,
+      "step": 2766
+    },
+    {
+      "cosine_similarity": -0.051417205663088335,
+      "epoch": 2.5787511649580614,
+      "grad_norm": 1.0120196237427628,
+      "learning_rate": 3.120469451156369e-06,
+      "loss": 1.1052,
+      "step": 2767
+    },
+    {
+      "cosine_similarity": 0.1335442485969201,
+      "epoch": 2.5796831314072692,
+      "grad_norm": 1.3316071871492456,
+      "learning_rate": 3.1135657576803592e-06,
+      "loss": 1.4881,
+      "step": 2768
+    },
+    {
+      "cosine_similarity": 0.08285767251035518,
+      "epoch": 2.580615097856477,
+      "grad_norm": 1.166190988054179,
+      "learning_rate": 3.1066620642043495e-06,
+      "loss": 1.4993,
+      "step": 2769
+    },
+    {
+      "cosine_similarity": 0.0004358119085037158,
+      "epoch": 2.581547064305685,
+      "grad_norm": 1.1680020863136016,
+      "learning_rate": 3.0997583707283403e-06,
+      "loss": 1.6072,
+      "step": 2770
+    },
+    {
+      "cosine_similarity": 0.02801588524136717,
+      "epoch": 2.582479030754893,
+      "grad_norm": 1.1398900376183614,
+      "learning_rate": 3.0928546772523306e-06,
+      "loss": 1.4648,
+      "step": 2771
+    },
+    {
+      "cosine_similarity": 0.1428470887460836,
+      "epoch": 2.5834109972041004,
+      "grad_norm": 1.078231607060509,
+      "learning_rate": 3.085950983776321e-06,
+      "loss": 1.5115,
+      "step": 2772
+    },
+    {
+      "cosine_similarity": -0.0058921377432984475,
+      "epoch": 2.5843429636533086,
+      "grad_norm": 0.9296784601644943,
+      "learning_rate": 3.079047290300311e-06,
+      "loss": 1.3493,
+      "step": 2773
+    },
+    {
+      "cosine_similarity": -0.058813143801127635,
+      "epoch": 2.5852749301025164,
+      "grad_norm": 1.074666881129382,
+      "learning_rate": 3.0721435968243015e-06,
+      "loss": 1.2508,
+      "step": 2774
+    },
+    {
+      "cosine_similarity": 0.11180166309867921,
+      "epoch": 2.586206896551724,
+      "grad_norm": 1.049318733718672,
+      "learning_rate": 3.0652399033482914e-06,
+      "loss": 1.2972,
+      "step": 2775
+    },
+    {
+      "cosine_similarity": -0.10122246644409867,
+      "epoch": 2.587138863000932,
+      "grad_norm": 1.0036703082570555,
+      "learning_rate": 3.0583362098722817e-06,
+      "loss": 1.3649,
+      "step": 2776
+    },
+    {
+      "cosine_similarity": -0.004799805893976944,
+      "epoch": 2.5880708294501398,
+      "grad_norm": 1.2723049093476675,
+      "learning_rate": 3.051432516396272e-06,
+      "loss": 1.5598,
+      "step": 2777
+    },
+    {
+      "cosine_similarity": 0.05178547445729181,
+      "epoch": 2.5890027958993476,
+      "grad_norm": 1.2959776947569264,
+      "learning_rate": 3.0445288229202623e-06,
+      "loss": 1.5056,
+      "step": 2778
+    },
+    {
+      "cosine_similarity": -0.010395013979347927,
+      "epoch": 2.5899347623485554,
+      "grad_norm": 1.1295423740367192,
+      "learning_rate": 3.0376251294442526e-06,
+      "loss": 1.4494,
+      "step": 2779
+    },
+    {
+      "cosine_similarity": -0.03817986815192157,
+      "epoch": 2.590866728797763,
+      "grad_norm": 1.0001559214387246,
+      "learning_rate": 3.0307214359682433e-06,
+      "loss": 1.3877,
+      "step": 2780
+    },
+    {
+      "cosine_similarity": 0.055958499563850274,
+      "epoch": 2.591798695246971,
+      "grad_norm": 1.3968500318465233,
+      "learning_rate": 3.0238177424922337e-06,
+      "loss": 1.5611,
+      "step": 2781
+    },
+    {
+      "cosine_similarity": 0.10950139614689168,
+      "epoch": 2.592730661696179,
+      "grad_norm": 1.249807263727298,
+      "learning_rate": 3.016914049016224e-06,
+      "loss": 1.3296,
+      "step": 2782
+    },
+    {
+      "cosine_similarity": -0.01118459346359257,
+      "epoch": 2.5936626281453865,
+      "grad_norm": 1.1168618177886762,
+      "learning_rate": 3.0100103555402143e-06,
+      "loss": 1.634,
+      "step": 2783
+    },
+    {
+      "cosine_similarity": 0.1849414271795837,
+      "epoch": 2.5945945945945947,
+      "grad_norm": 1.253074643540388,
+      "learning_rate": 3.0031066620642046e-06,
+      "loss": 1.5784,
+      "step": 2784
+    },
+    {
+      "cosine_similarity": -0.034992100786561495,
+      "epoch": 2.5955265610438025,
+      "grad_norm": 1.0958442423718275,
+      "learning_rate": 2.996202968588195e-06,
+      "loss": 1.3107,
+      "step": 2785
+    },
+    {
+      "cosine_similarity": 0.05984185623072006,
+      "epoch": 2.5964585274930103,
+      "grad_norm": 1.3143342936442728,
+      "learning_rate": 2.989299275112185e-06,
+      "loss": 1.609,
+      "step": 2786
+    },
+    {
+      "cosine_similarity": 0.06681684031213803,
+      "epoch": 2.597390493942218,
+      "grad_norm": 1.079393842375692,
+      "learning_rate": 2.9823955816361755e-06,
+      "loss": 1.618,
+      "step": 2787
+    },
+    {
+      "cosine_similarity": 0.007084305837140904,
+      "epoch": 2.598322460391426,
+      "grad_norm": 1.004249857870475,
+      "learning_rate": 2.975491888160166e-06,
+      "loss": 1.471,
+      "step": 2788
+    },
+    {
+      "cosine_similarity": -0.20567454828222972,
+      "epoch": 2.5992544268406337,
+      "grad_norm": 1.4871401433507594,
+      "learning_rate": 2.968588194684156e-06,
+      "loss": 1.4984,
+      "step": 2789
+    },
+    {
+      "cosine_similarity": 0.11836705101401516,
+      "epoch": 2.6001863932898415,
+      "grad_norm": 1.3298634041166668,
+      "learning_rate": 2.9616845012081464e-06,
+      "loss": 1.4423,
+      "step": 2790
+    },
+    {
+      "cosine_similarity": 0.04240034758605309,
+      "epoch": 2.6011183597390493,
+      "grad_norm": 1.1763402705498602,
+      "learning_rate": 2.954780807732137e-06,
+      "loss": 1.4807,
+      "step": 2791
+    },
+    {
+      "cosine_similarity": 0.12819061395354925,
+      "epoch": 2.602050326188257,
+      "grad_norm": 1.1761023347499355,
+      "learning_rate": 2.9478771142561275e-06,
+      "loss": 1.5271,
+      "step": 2792
+    },
+    {
+      "cosine_similarity": 0.05515442940029592,
+      "epoch": 2.6029822926374653,
+      "grad_norm": 1.3248796608473792,
+      "learning_rate": 2.940973420780118e-06,
+      "loss": 1.633,
+      "step": 2793
+    },
+    {
+      "cosine_similarity": 0.07912835547778632,
+      "epoch": 2.6039142590866726,
+      "grad_norm": 1.3909864377132146,
+      "learning_rate": 2.934069727304108e-06,
+      "loss": 1.4637,
+      "step": 2794
+    },
+    {
+      "cosine_similarity": 0.06711880568471898,
+      "epoch": 2.604846225535881,
+      "grad_norm": 1.0393389908769446,
+      "learning_rate": 2.9271660338280984e-06,
+      "loss": 1.392,
+      "step": 2795
+    },
+    {
+      "cosine_similarity": 0.030489608773161907,
+      "epoch": 2.6057781919850886,
+      "grad_norm": 1.4943314262822842,
+      "learning_rate": 2.9202623403520887e-06,
+      "loss": 1.6422,
+      "step": 2796
+    },
+    {
+      "cosine_similarity": -0.09610209772507072,
+      "epoch": 2.6067101584342964,
+      "grad_norm": 1.0089428618947294,
+      "learning_rate": 2.913358646876079e-06,
+      "loss": 1.4482,
+      "step": 2797
+    },
+    {
+      "cosine_similarity": 0.03403762729039837,
+      "epoch": 2.607642124883504,
+      "grad_norm": 1.2893721859897647,
+      "learning_rate": 2.906454953400069e-06,
+      "loss": 1.7125,
+      "step": 2798
+    },
+    {
+      "cosine_similarity": 0.0867159976609385,
+      "epoch": 2.608574091332712,
+      "grad_norm": 1.3526721682495233,
+      "learning_rate": 2.8995512599240592e-06,
+      "loss": 1.358,
+      "step": 2799
+    },
+    {
+      "cosine_similarity": 0.1611330052059222,
+      "epoch": 2.60950605778192,
+      "grad_norm": 1.2190498638843512,
+      "learning_rate": 2.8926475664480495e-06,
+      "loss": 1.4727,
+      "step": 2800
+    },
+    {
+      "cosine_similarity": -0.12489899282817718,
+      "epoch": 2.6104380242311276,
+      "grad_norm": 1.2132987351156905,
+      "learning_rate": 2.8857438729720407e-06,
+      "loss": 1.5309,
+      "step": 2801
+    },
+    {
+      "cosine_similarity": -0.005221999192132776,
+      "epoch": 2.6113699906803354,
+      "grad_norm": 1.4006199284552705,
+      "learning_rate": 2.878840179496031e-06,
+      "loss": 1.3691,
+      "step": 2802
+    },
+    {
+      "cosine_similarity": -0.012994735087719565,
+      "epoch": 2.612301957129543,
+      "grad_norm": 1.236693628176731,
+      "learning_rate": 2.871936486020021e-06,
+      "loss": 1.3647,
+      "step": 2803
+    },
+    {
+      "cosine_similarity": -0.09162728922370646,
+      "epoch": 2.6132339235787514,
+      "grad_norm": 1.3059497326275011,
+      "learning_rate": 2.865032792544011e-06,
+      "loss": 1.4788,
+      "step": 2804
+    },
+    {
+      "cosine_similarity": -0.03074098987553991,
+      "epoch": 2.6141658900279587,
+      "grad_norm": 1.1260167286458678,
+      "learning_rate": 2.8581290990680015e-06,
+      "loss": 1.4133,
+      "step": 2805
+    },
+    {
+      "cosine_similarity": 0.011986753927461229,
+      "epoch": 2.615097856477167,
+      "grad_norm": 1.2636845246803423,
+      "learning_rate": 2.851225405591992e-06,
+      "loss": 1.9322,
+      "step": 2806
+    },
+    {
+      "cosine_similarity": -0.07187475746124482,
+      "epoch": 2.6160298229263748,
+      "grad_norm": 1.0305859704638556,
+      "learning_rate": 2.844321712115982e-06,
+      "loss": 1.3191,
+      "step": 2807
+    },
+    {
+      "cosine_similarity": 0.13211322632040273,
+      "epoch": 2.6169617893755825,
+      "grad_norm": 1.1682578425455175,
+      "learning_rate": 2.8374180186399724e-06,
+      "loss": 1.4855,
+      "step": 2808
+    },
+    {
+      "cosine_similarity": -0.047664806567106116,
+      "epoch": 2.6178937558247903,
+      "grad_norm": 1.06135538874305,
+      "learning_rate": 2.8305143251639627e-06,
+      "loss": 1.7264,
+      "step": 2809
+    },
+    {
+      "cosine_similarity": -0.03854304784745787,
+      "epoch": 2.618825722273998,
+      "grad_norm": 1.0432852790893739,
+      "learning_rate": 2.823610631687953e-06,
+      "loss": 1.165,
+      "step": 2810
+    },
+    {
+      "cosine_similarity": -0.027276626588982406,
+      "epoch": 2.619757688723206,
+      "grad_norm": 1.4089552241706045,
+      "learning_rate": 2.8167069382119438e-06,
+      "loss": 1.3601,
+      "step": 2811
+    },
+    {
+      "cosine_similarity": 0.11053835892025585,
+      "epoch": 2.6206896551724137,
+      "grad_norm": 1.045278171468533,
+      "learning_rate": 2.809803244735934e-06,
+      "loss": 1.2489,
+      "step": 2812
+    },
+    {
+      "cosine_similarity": 0.02878908481714491,
+      "epoch": 2.6216216216216215,
+      "grad_norm": 1.0866361819919859,
+      "learning_rate": 2.8028995512599244e-06,
+      "loss": 1.0513,
+      "step": 2813
+    },
+    {
+      "cosine_similarity": 0.009861154811581016,
+      "epoch": 2.6225535880708293,
+      "grad_norm": 1.0690650260798134,
+      "learning_rate": 2.7959958577839147e-06,
+      "loss": 1.5499,
+      "step": 2814
+    },
+    {
+      "cosine_similarity": 0.11615631618954833,
+      "epoch": 2.6234855545200375,
+      "grad_norm": 1.0403910051092735,
+      "learning_rate": 2.789092164307905e-06,
+      "loss": 1.4037,
+      "step": 2815
+    },
+    {
+      "cosine_similarity": 0.016055730045975287,
+      "epoch": 2.624417520969245,
+      "grad_norm": 1.0551066932808961,
+      "learning_rate": 2.7821884708318953e-06,
+      "loss": 1.3203,
+      "step": 2816
+    },
+    {
+      "cosine_similarity": 0.056955680871619514,
+      "epoch": 2.625349487418453,
+      "grad_norm": 1.191065314337494,
+      "learning_rate": 2.7752847773558856e-06,
+      "loss": 1.5868,
+      "step": 2817
+    },
+    {
+      "cosine_similarity": -0.1087322126187299,
+      "epoch": 2.626281453867661,
+      "grad_norm": 1.218578811962513,
+      "learning_rate": 2.768381083879876e-06,
+      "loss": 1.481,
+      "step": 2818
+    },
+    {
+      "cosine_similarity": 0.09559000611206526,
+      "epoch": 2.6272134203168687,
+      "grad_norm": 1.0617775000244032,
+      "learning_rate": 2.7614773904038662e-06,
+      "loss": 1.2781,
+      "step": 2819
+    },
+    {
+      "cosine_similarity": 0.04183715677882739,
+      "epoch": 2.6281453867660765,
+      "grad_norm": 1.2609611324481411,
+      "learning_rate": 2.7545736969278566e-06,
+      "loss": 1.2917,
+      "step": 2820
+    },
+    {
+      "cosine_similarity": 0.22340006768219686,
+      "epoch": 2.6290773532152842,
+      "grad_norm": 1.368919024774817,
+      "learning_rate": 2.7476700034518473e-06,
+      "loss": 1.7571,
+      "step": 2821
+    },
+    {
+      "cosine_similarity": 0.009872978729640866,
+      "epoch": 2.630009319664492,
+      "grad_norm": 1.2391961651890218,
+      "learning_rate": 2.7407663099758376e-06,
+      "loss": 1.5803,
+      "step": 2822
+    },
+    {
+      "cosine_similarity": -0.08146652981655782,
+      "epoch": 2.6309412861137,
+      "grad_norm": 1.2137854986382035,
+      "learning_rate": 2.733862616499828e-06,
+      "loss": 1.7512,
+      "step": 2823
+    },
+    {
+      "cosine_similarity": -0.05368725270527501,
+      "epoch": 2.6318732525629076,
+      "grad_norm": 1.2402430550804928,
+      "learning_rate": 2.7269589230238182e-06,
+      "loss": 1.5873,
+      "step": 2824
+    },
+    {
+      "cosine_similarity": -0.01591710583115711,
+      "epoch": 2.6328052190121154,
+      "grad_norm": 1.1449267508743413,
+      "learning_rate": 2.7200552295478085e-06,
+      "loss": 1.3749,
+      "step": 2825
+    },
+    {
+      "cosine_similarity": 0.1729311549763909,
+      "epoch": 2.6337371854613236,
+      "grad_norm": 1.1448688198006265,
+      "learning_rate": 2.7131515360717984e-06,
+      "loss": 1.4418,
+      "step": 2826
+    },
+    {
+      "cosine_similarity": -0.08724416717583391,
+      "epoch": 2.634669151910531,
+      "grad_norm": 1.158884702220611,
+      "learning_rate": 2.7062478425957887e-06,
+      "loss": 1.6933,
+      "step": 2827
+    },
+    {
+      "cosine_similarity": 0.13830458492748102,
+      "epoch": 2.635601118359739,
+      "grad_norm": 1.326039279449579,
+      "learning_rate": 2.699344149119779e-06,
+      "loss": 1.6285,
+      "step": 2828
+    },
+    {
+      "cosine_similarity": 0.15892092918991224,
+      "epoch": 2.636533084808947,
+      "grad_norm": 1.4382411525341525,
+      "learning_rate": 2.6924404556437693e-06,
+      "loss": 1.294,
+      "step": 2829
+    },
+    {
+      "cosine_similarity": -0.09062509626444032,
+      "epoch": 2.637465051258155,
+      "grad_norm": 1.3093718821980063,
+      "learning_rate": 2.6855367621677596e-06,
+      "loss": 1.8656,
+      "step": 2830
+    },
+    {
+      "cosine_similarity": 0.019544010312147583,
+      "epoch": 2.6383970177073626,
+      "grad_norm": 1.2470998478737338,
+      "learning_rate": 2.6786330686917504e-06,
+      "loss": 1.4072,
+      "step": 2831
+    },
+    {
+      "cosine_similarity": 0.18983615873808493,
+      "epoch": 2.6393289841565704,
+      "grad_norm": 1.4870973285618465,
+      "learning_rate": 2.6717293752157407e-06,
+      "loss": 1.5349,
+      "step": 2832
+    },
+    {
+      "cosine_similarity": 0.16705461281787362,
+      "epoch": 2.640260950605778,
+      "grad_norm": 1.0671482336836364,
+      "learning_rate": 2.664825681739731e-06,
+      "loss": 1.2799,
+      "step": 2833
+    },
+    {
+      "cosine_similarity": 0.007736610903679916,
+      "epoch": 2.641192917054986,
+      "grad_norm": 1.1000394042746144,
+      "learning_rate": 2.6579219882637213e-06,
+      "loss": 1.7028,
+      "step": 2834
+    },
+    {
+      "cosine_similarity": 0.12177359865189351,
+      "epoch": 2.6421248835041937,
+      "grad_norm": 1.0965553114797595,
+      "learning_rate": 2.6510182947877116e-06,
+      "loss": 1.2909,
+      "step": 2835
+    },
+    {
+      "cosine_similarity": 0.14356540533130235,
+      "epoch": 2.6430568499534015,
+      "grad_norm": 1.185524401425475,
+      "learning_rate": 2.644114601311702e-06,
+      "loss": 1.2559,
+      "step": 2836
+    },
+    {
+      "cosine_similarity": -0.04657681505017191,
+      "epoch": 2.6439888164026097,
+      "grad_norm": 1.1568369026674497,
+      "learning_rate": 2.6372109078356922e-06,
+      "loss": 1.6809,
+      "step": 2837
+    },
+    {
+      "cosine_similarity": -0.040839143415523506,
+      "epoch": 2.644920782851817,
+      "grad_norm": 1.0266427134902585,
+      "learning_rate": 2.6303072143596825e-06,
+      "loss": 1.5248,
+      "step": 2838
+    },
+    {
+      "cosine_similarity": 0.1919267114115722,
+      "epoch": 2.6458527493010253,
+      "grad_norm": 1.185461099472731,
+      "learning_rate": 2.623403520883673e-06,
+      "loss": 1.3351,
+      "step": 2839
+    },
+    {
+      "cosine_similarity": 0.00035871664341043097,
+      "epoch": 2.646784715750233,
+      "grad_norm": 1.1151972978705131,
+      "learning_rate": 2.616499827407663e-06,
+      "loss": 1.4642,
+      "step": 2840
+    },
+    {
+      "cosine_similarity": 0.11914072450925674,
+      "epoch": 2.647716682199441,
+      "grad_norm": 1.0997683005151289,
+      "learning_rate": 2.609596133931654e-06,
+      "loss": 1.4054,
+      "step": 2841
+    },
+    {
+      "cosine_similarity": 0.009679026366523381,
+      "epoch": 2.6486486486486487,
+      "grad_norm": 1.254397539581071,
+      "learning_rate": 2.602692440455644e-06,
+      "loss": 1.8988,
+      "step": 2842
+    },
+    {
+      "cosine_similarity": -0.02109366182532496,
+      "epoch": 2.6495806150978565,
+      "grad_norm": 1.10072605713392,
+      "learning_rate": 2.5957887469796345e-06,
+      "loss": 1.4862,
+      "step": 2843
+    },
+    {
+      "cosine_similarity": 0.10489286616301331,
+      "epoch": 2.6505125815470643,
+      "grad_norm": 1.281315717450355,
+      "learning_rate": 2.588885053503625e-06,
+      "loss": 1.342,
+      "step": 2844
+    },
+    {
+      "cosine_similarity": -0.13844378022600778,
+      "epoch": 2.651444547996272,
+      "grad_norm": 1.0851773288948388,
+      "learning_rate": 2.581981360027615e-06,
+      "loss": 1.4403,
+      "step": 2845
+    },
+    {
+      "cosine_similarity": 0.1316422976604795,
+      "epoch": 2.65237651444548,
+      "grad_norm": 1.1447633508255388,
+      "learning_rate": 2.5750776665516054e-06,
+      "loss": 1.5689,
+      "step": 2846
+    },
+    {
+      "cosine_similarity": 0.15700543610310844,
+      "epoch": 2.6533084808946876,
+      "grad_norm": 1.2437775639467035,
+      "learning_rate": 2.5681739730755957e-06,
+      "loss": 1.6924,
+      "step": 2847
+    },
+    {
+      "cosine_similarity": 0.16046767581841903,
+      "epoch": 2.654240447343896,
+      "grad_norm": 1.1465621172457536,
+      "learning_rate": 2.561270279599586e-06,
+      "loss": 1.4532,
+      "step": 2848
+    },
+    {
+      "cosine_similarity": 0.07864609288170994,
+      "epoch": 2.655172413793103,
+      "grad_norm": 1.1936066209031149,
+      "learning_rate": 2.554366586123576e-06,
+      "loss": 1.7303,
+      "step": 2849
+    },
+    {
+      "cosine_similarity": 0.12820571490786714,
+      "epoch": 2.6561043802423114,
+      "grad_norm": 1.1461177288976727,
+      "learning_rate": 2.5474628926475662e-06,
+      "loss": 1.6003,
+      "step": 2850
+    },
+    {
+      "cosine_similarity": 0.05723398850103756,
+      "epoch": 2.6570363466915192,
+      "grad_norm": 1.3778215958592328,
+      "learning_rate": 2.5405591991715574e-06,
+      "loss": 1.3652,
+      "step": 2851
+    },
+    {
+      "cosine_similarity": 0.1251357931247132,
+      "epoch": 2.657968313140727,
+      "grad_norm": 1.138567405827038,
+      "learning_rate": 2.5336555056955477e-06,
+      "loss": 1.2809,
+      "step": 2852
+    },
+    {
+      "cosine_similarity": 0.037431636992860506,
+      "epoch": 2.658900279589935,
+      "grad_norm": 1.2732431239264832,
+      "learning_rate": 2.5267518122195376e-06,
+      "loss": 1.4728,
+      "step": 2853
+    },
+    {
+      "cosine_similarity": 0.07006497694468918,
+      "epoch": 2.6598322460391426,
+      "grad_norm": 1.2167178327203336,
+      "learning_rate": 2.519848118743528e-06,
+      "loss": 1.5887,
+      "step": 2854
+    },
+    {
+      "cosine_similarity": 0.0730091852247393,
+      "epoch": 2.6607642124883504,
+      "grad_norm": 1.0840189333940482,
+      "learning_rate": 2.5129444252675182e-06,
+      "loss": 1.4774,
+      "step": 2855
+    },
+    {
+      "cosine_similarity": 0.10765610068695479,
+      "epoch": 2.661696178937558,
+      "grad_norm": 0.9265773710269527,
+      "learning_rate": 2.5060407317915085e-06,
+      "loss": 1.5543,
+      "step": 2856
+    },
+    {
+      "cosine_similarity": -0.12497288305203547,
+      "epoch": 2.662628145386766,
+      "grad_norm": 1.1975088047979514,
+      "learning_rate": 2.499137038315499e-06,
+      "loss": 1.4089,
+      "step": 2857
+    },
+    {
+      "cosine_similarity": -0.04290369443401657,
+      "epoch": 2.6635601118359737,
+      "grad_norm": 1.1534809881972785,
+      "learning_rate": 2.492233344839489e-06,
+      "loss": 1.4456,
+      "step": 2858
+    },
+    {
+      "cosine_similarity": -0.013809919151648605,
+      "epoch": 2.664492078285182,
+      "grad_norm": 1.4629440931499105,
+      "learning_rate": 2.48532965136348e-06,
+      "loss": 1.5797,
+      "step": 2859
+    },
+    {
+      "cosine_similarity": 0.1496074383987882,
+      "epoch": 2.6654240447343893,
+      "grad_norm": 1.2087424372093374,
+      "learning_rate": 2.47842595788747e-06,
+      "loss": 1.5948,
+      "step": 2860
+    },
+    {
+      "cosine_similarity": -0.008069202344656126,
+      "epoch": 2.6663560111835976,
+      "grad_norm": 1.3871374205691716,
+      "learning_rate": 2.4715222644114605e-06,
+      "loss": 1.2639,
+      "step": 2861
+    },
+    {
+      "cosine_similarity": -0.017464535538053324,
+      "epoch": 2.6672879776328053,
+      "grad_norm": 1.1824072411226731,
+      "learning_rate": 2.464618570935451e-06,
+      "loss": 1.5617,
+      "step": 2862
+    },
+    {
+      "cosine_similarity": 0.051602381919671654,
+      "epoch": 2.668219944082013,
+      "grad_norm": 1.1388565904299843,
+      "learning_rate": 2.4577148774594407e-06,
+      "loss": 1.5542,
+      "step": 2863
+    },
+    {
+      "cosine_similarity": 0.11069544925475076,
+      "epoch": 2.669151910531221,
+      "grad_norm": 1.0663801852925316,
+      "learning_rate": 2.4508111839834314e-06,
+      "loss": 1.3799,
+      "step": 2864
+    },
+    {
+      "cosine_similarity": -0.028252037845046686,
+      "epoch": 2.6700838769804287,
+      "grad_norm": 1.2849588412852704,
+      "learning_rate": 2.4439074905074217e-06,
+      "loss": 1.9729,
+      "step": 2865
+    },
+    {
+      "cosine_similarity": 0.04126086144982305,
+      "epoch": 2.6710158434296365,
+      "grad_norm": 1.1056423320267827,
+      "learning_rate": 2.437003797031412e-06,
+      "loss": 1.6875,
+      "step": 2866
+    },
+    {
+      "cosine_similarity": -0.06566705414470202,
+      "epoch": 2.6719478098788443,
+      "grad_norm": 1.157317355267638,
+      "learning_rate": 2.4301001035554023e-06,
+      "loss": 1.6132,
+      "step": 2867
+    },
+    {
+      "cosine_similarity": 0.03985117492790908,
+      "epoch": 2.672879776328052,
+      "grad_norm": 1.097122388545919,
+      "learning_rate": 2.4231964100793927e-06,
+      "loss": 1.3502,
+      "step": 2868
+    },
+    {
+      "cosine_similarity": -0.03884499266675594,
+      "epoch": 2.67381174277726,
+      "grad_norm": 1.0038011964896485,
+      "learning_rate": 2.416292716603383e-06,
+      "loss": 1.1549,
+      "step": 2869
+    },
+    {
+      "cosine_similarity": 0.014643160003133487,
+      "epoch": 2.674743709226468,
+      "grad_norm": 1.1194565812005826,
+      "learning_rate": 2.4093890231273733e-06,
+      "loss": 1.3963,
+      "step": 2870
+    },
+    {
+      "cosine_similarity": 0.07470432878263888,
+      "epoch": 2.6756756756756754,
+      "grad_norm": 1.2399045733439553,
+      "learning_rate": 2.4024853296513636e-06,
+      "loss": 1.5577,
+      "step": 2871
+    },
+    {
+      "cosine_similarity": -0.01422468606827065,
+      "epoch": 2.6766076421248837,
+      "grad_norm": 1.126580201769804,
+      "learning_rate": 2.395581636175354e-06,
+      "loss": 1.1381,
+      "step": 2872
+    },
+    {
+      "cosine_similarity": -0.08386103744849001,
+      "epoch": 2.6775396085740915,
+      "grad_norm": 1.264661929996738,
+      "learning_rate": 2.388677942699344e-06,
+      "loss": 1.6877,
+      "step": 2873
+    },
+    {
+      "cosine_similarity": 0.10405776003010142,
+      "epoch": 2.6784715750232992,
+      "grad_norm": 1.3037953366328807,
+      "learning_rate": 2.381774249223335e-06,
+      "loss": 1.6956,
+      "step": 2874
+    },
+    {
+      "cosine_similarity": 0.04455430881501886,
+      "epoch": 2.679403541472507,
+      "grad_norm": 1.161741800996259,
+      "learning_rate": 2.3748705557473252e-06,
+      "loss": 1.5817,
+      "step": 2875
+    },
+    {
+      "cosine_similarity": 0.012713472345954676,
+      "epoch": 2.680335507921715,
+      "grad_norm": 1.6696099108747333,
+      "learning_rate": 2.367966862271315e-06,
+      "loss": 1.4519,
+      "step": 2876
+    },
+    {
+      "cosine_similarity": -0.09680783827497234,
+      "epoch": 2.6812674743709226,
+      "grad_norm": 1.1002360307605945,
+      "learning_rate": 2.3610631687953054e-06,
+      "loss": 1.2247,
+      "step": 2877
+    },
+    {
+      "cosine_similarity": 0.037629930400541536,
+      "epoch": 2.6821994408201304,
+      "grad_norm": 1.0945170726378128,
+      "learning_rate": 2.3541594753192957e-06,
+      "loss": 1.4229,
+      "step": 2878
+    },
+    {
+      "cosine_similarity": 0.1568823951682388,
+      "epoch": 2.683131407269338,
+      "grad_norm": 1.1240667504534112,
+      "learning_rate": 2.3472557818432865e-06,
+      "loss": 1.298,
+      "step": 2879
+    },
+    {
+      "cosine_similarity": -0.06652936348898045,
+      "epoch": 2.684063373718546,
+      "grad_norm": 1.1127406464703327,
+      "learning_rate": 2.3403520883672768e-06,
+      "loss": 1.4456,
+      "step": 2880
+    },
+    {
+      "cosine_similarity": 0.12796757303377712,
+      "epoch": 2.684995340167754,
+      "grad_norm": 1.1644887204260184,
+      "learning_rate": 2.333448394891267e-06,
+      "loss": 1.6938,
+      "step": 2881
+    },
+    {
+      "cosine_similarity": 0.07371255574815547,
+      "epoch": 2.6859273066169616,
+      "grad_norm": 1.2562683131397216,
+      "learning_rate": 2.3265447014152574e-06,
+      "loss": 1.533,
+      "step": 2882
+    },
+    {
+      "cosine_similarity": 0.1097134484437083,
+      "epoch": 2.68685927306617,
+      "grad_norm": 1.0328828777914483,
+      "learning_rate": 2.3196410079392477e-06,
+      "loss": 1.6962,
+      "step": 2883
+    },
+    {
+      "cosine_similarity": 0.08114917993458275,
+      "epoch": 2.6877912395153776,
+      "grad_norm": 1.3737356377504162,
+      "learning_rate": 2.312737314463238e-06,
+      "loss": 1.4701,
+      "step": 2884
+    },
+    {
+      "cosine_similarity": -0.035929658655559416,
+      "epoch": 2.6887232059645854,
+      "grad_norm": 0.9701949140224863,
+      "learning_rate": 2.3058336209872283e-06,
+      "loss": 1.291,
+      "step": 2885
+    },
+    {
+      "cosine_similarity": 0.04381023561035433,
+      "epoch": 2.689655172413793,
+      "grad_norm": 1.0434361126371943,
+      "learning_rate": 2.2989299275112186e-06,
+      "loss": 1.4193,
+      "step": 2886
+    },
+    {
+      "cosine_similarity": -0.05669532138909379,
+      "epoch": 2.690587138863001,
+      "grad_norm": 1.102037264766202,
+      "learning_rate": 2.292026234035209e-06,
+      "loss": 1.3178,
+      "step": 2887
+    },
+    {
+      "cosine_similarity": 0.021394776829087987,
+      "epoch": 2.6915191053122087,
+      "grad_norm": 1.259701765985499,
+      "learning_rate": 2.2851225405591993e-06,
+      "loss": 1.3536,
+      "step": 2888
+    },
+    {
+      "cosine_similarity": -0.09258872476965416,
+      "epoch": 2.6924510717614165,
+      "grad_norm": 1.1278519005480439,
+      "learning_rate": 2.27821884708319e-06,
+      "loss": 1.3746,
+      "step": 2889
+    },
+    {
+      "cosine_similarity": 0.04734528681436611,
+      "epoch": 2.6933830382106243,
+      "grad_norm": 1.0742462656508647,
+      "learning_rate": 2.27131515360718e-06,
+      "loss": 1.499,
+      "step": 2890
+    },
+    {
+      "cosine_similarity": 0.07683930448144122,
+      "epoch": 2.694315004659832,
+      "grad_norm": 1.1114280263380723,
+      "learning_rate": 2.26441146013117e-06,
+      "loss": 1.3287,
+      "step": 2891
+    },
+    {
+      "cosine_similarity": 0.10904022344160771,
+      "epoch": 2.6952469711090403,
+      "grad_norm": 1.142622951893357,
+      "learning_rate": 2.2575077666551605e-06,
+      "loss": 1.4228,
+      "step": 2892
+    },
+    {
+      "cosine_similarity": 0.032264457011709914,
+      "epoch": 2.6961789375582477,
+      "grad_norm": 1.042889512640662,
+      "learning_rate": 2.250604073179151e-06,
+      "loss": 1.4423,
+      "step": 2893
+    },
+    {
+      "cosine_similarity": 0.047884709749022585,
+      "epoch": 2.697110904007456,
+      "grad_norm": 1.2398207547978568,
+      "learning_rate": 2.2437003797031415e-06,
+      "loss": 1.464,
+      "step": 2894
+    },
+    {
+      "cosine_similarity": 0.07974424208792526,
+      "epoch": 2.6980428704566637,
+      "grad_norm": 1.0466018434542033,
+      "learning_rate": 2.236796686227132e-06,
+      "loss": 1.4165,
+      "step": 2895
+    },
+    {
+      "cosine_similarity": -0.02328638459805271,
+      "epoch": 2.6989748369058715,
+      "grad_norm": 1.249648163552076,
+      "learning_rate": 2.229892992751122e-06,
+      "loss": 1.966,
+      "step": 2896
+    },
+    {
+      "cosine_similarity": -0.0028731133910468204,
+      "epoch": 2.6999068033550793,
+      "grad_norm": 1.0358374869836326,
+      "learning_rate": 2.2229892992751125e-06,
+      "loss": 1.3485,
+      "step": 2897
+    },
+    {
+      "cosine_similarity": -0.07613738794761213,
+      "epoch": 2.700838769804287,
+      "grad_norm": 1.3365770502423053,
+      "learning_rate": 2.2160856057991028e-06,
+      "loss": 1.7641,
+      "step": 2898
+    },
+    {
+      "cosine_similarity": -0.058284457907036,
+      "epoch": 2.701770736253495,
+      "grad_norm": 1.115499005315317,
+      "learning_rate": 2.209181912323093e-06,
+      "loss": 1.3963,
+      "step": 2899
+    },
+    {
+      "cosine_similarity": -0.0345008753979254,
+      "epoch": 2.7027027027027026,
+      "grad_norm": 1.1047386022233405,
+      "learning_rate": 2.2022782188470834e-06,
+      "loss": 1.4529,
+      "step": 2900
+    },
+    {
+      "cosine_similarity": -0.0625249904045022,
+      "epoch": 2.7036346691519104,
+      "grad_norm": 1.1523336543949805,
+      "learning_rate": 2.1953745253710737e-06,
+      "loss": 1.2811,
+      "step": 2901
+    },
+    {
+      "cosine_similarity": -0.018575295626854005,
+      "epoch": 2.704566635601118,
+      "grad_norm": 1.1241165662832089,
+      "learning_rate": 2.188470831895064e-06,
+      "loss": 1.4261,
+      "step": 2902
+    },
+    {
+      "cosine_similarity": 0.1589439880066191,
+      "epoch": 2.7054986020503264,
+      "grad_norm": 1.1445910543965332,
+      "learning_rate": 2.1815671384190543e-06,
+      "loss": 1.1888,
+      "step": 2903
+    },
+    {
+      "cosine_similarity": -0.027821692304183243,
+      "epoch": 2.706430568499534,
+      "grad_norm": 1.3522302604516014,
+      "learning_rate": 2.1746634449430446e-06,
+      "loss": 1.8137,
+      "step": 2904
+    },
+    {
+      "cosine_similarity": 0.022967694930767883,
+      "epoch": 2.707362534948742,
+      "grad_norm": 1.0620184609182157,
+      "learning_rate": 2.167759751467035e-06,
+      "loss": 1.3748,
+      "step": 2905
+    },
+    {
+      "cosine_similarity": 0.12139952737390197,
+      "epoch": 2.70829450139795,
+      "grad_norm": 1.273820194313384,
+      "learning_rate": 2.1608560579910252e-06,
+      "loss": 1.5661,
+      "step": 2906
+    },
+    {
+      "cosine_similarity": 0.07611376038834805,
+      "epoch": 2.7092264678471576,
+      "grad_norm": 1.6578281079599622,
+      "learning_rate": 2.1539523645150156e-06,
+      "loss": 1.5263,
+      "step": 2907
+    },
+    {
+      "cosine_similarity": 0.19314912658618416,
+      "epoch": 2.7101584342963654,
+      "grad_norm": 1.261445303463252,
+      "learning_rate": 2.147048671039006e-06,
+      "loss": 1.2691,
+      "step": 2908
+    },
+    {
+      "cosine_similarity": -0.06225954934425593,
+      "epoch": 2.711090400745573,
+      "grad_norm": 1.1435227850797207,
+      "learning_rate": 2.1401449775629966e-06,
+      "loss": 1.3186,
+      "step": 2909
+    },
+    {
+      "cosine_similarity": 0.015282348247867217,
+      "epoch": 2.712022367194781,
+      "grad_norm": 1.5110730517526454,
+      "learning_rate": 2.133241284086987e-06,
+      "loss": 1.6598,
+      "step": 2910
+    },
+    {
+      "cosine_similarity": -0.08009321488141637,
+      "epoch": 2.7129543336439887,
+      "grad_norm": 1.1359464044456473,
+      "learning_rate": 2.1263375906109772e-06,
+      "loss": 1.4626,
+      "step": 2911
+    },
+    {
+      "cosine_similarity": -0.06203765763856366,
+      "epoch": 2.7138863000931965,
+      "grad_norm": 1.2212958187821112,
+      "learning_rate": 2.1194338971349675e-06,
+      "loss": 1.461,
+      "step": 2912
+    },
+    {
+      "cosine_similarity": 0.021436523324678134,
+      "epoch": 2.7148182665424043,
+      "grad_norm": 1.0369302480323315,
+      "learning_rate": 2.1125302036589574e-06,
+      "loss": 1.3525,
+      "step": 2913
+    },
+    {
+      "cosine_similarity": 0.00628468354690679,
+      "epoch": 2.7157502329916126,
+      "grad_norm": 0.9375440153363319,
+      "learning_rate": 2.105626510182948e-06,
+      "loss": 1.1803,
+      "step": 2914
+    },
+    {
+      "cosine_similarity": 0.1223375892177181,
+      "epoch": 2.71668219944082,
+      "grad_norm": 1.5011516921553736,
+      "learning_rate": 2.0987228167069384e-06,
+      "loss": 1.5239,
+      "step": 2915
+    },
+    {
+      "cosine_similarity": 0.08191217573568425,
+      "epoch": 2.717614165890028,
+      "grad_norm": 1.086289571871512,
+      "learning_rate": 2.0918191232309288e-06,
+      "loss": 1.4154,
+      "step": 2916
+    },
+    {
+      "cosine_similarity": 0.10459999857461096,
+      "epoch": 2.718546132339236,
+      "grad_norm": 1.1733850785439066,
+      "learning_rate": 2.084915429754919e-06,
+      "loss": 1.3649,
+      "step": 2917
+    },
+    {
+      "cosine_similarity": 0.15659687752424248,
+      "epoch": 2.7194780987884437,
+      "grad_norm": 1.185474628753089,
+      "learning_rate": 2.0780117362789094e-06,
+      "loss": 1.7013,
+      "step": 2918
+    },
+    {
+      "cosine_similarity": 0.08322747790342662,
+      "epoch": 2.7204100652376515,
+      "grad_norm": 1.1788720949583569,
+      "learning_rate": 2.0711080428028997e-06,
+      "loss": 1.529,
+      "step": 2919
+    },
+    {
+      "cosine_similarity": -0.01328543945690336,
+      "epoch": 2.7213420316868593,
+      "grad_norm": 1.4609525337830345,
+      "learning_rate": 2.06420434932689e-06,
+      "loss": 1.5764,
+      "step": 2920
+    },
+    {
+      "cosine_similarity": -0.05818420226338711,
+      "epoch": 2.722273998136067,
+      "grad_norm": 1.171272776461478,
+      "learning_rate": 2.0573006558508803e-06,
+      "loss": 1.4143,
+      "step": 2921
+    },
+    {
+      "cosine_similarity": 0.005128592883509582,
+      "epoch": 2.723205964585275,
+      "grad_norm": 1.194871446183493,
+      "learning_rate": 2.0503969623748706e-06,
+      "loss": 1.4165,
+      "step": 2922
+    },
+    {
+      "cosine_similarity": 0.019498194985759235,
+      "epoch": 2.7241379310344827,
+      "grad_norm": 1.2969147135788652,
+      "learning_rate": 2.043493268898861e-06,
+      "loss": 1.4911,
+      "step": 2923
+    },
+    {
+      "cosine_similarity": 0.13728196506182097,
+      "epoch": 2.7250698974836904,
+      "grad_norm": 1.310832570769842,
+      "learning_rate": 2.0365895754228517e-06,
+      "loss": 1.4629,
+      "step": 2924
+    },
+    {
+      "cosine_similarity": 0.036600801644662895,
+      "epoch": 2.7260018639328987,
+      "grad_norm": 1.2135116253858544,
+      "learning_rate": 2.029685881946842e-06,
+      "loss": 1.5103,
+      "step": 2925
+    },
+    {
+      "cosine_similarity": 0.009691469607653866,
+      "epoch": 2.726933830382106,
+      "grad_norm": 1.053278436785956,
+      "learning_rate": 2.0227821884708323e-06,
+      "loss": 1.3002,
+      "step": 2926
+    },
+    {
+      "cosine_similarity": 0.07483886410227703,
+      "epoch": 2.7278657968313142,
+      "grad_norm": 1.4579542970746002,
+      "learning_rate": 2.015878494994822e-06,
+      "loss": 1.4771,
+      "step": 2927
+    },
+    {
+      "cosine_similarity": -0.001871594752502275,
+      "epoch": 2.728797763280522,
+      "grad_norm": 1.0140225879680453,
+      "learning_rate": 2.0089748015188125e-06,
+      "loss": 1.3172,
+      "step": 2928
+    },
+    {
+      "cosine_similarity": 0.021913346313973077,
+      "epoch": 2.72972972972973,
+      "grad_norm": 1.1876341950488263,
+      "learning_rate": 2.002071108042803e-06,
+      "loss": 1.2394,
+      "step": 2929
+    },
+    {
+      "cosine_similarity": 0.011018804248290784,
+      "epoch": 2.7306616961789376,
+      "grad_norm": 1.0876673711959413,
+      "learning_rate": 1.9951674145667935e-06,
+      "loss": 1.4882,
+      "step": 2930
+    },
+    {
+      "cosine_similarity": -0.0764178488895301,
+      "epoch": 2.7315936626281454,
+      "grad_norm": 1.25771292267329,
+      "learning_rate": 1.988263721090784e-06,
+      "loss": 1.6087,
+      "step": 2931
+    },
+    {
+      "cosine_similarity": -0.053249449142111216,
+      "epoch": 2.732525629077353,
+      "grad_norm": 1.0128167180464114,
+      "learning_rate": 1.981360027614774e-06,
+      "loss": 1.2204,
+      "step": 2932
+    },
+    {
+      "cosine_similarity": 0.0856392703834627,
+      "epoch": 2.733457595526561,
+      "grad_norm": 1.5692647799608475,
+      "learning_rate": 1.9744563341387644e-06,
+      "loss": 1.8177,
+      "step": 2933
+    },
+    {
+      "cosine_similarity": -0.014964829704920028,
+      "epoch": 2.7343895619757688,
+      "grad_norm": 1.3244457853808775,
+      "learning_rate": 1.9675526406627547e-06,
+      "loss": 1.9089,
+      "step": 2934
+    },
+    {
+      "cosine_similarity": 0.04347758629743943,
+      "epoch": 2.7353215284249766,
+      "grad_norm": 1.2415574297538445,
+      "learning_rate": 1.960648947186745e-06,
+      "loss": 1.126,
+      "step": 2935
+    },
+    {
+      "cosine_similarity": -0.034289384328068766,
+      "epoch": 2.736253494874185,
+      "grad_norm": 1.2798655083533752,
+      "learning_rate": 1.9537452537107354e-06,
+      "loss": 1.6417,
+      "step": 2936
+    },
+    {
+      "cosine_similarity": -0.010476276706024747,
+      "epoch": 2.737185461323392,
+      "grad_norm": 0.9772414051987539,
+      "learning_rate": 1.9468415602347257e-06,
+      "loss": 1.2765,
+      "step": 2937
+    },
+    {
+      "cosine_similarity": -0.18633463372027326,
+      "epoch": 2.7381174277726004,
+      "grad_norm": 1.3004449947745986,
+      "learning_rate": 1.939937866758716e-06,
+      "loss": 1.6456,
+      "step": 2938
+    },
+    {
+      "cosine_similarity": -3.411473533831527e-05,
+      "epoch": 2.739049394221808,
+      "grad_norm": 1.158247881082701,
+      "learning_rate": 1.9330341732827067e-06,
+      "loss": 1.5033,
+      "step": 2939
+    },
+    {
+      "cosine_similarity": 0.013364855014919635,
+      "epoch": 2.739981360671016,
+      "grad_norm": 0.914618700420997,
+      "learning_rate": 1.926130479806697e-06,
+      "loss": 1.1581,
+      "step": 2940
+    },
+    {
+      "cosine_similarity": -0.04169553681547663,
+      "epoch": 2.7409133271202237,
+      "grad_norm": 1.3217561881689046,
+      "learning_rate": 1.919226786330687e-06,
+      "loss": 1.578,
+      "step": 2941
+    },
+    {
+      "cosine_similarity": 0.05756636084809726,
+      "epoch": 2.7418452935694315,
+      "grad_norm": 3.0578371872642696,
+      "learning_rate": 1.9123230928546772e-06,
+      "loss": 1.3775,
+      "step": 2942
+    },
+    {
+      "cosine_similarity": 0.011851665530419329,
+      "epoch": 2.7427772600186393,
+      "grad_norm": 1.1353505723914752,
+      "learning_rate": 1.9054193993786677e-06,
+      "loss": 1.5024,
+      "step": 2943
+    },
+    {
+      "cosine_similarity": 0.0059178302486420035,
+      "epoch": 2.743709226467847,
+      "grad_norm": 1.0973194453101933,
+      "learning_rate": 1.8985157059026583e-06,
+      "loss": 1.4054,
+      "step": 2944
+    },
+    {
+      "cosine_similarity": 0.024671058061429774,
+      "epoch": 2.744641192917055,
+      "grad_norm": 0.9890774475537708,
+      "learning_rate": 1.8916120124266486e-06,
+      "loss": 1.1781,
+      "step": 2945
+    },
+    {
+      "cosine_similarity": -0.017707454895778452,
+      "epoch": 2.7455731593662627,
+      "grad_norm": 1.2251098788870018,
+      "learning_rate": 1.8847083189506389e-06,
+      "loss": 1.5505,
+      "step": 2946
+    },
+    {
+      "cosine_similarity": -0.07567668664353439,
+      "epoch": 2.746505125815471,
+      "grad_norm": 1.2021649075955665,
+      "learning_rate": 1.877804625474629e-06,
+      "loss": 1.5031,
+      "step": 2947
+    },
+    {
+      "cosine_similarity": -0.02244424953763402,
+      "epoch": 2.7474370922646782,
+      "grad_norm": 1.0388586309729881,
+      "learning_rate": 1.8709009319986193e-06,
+      "loss": 1.3386,
+      "step": 2948
+    },
+    {
+      "cosine_similarity": 0.02247356298174928,
+      "epoch": 2.7483690587138865,
+      "grad_norm": 1.0519827599494738,
+      "learning_rate": 1.8639972385226098e-06,
+      "loss": 1.3662,
+      "step": 2949
+    },
+    {
+      "cosine_similarity": 0.02586465062454552,
+      "epoch": 2.7493010251630943,
+      "grad_norm": 1.0175405447031558,
+      "learning_rate": 1.8570935450466001e-06,
+      "loss": 1.363,
+      "step": 2950
+    },
+    {
+      "cosine_similarity": 0.049956173760713124,
+      "epoch": 2.750232991612302,
+      "grad_norm": 1.181858236565536,
+      "learning_rate": 1.8501898515705904e-06,
+      "loss": 1.3756,
+      "step": 2951
+    },
+    {
+      "cosine_similarity": 0.024182157606118276,
+      "epoch": 2.75116495806151,
+      "grad_norm": 1.1475846091387831,
+      "learning_rate": 1.8432861580945807e-06,
+      "loss": 1.2627,
+      "step": 2952
+    },
+    {
+      "cosine_similarity": 0.030834230149853446,
+      "epoch": 2.7520969245107176,
+      "grad_norm": 1.0739714909683309,
+      "learning_rate": 1.836382464618571e-06,
+      "loss": 1.4095,
+      "step": 2953
+    },
+    {
+      "cosine_similarity": -0.006292109670531161,
+      "epoch": 2.7530288909599254,
+      "grad_norm": 1.2956190757911903,
+      "learning_rate": 1.8294787711425613e-06,
+      "loss": 1.5462,
+      "step": 2954
+    },
+    {
+      "cosine_similarity": -0.09094773217847649,
+      "epoch": 2.753960857409133,
+      "grad_norm": 1.3893345762163487,
+      "learning_rate": 1.8225750776665519e-06,
+      "loss": 1.5264,
+      "step": 2955
+    },
+    {
+      "cosine_similarity": 0.08849068788540111,
+      "epoch": 2.754892823858341,
+      "grad_norm": 1.1245460522707,
+      "learning_rate": 1.8156713841905422e-06,
+      "loss": 1.5536,
+      "step": 2956
+    },
+    {
+      "cosine_similarity": 0.07353689369574755,
+      "epoch": 2.755824790307549,
+      "grad_norm": 1.3156110445758369,
+      "learning_rate": 1.8087676907145325e-06,
+      "loss": 1.2688,
+      "step": 2957
+    },
+    {
+      "cosine_similarity": 0.1181365263372611,
+      "epoch": 2.756756756756757,
+      "grad_norm": 1.2487680710540106,
+      "learning_rate": 1.8018639972385226e-06,
+      "loss": 1.5424,
+      "step": 2958
+    },
+    {
+      "cosine_similarity": 0.08252159170867022,
+      "epoch": 2.7576887232059644,
+      "grad_norm": 1.343464107769563,
+      "learning_rate": 1.7949603037625129e-06,
+      "loss": 1.542,
+      "step": 2959
+    },
+    {
+      "cosine_similarity": -0.09273265264909007,
+      "epoch": 2.7586206896551726,
+      "grad_norm": 1.3362850185267434,
+      "learning_rate": 1.7880566102865034e-06,
+      "loss": 1.386,
+      "step": 2960
+    },
+    {
+      "cosine_similarity": -0.024416968457699098,
+      "epoch": 2.7595526561043804,
+      "grad_norm": 1.393103557793511,
+      "learning_rate": 1.7811529168104937e-06,
+      "loss": 1.6818,
+      "step": 2961
+    },
+    {
+      "cosine_similarity": -0.045189002878285,
+      "epoch": 2.760484622553588,
+      "grad_norm": 1.1920266318269521,
+      "learning_rate": 1.774249223334484e-06,
+      "loss": 1.5376,
+      "step": 2962
+    },
+    {
+      "cosine_similarity": 0.04888503943227135,
+      "epoch": 2.761416589002796,
+      "grad_norm": 1.2482998056245787,
+      "learning_rate": 1.7673455298584743e-06,
+      "loss": 1.4653,
+      "step": 2963
+    },
+    {
+      "cosine_similarity": 0.049082489120031596,
+      "epoch": 2.7623485554520038,
+      "grad_norm": 1.2228861587608753,
+      "learning_rate": 1.7604418363824646e-06,
+      "loss": 1.6399,
+      "step": 2964
+    },
+    {
+      "cosine_similarity": -0.05654069883629739,
+      "epoch": 2.7632805219012115,
+      "grad_norm": 1.1292906329125851,
+      "learning_rate": 1.7535381429064552e-06,
+      "loss": 1.4755,
+      "step": 2965
+    },
+    {
+      "cosine_similarity": 0.0712153091125918,
+      "epoch": 2.7642124883504193,
+      "grad_norm": 1.0927389543382846,
+      "learning_rate": 1.7466344494304455e-06,
+      "loss": 1.5345,
+      "step": 2966
+    },
+    {
+      "cosine_similarity": 0.01156524912814152,
+      "epoch": 2.765144454799627,
+      "grad_norm": 1.0979825483600398,
+      "learning_rate": 1.7397307559544358e-06,
+      "loss": 1.1022,
+      "step": 2967
+    },
+    {
+      "cosine_similarity": 0.025649227003249266,
+      "epoch": 2.766076421248835,
+      "grad_norm": 1.24046223461152,
+      "learning_rate": 1.732827062478426e-06,
+      "loss": 1.4979,
+      "step": 2968
+    },
+    {
+      "cosine_similarity": -0.03272891569495737,
+      "epoch": 2.767008387698043,
+      "grad_norm": 1.018086850659195,
+      "learning_rate": 1.7259233690024164e-06,
+      "loss": 1.6488,
+      "step": 2969
+    },
+    {
+      "cosine_similarity": -0.021686422382338644,
+      "epoch": 2.7679403541472505,
+      "grad_norm": 1.0205940985798085,
+      "learning_rate": 1.719019675526407e-06,
+      "loss": 1.1997,
+      "step": 2970
+    },
+    {
+      "cosine_similarity": 0.1518006003073711,
+      "epoch": 2.7688723205964587,
+      "grad_norm": 1.187181095674106,
+      "learning_rate": 1.7121159820503972e-06,
+      "loss": 1.3877,
+      "step": 2971
+    },
+    {
+      "cosine_similarity": -0.03720649619281785,
+      "epoch": 2.7698042870456665,
+      "grad_norm": 1.151169855723786,
+      "learning_rate": 1.7052122885743873e-06,
+      "loss": 2.0143,
+      "step": 2972
+    },
+    {
+      "cosine_similarity": 0.11407421306649076,
+      "epoch": 2.7707362534948743,
+      "grad_norm": 1.2279849603890003,
+      "learning_rate": 1.6983085950983776e-06,
+      "loss": 1.6895,
+      "step": 2973
+    },
+    {
+      "cosine_similarity": -0.024374570533854167,
+      "epoch": 2.771668219944082,
+      "grad_norm": 0.9939363480124037,
+      "learning_rate": 1.691404901622368e-06,
+      "loss": 1.4906,
+      "step": 2974
+    },
+    {
+      "cosine_similarity": 0.11842572393459656,
+      "epoch": 2.77260018639329,
+      "grad_norm": 1.141807245760369,
+      "learning_rate": 1.6845012081463585e-06,
+      "loss": 1.355,
+      "step": 2975
+    },
+    {
+      "cosine_similarity": 0.05271169383976318,
+      "epoch": 2.7735321528424977,
+      "grad_norm": 1.183762262837876,
+      "learning_rate": 1.6775975146703488e-06,
+      "loss": 1.3255,
+      "step": 2976
+    },
+    {
+      "cosine_similarity": -0.02871725038402711,
+      "epoch": 2.7744641192917054,
+      "grad_norm": 1.1050233124813362,
+      "learning_rate": 1.670693821194339e-06,
+      "loss": 1.4611,
+      "step": 2977
+    },
+    {
+      "cosine_similarity": 0.09229007813914328,
+      "epoch": 2.7753960857409132,
+      "grad_norm": 1.3522428351292854,
+      "learning_rate": 1.6637901277183294e-06,
+      "loss": 1.7514,
+      "step": 2978
+    },
+    {
+      "cosine_similarity": 0.027289668281216675,
+      "epoch": 2.776328052190121,
+      "grad_norm": 1.2001178943573894,
+      "learning_rate": 1.6568864342423197e-06,
+      "loss": 1.3886,
+      "step": 2979
+    },
+    {
+      "cosine_similarity": 0.06868589956161265,
+      "epoch": 2.7772600186393293,
+      "grad_norm": 1.2725990881881626,
+      "learning_rate": 1.6499827407663102e-06,
+      "loss": 1.5752,
+      "step": 2980
+    },
+    {
+      "cosine_similarity": 0.03467487133504286,
+      "epoch": 2.7781919850885366,
+      "grad_norm": 1.271538853386436,
+      "learning_rate": 1.6430790472903005e-06,
+      "loss": 1.6063,
+      "step": 2981
+    },
+    {
+      "cosine_similarity": 0.14795076398294654,
+      "epoch": 2.779123951537745,
+      "grad_norm": 1.841572405141723,
+      "learning_rate": 1.6361753538142908e-06,
+      "loss": 1.9636,
+      "step": 2982
+    },
+    {
+      "cosine_similarity": 0.06609063075826593,
+      "epoch": 2.7800559179869526,
+      "grad_norm": 1.056640701823969,
+      "learning_rate": 1.629271660338281e-06,
+      "loss": 1.1445,
+      "step": 2983
+    },
+    {
+      "cosine_similarity": 0.09887059799142235,
+      "epoch": 2.7809878844361604,
+      "grad_norm": 1.0837823948237322,
+      "learning_rate": 1.6223679668622713e-06,
+      "loss": 1.5403,
+      "step": 2984
+    },
+    {
+      "cosine_similarity": 0.039327448879543266,
+      "epoch": 2.781919850885368,
+      "grad_norm": 1.1110899913411625,
+      "learning_rate": 1.615464273386262e-06,
+      "loss": 1.4301,
+      "step": 2985
+    },
+    {
+      "cosine_similarity": 0.05533929107804775,
+      "epoch": 2.782851817334576,
+      "grad_norm": 1.3842755954932766,
+      "learning_rate": 1.608560579910252e-06,
+      "loss": 1.2451,
+      "step": 2986
+    },
+    {
+      "cosine_similarity": -0.01628080066615463,
+      "epoch": 2.7837837837837838,
+      "grad_norm": 1.2988240378141587,
+      "learning_rate": 1.6016568864342424e-06,
+      "loss": 1.4018,
+      "step": 2987
+    },
+    {
+      "cosine_similarity": -0.0441464549512499,
+      "epoch": 2.7847157502329916,
+      "grad_norm": 1.144336991715197,
+      "learning_rate": 1.5947531929582327e-06,
+      "loss": 1.5843,
+      "step": 2988
+    },
+    {
+      "cosine_similarity": 0.033702803700910236,
+      "epoch": 2.7856477166821993,
+      "grad_norm": 1.1957593356248146,
+      "learning_rate": 1.587849499482223e-06,
+      "loss": 1.3691,
+      "step": 2989
+    },
+    {
+      "cosine_similarity": -0.07990810972593908,
+      "epoch": 2.786579683131407,
+      "grad_norm": 1.2333905785359278,
+      "learning_rate": 1.5809458060062135e-06,
+      "loss": 1.3497,
+      "step": 2990
+    },
+    {
+      "cosine_similarity": 0.010575510918491435,
+      "epoch": 2.7875116495806154,
+      "grad_norm": 1.034442338346615,
+      "learning_rate": 1.5740421125302038e-06,
+      "loss": 1.1616,
+      "step": 2991
+    },
+    {
+      "cosine_similarity": -0.006015846454074027,
+      "epoch": 2.7884436160298227,
+      "grad_norm": 1.0872778352803554,
+      "learning_rate": 1.5671384190541941e-06,
+      "loss": 1.3753,
+      "step": 2992
+    },
+    {
+      "cosine_similarity": -0.01389624928620214,
+      "epoch": 2.789375582479031,
+      "grad_norm": 1.186403847838167,
+      "learning_rate": 1.5602347255781845e-06,
+      "loss": 1.498,
+      "step": 2993
+    },
+    {
+      "cosine_similarity": 0.075692474211298,
+      "epoch": 2.7903075489282387,
+      "grad_norm": 1.2694705008786342,
+      "learning_rate": 1.5533310321021748e-06,
+      "loss": 1.3726,
+      "step": 2994
+    },
+    {
+      "cosine_similarity": 0.041208874814155944,
+      "epoch": 2.7912395153774465,
+      "grad_norm": 1.1187543631809247,
+      "learning_rate": 1.5464273386261653e-06,
+      "loss": 1.504,
+      "step": 2995
+    },
+    {
+      "cosine_similarity": -0.048240289157545906,
+      "epoch": 2.7921714818266543,
+      "grad_norm": 1.2465954008350837,
+      "learning_rate": 1.5395236451501556e-06,
+      "loss": 1.298,
+      "step": 2996
+    },
+    {
+      "cosine_similarity": -0.027977521828205464,
+      "epoch": 2.793103448275862,
+      "grad_norm": 1.1172679767465996,
+      "learning_rate": 1.5326199516741457e-06,
+      "loss": 1.2566,
+      "step": 2997
+    },
+    {
+      "cosine_similarity": 0.07233673278785063,
+      "epoch": 2.79403541472507,
+      "grad_norm": 1.1889448354984922,
+      "learning_rate": 1.525716258198136e-06,
+      "loss": 1.4878,
+      "step": 2998
+    },
+    {
+      "cosine_similarity": 0.060252106179058386,
+      "epoch": 2.7949673811742777,
+      "grad_norm": 1.1854326521437057,
+      "learning_rate": 1.5188125647221263e-06,
+      "loss": 1.7171,
+      "step": 2999
+    },
+    {
+      "cosine_similarity": 0.0905611132110013,
+      "epoch": 2.7958993476234855,
+      "grad_norm": 1.0359084944823633,
+      "learning_rate": 1.5119088712461168e-06,
+      "loss": 1.4698,
+      "step": 3000
+    },
+    {
+      "cosine_similarity": -0.02189445007954178,
+      "epoch": 2.7968313140726933,
+      "grad_norm": 1.073461138864537,
+      "learning_rate": 1.5050051777701071e-06,
+      "loss": 1.3113,
+      "step": 3001
+    },
+    {
+      "cosine_similarity": 0.051517986278813196,
+      "epoch": 2.7977632805219015,
+      "grad_norm": 1.0042205405590336,
+      "learning_rate": 1.4981014842940974e-06,
+      "loss": 1.5335,
+      "step": 3002
+    },
+    {
+      "cosine_similarity": 0.29324228353068427,
+      "epoch": 2.798695246971109,
+      "grad_norm": 1.0920634174599952,
+      "learning_rate": 1.4911977908180878e-06,
+      "loss": 1.3503,
+      "step": 3003
+    },
+    {
+      "cosine_similarity": -0.00302271111606227,
+      "epoch": 2.799627213420317,
+      "grad_norm": 1.3663326316207354,
+      "learning_rate": 1.484294097342078e-06,
+      "loss": 1.3276,
+      "step": 3004
+    },
+    {
+      "cosine_similarity": 0.005371053662578279,
+      "epoch": 2.800559179869525,
+      "grad_norm": 1.1984068037676272,
+      "learning_rate": 1.4773904038660686e-06,
+      "loss": 1.3968,
+      "step": 3005
+    },
+    {
+      "cosine_similarity": 0.14050086934404132,
+      "epoch": 2.8014911463187326,
+      "grad_norm": 1.1304019725453296,
+      "learning_rate": 1.470486710390059e-06,
+      "loss": 1.4763,
+      "step": 3006
+    },
+    {
+      "cosine_similarity": -0.06774952698690416,
+      "epoch": 2.8024231127679404,
+      "grad_norm": 1.0591941968587588,
+      "learning_rate": 1.4635830169140492e-06,
+      "loss": 1.5854,
+      "step": 3007
+    },
+    {
+      "cosine_similarity": 0.006582598746136971,
+      "epoch": 2.803355079217148,
+      "grad_norm": 1.1992782895114547,
+      "learning_rate": 1.4566793234380395e-06,
+      "loss": 1.7396,
+      "step": 3008
+    },
+    {
+      "cosine_similarity": 0.13157653060665442,
+      "epoch": 2.804287045666356,
+      "grad_norm": 1.1980308025232351,
+      "learning_rate": 1.4497756299620296e-06,
+      "loss": 1.6044,
+      "step": 3009
+    },
+    {
+      "cosine_similarity": 0.06313328358827353,
+      "epoch": 2.805219012115564,
+      "grad_norm": 1.1352779677712066,
+      "learning_rate": 1.4428719364860203e-06,
+      "loss": 1.451,
+      "step": 3010
+    },
+    {
+      "cosine_similarity": 0.06826624113252819,
+      "epoch": 2.8061509785647716,
+      "grad_norm": 1.1006023947754675,
+      "learning_rate": 1.4359682430100104e-06,
+      "loss": 1.4469,
+      "step": 3011
+    },
+    {
+      "cosine_similarity": 0.02692113986156486,
+      "epoch": 2.8070829450139794,
+      "grad_norm": 0.9521211979577202,
+      "learning_rate": 1.4290645495340007e-06,
+      "loss": 1.1575,
+      "step": 3012
+    },
+    {
+      "cosine_similarity": -0.024980062074049576,
+      "epoch": 2.8080149114631876,
+      "grad_norm": 1.1822115354207472,
+      "learning_rate": 1.422160856057991e-06,
+      "loss": 1.5528,
+      "step": 3013
+    },
+    {
+      "cosine_similarity": 0.1999577015940989,
+      "epoch": 2.808946877912395,
+      "grad_norm": 1.319290974881397,
+      "learning_rate": 1.4152571625819814e-06,
+      "loss": 1.3563,
+      "step": 3014
+    },
+    {
+      "cosine_similarity": 0.019952107797062708,
+      "epoch": 2.809878844361603,
+      "grad_norm": 1.1948178858458727,
+      "learning_rate": 1.4083534691059719e-06,
+      "loss": 1.5544,
+      "step": 3015
+    },
+    {
+      "cosine_similarity": 0.013757673583353766,
+      "epoch": 2.810810810810811,
+      "grad_norm": 1.1669762587789079,
+      "learning_rate": 1.4014497756299622e-06,
+      "loss": 1.2786,
+      "step": 3016
+    },
+    {
+      "cosine_similarity": -0.0872290283990722,
+      "epoch": 2.8117427772600188,
+      "grad_norm": 1.1259969234719147,
+      "learning_rate": 1.3945460821539525e-06,
+      "loss": 1.433,
+      "step": 3017
+    },
+    {
+      "cosine_similarity": 0.021985813458200255,
+      "epoch": 2.8126747437092265,
+      "grad_norm": 1.5877290749378274,
+      "learning_rate": 1.3876423886779428e-06,
+      "loss": 1.6947,
+      "step": 3018
+    },
+    {
+      "cosine_similarity": 0.1715689029499229,
+      "epoch": 2.8136067101584343,
+      "grad_norm": 0.9995985317257463,
+      "learning_rate": 1.3807386952019331e-06,
+      "loss": 1.416,
+      "step": 3019
+    },
+    {
+      "cosine_similarity": -0.008443398042993032,
+      "epoch": 2.814538676607642,
+      "grad_norm": 1.3040930588092519,
+      "learning_rate": 1.3738350017259236e-06,
+      "loss": 1.3651,
+      "step": 3020
+    },
+    {
+      "cosine_similarity": 0.04289392917278872,
+      "epoch": 2.81547064305685,
+      "grad_norm": 1.435524716411907,
+      "learning_rate": 1.366931308249914e-06,
+      "loss": 1.6445,
+      "step": 3021
+    },
+    {
+      "cosine_similarity": 0.032142212534415834,
+      "epoch": 2.8164026095060577,
+      "grad_norm": 1.2477447928310492,
+      "learning_rate": 1.3600276147739043e-06,
+      "loss": 1.5296,
+      "step": 3022
+    },
+    {
+      "cosine_similarity": -0.1803910381634531,
+      "epoch": 2.8173345759552655,
+      "grad_norm": 1.0581283012129805,
+      "learning_rate": 1.3531239212978944e-06,
+      "loss": 1.2885,
+      "step": 3023
+    },
+    {
+      "cosine_similarity": 0.004663849161150091,
+      "epoch": 2.8182665424044733,
+      "grad_norm": 1.1562194061091846,
+      "learning_rate": 1.3462202278218847e-06,
+      "loss": 1.2748,
+      "step": 3024
+    },
+    {
+      "cosine_similarity": -0.037896463100017645,
+      "epoch": 2.819198508853681,
+      "grad_norm": 1.300965691092093,
+      "learning_rate": 1.3393165343458752e-06,
+      "loss": 1.6671,
+      "step": 3025
+    },
+    {
+      "cosine_similarity": 0.09790562149557479,
+      "epoch": 2.8201304753028893,
+      "grad_norm": 1.2863747085586157,
+      "learning_rate": 1.3324128408698655e-06,
+      "loss": 1.3746,
+      "step": 3026
+    },
+    {
+      "cosine_similarity": 0.13505605653731972,
+      "epoch": 2.821062441752097,
+      "grad_norm": 1.370085989796952,
+      "learning_rate": 1.3255091473938558e-06,
+      "loss": 1.7576,
+      "step": 3027
+    },
+    {
+      "cosine_similarity": 0.05203942167968731,
+      "epoch": 2.821994408201305,
+      "grad_norm": 1.3130266950125329,
+      "learning_rate": 1.3186054539178461e-06,
+      "loss": 1.4375,
+      "step": 3028
+    },
+    {
+      "cosine_similarity": 0.07286942304792751,
+      "epoch": 2.8229263746505127,
+      "grad_norm": 1.0197710447458874,
+      "learning_rate": 1.3117017604418364e-06,
+      "loss": 1.3582,
+      "step": 3029
+    },
+    {
+      "cosine_similarity": -0.020459310223308446,
+      "epoch": 2.8238583410997204,
+      "grad_norm": 1.1998342250075678,
+      "learning_rate": 1.304798066965827e-06,
+      "loss": 1.3979,
+      "step": 3030
+    },
+    {
+      "cosine_similarity": 0.16804401220813975,
+      "epoch": 2.8247903075489282,
+      "grad_norm": 1.2004539086581878,
+      "learning_rate": 1.2978943734898173e-06,
+      "loss": 1.3413,
+      "step": 3031
+    },
+    {
+      "cosine_similarity": -0.07207726849469308,
+      "epoch": 2.825722273998136,
+      "grad_norm": 1.3038203463884135,
+      "learning_rate": 1.2909906800138076e-06,
+      "loss": 1.4436,
+      "step": 3032
+    },
+    {
+      "cosine_similarity": 0.011985231540749082,
+      "epoch": 2.826654240447344,
+      "grad_norm": 1.1931606453343435,
+      "learning_rate": 1.2840869865377979e-06,
+      "loss": 1.3051,
+      "step": 3033
+    },
+    {
+      "cosine_similarity": 0.010449347817291213,
+      "epoch": 2.8275862068965516,
+      "grad_norm": 1.1640659970466518,
+      "learning_rate": 1.277183293061788e-06,
+      "loss": 1.2479,
+      "step": 3034
+    },
+    {
+      "cosine_similarity": 0.013157222162232482,
+      "epoch": 2.8285181733457594,
+      "grad_norm": 1.5718553975253633,
+      "learning_rate": 1.2702795995857787e-06,
+      "loss": 1.6869,
+      "step": 3035
+    },
+    {
+      "cosine_similarity": 0.16373281680598795,
+      "epoch": 2.829450139794967,
+      "grad_norm": 1.1996295845358969,
+      "learning_rate": 1.2633759061097688e-06,
+      "loss": 1.3587,
+      "step": 3036
+    },
+    {
+      "cosine_similarity": 0.06492172581281265,
+      "epoch": 2.8303821062441754,
+      "grad_norm": 1.754657105787047,
+      "learning_rate": 1.2564722126337591e-06,
+      "loss": 1.6404,
+      "step": 3037
+    },
+    {
+      "cosine_similarity": 0.02779730426678639,
+      "epoch": 2.831314072693383,
+      "grad_norm": 1.229625260090617,
+      "learning_rate": 1.2495685191577494e-06,
+      "loss": 1.5804,
+      "step": 3038
+    },
+    {
+      "cosine_similarity": -0.06470547289186158,
+      "epoch": 2.832246039142591,
+      "grad_norm": 1.27713485674507,
+      "learning_rate": 1.24266482568174e-06,
+      "loss": 1.4823,
+      "step": 3039
+    },
+    {
+      "cosine_similarity": 0.030134439607464328,
+      "epoch": 2.8331780055917988,
+      "grad_norm": 1.1463679780280598,
+      "learning_rate": 1.2357611322057302e-06,
+      "loss": 1.4588,
+      "step": 3040
+    },
+    {
+      "cosine_similarity": 0.1157118160354448,
+      "epoch": 2.8341099720410066,
+      "grad_norm": 1.1612571777787448,
+      "learning_rate": 1.2288574387297203e-06,
+      "loss": 1.3513,
+      "step": 3041
+    },
+    {
+      "cosine_similarity": -0.0885938450459886,
+      "epoch": 2.8350419384902144,
+      "grad_norm": 1.084165524178909,
+      "learning_rate": 1.2219537452537109e-06,
+      "loss": 1.2898,
+      "step": 3042
+    },
+    {
+      "cosine_similarity": 0.007990837484749349,
+      "epoch": 2.835973904939422,
+      "grad_norm": 1.0230864054247668,
+      "learning_rate": 1.2150500517777012e-06,
+      "loss": 1.4645,
+      "step": 3043
+    },
+    {
+      "cosine_similarity": 0.022423230286898202,
+      "epoch": 2.83690587138863,
+      "grad_norm": 1.266890604381814,
+      "learning_rate": 1.2081463583016915e-06,
+      "loss": 1.3251,
+      "step": 3044
+    },
+    {
+      "cosine_similarity": 0.0045708830462408674,
+      "epoch": 2.8378378378378377,
+      "grad_norm": 1.1760970819185645,
+      "learning_rate": 1.2012426648256818e-06,
+      "loss": 1.3587,
+      "step": 3045
+    },
+    {
+      "cosine_similarity": -0.11896183343592492,
+      "epoch": 2.8387698042870455,
+      "grad_norm": 1.0420587031563229,
+      "learning_rate": 1.194338971349672e-06,
+      "loss": 1.5454,
+      "step": 3046
+    },
+    {
+      "cosine_similarity": -0.07611004210949388,
+      "epoch": 2.8397017707362533,
+      "grad_norm": 1.0451536771111596,
+      "learning_rate": 1.1874352778736626e-06,
+      "loss": 1.3651,
+      "step": 3047
+    },
+    {
+      "cosine_similarity": 0.11639627206997662,
+      "epoch": 2.8406337371854615,
+      "grad_norm": 1.2209983446681583,
+      "learning_rate": 1.1805315843976527e-06,
+      "loss": 1.1616,
+      "step": 3048
+    },
+    {
+      "cosine_similarity": 0.11898215658279715,
+      "epoch": 2.8415657036346693,
+      "grad_norm": 1.3565606161324066,
+      "learning_rate": 1.1736278909216432e-06,
+      "loss": 1.4076,
+      "step": 3049
+    },
+    {
+      "cosine_similarity": -0.018361237894728783,
+      "epoch": 2.842497670083877,
+      "grad_norm": 1.3304543968652471,
+      "learning_rate": 1.1667241974456335e-06,
+      "loss": 1.3788,
+      "step": 3050
+    },
+    {
+      "cosine_similarity": 0.25127190870245775,
+      "epoch": 2.843429636533085,
+      "grad_norm": 1.4490464810987442,
+      "learning_rate": 1.1598205039696239e-06,
+      "loss": 1.3861,
+      "step": 3051
+    },
+    {
+      "cosine_similarity": 0.0030808435810770244,
+      "epoch": 2.8443616029822927,
+      "grad_norm": 1.3996706966354038,
+      "learning_rate": 1.1529168104936142e-06,
+      "loss": 1.3574,
+      "step": 3052
+    },
+    {
+      "cosine_similarity": 0.0021362050592958703,
+      "epoch": 2.8452935694315005,
+      "grad_norm": 1.162040568461673,
+      "learning_rate": 1.1460131170176045e-06,
+      "loss": 1.3364,
+      "step": 3053
+    },
+    {
+      "cosine_similarity": -0.02603255880936225,
+      "epoch": 2.8462255358807083,
+      "grad_norm": 1.208702857424272,
+      "learning_rate": 1.139109423541595e-06,
+      "loss": 1.4089,
+      "step": 3054
+    },
+    {
+      "cosine_similarity": 0.13325283770862026,
+      "epoch": 2.847157502329916,
+      "grad_norm": 1.518104374446231,
+      "learning_rate": 1.132205730065585e-06,
+      "loss": 1.5541,
+      "step": 3055
+    },
+    {
+      "cosine_similarity": 0.080764184606684,
+      "epoch": 2.848089468779124,
+      "grad_norm": 1.2326565570125276,
+      "learning_rate": 1.1253020365895754e-06,
+      "loss": 1.4814,
+      "step": 3056
+    },
+    {
+      "cosine_similarity": 0.06534643353880232,
+      "epoch": 2.8490214352283316,
+      "grad_norm": 1.2019256955218942,
+      "learning_rate": 1.118398343113566e-06,
+      "loss": 1.6534,
+      "step": 3057
+    },
+    {
+      "cosine_similarity": 0.02796268846396516,
+      "epoch": 2.8499534016775394,
+      "grad_norm": 1.3998303846773208,
+      "learning_rate": 1.1114946496375562e-06,
+      "loss": 1.5511,
+      "step": 3058
+    },
+    {
+      "cosine_similarity": -0.011241638650746764,
+      "epoch": 2.8508853681267476,
+      "grad_norm": 1.3702763225157073,
+      "learning_rate": 1.1045909561615465e-06,
+      "loss": 1.6055,
+      "step": 3059
+    },
+    {
+      "cosine_similarity": -0.006343047454820089,
+      "epoch": 2.8518173345759554,
+      "grad_norm": 1.2770559549527314,
+      "learning_rate": 1.0976872626855368e-06,
+      "loss": 1.4197,
+      "step": 3060
+    },
+    {
+      "cosine_similarity": -0.01428077560890953,
+      "epoch": 2.852749301025163,
+      "grad_norm": 1.1830725129829347,
+      "learning_rate": 1.0907835692095272e-06,
+      "loss": 1.3352,
+      "step": 3061
+    },
+    {
+      "cosine_similarity": -0.0279850209046797,
+      "epoch": 2.853681267474371,
+      "grad_norm": 1.2595518266236714,
+      "learning_rate": 1.0838798757335175e-06,
+      "loss": 1.4065,
+      "step": 3062
+    },
+    {
+      "cosine_similarity": 0.06354324300186691,
+      "epoch": 2.854613233923579,
+      "grad_norm": 1.3257177756713339,
+      "learning_rate": 1.0769761822575078e-06,
+      "loss": 1.5475,
+      "step": 3063
+    },
+    {
+      "cosine_similarity": -0.04405156430041624,
+      "epoch": 2.8555452003727866,
+      "grad_norm": 1.1341182054373442,
+      "learning_rate": 1.0700724887814983e-06,
+      "loss": 1.4356,
+      "step": 3064
+    },
+    {
+      "cosine_similarity": -0.008325548427323526,
+      "epoch": 2.8564771668219944,
+      "grad_norm": 1.3115110154231797,
+      "learning_rate": 1.0631687953054886e-06,
+      "loss": 1.4169,
+      "step": 3065
+    },
+    {
+      "cosine_similarity": -0.009934166212058363,
+      "epoch": 2.857409133271202,
+      "grad_norm": 1.3259872340047116,
+      "learning_rate": 1.0562651018294787e-06,
+      "loss": 1.6281,
+      "step": 3066
+    },
+    {
+      "cosine_similarity": 0.1091284533742939,
+      "epoch": 2.85834109972041,
+      "grad_norm": 1.175066012442556,
+      "learning_rate": 1.0493614083534692e-06,
+      "loss": 1.5231,
+      "step": 3067
+    },
+    {
+      "cosine_similarity": -0.01176809817089287,
+      "epoch": 2.8592730661696177,
+      "grad_norm": 1.1558268326441223,
+      "learning_rate": 1.0424577148774595e-06,
+      "loss": 1.3547,
+      "step": 3068
+    },
+    {
+      "cosine_similarity": 0.06597756902521607,
+      "epoch": 2.8602050326188255,
+      "grad_norm": 0.9852781587844951,
+      "learning_rate": 1.0355540214014498e-06,
+      "loss": 1.2129,
+      "step": 3069
+    },
+    {
+      "cosine_similarity": 0.015185277817152633,
+      "epoch": 2.8611369990680338,
+      "grad_norm": 1.3568579726352312,
+      "learning_rate": 1.0286503279254402e-06,
+      "loss": 1.8962,
+      "step": 3070
+    },
+    {
+      "cosine_similarity": 0.014475201160685775,
+      "epoch": 2.862068965517241,
+      "grad_norm": 1.2927116823179512,
+      "learning_rate": 1.0217466344494305e-06,
+      "loss": 1.5729,
+      "step": 3071
+    },
+    {
+      "cosine_similarity": 0.04044569461405313,
+      "epoch": 2.8630009319664493,
+      "grad_norm": 1.1997656463073665,
+      "learning_rate": 1.014842940973421e-06,
+      "loss": 1.5128,
+      "step": 3072
+    },
+    {
+      "cosine_similarity": 0.34116853162211885,
+      "epoch": 2.863932898415657,
+      "grad_norm": 1.4738802541236953,
+      "learning_rate": 1.007939247497411e-06,
+      "loss": 1.717,
+      "step": 3073
+    },
+    {
+      "cosine_similarity": 0.12476735047782488,
+      "epoch": 2.864864864864865,
+      "grad_norm": 1.244455066514939,
+      "learning_rate": 1.0010355540214016e-06,
+      "loss": 1.3664,
+      "step": 3074
+    },
+    {
+      "cosine_similarity": -0.06788992789857586,
+      "epoch": 2.8657968313140727,
+      "grad_norm": 1.1669356567531664,
+      "learning_rate": 9.94131860545392e-07,
+      "loss": 1.4339,
+      "step": 3075
+    },
+    {
+      "cosine_similarity": -0.060989591768969564,
+      "epoch": 2.8667287977632805,
+      "grad_norm": 1.177924419982618,
+      "learning_rate": 9.872281670693822e-07,
+      "loss": 1.5166,
+      "step": 3076
+    },
+    {
+      "cosine_similarity": 0.07463671022190377,
+      "epoch": 2.8676607642124883,
+      "grad_norm": 1.1139016334533136,
+      "learning_rate": 9.803244735933725e-07,
+      "loss": 1.5297,
+      "step": 3077
+    },
+    {
+      "cosine_similarity": -0.025466312056052407,
+      "epoch": 2.868592730661696,
+      "grad_norm": 1.2180647160258429,
+      "learning_rate": 9.734207801173628e-07,
+      "loss": 1.7015,
+      "step": 3078
+    },
+    {
+      "cosine_similarity": 0.0045517801775941885,
+      "epoch": 2.869524697110904,
+      "grad_norm": 0.9903747546561251,
+      "learning_rate": 9.665170866413534e-07,
+      "loss": 1.1286,
+      "step": 3079
+    },
+    {
+      "cosine_similarity": -0.005601931489799377,
+      "epoch": 2.8704566635601116,
+      "grad_norm": 1.232111649312903,
+      "learning_rate": 9.596133931653435e-07,
+      "loss": 1.6882,
+      "step": 3080
+    },
+    {
+      "cosine_similarity": 0.06223357040260325,
+      "epoch": 2.87138863000932,
+      "grad_norm": 1.259934366991241,
+      "learning_rate": 9.527096996893339e-07,
+      "loss": 1.5545,
+      "step": 3081
+    },
+    {
+      "cosine_similarity": -0.09346282787288611,
+      "epoch": 2.872320596458527,
+      "grad_norm": 1.2260597128974529,
+      "learning_rate": 9.458060062133243e-07,
+      "loss": 1.2701,
+      "step": 3082
+    },
+    {
+      "cosine_similarity": 0.034286977672397634,
+      "epoch": 2.8732525629077355,
+      "grad_norm": 1.2233551574768105,
+      "learning_rate": 9.389023127373145e-07,
+      "loss": 1.3405,
+      "step": 3083
+    },
+    {
+      "cosine_similarity": 0.056096373225824336,
+      "epoch": 2.8741845293569432,
+      "grad_norm": 1.0867723398433464,
+      "learning_rate": 9.319986192613049e-07,
+      "loss": 1.4634,
+      "step": 3084
+    },
+    {
+      "cosine_similarity": 0.03964870314677504,
+      "epoch": 2.875116495806151,
+      "grad_norm": 1.0017267566091304,
+      "learning_rate": 9.250949257852952e-07,
+      "loss": 1.1775,
+      "step": 3085
+    },
+    {
+      "cosine_similarity": 0.11689658709044369,
+      "epoch": 2.876048462255359,
+      "grad_norm": 1.3202688252928214,
+      "learning_rate": 9.181912323092855e-07,
+      "loss": 1.5416,
+      "step": 3086
+    },
+    {
+      "cosine_similarity": -0.15671824814263674,
+      "epoch": 2.8769804287045666,
+      "grad_norm": 1.2719865632157705,
+      "learning_rate": 9.112875388332759e-07,
+      "loss": 1.2792,
+      "step": 3087
+    },
+    {
+      "cosine_similarity": 0.12099273079988397,
+      "epoch": 2.8779123951537744,
+      "grad_norm": 1.0557655968340696,
+      "learning_rate": 9.043838453572662e-07,
+      "loss": 1.5559,
+      "step": 3088
+    },
+    {
+      "cosine_similarity": -0.0696283889047845,
+      "epoch": 2.878844361602982,
+      "grad_norm": 1.2239391227842829,
+      "learning_rate": 8.974801518812564e-07,
+      "loss": 1.2918,
+      "step": 3089
+    },
+    {
+      "cosine_similarity": -0.023808545336709598,
+      "epoch": 2.87977632805219,
+      "grad_norm": 1.1545996511972914,
+      "learning_rate": 8.905764584052469e-07,
+      "loss": 1.3438,
+      "step": 3090
+    },
+    {
+      "cosine_similarity": -0.13380171884825873,
+      "epoch": 2.8807082945013978,
+      "grad_norm": 1.0641895995265305,
+      "learning_rate": 8.836727649292372e-07,
+      "loss": 1.4619,
+      "step": 3091
+    },
+    {
+      "cosine_similarity": 0.006890528277762262,
+      "epoch": 2.881640260950606,
+      "grad_norm": 1.2399392437342471,
+      "learning_rate": 8.767690714532276e-07,
+      "loss": 1.3104,
+      "step": 3092
+    },
+    {
+      "cosine_similarity": 0.001704009414680283,
+      "epoch": 2.8825722273998133,
+      "grad_norm": 1.0680460961858038,
+      "learning_rate": 8.698653779772179e-07,
+      "loss": 1.3557,
+      "step": 3093
+    },
+    {
+      "cosine_similarity": 0.08325086069563935,
+      "epoch": 2.8835041938490216,
+      "grad_norm": 1.0573133014998513,
+      "learning_rate": 8.629616845012082e-07,
+      "loss": 1.2597,
+      "step": 3094
+    },
+    {
+      "cosine_similarity": -0.03176755861358869,
+      "epoch": 2.8844361602982294,
+      "grad_norm": 1.093104820071848,
+      "learning_rate": 8.560579910251986e-07,
+      "loss": 1.4909,
+      "step": 3095
+    },
+    {
+      "cosine_similarity": 0.08547774458766384,
+      "epoch": 2.885368126747437,
+      "grad_norm": 1.280420475801882,
+      "learning_rate": 8.491542975491888e-07,
+      "loss": 1.3022,
+      "step": 3096
+    },
+    {
+      "cosine_similarity": 0.1127258651688385,
+      "epoch": 2.886300093196645,
+      "grad_norm": 1.3230432494624316,
+      "learning_rate": 8.422506040731792e-07,
+      "loss": 1.5842,
+      "step": 3097
+    },
+    {
+      "cosine_similarity": 0.06312002908774349,
+      "epoch": 2.8872320596458527,
+      "grad_norm": 1.0542151071732617,
+      "learning_rate": 8.353469105971695e-07,
+      "loss": 1.3951,
+      "step": 3098
+    },
+    {
+      "cosine_similarity": 0.08281693250961725,
+      "epoch": 2.8881640260950605,
+      "grad_norm": 1.1267080617221545,
+      "learning_rate": 8.284432171211599e-07,
+      "loss": 1.3724,
+      "step": 3099
+    },
+    {
+      "cosine_similarity": -0.05301079113828359,
+      "epoch": 2.8890959925442683,
+      "grad_norm": 1.1999557729986883,
+      "learning_rate": 8.215395236451503e-07,
+      "loss": 1.4091,
+      "step": 3100
+    },
+    {
+      "cosine_similarity": 0.07908702272180597,
+      "epoch": 2.890027958993476,
+      "grad_norm": 1.2236742317621938,
+      "learning_rate": 8.146358301691405e-07,
+      "loss": 1.1905,
+      "step": 3101
+    },
+    {
+      "cosine_similarity": -0.23671730443609065,
+      "epoch": 2.890959925442684,
+      "grad_norm": 1.5298970515582924,
+      "learning_rate": 8.07732136693131e-07,
+      "loss": 1.5351,
+      "step": 3102
+    },
+    {
+      "cosine_similarity": 0.018091644293003563,
+      "epoch": 2.891891891891892,
+      "grad_norm": 1.1679874558434917,
+      "learning_rate": 8.008284432171212e-07,
+      "loss": 1.3062,
+      "step": 3103
+    },
+    {
+      "cosine_similarity": 0.17606430347147634,
+      "epoch": 2.8928238583410995,
+      "grad_norm": 1.3226520344172952,
+      "learning_rate": 7.939247497411115e-07,
+      "loss": 1.5104,
+      "step": 3104
+    },
+    {
+      "cosine_similarity": 0.025461232514251878,
+      "epoch": 2.8937558247903077,
+      "grad_norm": 0.8620887533288201,
+      "learning_rate": 7.870210562651019e-07,
+      "loss": 1.2935,
+      "step": 3105
+    },
+    {
+      "cosine_similarity": 0.09008870489380878,
+      "epoch": 2.8946877912395155,
+      "grad_norm": 1.129127634993316,
+      "learning_rate": 7.801173627890922e-07,
+      "loss": 1.4377,
+      "step": 3106
+    },
+    {
+      "cosine_similarity": 0.08809536871406462,
+      "epoch": 2.8956197576887233,
+      "grad_norm": 0.9280100953671364,
+      "learning_rate": 7.732136693130826e-07,
+      "loss": 1.229,
+      "step": 3107
+    },
+    {
+      "cosine_similarity": 0.11955332602399711,
+      "epoch": 2.896551724137931,
+      "grad_norm": 1.3477874493114623,
+      "learning_rate": 7.663099758370728e-07,
+      "loss": 1.7686,
+      "step": 3108
+    },
+    {
+      "cosine_similarity": 0.01902950934563361,
+      "epoch": 2.897483690587139,
+      "grad_norm": 1.2022042602279472,
+      "learning_rate": 7.594062823610632e-07,
+      "loss": 1.504,
+      "step": 3109
+    },
+    {
+      "cosine_similarity": -0.02955028247489822,
+      "epoch": 2.8984156570363466,
+      "grad_norm": 1.0492862708512165,
+      "learning_rate": 7.525025888850536e-07,
+      "loss": 1.4302,
+      "step": 3110
+    },
+    {
+      "cosine_similarity": 0.04507068156847293,
+      "epoch": 2.8993476234855544,
+      "grad_norm": 1.0888148410046166,
+      "learning_rate": 7.455988954090439e-07,
+      "loss": 1.2487,
+      "step": 3111
+    },
+    {
+      "cosine_similarity": -0.09828615804891046,
+      "epoch": 2.900279589934762,
+      "grad_norm": 1.0382957424701642,
+      "learning_rate": 7.386952019330343e-07,
+      "loss": 1.0602,
+      "step": 3112
+    },
+    {
+      "cosine_similarity": 0.03035967623800214,
+      "epoch": 2.90121155638397,
+      "grad_norm": 1.4019912171145785,
+      "learning_rate": 7.317915084570246e-07,
+      "loss": 1.5452,
+      "step": 3113
+    },
+    {
+      "cosine_similarity": 0.011466192579404195,
+      "epoch": 2.9021435228331782,
+      "grad_norm": 1.1809528677045966,
+      "learning_rate": 7.248878149810148e-07,
+      "loss": 1.2676,
+      "step": 3114
+    },
+    {
+      "cosine_similarity": -0.05157202678460121,
+      "epoch": 2.9030754892823856,
+      "grad_norm": 1.1781940939956401,
+      "learning_rate": 7.179841215050052e-07,
+      "loss": 1.49,
+      "step": 3115
+    },
+    {
+      "cosine_similarity": 0.015883427687096708,
+      "epoch": 2.904007455731594,
+      "grad_norm": 1.0482506792878685,
+      "learning_rate": 7.110804280289955e-07,
+      "loss": 1.5533,
+      "step": 3116
+    },
+    {
+      "cosine_similarity": 0.18065348084275823,
+      "epoch": 2.9049394221808016,
+      "grad_norm": 1.7695032610714445,
+      "learning_rate": 7.041767345529859e-07,
+      "loss": 1.3808,
+      "step": 3117
+    },
+    {
+      "cosine_similarity": 0.046155751598469456,
+      "epoch": 2.9058713886300094,
+      "grad_norm": 1.0954964024624034,
+      "learning_rate": 6.972730410769763e-07,
+      "loss": 1.5059,
+      "step": 3118
+    },
+    {
+      "cosine_similarity": -0.1433022318949784,
+      "epoch": 2.906803355079217,
+      "grad_norm": 1.089071453412882,
+      "learning_rate": 6.903693476009666e-07,
+      "loss": 1.4036,
+      "step": 3119
+    },
+    {
+      "cosine_similarity": 0.07031468307304739,
+      "epoch": 2.907735321528425,
+      "grad_norm": 1.3257823933778907,
+      "learning_rate": 6.83465654124957e-07,
+      "loss": 1.5547,
+      "step": 3120
+    },
+    {
+      "cosine_similarity": 0.016633922846942345,
+      "epoch": 2.9086672879776327,
+      "grad_norm": 1.2817062981337033,
+      "learning_rate": 6.765619606489472e-07,
+      "loss": 1.5146,
+      "step": 3121
+    },
+    {
+      "cosine_similarity": 0.11237422797114521,
+      "epoch": 2.9095992544268405,
+      "grad_norm": 1.1682845407944975,
+      "learning_rate": 6.696582671729376e-07,
+      "loss": 1.3413,
+      "step": 3122
+    },
+    {
+      "cosine_similarity": 0.08165690116077404,
+      "epoch": 2.9105312208760483,
+      "grad_norm": 1.2360394586314203,
+      "learning_rate": 6.627545736969279e-07,
+      "loss": 1.4845,
+      "step": 3123
+    },
+    {
+      "cosine_similarity": -0.140436438789676,
+      "epoch": 2.911463187325256,
+      "grad_norm": 1.196528249802189,
+      "learning_rate": 6.558508802209182e-07,
+      "loss": 1.3967,
+      "step": 3124
+    },
+    {
+      "cosine_similarity": 0.2005497875171469,
+      "epoch": 2.9123951537744643,
+      "grad_norm": 1.321941643476772,
+      "learning_rate": 6.489471867449086e-07,
+      "loss": 1.3613,
+      "step": 3125
+    },
+    {
+      "cosine_similarity": 0.024103233827663054,
+      "epoch": 2.9133271202236717,
+      "grad_norm": 1.0517915018749189,
+      "learning_rate": 6.420434932688989e-07,
+      "loss": 1.1747,
+      "step": 3126
+    },
+    {
+      "cosine_similarity": 0.044347151281286086,
+      "epoch": 2.91425908667288,
+      "grad_norm": 1.1641280599341808,
+      "learning_rate": 6.351397997928894e-07,
+      "loss": 1.5781,
+      "step": 3127
+    },
+    {
+      "cosine_similarity": -0.06894579166704183,
+      "epoch": 2.9151910531220877,
+      "grad_norm": 1.2264145192396922,
+      "learning_rate": 6.282361063168796e-07,
+      "loss": 1.4127,
+      "step": 3128
+    },
+    {
+      "cosine_similarity": -0.11440585159639656,
+      "epoch": 2.9161230195712955,
+      "grad_norm": 1.5275884986455028,
+      "learning_rate": 6.2133241284087e-07,
+      "loss": 1.3152,
+      "step": 3129
+    },
+    {
+      "cosine_similarity": -0.0515024358655498,
+      "epoch": 2.9170549860205033,
+      "grad_norm": 1.237139313793367,
+      "learning_rate": 6.144287193648602e-07,
+      "loss": 1.5636,
+      "step": 3130
+    },
+    {
+      "cosine_similarity": 0.07090020028555255,
+      "epoch": 2.917986952469711,
+      "grad_norm": 1.3224928028576026,
+      "learning_rate": 6.075250258888506e-07,
+      "loss": 1.2955,
+      "step": 3131
+    },
+    {
+      "cosine_similarity": 0.0013010338637127168,
+      "epoch": 2.918918918918919,
+      "grad_norm": 1.1419772114965312,
+      "learning_rate": 6.006213324128409e-07,
+      "loss": 1.5081,
+      "step": 3132
+    },
+    {
+      "cosine_similarity": 0.09873811348454745,
+      "epoch": 2.9198508853681266,
+      "grad_norm": 1.1547970521618047,
+      "learning_rate": 5.937176389368313e-07,
+      "loss": 1.6086,
+      "step": 3133
+    },
+    {
+      "cosine_similarity": 0.06414586829495253,
+      "epoch": 2.9207828518173344,
+      "grad_norm": 1.2440039297537004,
+      "learning_rate": 5.868139454608216e-07,
+      "loss": 1.3601,
+      "step": 3134
+    },
+    {
+      "cosine_similarity": -0.050089011047063266,
+      "epoch": 2.9217148182665422,
+      "grad_norm": 1.2115281298944036,
+      "learning_rate": 5.799102519848119e-07,
+      "loss": 1.4477,
+      "step": 3135
+    },
+    {
+      "cosine_similarity": -0.016967881342346783,
+      "epoch": 2.9226467847157505,
+      "grad_norm": 1.0041654789389403,
+      "learning_rate": 5.730065585088022e-07,
+      "loss": 1.0557,
+      "step": 3136
+    },
+    {
+      "cosine_similarity": -0.09816126974861623,
+      "epoch": 2.923578751164958,
+      "grad_norm": 1.069492599649633,
+      "learning_rate": 5.661028650327925e-07,
+      "loss": 1.1318,
+      "step": 3137
+    },
+    {
+      "cosine_similarity": -0.09376223723695307,
+      "epoch": 2.924510717614166,
+      "grad_norm": 1.1842335998219944,
+      "learning_rate": 5.59199171556783e-07,
+      "loss": 1.3916,
+      "step": 3138
+    },
+    {
+      "cosine_similarity": -0.019818161539361983,
+      "epoch": 2.925442684063374,
+      "grad_norm": 1.015633587393299,
+      "learning_rate": 5.522954780807733e-07,
+      "loss": 1.3809,
+      "step": 3139
+    },
+    {
+      "cosine_similarity": 0.0229144409142705,
+      "epoch": 2.9263746505125816,
+      "grad_norm": 1.083302767172673,
+      "learning_rate": 5.453917846047636e-07,
+      "loss": 1.2168,
+      "step": 3140
+    },
+    {
+      "cosine_similarity": 0.036932859050199,
+      "epoch": 2.9273066169617894,
+      "grad_norm": 1.155973429513987,
+      "learning_rate": 5.384880911287539e-07,
+      "loss": 1.595,
+      "step": 3141
+    },
+    {
+      "cosine_similarity": 0.04259589724979375,
+      "epoch": 2.928238583410997,
+      "grad_norm": 1.306732968793612,
+      "learning_rate": 5.315843976527443e-07,
+      "loss": 1.641,
+      "step": 3142
+    },
+    {
+      "cosine_similarity": 0.06508939761121313,
+      "epoch": 2.929170549860205,
+      "grad_norm": 1.0127013223068007,
+      "learning_rate": 5.246807041767346e-07,
+      "loss": 1.2746,
+      "step": 3143
+    },
+    {
+      "cosine_similarity": -0.017903232798907633,
+      "epoch": 2.9301025163094128,
+      "grad_norm": 1.2615920341386428,
+      "learning_rate": 5.177770107007249e-07,
+      "loss": 1.5662,
+      "step": 3144
+    },
+    {
+      "cosine_similarity": -0.019023015614805013,
+      "epoch": 2.9310344827586206,
+      "grad_norm": 1.384678948497805,
+      "learning_rate": 5.108733172247152e-07,
+      "loss": 1.6453,
+      "step": 3145
+    },
+    {
+      "cosine_similarity": 0.1333799015980071,
+      "epoch": 2.9319664492078283,
+      "grad_norm": 1.2214101600709708,
+      "learning_rate": 5.039696237487055e-07,
+      "loss": 1.6466,
+      "step": 3146
+    },
+    {
+      "cosine_similarity": -0.06624284859313403,
+      "epoch": 2.9328984156570366,
+      "grad_norm": 1.103654501203258,
+      "learning_rate": 4.97065930272696e-07,
+      "loss": 1.2976,
+      "step": 3147
+    },
+    {
+      "cosine_similarity": 0.1066055130632148,
+      "epoch": 2.933830382106244,
+      "grad_norm": 1.0384640875500064,
+      "learning_rate": 4.901622367966863e-07,
+      "loss": 1.696,
+      "step": 3148
+    },
+    {
+      "cosine_similarity": 0.06255021716780061,
+      "epoch": 2.934762348555452,
+      "grad_norm": 1.0109469817431018,
+      "learning_rate": 4.832585433206767e-07,
+      "loss": 1.5451,
+      "step": 3149
+    },
+    {
+      "cosine_similarity": 0.00421943129977012,
+      "epoch": 2.93569431500466,
+      "grad_norm": 1.3501455777264795,
+      "learning_rate": 4.7635484984466693e-07,
+      "loss": 1.5,
+      "step": 3150
+    },
+    {
+      "cosine_similarity": -0.1056869369541365,
+      "epoch": 2.9366262814538677,
+      "grad_norm": 1.1204119252908713,
+      "learning_rate": 4.6945115636865724e-07,
+      "loss": 1.4344,
+      "step": 3151
+    },
+    {
+      "cosine_similarity": 0.03409373382231813,
+      "epoch": 2.9375582479030755,
+      "grad_norm": 1.3179686405486302,
+      "learning_rate": 4.625474628926476e-07,
+      "loss": 1.2641,
+      "step": 3152
+    },
+    {
+      "cosine_similarity": -0.0526905804690324,
+      "epoch": 2.9384902143522833,
+      "grad_norm": 1.5042627253498102,
+      "learning_rate": 4.5564376941663797e-07,
+      "loss": 1.6259,
+      "step": 3153
+    },
+    {
+      "cosine_similarity": -0.0609940400638201,
+      "epoch": 2.939422180801491,
+      "grad_norm": 1.1335411665048585,
+      "learning_rate": 4.487400759406282e-07,
+      "loss": 1.185,
+      "step": 3154
+    },
+    {
+      "cosine_similarity": -0.10499146620422241,
+      "epoch": 2.940354147250699,
+      "grad_norm": 1.1510120480537005,
+      "learning_rate": 4.418363824646186e-07,
+      "loss": 1.6979,
+      "step": 3155
+    },
+    {
+      "cosine_similarity": 0.044062739478931835,
+      "epoch": 2.9412861136999067,
+      "grad_norm": 1.2947091130201178,
+      "learning_rate": 4.3493268898860895e-07,
+      "loss": 1.2985,
+      "step": 3156
+    },
+    {
+      "cosine_similarity": -0.01086828491287457,
+      "epoch": 2.9422180801491145,
+      "grad_norm": 1.49291483584275,
+      "learning_rate": 4.280289955125993e-07,
+      "loss": 1.7791,
+      "step": 3157
+    },
+    {
+      "cosine_similarity": 0.0666876404114256,
+      "epoch": 2.9431500465983227,
+      "grad_norm": 1.0710356852330698,
+      "learning_rate": 4.211253020365896e-07,
+      "loss": 1.3567,
+      "step": 3158
+    },
+    {
+      "cosine_similarity": -0.05646934157091234,
+      "epoch": 2.94408201304753,
+      "grad_norm": 1.088444887634355,
+      "learning_rate": 4.142216085605799e-07,
+      "loss": 1.6053,
+      "step": 3159
+    },
+    {
+      "cosine_similarity": 0.037625372640025266,
+      "epoch": 2.9450139794967383,
+      "grad_norm": 1.329775805984992,
+      "learning_rate": 4.0731791508457024e-07,
+      "loss": 1.2887,
+      "step": 3160
+    },
+    {
+      "cosine_similarity": 0.03095936060379507,
+      "epoch": 2.945945945945946,
+      "grad_norm": 1.3087480402944114,
+      "learning_rate": 4.004142216085606e-07,
+      "loss": 1.4904,
+      "step": 3161
+    },
+    {
+      "cosine_similarity": -0.07688295853605968,
+      "epoch": 2.946877912395154,
+      "grad_norm": 1.2616130438273336,
+      "learning_rate": 3.9351052813255096e-07,
+      "loss": 1.3576,
+      "step": 3162
+    },
+    {
+      "cosine_similarity": 0.06437721867638427,
+      "epoch": 2.9478098788443616,
+      "grad_norm": 1.2743381931643007,
+      "learning_rate": 3.866068346565413e-07,
+      "loss": 1.4255,
+      "step": 3163
+    },
+    {
+      "cosine_similarity": -0.06886352195324952,
+      "epoch": 2.9487418452935694,
+      "grad_norm": 1.5609418777157669,
+      "learning_rate": 3.797031411805316e-07,
+      "loss": 1.6155,
+      "step": 3164
+    },
+    {
+      "cosine_similarity": 0.04610962919212773,
+      "epoch": 2.949673811742777,
+      "grad_norm": 1.2474367742115025,
+      "learning_rate": 3.7279944770452194e-07,
+      "loss": 1.445,
+      "step": 3165
+    },
+    {
+      "cosine_similarity": 0.04961631311116946,
+      "epoch": 2.950605778191985,
+      "grad_norm": 1.161901141856809,
+      "learning_rate": 3.658957542285123e-07,
+      "loss": 1.3388,
+      "step": 3166
+    },
+    {
+      "cosine_similarity": 0.05281566810351998,
+      "epoch": 2.951537744641193,
+      "grad_norm": 1.1285809781012717,
+      "learning_rate": 3.589920607525026e-07,
+      "loss": 1.4287,
+      "step": 3167
+    },
+    {
+      "cosine_similarity": -0.13044243052746982,
+      "epoch": 2.9524697110904006,
+      "grad_norm": 1.115766210057448,
+      "learning_rate": 3.5208836727649297e-07,
+      "loss": 1.5671,
+      "step": 3168
+    },
+    {
+      "cosine_similarity": -0.010530090077197646,
+      "epoch": 2.953401677539609,
+      "grad_norm": 1.1782859769022216,
+      "learning_rate": 3.451846738004833e-07,
+      "loss": 1.6147,
+      "step": 3169
+    },
+    {
+      "cosine_similarity": 0.03767981209974362,
+      "epoch": 2.954333643988816,
+      "grad_norm": 1.1536873571809345,
+      "learning_rate": 3.382809803244736e-07,
+      "loss": 1.5845,
+      "step": 3170
+    },
+    {
+      "cosine_similarity": -0.0276163121055913,
+      "epoch": 2.9552656104380244,
+      "grad_norm": 1.0586453001440237,
+      "learning_rate": 3.3137728684846395e-07,
+      "loss": 1.3822,
+      "step": 3171
+    },
+    {
+      "cosine_similarity": 0.08644952283690363,
+      "epoch": 2.956197576887232,
+      "grad_norm": 1.0554603028766059,
+      "learning_rate": 3.244735933724543e-07,
+      "loss": 1.4115,
+      "step": 3172
+    },
+    {
+      "cosine_similarity": 0.03933246580243102,
+      "epoch": 2.95712954333644,
+      "grad_norm": 1.040561961402387,
+      "learning_rate": 3.175698998964447e-07,
+      "loss": 1.5781,
+      "step": 3173
+    },
+    {
+      "cosine_similarity": 0.0103945180148109,
+      "epoch": 2.9580615097856477,
+      "grad_norm": 1.282905918742886,
+      "learning_rate": 3.10666206420435e-07,
+      "loss": 1.7926,
+      "step": 3174
+    },
+    {
+      "cosine_similarity": 0.031163377826900863,
+      "epoch": 2.9589934762348555,
+      "grad_norm": 0.9997589007267516,
+      "learning_rate": 3.037625129444253e-07,
+      "loss": 1.2392,
+      "step": 3175
+    },
+    {
+      "cosine_similarity": 0.005882491375697616,
+      "epoch": 2.9599254426840633,
+      "grad_norm": 1.0102624391631077,
+      "learning_rate": 2.9685881946841566e-07,
+      "loss": 1.1999,
+      "step": 3176
+    },
+    {
+      "cosine_similarity": 0.14372423859368016,
+      "epoch": 2.960857409133271,
+      "grad_norm": 1.09909468352956,
+      "learning_rate": 2.8995512599240596e-07,
+      "loss": 1.4106,
+      "step": 3177
+    },
+    {
+      "cosine_similarity": 0.22921888261581413,
+      "epoch": 2.961789375582479,
+      "grad_norm": 1.3623034688321665,
+      "learning_rate": 2.8305143251639627e-07,
+      "loss": 1.4971,
+      "step": 3178
+    },
+    {
+      "cosine_similarity": -0.0369662671670227,
+      "epoch": 2.9627213420316867,
+      "grad_norm": 1.420300764269607,
+      "learning_rate": 2.7614773904038664e-07,
+      "loss": 1.6066,
+      "step": 3179
+    },
+    {
+      "cosine_similarity": 0.037203664781265085,
+      "epoch": 2.963653308480895,
+      "grad_norm": 1.1231394164441506,
+      "learning_rate": 2.6924404556437694e-07,
+      "loss": 1.4355,
+      "step": 3180
+    },
+    {
+      "cosine_similarity": 0.016406812272183456,
+      "epoch": 2.9645852749301023,
+      "grad_norm": 1.6313236127733848,
+      "learning_rate": 2.623403520883673e-07,
+      "loss": 1.7213,
+      "step": 3181
+    },
+    {
+      "cosine_similarity": -0.007087693005625542,
+      "epoch": 2.9655172413793105,
+      "grad_norm": 1.076942855385941,
+      "learning_rate": 2.554366586123576e-07,
+      "loss": 1.4799,
+      "step": 3182
+    },
+    {
+      "cosine_similarity": -0.009053852748307653,
+      "epoch": 2.9664492078285183,
+      "grad_norm": 1.2017108000670809,
+      "learning_rate": 2.48532965136348e-07,
+      "loss": 1.3482,
+      "step": 3183
+    },
+    {
+      "cosine_similarity": 0.06827110380170534,
+      "epoch": 2.967381174277726,
+      "grad_norm": 1.2037479047162647,
+      "learning_rate": 2.4162927166033834e-07,
+      "loss": 1.6466,
+      "step": 3184
+    },
+    {
+      "cosine_similarity": -0.03747045853167023,
+      "epoch": 2.968313140726934,
+      "grad_norm": 1.1792161163058148,
+      "learning_rate": 2.3472557818432862e-07,
+      "loss": 1.3116,
+      "step": 3185
+    },
+    {
+      "cosine_similarity": -0.07028470263165568,
+      "epoch": 2.9692451071761417,
+      "grad_norm": 1.4718604796534382,
+      "learning_rate": 2.2782188470831898e-07,
+      "loss": 1.564,
+      "step": 3186
+    },
+    {
+      "cosine_similarity": 0.09975400705890668,
+      "epoch": 2.9701770736253494,
+      "grad_norm": 1.4286464864934898,
+      "learning_rate": 2.209181912323093e-07,
+      "loss": 1.4771,
+      "step": 3187
+    },
+    {
+      "cosine_similarity": 0.05863867799261423,
+      "epoch": 2.9711090400745572,
+      "grad_norm": 1.2093842992629646,
+      "learning_rate": 2.1401449775629965e-07,
+      "loss": 1.5067,
+      "step": 3188
+    },
+    {
+      "cosine_similarity": 0.08133797255841582,
+      "epoch": 2.972041006523765,
+      "grad_norm": 1.333096931057469,
+      "learning_rate": 2.0711080428028996e-07,
+      "loss": 1.3415,
+      "step": 3189
+    },
+    {
+      "cosine_similarity": -0.07815464541030571,
+      "epoch": 2.972972972972973,
+      "grad_norm": 1.1195602679886107,
+      "learning_rate": 2.002071108042803e-07,
+      "loss": 1.2297,
+      "step": 3190
+    },
+    {
+      "cosine_similarity": 0.017435094634934273,
+      "epoch": 2.973904939422181,
+      "grad_norm": 1.6071545502668927,
+      "learning_rate": 1.9330341732827066e-07,
+      "loss": 1.5243,
+      "step": 3191
+    },
+    {
+      "cosine_similarity": 0.13705120550116312,
+      "epoch": 2.9748369058713884,
+      "grad_norm": 1.2735284277786862,
+      "learning_rate": 1.8639972385226097e-07,
+      "loss": 1.5497,
+      "step": 3192
+    },
+    {
+      "cosine_similarity": 0.010688575784587576,
+      "epoch": 2.9757688723205966,
+      "grad_norm": 1.0832424933453988,
+      "learning_rate": 1.794960303762513e-07,
+      "loss": 1.4282,
+      "step": 3193
+    },
+    {
+      "cosine_similarity": -0.05446310910909935,
+      "epoch": 2.9767008387698044,
+      "grad_norm": 1.2762862969483424,
+      "learning_rate": 1.7259233690024164e-07,
+      "loss": 1.5214,
+      "step": 3194
+    },
+    {
+      "cosine_similarity": -0.06989602970782566,
+      "epoch": 2.977632805219012,
+      "grad_norm": 1.2192294547569573,
+      "learning_rate": 1.6568864342423198e-07,
+      "loss": 1.2961,
+      "step": 3195
+    },
+    {
+      "cosine_similarity": 0.03657694990214051,
+      "epoch": 2.97856477166822,
+      "grad_norm": 1.2548459947315351,
+      "learning_rate": 1.5878494994822234e-07,
+      "loss": 1.5162,
+      "step": 3196
+    },
+    {
+      "cosine_similarity": -0.019017535202793142,
+      "epoch": 2.9794967381174278,
+      "grad_norm": 1.1383196255969352,
+      "learning_rate": 1.5188125647221265e-07,
+      "loss": 1.4715,
+      "step": 3197
+    },
+    {
+      "cosine_similarity": 0.12890580641275284,
+      "epoch": 2.9804287045666356,
+      "grad_norm": 1.3818236414057892,
+      "learning_rate": 1.4497756299620298e-07,
+      "loss": 1.5549,
+      "step": 3198
+    },
+    {
+      "cosine_similarity": 0.06207058936630164,
+      "epoch": 2.9813606710158433,
+      "grad_norm": 1.1949446740417442,
+      "learning_rate": 1.3807386952019332e-07,
+      "loss": 1.3688,
+      "step": 3199
+    },
+    {
+      "cosine_similarity": 0.1430093314734709,
+      "epoch": 2.982292637465051,
+      "grad_norm": 1.2658026466168657,
+      "learning_rate": 1.3117017604418365e-07,
+      "loss": 1.5801,
+      "step": 3200
+    },
+    {
+      "cosine_similarity": 0.06954663781041548,
+      "epoch": 2.983224603914259,
+      "grad_norm": 1.2283952307160253,
+      "learning_rate": 1.24266482568174e-07,
+      "loss": 1.4812,
+      "step": 3201
+    },
+    {
+      "cosine_similarity": -0.3108500894613517,
+      "epoch": 2.984156570363467,
+      "grad_norm": 1.3345374278415378,
+      "learning_rate": 1.1736278909216431e-07,
+      "loss": 1.5328,
+      "step": 3202
+    },
+    {
+      "cosine_similarity": -0.02402119258535602,
+      "epoch": 2.9850885368126745,
+      "grad_norm": 1.1735761090827737,
+      "learning_rate": 1.1045909561615465e-07,
+      "loss": 1.6108,
+      "step": 3203
+    },
+    {
+      "cosine_similarity": -0.03941495355582928,
+      "epoch": 2.9860205032618827,
+      "grad_norm": 1.0324179277079395,
+      "learning_rate": 1.0355540214014498e-07,
+      "loss": 1.3134,
+      "step": 3204
+    },
+    {
+      "cosine_similarity": 0.01577749400700886,
+      "epoch": 2.9869524697110905,
+      "grad_norm": 1.3640397451530963,
+      "learning_rate": 9.665170866413533e-08,
+      "loss": 1.6427,
+      "step": 3205
+    },
+    {
+      "cosine_similarity": 0.018907811072116712,
+      "epoch": 2.9878844361602983,
+      "grad_norm": 1.1119326722667906,
+      "learning_rate": 8.974801518812565e-08,
+      "loss": 1.1447,
+      "step": 3206
+    },
+    {
+      "cosine_similarity": -0.001970988585483598,
+      "epoch": 2.988816402609506,
+      "grad_norm": 0.9915948382339399,
+      "learning_rate": 8.284432171211599e-08,
+      "loss": 1.3254,
+      "step": 3207
+    },
+    {
+      "cosine_similarity": -0.00476900922568537,
+      "epoch": 2.989748369058714,
+      "grad_norm": 1.3053203566826839,
+      "learning_rate": 7.594062823610632e-08,
+      "loss": 1.6079,
+      "step": 3208
+    },
+    {
+      "cosine_similarity": 0.027654979129922628,
+      "epoch": 2.9906803355079217,
+      "grad_norm": 1.3372073464784269,
+      "learning_rate": 6.903693476009666e-08,
+      "loss": 1.499,
+      "step": 3209
+    },
+    {
+      "cosine_similarity": 0.0041062766155067374,
+      "epoch": 2.9916123019571295,
+      "grad_norm": 0.9783766115384867,
+      "learning_rate": 6.2133241284087e-08,
+      "loss": 1.4116,
+      "step": 3210
+    },
+    {
+      "cosine_similarity": -0.14305740101462464,
+      "epoch": 2.9925442684063372,
+      "grad_norm": 1.30299294640863,
+      "learning_rate": 5.522954780807732e-08,
+      "loss": 1.7541,
+      "step": 3211
+    },
+    {
+      "cosine_similarity": -0.02234158577729623,
+      "epoch": 2.993476234855545,
+      "grad_norm": 1.4132164818364692,
+      "learning_rate": 4.8325854332067665e-08,
+      "loss": 1.3496,
+      "step": 3212
+    },
+    {
+      "cosine_similarity": 0.08451626502685572,
+      "epoch": 2.9944082013047533,
+      "grad_norm": 1.2588436358906074,
+      "learning_rate": 4.1422160856057994e-08,
+      "loss": 1.5847,
+      "step": 3213
+    },
+    {
+      "cosine_similarity": -0.0761023137691961,
+      "epoch": 2.9953401677539606,
+      "grad_norm": 1.0998611362540618,
+      "learning_rate": 3.451846738004833e-08,
+      "loss": 1.5355,
+      "step": 3214
+    },
+    {
+      "cosine_similarity": 0.026831858387630737,
+      "epoch": 2.996272134203169,
+      "grad_norm": 1.0398157173956808,
+      "learning_rate": 2.761477390403866e-08,
+      "loss": 1.2187,
+      "step": 3215
+    },
+    {
+      "cosine_similarity": -0.05542289615917284,
+      "epoch": 2.9972041006523766,
+      "grad_norm": 1.3911227162212385,
+      "learning_rate": 2.0711080428028997e-08,
+      "loss": 1.2267,
+      "step": 3216
+    },
+    {
+      "cosine_similarity": 0.02088869697201596,
+      "epoch": 2.9981360671015844,
+      "grad_norm": 1.1746502918031991,
+      "learning_rate": 1.380738695201933e-08,
+      "loss": 1.3578,
+      "step": 3217
+    },
+    {
+      "cosine_similarity": 0.021571808538570117,
+      "epoch": 2.999068033550792,
+      "grad_norm": 1.1765546343791717,
+      "learning_rate": 6.903693476009665e-09,
+      "loss": 1.4175,
+      "step": 3218
+    },
+    {
+      "cosine_similarity": -0.0519376256970489,
+      "epoch": 3.0,
+      "grad_norm": 0.9383853840429492,
+      "learning_rate": 0.0,
+      "loss": 1.0484,
+      "step": 3219
+    },
+    {
+      "epoch": 3.0,
+      "step": 3219,
+      "total_flos": 0.0,
+      "train_loss": 1.1290979329826447,
+      "train_runtime": 79608.9956,
+      "train_samples_per_second": 0.647,
+      "train_steps_per_second": 0.04
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3219,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}