diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.15009081402388685,
+  "epoch": 0.20012108536518244,
   "eval_steps": 500,
-  "global_step": 2727,
+  "global_step": 3636,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -19096,6 +19096,6369 @@
       "learning_rate": 9.870891892373397e-06,
       "loss": 0.8069,
       "step": 2727
+    },
+    {
+      "epoch": 0.1501458528262425,
+      "grad_norm": 0.8530564308166504,
+      "learning_rate": 9.870794006093188e-06,
+      "loss": 0.9229,
+      "step": 2728
+    },
+    {
+      "epoch": 0.15020089162859815,
+      "grad_norm": 0.7640067934989929,
+      "learning_rate": 9.870696083205394e-06,
+      "loss": 0.761,
+      "step": 2729
+    },
+    {
+      "epoch": 0.1502559304309538,
+      "grad_norm": 0.8953121900558472,
+      "learning_rate": 9.87059812371075e-06,
+      "loss": 0.8537,
+      "step": 2730
+    },
+    {
+      "epoch": 0.15031096923330947,
+      "grad_norm": 0.7779926657676697,
+      "learning_rate": 9.870500127609996e-06,
+      "loss": 0.8184,
+      "step": 2731
+    },
+    {
+      "epoch": 0.15036600803566513,
+      "grad_norm": 0.9181544184684753,
+      "learning_rate": 9.870402094903865e-06,
+      "loss": 0.8583,
+      "step": 2732
+    },
+    {
+      "epoch": 0.1504210468380208,
+      "grad_norm": 0.7629374861717224,
+      "learning_rate": 9.870304025593097e-06,
+      "loss": 0.6741,
+      "step": 2733
+    },
+    {
+      "epoch": 0.15047608564037646,
+      "grad_norm": 1.1455601453781128,
+      "learning_rate": 9.87020591967843e-06,
+      "loss": 0.8602,
+      "step": 2734
+    },
+    {
+      "epoch": 0.15053112444273212,
+      "grad_norm": 0.83924800157547,
+      "learning_rate": 9.870107777160596e-06,
+      "loss": 0.8847,
+      "step": 2735
+    },
+    {
+      "epoch": 0.15058616324508778,
+      "grad_norm": 0.9293402433395386,
+      "learning_rate": 9.870009598040336e-06,
+      "loss": 0.9008,
+      "step": 2736
+    },
+    {
+      "epoch": 0.15064120204744344,
+      "grad_norm": 0.8198057413101196,
+      "learning_rate": 9.869911382318389e-06,
+      "loss": 0.8004,
+      "step": 2737
+    },
+    {
+      "epoch": 0.1506962408497991,
+      "grad_norm": 0.8139753341674805,
+      "learning_rate": 9.86981312999549e-06,
+      "loss": 0.7316,
+      "step": 2738
+    },
+    {
+      "epoch": 0.15075127965215476,
+      "grad_norm": 0.854184091091156,
+      "learning_rate": 9.86971484107238e-06,
+      "loss": 0.9424,
+      "step": 2739
+    },
+    {
+      "epoch": 0.15080631845451042,
+      "grad_norm": 0.8626797199249268,
+      "learning_rate": 9.869616515549797e-06,
+      "loss": 0.8882,
+      "step": 2740
+    },
+    {
+      "epoch": 0.15086135725686609,
+      "grad_norm": 0.8447514176368713,
+      "learning_rate": 9.869518153428479e-06,
+      "loss": 0.7762,
+      "step": 2741
+    },
+    {
+      "epoch": 0.15091639605922175,
+      "grad_norm": 1.1359349489212036,
+      "learning_rate": 9.869419754709166e-06,
+      "loss": 0.9233,
+      "step": 2742
+    },
+    {
+      "epoch": 0.1509714348615774,
+      "grad_norm": 0.8095758557319641,
+      "learning_rate": 9.869321319392597e-06,
+      "loss": 0.8833,
+      "step": 2743
+    },
+    {
+      "epoch": 0.15102647366393307,
+      "grad_norm": 0.8364169001579285,
+      "learning_rate": 9.869222847479514e-06,
+      "loss": 0.833,
+      "step": 2744
+    },
+    {
+      "epoch": 0.15108151246628873,
+      "grad_norm": 0.7664803266525269,
+      "learning_rate": 9.869124338970653e-06,
+      "loss": 0.8125,
+      "step": 2745
+    },
+    {
+      "epoch": 0.1511365512686444,
+      "grad_norm": 0.8129634857177734,
+      "learning_rate": 9.86902579386676e-06,
+      "loss": 0.8277,
+      "step": 2746
+    },
+    {
+      "epoch": 0.15119159007100005,
+      "grad_norm": 0.8195592164993286,
+      "learning_rate": 9.86892721216857e-06,
+      "loss": 0.8489,
+      "step": 2747
+    },
+    {
+      "epoch": 0.15124662887335572,
+      "grad_norm": 0.8116651177406311,
+      "learning_rate": 9.868828593876827e-06,
+      "loss": 0.7831,
+      "step": 2748
+    },
+    {
+      "epoch": 0.15130166767571138,
+      "grad_norm": 0.8200114369392395,
+      "learning_rate": 9.868729938992272e-06,
+      "loss": 0.8956,
+      "step": 2749
+    },
+    {
+      "epoch": 0.15135670647806704,
+      "grad_norm": 0.8521816730499268,
+      "learning_rate": 9.868631247515645e-06,
+      "loss": 0.804,
+      "step": 2750
+    },
+    {
+      "epoch": 0.1514117452804227,
+      "grad_norm": 1.0386497974395752,
+      "learning_rate": 9.868532519447691e-06,
+      "loss": 0.8563,
+      "step": 2751
+    },
+    {
+      "epoch": 0.15146678408277836,
+      "grad_norm": 0.8345486521720886,
+      "learning_rate": 9.868433754789149e-06,
+      "loss": 0.9838,
+      "step": 2752
+    },
+    {
+      "epoch": 0.15152182288513402,
+      "grad_norm": 0.7207526564598083,
+      "learning_rate": 9.868334953540762e-06,
+      "loss": 0.6711,
+      "step": 2753
+    },
+    {
+      "epoch": 0.15157686168748968,
+      "grad_norm": 0.8159164786338806,
+      "learning_rate": 9.86823611570327e-06,
+      "loss": 0.7591,
+      "step": 2754
+    },
+    {
+      "epoch": 0.15163190048984534,
+      "grad_norm": 0.9062225818634033,
+      "learning_rate": 9.868137241277422e-06,
+      "loss": 0.8009,
+      "step": 2755
+    },
+    {
+      "epoch": 0.151686939292201,
+      "grad_norm": 0.8136696219444275,
+      "learning_rate": 9.868038330263957e-06,
+      "loss": 0.7014,
+      "step": 2756
+    },
+    {
+      "epoch": 0.15174197809455667,
+      "grad_norm": 0.7237691283226013,
+      "learning_rate": 9.867939382663618e-06,
+      "loss": 0.7766,
+      "step": 2757
+    },
+    {
+      "epoch": 0.15179701689691233,
+      "grad_norm": 0.8913742303848267,
+      "learning_rate": 9.86784039847715e-06,
+      "loss": 0.9362,
+      "step": 2758
+    },
+    {
+      "epoch": 0.151852055699268,
+      "grad_norm": 0.7763763070106506,
+      "learning_rate": 9.867741377705296e-06,
+      "loss": 0.7843,
+      "step": 2759
+    },
+    {
+      "epoch": 0.15190709450162365,
+      "grad_norm": 0.8973854780197144,
+      "learning_rate": 9.867642320348803e-06,
+      "loss": 0.911,
+      "step": 2760
+    },
+    {
+      "epoch": 0.1519621333039793,
+      "grad_norm": 0.7979685664176941,
+      "learning_rate": 9.86754322640841e-06,
+      "loss": 0.81,
+      "step": 2761
+    },
+    {
+      "epoch": 0.15201717210633497,
+      "grad_norm": 0.7740911841392517,
+      "learning_rate": 9.867444095884867e-06,
+      "loss": 0.8197,
+      "step": 2762
+    },
+    {
+      "epoch": 0.15207221090869064,
+      "grad_norm": 0.8400475978851318,
+      "learning_rate": 9.867344928778916e-06,
+      "loss": 0.8809,
+      "step": 2763
+    },
+    {
+      "epoch": 0.1521272497110463,
+      "grad_norm": 0.8995040655136108,
+      "learning_rate": 9.867245725091305e-06,
+      "loss": 0.8382,
+      "step": 2764
+    },
+    {
+      "epoch": 0.15218228851340196,
+      "grad_norm": 0.8162381052970886,
+      "learning_rate": 9.867146484822779e-06,
+      "loss": 0.9238,
+      "step": 2765
+    },
+    {
+      "epoch": 0.15223732731575762,
+      "grad_norm": 0.7668827176094055,
+      "learning_rate": 9.867047207974079e-06,
+      "loss": 0.8345,
+      "step": 2766
+    },
+    {
+      "epoch": 0.15229236611811328,
+      "grad_norm": 0.8719204664230347,
+      "learning_rate": 9.866947894545957e-06,
+      "loss": 0.7899,
+      "step": 2767
+    },
+    {
+      "epoch": 0.15234740492046894,
+      "grad_norm": 0.9043570756912231,
+      "learning_rate": 9.866848544539159e-06,
+      "loss": 0.8783,
+      "step": 2768
+    },
+    {
+      "epoch": 0.1524024437228246,
+      "grad_norm": 0.8859694004058838,
+      "learning_rate": 9.866749157954428e-06,
+      "loss": 0.862,
+      "step": 2769
+    },
+    {
+      "epoch": 0.15245748252518027,
+      "grad_norm": 1.022719144821167,
+      "learning_rate": 9.866649734792514e-06,
+      "loss": 0.8943,
+      "step": 2770
+    },
+    {
+      "epoch": 0.1525125213275359,
+      "grad_norm": 0.8710635900497437,
+      "learning_rate": 9.866550275054163e-06,
+      "loss": 0.7002,
+      "step": 2771
+    },
+    {
+      "epoch": 0.15256756012989156,
+      "grad_norm": 0.8482942581176758,
+      "learning_rate": 9.866450778740122e-06,
+      "loss": 0.7529,
+      "step": 2772
+    },
+    {
+      "epoch": 0.15262259893224722,
+      "grad_norm": 0.9637784361839294,
+      "learning_rate": 9.866351245851142e-06,
+      "loss": 0.8147,
+      "step": 2773
+    },
+    {
+      "epoch": 0.15267763773460288,
+      "grad_norm": 1.0472246408462524,
+      "learning_rate": 9.866251676387967e-06,
+      "loss": 0.8019,
+      "step": 2774
+    },
+    {
+      "epoch": 0.15273267653695854,
+      "grad_norm": 0.7916847467422485,
+      "learning_rate": 9.866152070351347e-06,
+      "loss": 0.7698,
+      "step": 2775
+    },
+    {
+      "epoch": 0.1527877153393142,
+      "grad_norm": 0.8421853184700012,
+      "learning_rate": 9.86605242774203e-06,
+      "loss": 0.8085,
+      "step": 2776
+    },
+    {
+      "epoch": 0.15284275414166987,
+      "grad_norm": 0.7990233898162842,
+      "learning_rate": 9.865952748560768e-06,
+      "loss": 0.8878,
+      "step": 2777
+    },
+    {
+      "epoch": 0.15289779294402553,
+      "grad_norm": 0.8017451167106628,
+      "learning_rate": 9.865853032808305e-06,
+      "loss": 0.8707,
+      "step": 2778
+    },
+    {
+      "epoch": 0.1529528317463812,
+      "grad_norm": 0.739850640296936,
+      "learning_rate": 9.865753280485393e-06,
+      "loss": 0.7884,
+      "step": 2779
+    },
+    {
+      "epoch": 0.15300787054873685,
+      "grad_norm": 1.0682430267333984,
+      "learning_rate": 9.865653491592784e-06,
+      "loss": 0.8548,
+      "step": 2780
+    },
+    {
+      "epoch": 0.1530629093510925,
+      "grad_norm": 0.7766296863555908,
+      "learning_rate": 9.865553666131225e-06,
+      "loss": 0.7786,
+      "step": 2781
+    },
+    {
+      "epoch": 0.15311794815344817,
+      "grad_norm": 0.8903290629386902,
+      "learning_rate": 9.865453804101466e-06,
+      "loss": 0.8978,
+      "step": 2782
+    },
+    {
+      "epoch": 0.15317298695580384,
+      "grad_norm": 0.8624514937400818,
+      "learning_rate": 9.86535390550426e-06,
+      "loss": 0.8472,
+      "step": 2783
+    },
+    {
+      "epoch": 0.1532280257581595,
+      "grad_norm": 0.7765294909477234,
+      "learning_rate": 9.865253970340356e-06,
+      "loss": 0.7702,
+      "step": 2784
+    },
+    {
+      "epoch": 0.15328306456051516,
+      "grad_norm": 0.9349095225334167,
+      "learning_rate": 9.865153998610504e-06,
+      "loss": 0.9154,
+      "step": 2785
+    },
+    {
+      "epoch": 0.15333810336287082,
+      "grad_norm": 0.8435478210449219,
+      "learning_rate": 9.865053990315458e-06,
+      "loss": 0.8986,
+      "step": 2786
+    },
+    {
+      "epoch": 0.15339314216522648,
+      "grad_norm": 0.8003486394882202,
+      "learning_rate": 9.864953945455968e-06,
+      "loss": 0.767,
+      "step": 2787
+    },
+    {
+      "epoch": 0.15344818096758214,
+      "grad_norm": 0.8060823678970337,
+      "learning_rate": 9.86485386403279e-06,
+      "loss": 0.8332,
+      "step": 2788
+    },
+    {
+      "epoch": 0.1535032197699378,
+      "grad_norm": 0.7914995551109314,
+      "learning_rate": 9.864753746046668e-06,
+      "loss": 0.6706,
+      "step": 2789
+    },
+    {
+      "epoch": 0.15355825857229347,
+      "grad_norm": 0.7792215943336487,
+      "learning_rate": 9.86465359149836e-06,
+      "loss": 0.8721,
+      "step": 2790
+    },
+    {
+      "epoch": 0.15361329737464913,
+      "grad_norm": 0.8572536110877991,
+      "learning_rate": 9.864553400388619e-06,
+      "loss": 0.8378,
+      "step": 2791
+    },
+    {
+      "epoch": 0.1536683361770048,
+      "grad_norm": 0.7645615339279175,
+      "learning_rate": 9.864453172718195e-06,
+      "loss": 0.6909,
+      "step": 2792
+    },
+    {
+      "epoch": 0.15372337497936045,
+      "grad_norm": 0.7627308964729309,
+      "learning_rate": 9.864352908487846e-06,
+      "loss": 0.7918,
+      "step": 2793
+    },
+    {
+      "epoch": 0.1537784137817161,
+      "grad_norm": 1.0830100774765015,
+      "learning_rate": 9.86425260769832e-06,
+      "loss": 0.9007,
+      "step": 2794
+    },
+    {
+      "epoch": 0.15383345258407177,
+      "grad_norm": 0.7667998671531677,
+      "learning_rate": 9.864152270350374e-06,
+      "loss": 0.832,
+      "step": 2795
+    },
+    {
+      "epoch": 0.15388849138642743,
+      "grad_norm": 0.9967591762542725,
+      "learning_rate": 9.864051896444764e-06,
+      "loss": 0.8917,
+      "step": 2796
+    },
+    {
+      "epoch": 0.1539435301887831,
+      "grad_norm": 0.8948462605476379,
+      "learning_rate": 9.86395148598224e-06,
+      "loss": 0.983,
+      "step": 2797
+    },
+    {
+      "epoch": 0.15399856899113876,
+      "grad_norm": 0.7857423424720764,
+      "learning_rate": 9.863851038963556e-06,
+      "loss": 0.7826,
+      "step": 2798
+    },
+    {
+      "epoch": 0.15405360779349442,
+      "grad_norm": 0.8821337223052979,
+      "learning_rate": 9.863750555389473e-06,
+      "loss": 0.8918,
+      "step": 2799
+    },
+    {
+      "epoch": 0.15410864659585008,
+      "grad_norm": 0.7896875143051147,
+      "learning_rate": 9.863650035260742e-06,
+      "loss": 0.8199,
+      "step": 2800
+    },
+    {
+      "epoch": 0.15416368539820574,
+      "grad_norm": 0.8046941161155701,
+      "learning_rate": 9.86354947857812e-06,
+      "loss": 0.8572,
+      "step": 2801
+    },
+    {
+      "epoch": 0.1542187242005614,
+      "grad_norm": 0.7266830205917358,
+      "learning_rate": 9.863448885342361e-06,
+      "loss": 0.8315,
+      "step": 2802
+    },
+    {
+      "epoch": 0.15427376300291706,
+      "grad_norm": 0.9009475708007812,
+      "learning_rate": 9.863348255554222e-06,
+      "loss": 0.7928,
+      "step": 2803
+    },
+    {
+      "epoch": 0.15432880180527273,
+      "grad_norm": 0.963364839553833,
+      "learning_rate": 9.863247589214459e-06,
+      "loss": 0.918,
+      "step": 2804
+    },
+    {
+      "epoch": 0.1543838406076284,
+      "grad_norm": 0.8278035521507263,
+      "learning_rate": 9.863146886323829e-06,
+      "loss": 0.8497,
+      "step": 2805
+    },
+    {
+      "epoch": 0.15443887940998405,
+      "grad_norm": 0.7360561490058899,
+      "learning_rate": 9.86304614688309e-06,
+      "loss": 0.676,
+      "step": 2806
+    },
+    {
+      "epoch": 0.1544939182123397,
+      "grad_norm": 0.7679837346076965,
+      "learning_rate": 9.862945370892996e-06,
+      "loss": 0.8114,
+      "step": 2807
+    },
+    {
+      "epoch": 0.15454895701469537,
+      "grad_norm": 0.8550567030906677,
+      "learning_rate": 9.862844558354309e-06,
+      "loss": 0.8222,
+      "step": 2808
+    },
+    {
+      "epoch": 0.15460399581705103,
+      "grad_norm": 0.7852397561073303,
+      "learning_rate": 9.86274370926778e-06,
+      "loss": 0.7449,
+      "step": 2809
+    },
+    {
+      "epoch": 0.1546590346194067,
+      "grad_norm": 0.9120833277702332,
+      "learning_rate": 9.862642823634175e-06,
+      "loss": 0.8702,
+      "step": 2810
+    },
+    {
+      "epoch": 0.15471407342176235,
+      "grad_norm": 0.8729703426361084,
+      "learning_rate": 9.862541901454246e-06,
+      "loss": 0.8064,
+      "step": 2811
+    },
+    {
+      "epoch": 0.15476911222411802,
+      "grad_norm": 0.7935470342636108,
+      "learning_rate": 9.862440942728754e-06,
+      "loss": 0.8502,
+      "step": 2812
+    },
+    {
+      "epoch": 0.15482415102647368,
+      "grad_norm": 0.8640689849853516,
+      "learning_rate": 9.86233994745846e-06,
+      "loss": 0.8159,
+      "step": 2813
+    },
+    {
+      "epoch": 0.1548791898288293,
+      "grad_norm": 0.9959222078323364,
+      "learning_rate": 9.862238915644116e-06,
+      "loss": 0.7767,
+      "step": 2814
+    },
+    {
+      "epoch": 0.15493422863118497,
+      "grad_norm": 0.7889506220817566,
+      "learning_rate": 9.862137847286487e-06,
+      "loss": 0.8293,
+      "step": 2815
+    },
+    {
+      "epoch": 0.15498926743354063,
+      "grad_norm": 0.8764606714248657,
+      "learning_rate": 9.862036742386335e-06,
+      "loss": 0.856,
+      "step": 2816
+    },
+    {
+      "epoch": 0.1550443062358963,
+      "grad_norm": 0.743727445602417,
+      "learning_rate": 9.861935600944413e-06,
+      "loss": 0.7099,
+      "step": 2817
+    },
+    {
+      "epoch": 0.15509934503825196,
+      "grad_norm": 0.7866224050521851,
+      "learning_rate": 9.861834422961485e-06,
+      "loss": 0.8805,
+      "step": 2818
+    },
+    {
+      "epoch": 0.15515438384060762,
+      "grad_norm": 0.8333723545074463,
+      "learning_rate": 9.861733208438311e-06,
+      "loss": 0.8486,
+      "step": 2819
+    },
+    {
+      "epoch": 0.15520942264296328,
+      "grad_norm": 0.8261659741401672,
+      "learning_rate": 9.861631957375652e-06,
+      "loss": 0.8896,
+      "step": 2820
+    },
+    {
+      "epoch": 0.15526446144531894,
+      "grad_norm": 0.8381538987159729,
+      "learning_rate": 9.861530669774268e-06,
+      "loss": 0.8686,
+      "step": 2821
+    },
+    {
+      "epoch": 0.1553195002476746,
+      "grad_norm": 0.9184440970420837,
+      "learning_rate": 9.861429345634923e-06,
+      "loss": 0.9702,
+      "step": 2822
+    },
+    {
+      "epoch": 0.15537453905003026,
+      "grad_norm": 0.8170294165611267,
+      "learning_rate": 9.861327984958374e-06,
+      "loss": 0.8298,
+      "step": 2823
+    },
+    {
+      "epoch": 0.15542957785238593,
+      "grad_norm": 0.8361968398094177,
+      "learning_rate": 9.861226587745385e-06,
+      "loss": 0.8232,
+      "step": 2824
+    },
+    {
+      "epoch": 0.1554846166547416,
+      "grad_norm": 0.7437820434570312,
+      "learning_rate": 9.861125153996718e-06,
+      "loss": 0.8271,
+      "step": 2825
+    },
+    {
+      "epoch": 0.15553965545709725,
+      "grad_norm": 0.715887188911438,
+      "learning_rate": 9.861023683713137e-06,
+      "loss": 0.7726,
+      "step": 2826
+    },
+    {
+      "epoch": 0.1555946942594529,
+      "grad_norm": 0.8358462452888489,
+      "learning_rate": 9.860922176895403e-06,
+      "loss": 0.8247,
+      "step": 2827
+    },
+    {
+      "epoch": 0.15564973306180857,
+      "grad_norm": 0.8620158433914185,
+      "learning_rate": 9.860820633544278e-06,
+      "loss": 0.8804,
+      "step": 2828
+    },
+    {
+      "epoch": 0.15570477186416423,
+      "grad_norm": 0.9035346508026123,
+      "learning_rate": 9.860719053660527e-06,
+      "loss": 0.7973,
+      "step": 2829
+    },
+    {
+      "epoch": 0.1557598106665199,
+      "grad_norm": 0.8014782071113586,
+      "learning_rate": 9.860617437244914e-06,
+      "loss": 0.7914,
+      "step": 2830
+    },
+    {
+      "epoch": 0.15581484946887555,
+      "grad_norm": 0.7788864970207214,
+      "learning_rate": 9.8605157842982e-06,
+      "loss": 0.7377,
+      "step": 2831
+    },
+    {
+      "epoch": 0.15586988827123122,
+      "grad_norm": 0.7475222945213318,
+      "learning_rate": 9.860414094821152e-06,
+      "loss": 0.7173,
+      "step": 2832
+    },
+    {
+      "epoch": 0.15592492707358688,
+      "grad_norm": 0.8866652846336365,
+      "learning_rate": 9.86031236881453e-06,
+      "loss": 0.8231,
+      "step": 2833
+    },
+    {
+      "epoch": 0.15597996587594254,
+      "grad_norm": 0.8725677728652954,
+      "learning_rate": 9.860210606279102e-06,
+      "loss": 0.9025,
+      "step": 2834
+    },
+    {
+      "epoch": 0.1560350046782982,
+      "grad_norm": 0.7608423233032227,
+      "learning_rate": 9.860108807215634e-06,
+      "loss": 0.8385,
+      "step": 2835
+    },
+    {
+      "epoch": 0.15609004348065386,
+      "grad_norm": 0.8237566351890564,
+      "learning_rate": 9.860006971624887e-06,
+      "loss": 0.8635,
+      "step": 2836
+    },
+    {
+      "epoch": 0.15614508228300952,
+      "grad_norm": 0.8078347444534302,
+      "learning_rate": 9.859905099507629e-06,
+      "loss": 0.7916,
+      "step": 2837
+    },
+    {
+      "epoch": 0.15620012108536518,
+      "grad_norm": 0.8282070755958557,
+      "learning_rate": 9.859803190864626e-06,
+      "loss": 0.8141,
+      "step": 2838
+    },
+    {
+      "epoch": 0.15625515988772085,
+      "grad_norm": 0.7639191150665283,
+      "learning_rate": 9.859701245696642e-06,
+      "loss": 0.7457,
+      "step": 2839
+    },
+    {
+      "epoch": 0.1563101986900765,
+      "grad_norm": 0.8429144620895386,
+      "learning_rate": 9.859599264004446e-06,
+      "loss": 0.9176,
+      "step": 2840
+    },
+    {
+      "epoch": 0.15636523749243217,
+      "grad_norm": 0.7792791724205017,
+      "learning_rate": 9.859497245788801e-06,
+      "loss": 0.8738,
+      "step": 2841
+    },
+    {
+      "epoch": 0.15642027629478783,
+      "grad_norm": 0.9018417596817017,
+      "learning_rate": 9.859395191050476e-06,
+      "loss": 0.841,
+      "step": 2842
+    },
+    {
+      "epoch": 0.1564753150971435,
+      "grad_norm": 0.7113705277442932,
+      "learning_rate": 9.859293099790239e-06,
+      "loss": 0.6576,
+      "step": 2843
+    },
+    {
+      "epoch": 0.15653035389949915,
+      "grad_norm": 0.8376311659812927,
+      "learning_rate": 9.859190972008853e-06,
+      "loss": 0.8559,
+      "step": 2844
+    },
+    {
+      "epoch": 0.15658539270185481,
+      "grad_norm": 0.7689141035079956,
+      "learning_rate": 9.859088807707092e-06,
+      "loss": 0.7844,
+      "step": 2845
+    },
+    {
+      "epoch": 0.15664043150421048,
+      "grad_norm": 0.7559483647346497,
+      "learning_rate": 9.858986606885717e-06,
+      "loss": 0.8676,
+      "step": 2846
+    },
+    {
+      "epoch": 0.15669547030656614,
+      "grad_norm": 0.7743827700614929,
+      "learning_rate": 9.8588843695455e-06,
+      "loss": 0.7995,
+      "step": 2847
+    },
+    {
+      "epoch": 0.1567505091089218,
+      "grad_norm": 0.8631327152252197,
+      "learning_rate": 9.85878209568721e-06,
+      "loss": 0.801,
+      "step": 2848
+    },
+    {
+      "epoch": 0.15680554791127746,
+      "grad_norm": 0.7454009056091309,
+      "learning_rate": 9.858679785311613e-06,
+      "loss": 0.8172,
+      "step": 2849
+    },
+    {
+      "epoch": 0.15686058671363312,
+      "grad_norm": 0.7915313839912415,
+      "learning_rate": 9.858577438419479e-06,
+      "loss": 0.833,
+      "step": 2850
+    },
+    {
+      "epoch": 0.15691562551598878,
+      "grad_norm": 0.8472526669502258,
+      "learning_rate": 9.858475055011578e-06,
+      "loss": 0.8249,
+      "step": 2851
+    },
+    {
+      "epoch": 0.15697066431834444,
+      "grad_norm": 0.7967580556869507,
+      "learning_rate": 9.85837263508868e-06,
+      "loss": 0.7533,
+      "step": 2852
+    },
+    {
+      "epoch": 0.1570257031207001,
+      "grad_norm": 0.7476257085800171,
+      "learning_rate": 9.858270178651554e-06,
+      "loss": 0.7918,
+      "step": 2853
+    },
+    {
+      "epoch": 0.15708074192305577,
+      "grad_norm": 0.8736184239387512,
+      "learning_rate": 9.858167685700968e-06,
+      "loss": 0.8254,
+      "step": 2854
+    },
+    {
+      "epoch": 0.15713578072541143,
+      "grad_norm": 0.8734819889068604,
+      "learning_rate": 9.858065156237694e-06,
+      "loss": 0.749,
+      "step": 2855
+    },
+    {
+      "epoch": 0.1571908195277671,
+      "grad_norm": 1.0344874858856201,
+      "learning_rate": 9.857962590262506e-06,
+      "loss": 0.9578,
+      "step": 2856
+    },
+    {
+      "epoch": 0.15724585833012272,
+      "grad_norm": 0.81183922290802,
+      "learning_rate": 9.857859987776168e-06,
+      "loss": 0.8845,
+      "step": 2857
+    },
+    {
+      "epoch": 0.15730089713247838,
+      "grad_norm": 0.8252540230751038,
+      "learning_rate": 9.857757348779456e-06,
+      "loss": 0.7862,
+      "step": 2858
+    },
+    {
+      "epoch": 0.15735593593483405,
+      "grad_norm": 0.7468119859695435,
+      "learning_rate": 9.85765467327314e-06,
+      "loss": 0.7587,
+      "step": 2859
+    },
+    {
+      "epoch": 0.1574109747371897,
+      "grad_norm": 0.8095998167991638,
+      "learning_rate": 9.857551961257993e-06,
+      "loss": 0.7467,
+      "step": 2860
+    },
+    {
+      "epoch": 0.15746601353954537,
+      "grad_norm": 0.8908564448356628,
+      "learning_rate": 9.857449212734785e-06,
+      "loss": 0.8199,
+      "step": 2861
+    },
+    {
+      "epoch": 0.15752105234190103,
+      "grad_norm": 0.7605593204498291,
+      "learning_rate": 9.857346427704288e-06,
+      "loss": 0.7196,
+      "step": 2862
+    },
+    {
+      "epoch": 0.1575760911442567,
+      "grad_norm": 0.9250784516334534,
+      "learning_rate": 9.857243606167276e-06,
+      "loss": 0.7366,
+      "step": 2863
+    },
+    {
+      "epoch": 0.15763112994661235,
+      "grad_norm": 0.882796585559845,
+      "learning_rate": 9.85714074812452e-06,
+      "loss": 0.8422,
+      "step": 2864
+    },
+    {
+      "epoch": 0.15768616874896801,
+      "grad_norm": 1.0014574527740479,
+      "learning_rate": 9.857037853576797e-06,
+      "loss": 0.8762,
+      "step": 2865
+    },
+    {
+      "epoch": 0.15774120755132368,
+      "grad_norm": 0.86713045835495,
+      "learning_rate": 9.856934922524877e-06,
+      "loss": 0.9282,
+      "step": 2866
+    },
+    {
+      "epoch": 0.15779624635367934,
+      "grad_norm": 1.1457390785217285,
+      "learning_rate": 9.856831954969532e-06,
+      "loss": 0.7947,
+      "step": 2867
+    },
+    {
+      "epoch": 0.157851285156035,
+      "grad_norm": 0.8902556896209717,
+      "learning_rate": 9.85672895091154e-06,
+      "loss": 0.928,
+      "step": 2868
+    },
+    {
+      "epoch": 0.15790632395839066,
+      "grad_norm": 0.7978467345237732,
+      "learning_rate": 9.856625910351674e-06,
+      "loss": 0.7382,
+      "step": 2869
+    },
+    {
+      "epoch": 0.15796136276074632,
+      "grad_norm": 0.741457462310791,
+      "learning_rate": 9.856522833290705e-06,
+      "loss": 0.7736,
+      "step": 2870
+    },
+    {
+      "epoch": 0.15801640156310198,
+      "grad_norm": 0.8330628871917725,
+      "learning_rate": 9.856419719729413e-06,
+      "loss": 0.8396,
+      "step": 2871
+    },
+    {
+      "epoch": 0.15807144036545764,
+      "grad_norm": 0.8771876692771912,
+      "learning_rate": 9.85631656966857e-06,
+      "loss": 0.6669,
+      "step": 2872
+    },
+    {
+      "epoch": 0.1581264791678133,
+      "grad_norm": 0.8073394298553467,
+      "learning_rate": 9.85621338310895e-06,
+      "loss": 0.8206,
+      "step": 2873
+    },
+    {
+      "epoch": 0.15818151797016897,
+      "grad_norm": 1.1058349609375,
+      "learning_rate": 9.85611016005133e-06,
+      "loss": 0.9526,
+      "step": 2874
+    },
+    {
+      "epoch": 0.15823655677252463,
+      "grad_norm": 0.7734992504119873,
+      "learning_rate": 9.856006900496488e-06,
+      "loss": 0.7477,
+      "step": 2875
+    },
+    {
+      "epoch": 0.1582915955748803,
+      "grad_norm": 0.9053219556808472,
+      "learning_rate": 9.855903604445196e-06,
+      "loss": 0.8009,
+      "step": 2876
+    },
+    {
+      "epoch": 0.15834663437723595,
+      "grad_norm": 0.8774041533470154,
+      "learning_rate": 9.855800271898233e-06,
+      "loss": 0.854,
+      "step": 2877
+    },
+    {
+      "epoch": 0.1584016731795916,
+      "grad_norm": 0.8346550464630127,
+      "learning_rate": 9.855696902856376e-06,
+      "loss": 0.7976,
+      "step": 2878
+    },
+    {
+      "epoch": 0.15845671198194727,
+      "grad_norm": 0.7781139016151428,
+      "learning_rate": 9.855593497320401e-06,
+      "loss": 0.7693,
+      "step": 2879
+    },
+    {
+      "epoch": 0.15851175078430293,
+      "grad_norm": 0.8707864880561829,
+      "learning_rate": 9.855490055291084e-06,
+      "loss": 0.882,
+      "step": 2880
+    },
+    {
+      "epoch": 0.1585667895866586,
+      "grad_norm": 0.7982275485992432,
+      "learning_rate": 9.855386576769203e-06,
+      "loss": 0.8457,
+      "step": 2881
+    },
+    {
+      "epoch": 0.15862182838901426,
+      "grad_norm": 0.7577090263366699,
+      "learning_rate": 9.855283061755536e-06,
+      "loss": 0.754,
+      "step": 2882
+    },
+    {
+      "epoch": 0.15867686719136992,
+      "grad_norm": 0.7422069311141968,
+      "learning_rate": 9.855179510250863e-06,
+      "loss": 0.673,
+      "step": 2883
+    },
+    {
+      "epoch": 0.15873190599372558,
+      "grad_norm": 0.7730041742324829,
+      "learning_rate": 9.85507592225596e-06,
+      "loss": 0.7888,
+      "step": 2884
+    },
+    {
+      "epoch": 0.15878694479608124,
+      "grad_norm": 0.7370560169219971,
+      "learning_rate": 9.854972297771605e-06,
+      "loss": 0.7762,
+      "step": 2885
+    },
+    {
+      "epoch": 0.1588419835984369,
+      "grad_norm": 0.725074291229248,
+      "learning_rate": 9.854868636798577e-06,
+      "loss": 0.7951,
+      "step": 2886
+    },
+    {
+      "epoch": 0.15889702240079256,
+      "grad_norm": 0.8088375926017761,
+      "learning_rate": 9.854764939337657e-06,
+      "loss": 0.8557,
+      "step": 2887
+    },
+    {
+      "epoch": 0.15895206120314823,
+      "grad_norm": 0.8268256187438965,
+      "learning_rate": 9.854661205389624e-06,
+      "loss": 0.7641,
+      "step": 2888
+    },
+    {
+      "epoch": 0.1590071000055039,
+      "grad_norm": 0.8079462051391602,
+      "learning_rate": 9.854557434955257e-06,
+      "loss": 0.7947,
+      "step": 2889
+    },
+    {
+      "epoch": 0.15906213880785955,
+      "grad_norm": 0.8257912993431091,
+      "learning_rate": 9.854453628035335e-06,
+      "loss": 0.771,
+      "step": 2890
+    },
+    {
+      "epoch": 0.1591171776102152,
+      "grad_norm": 0.8901774287223816,
+      "learning_rate": 9.85434978463064e-06,
+      "loss": 0.9415,
+      "step": 2891
+    },
+    {
+      "epoch": 0.15917221641257087,
+      "grad_norm": 0.8283013105392456,
+      "learning_rate": 9.854245904741948e-06,
+      "loss": 0.7267,
+      "step": 2892
+    },
+    {
+      "epoch": 0.15922725521492653,
+      "grad_norm": 0.8665382266044617,
+      "learning_rate": 9.854141988370045e-06,
+      "loss": 0.8681,
+      "step": 2893
+    },
+    {
+      "epoch": 0.1592822940172822,
+      "grad_norm": 0.786494255065918,
+      "learning_rate": 9.854038035515712e-06,
+      "loss": 0.7614,
+      "step": 2894
+    },
+    {
+      "epoch": 0.15933733281963786,
+      "grad_norm": 1.0548759698867798,
+      "learning_rate": 9.853934046179727e-06,
+      "loss": 0.861,
+      "step": 2895
+    },
+    {
+      "epoch": 0.15939237162199352,
+      "grad_norm": 0.8565425276756287,
+      "learning_rate": 9.853830020362873e-06,
+      "loss": 0.7858,
+      "step": 2896
+    },
+    {
+      "epoch": 0.15944741042434918,
+      "grad_norm": 0.7982691526412964,
+      "learning_rate": 9.853725958065933e-06,
+      "loss": 0.8797,
+      "step": 2897
+    },
+    {
+      "epoch": 0.15950244922670484,
+      "grad_norm": 0.8613169193267822,
+      "learning_rate": 9.853621859289686e-06,
+      "loss": 0.9217,
+      "step": 2898
+    },
+    {
+      "epoch": 0.1595574880290605,
+      "grad_norm": 0.950639009475708,
+      "learning_rate": 9.853517724034918e-06,
+      "loss": 0.8315,
+      "step": 2899
+    },
+    {
+      "epoch": 0.15961252683141613,
+      "grad_norm": 0.7940176129341125,
+      "learning_rate": 9.853413552302409e-06,
+      "loss": 0.7713,
+      "step": 2900
+    },
+    {
+      "epoch": 0.1596675656337718,
+      "grad_norm": 0.7716153264045715,
+      "learning_rate": 9.853309344092944e-06,
+      "loss": 0.7922,
+      "step": 2901
+    },
+    {
+      "epoch": 0.15972260443612746,
+      "grad_norm": 0.7626190781593323,
+      "learning_rate": 9.853205099407303e-06,
+      "loss": 0.7278,
+      "step": 2902
+    },
+    {
+      "epoch": 0.15977764323848312,
+      "grad_norm": 0.8523116707801819,
+      "learning_rate": 9.853100818246274e-06,
+      "loss": 0.8136,
+      "step": 2903
+    },
+    {
+      "epoch": 0.15983268204083878,
+      "grad_norm": 0.7636643052101135,
+      "learning_rate": 9.852996500610637e-06,
+      "loss": 0.6984,
+      "step": 2904
+    },
+    {
+      "epoch": 0.15988772084319444,
+      "grad_norm": 0.799201250076294,
+      "learning_rate": 9.852892146501179e-06,
+      "loss": 0.8319,
+      "step": 2905
+    },
+    {
+      "epoch": 0.1599427596455501,
+      "grad_norm": 0.7743694186210632,
+      "learning_rate": 9.85278775591868e-06,
+      "loss": 0.81,
+      "step": 2906
+    },
+    {
+      "epoch": 0.15999779844790576,
+      "grad_norm": 0.8964856863021851,
+      "learning_rate": 9.85268332886393e-06,
+      "loss": 0.9227,
+      "step": 2907
+    },
+    {
+      "epoch": 0.16005283725026143,
+      "grad_norm": 0.8809369802474976,
+      "learning_rate": 9.852578865337708e-06,
+      "loss": 0.9285,
+      "step": 2908
+    },
+    {
+      "epoch": 0.1601078760526171,
+      "grad_norm": 0.8960002064704895,
+      "learning_rate": 9.852474365340806e-06,
+      "loss": 0.8611,
+      "step": 2909
+    },
+    {
+      "epoch": 0.16016291485497275,
+      "grad_norm": 0.7539754509925842,
+      "learning_rate": 9.852369828874002e-06,
+      "loss": 0.7455,
+      "step": 2910
+    },
+    {
+      "epoch": 0.1602179536573284,
+      "grad_norm": 0.8189692497253418,
+      "learning_rate": 9.852265255938088e-06,
+      "loss": 0.8321,
+      "step": 2911
+    },
+    {
+      "epoch": 0.16027299245968407,
+      "grad_norm": 0.8708549737930298,
+      "learning_rate": 9.852160646533844e-06,
+      "loss": 0.8373,
+      "step": 2912
+    },
+    {
+      "epoch": 0.16032803126203973,
+      "grad_norm": 0.7701451778411865,
+      "learning_rate": 9.852056000662063e-06,
+      "loss": 0.805,
+      "step": 2913
+    },
+    {
+      "epoch": 0.1603830700643954,
+      "grad_norm": 0.9111948609352112,
+      "learning_rate": 9.851951318323526e-06,
+      "loss": 0.8513,
+      "step": 2914
+    },
+    {
+      "epoch": 0.16043810886675106,
+      "grad_norm": 0.7863909602165222,
+      "learning_rate": 9.85184659951902e-06,
+      "loss": 0.7856,
+      "step": 2915
+    },
+    {
+      "epoch": 0.16049314766910672,
+      "grad_norm": 0.9000817537307739,
+      "learning_rate": 9.851741844249336e-06,
+      "loss": 0.9172,
+      "step": 2916
+    },
+    {
+      "epoch": 0.16054818647146238,
+      "grad_norm": 1.0953118801116943,
+      "learning_rate": 9.851637052515259e-06,
+      "loss": 0.8564,
+      "step": 2917
+    },
+    {
+      "epoch": 0.16060322527381804,
+      "grad_norm": 0.8405389785766602,
+      "learning_rate": 9.851532224317575e-06,
+      "loss": 0.8317,
+      "step": 2918
+    },
+    {
+      "epoch": 0.1606582640761737,
+      "grad_norm": 0.8524565100669861,
+      "learning_rate": 9.851427359657075e-06,
+      "loss": 0.8765,
+      "step": 2919
+    },
+    {
+      "epoch": 0.16071330287852936,
+      "grad_norm": 0.8234089016914368,
+      "learning_rate": 9.851322458534546e-06,
+      "loss": 0.7873,
+      "step": 2920
+    },
+    {
+      "epoch": 0.16076834168088502,
+      "grad_norm": 0.7879638671875,
+      "learning_rate": 9.851217520950775e-06,
+      "loss": 0.8394,
+      "step": 2921
+    },
+    {
+      "epoch": 0.16082338048324069,
+      "grad_norm": 0.8168820738792419,
+      "learning_rate": 9.851112546906552e-06,
+      "loss": 0.8223,
+      "step": 2922
+    },
+    {
+      "epoch": 0.16087841928559635,
+      "grad_norm": 0.9423845410346985,
+      "learning_rate": 9.851007536402666e-06,
+      "loss": 0.9256,
+      "step": 2923
+    },
+    {
+      "epoch": 0.160933458087952,
+      "grad_norm": 0.7875099778175354,
+      "learning_rate": 9.850902489439906e-06,
+      "loss": 0.8199,
+      "step": 2924
+    },
+    {
+      "epoch": 0.16098849689030767,
+      "grad_norm": 0.6934793591499329,
+      "learning_rate": 9.85079740601906e-06,
+      "loss": 0.671,
+      "step": 2925
+    },
+    {
+      "epoch": 0.16104353569266333,
+      "grad_norm": 0.8172206282615662,
+      "learning_rate": 9.85069228614092e-06,
+      "loss": 0.7633,
+      "step": 2926
+    },
+    {
+      "epoch": 0.161098574495019,
+      "grad_norm": 0.72749263048172,
+      "learning_rate": 9.850587129806274e-06,
+      "loss": 0.8719,
+      "step": 2927
+    },
+    {
+      "epoch": 0.16115361329737465,
+      "grad_norm": 0.8416743874549866,
+      "learning_rate": 9.850481937015917e-06,
+      "loss": 0.8438,
+      "step": 2928
+    },
+    {
+      "epoch": 0.16120865209973032,
+      "grad_norm": 0.7415444850921631,
+      "learning_rate": 9.850376707770633e-06,
+      "loss": 0.7673,
+      "step": 2929
+    },
+    {
+      "epoch": 0.16126369090208598,
+      "grad_norm": 0.9364289045333862,
+      "learning_rate": 9.850271442071217e-06,
+      "loss": 0.7224,
+      "step": 2930
+    },
+    {
+      "epoch": 0.16131872970444164,
+      "grad_norm": 0.7314212918281555,
+      "learning_rate": 9.85016613991846e-06,
+      "loss": 0.7759,
+      "step": 2931
+    },
+    {
+      "epoch": 0.1613737685067973,
+      "grad_norm": 0.8940219283103943,
+      "learning_rate": 9.850060801313151e-06,
+      "loss": 0.8432,
+      "step": 2932
+    },
+    {
+      "epoch": 0.16142880730915296,
+      "grad_norm": 0.7499691843986511,
+      "learning_rate": 9.849955426256084e-06,
+      "loss": 0.8171,
+      "step": 2933
+    },
+    {
+      "epoch": 0.16148384611150862,
+      "grad_norm": 0.8123053312301636,
+      "learning_rate": 9.84985001474805e-06,
+      "loss": 0.7839,
+      "step": 2934
+    },
+    {
+      "epoch": 0.16153888491386428,
+      "grad_norm": 0.819618821144104,
+      "learning_rate": 9.849744566789842e-06,
+      "loss": 0.9123,
+      "step": 2935
+    },
+    {
+      "epoch": 0.16159392371621994,
+      "grad_norm": 0.791088342666626,
+      "learning_rate": 9.849639082382251e-06,
+      "loss": 0.8347,
+      "step": 2936
+    },
+    {
+      "epoch": 0.1616489625185756,
+      "grad_norm": 0.8166706562042236,
+      "learning_rate": 9.849533561526072e-06,
+      "loss": 0.8309,
+      "step": 2937
+    },
+    {
+      "epoch": 0.16170400132093127,
+      "grad_norm": 0.7944774031639099,
+      "learning_rate": 9.849428004222098e-06,
+      "loss": 0.8387,
+      "step": 2938
+    },
+    {
+      "epoch": 0.16175904012328693,
+      "grad_norm": 0.7414719462394714,
+      "learning_rate": 9.849322410471119e-06,
+      "loss": 0.71,
+      "step": 2939
+    },
+    {
+      "epoch": 0.1618140789256426,
+      "grad_norm": 0.8983511924743652,
+      "learning_rate": 9.849216780273931e-06,
+      "loss": 0.8902,
+      "step": 2940
+    },
+    {
+      "epoch": 0.16186911772799825,
+      "grad_norm": 0.9058687686920166,
+      "learning_rate": 9.849111113631329e-06,
+      "loss": 0.8804,
+      "step": 2941
+    },
+    {
+      "epoch": 0.1619241565303539,
+      "grad_norm": 0.948871374130249,
+      "learning_rate": 9.849005410544105e-06,
+      "loss": 0.9871,
+      "step": 2942
+    },
+    {
+      "epoch": 0.16197919533270955,
+      "grad_norm": 0.8240115642547607,
+      "learning_rate": 9.848899671013055e-06,
+      "loss": 0.8708,
+      "step": 2943
+    },
+    {
+      "epoch": 0.1620342341350652,
+      "grad_norm": 0.879953145980835,
+      "learning_rate": 9.848793895038972e-06,
+      "loss": 0.9279,
+      "step": 2944
+    },
+    {
+      "epoch": 0.16208927293742087,
+      "grad_norm": 0.8464690446853638,
+      "learning_rate": 9.848688082622653e-06,
+      "loss": 0.8418,
+      "step": 2945
+    },
+    {
+      "epoch": 0.16214431173977653,
+      "grad_norm": 0.8990732431411743,
+      "learning_rate": 9.848582233764891e-06,
+      "loss": 0.8622,
+      "step": 2946
+    },
+    {
+      "epoch": 0.1621993505421322,
+      "grad_norm": 0.8054911494255066,
+      "learning_rate": 9.848476348466483e-06,
+      "loss": 0.8295,
+      "step": 2947
+    },
+    {
+      "epoch": 0.16225438934448785,
+      "grad_norm": 0.7904845476150513,
+      "learning_rate": 9.848370426728226e-06,
+      "loss": 0.7777,
+      "step": 2948
+    },
+    {
+      "epoch": 0.16230942814684352,
+      "grad_norm": 1.0143954753875732,
+      "learning_rate": 9.848264468550915e-06,
+      "loss": 0.8556,
+      "step": 2949
+    },
+    {
+      "epoch": 0.16236446694919918,
+      "grad_norm": 0.7201125621795654,
+      "learning_rate": 9.848158473935344e-06,
+      "loss": 0.7981,
+      "step": 2950
+    },
+    {
+      "epoch": 0.16241950575155484,
+      "grad_norm": 0.8322157263755798,
+      "learning_rate": 9.848052442882312e-06,
+      "loss": 0.8323,
+      "step": 2951
+    },
+    {
+      "epoch": 0.1624745445539105,
+      "grad_norm": 0.7740346193313599,
+      "learning_rate": 9.847946375392617e-06,
+      "loss": 0.8355,
+      "step": 2952
+    },
+    {
+      "epoch": 0.16252958335626616,
+      "grad_norm": 0.8955645561218262,
+      "learning_rate": 9.847840271467053e-06,
+      "loss": 0.7161,
+      "step": 2953
+    },
+    {
+      "epoch": 0.16258462215862182,
+      "grad_norm": 0.800364077091217,
+      "learning_rate": 9.847734131106421e-06,
+      "loss": 0.8165,
+      "step": 2954
+    },
+    {
+      "epoch": 0.16263966096097748,
+      "grad_norm": 0.8305484056472778,
+      "learning_rate": 9.847627954311516e-06,
+      "loss": 0.7846,
+      "step": 2955
+    },
+    {
+      "epoch": 0.16269469976333314,
+      "grad_norm": 0.7354590892791748,
+      "learning_rate": 9.847521741083136e-06,
+      "loss": 0.7743,
+      "step": 2956
+    },
+    {
+      "epoch": 0.1627497385656888,
+      "grad_norm": 0.8173812627792358,
+      "learning_rate": 9.847415491422083e-06,
+      "loss": 0.8626,
+      "step": 2957
+    },
+    {
+      "epoch": 0.16280477736804447,
+      "grad_norm": 0.7959356307983398,
+      "learning_rate": 9.84730920532915e-06,
+      "loss": 0.8016,
+      "step": 2958
+    },
+    {
+      "epoch": 0.16285981617040013,
+      "grad_norm": 0.8256500363349915,
+      "learning_rate": 9.84720288280514e-06,
+      "loss": 0.7407,
+      "step": 2959
+    },
+    {
+      "epoch": 0.1629148549727558,
+      "grad_norm": 0.8522148728370667,
+      "learning_rate": 9.84709652385085e-06,
+      "loss": 0.8342,
+      "step": 2960
+    },
+    {
+      "epoch": 0.16296989377511145,
+      "grad_norm": 0.7791039943695068,
+      "learning_rate": 9.84699012846708e-06,
+      "loss": 0.7239,
+      "step": 2961
+    },
+    {
+      "epoch": 0.1630249325774671,
+      "grad_norm": 0.84294193983078,
+      "learning_rate": 9.84688369665463e-06,
+      "loss": 0.7498,
+      "step": 2962
+    },
+    {
+      "epoch": 0.16307997137982277,
+      "grad_norm": 0.7948899865150452,
+      "learning_rate": 9.846777228414299e-06,
+      "loss": 0.7713,
+      "step": 2963
+    },
+    {
+      "epoch": 0.16313501018217844,
+      "grad_norm": 0.6673180460929871,
+      "learning_rate": 9.846670723746888e-06,
+      "loss": 0.6759,
+      "step": 2964
+    },
+    {
+      "epoch": 0.1631900489845341,
+      "grad_norm": 0.8141015768051147,
+      "learning_rate": 9.846564182653199e-06,
+      "loss": 0.7928,
+      "step": 2965
+    },
+    {
+      "epoch": 0.16324508778688976,
+      "grad_norm": 0.967830240726471,
+      "learning_rate": 9.846457605134028e-06,
+      "loss": 0.823,
+      "step": 2966
+    },
+    {
+      "epoch": 0.16330012658924542,
+      "grad_norm": 0.8099361658096313,
+      "learning_rate": 9.84635099119018e-06,
+      "loss": 0.8724,
+      "step": 2967
+    },
+    {
+      "epoch": 0.16335516539160108,
+      "grad_norm": 0.7913978099822998,
+      "learning_rate": 9.846244340822456e-06,
+      "loss": 0.7106,
+      "step": 2968
+    },
+    {
+      "epoch": 0.16341020419395674,
+      "grad_norm": 0.8308563828468323,
+      "learning_rate": 9.846137654031655e-06,
+      "loss": 0.7631,
+      "step": 2969
+    },
+    {
+      "epoch": 0.1634652429963124,
+      "grad_norm": 0.8634191751480103,
+      "learning_rate": 9.846030930818582e-06,
+      "loss": 0.7363,
+      "step": 2970
+    },
+    {
+      "epoch": 0.16352028179866807,
+      "grad_norm": 0.8936432600021362,
+      "learning_rate": 9.845924171184038e-06,
+      "loss": 0.8714,
+      "step": 2971
+    },
+    {
+      "epoch": 0.16357532060102373,
+      "grad_norm": 0.8776300549507141,
+      "learning_rate": 9.845817375128825e-06,
+      "loss": 0.914,
+      "step": 2972
+    },
+    {
+      "epoch": 0.1636303594033794,
+      "grad_norm": 0.8793039321899414,
+      "learning_rate": 9.845710542653745e-06,
+      "loss": 0.7999,
+      "step": 2973
+    },
+    {
+      "epoch": 0.16368539820573505,
+      "grad_norm": 0.8391743302345276,
+      "learning_rate": 9.845603673759603e-06,
+      "loss": 0.8124,
+      "step": 2974
+    },
+    {
+      "epoch": 0.1637404370080907,
+      "grad_norm": 0.8487186431884766,
+      "learning_rate": 9.845496768447199e-06,
+      "loss": 0.8551,
+      "step": 2975
+    },
+    {
+      "epoch": 0.16379547581044637,
+      "grad_norm": 0.7780638933181763,
+      "learning_rate": 9.845389826717339e-06,
+      "loss": 0.7281,
+      "step": 2976
+    },
+    {
+      "epoch": 0.16385051461280203,
+      "grad_norm": 0.7209637761116028,
+      "learning_rate": 9.845282848570825e-06,
+      "loss": 0.6737,
+      "step": 2977
+    },
+    {
+      "epoch": 0.1639055534151577,
+      "grad_norm": 0.8414756059646606,
+      "learning_rate": 9.845175834008464e-06,
+      "loss": 0.8003,
+      "step": 2978
+    },
+    {
+      "epoch": 0.16396059221751336,
+      "grad_norm": 1.2730679512023926,
+      "learning_rate": 9.845068783031057e-06,
+      "loss": 0.8243,
+      "step": 2979
+    },
+    {
+      "epoch": 0.16401563101986902,
+      "grad_norm": 0.8573475480079651,
+      "learning_rate": 9.844961695639413e-06,
+      "loss": 0.7844,
+      "step": 2980
+    },
+    {
+      "epoch": 0.16407066982222468,
+      "grad_norm": 0.8029958605766296,
+      "learning_rate": 9.84485457183433e-06,
+      "loss": 0.7722,
+      "step": 2981
+    },
+    {
+      "epoch": 0.16412570862458034,
+      "grad_norm": 0.7839805483818054,
+      "learning_rate": 9.844747411616619e-06,
+      "loss": 0.8146,
+      "step": 2982
+    },
+    {
+      "epoch": 0.164180747426936,
+      "grad_norm": 0.7563499212265015,
+      "learning_rate": 9.844640214987082e-06,
+      "loss": 0.6909,
+      "step": 2983
+    },
+    {
+      "epoch": 0.16423578622929166,
+      "grad_norm": 0.7199193239212036,
+      "learning_rate": 9.844532981946527e-06,
+      "loss": 0.702,
+      "step": 2984
+    },
+    {
+      "epoch": 0.16429082503164733,
+      "grad_norm": 0.7519383430480957,
+      "learning_rate": 9.844425712495758e-06,
+      "loss": 0.6493,
+      "step": 2985
+    },
+    {
+      "epoch": 0.16434586383400296,
+      "grad_norm": 0.7493193745613098,
+      "learning_rate": 9.844318406635584e-06,
+      "loss": 0.8318,
+      "step": 2986
+    },
+    {
+      "epoch": 0.16440090263635862,
+      "grad_norm": 0.7951106429100037,
+      "learning_rate": 9.84421106436681e-06,
+      "loss": 0.923,
+      "step": 2987
+    },
+    {
+      "epoch": 0.16445594143871428,
+      "grad_norm": 0.8350820541381836,
+      "learning_rate": 9.844103685690238e-06,
+      "loss": 0.8091,
+      "step": 2988
+    },
+    {
+      "epoch": 0.16451098024106994,
+      "grad_norm": 0.773932695388794,
+      "learning_rate": 9.843996270606683e-06,
+      "loss": 0.8016,
+      "step": 2989
+    },
+    {
+      "epoch": 0.1645660190434256,
+      "grad_norm": 0.8208432793617249,
+      "learning_rate": 9.843888819116947e-06,
+      "loss": 0.7704,
+      "step": 2990
+    },
+    {
+      "epoch": 0.16462105784578127,
+      "grad_norm": 0.8552223443984985,
+      "learning_rate": 9.84378133122184e-06,
+      "loss": 0.8519,
+      "step": 2991
+    },
+    {
+      "epoch": 0.16467609664813693,
+      "grad_norm": 0.8015661835670471,
+      "learning_rate": 9.84367380692217e-06,
+      "loss": 0.8389,
+      "step": 2992
+    },
+    {
+      "epoch": 0.1647311354504926,
+      "grad_norm": 0.7828749418258667,
+      "learning_rate": 9.843566246218743e-06,
+      "loss": 0.7385,
+      "step": 2993
+    },
+    {
+      "epoch": 0.16478617425284825,
+      "grad_norm": 0.7761647701263428,
+      "learning_rate": 9.84345864911237e-06,
+      "loss": 0.8419,
+      "step": 2994
+    },
+    {
+      "epoch": 0.1648412130552039,
+      "grad_norm": 0.8839839100837708,
+      "learning_rate": 9.843351015603857e-06,
+      "loss": 0.8069,
+      "step": 2995
+    },
+    {
+      "epoch": 0.16489625185755957,
+      "grad_norm": 0.8611735105514526,
+      "learning_rate": 9.843243345694014e-06,
+      "loss": 0.9406,
+      "step": 2996
+    },
+    {
+      "epoch": 0.16495129065991523,
+      "grad_norm": 0.9042683839797974,
+      "learning_rate": 9.84313563938365e-06,
+      "loss": 0.821,
+      "step": 2997
+    },
+    {
+      "epoch": 0.1650063294622709,
+      "grad_norm": 0.8333690762519836,
+      "learning_rate": 9.843027896673577e-06,
+      "loss": 0.781,
+      "step": 2998
+    },
+    {
+      "epoch": 0.16506136826462656,
+      "grad_norm": 0.819922924041748,
+      "learning_rate": 9.8429201175646e-06,
+      "loss": 0.869,
+      "step": 2999
+    },
+    {
+      "epoch": 0.16511640706698222,
+      "grad_norm": 0.8349948525428772,
+      "learning_rate": 9.842812302057534e-06,
+      "loss": 0.9271,
+      "step": 3000
+    },
+    {
+      "epoch": 0.16517144586933788,
+      "grad_norm": 0.8981684446334839,
+      "learning_rate": 9.842704450153187e-06,
+      "loss": 0.7384,
+      "step": 3001
+    },
+    {
+      "epoch": 0.16522648467169354,
+      "grad_norm": 0.839133083820343,
+      "learning_rate": 9.842596561852369e-06,
+      "loss": 0.9016,
+      "step": 3002
+    },
+    {
+      "epoch": 0.1652815234740492,
+      "grad_norm": 0.8303349614143372,
+      "learning_rate": 9.842488637155891e-06,
+      "loss": 0.7488,
+      "step": 3003
+    },
+    {
+      "epoch": 0.16533656227640486,
+      "grad_norm": 0.8748323917388916,
+      "learning_rate": 9.842380676064566e-06,
+      "loss": 0.8163,
+      "step": 3004
+    },
+    {
+      "epoch": 0.16539160107876053,
+      "grad_norm": 0.782844603061676,
+      "learning_rate": 9.842272678579203e-06,
+      "loss": 0.8465,
+      "step": 3005
+    },
+    {
+      "epoch": 0.1654466398811162,
+      "grad_norm": 0.8068844676017761,
+      "learning_rate": 9.842164644700615e-06,
+      "loss": 0.8856,
+      "step": 3006
+    },
+    {
+      "epoch": 0.16550167868347185,
+      "grad_norm": 0.9174006581306458,
+      "learning_rate": 9.842056574429615e-06,
+      "loss": 0.7748,
+      "step": 3007
+    },
+    {
+      "epoch": 0.1655567174858275,
+      "grad_norm": 0.7453809380531311,
+      "learning_rate": 9.841948467767012e-06,
+      "loss": 0.7565,
+      "step": 3008
+    },
+    {
+      "epoch": 0.16561175628818317,
+      "grad_norm": 0.8408182859420776,
+      "learning_rate": 9.841840324713622e-06,
+      "loss": 0.7345,
+      "step": 3009
+    },
+    {
+      "epoch": 0.16566679509053883,
+      "grad_norm": 0.8599638938903809,
+      "learning_rate": 9.841732145270254e-06,
+      "loss": 0.8163,
+      "step": 3010
+    },
+    {
+      "epoch": 0.1657218338928945,
+      "grad_norm": 0.877616822719574,
+      "learning_rate": 9.841623929437725e-06,
+      "loss": 0.8685,
+      "step": 3011
+    },
+    {
+      "epoch": 0.16577687269525015,
+      "grad_norm": 0.7765643000602722,
+      "learning_rate": 9.841515677216846e-06,
+      "loss": 0.7281,
+      "step": 3012
+    },
+    {
+      "epoch": 0.16583191149760582,
+      "grad_norm": 0.7891712784767151,
+      "learning_rate": 9.841407388608431e-06,
+      "loss": 0.8618,
+      "step": 3013
+    },
+    {
+      "epoch": 0.16588695029996148,
+      "grad_norm": 0.9215571284294128,
+      "learning_rate": 9.841299063613295e-06,
+      "loss": 0.8709,
+      "step": 3014
+    },
+    {
+      "epoch": 0.16594198910231714,
+      "grad_norm": 0.8428288698196411,
+      "learning_rate": 9.841190702232249e-06,
+      "loss": 0.8227,
+      "step": 3015
+    },
+    {
+      "epoch": 0.1659970279046728,
+      "grad_norm": 0.9294042587280273,
+      "learning_rate": 9.841082304466112e-06,
+      "loss": 0.8203,
+      "step": 3016
+    },
+    {
+      "epoch": 0.16605206670702846,
+      "grad_norm": 0.7530880570411682,
+      "learning_rate": 9.840973870315695e-06,
+      "loss": 0.7681,
+      "step": 3017
+    },
+    {
+      "epoch": 0.16610710550938412,
+      "grad_norm": 1.0149626731872559,
+      "learning_rate": 9.840865399781814e-06,
+      "loss": 0.9212,
+      "step": 3018
+    },
+    {
+      "epoch": 0.16616214431173978,
+      "grad_norm": 0.7595353722572327,
+      "learning_rate": 9.840756892865285e-06,
+      "loss": 0.795,
+      "step": 3019
+    },
+    {
+      "epoch": 0.16621718311409545,
+      "grad_norm": 0.7893253564834595,
+      "learning_rate": 9.840648349566924e-06,
+      "loss": 0.8147,
+      "step": 3020
+    },
+    {
+      "epoch": 0.1662722219164511,
+      "grad_norm": 0.8190789222717285,
+      "learning_rate": 9.840539769887543e-06,
+      "loss": 0.7233,
+      "step": 3021
+    },
+    {
+      "epoch": 0.16632726071880677,
+      "grad_norm": 0.7771229147911072,
+      "learning_rate": 9.840431153827963e-06,
+      "loss": 0.7172,
+      "step": 3022
+    },
+    {
+      "epoch": 0.16638229952116243,
+      "grad_norm": 0.7379328012466431,
+      "learning_rate": 9.840322501388998e-06,
+      "loss": 0.7603,
+      "step": 3023
+    },
+    {
+      "epoch": 0.1664373383235181,
+      "grad_norm": 0.9488499760627747,
+      "learning_rate": 9.840213812571464e-06,
+      "loss": 0.8025,
+      "step": 3024
+    },
+    {
+      "epoch": 0.16649237712587375,
+      "grad_norm": 0.7135865092277527,
+      "learning_rate": 9.84010508737618e-06,
+      "loss": 0.7412,
+      "step": 3025
+    },
+    {
+      "epoch": 0.16654741592822941,
+      "grad_norm": 1.6780112981796265,
+      "learning_rate": 9.83999632580396e-06,
+      "loss": 0.9231,
+      "step": 3026
+    },
+    {
+      "epoch": 0.16660245473058508,
+      "grad_norm": 0.8815935850143433,
+      "learning_rate": 9.839887527855623e-06,
+      "loss": 0.7903,
+      "step": 3027
+    },
+    {
+      "epoch": 0.16665749353294074,
+      "grad_norm": 0.8942261338233948,
+      "learning_rate": 9.83977869353199e-06,
+      "loss": 0.8328,
+      "step": 3028
+    },
+    {
+      "epoch": 0.16671253233529637,
+      "grad_norm": 0.7866815328598022,
+      "learning_rate": 9.839669822833873e-06,
+      "loss": 0.8483,
+      "step": 3029
+    },
+    {
+      "epoch": 0.16676757113765203,
+      "grad_norm": 0.8133070468902588,
+      "learning_rate": 9.839560915762094e-06,
+      "loss": 0.8665,
+      "step": 3030
+    },
+    {
+      "epoch": 0.1668226099400077,
+      "grad_norm": 0.7442927360534668,
+      "learning_rate": 9.839451972317469e-06,
+      "loss": 0.6296,
+      "step": 3031
+    },
+    {
+      "epoch": 0.16687764874236335,
+      "grad_norm": 0.7505021691322327,
+      "learning_rate": 9.83934299250082e-06,
+      "loss": 0.7976,
+      "step": 3032
+    },
+    {
+      "epoch": 0.16693268754471902,
+      "grad_norm": 0.8310422897338867,
+      "learning_rate": 9.839233976312964e-06,
+      "loss": 0.9022,
+      "step": 3033
+    },
+    {
+      "epoch": 0.16698772634707468,
+      "grad_norm": 0.8175413012504578,
+      "learning_rate": 9.839124923754721e-06,
+      "loss": 0.8653,
+      "step": 3034
+    },
+    {
+      "epoch": 0.16704276514943034,
+      "grad_norm": 0.7963089346885681,
+      "learning_rate": 9.839015834826912e-06,
+      "loss": 0.7888,
+      "step": 3035
+    },
+    {
+      "epoch": 0.167097803951786,
+      "grad_norm": 0.8923391699790955,
+      "learning_rate": 9.838906709530353e-06,
+      "loss": 0.9396,
+      "step": 3036
+    },
+    {
+      "epoch": 0.16715284275414166,
+      "grad_norm": 0.7851678133010864,
+      "learning_rate": 9.838797547865869e-06,
+      "loss": 0.8163,
+      "step": 3037
+    },
+    {
+      "epoch": 0.16720788155649732,
+      "grad_norm": 0.817877471446991,
+      "learning_rate": 9.838688349834275e-06,
+      "loss": 0.8928,
+      "step": 3038
+    },
+    {
+      "epoch": 0.16726292035885298,
+      "grad_norm": 0.7603926062583923,
+      "learning_rate": 9.838579115436395e-06,
+      "loss": 0.7418,
+      "step": 3039
+    },
+    {
+      "epoch": 0.16731795916120865,
+      "grad_norm": 0.8086446523666382,
+      "learning_rate": 9.83846984467305e-06,
+      "loss": 0.8017,
+      "step": 3040
+    },
+    {
+      "epoch": 0.1673729979635643,
+      "grad_norm": 1.4745439291000366,
+      "learning_rate": 9.838360537545061e-06,
+      "loss": 0.7964,
+      "step": 3041
+    },
+    {
+      "epoch": 0.16742803676591997,
+      "grad_norm": 0.778404176235199,
+      "learning_rate": 9.83825119405325e-06,
+      "loss": 0.7395,
+      "step": 3042
+    },
+    {
+      "epoch": 0.16748307556827563,
+      "grad_norm": 0.8245886564254761,
+      "learning_rate": 9.838141814198439e-06,
+      "loss": 0.8697,
+      "step": 3043
+    },
+    {
+      "epoch": 0.1675381143706313,
+      "grad_norm": 0.8395472764968872,
+      "learning_rate": 9.838032397981448e-06,
+      "loss": 0.8545,
+      "step": 3044
+    },
+    {
+      "epoch": 0.16759315317298695,
+      "grad_norm": 0.8973744511604309,
+      "learning_rate": 9.8379229454031e-06,
+      "loss": 0.8999,
+      "step": 3045
+    },
+    {
+      "epoch": 0.16764819197534261,
+      "grad_norm": 1.2034368515014648,
+      "learning_rate": 9.837813456464219e-06,
+      "loss": 0.9039,
+      "step": 3046
+    },
+    {
+      "epoch": 0.16770323077769828,
+      "grad_norm": 0.862167477607727,
+      "learning_rate": 9.837703931165625e-06,
+      "loss": 0.889,
+      "step": 3047
+    },
+    {
+      "epoch": 0.16775826958005394,
+      "grad_norm": 0.7624714970588684,
+      "learning_rate": 9.837594369508146e-06,
+      "loss": 0.7072,
+      "step": 3048
+    },
+    {
+      "epoch": 0.1678133083824096,
+      "grad_norm": 0.7771621346473694,
+      "learning_rate": 9.8374847714926e-06,
+      "loss": 0.8769,
+      "step": 3049
+    },
+    {
+      "epoch": 0.16786834718476526,
+      "grad_norm": 0.7834492921829224,
+      "learning_rate": 9.837375137119816e-06,
+      "loss": 0.841,
+      "step": 3050
+    },
+    {
+      "epoch": 0.16792338598712092,
+      "grad_norm": 0.8175067901611328,
+      "learning_rate": 9.837265466390612e-06,
+      "loss": 0.8149,
+      "step": 3051
+    },
+    {
+      "epoch": 0.16797842478947658,
+      "grad_norm": 0.7474493384361267,
+      "learning_rate": 9.83715575930582e-06,
+      "loss": 0.7716,
+      "step": 3052
+    },
+    {
+      "epoch": 0.16803346359183224,
+      "grad_norm": 1.1263303756713867,
+      "learning_rate": 9.837046015866257e-06,
+      "loss": 0.8026,
+      "step": 3053
+    },
+    {
+      "epoch": 0.1680885023941879,
+      "grad_norm": 0.8741740584373474,
+      "learning_rate": 9.836936236072752e-06,
+      "loss": 0.8795,
+      "step": 3054
+    },
+    {
+      "epoch": 0.16814354119654357,
+      "grad_norm": 0.8108506798744202,
+      "learning_rate": 9.83682641992613e-06,
+      "loss": 0.7682,
+      "step": 3055
+    },
+    {
+      "epoch": 0.16819857999889923,
+      "grad_norm": 0.9380543231964111,
+      "learning_rate": 9.836716567427213e-06,
+      "loss": 0.8739,
+      "step": 3056
+    },
+    {
+      "epoch": 0.1682536188012549,
+      "grad_norm": 0.7755940556526184,
+      "learning_rate": 9.83660667857683e-06,
+      "loss": 0.7287,
+      "step": 3057
+    },
+    {
+      "epoch": 0.16830865760361055,
+      "grad_norm": 0.808907151222229,
+      "learning_rate": 9.836496753375807e-06,
+      "loss": 0.7988,
+      "step": 3058
+    },
+    {
+      "epoch": 0.1683636964059662,
+      "grad_norm": 1.1496524810791016,
+      "learning_rate": 9.836386791824967e-06,
+      "loss": 0.8621,
+      "step": 3059
+    },
+    {
+      "epoch": 0.16841873520832187,
+      "grad_norm": 0.8550384640693665,
+      "learning_rate": 9.83627679392514e-06,
+      "loss": 0.913,
+      "step": 3060
+    },
+    {
+      "epoch": 0.16847377401067753,
+      "grad_norm": 0.761142909526825,
+      "learning_rate": 9.83616675967715e-06,
+      "loss": 0.7271,
+      "step": 3061
+    },
+    {
+      "epoch": 0.1685288128130332,
+      "grad_norm": 0.8496200442314148,
+      "learning_rate": 9.836056689081828e-06,
+      "loss": 0.7885,
+      "step": 3062
+    },
+    {
+      "epoch": 0.16858385161538886,
+      "grad_norm": 0.8310382962226868,
+      "learning_rate": 9.835946582139996e-06,
+      "loss": 0.858,
+      "step": 3063
+    },
+    {
+      "epoch": 0.16863889041774452,
+      "grad_norm": 0.7870821952819824,
+      "learning_rate": 9.835836438852485e-06,
+      "loss": 0.7791,
+      "step": 3064
+    },
+    {
+      "epoch": 0.16869392922010018,
+      "grad_norm": 0.7170534729957581,
+      "learning_rate": 9.83572625922012e-06,
+      "loss": 0.6666,
+      "step": 3065
+    },
+    {
+      "epoch": 0.16874896802245584,
+      "grad_norm": 0.9764187932014465,
+      "learning_rate": 9.835616043243732e-06,
+      "loss": 0.8341,
+      "step": 3066
+    },
+    {
+      "epoch": 0.1688040068248115,
+      "grad_norm": 0.7453315258026123,
+      "learning_rate": 9.83550579092415e-06,
+      "loss": 0.7032,
+      "step": 3067
+    },
+    {
+      "epoch": 0.16885904562716716,
+      "grad_norm": 0.9205759763717651,
+      "learning_rate": 9.835395502262196e-06,
+      "loss": 0.813,
+      "step": 3068
+    },
+    {
+      "epoch": 0.16891408442952283,
+      "grad_norm": 0.8152205944061279,
+      "learning_rate": 9.835285177258708e-06,
+      "loss": 0.8275,
+      "step": 3069
+    },
+    {
+      "epoch": 0.1689691232318785,
+      "grad_norm": 0.8065707087516785,
+      "learning_rate": 9.83517481591451e-06,
+      "loss": 0.8648,
+      "step": 3070
+    },
+    {
+      "epoch": 0.16902416203423415,
+      "grad_norm": 0.7774410247802734,
+      "learning_rate": 9.835064418230432e-06,
+      "loss": 0.7818,
+      "step": 3071
+    },
+    {
+      "epoch": 0.16907920083658978,
+      "grad_norm": 0.8591069579124451,
+      "learning_rate": 9.834953984207305e-06,
+      "loss": 0.8055,
+      "step": 3072
+    },
+    {
+      "epoch": 0.16913423963894544,
+      "grad_norm": 0.7421612739562988,
+      "learning_rate": 9.834843513845958e-06,
+      "loss": 0.7543,
+      "step": 3073
+    },
+    {
+      "epoch": 0.1691892784413011,
+      "grad_norm": 0.7855183482170105,
+      "learning_rate": 9.83473300714722e-06,
+      "loss": 0.7011,
+      "step": 3074
+    },
+    {
+      "epoch": 0.16924431724365677,
+      "grad_norm": 0.8061636686325073,
+      "learning_rate": 9.834622464111924e-06,
+      "loss": 0.8096,
+      "step": 3075
+    },
+    {
+      "epoch": 0.16929935604601243,
+      "grad_norm": 0.8048406839370728,
+      "learning_rate": 9.834511884740898e-06,
+      "loss": 0.8166,
+      "step": 3076
+    },
+    {
+      "epoch": 0.1693543948483681,
+      "grad_norm": 0.8776549696922302,
+      "learning_rate": 9.834401269034977e-06,
+      "loss": 0.8169,
+      "step": 3077
+    },
+    {
+      "epoch": 0.16940943365072375,
+      "grad_norm": 1.0208356380462646,
+      "learning_rate": 9.83429061699499e-06,
+      "loss": 0.6976,
+      "step": 3078
+    },
+    {
+      "epoch": 0.1694644724530794,
+      "grad_norm": 0.7641016840934753,
+      "learning_rate": 9.834179928621767e-06,
+      "loss": 0.7109,
+      "step": 3079
+    },
+    {
+      "epoch": 0.16951951125543507,
+      "grad_norm": 0.7648905515670776,
+      "learning_rate": 9.834069203916143e-06,
+      "loss": 0.7927,
+      "step": 3080
+    },
+    {
+      "epoch": 0.16957455005779073,
+      "grad_norm": 0.7898744344711304,
+      "learning_rate": 9.833958442878948e-06,
+      "loss": 0.7911,
+      "step": 3081
+    },
+    {
+      "epoch": 0.1696295888601464,
+      "grad_norm": 0.8812462687492371,
+      "learning_rate": 9.833847645511016e-06,
+      "loss": 0.8381,
+      "step": 3082
+    },
+    {
+      "epoch": 0.16968462766250206,
+      "grad_norm": 0.8141197562217712,
+      "learning_rate": 9.833736811813179e-06,
+      "loss": 0.7422,
+      "step": 3083
+    },
+    {
+      "epoch": 0.16973966646485772,
+      "grad_norm": 0.7860949635505676,
+      "learning_rate": 9.83362594178627e-06,
+      "loss": 0.7568,
+      "step": 3084
+    },
+    {
+      "epoch": 0.16979470526721338,
+      "grad_norm": 0.6688396334648132,
+      "learning_rate": 9.833515035431123e-06,
+      "loss": 0.7143,
+      "step": 3085
+    },
+    {
+      "epoch": 0.16984974406956904,
+      "grad_norm": 0.7525103092193604,
+      "learning_rate": 9.833404092748569e-06,
+      "loss": 0.8026,
+      "step": 3086
+    },
+    {
+      "epoch": 0.1699047828719247,
+      "grad_norm": 0.8505181670188904,
+      "learning_rate": 9.833293113739444e-06,
+      "loss": 0.8894,
+      "step": 3087
+    },
+    {
+      "epoch": 0.16995982167428036,
+      "grad_norm": 0.8432300090789795,
+      "learning_rate": 9.833182098404583e-06,
+      "loss": 0.7801,
+      "step": 3088
+    },
+    {
+      "epoch": 0.17001486047663603,
+      "grad_norm": 0.7655903100967407,
+      "learning_rate": 9.833071046744819e-06,
+      "loss": 0.7838,
+      "step": 3089
+    },
+    {
+      "epoch": 0.1700698992789917,
+      "grad_norm": 0.8436369895935059,
+      "learning_rate": 9.832959958760986e-06,
+      "loss": 0.8636,
+      "step": 3090
+    },
+    {
+      "epoch": 0.17012493808134735,
+      "grad_norm": 0.7880234122276306,
+      "learning_rate": 9.83284883445392e-06,
+      "loss": 0.7701,
+      "step": 3091
+    },
+    {
+      "epoch": 0.170179976883703,
+      "grad_norm": 0.7713757753372192,
+      "learning_rate": 9.832737673824455e-06,
+      "loss": 0.8652,
+      "step": 3092
+    },
+    {
+      "epoch": 0.17023501568605867,
+      "grad_norm": 0.7905295491218567,
+      "learning_rate": 9.832626476873428e-06,
+      "loss": 0.8666,
+      "step": 3093
+    },
+    {
+      "epoch": 0.17029005448841433,
+      "grad_norm": 0.7589883804321289,
+      "learning_rate": 9.832515243601675e-06,
+      "loss": 0.8051,
+      "step": 3094
+    },
+    {
+      "epoch": 0.17034509329077,
+      "grad_norm": 0.9068838953971863,
+      "learning_rate": 9.83240397401003e-06,
+      "loss": 0.9037,
+      "step": 3095
+    },
+    {
+      "epoch": 0.17040013209312566,
+      "grad_norm": 0.7465278506278992,
+      "learning_rate": 9.83229266809933e-06,
+      "loss": 0.7425,
+      "step": 3096
+    },
+    {
+      "epoch": 0.17045517089548132,
+      "grad_norm": 0.8111177086830139,
+      "learning_rate": 9.83218132587041e-06,
+      "loss": 0.8034,
+      "step": 3097
+    },
+    {
+      "epoch": 0.17051020969783698,
+      "grad_norm": 1.1007672548294067,
+      "learning_rate": 9.832069947324112e-06,
+      "loss": 0.9139,
+      "step": 3098
+    },
+    {
+      "epoch": 0.17056524850019264,
+      "grad_norm": 0.881179690361023,
+      "learning_rate": 9.831958532461269e-06,
+      "loss": 0.9062,
+      "step": 3099
+    },
+    {
+      "epoch": 0.1706202873025483,
+      "grad_norm": 0.8012413382530212,
+      "learning_rate": 9.831847081282718e-06,
+      "loss": 0.7956,
+      "step": 3100
+    },
+    {
+      "epoch": 0.17067532610490396,
+      "grad_norm": 0.741731584072113,
+      "learning_rate": 9.831735593789298e-06,
+      "loss": 0.8754,
+      "step": 3101
+    },
+    {
+      "epoch": 0.17073036490725962,
+      "grad_norm": 0.8945604562759399,
+      "learning_rate": 9.831624069981848e-06,
+      "loss": 0.8293,
+      "step": 3102
+    },
+    {
+      "epoch": 0.17078540370961529,
+      "grad_norm": 0.7865545749664307,
+      "learning_rate": 9.831512509861203e-06,
+      "loss": 0.7812,
+      "step": 3103
+    },
+    {
+      "epoch": 0.17084044251197095,
+      "grad_norm": 0.832847535610199,
+      "learning_rate": 9.831400913428205e-06,
+      "loss": 0.8925,
+      "step": 3104
+    },
+    {
+      "epoch": 0.1708954813143266,
+      "grad_norm": 0.7374216914176941,
+      "learning_rate": 9.83128928068369e-06,
+      "loss": 0.8275,
+      "step": 3105
+    },
+    {
+      "epoch": 0.17095052011668227,
+      "grad_norm": 0.748725414276123,
+      "learning_rate": 9.831177611628497e-06,
+      "loss": 0.8364,
+      "step": 3106
+    },
+    {
+      "epoch": 0.17100555891903793,
+      "grad_norm": 0.810276448726654,
+      "learning_rate": 9.831065906263468e-06,
+      "loss": 0.861,
+      "step": 3107
+    },
+    {
+      "epoch": 0.1710605977213936,
+      "grad_norm": 0.7607758641242981,
+      "learning_rate": 9.83095416458944e-06,
+      "loss": 0.7989,
+      "step": 3108
+    },
+    {
+      "epoch": 0.17111563652374925,
+      "grad_norm": 0.7206127047538757,
+      "learning_rate": 9.830842386607253e-06,
+      "loss": 0.7187,
+      "step": 3109
+    },
+    {
+      "epoch": 0.17117067532610492,
+      "grad_norm": 0.7775895595550537,
+      "learning_rate": 9.83073057231775e-06,
+      "loss": 0.8008,
+      "step": 3110
+    },
+    {
+      "epoch": 0.17122571412846058,
+      "grad_norm": 0.8351094722747803,
+      "learning_rate": 9.830618721721768e-06,
+      "loss": 0.8025,
+      "step": 3111
+    },
+    {
+      "epoch": 0.17128075293081624,
+      "grad_norm": 0.8090646266937256,
+      "learning_rate": 9.830506834820148e-06,
+      "loss": 0.8012,
+      "step": 3112
+    },
+    {
+      "epoch": 0.1713357917331719,
+      "grad_norm": 0.7762801051139832,
+      "learning_rate": 9.830394911613733e-06,
+      "loss": 0.8428,
+      "step": 3113
+    },
+    {
+      "epoch": 0.17139083053552756,
+      "grad_norm": 0.8117541074752808,
+      "learning_rate": 9.83028295210336e-06,
+      "loss": 0.8566,
+      "step": 3114
+    },
+    {
+      "epoch": 0.1714458693378832,
+      "grad_norm": 0.8786184787750244,
+      "learning_rate": 9.830170956289876e-06,
+      "loss": 0.8386,
+      "step": 3115
+    },
+    {
+      "epoch": 0.17150090814023886,
+      "grad_norm": 1.0181046724319458,
+      "learning_rate": 9.83005892417412e-06,
+      "loss": 0.8555,
+      "step": 3116
+    },
+    {
+      "epoch": 0.17155594694259452,
+      "grad_norm": 0.8236173391342163,
+      "learning_rate": 9.829946855756934e-06,
+      "loss": 0.7933,
+      "step": 3117
+    },
+    {
+      "epoch": 0.17161098574495018,
+      "grad_norm": 0.8058149814605713,
+      "learning_rate": 9.829834751039157e-06,
+      "loss": 0.842,
+      "step": 3118
+    },
+    {
+      "epoch": 0.17166602454730584,
+      "grad_norm": 0.7419908046722412,
+      "learning_rate": 9.82972261002164e-06,
+      "loss": 0.8397,
+      "step": 3119
+    },
+    {
+      "epoch": 0.1717210633496615,
+      "grad_norm": 0.7528164982795715,
+      "learning_rate": 9.829610432705216e-06,
+      "loss": 0.7931,
+      "step": 3120
+    },
+    {
+      "epoch": 0.17177610215201716,
+      "grad_norm": 0.7357296943664551,
+      "learning_rate": 9.829498219090736e-06,
+      "loss": 0.8089,
+      "step": 3121
+    },
+    {
+      "epoch": 0.17183114095437282,
+      "grad_norm": 0.7635773420333862,
+      "learning_rate": 9.829385969179039e-06,
+      "loss": 0.7442,
+      "step": 3122
+    },
+    {
+      "epoch": 0.17188617975672849,
+      "grad_norm": 0.8200171589851379,
+      "learning_rate": 9.82927368297097e-06,
+      "loss": 0.757,
+      "step": 3123
+    },
+    {
+      "epoch": 0.17194121855908415,
+      "grad_norm": 0.8367171287536621,
+      "learning_rate": 9.829161360467374e-06,
+      "loss": 0.915,
+      "step": 3124
+    },
+    {
+      "epoch": 0.1719962573614398,
+      "grad_norm": 0.8460778594017029,
+      "learning_rate": 9.829049001669091e-06,
+      "loss": 0.8568,
+      "step": 3125
+    },
+    {
+      "epoch": 0.17205129616379547,
+      "grad_norm": 0.7301799058914185,
+      "learning_rate": 9.82893660657697e-06,
+      "loss": 0.8041,
+      "step": 3126
+    },
+    {
+      "epoch": 0.17210633496615113,
+      "grad_norm": 0.7858132123947144,
+      "learning_rate": 9.828824175191854e-06,
+      "loss": 0.8367,
+      "step": 3127
+    },
+    {
+      "epoch": 0.1721613737685068,
+      "grad_norm": 0.8118360042572021,
+      "learning_rate": 9.82871170751459e-06,
+      "loss": 0.85,
+      "step": 3128
+    },
+    {
+      "epoch": 0.17221641257086245,
+      "grad_norm": 0.9020261764526367,
+      "learning_rate": 9.828599203546019e-06,
+      "loss": 0.789,
+      "step": 3129
+    },
+    {
+      "epoch": 0.17227145137321812,
+      "grad_norm": 0.8194546699523926,
+      "learning_rate": 9.828486663286989e-06,
+      "loss": 0.8644,
+      "step": 3130
+    },
+    {
+      "epoch": 0.17232649017557378,
+      "grad_norm": 0.7764905095100403,
+      "learning_rate": 9.828374086738345e-06,
+      "loss": 0.7961,
+      "step": 3131
+    },
+    {
+      "epoch": 0.17238152897792944,
+      "grad_norm": 0.7712632417678833,
+      "learning_rate": 9.828261473900935e-06,
+      "loss": 0.8082,
+      "step": 3132
+    },
+    {
+      "epoch": 0.1724365677802851,
+      "grad_norm": 0.7100280523300171,
+      "learning_rate": 9.828148824775604e-06,
+      "loss": 0.7514,
+      "step": 3133
+    },
+    {
+      "epoch": 0.17249160658264076,
+      "grad_norm": 0.7812890410423279,
+      "learning_rate": 9.8280361393632e-06,
+      "loss": 0.7125,
+      "step": 3134
+    },
+    {
+      "epoch": 0.17254664538499642,
+      "grad_norm": 0.8772642612457275,
+      "learning_rate": 9.827923417664568e-06,
+      "loss": 0.8355,
+      "step": 3135
+    },
+    {
+      "epoch": 0.17260168418735208,
+      "grad_norm": 0.9161205291748047,
+      "learning_rate": 9.827810659680555e-06,
+      "loss": 0.7511,
+      "step": 3136
+    },
+    {
+      "epoch": 0.17265672298970774,
+      "grad_norm": 0.7628560662269592,
+      "learning_rate": 9.82769786541201e-06,
+      "loss": 0.882,
+      "step": 3137
+    },
+    {
+      "epoch": 0.1727117617920634,
+      "grad_norm": 0.8203405737876892,
+      "learning_rate": 9.827585034859781e-06,
+      "loss": 0.8172,
+      "step": 3138
+    },
+    {
+      "epoch": 0.17276680059441907,
+      "grad_norm": 0.8318095207214355,
+      "learning_rate": 9.827472168024715e-06,
+      "loss": 0.7784,
+      "step": 3139
+    },
+    {
+      "epoch": 0.17282183939677473,
+      "grad_norm": 0.9137747287750244,
+      "learning_rate": 9.827359264907658e-06,
+      "loss": 0.8643,
+      "step": 3140
+    },
+    {
+      "epoch": 0.1728768781991304,
+      "grad_norm": 0.9441068768501282,
+      "learning_rate": 9.827246325509463e-06,
+      "loss": 0.7936,
+      "step": 3141
+    },
+    {
+      "epoch": 0.17293191700148605,
+      "grad_norm": 0.7402390837669373,
+      "learning_rate": 9.827133349830977e-06,
+      "loss": 0.7813,
+      "step": 3142
+    },
+    {
+      "epoch": 0.1729869558038417,
+      "grad_norm": 0.8328836560249329,
+      "learning_rate": 9.827020337873048e-06,
+      "loss": 0.7676,
+      "step": 3143
+    },
+    {
+      "epoch": 0.17304199460619737,
+      "grad_norm": 0.8106881380081177,
+      "learning_rate": 9.826907289636526e-06,
+      "loss": 0.9037,
+      "step": 3144
+    },
+    {
+      "epoch": 0.17309703340855304,
+      "grad_norm": 0.8457425236701965,
+      "learning_rate": 9.826794205122263e-06,
+      "loss": 0.78,
+      "step": 3145
+    },
+    {
+      "epoch": 0.1731520722109087,
+      "grad_norm": 0.9335517883300781,
+      "learning_rate": 9.826681084331105e-06,
+      "loss": 0.9197,
+      "step": 3146
+    },
+    {
+      "epoch": 0.17320711101326436,
+      "grad_norm": 0.9098715782165527,
+      "learning_rate": 9.826567927263904e-06,
+      "loss": 0.932,
+      "step": 3147
+    },
+    {
+      "epoch": 0.17326214981562002,
+      "grad_norm": 0.767234206199646,
+      "learning_rate": 9.826454733921512e-06,
+      "loss": 0.8717,
+      "step": 3148
+    },
+    {
+      "epoch": 0.17331718861797568,
+      "grad_norm": 0.8114444017410278,
+      "learning_rate": 9.826341504304775e-06,
+      "loss": 0.8744,
+      "step": 3149
+    },
+    {
+      "epoch": 0.17337222742033134,
+      "grad_norm": 0.7948976755142212,
+      "learning_rate": 9.82622823841455e-06,
+      "loss": 0.7947,
+      "step": 3150
+    },
+    {
+      "epoch": 0.173427266222687,
+      "grad_norm": 0.7808204889297485,
+      "learning_rate": 9.826114936251684e-06,
+      "loss": 0.8151,
+      "step": 3151
+    },
+    {
+      "epoch": 0.17348230502504267,
+      "grad_norm": 0.733860969543457,
+      "learning_rate": 9.82600159781703e-06,
+      "loss": 0.8018,
+      "step": 3152
+    },
+    {
+      "epoch": 0.17353734382739833,
+      "grad_norm": 0.7630699276924133,
+      "learning_rate": 9.825888223111442e-06,
+      "loss": 0.7937,
+      "step": 3153
+    },
+    {
+      "epoch": 0.173592382629754,
+      "grad_norm": 0.7892931699752808,
+      "learning_rate": 9.825774812135766e-06,
+      "loss": 0.782,
+      "step": 3154
+    },
+    {
+      "epoch": 0.17364742143210965,
+      "grad_norm": 0.6642436385154724,
+      "learning_rate": 9.825661364890862e-06,
+      "loss": 0.6611,
+      "step": 3155
+    },
+    {
+      "epoch": 0.1737024602344653,
+      "grad_norm": 0.7755968570709229,
+      "learning_rate": 9.825547881377577e-06,
+      "loss": 0.7835,
+      "step": 3156
+    },
+    {
+      "epoch": 0.17375749903682097,
+      "grad_norm": 0.8406579494476318,
+      "learning_rate": 9.825434361596766e-06,
+      "loss": 0.9178,
+      "step": 3157
+    },
+    {
+      "epoch": 0.1738125378391766,
+      "grad_norm": 0.8887308835983276,
+      "learning_rate": 9.825320805549284e-06,
+      "loss": 0.7951,
+      "step": 3158
+    },
+    {
+      "epoch": 0.17386757664153227,
+      "grad_norm": 0.85418701171875,
+      "learning_rate": 9.825207213235978e-06,
+      "loss": 0.8671,
+      "step": 3159
+    },
+    {
+      "epoch": 0.17392261544388793,
+      "grad_norm": 0.8831202983856201,
+      "learning_rate": 9.82509358465771e-06,
+      "loss": 0.8708,
+      "step": 3160
+    },
+    {
+      "epoch": 0.1739776542462436,
+      "grad_norm": 0.9041616320610046,
+      "learning_rate": 9.82497991981533e-06,
+      "loss": 0.8981,
+      "step": 3161
+    },
+    {
+      "epoch": 0.17403269304859925,
+      "grad_norm": 0.8169258832931519,
+      "learning_rate": 9.824866218709692e-06,
+      "loss": 0.8857,
+      "step": 3162
+    },
+    {
+      "epoch": 0.1740877318509549,
+      "grad_norm": 0.8714475631713867,
+      "learning_rate": 9.824752481341651e-06,
+      "loss": 0.8552,
+      "step": 3163
+    },
+    {
+      "epoch": 0.17414277065331057,
+      "grad_norm": 0.8261111378669739,
+      "learning_rate": 9.824638707712061e-06,
+      "loss": 0.808,
+      "step": 3164
+    },
+    {
+      "epoch": 0.17419780945566624,
+      "grad_norm": 0.7542527914047241,
+      "learning_rate": 9.82452489782178e-06,
+      "loss": 0.8078,
+      "step": 3165
+    },
+    {
+      "epoch": 0.1742528482580219,
+      "grad_norm": 1.309218168258667,
+      "learning_rate": 9.824411051671658e-06,
+      "loss": 0.9325,
+      "step": 3166
+    },
+    {
+      "epoch": 0.17430788706037756,
+      "grad_norm": 0.8528563380241394,
+      "learning_rate": 9.824297169262555e-06,
+      "loss": 0.8493,
+      "step": 3167
+    },
+    {
+      "epoch": 0.17436292586273322,
+      "grad_norm": 0.7777062058448792,
+      "learning_rate": 9.824183250595328e-06,
+      "loss": 0.7002,
+      "step": 3168
+    },
+    {
+      "epoch": 0.17441796466508888,
+      "grad_norm": 0.7385506629943848,
+      "learning_rate": 9.824069295670828e-06,
+      "loss": 0.8396,
+      "step": 3169
+    },
+    {
+      "epoch": 0.17447300346744454,
+      "grad_norm": 0.8316949605941772,
+      "learning_rate": 9.823955304489918e-06,
+      "loss": 0.8769,
+      "step": 3170
+    },
+    {
+      "epoch": 0.1745280422698002,
+      "grad_norm": 0.8149139285087585,
+      "learning_rate": 9.823841277053448e-06,
+      "loss": 0.8009,
+      "step": 3171
+    },
+    {
+      "epoch": 0.17458308107215587,
+      "grad_norm": 0.8761584162712097,
+      "learning_rate": 9.82372721336228e-06,
+      "loss": 0.7366,
+      "step": 3172
+    },
+    {
+      "epoch": 0.17463811987451153,
+      "grad_norm": 0.7104084491729736,
+      "learning_rate": 9.82361311341727e-06,
+      "loss": 0.6704,
+      "step": 3173
+    },
+    {
+      "epoch": 0.1746931586768672,
+      "grad_norm": 0.791806697845459,
+      "learning_rate": 9.823498977219273e-06,
+      "loss": 0.9054,
+      "step": 3174
+    },
+    {
+      "epoch": 0.17474819747922285,
+      "grad_norm": 0.7675086855888367,
+      "learning_rate": 9.82338480476915e-06,
+      "loss": 0.751,
+      "step": 3175
+    },
+    {
+      "epoch": 0.1748032362815785,
+      "grad_norm": 0.7380725145339966,
+      "learning_rate": 9.823270596067759e-06,
+      "loss": 0.7618,
+      "step": 3176
+    },
+    {
+      "epoch": 0.17485827508393417,
+      "grad_norm": 0.7311519384384155,
+      "learning_rate": 9.823156351115954e-06,
+      "loss": 0.7424,
+      "step": 3177
+    },
+    {
+      "epoch": 0.17491331388628983,
+      "grad_norm": 0.7888365387916565,
+      "learning_rate": 9.8230420699146e-06,
+      "loss": 0.7717,
+      "step": 3178
+    },
+    {
+      "epoch": 0.1749683526886455,
+      "grad_norm": 0.9329265356063843,
+      "learning_rate": 9.822927752464552e-06,
+      "loss": 0.8256,
+      "step": 3179
+    },
+    {
+      "epoch": 0.17502339149100116,
+      "grad_norm": 0.711794912815094,
+      "learning_rate": 9.822813398766671e-06,
+      "loss": 0.7373,
+      "step": 3180
+    },
+    {
+      "epoch": 0.17507843029335682,
+      "grad_norm": 0.8713497519493103,
+      "learning_rate": 9.822699008821813e-06,
+      "loss": 0.8135,
+      "step": 3181
+    },
+    {
+      "epoch": 0.17513346909571248,
+      "grad_norm": 0.6923471689224243,
+      "learning_rate": 9.822584582630841e-06,
+      "loss": 0.7589,
+      "step": 3182
+    },
+    {
+      "epoch": 0.17518850789806814,
+      "grad_norm": 0.8648017048835754,
+      "learning_rate": 9.822470120194616e-06,
+      "loss": 0.7828,
+      "step": 3183
+    },
+    {
+      "epoch": 0.1752435467004238,
+      "grad_norm": 0.8407077789306641,
+      "learning_rate": 9.822355621513994e-06,
+      "loss": 0.8537,
+      "step": 3184
+    },
+    {
+      "epoch": 0.17529858550277946,
+      "grad_norm": 0.8076738119125366,
+      "learning_rate": 9.822241086589841e-06,
+      "loss": 0.7827,
+      "step": 3185
+    },
+    {
+      "epoch": 0.17535362430513513,
+      "grad_norm": 0.8402661085128784,
+      "learning_rate": 9.822126515423011e-06,
+      "loss": 0.8247,
+      "step": 3186
+    },
+    {
+      "epoch": 0.1754086631074908,
+      "grad_norm": 0.8911813497543335,
+      "learning_rate": 9.822011908014373e-06,
+      "loss": 0.8996,
+      "step": 3187
+    },
+    {
+      "epoch": 0.17546370190984645,
+      "grad_norm": 0.8060111999511719,
+      "learning_rate": 9.821897264364782e-06,
+      "loss": 0.796,
+      "step": 3188
+    },
+    {
+      "epoch": 0.1755187407122021,
+      "grad_norm": 0.8476423621177673,
+      "learning_rate": 9.8217825844751e-06,
+      "loss": 0.8657,
+      "step": 3189
+    },
+    {
+      "epoch": 0.17557377951455777,
+      "grad_norm": 0.7614054083824158,
+      "learning_rate": 9.821667868346194e-06,
+      "loss": 0.8583,
+      "step": 3190
+    },
+    {
+      "epoch": 0.17562881831691343,
+      "grad_norm": 0.8312287330627441,
+      "learning_rate": 9.821553115978923e-06,
+      "loss": 0.7718,
+      "step": 3191
+    },
+    {
+      "epoch": 0.1756838571192691,
+      "grad_norm": 0.8199487328529358,
+      "learning_rate": 9.82143832737415e-06,
+      "loss": 0.7617,
+      "step": 3192
+    },
+    {
+      "epoch": 0.17573889592162475,
+      "grad_norm": 0.7529115080833435,
+      "learning_rate": 9.821323502532733e-06,
+      "loss": 0.7587,
+      "step": 3193
+    },
+    {
+      "epoch": 0.17579393472398042,
+      "grad_norm": 0.9205463528633118,
+      "learning_rate": 9.821208641455542e-06,
+      "loss": 0.7871,
+      "step": 3194
+    },
+    {
+      "epoch": 0.17584897352633608,
+      "grad_norm": 0.8055161833763123,
+      "learning_rate": 9.821093744143437e-06,
+      "loss": 0.8133,
+      "step": 3195
+    },
+    {
+      "epoch": 0.17590401232869174,
+      "grad_norm": 0.7322981953620911,
+      "learning_rate": 9.82097881059728e-06,
+      "loss": 0.7442,
+      "step": 3196
+    },
+    {
+      "epoch": 0.1759590511310474,
+      "grad_norm": 1.0465941429138184,
+      "learning_rate": 9.82086384081794e-06,
+      "loss": 1.0073,
+      "step": 3197
+    },
+    {
+      "epoch": 0.17601408993340306,
+      "grad_norm": 0.7607331275939941,
+      "learning_rate": 9.820748834806278e-06,
+      "loss": 0.8128,
+      "step": 3198
+    },
+    {
+      "epoch": 0.17606912873575872,
+      "grad_norm": 0.7901879549026489,
+      "learning_rate": 9.820633792563156e-06,
+      "loss": 0.7928,
+      "step": 3199
+    },
+    {
+      "epoch": 0.17612416753811436,
+      "grad_norm": 0.8010839223861694,
+      "learning_rate": 9.820518714089442e-06,
+      "loss": 0.7025,
+      "step": 3200
+    },
+    {
+      "epoch": 0.17617920634047002,
+      "grad_norm": 0.8511317372322083,
+      "learning_rate": 9.820403599385999e-06,
+      "loss": 0.7947,
+      "step": 3201
+    },
+    {
+      "epoch": 0.17623424514282568,
+      "grad_norm": 0.7978847026824951,
+      "learning_rate": 9.820288448453693e-06,
+      "loss": 0.7395,
+      "step": 3202
+    },
+    {
+      "epoch": 0.17628928394518134,
+      "grad_norm": 0.6991232633590698,
+      "learning_rate": 9.820173261293388e-06,
+      "loss": 0.7113,
+      "step": 3203
+    },
+    {
+      "epoch": 0.176344322747537,
+      "grad_norm": 0.8966444730758667,
+      "learning_rate": 9.820058037905954e-06,
+      "loss": 0.7399,
+      "step": 3204
+    },
+    {
+      "epoch": 0.17639936154989266,
+      "grad_norm": 0.8042632341384888,
+      "learning_rate": 9.819942778292253e-06,
+      "loss": 0.8183,
+      "step": 3205
+    },
+    {
+      "epoch": 0.17645440035224833,
+      "grad_norm": 0.8047537803649902,
+      "learning_rate": 9.81982748245315e-06,
+      "loss": 0.852,
+      "step": 3206
+    },
+    {
+      "epoch": 0.176509439154604,
+      "grad_norm": 0.8277122378349304,
+      "learning_rate": 9.819712150389517e-06,
+      "loss": 0.8828,
+      "step": 3207
+    },
+    {
+      "epoch": 0.17656447795695965,
+      "grad_norm": 0.8677185773849487,
+      "learning_rate": 9.819596782102216e-06,
+      "loss": 0.8416,
+      "step": 3208
+    },
+    {
+      "epoch": 0.1766195167593153,
+      "grad_norm": 0.8750975728034973,
+      "learning_rate": 9.819481377592115e-06,
+      "loss": 0.9289,
+      "step": 3209
+    },
+    {
+      "epoch": 0.17667455556167097,
+      "grad_norm": 0.7665122151374817,
+      "learning_rate": 9.819365936860084e-06,
+      "loss": 0.8653,
+      "step": 3210
+    },
+    {
+      "epoch": 0.17672959436402663,
+      "grad_norm": 0.9341353178024292,
+      "learning_rate": 9.819250459906989e-06,
+      "loss": 0.7225,
+      "step": 3211
+    },
+    {
+      "epoch": 0.1767846331663823,
+      "grad_norm": 0.7007241249084473,
+      "learning_rate": 9.819134946733696e-06,
+      "loss": 0.7429,
+      "step": 3212
+    },
+    {
+      "epoch": 0.17683967196873795,
+      "grad_norm": 0.8001461029052734,
+      "learning_rate": 9.819019397341074e-06,
+      "loss": 0.759,
+      "step": 3213
+    },
+    {
+      "epoch": 0.17689471077109362,
+      "grad_norm": 0.8936446905136108,
+      "learning_rate": 9.818903811729993e-06,
+      "loss": 0.8248,
+      "step": 3214
+    },
+    {
+      "epoch": 0.17694974957344928,
+      "grad_norm": 0.805570125579834,
+      "learning_rate": 9.818788189901321e-06,
+      "loss": 0.9214,
+      "step": 3215
+    },
+    {
+      "epoch": 0.17700478837580494,
+      "grad_norm": 0.7762455940246582,
+      "learning_rate": 9.818672531855926e-06,
+      "loss": 0.7848,
+      "step": 3216
+    },
+    {
+      "epoch": 0.1770598271781606,
+      "grad_norm": 0.8391497731208801,
+      "learning_rate": 9.81855683759468e-06,
+      "loss": 0.7543,
+      "step": 3217
+    },
+    {
+      "epoch": 0.17711486598051626,
+      "grad_norm": 0.8489046692848206,
+      "learning_rate": 9.818441107118449e-06,
+      "loss": 0.7908,
+      "step": 3218
+    },
+    {
+      "epoch": 0.17716990478287192,
+      "grad_norm": 1.0949461460113525,
+      "learning_rate": 9.818325340428105e-06,
+      "loss": 0.8255,
+      "step": 3219
+    },
+    {
+      "epoch": 0.17722494358522758,
+      "grad_norm": 0.8710842132568359,
+      "learning_rate": 9.81820953752452e-06,
+      "loss": 0.859,
+      "step": 3220
+    },
+    {
+      "epoch": 0.17727998238758325,
+      "grad_norm": 0.7936064600944519,
+      "learning_rate": 9.818093698408558e-06,
+      "loss": 0.8475,
+      "step": 3221
+    },
+    {
+      "epoch": 0.1773350211899389,
+      "grad_norm": 0.790341854095459,
+      "learning_rate": 9.817977823081095e-06,
+      "loss": 0.8137,
+      "step": 3222
+    },
+    {
+      "epoch": 0.17739005999229457,
+      "grad_norm": 0.8154531717300415,
+      "learning_rate": 9.817861911543002e-06,
+      "loss": 0.8687,
+      "step": 3223
+    },
+    {
+      "epoch": 0.17744509879465023,
+      "grad_norm": 0.8346067070960999,
+      "learning_rate": 9.817745963795144e-06,
+      "loss": 0.8905,
+      "step": 3224
+    },
+    {
+      "epoch": 0.1775001375970059,
+      "grad_norm": 0.7137764096260071,
+      "learning_rate": 9.817629979838401e-06,
+      "loss": 0.7715,
+      "step": 3225
+    },
+    {
+      "epoch": 0.17755517639936155,
+      "grad_norm": 0.7237628102302551,
+      "learning_rate": 9.81751395967364e-06,
+      "loss": 0.7824,
+      "step": 3226
+    },
+    {
+      "epoch": 0.17761021520171721,
+      "grad_norm": 0.9481163024902344,
+      "learning_rate": 9.817397903301733e-06,
+      "loss": 0.7451,
+      "step": 3227
+    },
+    {
+      "epoch": 0.17766525400407288,
+      "grad_norm": 0.9472424387931824,
+      "learning_rate": 9.817281810723552e-06,
+      "loss": 0.8774,
+      "step": 3228
+    },
+    {
+      "epoch": 0.17772029280642854,
+      "grad_norm": 0.9295538663864136,
+      "learning_rate": 9.81716568193997e-06,
+      "loss": 0.8507,
+      "step": 3229
+    },
+    {
+      "epoch": 0.1777753316087842,
+      "grad_norm": 0.7668172717094421,
+      "learning_rate": 9.817049516951863e-06,
+      "loss": 0.8547,
+      "step": 3230
+    },
+    {
+      "epoch": 0.17783037041113986,
+      "grad_norm": 0.8640413880348206,
+      "learning_rate": 9.8169333157601e-06,
+      "loss": 0.8485,
+      "step": 3231
+    },
+    {
+      "epoch": 0.17788540921349552,
+      "grad_norm": 0.9901431798934937,
+      "learning_rate": 9.816817078365554e-06,
+      "loss": 0.9236,
+      "step": 3232
+    },
+    {
+      "epoch": 0.17794044801585118,
+      "grad_norm": 1.0242371559143066,
+      "learning_rate": 9.816700804769104e-06,
+      "loss": 0.8096,
+      "step": 3233
+    },
+    {
+      "epoch": 0.17799548681820684,
+      "grad_norm": 0.910498857498169,
+      "learning_rate": 9.816584494971617e-06,
+      "loss": 0.829,
+      "step": 3234
+    },
+    {
+      "epoch": 0.1780505256205625,
+      "grad_norm": 0.8254473805427551,
+      "learning_rate": 9.816468148973972e-06,
+      "loss": 0.7828,
+      "step": 3235
+    },
+    {
+      "epoch": 0.17810556442291817,
+      "grad_norm": 0.7971221804618835,
+      "learning_rate": 9.816351766777039e-06,
+      "loss": 0.8057,
+      "step": 3236
+    },
+    {
+      "epoch": 0.17816060322527383,
+      "grad_norm": 0.8151674270629883,
+      "learning_rate": 9.816235348381697e-06,
+      "loss": 0.7801,
+      "step": 3237
+    },
+    {
+      "epoch": 0.1782156420276295,
+      "grad_norm": 0.7587556838989258,
+      "learning_rate": 9.81611889378882e-06,
+      "loss": 0.7814,
+      "step": 3238
+    },
+    {
+      "epoch": 0.17827068082998515,
+      "grad_norm": 0.8843516111373901,
+      "learning_rate": 9.816002402999283e-06,
+      "loss": 0.8873,
+      "step": 3239
+    },
+    {
+      "epoch": 0.1783257196323408,
+      "grad_norm": 0.917859673500061,
+      "learning_rate": 9.81588587601396e-06,
+      "loss": 0.8963,
+      "step": 3240
+    },
+    {
+      "epoch": 0.17838075843469647,
+      "grad_norm": 0.8256439566612244,
+      "learning_rate": 9.815769312833727e-06,
+      "loss": 0.9157,
+      "step": 3241
+    },
+    {
+      "epoch": 0.17843579723705214,
+      "grad_norm": 0.8364603519439697,
+      "learning_rate": 9.815652713459462e-06,
+      "loss": 0.8253,
+      "step": 3242
+    },
+    {
+      "epoch": 0.17849083603940777,
+      "grad_norm": 0.7717131972312927,
+      "learning_rate": 9.81553607789204e-06,
+      "loss": 0.7211,
+      "step": 3243
+    },
+    {
+      "epoch": 0.17854587484176343,
+      "grad_norm": 0.8069111704826355,
+      "learning_rate": 9.815419406132338e-06,
+      "loss": 0.8986,
+      "step": 3244
+    },
+    {
+      "epoch": 0.1786009136441191,
+      "grad_norm": 0.9176943302154541,
+      "learning_rate": 9.815302698181233e-06,
+      "loss": 0.8084,
+      "step": 3245
+    },
+    {
+      "epoch": 0.17865595244647475,
+      "grad_norm": 0.769183874130249,
+      "learning_rate": 9.815185954039601e-06,
+      "loss": 0.8084,
+      "step": 3246
+    },
+    {
+      "epoch": 0.17871099124883041,
+      "grad_norm": 0.8070697784423828,
+      "learning_rate": 9.815069173708321e-06,
+      "loss": 0.8371,
+      "step": 3247
+    },
+    {
+      "epoch": 0.17876603005118608,
+      "grad_norm": 0.7837347388267517,
+      "learning_rate": 9.81495235718827e-06,
+      "loss": 0.8015,
+      "step": 3248
+    },
+    {
+      "epoch": 0.17882106885354174,
+      "grad_norm": 0.9248430728912354,
+      "learning_rate": 9.814835504480327e-06,
+      "loss": 0.8396,
+      "step": 3249
+    },
+    {
+      "epoch": 0.1788761076558974,
+      "grad_norm": 0.7914367914199829,
+      "learning_rate": 9.814718615585367e-06,
+      "loss": 0.8068,
+      "step": 3250
+    },
+    {
+      "epoch": 0.17893114645825306,
+      "grad_norm": 0.8612570762634277,
+      "learning_rate": 9.814601690504273e-06,
+      "loss": 0.8227,
+      "step": 3251
+    },
+    {
+      "epoch": 0.17898618526060872,
+      "grad_norm": 0.7476248741149902,
+      "learning_rate": 9.81448472923792e-06,
+      "loss": 0.8609,
+      "step": 3252
+    },
+    {
+      "epoch": 0.17904122406296438,
+      "grad_norm": 0.7455218434333801,
+      "learning_rate": 9.81436773178719e-06,
+      "loss": 0.7992,
+      "step": 3253
+    },
+    {
+      "epoch": 0.17909626286532004,
+      "grad_norm": 0.7917896509170532,
+      "learning_rate": 9.814250698152958e-06,
+      "loss": 0.8383,
+      "step": 3254
+    },
+    {
+      "epoch": 0.1791513016676757,
+      "grad_norm": 0.6926130652427673,
+      "learning_rate": 9.81413362833611e-06,
+      "loss": 0.709,
+      "step": 3255
+    },
+    {
+      "epoch": 0.17920634047003137,
+      "grad_norm": 0.8219630718231201,
+      "learning_rate": 9.814016522337519e-06,
+      "loss": 0.9387,
+      "step": 3256
+    },
+    {
+      "epoch": 0.17926137927238703,
+      "grad_norm": 0.8588619828224182,
+      "learning_rate": 9.81389938015807e-06,
+      "loss": 0.8354,
+      "step": 3257
+    },
+    {
+      "epoch": 0.1793164180747427,
+      "grad_norm": 0.7868718504905701,
+      "learning_rate": 9.81378220179864e-06,
+      "loss": 0.8464,
+      "step": 3258
+    },
+    {
+      "epoch": 0.17937145687709835,
+      "grad_norm": 0.789479672908783,
+      "learning_rate": 9.813664987260114e-06,
+      "loss": 0.8577,
+      "step": 3259
+    },
+    {
+      "epoch": 0.179426495679454,
+      "grad_norm": 0.8280717730522156,
+      "learning_rate": 9.81354773654337e-06,
+      "loss": 0.765,
+      "step": 3260
+    },
+    {
+      "epoch": 0.17948153448180967,
+      "grad_norm": 0.7660181522369385,
+      "learning_rate": 9.813430449649289e-06,
+      "loss": 0.7116,
+      "step": 3261
+    },
+    {
+      "epoch": 0.17953657328416534,
+      "grad_norm": 0.8043892979621887,
+      "learning_rate": 9.813313126578754e-06,
+      "loss": 0.8398,
+      "step": 3262
+    },
+    {
+      "epoch": 0.179591612086521,
+      "grad_norm": 0.8708420991897583,
+      "learning_rate": 9.813195767332647e-06,
+      "loss": 0.8246,
+      "step": 3263
+    },
+    {
+      "epoch": 0.17964665088887666,
+      "grad_norm": 1.1456964015960693,
+      "learning_rate": 9.813078371911846e-06,
+      "loss": 0.8798,
+      "step": 3264
+    },
+    {
+      "epoch": 0.17970168969123232,
+      "grad_norm": 0.9668154716491699,
+      "learning_rate": 9.812960940317238e-06,
+      "loss": 0.9645,
+      "step": 3265
+    },
+    {
+      "epoch": 0.17975672849358798,
+      "grad_norm": 0.862050473690033,
+      "learning_rate": 9.812843472549705e-06,
+      "loss": 0.8675,
+      "step": 3266
+    },
+    {
+      "epoch": 0.17981176729594364,
+      "grad_norm": 0.7776491641998291,
+      "learning_rate": 9.812725968610126e-06,
+      "loss": 0.7727,
+      "step": 3267
+    },
+    {
+      "epoch": 0.1798668060982993,
+      "grad_norm": 0.7197048664093018,
+      "learning_rate": 9.812608428499389e-06,
+      "loss": 0.6877,
+      "step": 3268
+    },
+    {
+      "epoch": 0.17992184490065496,
+      "grad_norm": 0.7995713353157043,
+      "learning_rate": 9.812490852218375e-06,
+      "loss": 0.8576,
+      "step": 3269
+    },
+    {
+      "epoch": 0.17997688370301063,
+      "grad_norm": 0.8300820589065552,
+      "learning_rate": 9.812373239767967e-06,
+      "loss": 0.8119,
+      "step": 3270
+    },
+    {
+      "epoch": 0.1800319225053663,
+      "grad_norm": 0.8625856041908264,
+      "learning_rate": 9.812255591149052e-06,
+      "loss": 0.7547,
+      "step": 3271
+    },
+    {
+      "epoch": 0.18008696130772195,
+      "grad_norm": 1.016419768333435,
+      "learning_rate": 9.812137906362511e-06,
+      "loss": 0.8457,
+      "step": 3272
+    },
+    {
+      "epoch": 0.1801420001100776,
+      "grad_norm": 0.7303110361099243,
+      "learning_rate": 9.812020185409229e-06,
+      "loss": 0.7954,
+      "step": 3273
+    },
+    {
+      "epoch": 0.18019703891243327,
+      "grad_norm": 0.8632498383522034,
+      "learning_rate": 9.811902428290093e-06,
+      "loss": 0.8952,
+      "step": 3274
+    },
+    {
+      "epoch": 0.18025207771478893,
+      "grad_norm": 0.7666932940483093,
+      "learning_rate": 9.811784635005984e-06,
+      "loss": 0.746,
+      "step": 3275
+    },
+    {
+      "epoch": 0.1803071165171446,
+      "grad_norm": 0.8962032198905945,
+      "learning_rate": 9.811666805557791e-06,
+      "loss": 0.8654,
+      "step": 3276
+    },
+    {
+      "epoch": 0.18036215531950026,
+      "grad_norm": 0.9399656057357788,
+      "learning_rate": 9.811548939946397e-06,
+      "loss": 0.8062,
+      "step": 3277
+    },
+    {
+      "epoch": 0.18041719412185592,
+      "grad_norm": 0.7469807863235474,
+      "learning_rate": 9.811431038172692e-06,
+      "loss": 0.79,
+      "step": 3278
+    },
+    {
+      "epoch": 0.18047223292421158,
+      "grad_norm": 0.7661105394363403,
+      "learning_rate": 9.811313100237556e-06,
+      "loss": 0.7768,
+      "step": 3279
+    },
+    {
+      "epoch": 0.18052727172656724,
+      "grad_norm": 0.7567458748817444,
+      "learning_rate": 9.811195126141881e-06,
+      "loss": 0.7329,
+      "step": 3280
+    },
+    {
+      "epoch": 0.1805823105289229,
+      "grad_norm": 0.7187278866767883,
+      "learning_rate": 9.811077115886552e-06,
+      "loss": 0.6511,
+      "step": 3281
+    },
+    {
+      "epoch": 0.18063734933127856,
+      "grad_norm": 0.7641230821609497,
+      "learning_rate": 9.810959069472452e-06,
+      "loss": 0.7704,
+      "step": 3282
+    },
+    {
+      "epoch": 0.18069238813363422,
+      "grad_norm": 0.7790704369544983,
+      "learning_rate": 9.810840986900474e-06,
+      "loss": 0.8142,
+      "step": 3283
+    },
+    {
+      "epoch": 0.18074742693598989,
+      "grad_norm": 0.8102816343307495,
+      "learning_rate": 9.810722868171502e-06,
+      "loss": 0.765,
+      "step": 3284
+    },
+    {
+      "epoch": 0.18080246573834555,
+      "grad_norm": 0.7251957058906555,
+      "learning_rate": 9.810604713286424e-06,
+      "loss": 0.7836,
+      "step": 3285
+    },
+    {
+      "epoch": 0.18085750454070118,
+      "grad_norm": 0.845348060131073,
+      "learning_rate": 9.81048652224613e-06,
+      "loss": 0.8386,
+      "step": 3286
+    },
+    {
+      "epoch": 0.18091254334305684,
+      "grad_norm": 0.8397864103317261,
+      "learning_rate": 9.810368295051507e-06,
+      "loss": 0.805,
+      "step": 3287
+    },
+    {
+      "epoch": 0.1809675821454125,
+      "grad_norm": 1.0739909410476685,
+      "learning_rate": 9.810250031703444e-06,
+      "loss": 0.8735,
+      "step": 3288
+    },
+    {
+      "epoch": 0.18102262094776816,
+      "grad_norm": 0.752091646194458,
+      "learning_rate": 9.810131732202826e-06,
+      "loss": 0.7814,
+      "step": 3289
+    },
+    {
+      "epoch": 0.18107765975012383,
+      "grad_norm": 0.7826841473579407,
+      "learning_rate": 9.810013396550548e-06,
+      "loss": 0.7761,
+      "step": 3290
+    },
+    {
+      "epoch": 0.1811326985524795,
+      "grad_norm": 0.6979131102561951,
+      "learning_rate": 9.809895024747498e-06,
+      "loss": 0.672,
+      "step": 3291
+    },
+    {
+      "epoch": 0.18118773735483515,
+      "grad_norm": 0.8571394085884094,
+      "learning_rate": 9.809776616794562e-06,
+      "loss": 0.8795,
+      "step": 3292
+    },
+    {
+      "epoch": 0.1812427761571908,
+      "grad_norm": 0.8287902474403381,
+      "learning_rate": 9.809658172692634e-06,
+      "loss": 0.9032,
+      "step": 3293
+    },
+    {
+      "epoch": 0.18129781495954647,
+      "grad_norm": 0.7884420156478882,
+      "learning_rate": 9.809539692442602e-06,
+      "loss": 0.87,
+      "step": 3294
+    },
+    {
+      "epoch": 0.18135285376190213,
+      "grad_norm": 0.8955305218696594,
+      "learning_rate": 9.809421176045358e-06,
+      "loss": 0.7982,
+      "step": 3295
+    },
+    {
+      "epoch": 0.1814078925642578,
+      "grad_norm": 0.7893335819244385,
+      "learning_rate": 9.809302623501791e-06,
+      "loss": 0.7792,
+      "step": 3296
+    },
+    {
+      "epoch": 0.18146293136661346,
+      "grad_norm": 0.8077870011329651,
+      "learning_rate": 9.809184034812794e-06,
+      "loss": 0.829,
+      "step": 3297
+    },
+    {
+      "epoch": 0.18151797016896912,
+      "grad_norm": 0.8282631635665894,
+      "learning_rate": 9.809065409979256e-06,
+      "loss": 0.8502,
+      "step": 3298
+    },
+    {
+      "epoch": 0.18157300897132478,
+      "grad_norm": 0.7988418936729431,
+      "learning_rate": 9.808946749002068e-06,
+      "loss": 0.7853,
+      "step": 3299
+    },
+    {
+      "epoch": 0.18162804777368044,
+      "grad_norm": 0.7776056528091431,
+      "learning_rate": 9.808828051882127e-06,
+      "loss": 0.7843,
+      "step": 3300
+    },
+    {
+      "epoch": 0.1816830865760361,
+      "grad_norm": 0.8772258758544922,
+      "learning_rate": 9.80870931862032e-06,
+      "loss": 0.896,
+      "step": 3301
+    },
+    {
+      "epoch": 0.18173812537839176,
+      "grad_norm": 0.8080328702926636,
+      "learning_rate": 9.80859054921754e-06,
+      "loss": 0.8464,
+      "step": 3302
+    },
+    {
+      "epoch": 0.18179316418074742,
+      "grad_norm": 0.862707257270813,
+      "learning_rate": 9.808471743674682e-06,
+      "loss": 0.8732,
+      "step": 3303
+    },
+    {
+      "epoch": 0.18184820298310309,
+      "grad_norm": 1.1964820623397827,
+      "learning_rate": 9.808352901992637e-06,
+      "loss": 0.9911,
+      "step": 3304
+    },
+    {
+      "epoch": 0.18190324178545875,
+      "grad_norm": 0.8597685694694519,
+      "learning_rate": 9.808234024172298e-06,
+      "loss": 0.8724,
+      "step": 3305
+    },
+    {
+      "epoch": 0.1819582805878144,
+      "grad_norm": 0.8068556189537048,
+      "learning_rate": 9.80811511021456e-06,
+      "loss": 0.8116,
+      "step": 3306
+    },
+    {
+      "epoch": 0.18201331939017007,
+      "grad_norm": 1.0014268159866333,
+      "learning_rate": 9.807996160120317e-06,
+      "loss": 0.8585,
+      "step": 3307
+    },
+    {
+      "epoch": 0.18206835819252573,
+      "grad_norm": 0.8541132807731628,
+      "learning_rate": 9.80787717389046e-06,
+      "loss": 0.8505,
+      "step": 3308
+    },
+    {
+      "epoch": 0.1821233969948814,
+      "grad_norm": 0.7973629832267761,
+      "learning_rate": 9.807758151525886e-06,
+      "loss": 0.8312,
+      "step": 3309
+    },
+    {
+      "epoch": 0.18217843579723705,
+      "grad_norm": 0.82973712682724,
+      "learning_rate": 9.807639093027488e-06,
+      "loss": 0.894,
+      "step": 3310
+    },
+    {
+      "epoch": 0.18223347459959272,
+      "grad_norm": 0.7729674577713013,
+      "learning_rate": 9.807519998396162e-06,
+      "loss": 0.7459,
+      "step": 3311
+    },
+    {
+      "epoch": 0.18228851340194838,
+      "grad_norm": 0.8106189370155334,
+      "learning_rate": 9.807400867632804e-06,
+      "loss": 0.914,
+      "step": 3312
+    },
+    {
+      "epoch": 0.18234355220430404,
+      "grad_norm": 0.7672377228736877,
+      "learning_rate": 9.807281700738305e-06,
+      "loss": 0.8475,
+      "step": 3313
+    },
+    {
+      "epoch": 0.1823985910066597,
+      "grad_norm": 0.8776688575744629,
+      "learning_rate": 9.807162497713566e-06,
+      "loss": 0.7641,
+      "step": 3314
+    },
+    {
+      "epoch": 0.18245362980901536,
+      "grad_norm": 0.8781917691230774,
+      "learning_rate": 9.807043258559479e-06,
+      "loss": 0.86,
+      "step": 3315
+    },
+    {
+      "epoch": 0.18250866861137102,
+      "grad_norm": 0.819362998008728,
+      "learning_rate": 9.806923983276942e-06,
+      "loss": 0.8829,
+      "step": 3316
+    },
+    {
+      "epoch": 0.18256370741372668,
+      "grad_norm": 0.8065270185470581,
+      "learning_rate": 9.80680467186685e-06,
+      "loss": 0.7335,
+      "step": 3317
+    },
+    {
+      "epoch": 0.18261874621608234,
+      "grad_norm": 0.8692485690116882,
+      "learning_rate": 9.806685324330102e-06,
+      "loss": 0.8582,
+      "step": 3318
+    },
+    {
+      "epoch": 0.182673785018438,
+      "grad_norm": 0.7910160422325134,
+      "learning_rate": 9.806565940667594e-06,
+      "loss": 0.8569,
+      "step": 3319
+    },
+    {
+      "epoch": 0.18272882382079367,
+      "grad_norm": 0.8282253742218018,
+      "learning_rate": 9.806446520880225e-06,
+      "loss": 0.7791,
+      "step": 3320
+    },
+    {
+      "epoch": 0.18278386262314933,
+      "grad_norm": 0.7513861060142517,
+      "learning_rate": 9.806327064968887e-06,
+      "loss": 0.7287,
+      "step": 3321
+    },
+    {
+      "epoch": 0.182838901425505,
+      "grad_norm": 0.8141188621520996,
+      "learning_rate": 9.806207572934483e-06,
+      "loss": 0.7772,
+      "step": 3322
+    },
+    {
+      "epoch": 0.18289394022786065,
+      "grad_norm": 0.7963125705718994,
+      "learning_rate": 9.806088044777909e-06,
+      "loss": 0.7993,
+      "step": 3323
+    },
+    {
+      "epoch": 0.1829489790302163,
+      "grad_norm": 0.8527218103408813,
+      "learning_rate": 9.805968480500063e-06,
+      "loss": 0.822,
+      "step": 3324
+    },
+    {
+      "epoch": 0.18300401783257197,
+      "grad_norm": 0.822467565536499,
+      "learning_rate": 9.805848880101845e-06,
+      "loss": 0.8606,
+      "step": 3325
+    },
+    {
+      "epoch": 0.18305905663492764,
+      "grad_norm": 0.8197154402732849,
+      "learning_rate": 9.805729243584154e-06,
+      "loss": 0.9004,
+      "step": 3326
+    },
+    {
+      "epoch": 0.1831140954372833,
+      "grad_norm": 0.8379594683647156,
+      "learning_rate": 9.805609570947887e-06,
+      "loss": 0.8467,
+      "step": 3327
+    },
+    {
+      "epoch": 0.18316913423963896,
+      "grad_norm": 0.7787355184555054,
+      "learning_rate": 9.805489862193947e-06,
+      "loss": 0.8221,
+      "step": 3328
+    },
+    {
+      "epoch": 0.1832241730419946,
+      "grad_norm": 0.8464100956916809,
+      "learning_rate": 9.80537011732323e-06,
+      "loss": 0.7722,
+      "step": 3329
+    },
+    {
+      "epoch": 0.18327921184435025,
+      "grad_norm": 0.8351306319236755,
+      "learning_rate": 9.805250336336637e-06,
+      "loss": 0.7638,
+      "step": 3330
+    },
+    {
+      "epoch": 0.18333425064670592,
+      "grad_norm": 0.8098864555358887,
+      "learning_rate": 9.805130519235068e-06,
+      "loss": 0.8448,
+      "step": 3331
+    },
+    {
+      "epoch": 0.18338928944906158,
+      "grad_norm": 0.8290563821792603,
+      "learning_rate": 9.805010666019427e-06,
+      "loss": 0.6574,
+      "step": 3332
+    },
+    {
+      "epoch": 0.18344432825141724,
+      "grad_norm": 0.7748262882232666,
+      "learning_rate": 9.804890776690611e-06,
+      "loss": 0.8002,
+      "step": 3333
+    },
+    {
+      "epoch": 0.1834993670537729,
+      "grad_norm": 0.8422787189483643,
+      "learning_rate": 9.80477085124952e-06,
+      "loss": 0.8452,
+      "step": 3334
+    },
+    {
+      "epoch": 0.18355440585612856,
+      "grad_norm": 0.7776510119438171,
+      "learning_rate": 9.804650889697061e-06,
+      "loss": 0.8774,
+      "step": 3335
+    },
+    {
+      "epoch": 0.18360944465848422,
+      "grad_norm": 0.8449370861053467,
+      "learning_rate": 9.80453089203413e-06,
+      "loss": 0.8233,
+      "step": 3336
+    },
+    {
+      "epoch": 0.18366448346083988,
+      "grad_norm": 0.8254217505455017,
+      "learning_rate": 9.804410858261632e-06,
+      "loss": 0.8778,
+      "step": 3337
+    },
+    {
+      "epoch": 0.18371952226319554,
+      "grad_norm": 0.8673515915870667,
+      "learning_rate": 9.804290788380466e-06,
+      "loss": 0.8005,
+      "step": 3338
+    },
+    {
+      "epoch": 0.1837745610655512,
+      "grad_norm": 0.8106067776679993,
+      "learning_rate": 9.804170682391538e-06,
+      "loss": 0.86,
+      "step": 3339
+    },
+    {
+      "epoch": 0.18382959986790687,
+      "grad_norm": 0.8211669325828552,
+      "learning_rate": 9.804050540295749e-06,
+      "loss": 0.8013,
+      "step": 3340
+    },
+    {
+      "epoch": 0.18388463867026253,
+      "grad_norm": 0.7866180539131165,
+      "learning_rate": 9.803930362094003e-06,
+      "loss": 0.8108,
+      "step": 3341
+    },
+    {
+      "epoch": 0.1839396774726182,
+      "grad_norm": 0.8192055225372314,
+      "learning_rate": 9.8038101477872e-06,
+      "loss": 0.7586,
+      "step": 3342
+    },
+    {
+      "epoch": 0.18399471627497385,
+      "grad_norm": 0.940910279750824,
+      "learning_rate": 9.803689897376248e-06,
+      "loss": 0.8174,
+      "step": 3343
+    },
+    {
+      "epoch": 0.1840497550773295,
+      "grad_norm": 0.7979292869567871,
+      "learning_rate": 9.803569610862048e-06,
+      "loss": 0.8341,
+      "step": 3344
+    },
+    {
+      "epoch": 0.18410479387968517,
+      "grad_norm": 0.7577546238899231,
+      "learning_rate": 9.803449288245504e-06,
+      "loss": 0.7775,
+      "step": 3345
+    },
+    {
+      "epoch": 0.18415983268204084,
+      "grad_norm": 0.7255160212516785,
+      "learning_rate": 9.80332892952752e-06,
+      "loss": 0.7648,
+      "step": 3346
+    },
+    {
+      "epoch": 0.1842148714843965,
+      "grad_norm": 0.8269388675689697,
+      "learning_rate": 9.803208534709004e-06,
+      "loss": 0.8902,
+      "step": 3347
+    },
+    {
+      "epoch": 0.18426991028675216,
+      "grad_norm": 0.783867359161377,
+      "learning_rate": 9.803088103790857e-06,
+      "loss": 0.8191,
+      "step": 3348
+    },
+    {
+      "epoch": 0.18432494908910782,
+      "grad_norm": 0.7658863663673401,
+      "learning_rate": 9.802967636773986e-06,
+      "loss": 0.7505,
+      "step": 3349
+    },
+    {
+      "epoch": 0.18437998789146348,
+      "grad_norm": 0.701225757598877,
+      "learning_rate": 9.802847133659294e-06,
+      "loss": 0.7159,
+      "step": 3350
+    },
+    {
+      "epoch": 0.18443502669381914,
+      "grad_norm": 0.9224311709403992,
+      "learning_rate": 9.802726594447692e-06,
+      "loss": 0.7766,
+      "step": 3351
+    },
+    {
+      "epoch": 0.1844900654961748,
+      "grad_norm": 0.8835979700088501,
+      "learning_rate": 9.80260601914008e-06,
+      "loss": 0.9304,
+      "step": 3352
+    },
+    {
+      "epoch": 0.18454510429853047,
+      "grad_norm": 0.7918481826782227,
+      "learning_rate": 9.802485407737368e-06,
+      "loss": 0.7691,
+      "step": 3353
+    },
+    {
+      "epoch": 0.18460014310088613,
+      "grad_norm": 0.8855286240577698,
+      "learning_rate": 9.80236476024046e-06,
+      "loss": 0.9213,
+      "step": 3354
+    },
+    {
+      "epoch": 0.1846551819032418,
+      "grad_norm": 0.7863314747810364,
+      "learning_rate": 9.802244076650264e-06,
+      "loss": 0.7675,
+      "step": 3355
+    },
+    {
+      "epoch": 0.18471022070559745,
+      "grad_norm": 0.8230198621749878,
+      "learning_rate": 9.802123356967687e-06,
+      "loss": 0.7243,
+      "step": 3356
+    },
+    {
+      "epoch": 0.1847652595079531,
+      "grad_norm": 0.8038737773895264,
+      "learning_rate": 9.80200260119364e-06,
+      "loss": 0.8094,
+      "step": 3357
+    },
+    {
+      "epoch": 0.18482029831030877,
+      "grad_norm": 0.7656993269920349,
+      "learning_rate": 9.801881809329022e-06,
+      "loss": 0.7736,
+      "step": 3358
+    },
+    {
+      "epoch": 0.18487533711266443,
+      "grad_norm": 0.8222082853317261,
+      "learning_rate": 9.801760981374747e-06,
+      "loss": 0.844,
+      "step": 3359
+    },
+    {
+      "epoch": 0.1849303759150201,
+      "grad_norm": 0.7632889747619629,
+      "learning_rate": 9.801640117331723e-06,
+      "loss": 0.8354,
+      "step": 3360
+    },
+    {
+      "epoch": 0.18498541471737576,
+      "grad_norm": 0.8308513760566711,
+      "learning_rate": 9.801519217200857e-06,
+      "loss": 0.8277,
+      "step": 3361
+    },
+    {
+      "epoch": 0.18504045351973142,
+      "grad_norm": 0.7865434885025024,
+      "learning_rate": 9.801398280983057e-06,
+      "loss": 0.8614,
+      "step": 3362
+    },
+    {
+      "epoch": 0.18509549232208708,
+      "grad_norm": 0.7249410152435303,
+      "learning_rate": 9.801277308679232e-06,
+      "loss": 0.7259,
+      "step": 3363
+    },
+    {
+      "epoch": 0.18515053112444274,
+      "grad_norm": 0.7604461908340454,
+      "learning_rate": 9.801156300290293e-06,
+      "loss": 0.8507,
+      "step": 3364
+    },
+    {
+      "epoch": 0.1852055699267984,
+      "grad_norm": 0.8725959062576294,
+      "learning_rate": 9.801035255817149e-06,
+      "loss": 0.7688,
+      "step": 3365
+    },
+    {
+      "epoch": 0.18526060872915406,
+      "grad_norm": 0.7798827290534973,
+      "learning_rate": 9.800914175260708e-06,
+      "loss": 0.8788,
+      "step": 3366
+    },
+    {
+      "epoch": 0.18531564753150973,
+      "grad_norm": 0.7060996890068054,
+      "learning_rate": 9.800793058621882e-06,
+      "loss": 0.8183,
+      "step": 3367
+    },
+    {
+      "epoch": 0.1853706863338654,
+      "grad_norm": 0.7558063268661499,
+      "learning_rate": 9.80067190590158e-06,
+      "loss": 0.7834,
+      "step": 3368
+    },
+    {
+      "epoch": 0.18542572513622105,
+      "grad_norm": 0.7411057353019714,
+      "learning_rate": 9.800550717100714e-06,
+      "loss": 0.8298,
+      "step": 3369
+    },
+    {
+      "epoch": 0.1854807639385767,
+      "grad_norm": 0.8466144800186157,
+      "learning_rate": 9.800429492220193e-06,
+      "loss": 0.8297,
+      "step": 3370
+    },
+    {
+      "epoch": 0.18553580274093237,
+      "grad_norm": 0.7302330136299133,
+      "learning_rate": 9.800308231260928e-06,
+      "loss": 0.72,
+      "step": 3371
+    },
+    {
+      "epoch": 0.185590841543288,
+      "grad_norm": 0.8140530586242676,
+      "learning_rate": 9.800186934223832e-06,
+      "loss": 0.9287,
+      "step": 3372
+    },
+    {
+      "epoch": 0.18564588034564367,
+      "grad_norm": 0.8246129751205444,
+      "learning_rate": 9.800065601109817e-06,
+      "loss": 0.7891,
+      "step": 3373
+    },
+    {
+      "epoch": 0.18570091914799933,
+      "grad_norm": 0.8746623396873474,
+      "learning_rate": 9.799944231919794e-06,
+      "loss": 0.8549,
+      "step": 3374
+    },
+    {
+      "epoch": 0.185755957950355,
+      "grad_norm": 0.9977195858955383,
+      "learning_rate": 9.799822826654672e-06,
+      "loss": 0.821,
+      "step": 3375
+    },
+    {
+      "epoch": 0.18581099675271065,
+      "grad_norm": 0.8937395811080933,
+      "learning_rate": 9.79970138531537e-06,
+      "loss": 0.8639,
+      "step": 3376
+    },
+    {
+      "epoch": 0.1858660355550663,
+      "grad_norm": 1.039695143699646,
+      "learning_rate": 9.799579907902794e-06,
+      "loss": 1.0425,
+      "step": 3377
+    },
+    {
+      "epoch": 0.18592107435742197,
+      "grad_norm": 0.7847749590873718,
+      "learning_rate": 9.799458394417863e-06,
+      "loss": 0.8505,
+      "step": 3378
+    },
+    {
+      "epoch": 0.18597611315977763,
+      "grad_norm": 0.760334312915802,
+      "learning_rate": 9.799336844861486e-06,
+      "loss": 0.7418,
+      "step": 3379
+    },
+    {
+      "epoch": 0.1860311519621333,
+      "grad_norm": 0.7599604725837708,
+      "learning_rate": 9.799215259234578e-06,
+      "loss": 0.8305,
+      "step": 3380
+    },
+    {
+      "epoch": 0.18608619076448896,
+      "grad_norm": 0.846767246723175,
+      "learning_rate": 9.799093637538054e-06,
+      "loss": 0.7526,
+      "step": 3381
+    },
+    {
+      "epoch": 0.18614122956684462,
+      "grad_norm": 0.7840956449508667,
+      "learning_rate": 9.798971979772825e-06,
+      "loss": 0.8009,
+      "step": 3382
+    },
+    {
+      "epoch": 0.18619626836920028,
+      "grad_norm": 0.7826499342918396,
+      "learning_rate": 9.798850285939809e-06,
+      "loss": 0.821,
+      "step": 3383
+    },
+    {
+      "epoch": 0.18625130717155594,
+      "grad_norm": 0.7829813361167908,
+      "learning_rate": 9.798728556039918e-06,
+      "loss": 0.8053,
+      "step": 3384
+    },
+    {
+      "epoch": 0.1863063459739116,
+      "grad_norm": 0.7267470359802246,
+      "learning_rate": 9.798606790074067e-06,
+      "loss": 0.6797,
+      "step": 3385
+    },
+    {
+      "epoch": 0.18636138477626726,
+      "grad_norm": 0.8560196757316589,
+      "learning_rate": 9.798484988043173e-06,
+      "loss": 0.8476,
+      "step": 3386
+    },
+    {
+      "epoch": 0.18641642357862293,
+      "grad_norm": 0.7920921444892883,
+      "learning_rate": 9.798363149948148e-06,
+      "loss": 0.8832,
+      "step": 3387
+    },
+    {
+      "epoch": 0.1864714623809786,
+      "grad_norm": 0.8414384126663208,
+      "learning_rate": 9.798241275789912e-06,
+      "loss": 0.8607,
+      "step": 3388
+    },
+    {
+      "epoch": 0.18652650118333425,
+      "grad_norm": 0.7255431413650513,
+      "learning_rate": 9.798119365569378e-06,
+      "loss": 0.6426,
+      "step": 3389
+    },
+    {
+      "epoch": 0.1865815399856899,
+      "grad_norm": 0.8842852711677551,
+      "learning_rate": 9.797997419287465e-06,
+      "loss": 0.9058,
+      "step": 3390
+    },
+    {
+      "epoch": 0.18663657878804557,
+      "grad_norm": 0.7178265452384949,
+      "learning_rate": 9.797875436945086e-06,
+      "loss": 0.8134,
+      "step": 3391
+    },
+    {
+      "epoch": 0.18669161759040123,
+      "grad_norm": 0.7275096774101257,
+      "learning_rate": 9.797753418543161e-06,
+      "loss": 0.6858,
+      "step": 3392
+    },
+    {
+      "epoch": 0.1867466563927569,
+      "grad_norm": 0.7587800025939941,
+      "learning_rate": 9.797631364082605e-06,
+      "loss": 0.7437,
+      "step": 3393
+    },
+    {
+      "epoch": 0.18680169519511255,
+      "grad_norm": 0.9769744873046875,
+      "learning_rate": 9.797509273564336e-06,
+      "loss": 0.8024,
+      "step": 3394
+    },
+    {
+      "epoch": 0.18685673399746822,
+      "grad_norm": 0.7662433385848999,
+      "learning_rate": 9.79738714698927e-06,
+      "loss": 0.8122,
+      "step": 3395
+    },
+    {
+      "epoch": 0.18691177279982388,
+      "grad_norm": 0.8620306849479675,
+      "learning_rate": 9.797264984358328e-06,
+      "loss": 0.7952,
+      "step": 3396
+    },
+    {
+      "epoch": 0.18696681160217954,
+      "grad_norm": 0.7542591094970703,
+      "learning_rate": 9.797142785672427e-06,
+      "loss": 0.8315,
+      "step": 3397
+    },
+    {
+      "epoch": 0.1870218504045352,
+      "grad_norm": 0.7273713946342468,
+      "learning_rate": 9.797020550932483e-06,
+      "loss": 0.7316,
+      "step": 3398
+    },
+    {
+      "epoch": 0.18707688920689086,
+      "grad_norm": 1.031592845916748,
+      "learning_rate": 9.796898280139417e-06,
+      "loss": 0.7478,
+      "step": 3399
+    },
+    {
+      "epoch": 0.18713192800924652,
+      "grad_norm": 0.791407585144043,
+      "learning_rate": 9.796775973294147e-06,
+      "loss": 0.7742,
+      "step": 3400
+    },
+    {
+      "epoch": 0.18718696681160218,
+      "grad_norm": 0.8311418294906616,
+      "learning_rate": 9.796653630397595e-06,
+      "loss": 0.8182,
+      "step": 3401
+    },
+    {
+      "epoch": 0.18724200561395785,
+      "grad_norm": 0.7960993051528931,
+      "learning_rate": 9.796531251450678e-06,
+      "loss": 0.7606,
+      "step": 3402
+    },
+    {
+      "epoch": 0.1872970444163135,
+      "grad_norm": 0.8671618103981018,
+      "learning_rate": 9.796408836454316e-06,
+      "loss": 0.7136,
+      "step": 3403
+    },
+    {
+      "epoch": 0.18735208321866917,
+      "grad_norm": 1.1071348190307617,
+      "learning_rate": 9.796286385409428e-06,
+      "loss": 0.7729,
+      "step": 3404
+    },
+    {
+      "epoch": 0.18740712202102483,
+      "grad_norm": 0.738217294216156,
+      "learning_rate": 9.796163898316935e-06,
+      "loss": 0.7425,
+      "step": 3405
+    },
+    {
+      "epoch": 0.1874621608233805,
+      "grad_norm": 0.7567199468612671,
+      "learning_rate": 9.796041375177758e-06,
+      "loss": 0.8442,
+      "step": 3406
+    },
+    {
+      "epoch": 0.18751719962573615,
+      "grad_norm": 0.7942413091659546,
+      "learning_rate": 9.79591881599282e-06,
+      "loss": 0.852,
+      "step": 3407
+    },
+    {
+      "epoch": 0.18757223842809181,
+      "grad_norm": 0.7529355883598328,
+      "learning_rate": 9.795796220763038e-06,
+      "loss": 0.8086,
+      "step": 3408
+    },
+    {
+      "epoch": 0.18762727723044748,
+      "grad_norm": 0.7645192742347717,
+      "learning_rate": 9.795673589489337e-06,
+      "loss": 0.831,
+      "step": 3409
+    },
+    {
+      "epoch": 0.18768231603280314,
+      "grad_norm": 0.694791853427887,
+      "learning_rate": 9.795550922172635e-06,
+      "loss": 0.6919,
+      "step": 3410
+    },
+    {
+      "epoch": 0.1877373548351588,
+      "grad_norm": 0.7041944265365601,
+      "learning_rate": 9.795428218813858e-06,
+      "loss": 0.7284,
+      "step": 3411
+    },
+    {
+      "epoch": 0.18779239363751446,
+      "grad_norm": 0.8972276449203491,
+      "learning_rate": 9.795305479413924e-06,
+      "loss": 0.7156,
+      "step": 3412
+    },
+    {
+      "epoch": 0.18784743243987012,
+      "grad_norm": 0.9730873107910156,
+      "learning_rate": 9.795182703973758e-06,
+      "loss": 0.8739,
+      "step": 3413
+    },
+    {
+      "epoch": 0.18790247124222578,
+      "grad_norm": 0.8137956261634827,
+      "learning_rate": 9.795059892494283e-06,
+      "loss": 0.8189,
+      "step": 3414
+    },
+    {
+      "epoch": 0.18795751004458142,
+      "grad_norm": 0.8171416521072388,
+      "learning_rate": 9.794937044976422e-06,
+      "loss": 0.9449,
+      "step": 3415
+    },
+    {
+      "epoch": 0.18801254884693708,
+      "grad_norm": 0.7929911017417908,
+      "learning_rate": 9.794814161421098e-06,
+      "loss": 0.8034,
+      "step": 3416
+    },
+    {
+      "epoch": 0.18806758764929274,
+      "grad_norm": 1.1045749187469482,
+      "learning_rate": 9.794691241829233e-06,
+      "loss": 0.875,
+      "step": 3417
+    },
+    {
+      "epoch": 0.1881226264516484,
+      "grad_norm": 0.8141040205955505,
+      "learning_rate": 9.794568286201752e-06,
+      "loss": 0.787,
+      "step": 3418
+    },
+    {
+      "epoch": 0.18817766525400406,
+      "grad_norm": 0.7615541815757751,
+      "learning_rate": 9.79444529453958e-06,
+      "loss": 0.8491,
+      "step": 3419
+    },
+    {
+      "epoch": 0.18823270405635972,
+      "grad_norm": 0.848419189453125,
+      "learning_rate": 9.79432226684364e-06,
+      "loss": 0.7445,
+      "step": 3420
+    },
+    {
+      "epoch": 0.18828774285871538,
+      "grad_norm": 0.8075067400932312,
+      "learning_rate": 9.794199203114858e-06,
+      "loss": 0.6581,
+      "step": 3421
+    },
+    {
+      "epoch": 0.18834278166107105,
+      "grad_norm": 0.8473401069641113,
+      "learning_rate": 9.794076103354158e-06,
+      "loss": 0.839,
+      "step": 3422
+    },
+    {
+      "epoch": 0.1883978204634267,
+      "grad_norm": 0.8211609721183777,
+      "learning_rate": 9.793952967562463e-06,
+      "loss": 0.7709,
+      "step": 3423
+    },
+    {
+      "epoch": 0.18845285926578237,
+      "grad_norm": 0.7527804374694824,
+      "learning_rate": 9.793829795740703e-06,
+      "loss": 0.7315,
+      "step": 3424
+    },
+    {
+      "epoch": 0.18850789806813803,
+      "grad_norm": 0.7971188426017761,
+      "learning_rate": 9.793706587889802e-06,
+      "loss": 0.7507,
+      "step": 3425
+    },
+    {
+      "epoch": 0.1885629368704937,
+      "grad_norm": 1.024066686630249,
+      "learning_rate": 9.793583344010684e-06,
+      "loss": 0.9043,
+      "step": 3426
+    },
+    {
+      "epoch": 0.18861797567284935,
+      "grad_norm": 0.7428625226020813,
+      "learning_rate": 9.793460064104276e-06,
+      "loss": 0.7435,
+      "step": 3427
+    },
+    {
+      "epoch": 0.18867301447520501,
+      "grad_norm": 0.8438264727592468,
+      "learning_rate": 9.793336748171507e-06,
+      "loss": 0.8618,
+      "step": 3428
+    },
+    {
+      "epoch": 0.18872805327756068,
+      "grad_norm": 0.7846877574920654,
+      "learning_rate": 9.793213396213302e-06,
+      "loss": 0.8064,
+      "step": 3429
+    },
+    {
+      "epoch": 0.18878309207991634,
+      "grad_norm": 0.7527204751968384,
+      "learning_rate": 9.793090008230587e-06,
+      "loss": 0.7596,
+      "step": 3430
+    },
+    {
+      "epoch": 0.188838130882272,
+      "grad_norm": 1.1236757040023804,
+      "learning_rate": 9.792966584224292e-06,
+      "loss": 0.8292,
+      "step": 3431
+    },
+    {
+      "epoch": 0.18889316968462766,
+      "grad_norm": 0.8128102421760559,
+      "learning_rate": 9.792843124195343e-06,
+      "loss": 0.8073,
+      "step": 3432
+    },
+    {
+      "epoch": 0.18894820848698332,
+      "grad_norm": 0.7668742537498474,
+      "learning_rate": 9.792719628144667e-06,
+      "loss": 0.7848,
+      "step": 3433
+    },
+    {
+      "epoch": 0.18900324728933898,
+      "grad_norm": 1.8663485050201416,
+      "learning_rate": 9.792596096073193e-06,
+      "loss": 0.9388,
+      "step": 3434
+    },
+    {
+      "epoch": 0.18905828609169464,
+      "grad_norm": 0.8066239356994629,
+      "learning_rate": 9.792472527981852e-06,
+      "loss": 0.6647,
+      "step": 3435
+    },
+    {
+      "epoch": 0.1891133248940503,
+      "grad_norm": 0.8268817067146301,
+      "learning_rate": 9.792348923871567e-06,
+      "loss": 0.9676,
+      "step": 3436
+    },
+    {
+      "epoch": 0.18916836369640597,
+      "grad_norm": 0.7165037393569946,
+      "learning_rate": 9.792225283743272e-06,
+      "loss": 0.6937,
+      "step": 3437
+    },
+    {
+      "epoch": 0.18922340249876163,
+      "grad_norm": 0.7850403785705566,
+      "learning_rate": 9.792101607597895e-06,
+      "loss": 0.7782,
+      "step": 3438
+    },
+    {
+      "epoch": 0.1892784413011173,
+      "grad_norm": 0.8839808702468872,
+      "learning_rate": 9.791977895436365e-06,
+      "loss": 0.7639,
+      "step": 3439
+    },
+    {
+      "epoch": 0.18933348010347295,
+      "grad_norm": 0.8260362148284912,
+      "learning_rate": 9.791854147259611e-06,
+      "loss": 0.8201,
+      "step": 3440
+    },
+    {
+      "epoch": 0.1893885189058286,
+      "grad_norm": 0.8792916536331177,
+      "learning_rate": 9.791730363068564e-06,
+      "loss": 0.8251,
+      "step": 3441
+    },
+    {
+      "epoch": 0.18944355770818427,
+      "grad_norm": 0.8192774653434753,
+      "learning_rate": 9.791606542864154e-06,
+      "loss": 0.7944,
+      "step": 3442
+    },
+    {
+      "epoch": 0.18949859651053994,
+      "grad_norm": 0.751470685005188,
+      "learning_rate": 9.791482686647313e-06,
+      "loss": 0.7563,
+      "step": 3443
+    },
+    {
+      "epoch": 0.1895536353128956,
+      "grad_norm": 0.8902072906494141,
+      "learning_rate": 9.79135879441897e-06,
+      "loss": 0.7719,
+      "step": 3444
+    },
+    {
+      "epoch": 0.18960867411525126,
+      "grad_norm": 0.7166435122489929,
+      "learning_rate": 9.791234866180058e-06,
+      "loss": 0.7871,
+      "step": 3445
+    },
+    {
+      "epoch": 0.18966371291760692,
+      "grad_norm": 0.763416588306427,
+      "learning_rate": 9.791110901931505e-06,
+      "loss": 0.8226,
+      "step": 3446
+    },
+    {
+      "epoch": 0.18971875171996258,
+      "grad_norm": 0.806633472442627,
+      "learning_rate": 9.790986901674246e-06,
+      "loss": 0.7828,
+      "step": 3447
+    },
+    {
+      "epoch": 0.18977379052231824,
+      "grad_norm": 0.8139312863349915,
+      "learning_rate": 9.790862865409213e-06,
+      "loss": 0.8441,
+      "step": 3448
+    },
+    {
+      "epoch": 0.1898288293246739,
+      "grad_norm": 0.8362452387809753,
+      "learning_rate": 9.790738793137335e-06,
+      "loss": 0.8765,
+      "step": 3449
+    },
+    {
+      "epoch": 0.18988386812702956,
+      "grad_norm": 0.7736263871192932,
+      "learning_rate": 9.790614684859549e-06,
+      "loss": 0.8373,
+      "step": 3450
+    },
+    {
+      "epoch": 0.18993890692938523,
+      "grad_norm": 0.8742800354957581,
+      "learning_rate": 9.790490540576784e-06,
+      "loss": 0.8976,
+      "step": 3451
+    },
+    {
+      "epoch": 0.1899939457317409,
+      "grad_norm": 0.701505720615387,
+      "learning_rate": 9.790366360289974e-06,
+      "loss": 0.7799,
+      "step": 3452
+    },
+    {
+      "epoch": 0.19004898453409655,
+      "grad_norm": 0.7771356701850891,
+      "learning_rate": 9.790242144000055e-06,
+      "loss": 0.7617,
+      "step": 3453
+    },
+    {
+      "epoch": 0.1901040233364522,
+      "grad_norm": 0.897576093673706,
+      "learning_rate": 9.790117891707955e-06,
+      "loss": 0.7817,
+      "step": 3454
+    },
+    {
+      "epoch": 0.19015906213880787,
+      "grad_norm": 0.7296561002731323,
+      "learning_rate": 9.789993603414613e-06,
+      "loss": 0.8344,
+      "step": 3455
+    },
+    {
+      "epoch": 0.19021410094116353,
+      "grad_norm": 0.8099396228790283,
+      "learning_rate": 9.789869279120962e-06,
+      "loss": 0.7369,
+      "step": 3456
+    },
+    {
+      "epoch": 0.1902691397435192,
+      "grad_norm": 0.7802554368972778,
+      "learning_rate": 9.789744918827935e-06,
+      "loss": 0.8383,
+      "step": 3457
+    },
+    {
+      "epoch": 0.19032417854587483,
+      "grad_norm": 0.7508029341697693,
+      "learning_rate": 9.789620522536467e-06,
+      "loss": 0.825,
+      "step": 3458
+    },
+    {
+      "epoch": 0.1903792173482305,
+      "grad_norm": 0.7782164216041565,
+      "learning_rate": 9.789496090247494e-06,
+      "loss": 0.7737,
+      "step": 3459
+    },
+    {
+      "epoch": 0.19043425615058615,
+      "grad_norm": 0.7711489796638489,
+      "learning_rate": 9.78937162196195e-06,
+      "loss": 0.7694,
+      "step": 3460
+    },
+    {
+      "epoch": 0.1904892949529418,
+      "grad_norm": 0.821579098701477,
+      "learning_rate": 9.789247117680769e-06,
+      "loss": 0.7493,
+      "step": 3461
+    },
+    {
+      "epoch": 0.19054433375529747,
+      "grad_norm": 0.6700833439826965,
+      "learning_rate": 9.789122577404892e-06,
+      "loss": 0.7696,
+      "step": 3462
+    },
+    {
+      "epoch": 0.19059937255765314,
+      "grad_norm": 0.854340136051178,
+      "learning_rate": 9.78899800113525e-06,
+      "loss": 0.9503,
+      "step": 3463
+    },
+    {
+      "epoch": 0.1906544113600088,
+      "grad_norm": 0.8095537424087524,
+      "learning_rate": 9.78887338887278e-06,
+      "loss": 0.8435,
+      "step": 3464
+    },
+    {
+      "epoch": 0.19070945016236446,
+      "grad_norm": 0.8156480193138123,
+      "learning_rate": 9.78874874061842e-06,
+      "loss": 0.8561,
+      "step": 3465
+    },
+    {
+      "epoch": 0.19076448896472012,
+      "grad_norm": 0.8065482378005981,
+      "learning_rate": 9.788624056373108e-06,
+      "loss": 0.7793,
+      "step": 3466
+    },
+    {
+      "epoch": 0.19081952776707578,
+      "grad_norm": 0.789601743221283,
+      "learning_rate": 9.788499336137778e-06,
+      "loss": 0.7523,
+      "step": 3467
+    },
+    {
+      "epoch": 0.19087456656943144,
+      "grad_norm": 0.8322301506996155,
+      "learning_rate": 9.788374579913369e-06,
+      "loss": 0.9034,
+      "step": 3468
+    },
+    {
+      "epoch": 0.1909296053717871,
+      "grad_norm": 0.8194506764411926,
+      "learning_rate": 9.788249787700818e-06,
+      "loss": 0.8601,
+      "step": 3469
+    },
+    {
+      "epoch": 0.19098464417414276,
+      "grad_norm": 0.8419962525367737,
+      "learning_rate": 9.788124959501065e-06,
+      "loss": 0.869,
+      "step": 3470
+    },
+    {
+      "epoch": 0.19103968297649843,
+      "grad_norm": 0.760637104511261,
+      "learning_rate": 9.788000095315044e-06,
+      "loss": 0.7293,
+      "step": 3471
+    },
+    {
+      "epoch": 0.1910947217788541,
+      "grad_norm": 1.3964574337005615,
+      "learning_rate": 9.787875195143697e-06,
+      "loss": 0.8032,
+      "step": 3472
+    },
+    {
+      "epoch": 0.19114976058120975,
+      "grad_norm": 0.8205012679100037,
+      "learning_rate": 9.787750258987962e-06,
+      "loss": 0.8868,
+      "step": 3473
+    },
+    {
+      "epoch": 0.1912047993835654,
+      "grad_norm": 0.8183104991912842,
+      "learning_rate": 9.78762528684878e-06,
+      "loss": 0.7531,
+      "step": 3474
+    },
+    {
+      "epoch": 0.19125983818592107,
+      "grad_norm": 0.7659775018692017,
+      "learning_rate": 9.787500278727083e-06,
+      "loss": 0.8081,
+      "step": 3475
+    },
+    {
+      "epoch": 0.19131487698827673,
+      "grad_norm": 0.8262091279029846,
+      "learning_rate": 9.787375234623819e-06,
+      "loss": 0.82,
+      "step": 3476
+    },
+    {
+      "epoch": 0.1913699157906324,
+      "grad_norm": 0.857761025428772,
+      "learning_rate": 9.787250154539923e-06,
+      "loss": 0.9133,
+      "step": 3477
+    },
+    {
+      "epoch": 0.19142495459298806,
+      "grad_norm": 0.7551915645599365,
+      "learning_rate": 9.787125038476334e-06,
+      "loss": 0.7822,
+      "step": 3478
+    },
+    {
+      "epoch": 0.19147999339534372,
+      "grad_norm": 0.7777357697486877,
+      "learning_rate": 9.786999886433998e-06,
+      "loss": 0.7676,
+      "step": 3479
+    },
+    {
+      "epoch": 0.19153503219769938,
+      "grad_norm": 0.8389080166816711,
+      "learning_rate": 9.786874698413852e-06,
+      "loss": 0.7901,
+      "step": 3480
+    },
+    {
+      "epoch": 0.19159007100005504,
+      "grad_norm": 0.7894837856292725,
+      "learning_rate": 9.786749474416836e-06,
+      "loss": 0.8393,
+      "step": 3481
+    },
+    {
+      "epoch": 0.1916451098024107,
+      "grad_norm": 1.9752860069274902,
+      "learning_rate": 9.786624214443893e-06,
+      "loss": 0.7611,
+      "step": 3482
+    },
+    {
+      "epoch": 0.19170014860476636,
+      "grad_norm": 0.8023802042007446,
+      "learning_rate": 9.786498918495963e-06,
+      "loss": 0.8426,
+      "step": 3483
+    },
+    {
+      "epoch": 0.19175518740712202,
+      "grad_norm": 0.7232086658477783,
+      "learning_rate": 9.78637358657399e-06,
+      "loss": 0.6611,
+      "step": 3484
+    },
+    {
+      "epoch": 0.19181022620947769,
+      "grad_norm": 0.8198665380477905,
+      "learning_rate": 9.786248218678912e-06,
+      "loss": 0.8795,
+      "step": 3485
+    },
+    {
+      "epoch": 0.19186526501183335,
+      "grad_norm": 0.942404568195343,
+      "learning_rate": 9.786122814811675e-06,
+      "loss": 0.9146,
+      "step": 3486
+    },
+    {
+      "epoch": 0.191920303814189,
+      "grad_norm": 0.7602691054344177,
+      "learning_rate": 9.78599737497322e-06,
+      "loss": 0.7514,
+      "step": 3487
+    },
+    {
+      "epoch": 0.19197534261654467,
+      "grad_norm": 0.7981933951377869,
+      "learning_rate": 9.785871899164489e-06,
+      "loss": 0.7722,
+      "step": 3488
+    },
+    {
+      "epoch": 0.19203038141890033,
+      "grad_norm": 0.8617631793022156,
+      "learning_rate": 9.785746387386427e-06,
+      "loss": 0.8989,
+      "step": 3489
+    },
+    {
+      "epoch": 0.192085420221256,
+      "grad_norm": 0.7691803574562073,
+      "learning_rate": 9.785620839639976e-06,
+      "loss": 0.7929,
+      "step": 3490
+    },
+    {
+      "epoch": 0.19214045902361165,
+      "grad_norm": 1.3053189516067505,
+      "learning_rate": 9.785495255926078e-06,
+      "loss": 0.8478,
+      "step": 3491
+    },
+    {
+      "epoch": 0.19219549782596732,
+      "grad_norm": 0.807064950466156,
+      "learning_rate": 9.785369636245681e-06,
+      "loss": 0.7452,
+      "step": 3492
+    },
+    {
+      "epoch": 0.19225053662832298,
+      "grad_norm": 0.8182778358459473,
+      "learning_rate": 9.785243980599726e-06,
+      "loss": 0.8371,
+      "step": 3493
+    },
+    {
+      "epoch": 0.19230557543067864,
+      "grad_norm": 0.7654449343681335,
+      "learning_rate": 9.785118288989157e-06,
+      "loss": 0.8321,
+      "step": 3494
+    },
+    {
+      "epoch": 0.1923606142330343,
+      "grad_norm": 0.7192448973655701,
+      "learning_rate": 9.784992561414922e-06,
+      "loss": 0.7451,
+      "step": 3495
+    },
+    {
+      "epoch": 0.19241565303538996,
+      "grad_norm": 0.8639407753944397,
+      "learning_rate": 9.784866797877964e-06,
+      "loss": 0.9272,
+      "step": 3496
+    },
+    {
+      "epoch": 0.19247069183774562,
+      "grad_norm": 0.8329927921295166,
+      "learning_rate": 9.784740998379225e-06,
+      "loss": 0.8034,
+      "step": 3497
+    },
+    {
+      "epoch": 0.19252573064010128,
+      "grad_norm": 0.7975476980209351,
+      "learning_rate": 9.784615162919656e-06,
+      "loss": 0.6885,
+      "step": 3498
+    },
+    {
+      "epoch": 0.19258076944245694,
+      "grad_norm": 0.8077559471130371,
+      "learning_rate": 9.7844892915002e-06,
+      "loss": 0.8745,
+      "step": 3499
+    },
+    {
+      "epoch": 0.1926358082448126,
+      "grad_norm": 0.7957825660705566,
+      "learning_rate": 9.7843633841218e-06,
+      "loss": 0.7612,
+      "step": 3500
+    },
+    {
+      "epoch": 0.19269084704716824,
+      "grad_norm": 0.8478250503540039,
+      "learning_rate": 9.784237440785408e-06,
+      "loss": 0.8675,
+      "step": 3501
+    },
+    {
+      "epoch": 0.1927458858495239,
+      "grad_norm": 0.7289726138114929,
+      "learning_rate": 9.78411146149197e-06,
+      "loss": 0.7126,
+      "step": 3502
+    },
+    {
+      "epoch": 0.19280092465187956,
+      "grad_norm": 0.7608509063720703,
+      "learning_rate": 9.783985446242427e-06,
+      "loss": 0.7049,
+      "step": 3503
+    },
+    {
+      "epoch": 0.19285596345423522,
+      "grad_norm": 0.8985201120376587,
+      "learning_rate": 9.783859395037733e-06,
+      "loss": 0.8067,
+      "step": 3504
+    },
+    {
+      "epoch": 0.19291100225659089,
+      "grad_norm": 0.7563273906707764,
+      "learning_rate": 9.78373330787883e-06,
+      "loss": 0.7018,
+      "step": 3505
+    },
+    {
+      "epoch": 0.19296604105894655,
+      "grad_norm": 0.8022900223731995,
+      "learning_rate": 9.78360718476667e-06,
+      "loss": 0.8346,
+      "step": 3506
+    },
+    {
+      "epoch": 0.1930210798613022,
+      "grad_norm": 0.897566020488739,
+      "learning_rate": 9.783481025702197e-06,
+      "loss": 0.9465,
+      "step": 3507
+    },
+    {
+      "epoch": 0.19307611866365787,
+      "grad_norm": 0.9550303220748901,
+      "learning_rate": 9.783354830686363e-06,
+      "loss": 0.8904,
+      "step": 3508
+    },
+    {
+      "epoch": 0.19313115746601353,
+      "grad_norm": 0.8152582049369812,
+      "learning_rate": 9.783228599720114e-06,
+      "loss": 0.7776,
+      "step": 3509
+    },
+    {
+      "epoch": 0.1931861962683692,
+      "grad_norm": 0.7421940565109253,
+      "learning_rate": 9.783102332804398e-06,
+      "loss": 0.6847,
+      "step": 3510
+    },
+    {
+      "epoch": 0.19324123507072485,
+      "grad_norm": 0.7414368391036987,
+      "learning_rate": 9.782976029940167e-06,
+      "loss": 0.8435,
+      "step": 3511
+    },
+    {
+      "epoch": 0.19329627387308052,
+      "grad_norm": 0.7845529317855835,
+      "learning_rate": 9.782849691128366e-06,
+      "loss": 0.8255,
+      "step": 3512
+    },
+    {
+      "epoch": 0.19335131267543618,
+      "grad_norm": 0.7779788970947266,
+      "learning_rate": 9.78272331636995e-06,
+      "loss": 0.7801,
+      "step": 3513
+    },
+    {
+      "epoch": 0.19340635147779184,
+      "grad_norm": 0.7537885904312134,
+      "learning_rate": 9.782596905665865e-06,
+      "loss": 0.7501,
+      "step": 3514
+    },
+    {
+      "epoch": 0.1934613902801475,
+      "grad_norm": 0.7585812211036682,
+      "learning_rate": 9.782470459017059e-06,
+      "loss": 0.8425,
+      "step": 3515
+    },
+    {
+      "epoch": 0.19351642908250316,
+      "grad_norm": 0.7923589944839478,
+      "learning_rate": 9.78234397642449e-06,
+      "loss": 0.8412,
+      "step": 3516
+    },
+    {
+      "epoch": 0.19357146788485882,
+      "grad_norm": 0.8710628151893616,
+      "learning_rate": 9.7822174578891e-06,
+      "loss": 0.8014,
+      "step": 3517
+    },
+    {
+      "epoch": 0.19362650668721448,
+      "grad_norm": 0.7646920084953308,
+      "learning_rate": 9.782090903411845e-06,
+      "loss": 0.8256,
+      "step": 3518
+    },
+    {
+      "epoch": 0.19368154548957014,
+      "grad_norm": 0.7560480833053589,
+      "learning_rate": 9.781964312993675e-06,
+      "loss": 0.7816,
+      "step": 3519
+    },
+    {
+      "epoch": 0.1937365842919258,
+      "grad_norm": 0.7438123226165771,
+      "learning_rate": 9.78183768663554e-06,
+      "loss": 0.8319,
+      "step": 3520
+    },
+    {
+      "epoch": 0.19379162309428147,
+      "grad_norm": 0.7239874601364136,
+      "learning_rate": 9.781711024338394e-06,
+      "loss": 0.6968,
+      "step": 3521
+    },
+    {
+      "epoch": 0.19384666189663713,
+      "grad_norm": 0.881197988986969,
+      "learning_rate": 9.781584326103188e-06,
+      "loss": 0.9493,
+      "step": 3522
+    },
+    {
+      "epoch": 0.1939017006989928,
+      "grad_norm": 0.7903854846954346,
+      "learning_rate": 9.781457591930874e-06,
+      "loss": 0.8312,
+      "step": 3523
+    },
+    {
+      "epoch": 0.19395673950134845,
+      "grad_norm": 0.7375456094741821,
+      "learning_rate": 9.781330821822405e-06,
+      "loss": 0.7434,
+      "step": 3524
+    },
+    {
+      "epoch": 0.1940117783037041,
+      "grad_norm": 0.7101724743843079,
+      "learning_rate": 9.781204015778733e-06,
+      "loss": 0.75,
+      "step": 3525
+    },
+    {
+      "epoch": 0.19406681710605977,
+      "grad_norm": 0.8267471194267273,
+      "learning_rate": 9.781077173800812e-06,
+      "loss": 0.8807,
+      "step": 3526
+    },
+    {
+      "epoch": 0.19412185590841544,
+      "grad_norm": 0.9014178514480591,
+      "learning_rate": 9.780950295889594e-06,
+      "loss": 0.7836,
+      "step": 3527
+    },
+    {
+      "epoch": 0.1941768947107711,
+      "grad_norm": 0.7579739689826965,
+      "learning_rate": 9.780823382046034e-06,
+      "loss": 0.8331,
+      "step": 3528
+    },
+    {
+      "epoch": 0.19423193351312676,
+      "grad_norm": 0.8308925032615662,
+      "learning_rate": 9.780696432271084e-06,
+      "loss": 0.794,
+      "step": 3529
+    },
+    {
+      "epoch": 0.19428697231548242,
+      "grad_norm": 0.7461574673652649,
+      "learning_rate": 9.780569446565701e-06,
+      "loss": 0.8155,
+      "step": 3530
+    },
+    {
+      "epoch": 0.19434201111783808,
+      "grad_norm": 0.8658885359764099,
+      "learning_rate": 9.780442424930836e-06,
+      "loss": 0.7907,
+      "step": 3531
+    },
+    {
+      "epoch": 0.19439704992019374,
+      "grad_norm": 0.7243279218673706,
+      "learning_rate": 9.780315367367449e-06,
+      "loss": 0.7985,
+      "step": 3532
+    },
+    {
+      "epoch": 0.1944520887225494,
+      "grad_norm": 0.8482224345207214,
+      "learning_rate": 9.780188273876486e-06,
+      "loss": 0.9095,
+      "step": 3533
+    },
+    {
+      "epoch": 0.19450712752490507,
+      "grad_norm": 0.8675364255905151,
+      "learning_rate": 9.78006114445891e-06,
+      "loss": 0.759,
+      "step": 3534
+    },
+    {
+      "epoch": 0.19456216632726073,
+      "grad_norm": 0.8388474583625793,
+      "learning_rate": 9.779933979115675e-06,
+      "loss": 0.8331,
+      "step": 3535
+    },
+    {
+      "epoch": 0.1946172051296164,
+      "grad_norm": 0.8050872683525085,
+      "learning_rate": 9.779806777847735e-06,
+      "loss": 0.861,
+      "step": 3536
+    },
+    {
+      "epoch": 0.19467224393197205,
+      "grad_norm": 0.8401390910148621,
+      "learning_rate": 9.779679540656046e-06,
+      "loss": 0.755,
+      "step": 3537
+    },
+    {
+      "epoch": 0.1947272827343277,
+      "grad_norm": 0.865160346031189,
+      "learning_rate": 9.779552267541566e-06,
+      "loss": 0.7515,
+      "step": 3538
+    },
+    {
+      "epoch": 0.19478232153668337,
+      "grad_norm": 0.923086941242218,
+      "learning_rate": 9.77942495850525e-06,
+      "loss": 0.8032,
+      "step": 3539
+    },
+    {
+      "epoch": 0.19483736033903903,
+      "grad_norm": 0.8402467966079712,
+      "learning_rate": 9.779297613548056e-06,
+      "loss": 0.9198,
+      "step": 3540
+    },
+    {
+      "epoch": 0.1948923991413947,
+      "grad_norm": 0.7875306606292725,
+      "learning_rate": 9.779170232670939e-06,
+      "loss": 0.712,
+      "step": 3541
+    },
+    {
+      "epoch": 0.19494743794375036,
+      "grad_norm": 0.7996379137039185,
+      "learning_rate": 9.779042815874858e-06,
+      "loss": 0.8126,
+      "step": 3542
+    },
+    {
+      "epoch": 0.19500247674610602,
+      "grad_norm": 0.7644525766372681,
+      "learning_rate": 9.778915363160773e-06,
+      "loss": 0.8602,
+      "step": 3543
+    },
+    {
+      "epoch": 0.19505751554846165,
+      "grad_norm": 0.8068630695343018,
+      "learning_rate": 9.778787874529635e-06,
+      "loss": 0.736,
+      "step": 3544
+    },
+    {
+      "epoch": 0.1951125543508173,
+      "grad_norm": 0.7889519929885864,
+      "learning_rate": 9.77866034998241e-06,
+      "loss": 0.755,
+      "step": 3545
+    },
+    {
+      "epoch": 0.19516759315317297,
+      "grad_norm": 0.7895978689193726,
+      "learning_rate": 9.778532789520053e-06,
+      "loss": 0.8213,
+      "step": 3546
+    },
+    {
+      "epoch": 0.19522263195552864,
+      "grad_norm": 0.8571796417236328,
+      "learning_rate": 9.77840519314352e-06,
+      "loss": 0.8193,
+      "step": 3547
+    },
+    {
+      "epoch": 0.1952776707578843,
+      "grad_norm": 0.6880007982254028,
+      "learning_rate": 9.778277560853775e-06,
+      "loss": 0.6354,
+      "step": 3548
+    },
+    {
+      "epoch": 0.19533270956023996,
+      "grad_norm": 0.8155353665351868,
+      "learning_rate": 9.778149892651775e-06,
+      "loss": 0.8518,
+      "step": 3549
+    },
+    {
+      "epoch": 0.19538774836259562,
+      "grad_norm": 0.851021945476532,
+      "learning_rate": 9.778022188538479e-06,
+      "loss": 0.8506,
+      "step": 3550
+    },
+    {
+      "epoch": 0.19544278716495128,
+      "grad_norm": 0.8910510540008545,
+      "learning_rate": 9.777894448514847e-06,
+      "loss": 0.8825,
+      "step": 3551
+    },
+    {
+      "epoch": 0.19549782596730694,
+      "grad_norm": 0.8156018853187561,
+      "learning_rate": 9.777766672581838e-06,
+      "loss": 0.8262,
+      "step": 3552
+    },
+    {
+      "epoch": 0.1955528647696626,
+      "grad_norm": 0.756340503692627,
+      "learning_rate": 9.777638860740415e-06,
+      "loss": 0.7094,
+      "step": 3553
+    },
+    {
+      "epoch": 0.19560790357201827,
+      "grad_norm": 0.7604243159294128,
+      "learning_rate": 9.777511012991538e-06,
+      "loss": 0.8089,
+      "step": 3554
+    },
+    {
+      "epoch": 0.19566294237437393,
+      "grad_norm": 0.7609277963638306,
+      "learning_rate": 9.777383129336167e-06,
+      "loss": 0.7853,
+      "step": 3555
+    },
+    {
+      "epoch": 0.1957179811767296,
+      "grad_norm": 1.3562177419662476,
+      "learning_rate": 9.77725520977526e-06,
+      "loss": 0.7051,
+      "step": 3556
+    },
+    {
+      "epoch": 0.19577301997908525,
+      "grad_norm": 0.7428582310676575,
+      "learning_rate": 9.777127254309784e-06,
+      "loss": 0.734,
+      "step": 3557
+    },
+    {
+      "epoch": 0.1958280587814409,
+      "grad_norm": 0.6941032409667969,
+      "learning_rate": 9.776999262940698e-06,
+      "loss": 0.7862,
+      "step": 3558
+    },
+    {
+      "epoch": 0.19588309758379657,
+      "grad_norm": 0.8249906301498413,
+      "learning_rate": 9.776871235668966e-06,
+      "loss": 0.8324,
+      "step": 3559
+    },
+    {
+      "epoch": 0.19593813638615223,
+      "grad_norm": 0.6778795719146729,
+      "learning_rate": 9.776743172495546e-06,
+      "loss": 0.743,
+      "step": 3560
+    },
+    {
+      "epoch": 0.1959931751885079,
+      "grad_norm": 0.8454411625862122,
+      "learning_rate": 9.776615073421405e-06,
+      "loss": 0.8625,
+      "step": 3561
+    },
+    {
+      "epoch": 0.19604821399086356,
+      "grad_norm": 0.8303809762001038,
+      "learning_rate": 9.776486938447503e-06,
+      "loss": 0.8806,
+      "step": 3562
+    },
+    {
+      "epoch": 0.19610325279321922,
+      "grad_norm": 0.8814080357551575,
+      "learning_rate": 9.776358767574803e-06,
+      "loss": 0.9096,
+      "step": 3563
+    },
+    {
+      "epoch": 0.19615829159557488,
+      "grad_norm": 0.7860022187232971,
+      "learning_rate": 9.77623056080427e-06,
+      "loss": 0.8101,
+      "step": 3564
+    },
+    {
+      "epoch": 0.19621333039793054,
+      "grad_norm": 0.7604898810386658,
+      "learning_rate": 9.776102318136866e-06,
+      "loss": 0.8121,
+      "step": 3565
+    },
+    {
+      "epoch": 0.1962683692002862,
+      "grad_norm": 0.810708224773407,
+      "learning_rate": 9.775974039573555e-06,
+      "loss": 0.8334,
+      "step": 3566
+    },
+    {
+      "epoch": 0.19632340800264186,
+      "grad_norm": 1.0174707174301147,
+      "learning_rate": 9.775845725115301e-06,
+      "loss": 0.8147,
+      "step": 3567
+    },
+    {
+      "epoch": 0.19637844680499753,
+      "grad_norm": 0.825137734413147,
+      "learning_rate": 9.77571737476307e-06,
+      "loss": 0.816,
+      "step": 3568
+    },
+    {
+      "epoch": 0.1964334856073532,
+      "grad_norm": 0.9023691415786743,
+      "learning_rate": 9.775588988517826e-06,
+      "loss": 0.9157,
+      "step": 3569
+    },
+    {
+      "epoch": 0.19648852440970885,
+      "grad_norm": 0.7287655472755432,
+      "learning_rate": 9.775460566380534e-06,
+      "loss": 0.7414,
+      "step": 3570
+    },
+    {
+      "epoch": 0.1965435632120645,
+      "grad_norm": 0.8675361275672913,
+      "learning_rate": 9.775332108352158e-06,
+      "loss": 0.7212,
+      "step": 3571
+    },
+    {
+      "epoch": 0.19659860201442017,
+      "grad_norm": 0.8633139729499817,
+      "learning_rate": 9.775203614433664e-06,
+      "loss": 0.7254,
+      "step": 3572
+    },
+    {
+      "epoch": 0.19665364081677583,
+      "grad_norm": 0.8628275394439697,
+      "learning_rate": 9.775075084626017e-06,
+      "loss": 0.7403,
+      "step": 3573
+    },
+    {
+      "epoch": 0.1967086796191315,
+      "grad_norm": 0.86918044090271,
+      "learning_rate": 9.774946518930184e-06,
+      "loss": 0.8208,
+      "step": 3574
+    },
+    {
+      "epoch": 0.19676371842148715,
+      "grad_norm": 1.3616218566894531,
+      "learning_rate": 9.774817917347132e-06,
+      "loss": 0.7432,
+      "step": 3575
+    },
+    {
+      "epoch": 0.19681875722384282,
+      "grad_norm": 0.929084062576294,
+      "learning_rate": 9.774689279877827e-06,
+      "loss": 0.9567,
+      "step": 3576
+    },
+    {
+      "epoch": 0.19687379602619848,
+      "grad_norm": 0.7732542753219604,
+      "learning_rate": 9.774560606523234e-06,
+      "loss": 0.8682,
+      "step": 3577
+    },
+    {
+      "epoch": 0.19692883482855414,
+      "grad_norm": 0.7933471202850342,
+      "learning_rate": 9.774431897284323e-06,
+      "loss": 0.7112,
+      "step": 3578
+    },
+    {
+      "epoch": 0.1969838736309098,
+      "grad_norm": 0.8229583501815796,
+      "learning_rate": 9.77430315216206e-06,
+      "loss": 0.762,
+      "step": 3579
+    },
+    {
+      "epoch": 0.19703891243326546,
+      "grad_norm": 0.7571341395378113,
+      "learning_rate": 9.774174371157412e-06,
+      "loss": 0.7627,
+      "step": 3580
+    },
+    {
+      "epoch": 0.19709395123562112,
+      "grad_norm": 1.1551839113235474,
+      "learning_rate": 9.774045554271347e-06,
+      "loss": 0.8621,
+      "step": 3581
+    },
+    {
+      "epoch": 0.19714899003797678,
+      "grad_norm": 0.8546237349510193,
+      "learning_rate": 9.773916701504833e-06,
+      "loss": 0.8183,
+      "step": 3582
+    },
+    {
+      "epoch": 0.19720402884033245,
+      "grad_norm": 0.7297555804252625,
+      "learning_rate": 9.773787812858841e-06,
+      "loss": 0.8098,
+      "step": 3583
+    },
+    {
+      "epoch": 0.1972590676426881,
+      "grad_norm": 0.7846053838729858,
+      "learning_rate": 9.773658888334336e-06,
+      "loss": 0.7874,
+      "step": 3584
+    },
+    {
+      "epoch": 0.19731410644504377,
+      "grad_norm": 0.8949562907218933,
+      "learning_rate": 9.773529927932288e-06,
+      "loss": 0.8651,
+      "step": 3585
+    },
+    {
+      "epoch": 0.19736914524739943,
+      "grad_norm": 0.8041829466819763,
+      "learning_rate": 9.773400931653668e-06,
+      "loss": 0.7519,
+      "step": 3586
+    },
+    {
+      "epoch": 0.19742418404975506,
+      "grad_norm": 0.8090983033180237,
+      "learning_rate": 9.773271899499444e-06,
+      "loss": 0.8606,
+      "step": 3587
+    },
+    {
+      "epoch": 0.19747922285211073,
+      "grad_norm": 0.7954100966453552,
+      "learning_rate": 9.773142831470587e-06,
+      "loss": 0.9028,
+      "step": 3588
+    },
+    {
+      "epoch": 0.1975342616544664,
+      "grad_norm": 0.6865562796592712,
+      "learning_rate": 9.773013727568066e-06,
+      "loss": 0.7323,
+      "step": 3589
+    },
+    {
+      "epoch": 0.19758930045682205,
+      "grad_norm": 0.9144858717918396,
+      "learning_rate": 9.772884587792851e-06,
+      "loss": 0.8178,
+      "step": 3590
+    },
+    {
+      "epoch": 0.1976443392591777,
+      "grad_norm": 0.8096563220024109,
+      "learning_rate": 9.772755412145913e-06,
+      "loss": 0.7749,
+      "step": 3591
+    },
+    {
+      "epoch": 0.19769937806153337,
+      "grad_norm": 1.4496957063674927,
+      "learning_rate": 9.772626200628222e-06,
+      "loss": 0.7981,
+      "step": 3592
+    },
+    {
+      "epoch": 0.19775441686388903,
+      "grad_norm": 0.7699438333511353,
+      "learning_rate": 9.77249695324075e-06,
+      "loss": 0.7683,
+      "step": 3593
+    },
+    {
+      "epoch": 0.1978094556662447,
+      "grad_norm": 0.7883017063140869,
+      "learning_rate": 9.77236766998447e-06,
+      "loss": 0.7668,
+      "step": 3594
+    },
+    {
+      "epoch": 0.19786449446860035,
+      "grad_norm": 0.7552568912506104,
+      "learning_rate": 9.772238350860352e-06,
+      "loss": 0.7914,
+      "step": 3595
+    },
+    {
+      "epoch": 0.19791953327095602,
+      "grad_norm": 0.8585009574890137,
+      "learning_rate": 9.772108995869366e-06,
+      "loss": 0.9888,
+      "step": 3596
+    },
+    {
+      "epoch": 0.19797457207331168,
+      "grad_norm": 0.9459839463233948,
+      "learning_rate": 9.77197960501249e-06,
+      "loss": 0.9923,
+      "step": 3597
+    },
+    {
+      "epoch": 0.19802961087566734,
+      "grad_norm": 0.844771683216095,
+      "learning_rate": 9.77185017829069e-06,
+      "loss": 0.8427,
+      "step": 3598
+    },
+    {
+      "epoch": 0.198084649678023,
+      "grad_norm": 0.749700665473938,
+      "learning_rate": 9.77172071570494e-06,
+      "loss": 0.8111,
+      "step": 3599
+    },
+    {
+      "epoch": 0.19813968848037866,
+      "grad_norm": 0.7297450304031372,
+      "learning_rate": 9.771591217256216e-06,
+      "loss": 0.7783,
+      "step": 3600
+    },
+    {
+      "epoch": 0.19819472728273432,
+      "grad_norm": 0.7928450703620911,
+      "learning_rate": 9.77146168294549e-06,
+      "loss": 0.8755,
+      "step": 3601
+    },
+    {
+      "epoch": 0.19824976608508998,
+      "grad_norm": 0.7236143946647644,
+      "learning_rate": 9.771332112773734e-06,
+      "loss": 0.7159,
+      "step": 3602
+    },
+    {
+      "epoch": 0.19830480488744565,
+      "grad_norm": 0.8170965313911438,
+      "learning_rate": 9.771202506741926e-06,
+      "loss": 0.9093,
+      "step": 3603
+    },
+    {
+      "epoch": 0.1983598436898013,
+      "grad_norm": 0.8834578990936279,
+      "learning_rate": 9.771072864851035e-06,
+      "loss": 0.8961,
+      "step": 3604
+    },
+    {
+      "epoch": 0.19841488249215697,
+      "grad_norm": 1.3750289678573608,
+      "learning_rate": 9.770943187102037e-06,
+      "loss": 0.8175,
+      "step": 3605
+    },
+    {
+      "epoch": 0.19846992129451263,
+      "grad_norm": 0.7016286253929138,
+      "learning_rate": 9.770813473495909e-06,
+      "loss": 0.7171,
+      "step": 3606
+    },
+    {
+      "epoch": 0.1985249600968683,
+      "grad_norm": 0.7792307734489441,
+      "learning_rate": 9.770683724033622e-06,
+      "loss": 0.6892,
+      "step": 3607
+    },
+    {
+      "epoch": 0.19857999889922395,
+      "grad_norm": 0.789820671081543,
+      "learning_rate": 9.770553938716153e-06,
+      "loss": 0.8531,
+      "step": 3608
+    },
+    {
+      "epoch": 0.19863503770157961,
+      "grad_norm": 0.7585997581481934,
+      "learning_rate": 9.77042411754448e-06,
+      "loss": 0.8195,
+      "step": 3609
+    },
+    {
+      "epoch": 0.19869007650393528,
+      "grad_norm": 0.8989273905754089,
+      "learning_rate": 9.770294260519573e-06,
+      "loss": 0.891,
+      "step": 3610
+    },
+    {
+      "epoch": 0.19874511530629094,
+      "grad_norm": 0.8044012188911438,
+      "learning_rate": 9.770164367642414e-06,
+      "loss": 0.8428,
+      "step": 3611
+    },
+    {
+      "epoch": 0.1988001541086466,
+      "grad_norm": 0.7847021222114563,
+      "learning_rate": 9.770034438913975e-06,
+      "loss": 0.8302,
+      "step": 3612
+    },
+    {
+      "epoch": 0.19885519291100226,
+      "grad_norm": 0.9260531663894653,
+      "learning_rate": 9.769904474335234e-06,
+      "loss": 0.8187,
+      "step": 3613
+    },
+    {
+      "epoch": 0.19891023171335792,
+      "grad_norm": 0.7491805553436279,
+      "learning_rate": 9.769774473907168e-06,
+      "loss": 0.8374,
+      "step": 3614
+    },
+    {
+      "epoch": 0.19896527051571358,
+      "grad_norm": 1.1665992736816406,
+      "learning_rate": 9.769644437630754e-06,
+      "loss": 0.8154,
+      "step": 3615
+    },
+    {
+      "epoch": 0.19902030931806924,
+      "grad_norm": 0.9162279963493347,
+      "learning_rate": 9.769514365506968e-06,
+      "loss": 0.8883,
+      "step": 3616
+    },
+    {
+      "epoch": 0.1990753481204249,
+      "grad_norm": 0.8980437517166138,
+      "learning_rate": 9.769384257536791e-06,
+      "loss": 0.8948,
+      "step": 3617
+    },
+    {
+      "epoch": 0.19913038692278057,
+      "grad_norm": 0.7544137835502625,
+      "learning_rate": 9.769254113721197e-06,
+      "loss": 0.7763,
+      "step": 3618
+    },
+    {
+      "epoch": 0.19918542572513623,
+      "grad_norm": 0.8393334746360779,
+      "learning_rate": 9.769123934061168e-06,
+      "loss": 0.8361,
+      "step": 3619
+    },
+    {
+      "epoch": 0.1992404645274919,
+      "grad_norm": 0.8184031248092651,
+      "learning_rate": 9.768993718557678e-06,
+      "loss": 0.8104,
+      "step": 3620
+    },
+    {
+      "epoch": 0.19929550332984755,
+      "grad_norm": 0.8023706674575806,
+      "learning_rate": 9.76886346721171e-06,
+      "loss": 0.7824,
+      "step": 3621
+    },
+    {
+      "epoch": 0.1993505421322032,
+      "grad_norm": 0.9354264736175537,
+      "learning_rate": 9.768733180024238e-06,
+      "loss": 0.7782,
+      "step": 3622
+    },
+    {
+      "epoch": 0.19940558093455887,
+      "grad_norm": 0.7037177681922913,
+      "learning_rate": 9.768602856996244e-06,
+      "loss": 0.8054,
+      "step": 3623
+    },
+    {
+      "epoch": 0.19946061973691454,
+      "grad_norm": 0.7926928997039795,
+      "learning_rate": 9.768472498128709e-06,
+      "loss": 0.8864,
+      "step": 3624
+    },
+    {
+      "epoch": 0.1995156585392702,
+      "grad_norm": 0.7963769435882568,
+      "learning_rate": 9.76834210342261e-06,
+      "loss": 0.8505,
+      "step": 3625
+    },
+    {
+      "epoch": 0.19957069734162586,
+      "grad_norm": 0.8553926944732666,
+      "learning_rate": 9.768211672878929e-06,
+      "loss": 0.8519,
+      "step": 3626
+    },
+    {
+      "epoch": 0.19962573614398152,
+      "grad_norm": 0.8147156834602356,
+      "learning_rate": 9.768081206498644e-06,
+      "loss": 0.8091,
+      "step": 3627
+    },
+    {
+      "epoch": 0.19968077494633718,
+      "grad_norm": 0.8226443529129028,
+      "learning_rate": 9.767950704282739e-06,
+      "loss": 0.8561,
+      "step": 3628
+    },
+    {
+      "epoch": 0.19973581374869284,
+      "grad_norm": 0.7246909141540527,
+      "learning_rate": 9.76782016623219e-06,
+      "loss": 0.7318,
+      "step": 3629
+    },
+    {
+      "epoch": 0.19979085255104848,
+      "grad_norm": 1.0527293682098389,
+      "learning_rate": 9.767689592347983e-06,
+      "loss": 0.7699,
+      "step": 3630
+    },
+    {
+      "epoch": 0.19984589135340414,
+      "grad_norm": 0.7433847188949585,
+      "learning_rate": 9.767558982631097e-06,
+      "loss": 0.8619,
+      "step": 3631
+    },
+    {
+      "epoch": 0.1999009301557598,
+      "grad_norm": 0.7901468873023987,
+      "learning_rate": 9.767428337082513e-06,
+      "loss": 0.8365,
+      "step": 3632
+    },
+    {
+      "epoch": 0.19995596895811546,
+      "grad_norm": 0.7766845226287842,
+      "learning_rate": 9.767297655703215e-06,
+      "loss": 0.7767,
+      "step": 3633
+    },
+    {
+      "epoch": 0.20001100776047112,
+      "grad_norm": 0.7785109281539917,
+      "learning_rate": 9.767166938494183e-06,
+      "loss": 0.7114,
+      "step": 3634
+    },
+    {
+      "epoch": 0.20006604656282678,
+      "grad_norm": 0.8068187832832336,
+      "learning_rate": 9.767036185456402e-06,
+      "loss": 0.8142,
+      "step": 3635
+    },
+    {
+      "epoch": 0.20012108536518244,
+      "grad_norm": 0.7893292307853699,
+      "learning_rate": 9.766905396590851e-06,
+      "loss": 0.8658,
+      "step": 3636
     }
   ],
   "logging_steps": 1,
@@ -19115,7 +25478,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 8.047544142365983e+18,
+  "total_flos": 1.0730058856487977e+19,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null