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": 0.4379225568531039,
+  "epoch": 0.53779963122311,
   "eval_steps": 500,
-  "global_step": 5700,
+  "global_step": 7000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -19957,6 +19957,4556 @@
       "learning_rate": 1.8248309772587588e-05,
       "loss": 1.3252,
       "step": 5700
+    },
+    {
+      "epoch": 0.4380762138905962,
+      "grad_norm": 2.912013530731201,
+      "learning_rate": 1.8247695144437614e-05,
+      "loss": 1.3125,
+      "step": 5702
+    },
+    {
+      "epoch": 0.4382298709280885,
+      "grad_norm": 3.1190555095672607,
+      "learning_rate": 1.8247080516287648e-05,
+      "loss": 1.5832,
+      "step": 5704
+    },
+    {
+      "epoch": 0.43838352796558083,
+      "grad_norm": 3.2913010120391846,
+      "learning_rate": 1.8246465888137677e-05,
+      "loss": 1.4592,
+      "step": 5706
+    },
+    {
+      "epoch": 0.43853718500307315,
+      "grad_norm": 3.0158445835113525,
+      "learning_rate": 1.8245851259987707e-05,
+      "loss": 1.258,
+      "step": 5708
+    },
+    {
+      "epoch": 0.43869084204056547,
+      "grad_norm": 2.7523088455200195,
+      "learning_rate": 1.824523663183774e-05,
+      "loss": 1.3033,
+      "step": 5710
+    },
+    {
+      "epoch": 0.4388444990780578,
+      "grad_norm": 3.2428712844848633,
+      "learning_rate": 1.824462200368777e-05,
+      "loss": 1.5227,
+      "step": 5712
+    },
+    {
+      "epoch": 0.4389981561155501,
+      "grad_norm": 3.1375772953033447,
+      "learning_rate": 1.8244007375537803e-05,
+      "loss": 1.2947,
+      "step": 5714
+    },
+    {
+      "epoch": 0.4391518131530424,
+      "grad_norm": 3.0907890796661377,
+      "learning_rate": 1.8243392747387832e-05,
+      "loss": 1.5986,
+      "step": 5716
+    },
+    {
+      "epoch": 0.43930547019053473,
+      "grad_norm": 2.9191091060638428,
+      "learning_rate": 1.8242778119237862e-05,
+      "loss": 1.3305,
+      "step": 5718
+    },
+    {
+      "epoch": 0.43945912722802705,
+      "grad_norm": 2.765789747238159,
+      "learning_rate": 1.8242163491087895e-05,
+      "loss": 1.3996,
+      "step": 5720
+    },
+    {
+      "epoch": 0.43961278426551936,
+      "grad_norm": 3.196028470993042,
+      "learning_rate": 1.8241548862937925e-05,
+      "loss": 1.4643,
+      "step": 5722
+    },
+    {
+      "epoch": 0.4397664413030117,
+      "grad_norm": 3.159723997116089,
+      "learning_rate": 1.8240934234787955e-05,
+      "loss": 1.2982,
+      "step": 5724
+    },
+    {
+      "epoch": 0.439920098340504,
+      "grad_norm": 4.066509246826172,
+      "learning_rate": 1.8240319606637988e-05,
+      "loss": 1.4798,
+      "step": 5726
+    },
+    {
+      "epoch": 0.4400737553779963,
+      "grad_norm": 3.310293436050415,
+      "learning_rate": 1.8239704978488014e-05,
+      "loss": 1.3133,
+      "step": 5728
+    },
+    {
+      "epoch": 0.44022741241548863,
+      "grad_norm": 3.166078805923462,
+      "learning_rate": 1.8239090350338047e-05,
+      "loss": 1.3163,
+      "step": 5730
+    },
+    {
+      "epoch": 0.44038106945298094,
+      "grad_norm": 3.0429303646087646,
+      "learning_rate": 1.8238475722188077e-05,
+      "loss": 1.3178,
+      "step": 5732
+    },
+    {
+      "epoch": 0.44053472649047326,
+      "grad_norm": 2.984344959259033,
+      "learning_rate": 1.8237861094038106e-05,
+      "loss": 1.4449,
+      "step": 5734
+    },
+    {
+      "epoch": 0.4406883835279656,
+      "grad_norm": 3.379456043243408,
+      "learning_rate": 1.823724646588814e-05,
+      "loss": 1.3448,
+      "step": 5736
+    },
+    {
+      "epoch": 0.4408420405654579,
+      "grad_norm": 3.0715973377227783,
+      "learning_rate": 1.823663183773817e-05,
+      "loss": 1.4384,
+      "step": 5738
+    },
+    {
+      "epoch": 0.4409956976029502,
+      "grad_norm": 3.0291173458099365,
+      "learning_rate": 1.8236017209588202e-05,
+      "loss": 1.452,
+      "step": 5740
+    },
+    {
+      "epoch": 0.4411493546404425,
+      "grad_norm": 3.0475244522094727,
+      "learning_rate": 1.8235402581438232e-05,
+      "loss": 1.3113,
+      "step": 5742
+    },
+    {
+      "epoch": 0.44130301167793484,
+      "grad_norm": 3.0464026927948,
+      "learning_rate": 1.823478795328826e-05,
+      "loss": 1.2598,
+      "step": 5744
+    },
+    {
+      "epoch": 0.44145666871542716,
+      "grad_norm": 3.1016945838928223,
+      "learning_rate": 1.8234173325138295e-05,
+      "loss": 1.396,
+      "step": 5746
+    },
+    {
+      "epoch": 0.4416103257529195,
+      "grad_norm": 3.356527328491211,
+      "learning_rate": 1.8233558696988324e-05,
+      "loss": 1.4148,
+      "step": 5748
+    },
+    {
+      "epoch": 0.4417639827904118,
+      "grad_norm": 2.8270328044891357,
+      "learning_rate": 1.8232944068838354e-05,
+      "loss": 1.346,
+      "step": 5750
+    },
+    {
+      "epoch": 0.4419176398279041,
+      "grad_norm": 3.364140510559082,
+      "learning_rate": 1.8232329440688387e-05,
+      "loss": 1.3203,
+      "step": 5752
+    },
+    {
+      "epoch": 0.4420712968653964,
+      "grad_norm": 3.246849298477173,
+      "learning_rate": 1.8231714812538413e-05,
+      "loss": 1.5842,
+      "step": 5754
+    },
+    {
+      "epoch": 0.44222495390288874,
+      "grad_norm": 3.0349695682525635,
+      "learning_rate": 1.8231100184388446e-05,
+      "loss": 1.3516,
+      "step": 5756
+    },
+    {
+      "epoch": 0.44237861094038106,
+      "grad_norm": 3.3905465602874756,
+      "learning_rate": 1.8230485556238476e-05,
+      "loss": 1.3804,
+      "step": 5758
+    },
+    {
+      "epoch": 0.4425322679778734,
+      "grad_norm": 3.100095510482788,
+      "learning_rate": 1.822987092808851e-05,
+      "loss": 1.3149,
+      "step": 5760
+    },
+    {
+      "epoch": 0.4426859250153657,
+      "grad_norm": 2.685727596282959,
+      "learning_rate": 1.822925629993854e-05,
+      "loss": 1.4374,
+      "step": 5762
+    },
+    {
+      "epoch": 0.442839582052858,
+      "grad_norm": 3.614854335784912,
+      "learning_rate": 1.822864167178857e-05,
+      "loss": 1.4598,
+      "step": 5764
+    },
+    {
+      "epoch": 0.4429932390903503,
+      "grad_norm": 3.2616031169891357,
+      "learning_rate": 1.82280270436386e-05,
+      "loss": 1.3823,
+      "step": 5766
+    },
+    {
+      "epoch": 0.44314689612784264,
+      "grad_norm": 3.1923673152923584,
+      "learning_rate": 1.822741241548863e-05,
+      "loss": 1.3283,
+      "step": 5768
+    },
+    {
+      "epoch": 0.44330055316533495,
+      "grad_norm": 2.938978433609009,
+      "learning_rate": 1.822679778733866e-05,
+      "loss": 1.3428,
+      "step": 5770
+    },
+    {
+      "epoch": 0.44345421020282727,
+      "grad_norm": 2.983696222305298,
+      "learning_rate": 1.8226183159188694e-05,
+      "loss": 1.4124,
+      "step": 5772
+    },
+    {
+      "epoch": 0.4436078672403196,
+      "grad_norm": 2.893930673599243,
+      "learning_rate": 1.8225568531038724e-05,
+      "loss": 1.3289,
+      "step": 5774
+    },
+    {
+      "epoch": 0.4437615242778119,
+      "grad_norm": 2.9682881832122803,
+      "learning_rate": 1.8224953902888753e-05,
+      "loss": 1.3341,
+      "step": 5776
+    },
+    {
+      "epoch": 0.4439151813153042,
+      "grad_norm": 3.4092354774475098,
+      "learning_rate": 1.8224339274738786e-05,
+      "loss": 1.4816,
+      "step": 5778
+    },
+    {
+      "epoch": 0.44406883835279654,
+      "grad_norm": 3.190753698348999,
+      "learning_rate": 1.8223724646588816e-05,
+      "loss": 1.4722,
+      "step": 5780
+    },
+    {
+      "epoch": 0.44422249539028885,
+      "grad_norm": 3.21653413772583,
+      "learning_rate": 1.8223110018438846e-05,
+      "loss": 1.3672,
+      "step": 5782
+    },
+    {
+      "epoch": 0.44437615242778117,
+      "grad_norm": 2.8916025161743164,
+      "learning_rate": 1.8222495390288876e-05,
+      "loss": 1.3928,
+      "step": 5784
+    },
+    {
+      "epoch": 0.4445298094652735,
+      "grad_norm": 3.229156494140625,
+      "learning_rate": 1.822188076213891e-05,
+      "loss": 1.3865,
+      "step": 5786
+    },
+    {
+      "epoch": 0.4446834665027658,
+      "grad_norm": 3.3504526615142822,
+      "learning_rate": 1.8221266133988938e-05,
+      "loss": 1.4361,
+      "step": 5788
+    },
+    {
+      "epoch": 0.4448371235402581,
+      "grad_norm": 3.143098831176758,
+      "learning_rate": 1.8220651505838968e-05,
+      "loss": 1.3759,
+      "step": 5790
+    },
+    {
+      "epoch": 0.44499078057775043,
+      "grad_norm": 3.4050183296203613,
+      "learning_rate": 1.8220036877689e-05,
+      "loss": 1.4189,
+      "step": 5792
+    },
+    {
+      "epoch": 0.4451444376152428,
+      "grad_norm": 2.914482593536377,
+      "learning_rate": 1.821942224953903e-05,
+      "loss": 1.4096,
+      "step": 5794
+    },
+    {
+      "epoch": 0.4452980946527351,
+      "grad_norm": 3.0142135620117188,
+      "learning_rate": 1.821880762138906e-05,
+      "loss": 1.3218,
+      "step": 5796
+    },
+    {
+      "epoch": 0.44545175169022744,
+      "grad_norm": 3.1696937084198,
+      "learning_rate": 1.8218192993239093e-05,
+      "loss": 1.3977,
+      "step": 5798
+    },
+    {
+      "epoch": 0.44560540872771975,
+      "grad_norm": 3.157047986984253,
+      "learning_rate": 1.8217578365089123e-05,
+      "loss": 1.4031,
+      "step": 5800
+    },
+    {
+      "epoch": 0.44575906576521207,
+      "grad_norm": 3.1851911544799805,
+      "learning_rate": 1.8216963736939153e-05,
+      "loss": 1.4248,
+      "step": 5802
+    },
+    {
+      "epoch": 0.4459127228027044,
+      "grad_norm": 2.895350456237793,
+      "learning_rate": 1.8216349108789183e-05,
+      "loss": 1.3526,
+      "step": 5804
+    },
+    {
+      "epoch": 0.4460663798401967,
+      "grad_norm": 3.438236713409424,
+      "learning_rate": 1.8215734480639216e-05,
+      "loss": 1.4285,
+      "step": 5806
+    },
+    {
+      "epoch": 0.446220036877689,
+      "grad_norm": 3.439058780670166,
+      "learning_rate": 1.8215119852489245e-05,
+      "loss": 1.3343,
+      "step": 5808
+    },
+    {
+      "epoch": 0.44637369391518134,
+      "grad_norm": 3.126304864883423,
+      "learning_rate": 1.8214505224339275e-05,
+      "loss": 1.4201,
+      "step": 5810
+    },
+    {
+      "epoch": 0.44652735095267365,
+      "grad_norm": 2.642670154571533,
+      "learning_rate": 1.8213890596189308e-05,
+      "loss": 1.3266,
+      "step": 5812
+    },
+    {
+      "epoch": 0.44668100799016597,
+      "grad_norm": 3.0952022075653076,
+      "learning_rate": 1.8213275968039338e-05,
+      "loss": 1.4502,
+      "step": 5814
+    },
+    {
+      "epoch": 0.4468346650276583,
+      "grad_norm": 3.7239139080047607,
+      "learning_rate": 1.8212661339889367e-05,
+      "loss": 1.5575,
+      "step": 5816
+    },
+    {
+      "epoch": 0.4469883220651506,
+      "grad_norm": 3.190653085708618,
+      "learning_rate": 1.82120467117394e-05,
+      "loss": 1.4421,
+      "step": 5818
+    },
+    {
+      "epoch": 0.4471419791026429,
+      "grad_norm": 3.4979770183563232,
+      "learning_rate": 1.821143208358943e-05,
+      "loss": 1.3664,
+      "step": 5820
+    },
+    {
+      "epoch": 0.44729563614013523,
+      "grad_norm": 3.044233798980713,
+      "learning_rate": 1.821081745543946e-05,
+      "loss": 1.3876,
+      "step": 5822
+    },
+    {
+      "epoch": 0.44744929317762755,
+      "grad_norm": 3.0839719772338867,
+      "learning_rate": 1.8210202827289493e-05,
+      "loss": 1.294,
+      "step": 5824
+    },
+    {
+      "epoch": 0.44760295021511987,
+      "grad_norm": 3.3031883239746094,
+      "learning_rate": 1.8209588199139523e-05,
+      "loss": 1.4067,
+      "step": 5826
+    },
+    {
+      "epoch": 0.4477566072526122,
+      "grad_norm": 3.392949104309082,
+      "learning_rate": 1.8208973570989552e-05,
+      "loss": 1.4638,
+      "step": 5828
+    },
+    {
+      "epoch": 0.4479102642901045,
+      "grad_norm": 3.3899407386779785,
+      "learning_rate": 1.8208358942839582e-05,
+      "loss": 1.3866,
+      "step": 5830
+    },
+    {
+      "epoch": 0.4480639213275968,
+      "grad_norm": 2.626540422439575,
+      "learning_rate": 1.8207744314689615e-05,
+      "loss": 1.2544,
+      "step": 5832
+    },
+    {
+      "epoch": 0.44821757836508913,
+      "grad_norm": 3.6084189414978027,
+      "learning_rate": 1.8207129686539645e-05,
+      "loss": 1.3872,
+      "step": 5834
+    },
+    {
+      "epoch": 0.44837123540258145,
+      "grad_norm": 2.835602045059204,
+      "learning_rate": 1.8206515058389674e-05,
+      "loss": 1.3862,
+      "step": 5836
+    },
+    {
+      "epoch": 0.44852489244007376,
+      "grad_norm": 2.9570329189300537,
+      "learning_rate": 1.8205900430239707e-05,
+      "loss": 1.4073,
+      "step": 5838
+    },
+    {
+      "epoch": 0.4486785494775661,
+      "grad_norm": 2.910385847091675,
+      "learning_rate": 1.8205285802089737e-05,
+      "loss": 1.3464,
+      "step": 5840
+    },
+    {
+      "epoch": 0.4488322065150584,
+      "grad_norm": 3.0790252685546875,
+      "learning_rate": 1.8204671173939767e-05,
+      "loss": 1.3937,
+      "step": 5842
+    },
+    {
+      "epoch": 0.4489858635525507,
+      "grad_norm": 3.0580132007598877,
+      "learning_rate": 1.82040565457898e-05,
+      "loss": 1.2775,
+      "step": 5844
+    },
+    {
+      "epoch": 0.44913952059004303,
+      "grad_norm": 2.8806533813476562,
+      "learning_rate": 1.820344191763983e-05,
+      "loss": 1.4138,
+      "step": 5846
+    },
+    {
+      "epoch": 0.44929317762753535,
+      "grad_norm": 3.8401284217834473,
+      "learning_rate": 1.820282728948986e-05,
+      "loss": 1.3532,
+      "step": 5848
+    },
+    {
+      "epoch": 0.44944683466502766,
+      "grad_norm": 3.237717628479004,
+      "learning_rate": 1.8202212661339892e-05,
+      "loss": 1.3316,
+      "step": 5850
+    },
+    {
+      "epoch": 0.44960049170252,
+      "grad_norm": 2.958611011505127,
+      "learning_rate": 1.8201598033189922e-05,
+      "loss": 1.2768,
+      "step": 5852
+    },
+    {
+      "epoch": 0.4497541487400123,
+      "grad_norm": 3.4589743614196777,
+      "learning_rate": 1.8200983405039952e-05,
+      "loss": 1.5463,
+      "step": 5854
+    },
+    {
+      "epoch": 0.4499078057775046,
+      "grad_norm": 2.9172708988189697,
+      "learning_rate": 1.820036877688998e-05,
+      "loss": 1.4517,
+      "step": 5856
+    },
+    {
+      "epoch": 0.4500614628149969,
+      "grad_norm": 3.213388681411743,
+      "learning_rate": 1.8199754148740014e-05,
+      "loss": 1.2623,
+      "step": 5858
+    },
+    {
+      "epoch": 0.45021511985248924,
+      "grad_norm": 2.901362180709839,
+      "learning_rate": 1.8199139520590044e-05,
+      "loss": 1.2491,
+      "step": 5860
+    },
+    {
+      "epoch": 0.45036877688998156,
+      "grad_norm": 3.0259957313537598,
+      "learning_rate": 1.8198524892440074e-05,
+      "loss": 1.5275,
+      "step": 5862
+    },
+    {
+      "epoch": 0.4505224339274739,
+      "grad_norm": 3.00325870513916,
+      "learning_rate": 1.8197910264290107e-05,
+      "loss": 1.2732,
+      "step": 5864
+    },
+    {
+      "epoch": 0.4506760909649662,
+      "grad_norm": 3.229722023010254,
+      "learning_rate": 1.8197295636140137e-05,
+      "loss": 1.4167,
+      "step": 5866
+    },
+    {
+      "epoch": 0.4508297480024585,
+      "grad_norm": 3.2484633922576904,
+      "learning_rate": 1.8196681007990166e-05,
+      "loss": 1.3736,
+      "step": 5868
+    },
+    {
+      "epoch": 0.4509834050399508,
+      "grad_norm": 2.872192859649658,
+      "learning_rate": 1.81960663798402e-05,
+      "loss": 1.4017,
+      "step": 5870
+    },
+    {
+      "epoch": 0.45113706207744314,
+      "grad_norm": 3.5792553424835205,
+      "learning_rate": 1.819545175169023e-05,
+      "loss": 1.3938,
+      "step": 5872
+    },
+    {
+      "epoch": 0.45129071911493546,
+      "grad_norm": 3.1069889068603516,
+      "learning_rate": 1.819483712354026e-05,
+      "loss": 1.3554,
+      "step": 5874
+    },
+    {
+      "epoch": 0.4514443761524278,
+      "grad_norm": 3.3849682807922363,
+      "learning_rate": 1.8194222495390292e-05,
+      "loss": 1.3308,
+      "step": 5876
+    },
+    {
+      "epoch": 0.4515980331899201,
+      "grad_norm": 3.07407808303833,
+      "learning_rate": 1.819360786724032e-05,
+      "loss": 1.3343,
+      "step": 5878
+    },
+    {
+      "epoch": 0.4517516902274124,
+      "grad_norm": 3.1967318058013916,
+      "learning_rate": 1.819299323909035e-05,
+      "loss": 1.3538,
+      "step": 5880
+    },
+    {
+      "epoch": 0.4519053472649047,
+      "grad_norm": 3.133248805999756,
+      "learning_rate": 1.819237861094038e-05,
+      "loss": 1.4287,
+      "step": 5882
+    },
+    {
+      "epoch": 0.45205900430239704,
+      "grad_norm": 3.287682056427002,
+      "learning_rate": 1.8191763982790414e-05,
+      "loss": 1.3651,
+      "step": 5884
+    },
+    {
+      "epoch": 0.45221266133988935,
+      "grad_norm": 3.5104236602783203,
+      "learning_rate": 1.8191149354640444e-05,
+      "loss": 1.3389,
+      "step": 5886
+    },
+    {
+      "epoch": 0.45236631837738167,
+      "grad_norm": 3.1581168174743652,
+      "learning_rate": 1.8190534726490473e-05,
+      "loss": 1.493,
+      "step": 5888
+    },
+    {
+      "epoch": 0.452519975414874,
+      "grad_norm": 3.408336877822876,
+      "learning_rate": 1.8189920098340506e-05,
+      "loss": 1.3866,
+      "step": 5890
+    },
+    {
+      "epoch": 0.4526736324523663,
+      "grad_norm": 3.840761661529541,
+      "learning_rate": 1.8189305470190536e-05,
+      "loss": 1.5172,
+      "step": 5892
+    },
+    {
+      "epoch": 0.4528272894898586,
+      "grad_norm": 3.4090771675109863,
+      "learning_rate": 1.8188690842040566e-05,
+      "loss": 1.3251,
+      "step": 5894
+    },
+    {
+      "epoch": 0.45298094652735094,
+      "grad_norm": 2.792132616043091,
+      "learning_rate": 1.81880762138906e-05,
+      "loss": 1.4536,
+      "step": 5896
+    },
+    {
+      "epoch": 0.45313460356484325,
+      "grad_norm": 2.6732709407806396,
+      "learning_rate": 1.818746158574063e-05,
+      "loss": 1.3353,
+      "step": 5898
+    },
+    {
+      "epoch": 0.45328826060233557,
+      "grad_norm": 3.2554478645324707,
+      "learning_rate": 1.8186846957590658e-05,
+      "loss": 1.3426,
+      "step": 5900
+    },
+    {
+      "epoch": 0.4534419176398279,
+      "grad_norm": 3.0684661865234375,
+      "learning_rate": 1.8186232329440688e-05,
+      "loss": 1.2457,
+      "step": 5902
+    },
+    {
+      "epoch": 0.4535955746773202,
+      "grad_norm": 2.9400198459625244,
+      "learning_rate": 1.818561770129072e-05,
+      "loss": 1.386,
+      "step": 5904
+    },
+    {
+      "epoch": 0.4537492317148125,
+      "grad_norm": 3.3350679874420166,
+      "learning_rate": 1.8185003073140754e-05,
+      "loss": 1.4366,
+      "step": 5906
+    },
+    {
+      "epoch": 0.45390288875230483,
+      "grad_norm": 2.902355670928955,
+      "learning_rate": 1.818438844499078e-05,
+      "loss": 1.2785,
+      "step": 5908
+    },
+    {
+      "epoch": 0.45405654578979715,
+      "grad_norm": 2.7429327964782715,
+      "learning_rate": 1.8183773816840813e-05,
+      "loss": 1.2129,
+      "step": 5910
+    },
+    {
+      "epoch": 0.45421020282728947,
+      "grad_norm": 3.4028749465942383,
+      "learning_rate": 1.8183159188690843e-05,
+      "loss": 1.5344,
+      "step": 5912
+    },
+    {
+      "epoch": 0.4543638598647818,
+      "grad_norm": 3.4940731525421143,
+      "learning_rate": 1.8182544560540873e-05,
+      "loss": 1.3376,
+      "step": 5914
+    },
+    {
+      "epoch": 0.4545175169022741,
+      "grad_norm": 3.099562644958496,
+      "learning_rate": 1.8181929932390906e-05,
+      "loss": 1.3071,
+      "step": 5916
+    },
+    {
+      "epoch": 0.4546711739397664,
+      "grad_norm": 2.8082468509674072,
+      "learning_rate": 1.8181315304240935e-05,
+      "loss": 1.273,
+      "step": 5918
+    },
+    {
+      "epoch": 0.45482483097725873,
+      "grad_norm": 2.951359510421753,
+      "learning_rate": 1.8180700676090965e-05,
+      "loss": 1.2602,
+      "step": 5920
+    },
+    {
+      "epoch": 0.45497848801475105,
+      "grad_norm": 3.421891212463379,
+      "learning_rate": 1.8180086047940998e-05,
+      "loss": 1.4899,
+      "step": 5922
+    },
+    {
+      "epoch": 0.4551321450522434,
+      "grad_norm": 3.0956802368164062,
+      "learning_rate": 1.8179471419791028e-05,
+      "loss": 1.4078,
+      "step": 5924
+    },
+    {
+      "epoch": 0.45528580208973574,
+      "grad_norm": 3.26657772064209,
+      "learning_rate": 1.817885679164106e-05,
+      "loss": 1.4338,
+      "step": 5926
+    },
+    {
+      "epoch": 0.45543945912722805,
+      "grad_norm": 3.3916943073272705,
+      "learning_rate": 1.8178242163491087e-05,
+      "loss": 1.4113,
+      "step": 5928
+    },
+    {
+      "epoch": 0.45559311616472037,
+      "grad_norm": 3.281963348388672,
+      "learning_rate": 1.817762753534112e-05,
+      "loss": 1.3594,
+      "step": 5930
+    },
+    {
+      "epoch": 0.4557467732022127,
+      "grad_norm": 2.989858627319336,
+      "learning_rate": 1.817701290719115e-05,
+      "loss": 1.4444,
+      "step": 5932
+    },
+    {
+      "epoch": 0.455900430239705,
+      "grad_norm": 2.8269715309143066,
+      "learning_rate": 1.817639827904118e-05,
+      "loss": 1.3883,
+      "step": 5934
+    },
+    {
+      "epoch": 0.4560540872771973,
+      "grad_norm": 3.383490800857544,
+      "learning_rate": 1.8175783650891213e-05,
+      "loss": 1.4248,
+      "step": 5936
+    },
+    {
+      "epoch": 0.45620774431468963,
+      "grad_norm": 3.237833023071289,
+      "learning_rate": 1.8175169022741242e-05,
+      "loss": 1.3974,
+      "step": 5938
+    },
+    {
+      "epoch": 0.45636140135218195,
+      "grad_norm": 3.240793466567993,
+      "learning_rate": 1.8174554394591272e-05,
+      "loss": 1.3285,
+      "step": 5940
+    },
+    {
+      "epoch": 0.45651505838967427,
+      "grad_norm": 3.475192070007324,
+      "learning_rate": 1.8173939766441305e-05,
+      "loss": 1.384,
+      "step": 5942
+    },
+    {
+      "epoch": 0.4566687154271666,
+      "grad_norm": 3.198943853378296,
+      "learning_rate": 1.8173325138291335e-05,
+      "loss": 1.4785,
+      "step": 5944
+    },
+    {
+      "epoch": 0.4568223724646589,
+      "grad_norm": 3.021594524383545,
+      "learning_rate": 1.8172710510141368e-05,
+      "loss": 1.3124,
+      "step": 5946
+    },
+    {
+      "epoch": 0.4569760295021512,
+      "grad_norm": 3.0245521068573,
+      "learning_rate": 1.8172095881991398e-05,
+      "loss": 1.439,
+      "step": 5948
+    },
+    {
+      "epoch": 0.45712968653964353,
+      "grad_norm": 3.3448681831359863,
+      "learning_rate": 1.8171481253841427e-05,
+      "loss": 1.401,
+      "step": 5950
+    },
+    {
+      "epoch": 0.45728334357713585,
+      "grad_norm": 3.0003669261932373,
+      "learning_rate": 1.817086662569146e-05,
+      "loss": 1.4555,
+      "step": 5952
+    },
+    {
+      "epoch": 0.45743700061462816,
+      "grad_norm": 3.3074393272399902,
+      "learning_rate": 1.8170251997541487e-05,
+      "loss": 1.3984,
+      "step": 5954
+    },
+    {
+      "epoch": 0.4575906576521205,
+      "grad_norm": 3.2424259185791016,
+      "learning_rate": 1.816963736939152e-05,
+      "loss": 1.3531,
+      "step": 5956
+    },
+    {
+      "epoch": 0.4577443146896128,
+      "grad_norm": 3.414992094039917,
+      "learning_rate": 1.816902274124155e-05,
+      "loss": 1.4555,
+      "step": 5958
+    },
+    {
+      "epoch": 0.4578979717271051,
+      "grad_norm": 3.0615060329437256,
+      "learning_rate": 1.816840811309158e-05,
+      "loss": 1.5183,
+      "step": 5960
+    },
+    {
+      "epoch": 0.45805162876459743,
+      "grad_norm": 3.127685070037842,
+      "learning_rate": 1.8167793484941612e-05,
+      "loss": 1.4996,
+      "step": 5962
+    },
+    {
+      "epoch": 0.45820528580208975,
+      "grad_norm": 2.8747687339782715,
+      "learning_rate": 1.8167178856791642e-05,
+      "loss": 1.363,
+      "step": 5964
+    },
+    {
+      "epoch": 0.45835894283958206,
+      "grad_norm": 3.215275764465332,
+      "learning_rate": 1.8166564228641675e-05,
+      "loss": 1.32,
+      "step": 5966
+    },
+    {
+      "epoch": 0.4585125998770744,
+      "grad_norm": 2.8502352237701416,
+      "learning_rate": 1.8165949600491705e-05,
+      "loss": 1.3669,
+      "step": 5968
+    },
+    {
+      "epoch": 0.4586662569145667,
+      "grad_norm": 3.226792812347412,
+      "learning_rate": 1.8165334972341734e-05,
+      "loss": 1.4084,
+      "step": 5970
+    },
+    {
+      "epoch": 0.458819913952059,
+      "grad_norm": 2.946282386779785,
+      "learning_rate": 1.8164720344191767e-05,
+      "loss": 1.3186,
+      "step": 5972
+    },
+    {
+      "epoch": 0.4589735709895513,
+      "grad_norm": 2.6620097160339355,
+      "learning_rate": 1.8164105716041797e-05,
+      "loss": 1.219,
+      "step": 5974
+    },
+    {
+      "epoch": 0.45912722802704364,
+      "grad_norm": 3.052964925765991,
+      "learning_rate": 1.8163491087891827e-05,
+      "loss": 1.2451,
+      "step": 5976
+    },
+    {
+      "epoch": 0.45928088506453596,
+      "grad_norm": 2.9971115589141846,
+      "learning_rate": 1.816287645974186e-05,
+      "loss": 1.3349,
+      "step": 5978
+    },
+    {
+      "epoch": 0.4594345421020283,
+      "grad_norm": 2.765615701675415,
+      "learning_rate": 1.8162261831591886e-05,
+      "loss": 1.3993,
+      "step": 5980
+    },
+    {
+      "epoch": 0.4595881991395206,
+      "grad_norm": 3.110050916671753,
+      "learning_rate": 1.816164720344192e-05,
+      "loss": 1.3793,
+      "step": 5982
+    },
+    {
+      "epoch": 0.4597418561770129,
+      "grad_norm": 3.004164934158325,
+      "learning_rate": 1.816103257529195e-05,
+      "loss": 1.3538,
+      "step": 5984
+    },
+    {
+      "epoch": 0.4598955132145052,
+      "grad_norm": 2.9276111125946045,
+      "learning_rate": 1.8160417947141982e-05,
+      "loss": 1.4343,
+      "step": 5986
+    },
+    {
+      "epoch": 0.46004917025199754,
+      "grad_norm": 3.2313802242279053,
+      "learning_rate": 1.815980331899201e-05,
+      "loss": 1.4701,
+      "step": 5988
+    },
+    {
+      "epoch": 0.46020282728948986,
+      "grad_norm": 3.9364774227142334,
+      "learning_rate": 1.815918869084204e-05,
+      "loss": 1.4805,
+      "step": 5990
+    },
+    {
+      "epoch": 0.4603564843269822,
+      "grad_norm": 3.274184465408325,
+      "learning_rate": 1.8158574062692074e-05,
+      "loss": 1.3289,
+      "step": 5992
+    },
+    {
+      "epoch": 0.4605101413644745,
+      "grad_norm": 3.1217334270477295,
+      "learning_rate": 1.8157959434542104e-05,
+      "loss": 1.4466,
+      "step": 5994
+    },
+    {
+      "epoch": 0.4606637984019668,
+      "grad_norm": 2.6427552700042725,
+      "learning_rate": 1.8157344806392134e-05,
+      "loss": 1.3293,
+      "step": 5996
+    },
+    {
+      "epoch": 0.4608174554394591,
+      "grad_norm": 3.3423068523406982,
+      "learning_rate": 1.8156730178242167e-05,
+      "loss": 1.3001,
+      "step": 5998
+    },
+    {
+      "epoch": 0.46097111247695144,
+      "grad_norm": 3.421719551086426,
+      "learning_rate": 1.8156115550092193e-05,
+      "loss": 1.2599,
+      "step": 6000
+    },
+    {
+      "epoch": 0.46112476951444376,
+      "grad_norm": 2.9063069820404053,
+      "learning_rate": 1.8155500921942226e-05,
+      "loss": 1.2448,
+      "step": 6002
+    },
+    {
+      "epoch": 0.46127842655193607,
+      "grad_norm": 3.389843225479126,
+      "learning_rate": 1.815488629379226e-05,
+      "loss": 1.465,
+      "step": 6004
+    },
+    {
+      "epoch": 0.4614320835894284,
+      "grad_norm": 3.1748673915863037,
+      "learning_rate": 1.8154271665642286e-05,
+      "loss": 1.442,
+      "step": 6006
+    },
+    {
+      "epoch": 0.4615857406269207,
+      "grad_norm": 3.1274852752685547,
+      "learning_rate": 1.815365703749232e-05,
+      "loss": 1.3625,
+      "step": 6008
+    },
+    {
+      "epoch": 0.461739397664413,
+      "grad_norm": 2.8795769214630127,
+      "learning_rate": 1.815304240934235e-05,
+      "loss": 1.3289,
+      "step": 6010
+    },
+    {
+      "epoch": 0.46189305470190534,
+      "grad_norm": 2.991797924041748,
+      "learning_rate": 1.815242778119238e-05,
+      "loss": 1.3398,
+      "step": 6012
+    },
+    {
+      "epoch": 0.46204671173939765,
+      "grad_norm": 2.745926856994629,
+      "learning_rate": 1.815181315304241e-05,
+      "loss": 1.3411,
+      "step": 6014
+    },
+    {
+      "epoch": 0.46220036877688997,
+      "grad_norm": 3.4233834743499756,
+      "learning_rate": 1.815119852489244e-05,
+      "loss": 1.476,
+      "step": 6016
+    },
+    {
+      "epoch": 0.4623540258143823,
+      "grad_norm": 3.179094076156616,
+      "learning_rate": 1.8150583896742474e-05,
+      "loss": 1.3088,
+      "step": 6018
+    },
+    {
+      "epoch": 0.4625076828518746,
+      "grad_norm": 3.251293897628784,
+      "learning_rate": 1.8149969268592504e-05,
+      "loss": 1.3724,
+      "step": 6020
+    },
+    {
+      "epoch": 0.4626613398893669,
+      "grad_norm": 3.3762452602386475,
+      "learning_rate": 1.8149354640442533e-05,
+      "loss": 1.4215,
+      "step": 6022
+    },
+    {
+      "epoch": 0.46281499692685923,
+      "grad_norm": 2.5178112983703613,
+      "learning_rate": 1.8148740012292566e-05,
+      "loss": 1.2476,
+      "step": 6024
+    },
+    {
+      "epoch": 0.46296865396435155,
+      "grad_norm": 3.1534478664398193,
+      "learning_rate": 1.8148125384142593e-05,
+      "loss": 1.2818,
+      "step": 6026
+    },
+    {
+      "epoch": 0.46312231100184387,
+      "grad_norm": 2.7992680072784424,
+      "learning_rate": 1.8147510755992626e-05,
+      "loss": 1.3746,
+      "step": 6028
+    },
+    {
+      "epoch": 0.4632759680393362,
+      "grad_norm": 3.016869068145752,
+      "learning_rate": 1.8146896127842655e-05,
+      "loss": 1.1759,
+      "step": 6030
+    },
+    {
+      "epoch": 0.4634296250768285,
+      "grad_norm": 3.1666297912597656,
+      "learning_rate": 1.814628149969269e-05,
+      "loss": 1.2996,
+      "step": 6032
+    },
+    {
+      "epoch": 0.4635832821143208,
+      "grad_norm": 3.2005155086517334,
+      "learning_rate": 1.8145666871542718e-05,
+      "loss": 1.354,
+      "step": 6034
+    },
+    {
+      "epoch": 0.46373693915181313,
+      "grad_norm": 2.9398839473724365,
+      "learning_rate": 1.8145052243392748e-05,
+      "loss": 1.3371,
+      "step": 6036
+    },
+    {
+      "epoch": 0.46389059618930545,
+      "grad_norm": 2.675481081008911,
+      "learning_rate": 1.814443761524278e-05,
+      "loss": 1.4086,
+      "step": 6038
+    },
+    {
+      "epoch": 0.46404425322679776,
+      "grad_norm": 2.6939587593078613,
+      "learning_rate": 1.814382298709281e-05,
+      "loss": 1.4703,
+      "step": 6040
+    },
+    {
+      "epoch": 0.4641979102642901,
+      "grad_norm": 3.3314614295959473,
+      "learning_rate": 1.814320835894284e-05,
+      "loss": 1.3802,
+      "step": 6042
+    },
+    {
+      "epoch": 0.4643515673017824,
+      "grad_norm": 2.7407829761505127,
+      "learning_rate": 1.8142593730792873e-05,
+      "loss": 1.2443,
+      "step": 6044
+    },
+    {
+      "epoch": 0.4645052243392747,
+      "grad_norm": 3.5611183643341064,
+      "learning_rate": 1.8141979102642903e-05,
+      "loss": 1.3102,
+      "step": 6046
+    },
+    {
+      "epoch": 0.46465888137676703,
+      "grad_norm": 3.135925054550171,
+      "learning_rate": 1.8141364474492933e-05,
+      "loss": 1.2848,
+      "step": 6048
+    },
+    {
+      "epoch": 0.46481253841425935,
+      "grad_norm": 2.807861566543579,
+      "learning_rate": 1.8140749846342966e-05,
+      "loss": 1.498,
+      "step": 6050
+    },
+    {
+      "epoch": 0.46496619545175166,
+      "grad_norm": 2.9039602279663086,
+      "learning_rate": 1.8140135218192995e-05,
+      "loss": 1.3336,
+      "step": 6052
+    },
+    {
+      "epoch": 0.46511985248924403,
+      "grad_norm": 3.3446543216705322,
+      "learning_rate": 1.8139520590043025e-05,
+      "loss": 1.3786,
+      "step": 6054
+    },
+    {
+      "epoch": 0.46527350952673635,
+      "grad_norm": 3.125364065170288,
+      "learning_rate": 1.8138905961893055e-05,
+      "loss": 1.3462,
+      "step": 6056
+    },
+    {
+      "epoch": 0.46542716656422867,
+      "grad_norm": 3.407083511352539,
+      "learning_rate": 1.8138291333743088e-05,
+      "loss": 1.4261,
+      "step": 6058
+    },
+    {
+      "epoch": 0.465580823601721,
+      "grad_norm": 2.6646227836608887,
+      "learning_rate": 1.8137676705593118e-05,
+      "loss": 1.2944,
+      "step": 6060
+    },
+    {
+      "epoch": 0.4657344806392133,
+      "grad_norm": 3.1010098457336426,
+      "learning_rate": 1.8137062077443147e-05,
+      "loss": 1.4409,
+      "step": 6062
+    },
+    {
+      "epoch": 0.4658881376767056,
+      "grad_norm": 2.994729518890381,
+      "learning_rate": 1.813644744929318e-05,
+      "loss": 1.2887,
+      "step": 6064
+    },
+    {
+      "epoch": 0.46604179471419793,
+      "grad_norm": 2.98126482963562,
+      "learning_rate": 1.813583282114321e-05,
+      "loss": 1.4378,
+      "step": 6066
+    },
+    {
+      "epoch": 0.46619545175169025,
+      "grad_norm": 3.0406136512756348,
+      "learning_rate": 1.813521819299324e-05,
+      "loss": 1.4089,
+      "step": 6068
+    },
+    {
+      "epoch": 0.46634910878918256,
+      "grad_norm": 2.912886619567871,
+      "learning_rate": 1.8134603564843273e-05,
+      "loss": 1.297,
+      "step": 6070
+    },
+    {
+      "epoch": 0.4665027658266749,
+      "grad_norm": 3.1055288314819336,
+      "learning_rate": 1.8133988936693302e-05,
+      "loss": 1.3387,
+      "step": 6072
+    },
+    {
+      "epoch": 0.4666564228641672,
+      "grad_norm": 2.698050022125244,
+      "learning_rate": 1.8133374308543332e-05,
+      "loss": 1.3589,
+      "step": 6074
+    },
+    {
+      "epoch": 0.4668100799016595,
+      "grad_norm": 3.304744243621826,
+      "learning_rate": 1.8132759680393365e-05,
+      "loss": 1.3344,
+      "step": 6076
+    },
+    {
+      "epoch": 0.46696373693915183,
+      "grad_norm": 3.2374093532562256,
+      "learning_rate": 1.8132145052243395e-05,
+      "loss": 1.3627,
+      "step": 6078
+    },
+    {
+      "epoch": 0.46711739397664415,
+      "grad_norm": 3.138913631439209,
+      "learning_rate": 1.8131530424093425e-05,
+      "loss": 1.3703,
+      "step": 6080
+    },
+    {
+      "epoch": 0.46727105101413646,
+      "grad_norm": 3.517970085144043,
+      "learning_rate": 1.8130915795943454e-05,
+      "loss": 1.3284,
+      "step": 6082
+    },
+    {
+      "epoch": 0.4674247080516288,
+      "grad_norm": 2.9830198287963867,
+      "learning_rate": 1.8130301167793487e-05,
+      "loss": 1.3917,
+      "step": 6084
+    },
+    {
+      "epoch": 0.4675783650891211,
+      "grad_norm": 2.6390559673309326,
+      "learning_rate": 1.8129686539643517e-05,
+      "loss": 1.3801,
+      "step": 6086
+    },
+    {
+      "epoch": 0.4677320221266134,
+      "grad_norm": 3.0859718322753906,
+      "learning_rate": 1.8129071911493547e-05,
+      "loss": 1.4656,
+      "step": 6088
+    },
+    {
+      "epoch": 0.46788567916410573,
+      "grad_norm": 2.984755516052246,
+      "learning_rate": 1.812845728334358e-05,
+      "loss": 1.4782,
+      "step": 6090
+    },
+    {
+      "epoch": 0.46803933620159804,
+      "grad_norm": 3.045100450515747,
+      "learning_rate": 1.812784265519361e-05,
+      "loss": 1.4161,
+      "step": 6092
+    },
+    {
+      "epoch": 0.46819299323909036,
+      "grad_norm": 3.148865222930908,
+      "learning_rate": 1.812722802704364e-05,
+      "loss": 1.3084,
+      "step": 6094
+    },
+    {
+      "epoch": 0.4683466502765827,
+      "grad_norm": 3.177959680557251,
+      "learning_rate": 1.8126613398893672e-05,
+      "loss": 1.3234,
+      "step": 6096
+    },
+    {
+      "epoch": 0.468500307314075,
+      "grad_norm": 2.8266592025756836,
+      "learning_rate": 1.8125998770743702e-05,
+      "loss": 1.4313,
+      "step": 6098
+    },
+    {
+      "epoch": 0.4686539643515673,
+      "grad_norm": 3.3804705142974854,
+      "learning_rate": 1.812538414259373e-05,
+      "loss": 1.3802,
+      "step": 6100
+    },
+    {
+      "epoch": 0.4688076213890596,
+      "grad_norm": 3.3395400047302246,
+      "learning_rate": 1.8124769514443765e-05,
+      "loss": 1.2999,
+      "step": 6102
+    },
+    {
+      "epoch": 0.46896127842655194,
+      "grad_norm": 2.902346134185791,
+      "learning_rate": 1.8124154886293794e-05,
+      "loss": 1.3411,
+      "step": 6104
+    },
+    {
+      "epoch": 0.46911493546404426,
+      "grad_norm": 3.264467477798462,
+      "learning_rate": 1.8123540258143824e-05,
+      "loss": 1.4063,
+      "step": 6106
+    },
+    {
+      "epoch": 0.4692685925015366,
+      "grad_norm": 2.926862955093384,
+      "learning_rate": 1.8122925629993854e-05,
+      "loss": 1.2848,
+      "step": 6108
+    },
+    {
+      "epoch": 0.4694222495390289,
+      "grad_norm": 3.159416913986206,
+      "learning_rate": 1.8122311001843887e-05,
+      "loss": 1.526,
+      "step": 6110
+    },
+    {
+      "epoch": 0.4695759065765212,
+      "grad_norm": 2.464360237121582,
+      "learning_rate": 1.8121696373693916e-05,
+      "loss": 1.2014,
+      "step": 6112
+    },
+    {
+      "epoch": 0.4697295636140135,
+      "grad_norm": 3.0088672637939453,
+      "learning_rate": 1.8121081745543946e-05,
+      "loss": 1.4058,
+      "step": 6114
+    },
+    {
+      "epoch": 0.46988322065150584,
+      "grad_norm": 2.786822557449341,
+      "learning_rate": 1.812046711739398e-05,
+      "loss": 1.2714,
+      "step": 6116
+    },
+    {
+      "epoch": 0.47003687768899816,
+      "grad_norm": 2.730426549911499,
+      "learning_rate": 1.811985248924401e-05,
+      "loss": 1.1898,
+      "step": 6118
+    },
+    {
+      "epoch": 0.47019053472649047,
+      "grad_norm": 2.9403066635131836,
+      "learning_rate": 1.811923786109404e-05,
+      "loss": 1.4679,
+      "step": 6120
+    },
+    {
+      "epoch": 0.4703441917639828,
+      "grad_norm": 2.8969335556030273,
+      "learning_rate": 1.811862323294407e-05,
+      "loss": 1.3628,
+      "step": 6122
+    },
+    {
+      "epoch": 0.4704978488014751,
+      "grad_norm": 3.1321475505828857,
+      "learning_rate": 1.81180086047941e-05,
+      "loss": 1.2599,
+      "step": 6124
+    },
+    {
+      "epoch": 0.4706515058389674,
+      "grad_norm": 2.7310781478881836,
+      "learning_rate": 1.811739397664413e-05,
+      "loss": 1.343,
+      "step": 6126
+    },
+    {
+      "epoch": 0.47080516287645974,
+      "grad_norm": 3.2368597984313965,
+      "learning_rate": 1.811677934849416e-05,
+      "loss": 1.452,
+      "step": 6128
+    },
+    {
+      "epoch": 0.47095881991395205,
+      "grad_norm": 3.1757235527038574,
+      "learning_rate": 1.8116164720344194e-05,
+      "loss": 1.2629,
+      "step": 6130
+    },
+    {
+      "epoch": 0.47111247695144437,
+      "grad_norm": 3.524467706680298,
+      "learning_rate": 1.8115550092194223e-05,
+      "loss": 1.3946,
+      "step": 6132
+    },
+    {
+      "epoch": 0.4712661339889367,
+      "grad_norm": 2.9439032077789307,
+      "learning_rate": 1.8114935464044253e-05,
+      "loss": 1.3863,
+      "step": 6134
+    },
+    {
+      "epoch": 0.471419791026429,
+      "grad_norm": 3.1204442977905273,
+      "learning_rate": 1.8114320835894286e-05,
+      "loss": 1.3305,
+      "step": 6136
+    },
+    {
+      "epoch": 0.4715734480639213,
+      "grad_norm": 2.8491554260253906,
+      "learning_rate": 1.8113706207744316e-05,
+      "loss": 1.3718,
+      "step": 6138
+    },
+    {
+      "epoch": 0.47172710510141364,
+      "grad_norm": 3.3984553813934326,
+      "learning_rate": 1.8113091579594346e-05,
+      "loss": 1.3786,
+      "step": 6140
+    },
+    {
+      "epoch": 0.47188076213890595,
+      "grad_norm": 2.843414545059204,
+      "learning_rate": 1.811247695144438e-05,
+      "loss": 1.4403,
+      "step": 6142
+    },
+    {
+      "epoch": 0.47203441917639827,
+      "grad_norm": 3.248155355453491,
+      "learning_rate": 1.811186232329441e-05,
+      "loss": 1.3445,
+      "step": 6144
+    },
+    {
+      "epoch": 0.4721880762138906,
+      "grad_norm": 2.920718193054199,
+      "learning_rate": 1.8111247695144438e-05,
+      "loss": 1.4953,
+      "step": 6146
+    },
+    {
+      "epoch": 0.4723417332513829,
+      "grad_norm": 2.642434597015381,
+      "learning_rate": 1.811063306699447e-05,
+      "loss": 1.4154,
+      "step": 6148
+    },
+    {
+      "epoch": 0.4724953902888752,
+      "grad_norm": 2.8133106231689453,
+      "learning_rate": 1.81100184388445e-05,
+      "loss": 1.2154,
+      "step": 6150
+    },
+    {
+      "epoch": 0.47264904732636753,
+      "grad_norm": 2.921673059463501,
+      "learning_rate": 1.810940381069453e-05,
+      "loss": 1.4372,
+      "step": 6152
+    },
+    {
+      "epoch": 0.47280270436385985,
+      "grad_norm": 2.9900777339935303,
+      "learning_rate": 1.810878918254456e-05,
+      "loss": 1.3489,
+      "step": 6154
+    },
+    {
+      "epoch": 0.47295636140135217,
+      "grad_norm": 3.1354458332061768,
+      "learning_rate": 1.8108174554394593e-05,
+      "loss": 1.3848,
+      "step": 6156
+    },
+    {
+      "epoch": 0.4731100184388445,
+      "grad_norm": 2.98395037651062,
+      "learning_rate": 1.8107559926244623e-05,
+      "loss": 1.1782,
+      "step": 6158
+    },
+    {
+      "epoch": 0.4732636754763368,
+      "grad_norm": 3.0558385848999023,
+      "learning_rate": 1.8106945298094653e-05,
+      "loss": 1.4362,
+      "step": 6160
+    },
+    {
+      "epoch": 0.4734173325138291,
+      "grad_norm": 2.871683120727539,
+      "learning_rate": 1.8106330669944686e-05,
+      "loss": 1.361,
+      "step": 6162
+    },
+    {
+      "epoch": 0.47357098955132143,
+      "grad_norm": 3.1952767372131348,
+      "learning_rate": 1.8105716041794715e-05,
+      "loss": 1.371,
+      "step": 6164
+    },
+    {
+      "epoch": 0.47372464658881375,
+      "grad_norm": 3.256314754486084,
+      "learning_rate": 1.8105101413644745e-05,
+      "loss": 1.5385,
+      "step": 6166
+    },
+    {
+      "epoch": 0.47387830362630606,
+      "grad_norm": 3.1399924755096436,
+      "learning_rate": 1.8104486785494778e-05,
+      "loss": 1.4099,
+      "step": 6168
+    },
+    {
+      "epoch": 0.4740319606637984,
+      "grad_norm": 2.872492551803589,
+      "learning_rate": 1.8103872157344808e-05,
+      "loss": 1.3904,
+      "step": 6170
+    },
+    {
+      "epoch": 0.4741856177012907,
+      "grad_norm": 3.3130695819854736,
+      "learning_rate": 1.8103257529194837e-05,
+      "loss": 1.4766,
+      "step": 6172
+    },
+    {
+      "epoch": 0.474339274738783,
+      "grad_norm": 3.3041231632232666,
+      "learning_rate": 1.810264290104487e-05,
+      "loss": 1.3754,
+      "step": 6174
+    },
+    {
+      "epoch": 0.47449293177627533,
+      "grad_norm": 2.94787859916687,
+      "learning_rate": 1.81020282728949e-05,
+      "loss": 1.407,
+      "step": 6176
+    },
+    {
+      "epoch": 0.47464658881376764,
+      "grad_norm": 2.9895570278167725,
+      "learning_rate": 1.8101413644744933e-05,
+      "loss": 1.4253,
+      "step": 6178
+    },
+    {
+      "epoch": 0.47480024585125996,
+      "grad_norm": 3.096012830734253,
+      "learning_rate": 1.810079901659496e-05,
+      "loss": 1.4598,
+      "step": 6180
+    },
+    {
+      "epoch": 0.4749539028887523,
+      "grad_norm": 3.4266176223754883,
+      "learning_rate": 1.8100184388444993e-05,
+      "loss": 1.3014,
+      "step": 6182
+    },
+    {
+      "epoch": 0.47510755992624465,
+      "grad_norm": 3.3502449989318848,
+      "learning_rate": 1.8099569760295022e-05,
+      "loss": 1.403,
+      "step": 6184
+    },
+    {
+      "epoch": 0.47526121696373697,
+      "grad_norm": 2.791080951690674,
+      "learning_rate": 1.8098955132145052e-05,
+      "loss": 1.3161,
+      "step": 6186
+    },
+    {
+      "epoch": 0.4754148740012293,
+      "grad_norm": 3.389315366744995,
+      "learning_rate": 1.8098340503995085e-05,
+      "loss": 1.3105,
+      "step": 6188
+    },
+    {
+      "epoch": 0.4755685310387216,
+      "grad_norm": 3.1894092559814453,
+      "learning_rate": 1.8097725875845115e-05,
+      "loss": 1.334,
+      "step": 6190
+    },
+    {
+      "epoch": 0.4757221880762139,
+      "grad_norm": 3.219374418258667,
+      "learning_rate": 1.8097111247695144e-05,
+      "loss": 1.2583,
+      "step": 6192
+    },
+    {
+      "epoch": 0.47587584511370623,
+      "grad_norm": 2.9629125595092773,
+      "learning_rate": 1.8096496619545178e-05,
+      "loss": 1.472,
+      "step": 6194
+    },
+    {
+      "epoch": 0.47602950215119855,
+      "grad_norm": 2.8309929370880127,
+      "learning_rate": 1.8095881991395207e-05,
+      "loss": 1.4128,
+      "step": 6196
+    },
+    {
+      "epoch": 0.47618315918869086,
+      "grad_norm": 3.1029868125915527,
+      "learning_rate": 1.809526736324524e-05,
+      "loss": 1.4316,
+      "step": 6198
+    },
+    {
+      "epoch": 0.4763368162261832,
+      "grad_norm": 3.910332679748535,
+      "learning_rate": 1.809465273509527e-05,
+      "loss": 1.4264,
+      "step": 6200
+    },
+    {
+      "epoch": 0.4764904732636755,
+      "grad_norm": 3.1213109493255615,
+      "learning_rate": 1.80940381069453e-05,
+      "loss": 1.3166,
+      "step": 6202
+    },
+    {
+      "epoch": 0.4766441303011678,
+      "grad_norm": 3.3485963344573975,
+      "learning_rate": 1.8093423478795333e-05,
+      "loss": 1.4312,
+      "step": 6204
+    },
+    {
+      "epoch": 0.47679778733866013,
+      "grad_norm": 2.8326988220214844,
+      "learning_rate": 1.809280885064536e-05,
+      "loss": 1.3337,
+      "step": 6206
+    },
+    {
+      "epoch": 0.47695144437615244,
+      "grad_norm": 3.0213944911956787,
+      "learning_rate": 1.8092194222495392e-05,
+      "loss": 1.3796,
+      "step": 6208
+    },
+    {
+      "epoch": 0.47710510141364476,
+      "grad_norm": 3.090485095977783,
+      "learning_rate": 1.8091579594345422e-05,
+      "loss": 1.3794,
+      "step": 6210
+    },
+    {
+      "epoch": 0.4772587584511371,
+      "grad_norm": 2.931671619415283,
+      "learning_rate": 1.809096496619545e-05,
+      "loss": 1.4677,
+      "step": 6212
+    },
+    {
+      "epoch": 0.4774124154886294,
+      "grad_norm": 3.267240524291992,
+      "learning_rate": 1.8090350338045485e-05,
+      "loss": 1.3751,
+      "step": 6214
+    },
+    {
+      "epoch": 0.4775660725261217,
+      "grad_norm": 2.553067207336426,
+      "learning_rate": 1.8089735709895514e-05,
+      "loss": 1.3013,
+      "step": 6216
+    },
+    {
+      "epoch": 0.477719729563614,
+      "grad_norm": 3.1998229026794434,
+      "learning_rate": 1.8089121081745547e-05,
+      "loss": 1.3474,
+      "step": 6218
+    },
+    {
+      "epoch": 0.47787338660110634,
+      "grad_norm": 11.610280990600586,
+      "learning_rate": 1.8088506453595577e-05,
+      "loss": 1.2955,
+      "step": 6220
+    },
+    {
+      "epoch": 0.47802704363859866,
+      "grad_norm": 2.8859567642211914,
+      "learning_rate": 1.8087891825445607e-05,
+      "loss": 1.3495,
+      "step": 6222
+    },
+    {
+      "epoch": 0.478180700676091,
+      "grad_norm": 2.9452872276306152,
+      "learning_rate": 1.808727719729564e-05,
+      "loss": 1.276,
+      "step": 6224
+    },
+    {
+      "epoch": 0.4783343577135833,
+      "grad_norm": 3.28800892829895,
+      "learning_rate": 1.8086662569145666e-05,
+      "loss": 1.4918,
+      "step": 6226
+    },
+    {
+      "epoch": 0.4784880147510756,
+      "grad_norm": 3.066276788711548,
+      "learning_rate": 1.80860479409957e-05,
+      "loss": 1.3771,
+      "step": 6228
+    },
+    {
+      "epoch": 0.4786416717885679,
+      "grad_norm": 2.8848133087158203,
+      "learning_rate": 1.8085433312845732e-05,
+      "loss": 1.3354,
+      "step": 6230
+    },
+    {
+      "epoch": 0.47879532882606024,
+      "grad_norm": 2.9447460174560547,
+      "learning_rate": 1.808481868469576e-05,
+      "loss": 1.4236,
+      "step": 6232
+    },
+    {
+      "epoch": 0.47894898586355256,
+      "grad_norm": 3.061072826385498,
+      "learning_rate": 1.808420405654579e-05,
+      "loss": 1.2652,
+      "step": 6234
+    },
+    {
+      "epoch": 0.4791026429010449,
+      "grad_norm": 3.3094377517700195,
+      "learning_rate": 1.808358942839582e-05,
+      "loss": 1.3274,
+      "step": 6236
+    },
+    {
+      "epoch": 0.4792562999385372,
+      "grad_norm": 2.868401050567627,
+      "learning_rate": 1.8082974800245854e-05,
+      "loss": 1.361,
+      "step": 6238
+    },
+    {
+      "epoch": 0.4794099569760295,
+      "grad_norm": 2.7703821659088135,
+      "learning_rate": 1.8082360172095884e-05,
+      "loss": 1.3177,
+      "step": 6240
+    },
+    {
+      "epoch": 0.4795636140135218,
+      "grad_norm": 2.5703728199005127,
+      "learning_rate": 1.8081745543945914e-05,
+      "loss": 1.3978,
+      "step": 6242
+    },
+    {
+      "epoch": 0.47971727105101414,
+      "grad_norm": 3.038760185241699,
+      "learning_rate": 1.8081130915795947e-05,
+      "loss": 1.3651,
+      "step": 6244
+    },
+    {
+      "epoch": 0.47987092808850645,
+      "grad_norm": 3.6976478099823,
+      "learning_rate": 1.8080516287645976e-05,
+      "loss": 1.4703,
+      "step": 6246
+    },
+    {
+      "epoch": 0.48002458512599877,
+      "grad_norm": 3.0694258213043213,
+      "learning_rate": 1.8079901659496006e-05,
+      "loss": 1.3212,
+      "step": 6248
+    },
+    {
+      "epoch": 0.4801782421634911,
+      "grad_norm": 3.3709535598754883,
+      "learning_rate": 1.807928703134604e-05,
+      "loss": 1.36,
+      "step": 6250
+    },
+    {
+      "epoch": 0.4803318992009834,
+      "grad_norm": 3.294551134109497,
+      "learning_rate": 1.8078672403196065e-05,
+      "loss": 1.5437,
+      "step": 6252
+    },
+    {
+      "epoch": 0.4804855562384757,
+      "grad_norm": 3.8295395374298096,
+      "learning_rate": 1.80780577750461e-05,
+      "loss": 1.2844,
+      "step": 6254
+    },
+    {
+      "epoch": 0.48063921327596804,
+      "grad_norm": 2.620631217956543,
+      "learning_rate": 1.8077443146896128e-05,
+      "loss": 1.2676,
+      "step": 6256
+    },
+    {
+      "epoch": 0.48079287031346035,
+      "grad_norm": 2.7182457447052,
+      "learning_rate": 1.8076828518746158e-05,
+      "loss": 1.4544,
+      "step": 6258
+    },
+    {
+      "epoch": 0.48094652735095267,
+      "grad_norm": 3.396521806716919,
+      "learning_rate": 1.807621389059619e-05,
+      "loss": 1.3121,
+      "step": 6260
+    },
+    {
+      "epoch": 0.481100184388445,
+      "grad_norm": 2.63437557220459,
+      "learning_rate": 1.807559926244622e-05,
+      "loss": 1.234,
+      "step": 6262
+    },
+    {
+      "epoch": 0.4812538414259373,
+      "grad_norm": 2.776506185531616,
+      "learning_rate": 1.8074984634296254e-05,
+      "loss": 1.4487,
+      "step": 6264
+    },
+    {
+      "epoch": 0.4814074984634296,
+      "grad_norm": 3.227975845336914,
+      "learning_rate": 1.8074370006146283e-05,
+      "loss": 1.2383,
+      "step": 6266
+    },
+    {
+      "epoch": 0.48156115550092193,
+      "grad_norm": 2.9529471397399902,
+      "learning_rate": 1.8073755377996313e-05,
+      "loss": 1.3515,
+      "step": 6268
+    },
+    {
+      "epoch": 0.48171481253841425,
+      "grad_norm": 3.3522536754608154,
+      "learning_rate": 1.8073140749846346e-05,
+      "loss": 1.3238,
+      "step": 6270
+    },
+    {
+      "epoch": 0.48186846957590657,
+      "grad_norm": 3.3435351848602295,
+      "learning_rate": 1.8072526121696376e-05,
+      "loss": 1.4359,
+      "step": 6272
+    },
+    {
+      "epoch": 0.4820221266133989,
+      "grad_norm": 2.8637821674346924,
+      "learning_rate": 1.8071911493546406e-05,
+      "loss": 1.2188,
+      "step": 6274
+    },
+    {
+      "epoch": 0.4821757836508912,
+      "grad_norm": 2.939358949661255,
+      "learning_rate": 1.807129686539644e-05,
+      "loss": 1.2918,
+      "step": 6276
+    },
+    {
+      "epoch": 0.4823294406883835,
+      "grad_norm": 3.1831789016723633,
+      "learning_rate": 1.8070682237246465e-05,
+      "loss": 1.4176,
+      "step": 6278
+    },
+    {
+      "epoch": 0.48248309772587583,
+      "grad_norm": 3.3042044639587402,
+      "learning_rate": 1.8070067609096498e-05,
+      "loss": 1.3979,
+      "step": 6280
+    },
+    {
+      "epoch": 0.48263675476336815,
+      "grad_norm": 3.7356269359588623,
+      "learning_rate": 1.8069452980946528e-05,
+      "loss": 1.3986,
+      "step": 6282
+    },
+    {
+      "epoch": 0.48279041180086046,
+      "grad_norm": 3.174906015396118,
+      "learning_rate": 1.806883835279656e-05,
+      "loss": 1.2884,
+      "step": 6284
+    },
+    {
+      "epoch": 0.4829440688383528,
+      "grad_norm": 2.9945216178894043,
+      "learning_rate": 1.806822372464659e-05,
+      "loss": 1.4439,
+      "step": 6286
+    },
+    {
+      "epoch": 0.4830977258758451,
+      "grad_norm": 3.113851308822632,
+      "learning_rate": 1.806760909649662e-05,
+      "loss": 1.3341,
+      "step": 6288
+    },
+    {
+      "epoch": 0.4832513829133374,
+      "grad_norm": 3.578928232192993,
+      "learning_rate": 1.8066994468346653e-05,
+      "loss": 1.4566,
+      "step": 6290
+    },
+    {
+      "epoch": 0.48340503995082973,
+      "grad_norm": 3.227860450744629,
+      "learning_rate": 1.8066379840196683e-05,
+      "loss": 1.3845,
+      "step": 6292
+    },
+    {
+      "epoch": 0.48355869698832205,
+      "grad_norm": 3.249185562133789,
+      "learning_rate": 1.8065765212046713e-05,
+      "loss": 1.3953,
+      "step": 6294
+    },
+    {
+      "epoch": 0.48371235402581436,
+      "grad_norm": 2.9868662357330322,
+      "learning_rate": 1.8065150583896746e-05,
+      "loss": 1.2176,
+      "step": 6296
+    },
+    {
+      "epoch": 0.4838660110633067,
+      "grad_norm": 2.748054265975952,
+      "learning_rate": 1.8064535955746775e-05,
+      "loss": 1.251,
+      "step": 6298
+    },
+    {
+      "epoch": 0.484019668100799,
+      "grad_norm": 3.2431583404541016,
+      "learning_rate": 1.8063921327596805e-05,
+      "loss": 1.3409,
+      "step": 6300
+    },
+    {
+      "epoch": 0.4841733251382913,
+      "grad_norm": 3.484886407852173,
+      "learning_rate": 1.8063306699446838e-05,
+      "loss": 1.3708,
+      "step": 6302
+    },
+    {
+      "epoch": 0.4843269821757836,
+      "grad_norm": 3.2029964923858643,
+      "learning_rate": 1.8062692071296868e-05,
+      "loss": 1.4806,
+      "step": 6304
+    },
+    {
+      "epoch": 0.48448063921327594,
+      "grad_norm": 2.78397274017334,
+      "learning_rate": 1.8062077443146897e-05,
+      "loss": 1.2184,
+      "step": 6306
+    },
+    {
+      "epoch": 0.48463429625076826,
+      "grad_norm": 2.9794437885284424,
+      "learning_rate": 1.8061462814996927e-05,
+      "loss": 1.4035,
+      "step": 6308
+    },
+    {
+      "epoch": 0.4847879532882606,
+      "grad_norm": 2.6800053119659424,
+      "learning_rate": 1.806084818684696e-05,
+      "loss": 1.3018,
+      "step": 6310
+    },
+    {
+      "epoch": 0.4849416103257529,
+      "grad_norm": 2.774409055709839,
+      "learning_rate": 1.806023355869699e-05,
+      "loss": 1.4247,
+      "step": 6312
+    },
+    {
+      "epoch": 0.48509526736324526,
+      "grad_norm": 3.3923118114471436,
+      "learning_rate": 1.805961893054702e-05,
+      "loss": 1.1499,
+      "step": 6314
+    },
+    {
+      "epoch": 0.4852489244007376,
+      "grad_norm": 2.7925145626068115,
+      "learning_rate": 1.8059004302397053e-05,
+      "loss": 1.2022,
+      "step": 6316
+    },
+    {
+      "epoch": 0.4854025814382299,
+      "grad_norm": 3.3250386714935303,
+      "learning_rate": 1.8058389674247082e-05,
+      "loss": 1.3755,
+      "step": 6318
+    },
+    {
+      "epoch": 0.4855562384757222,
+      "grad_norm": 3.531944990158081,
+      "learning_rate": 1.8057775046097112e-05,
+      "loss": 1.4049,
+      "step": 6320
+    },
+    {
+      "epoch": 0.48570989551321453,
+      "grad_norm": 3.0159146785736084,
+      "learning_rate": 1.8057160417947145e-05,
+      "loss": 1.4002,
+      "step": 6322
+    },
+    {
+      "epoch": 0.48586355255070685,
+      "grad_norm": 3.2475900650024414,
+      "learning_rate": 1.8056545789797175e-05,
+      "loss": 1.3624,
+      "step": 6324
+    },
+    {
+      "epoch": 0.48601720958819916,
+      "grad_norm": 3.2522149085998535,
+      "learning_rate": 1.8055931161647204e-05,
+      "loss": 1.3565,
+      "step": 6326
+    },
+    {
+      "epoch": 0.4861708666256915,
+      "grad_norm": 3.1061174869537354,
+      "learning_rate": 1.8055316533497237e-05,
+      "loss": 1.3549,
+      "step": 6328
+    },
+    {
+      "epoch": 0.4863245236631838,
+      "grad_norm": 3.162954330444336,
+      "learning_rate": 1.8054701905347267e-05,
+      "loss": 1.4947,
+      "step": 6330
+    },
+    {
+      "epoch": 0.4864781807006761,
+      "grad_norm": 3.179232597351074,
+      "learning_rate": 1.8054087277197297e-05,
+      "loss": 1.32,
+      "step": 6332
+    },
+    {
+      "epoch": 0.4866318377381684,
+      "grad_norm": 2.8867878913879395,
+      "learning_rate": 1.8053472649047327e-05,
+      "loss": 1.2925,
+      "step": 6334
+    },
+    {
+      "epoch": 0.48678549477566074,
+      "grad_norm": 3.4606575965881348,
+      "learning_rate": 1.805285802089736e-05,
+      "loss": 1.4834,
+      "step": 6336
+    },
+    {
+      "epoch": 0.48693915181315306,
+      "grad_norm": 3.118943214416504,
+      "learning_rate": 1.805224339274739e-05,
+      "loss": 1.227,
+      "step": 6338
+    },
+    {
+      "epoch": 0.4870928088506454,
+      "grad_norm": 3.1794466972351074,
+      "learning_rate": 1.805162876459742e-05,
+      "loss": 1.4092,
+      "step": 6340
+    },
+    {
+      "epoch": 0.4872464658881377,
+      "grad_norm": 2.719515562057495,
+      "learning_rate": 1.8051014136447452e-05,
+      "loss": 1.3534,
+      "step": 6342
+    },
+    {
+      "epoch": 0.48740012292563,
+      "grad_norm": 3.8740384578704834,
+      "learning_rate": 1.8050399508297482e-05,
+      "loss": 1.4607,
+      "step": 6344
+    },
+    {
+      "epoch": 0.4875537799631223,
+      "grad_norm": 3.121920585632324,
+      "learning_rate": 1.804978488014751e-05,
+      "loss": 1.365,
+      "step": 6346
+    },
+    {
+      "epoch": 0.48770743700061464,
+      "grad_norm": 2.8695993423461914,
+      "learning_rate": 1.8049170251997544e-05,
+      "loss": 1.2682,
+      "step": 6348
+    },
+    {
+      "epoch": 0.48786109403810696,
+      "grad_norm": 2.9940059185028076,
+      "learning_rate": 1.8048555623847574e-05,
+      "loss": 1.3376,
+      "step": 6350
+    },
+    {
+      "epoch": 0.4880147510755993,
+      "grad_norm": 3.0862960815429688,
+      "learning_rate": 1.8047940995697604e-05,
+      "loss": 1.3186,
+      "step": 6352
+    },
+    {
+      "epoch": 0.4881684081130916,
+      "grad_norm": 3.2420318126678467,
+      "learning_rate": 1.8047326367547634e-05,
+      "loss": 1.3098,
+      "step": 6354
+    },
+    {
+      "epoch": 0.4883220651505839,
+      "grad_norm": 2.9130239486694336,
+      "learning_rate": 1.8046711739397667e-05,
+      "loss": 1.3989,
+      "step": 6356
+    },
+    {
+      "epoch": 0.4884757221880762,
+      "grad_norm": 3.1051688194274902,
+      "learning_rate": 1.8046097111247696e-05,
+      "loss": 1.3173,
+      "step": 6358
+    },
+    {
+      "epoch": 0.48862937922556854,
+      "grad_norm": 2.9414143562316895,
+      "learning_rate": 1.8045482483097726e-05,
+      "loss": 1.3141,
+      "step": 6360
+    },
+    {
+      "epoch": 0.48878303626306085,
+      "grad_norm": 2.926856279373169,
+      "learning_rate": 1.804486785494776e-05,
+      "loss": 1.3659,
+      "step": 6362
+    },
+    {
+      "epoch": 0.48893669330055317,
+      "grad_norm": 3.0181140899658203,
+      "learning_rate": 1.804425322679779e-05,
+      "loss": 1.4083,
+      "step": 6364
+    },
+    {
+      "epoch": 0.4890903503380455,
+      "grad_norm": 2.822953462600708,
+      "learning_rate": 1.804363859864782e-05,
+      "loss": 1.3084,
+      "step": 6366
+    },
+    {
+      "epoch": 0.4892440073755378,
+      "grad_norm": 3.2265994548797607,
+      "learning_rate": 1.804302397049785e-05,
+      "loss": 1.4421,
+      "step": 6368
+    },
+    {
+      "epoch": 0.4893976644130301,
+      "grad_norm": 2.932751417160034,
+      "learning_rate": 1.804240934234788e-05,
+      "loss": 1.3666,
+      "step": 6370
+    },
+    {
+      "epoch": 0.48955132145052244,
+      "grad_norm": 3.190852403640747,
+      "learning_rate": 1.804179471419791e-05,
+      "loss": 1.4584,
+      "step": 6372
+    },
+    {
+      "epoch": 0.48970497848801475,
+      "grad_norm": 2.9671835899353027,
+      "learning_rate": 1.8041180086047944e-05,
+      "loss": 1.4951,
+      "step": 6374
+    },
+    {
+      "epoch": 0.48985863552550707,
+      "grad_norm": 3.1517322063446045,
+      "learning_rate": 1.8040565457897974e-05,
+      "loss": 1.4506,
+      "step": 6376
+    },
+    {
+      "epoch": 0.4900122925629994,
+      "grad_norm": 3.0485892295837402,
+      "learning_rate": 1.8039950829748003e-05,
+      "loss": 1.3996,
+      "step": 6378
+    },
+    {
+      "epoch": 0.4901659496004917,
+      "grad_norm": 2.753948450088501,
+      "learning_rate": 1.8039336201598033e-05,
+      "loss": 1.4296,
+      "step": 6380
+    },
+    {
+      "epoch": 0.490319606637984,
+      "grad_norm": 2.8919365406036377,
+      "learning_rate": 1.8038721573448066e-05,
+      "loss": 1.3162,
+      "step": 6382
+    },
+    {
+      "epoch": 0.49047326367547633,
+      "grad_norm": 2.782630205154419,
+      "learning_rate": 1.8038106945298096e-05,
+      "loss": 1.3588,
+      "step": 6384
+    },
+    {
+      "epoch": 0.49062692071296865,
+      "grad_norm": 2.8316001892089844,
+      "learning_rate": 1.8037492317148125e-05,
+      "loss": 1.4375,
+      "step": 6386
+    },
+    {
+      "epoch": 0.49078057775046097,
+      "grad_norm": 2.9996912479400635,
+      "learning_rate": 1.803687768899816e-05,
+      "loss": 1.3141,
+      "step": 6388
+    },
+    {
+      "epoch": 0.4909342347879533,
+      "grad_norm": 2.8670809268951416,
+      "learning_rate": 1.8036263060848188e-05,
+      "loss": 1.424,
+      "step": 6390
+    },
+    {
+      "epoch": 0.4910878918254456,
+      "grad_norm": 3.0059220790863037,
+      "learning_rate": 1.8035648432698218e-05,
+      "loss": 1.3933,
+      "step": 6392
+    },
+    {
+      "epoch": 0.4912415488629379,
+      "grad_norm": 3.1974833011627197,
+      "learning_rate": 1.803503380454825e-05,
+      "loss": 1.3125,
+      "step": 6394
+    },
+    {
+      "epoch": 0.49139520590043023,
+      "grad_norm": 2.683246612548828,
+      "learning_rate": 1.803441917639828e-05,
+      "loss": 1.258,
+      "step": 6396
+    },
+    {
+      "epoch": 0.49154886293792255,
+      "grad_norm": 3.330538034439087,
+      "learning_rate": 1.803380454824831e-05,
+      "loss": 1.4212,
+      "step": 6398
+    },
+    {
+      "epoch": 0.49170251997541486,
+      "grad_norm": 3.116828680038452,
+      "learning_rate": 1.8033189920098343e-05,
+      "loss": 1.3991,
+      "step": 6400
+    },
+    {
+      "epoch": 0.4918561770129072,
+      "grad_norm": 3.45032000541687,
+      "learning_rate": 1.8032575291948373e-05,
+      "loss": 1.3591,
+      "step": 6402
+    },
+    {
+      "epoch": 0.4920098340503995,
+      "grad_norm": 3.455242872238159,
+      "learning_rate": 1.8031960663798403e-05,
+      "loss": 1.3312,
+      "step": 6404
+    },
+    {
+      "epoch": 0.4921634910878918,
+      "grad_norm": 2.821232557296753,
+      "learning_rate": 1.8031346035648432e-05,
+      "loss": 1.2594,
+      "step": 6406
+    },
+    {
+      "epoch": 0.49231714812538413,
+      "grad_norm": 3.163733720779419,
+      "learning_rate": 1.8030731407498465e-05,
+      "loss": 1.3575,
+      "step": 6408
+    },
+    {
+      "epoch": 0.49247080516287645,
+      "grad_norm": 3.2715537548065186,
+      "learning_rate": 1.8030116779348495e-05,
+      "loss": 1.4909,
+      "step": 6410
+    },
+    {
+      "epoch": 0.49262446220036876,
+      "grad_norm": 3.1828911304473877,
+      "learning_rate": 1.8029502151198525e-05,
+      "loss": 1.3965,
+      "step": 6412
+    },
+    {
+      "epoch": 0.4927781192378611,
+      "grad_norm": 3.2268998622894287,
+      "learning_rate": 1.8028887523048558e-05,
+      "loss": 1.3785,
+      "step": 6414
+    },
+    {
+      "epoch": 0.4929317762753534,
+      "grad_norm": 3.106019973754883,
+      "learning_rate": 1.8028272894898588e-05,
+      "loss": 1.2841,
+      "step": 6416
+    },
+    {
+      "epoch": 0.4930854333128457,
+      "grad_norm": 3.220978260040283,
+      "learning_rate": 1.8027658266748617e-05,
+      "loss": 1.5033,
+      "step": 6418
+    },
+    {
+      "epoch": 0.493239090350338,
+      "grad_norm": 3.288722276687622,
+      "learning_rate": 1.802704363859865e-05,
+      "loss": 1.3628,
+      "step": 6420
+    },
+    {
+      "epoch": 0.49339274738783034,
+      "grad_norm": 3.134910821914673,
+      "learning_rate": 1.802642901044868e-05,
+      "loss": 1.3892,
+      "step": 6422
+    },
+    {
+      "epoch": 0.49354640442532266,
+      "grad_norm": 2.868943929672241,
+      "learning_rate": 1.802581438229871e-05,
+      "loss": 1.4807,
+      "step": 6424
+    },
+    {
+      "epoch": 0.493700061462815,
+      "grad_norm": 3.257479190826416,
+      "learning_rate": 1.8025199754148743e-05,
+      "loss": 1.3037,
+      "step": 6426
+    },
+    {
+      "epoch": 0.4938537185003073,
+      "grad_norm": 2.6661643981933594,
+      "learning_rate": 1.8024585125998772e-05,
+      "loss": 1.3603,
+      "step": 6428
+    },
+    {
+      "epoch": 0.4940073755377996,
+      "grad_norm": 3.075345039367676,
+      "learning_rate": 1.8023970497848806e-05,
+      "loss": 1.3357,
+      "step": 6430
+    },
+    {
+      "epoch": 0.4941610325752919,
+      "grad_norm": 3.3657965660095215,
+      "learning_rate": 1.8023355869698832e-05,
+      "loss": 1.3358,
+      "step": 6432
+    },
+    {
+      "epoch": 0.49431468961278424,
+      "grad_norm": 3.102015495300293,
+      "learning_rate": 1.8022741241548865e-05,
+      "loss": 1.4151,
+      "step": 6434
+    },
+    {
+      "epoch": 0.49446834665027656,
+      "grad_norm": 2.722320795059204,
+      "learning_rate": 1.8022126613398895e-05,
+      "loss": 1.3975,
+      "step": 6436
+    },
+    {
+      "epoch": 0.4946220036877689,
+      "grad_norm": 3.373051166534424,
+      "learning_rate": 1.8021511985248924e-05,
+      "loss": 1.2497,
+      "step": 6438
+    },
+    {
+      "epoch": 0.4947756607252612,
+      "grad_norm": 2.7113919258117676,
+      "learning_rate": 1.8020897357098957e-05,
+      "loss": 1.3809,
+      "step": 6440
+    },
+    {
+      "epoch": 0.4949293177627535,
+      "grad_norm": 3.1061747074127197,
+      "learning_rate": 1.8020282728948987e-05,
+      "loss": 1.4414,
+      "step": 6442
+    },
+    {
+      "epoch": 0.4950829748002459,
+      "grad_norm": 3.091012477874756,
+      "learning_rate": 1.8019668100799017e-05,
+      "loss": 1.3563,
+      "step": 6444
+    },
+    {
+      "epoch": 0.4952366318377382,
+      "grad_norm": 2.845564126968384,
+      "learning_rate": 1.801905347264905e-05,
+      "loss": 1.3844,
+      "step": 6446
+    },
+    {
+      "epoch": 0.4953902888752305,
+      "grad_norm": 3.5024054050445557,
+      "learning_rate": 1.801843884449908e-05,
+      "loss": 1.4408,
+      "step": 6448
+    },
+    {
+      "epoch": 0.4955439459127228,
+      "grad_norm": 2.915093421936035,
+      "learning_rate": 1.8017824216349113e-05,
+      "loss": 1.2814,
+      "step": 6450
+    },
+    {
+      "epoch": 0.49569760295021514,
+      "grad_norm": 3.042811632156372,
+      "learning_rate": 1.801720958819914e-05,
+      "loss": 1.5103,
+      "step": 6452
+    },
+    {
+      "epoch": 0.49585125998770746,
+      "grad_norm": 2.851787805557251,
+      "learning_rate": 1.8016594960049172e-05,
+      "loss": 1.4586,
+      "step": 6454
+    },
+    {
+      "epoch": 0.4960049170251998,
+      "grad_norm": 3.1319057941436768,
+      "learning_rate": 1.80159803318992e-05,
+      "loss": 1.2987,
+      "step": 6456
+    },
+    {
+      "epoch": 0.4961585740626921,
+      "grad_norm": 2.650089740753174,
+      "learning_rate": 1.801536570374923e-05,
+      "loss": 1.3441,
+      "step": 6458
+    },
+    {
+      "epoch": 0.4963122311001844,
+      "grad_norm": 3.493260383605957,
+      "learning_rate": 1.8014751075599264e-05,
+      "loss": 1.6013,
+      "step": 6460
+    },
+    {
+      "epoch": 0.4964658881376767,
+      "grad_norm": 2.9392683506011963,
+      "learning_rate": 1.8014136447449294e-05,
+      "loss": 1.3622,
+      "step": 6462
+    },
+    {
+      "epoch": 0.49661954517516904,
+      "grad_norm": 3.106326103210449,
+      "learning_rate": 1.8013521819299324e-05,
+      "loss": 1.3347,
+      "step": 6464
+    },
+    {
+      "epoch": 0.49677320221266136,
+      "grad_norm": 2.914437770843506,
+      "learning_rate": 1.8012907191149357e-05,
+      "loss": 1.2599,
+      "step": 6466
+    },
+    {
+      "epoch": 0.4969268592501537,
+      "grad_norm": 2.910841464996338,
+      "learning_rate": 1.8012292562999386e-05,
+      "loss": 1.3622,
+      "step": 6468
+    },
+    {
+      "epoch": 0.497080516287646,
+      "grad_norm": 2.89884352684021,
+      "learning_rate": 1.801167793484942e-05,
+      "loss": 1.3582,
+      "step": 6470
+    },
+    {
+      "epoch": 0.4972341733251383,
+      "grad_norm": 2.8821210861206055,
+      "learning_rate": 1.801106330669945e-05,
+      "loss": 1.4164,
+      "step": 6472
+    },
+    {
+      "epoch": 0.4973878303626306,
+      "grad_norm": 2.601987838745117,
+      "learning_rate": 1.801044867854948e-05,
+      "loss": 1.5184,
+      "step": 6474
+    },
+    {
+      "epoch": 0.49754148740012294,
+      "grad_norm": 2.958704710006714,
+      "learning_rate": 1.8009834050399512e-05,
+      "loss": 1.3903,
+      "step": 6476
+    },
+    {
+      "epoch": 0.49769514443761526,
+      "grad_norm": 2.9979889392852783,
+      "learning_rate": 1.8009219422249538e-05,
+      "loss": 1.2592,
+      "step": 6478
+    },
+    {
+      "epoch": 0.49784880147510757,
+      "grad_norm": 2.916813611984253,
+      "learning_rate": 1.800860479409957e-05,
+      "loss": 1.3302,
+      "step": 6480
+    },
+    {
+      "epoch": 0.4980024585125999,
+      "grad_norm": 3.3828582763671875,
+      "learning_rate": 1.80079901659496e-05,
+      "loss": 1.4222,
+      "step": 6482
+    },
+    {
+      "epoch": 0.4981561155500922,
+      "grad_norm": 3.1152589321136475,
+      "learning_rate": 1.800737553779963e-05,
+      "loss": 1.3731,
+      "step": 6484
+    },
+    {
+      "epoch": 0.4983097725875845,
+      "grad_norm": 2.974968671798706,
+      "learning_rate": 1.8006760909649664e-05,
+      "loss": 1.2744,
+      "step": 6486
+    },
+    {
+      "epoch": 0.49846342962507684,
+      "grad_norm": 3.652846336364746,
+      "learning_rate": 1.8006146281499693e-05,
+      "loss": 1.4291,
+      "step": 6488
+    },
+    {
+      "epoch": 0.49861708666256915,
+      "grad_norm": 2.827791213989258,
+      "learning_rate": 1.8005531653349727e-05,
+      "loss": 1.2982,
+      "step": 6490
+    },
+    {
+      "epoch": 0.49877074370006147,
+      "grad_norm": 3.1473135948181152,
+      "learning_rate": 1.8004917025199756e-05,
+      "loss": 1.3051,
+      "step": 6492
+    },
+    {
+      "epoch": 0.4989244007375538,
+      "grad_norm": 3.156839370727539,
+      "learning_rate": 1.8004302397049786e-05,
+      "loss": 1.4366,
+      "step": 6494
+    },
+    {
+      "epoch": 0.4990780577750461,
+      "grad_norm": 2.8511626720428467,
+      "learning_rate": 1.800368776889982e-05,
+      "loss": 1.3214,
+      "step": 6496
+    },
+    {
+      "epoch": 0.4992317148125384,
+      "grad_norm": 2.7867062091827393,
+      "learning_rate": 1.800307314074985e-05,
+      "loss": 1.3791,
+      "step": 6498
+    },
+    {
+      "epoch": 0.49938537185003073,
+      "grad_norm": 3.0567257404327393,
+      "learning_rate": 1.800245851259988e-05,
+      "loss": 1.3549,
+      "step": 6500
+    },
+    {
+      "epoch": 0.49953902888752305,
+      "grad_norm": 2.9443788528442383,
+      "learning_rate": 1.800184388444991e-05,
+      "loss": 1.3798,
+      "step": 6502
+    },
+    {
+      "epoch": 0.49969268592501537,
+      "grad_norm": 2.903205156326294,
+      "learning_rate": 1.8001229256299938e-05,
+      "loss": 1.3004,
+      "step": 6504
+    },
+    {
+      "epoch": 0.4998463429625077,
+      "grad_norm": 3.2189865112304688,
+      "learning_rate": 1.800061462814997e-05,
+      "loss": 1.3023,
+      "step": 6506
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1087801456451416,
+      "learning_rate": 1.8e-05,
+      "loss": 1.508,
+      "step": 6508
+    },
+    {
+      "epoch": 0.5001536570374924,
+      "grad_norm": 3.1676154136657715,
+      "learning_rate": 1.799938537185003e-05,
+      "loss": 1.4198,
+      "step": 6510
+    },
+    {
+      "epoch": 0.5003073140749846,
+      "grad_norm": 2.717998504638672,
+      "learning_rate": 1.7998770743700063e-05,
+      "loss": 1.3673,
+      "step": 6512
+    },
+    {
+      "epoch": 0.500460971112477,
+      "grad_norm": 2.779644012451172,
+      "learning_rate": 1.7998156115550093e-05,
+      "loss": 1.493,
+      "step": 6514
+    },
+    {
+      "epoch": 0.5006146281499693,
+      "grad_norm": 2.706282377243042,
+      "learning_rate": 1.7997541487400126e-05,
+      "loss": 1.2036,
+      "step": 6516
+    },
+    {
+      "epoch": 0.5007682851874616,
+      "grad_norm": 2.846618175506592,
+      "learning_rate": 1.7996926859250156e-05,
+      "loss": 1.431,
+      "step": 6518
+    },
+    {
+      "epoch": 0.5009219422249539,
+      "grad_norm": 2.9443068504333496,
+      "learning_rate": 1.7996312231100185e-05,
+      "loss": 1.2797,
+      "step": 6520
+    },
+    {
+      "epoch": 0.5010755992624463,
+      "grad_norm": 3.471630334854126,
+      "learning_rate": 1.799569760295022e-05,
+      "loss": 1.436,
+      "step": 6522
+    },
+    {
+      "epoch": 0.5012292562999385,
+      "grad_norm": 3.2783944606781006,
+      "learning_rate": 1.7995082974800248e-05,
+      "loss": 1.4904,
+      "step": 6524
+    },
+    {
+      "epoch": 0.5013829133374309,
+      "grad_norm": 3.51324725151062,
+      "learning_rate": 1.7994468346650278e-05,
+      "loss": 1.3418,
+      "step": 6526
+    },
+    {
+      "epoch": 0.5015365703749232,
+      "grad_norm": 2.648725748062134,
+      "learning_rate": 1.799385371850031e-05,
+      "loss": 1.2583,
+      "step": 6528
+    },
+    {
+      "epoch": 0.5016902274124155,
+      "grad_norm": 2.8676857948303223,
+      "learning_rate": 1.7993239090350337e-05,
+      "loss": 1.3348,
+      "step": 6530
+    },
+    {
+      "epoch": 0.5018438844499078,
+      "grad_norm": 3.1224286556243896,
+      "learning_rate": 1.799262446220037e-05,
+      "loss": 1.3221,
+      "step": 6532
+    },
+    {
+      "epoch": 0.5019975414874002,
+      "grad_norm": 3.003473997116089,
+      "learning_rate": 1.79920098340504e-05,
+      "loss": 1.3466,
+      "step": 6534
+    },
+    {
+      "epoch": 0.5021511985248924,
+      "grad_norm": 2.8612453937530518,
+      "learning_rate": 1.7991395205900433e-05,
+      "loss": 1.2505,
+      "step": 6536
+    },
+    {
+      "epoch": 0.5023048555623848,
+      "grad_norm": 2.9750072956085205,
+      "learning_rate": 1.7990780577750463e-05,
+      "loss": 1.3482,
+      "step": 6538
+    },
+    {
+      "epoch": 0.5024585125998771,
+      "grad_norm": 2.6988656520843506,
+      "learning_rate": 1.7990165949600492e-05,
+      "loss": 1.2996,
+      "step": 6540
+    },
+    {
+      "epoch": 0.5026121696373694,
+      "grad_norm": 3.0278279781341553,
+      "learning_rate": 1.7989551321450525e-05,
+      "loss": 1.3919,
+      "step": 6542
+    },
+    {
+      "epoch": 0.5027658266748617,
+      "grad_norm": 2.9088683128356934,
+      "learning_rate": 1.7988936693300555e-05,
+      "loss": 1.4706,
+      "step": 6544
+    },
+    {
+      "epoch": 0.5029194837123541,
+      "grad_norm": 2.8318252563476562,
+      "learning_rate": 1.7988322065150585e-05,
+      "loss": 1.4209,
+      "step": 6546
+    },
+    {
+      "epoch": 0.5030731407498463,
+      "grad_norm": 2.994474411010742,
+      "learning_rate": 1.7987707437000618e-05,
+      "loss": 1.3159,
+      "step": 6548
+    },
+    {
+      "epoch": 0.5032267977873387,
+      "grad_norm": 2.9801907539367676,
+      "learning_rate": 1.7987092808850644e-05,
+      "loss": 1.2645,
+      "step": 6550
+    },
+    {
+      "epoch": 0.503380454824831,
+      "grad_norm": 3.4261326789855957,
+      "learning_rate": 1.7986478180700677e-05,
+      "loss": 1.4416,
+      "step": 6552
+    },
+    {
+      "epoch": 0.5035341118623233,
+      "grad_norm": 2.65868878364563,
+      "learning_rate": 1.798586355255071e-05,
+      "loss": 1.2989,
+      "step": 6554
+    },
+    {
+      "epoch": 0.5036877688998156,
+      "grad_norm": 3.106370210647583,
+      "learning_rate": 1.798524892440074e-05,
+      "loss": 1.3748,
+      "step": 6556
+    },
+    {
+      "epoch": 0.503841425937308,
+      "grad_norm": 2.6410956382751465,
+      "learning_rate": 1.798463429625077e-05,
+      "loss": 1.3152,
+      "step": 6558
+    },
+    {
+      "epoch": 0.5039950829748002,
+      "grad_norm": 2.8190648555755615,
+      "learning_rate": 1.79840196681008e-05,
+      "loss": 1.4078,
+      "step": 6560
+    },
+    {
+      "epoch": 0.5041487400122926,
+      "grad_norm": 2.9152209758758545,
+      "learning_rate": 1.7983405039950832e-05,
+      "loss": 1.2513,
+      "step": 6562
+    },
+    {
+      "epoch": 0.5043023970497849,
+      "grad_norm": 3.070107936859131,
+      "learning_rate": 1.7982790411800862e-05,
+      "loss": 1.4488,
+      "step": 6564
+    },
+    {
+      "epoch": 0.5044560540872772,
+      "grad_norm": 2.9198601245880127,
+      "learning_rate": 1.7982175783650892e-05,
+      "loss": 1.2694,
+      "step": 6566
+    },
+    {
+      "epoch": 0.5046097111247695,
+      "grad_norm": 3.084261417388916,
+      "learning_rate": 1.7981561155500925e-05,
+      "loss": 1.3638,
+      "step": 6568
+    },
+    {
+      "epoch": 0.5047633681622619,
+      "grad_norm": 3.0732178688049316,
+      "learning_rate": 1.7980946527350955e-05,
+      "loss": 1.2477,
+      "step": 6570
+    },
+    {
+      "epoch": 0.5049170251997541,
+      "grad_norm": 3.3027431964874268,
+      "learning_rate": 1.7980331899200984e-05,
+      "loss": 1.335,
+      "step": 6572
+    },
+    {
+      "epoch": 0.5050706822372465,
+      "grad_norm": 3.3048105239868164,
+      "learning_rate": 1.7979717271051017e-05,
+      "loss": 1.4374,
+      "step": 6574
+    },
+    {
+      "epoch": 0.5052243392747388,
+      "grad_norm": 3.292642831802368,
+      "learning_rate": 1.7979102642901047e-05,
+      "loss": 1.353,
+      "step": 6576
+    },
+    {
+      "epoch": 0.5053779963122311,
+      "grad_norm": 3.257822036743164,
+      "learning_rate": 1.7978488014751077e-05,
+      "loss": 1.3081,
+      "step": 6578
+    },
+    {
+      "epoch": 0.5055316533497234,
+      "grad_norm": 2.796616792678833,
+      "learning_rate": 1.7977873386601106e-05,
+      "loss": 1.3497,
+      "step": 6580
+    },
+    {
+      "epoch": 0.5056853103872158,
+      "grad_norm": 3.194371223449707,
+      "learning_rate": 1.797725875845114e-05,
+      "loss": 1.3587,
+      "step": 6582
+    },
+    {
+      "epoch": 0.505838967424708,
+      "grad_norm": 2.6798551082611084,
+      "learning_rate": 1.797664413030117e-05,
+      "loss": 1.3413,
+      "step": 6584
+    },
+    {
+      "epoch": 0.5059926244622004,
+      "grad_norm": 3.123992681503296,
+      "learning_rate": 1.79760295021512e-05,
+      "loss": 1.2649,
+      "step": 6586
+    },
+    {
+      "epoch": 0.5061462814996927,
+      "grad_norm": 3.01867413520813,
+      "learning_rate": 1.7975414874001232e-05,
+      "loss": 1.3729,
+      "step": 6588
+    },
+    {
+      "epoch": 0.506299938537185,
+      "grad_norm": 3.5075738430023193,
+      "learning_rate": 1.797480024585126e-05,
+      "loss": 1.3458,
+      "step": 6590
+    },
+    {
+      "epoch": 0.5064535955746773,
+      "grad_norm": 3.0404345989227295,
+      "learning_rate": 1.797418561770129e-05,
+      "loss": 1.4686,
+      "step": 6592
+    },
+    {
+      "epoch": 0.5066072526121697,
+      "grad_norm": 2.7560205459594727,
+      "learning_rate": 1.7973570989551324e-05,
+      "loss": 1.3492,
+      "step": 6594
+    },
+    {
+      "epoch": 0.5067609096496619,
+      "grad_norm": 3.127852439880371,
+      "learning_rate": 1.7972956361401354e-05,
+      "loss": 1.3995,
+      "step": 6596
+    },
+    {
+      "epoch": 0.5069145666871543,
+      "grad_norm": 2.963434934616089,
+      "learning_rate": 1.7972341733251384e-05,
+      "loss": 1.3773,
+      "step": 6598
+    },
+    {
+      "epoch": 0.5070682237246465,
+      "grad_norm": 3.146491050720215,
+      "learning_rate": 1.7971727105101417e-05,
+      "loss": 1.282,
+      "step": 6600
+    },
+    {
+      "epoch": 0.5072218807621389,
+      "grad_norm": 2.8593227863311768,
+      "learning_rate": 1.7971112476951446e-05,
+      "loss": 1.2804,
+      "step": 6602
+    },
+    {
+      "epoch": 0.5073755377996312,
+      "grad_norm": 3.011699914932251,
+      "learning_rate": 1.7970497848801476e-05,
+      "loss": 1.3739,
+      "step": 6604
+    },
+    {
+      "epoch": 0.5075291948371236,
+      "grad_norm": 3.7797420024871826,
+      "learning_rate": 1.7969883220651506e-05,
+      "loss": 1.4324,
+      "step": 6606
+    },
+    {
+      "epoch": 0.5076828518746158,
+      "grad_norm": 2.8261046409606934,
+      "learning_rate": 1.796926859250154e-05,
+      "loss": 1.3562,
+      "step": 6608
+    },
+    {
+      "epoch": 0.5078365089121082,
+      "grad_norm": 2.9160585403442383,
+      "learning_rate": 1.796865396435157e-05,
+      "loss": 1.5061,
+      "step": 6610
+    },
+    {
+      "epoch": 0.5079901659496004,
+      "grad_norm": 3.1277146339416504,
+      "learning_rate": 1.7968039336201598e-05,
+      "loss": 1.3759,
+      "step": 6612
+    },
+    {
+      "epoch": 0.5081438229870928,
+      "grad_norm": 3.4111275672912598,
+      "learning_rate": 1.796742470805163e-05,
+      "loss": 1.3682,
+      "step": 6614
+    },
+    {
+      "epoch": 0.5082974800245851,
+      "grad_norm": 3.055345296859741,
+      "learning_rate": 1.796681007990166e-05,
+      "loss": 1.4154,
+      "step": 6616
+    },
+    {
+      "epoch": 0.5084511370620775,
+      "grad_norm": 3.3309273719787598,
+      "learning_rate": 1.796619545175169e-05,
+      "loss": 1.3364,
+      "step": 6618
+    },
+    {
+      "epoch": 0.5086047940995697,
+      "grad_norm": 2.7727885246276855,
+      "learning_rate": 1.7965580823601724e-05,
+      "loss": 1.2019,
+      "step": 6620
+    },
+    {
+      "epoch": 0.5087584511370621,
+      "grad_norm": 3.5600733757019043,
+      "learning_rate": 1.7964966195451753e-05,
+      "loss": 1.4411,
+      "step": 6622
+    },
+    {
+      "epoch": 0.5089121081745543,
+      "grad_norm": 3.1073739528656006,
+      "learning_rate": 1.7964351567301783e-05,
+      "loss": 1.4255,
+      "step": 6624
+    },
+    {
+      "epoch": 0.5090657652120467,
+      "grad_norm": 2.9542529582977295,
+      "learning_rate": 1.7963736939151816e-05,
+      "loss": 1.2744,
+      "step": 6626
+    },
+    {
+      "epoch": 0.509219422249539,
+      "grad_norm": 3.2199599742889404,
+      "learning_rate": 1.7963122311001846e-05,
+      "loss": 1.289,
+      "step": 6628
+    },
+    {
+      "epoch": 0.5093730792870313,
+      "grad_norm": 3.095400333404541,
+      "learning_rate": 1.7962507682851876e-05,
+      "loss": 1.3363,
+      "step": 6630
+    },
+    {
+      "epoch": 0.5095267363245236,
+      "grad_norm": 3.337625503540039,
+      "learning_rate": 1.7961893054701905e-05,
+      "loss": 1.4096,
+      "step": 6632
+    },
+    {
+      "epoch": 0.509680393362016,
+      "grad_norm": 3.3786509037017822,
+      "learning_rate": 1.7961278426551938e-05,
+      "loss": 1.3938,
+      "step": 6634
+    },
+    {
+      "epoch": 0.5098340503995082,
+      "grad_norm": 2.652902364730835,
+      "learning_rate": 1.7960663798401968e-05,
+      "loss": 1.2738,
+      "step": 6636
+    },
+    {
+      "epoch": 0.5099877074370006,
+      "grad_norm": 3.4588985443115234,
+      "learning_rate": 1.7960049170251998e-05,
+      "loss": 1.2758,
+      "step": 6638
+    },
+    {
+      "epoch": 0.510141364474493,
+      "grad_norm": 3.1711127758026123,
+      "learning_rate": 1.795943454210203e-05,
+      "loss": 1.3747,
+      "step": 6640
+    },
+    {
+      "epoch": 0.5102950215119852,
+      "grad_norm": 3.4093017578125,
+      "learning_rate": 1.795881991395206e-05,
+      "loss": 1.2861,
+      "step": 6642
+    },
+    {
+      "epoch": 0.5104486785494776,
+      "grad_norm": 3.0328543186187744,
+      "learning_rate": 1.795820528580209e-05,
+      "loss": 1.5437,
+      "step": 6644
+    },
+    {
+      "epoch": 0.5106023355869699,
+      "grad_norm": 2.8662827014923096,
+      "learning_rate": 1.7957590657652123e-05,
+      "loss": 1.3798,
+      "step": 6646
+    },
+    {
+      "epoch": 0.5107559926244623,
+      "grad_norm": 2.6307129859924316,
+      "learning_rate": 1.7956976029502153e-05,
+      "loss": 1.1287,
+      "step": 6648
+    },
+    {
+      "epoch": 0.5109096496619545,
+      "grad_norm": 3.1315627098083496,
+      "learning_rate": 1.7956361401352183e-05,
+      "loss": 1.4458,
+      "step": 6650
+    },
+    {
+      "epoch": 0.5110633066994469,
+      "grad_norm": 3.4800686836242676,
+      "learning_rate": 1.7955746773202216e-05,
+      "loss": 1.5149,
+      "step": 6652
+    },
+    {
+      "epoch": 0.5112169637369391,
+      "grad_norm": 3.0013492107391357,
+      "learning_rate": 1.7955132145052245e-05,
+      "loss": 1.3806,
+      "step": 6654
+    },
+    {
+      "epoch": 0.5113706207744315,
+      "grad_norm": 3.450124502182007,
+      "learning_rate": 1.7954517516902275e-05,
+      "loss": 1.4322,
+      "step": 6656
+    },
+    {
+      "epoch": 0.5115242778119238,
+      "grad_norm": 2.9068808555603027,
+      "learning_rate": 1.7953902888752305e-05,
+      "loss": 1.3998,
+      "step": 6658
+    },
+    {
+      "epoch": 0.5116779348494161,
+      "grad_norm": 2.6469430923461914,
+      "learning_rate": 1.7953288260602338e-05,
+      "loss": 1.4466,
+      "step": 6660
+    },
+    {
+      "epoch": 0.5118315918869084,
+      "grad_norm": 2.9947774410247803,
+      "learning_rate": 1.7952673632452367e-05,
+      "loss": 1.3114,
+      "step": 6662
+    },
+    {
+      "epoch": 0.5119852489244008,
+      "grad_norm": 3.5101516246795654,
+      "learning_rate": 1.7952059004302397e-05,
+      "loss": 1.3693,
+      "step": 6664
+    },
+    {
+      "epoch": 0.512138905961893,
+      "grad_norm": 3.2445108890533447,
+      "learning_rate": 1.795144437615243e-05,
+      "loss": 1.3274,
+      "step": 6666
+    },
+    {
+      "epoch": 0.5122925629993854,
+      "grad_norm": 3.050546169281006,
+      "learning_rate": 1.795082974800246e-05,
+      "loss": 1.3715,
+      "step": 6668
+    },
+    {
+      "epoch": 0.5124462200368777,
+      "grad_norm": 3.0119810104370117,
+      "learning_rate": 1.795021511985249e-05,
+      "loss": 1.3978,
+      "step": 6670
+    },
+    {
+      "epoch": 0.51259987707437,
+      "grad_norm": 3.514958381652832,
+      "learning_rate": 1.7949600491702523e-05,
+      "loss": 1.4204,
+      "step": 6672
+    },
+    {
+      "epoch": 0.5127535341118623,
+      "grad_norm": 2.9083685874938965,
+      "learning_rate": 1.7948985863552552e-05,
+      "loss": 1.2959,
+      "step": 6674
+    },
+    {
+      "epoch": 0.5129071911493547,
+      "grad_norm": 2.6381547451019287,
+      "learning_rate": 1.7948371235402582e-05,
+      "loss": 1.3359,
+      "step": 6676
+    },
+    {
+      "epoch": 0.5130608481868469,
+      "grad_norm": 3.478752374649048,
+      "learning_rate": 1.794775660725261e-05,
+      "loss": 1.4583,
+      "step": 6678
+    },
+    {
+      "epoch": 0.5132145052243393,
+      "grad_norm": 2.834408760070801,
+      "learning_rate": 1.7947141979102645e-05,
+      "loss": 1.3429,
+      "step": 6680
+    },
+    {
+      "epoch": 0.5133681622618316,
+      "grad_norm": 3.0341265201568604,
+      "learning_rate": 1.7946527350952674e-05,
+      "loss": 1.3716,
+      "step": 6682
+    },
+    {
+      "epoch": 0.513521819299324,
+      "grad_norm": 2.8251383304595947,
+      "learning_rate": 1.7945912722802704e-05,
+      "loss": 1.5223,
+      "step": 6684
+    },
+    {
+      "epoch": 0.5136754763368162,
+      "grad_norm": 3.265641212463379,
+      "learning_rate": 1.7945298094652737e-05,
+      "loss": 1.269,
+      "step": 6686
+    },
+    {
+      "epoch": 0.5138291333743086,
+      "grad_norm": 3.247816562652588,
+      "learning_rate": 1.7944683466502767e-05,
+      "loss": 1.2121,
+      "step": 6688
+    },
+    {
+      "epoch": 0.5139827904118008,
+      "grad_norm": 3.2081658840179443,
+      "learning_rate": 1.7944068838352797e-05,
+      "loss": 1.2807,
+      "step": 6690
+    },
+    {
+      "epoch": 0.5141364474492932,
+      "grad_norm": 3.1236660480499268,
+      "learning_rate": 1.794345421020283e-05,
+      "loss": 1.5633,
+      "step": 6692
+    },
+    {
+      "epoch": 0.5142901044867855,
+      "grad_norm": 2.951101064682007,
+      "learning_rate": 1.794283958205286e-05,
+      "loss": 1.3899,
+      "step": 6694
+    },
+    {
+      "epoch": 0.5144437615242778,
+      "grad_norm": 3.1760647296905518,
+      "learning_rate": 1.794222495390289e-05,
+      "loss": 1.2689,
+      "step": 6696
+    },
+    {
+      "epoch": 0.5145974185617701,
+      "grad_norm": 3.310821294784546,
+      "learning_rate": 1.7941610325752922e-05,
+      "loss": 1.4301,
+      "step": 6698
+    },
+    {
+      "epoch": 0.5147510755992625,
+      "grad_norm": 3.1752665042877197,
+      "learning_rate": 1.7940995697602952e-05,
+      "loss": 1.2948,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5149047326367547,
+      "grad_norm": 3.2648708820343018,
+      "learning_rate": 1.7940381069452985e-05,
+      "loss": 1.3254,
+      "step": 6702
+    },
+    {
+      "epoch": 0.5150583896742471,
+      "grad_norm": 3.0156726837158203,
+      "learning_rate": 1.793976644130301e-05,
+      "loss": 1.4047,
+      "step": 6704
+    },
+    {
+      "epoch": 0.5152120467117394,
+      "grad_norm": 3.873260974884033,
+      "learning_rate": 1.7939151813153044e-05,
+      "loss": 1.4925,
+      "step": 6706
+    },
+    {
+      "epoch": 0.5153657037492317,
+      "grad_norm": 3.176854372024536,
+      "learning_rate": 1.7938537185003074e-05,
+      "loss": 1.301,
+      "step": 6708
+    },
+    {
+      "epoch": 0.515519360786724,
+      "grad_norm": 2.7950966358184814,
+      "learning_rate": 1.7937922556853104e-05,
+      "loss": 1.3525,
+      "step": 6710
+    },
+    {
+      "epoch": 0.5156730178242164,
+      "grad_norm": 3.167912244796753,
+      "learning_rate": 1.7937307928703137e-05,
+      "loss": 1.5551,
+      "step": 6712
+    },
+    {
+      "epoch": 0.5158266748617086,
+      "grad_norm": 3.4793012142181396,
+      "learning_rate": 1.7936693300553166e-05,
+      "loss": 1.5019,
+      "step": 6714
+    },
+    {
+      "epoch": 0.515980331899201,
+      "grad_norm": 2.9153974056243896,
+      "learning_rate": 1.7936078672403196e-05,
+      "loss": 1.407,
+      "step": 6716
+    },
+    {
+      "epoch": 0.5161339889366933,
+      "grad_norm": 2.8722434043884277,
+      "learning_rate": 1.793546404425323e-05,
+      "loss": 1.3882,
+      "step": 6718
+    },
+    {
+      "epoch": 0.5162876459741856,
+      "grad_norm": 2.9194371700286865,
+      "learning_rate": 1.793484941610326e-05,
+      "loss": 1.4366,
+      "step": 6720
+    },
+    {
+      "epoch": 0.5164413030116779,
+      "grad_norm": 3.2603633403778076,
+      "learning_rate": 1.7934234787953292e-05,
+      "loss": 1.3387,
+      "step": 6722
+    },
+    {
+      "epoch": 0.5165949600491703,
+      "grad_norm": 3.0995850563049316,
+      "learning_rate": 1.793362015980332e-05,
+      "loss": 1.3719,
+      "step": 6724
+    },
+    {
+      "epoch": 0.5167486170866625,
+      "grad_norm": 2.8780972957611084,
+      "learning_rate": 1.793300553165335e-05,
+      "loss": 1.3721,
+      "step": 6726
+    },
+    {
+      "epoch": 0.5169022741241549,
+      "grad_norm": 3.1328043937683105,
+      "learning_rate": 1.7932390903503384e-05,
+      "loss": 1.2831,
+      "step": 6728
+    },
+    {
+      "epoch": 0.5170559311616472,
+      "grad_norm": 3.0199432373046875,
+      "learning_rate": 1.793177627535341e-05,
+      "loss": 1.3505,
+      "step": 6730
+    },
+    {
+      "epoch": 0.5172095881991395,
+      "grad_norm": 2.7379181385040283,
+      "learning_rate": 1.7931161647203444e-05,
+      "loss": 1.3088,
+      "step": 6732
+    },
+    {
+      "epoch": 0.5173632452366318,
+      "grad_norm": 2.929868459701538,
+      "learning_rate": 1.7930547019053473e-05,
+      "loss": 1.3396,
+      "step": 6734
+    },
+    {
+      "epoch": 0.5175169022741242,
+      "grad_norm": 2.787508249282837,
+      "learning_rate": 1.7929932390903503e-05,
+      "loss": 1.3644,
+      "step": 6736
+    },
+    {
+      "epoch": 0.5176705593116164,
+      "grad_norm": 2.7956087589263916,
+      "learning_rate": 1.7929317762753536e-05,
+      "loss": 1.2359,
+      "step": 6738
+    },
+    {
+      "epoch": 0.5178242163491088,
+      "grad_norm": 2.53226900100708,
+      "learning_rate": 1.7928703134603566e-05,
+      "loss": 1.307,
+      "step": 6740
+    },
+    {
+      "epoch": 0.5179778733866011,
+      "grad_norm": 2.9932830333709717,
+      "learning_rate": 1.79280885064536e-05,
+      "loss": 1.3173,
+      "step": 6742
+    },
+    {
+      "epoch": 0.5181315304240934,
+      "grad_norm": 2.900578260421753,
+      "learning_rate": 1.792747387830363e-05,
+      "loss": 1.4958,
+      "step": 6744
+    },
+    {
+      "epoch": 0.5182851874615857,
+      "grad_norm": 3.190718173980713,
+      "learning_rate": 1.7926859250153658e-05,
+      "loss": 1.3313,
+      "step": 6746
+    },
+    {
+      "epoch": 0.5184388444990781,
+      "grad_norm": 3.0104870796203613,
+      "learning_rate": 1.792624462200369e-05,
+      "loss": 1.3749,
+      "step": 6748
+    },
+    {
+      "epoch": 0.5185925015365703,
+      "grad_norm": 3.335132598876953,
+      "learning_rate": 1.792562999385372e-05,
+      "loss": 1.3441,
+      "step": 6750
+    },
+    {
+      "epoch": 0.5187461585740627,
+      "grad_norm": 2.9974780082702637,
+      "learning_rate": 1.792501536570375e-05,
+      "loss": 1.4476,
+      "step": 6752
+    },
+    {
+      "epoch": 0.518899815611555,
+      "grad_norm": 2.8957419395446777,
+      "learning_rate": 1.7924400737553784e-05,
+      "loss": 1.3322,
+      "step": 6754
+    },
+    {
+      "epoch": 0.5190534726490473,
+      "grad_norm": 3.018171787261963,
+      "learning_rate": 1.792378610940381e-05,
+      "loss": 1.3581,
+      "step": 6756
+    },
+    {
+      "epoch": 0.5192071296865396,
+      "grad_norm": 2.8184151649475098,
+      "learning_rate": 1.7923171481253843e-05,
+      "loss": 1.4037,
+      "step": 6758
+    },
+    {
+      "epoch": 0.519360786724032,
+      "grad_norm": 2.97739315032959,
+      "learning_rate": 1.7922556853103873e-05,
+      "loss": 1.2547,
+      "step": 6760
+    },
+    {
+      "epoch": 0.5195144437615242,
+      "grad_norm": 3.4361817836761475,
+      "learning_rate": 1.7921942224953906e-05,
+      "loss": 1.3367,
+      "step": 6762
+    },
+    {
+      "epoch": 0.5196681007990166,
+      "grad_norm": 25.288660049438477,
+      "learning_rate": 1.7921327596803935e-05,
+      "loss": 1.4986,
+      "step": 6764
+    },
+    {
+      "epoch": 0.5198217578365089,
+      "grad_norm": 3.100538492202759,
+      "learning_rate": 1.7920712968653965e-05,
+      "loss": 1.3149,
+      "step": 6766
+    },
+    {
+      "epoch": 0.5199754148740012,
+      "grad_norm": 2.8900961875915527,
+      "learning_rate": 1.7920098340503998e-05,
+      "loss": 1.352,
+      "step": 6768
+    },
+    {
+      "epoch": 0.5201290719114936,
+      "grad_norm": 2.9886999130249023,
+      "learning_rate": 1.7919483712354028e-05,
+      "loss": 1.2732,
+      "step": 6770
+    },
+    {
+      "epoch": 0.5202827289489859,
+      "grad_norm": 2.7704336643218994,
+      "learning_rate": 1.7918869084204058e-05,
+      "loss": 1.1976,
+      "step": 6772
+    },
+    {
+      "epoch": 0.5204363859864782,
+      "grad_norm": 2.448591709136963,
+      "learning_rate": 1.791825445605409e-05,
+      "loss": 1.2853,
+      "step": 6774
+    },
+    {
+      "epoch": 0.5205900430239705,
+      "grad_norm": 2.9498369693756104,
+      "learning_rate": 1.7917639827904117e-05,
+      "loss": 1.306,
+      "step": 6776
+    },
+    {
+      "epoch": 0.5207437000614629,
+      "grad_norm": 3.2364501953125,
+      "learning_rate": 1.791702519975415e-05,
+      "loss": 1.379,
+      "step": 6778
+    },
+    {
+      "epoch": 0.5208973570989551,
+      "grad_norm": 3.1900076866149902,
+      "learning_rate": 1.791641057160418e-05,
+      "loss": 1.4177,
+      "step": 6780
+    },
+    {
+      "epoch": 0.5210510141364475,
+      "grad_norm": 3.0492565631866455,
+      "learning_rate": 1.791579594345421e-05,
+      "loss": 1.4582,
+      "step": 6782
+    },
+    {
+      "epoch": 0.5212046711739398,
+      "grad_norm": 2.8004837036132812,
+      "learning_rate": 1.7915181315304242e-05,
+      "loss": 1.4414,
+      "step": 6784
+    },
+    {
+      "epoch": 0.5213583282114321,
+      "grad_norm": 2.4796876907348633,
+      "learning_rate": 1.7914566687154272e-05,
+      "loss": 1.388,
+      "step": 6786
+    },
+    {
+      "epoch": 0.5215119852489244,
+      "grad_norm": 2.892904281616211,
+      "learning_rate": 1.7913952059004305e-05,
+      "loss": 1.4414,
+      "step": 6788
+    },
+    {
+      "epoch": 0.5216656422864168,
+      "grad_norm": 3.042837381362915,
+      "learning_rate": 1.7913337430854335e-05,
+      "loss": 1.4302,
+      "step": 6790
+    },
+    {
+      "epoch": 0.521819299323909,
+      "grad_norm": 3.3126227855682373,
+      "learning_rate": 1.7912722802704365e-05,
+      "loss": 1.4458,
+      "step": 6792
+    },
+    {
+      "epoch": 0.5219729563614014,
+      "grad_norm": 2.8981101512908936,
+      "learning_rate": 1.7912108174554398e-05,
+      "loss": 1.4403,
+      "step": 6794
+    },
+    {
+      "epoch": 0.5221266133988937,
+      "grad_norm": 3.016014337539673,
+      "learning_rate": 1.7911493546404427e-05,
+      "loss": 1.235,
+      "step": 6796
+    },
+    {
+      "epoch": 0.522280270436386,
+      "grad_norm": 2.693057060241699,
+      "learning_rate": 1.7910878918254457e-05,
+      "loss": 1.3,
+      "step": 6798
+    },
+    {
+      "epoch": 0.5224339274738783,
+      "grad_norm": 2.8503472805023193,
+      "learning_rate": 1.791026429010449e-05,
+      "loss": 1.2262,
+      "step": 6800
+    },
+    {
+      "epoch": 0.5225875845113707,
+      "grad_norm": 3.0016205310821533,
+      "learning_rate": 1.7909649661954516e-05,
+      "loss": 1.328,
+      "step": 6802
+    },
+    {
+      "epoch": 0.5227412415488629,
+      "grad_norm": 2.7652969360351562,
+      "learning_rate": 1.790903503380455e-05,
+      "loss": 1.1953,
+      "step": 6804
+    },
+    {
+      "epoch": 0.5228948985863553,
+      "grad_norm": 3.02223539352417,
+      "learning_rate": 1.790842040565458e-05,
+      "loss": 1.1621,
+      "step": 6806
+    },
+    {
+      "epoch": 0.5230485556238476,
+      "grad_norm": 3.1197988986968994,
+      "learning_rate": 1.7907805777504612e-05,
+      "loss": 1.4014,
+      "step": 6808
+    },
+    {
+      "epoch": 0.5232022126613399,
+      "grad_norm": 2.959120273590088,
+      "learning_rate": 1.7907191149354642e-05,
+      "loss": 1.4844,
+      "step": 6810
+    },
+    {
+      "epoch": 0.5233558696988322,
+      "grad_norm": 3.0649282932281494,
+      "learning_rate": 1.790657652120467e-05,
+      "loss": 1.4225,
+      "step": 6812
+    },
+    {
+      "epoch": 0.5235095267363246,
+      "grad_norm": 2.9492135047912598,
+      "learning_rate": 1.7905961893054705e-05,
+      "loss": 1.4206,
+      "step": 6814
+    },
+    {
+      "epoch": 0.5236631837738168,
+      "grad_norm": 3.1265509128570557,
+      "learning_rate": 1.7905347264904734e-05,
+      "loss": 1.317,
+      "step": 6816
+    },
+    {
+      "epoch": 0.5238168408113092,
+      "grad_norm": 2.831228256225586,
+      "learning_rate": 1.7904732636754764e-05,
+      "loss": 1.4686,
+      "step": 6818
+    },
+    {
+      "epoch": 0.5239704978488015,
+      "grad_norm": 3.228123903274536,
+      "learning_rate": 1.7904118008604797e-05,
+      "loss": 1.4721,
+      "step": 6820
+    },
+    {
+      "epoch": 0.5241241548862938,
+      "grad_norm": 3.5068886280059814,
+      "learning_rate": 1.7903503380454827e-05,
+      "loss": 1.4167,
+      "step": 6822
+    },
+    {
+      "epoch": 0.5242778119237861,
+      "grad_norm": 2.8935916423797607,
+      "learning_rate": 1.7902888752304856e-05,
+      "loss": 1.362,
+      "step": 6824
+    },
+    {
+      "epoch": 0.5244314689612785,
+      "grad_norm": 3.246994972229004,
+      "learning_rate": 1.790227412415489e-05,
+      "loss": 1.3156,
+      "step": 6826
+    },
+    {
+      "epoch": 0.5245851259987707,
+      "grad_norm": 3.070075035095215,
+      "learning_rate": 1.790165949600492e-05,
+      "loss": 1.3155,
+      "step": 6828
+    },
+    {
+      "epoch": 0.5247387830362631,
+      "grad_norm": 3.2165653705596924,
+      "learning_rate": 1.790104486785495e-05,
+      "loss": 1.4158,
+      "step": 6830
+    },
+    {
+      "epoch": 0.5248924400737554,
+      "grad_norm": 3.122731924057007,
+      "learning_rate": 1.790043023970498e-05,
+      "loss": 1.312,
+      "step": 6832
+    },
+    {
+      "epoch": 0.5250460971112477,
+      "grad_norm": 2.7937536239624023,
+      "learning_rate": 1.789981561155501e-05,
+      "loss": 1.4255,
+      "step": 6834
+    },
+    {
+      "epoch": 0.52519975414874,
+      "grad_norm": 2.8714072704315186,
+      "learning_rate": 1.789920098340504e-05,
+      "loss": 1.3561,
+      "step": 6836
+    },
+    {
+      "epoch": 0.5253534111862324,
+      "grad_norm": 3.3876631259918213,
+      "learning_rate": 1.789858635525507e-05,
+      "loss": 1.2653,
+      "step": 6838
+    },
+    {
+      "epoch": 0.5255070682237246,
+      "grad_norm": 3.067852258682251,
+      "learning_rate": 1.7897971727105104e-05,
+      "loss": 1.462,
+      "step": 6840
+    },
+    {
+      "epoch": 0.525660725261217,
+      "grad_norm": 2.623149871826172,
+      "learning_rate": 1.7897357098955134e-05,
+      "loss": 1.358,
+      "step": 6842
+    },
+    {
+      "epoch": 0.5258143822987092,
+      "grad_norm": 3.001631021499634,
+      "learning_rate": 1.7896742470805163e-05,
+      "loss": 1.4515,
+      "step": 6844
+    },
+    {
+      "epoch": 0.5259680393362016,
+      "grad_norm": 2.7457780838012695,
+      "learning_rate": 1.7896127842655197e-05,
+      "loss": 1.2601,
+      "step": 6846
+    },
+    {
+      "epoch": 0.5261216963736939,
+      "grad_norm": 3.05543851852417,
+      "learning_rate": 1.7895513214505226e-05,
+      "loss": 1.3837,
+      "step": 6848
+    },
+    {
+      "epoch": 0.5262753534111863,
+      "grad_norm": 2.7177581787109375,
+      "learning_rate": 1.7894898586355256e-05,
+      "loss": 1.2661,
+      "step": 6850
+    },
+    {
+      "epoch": 0.5264290104486785,
+      "grad_norm": 2.9022562503814697,
+      "learning_rate": 1.789428395820529e-05,
+      "loss": 1.3782,
+      "step": 6852
+    },
+    {
+      "epoch": 0.5265826674861709,
+      "grad_norm": 3.01700758934021,
+      "learning_rate": 1.789366933005532e-05,
+      "loss": 1.3188,
+      "step": 6854
+    },
+    {
+      "epoch": 0.5267363245236631,
+      "grad_norm": 2.9770376682281494,
+      "learning_rate": 1.789305470190535e-05,
+      "loss": 1.4046,
+      "step": 6856
+    },
+    {
+      "epoch": 0.5268899815611555,
+      "grad_norm": 3.635504961013794,
+      "learning_rate": 1.7892440073755378e-05,
+      "loss": 1.3997,
+      "step": 6858
+    },
+    {
+      "epoch": 0.5270436385986478,
+      "grad_norm": 3.3790907859802246,
+      "learning_rate": 1.789182544560541e-05,
+      "loss": 1.3248,
+      "step": 6860
+    },
+    {
+      "epoch": 0.5271972956361402,
+      "grad_norm": 3.0795631408691406,
+      "learning_rate": 1.789121081745544e-05,
+      "loss": 1.2888,
+      "step": 6862
+    },
+    {
+      "epoch": 0.5273509526736324,
+      "grad_norm": 2.7538647651672363,
+      "learning_rate": 1.789059618930547e-05,
+      "loss": 1.5236,
+      "step": 6864
+    },
+    {
+      "epoch": 0.5275046097111248,
+      "grad_norm": 3.4936375617980957,
+      "learning_rate": 1.7889981561155504e-05,
+      "loss": 1.4455,
+      "step": 6866
+    },
+    {
+      "epoch": 0.527658266748617,
+      "grad_norm": 3.7572572231292725,
+      "learning_rate": 1.7889366933005533e-05,
+      "loss": 1.4181,
+      "step": 6868
+    },
+    {
+      "epoch": 0.5278119237861094,
+      "grad_norm": 3.407961130142212,
+      "learning_rate": 1.7888752304855563e-05,
+      "loss": 1.3646,
+      "step": 6870
+    },
+    {
+      "epoch": 0.5279655808236017,
+      "grad_norm": 2.7877607345581055,
+      "learning_rate": 1.7888137676705596e-05,
+      "loss": 1.259,
+      "step": 6872
+    },
+    {
+      "epoch": 0.528119237861094,
+      "grad_norm": 2.999338150024414,
+      "learning_rate": 1.7887523048555626e-05,
+      "loss": 1.3156,
+      "step": 6874
+    },
+    {
+      "epoch": 0.5282728948985863,
+      "grad_norm": 3.449953079223633,
+      "learning_rate": 1.7886908420405655e-05,
+      "loss": 1.4194,
+      "step": 6876
+    },
+    {
+      "epoch": 0.5284265519360787,
+      "grad_norm": 3.0166075229644775,
+      "learning_rate": 1.788629379225569e-05,
+      "loss": 1.3544,
+      "step": 6878
+    },
+    {
+      "epoch": 0.5285802089735709,
+      "grad_norm": 2.9438271522521973,
+      "learning_rate": 1.7885679164105718e-05,
+      "loss": 1.4669,
+      "step": 6880
+    },
+    {
+      "epoch": 0.5287338660110633,
+      "grad_norm": 2.558922529220581,
+      "learning_rate": 1.7885064535955748e-05,
+      "loss": 1.2715,
+      "step": 6882
+    },
+    {
+      "epoch": 0.5288875230485556,
+      "grad_norm": 2.997262954711914,
+      "learning_rate": 1.7884449907805777e-05,
+      "loss": 1.3229,
+      "step": 6884
+    },
+    {
+      "epoch": 0.529041180086048,
+      "grad_norm": 2.7784371376037598,
+      "learning_rate": 1.788383527965581e-05,
+      "loss": 1.2801,
+      "step": 6886
+    },
+    {
+      "epoch": 0.5291948371235402,
+      "grad_norm": 3.280125379562378,
+      "learning_rate": 1.788322065150584e-05,
+      "loss": 1.3505,
+      "step": 6888
+    },
+    {
+      "epoch": 0.5293484941610326,
+      "grad_norm": 3.9201979637145996,
+      "learning_rate": 1.788260602335587e-05,
+      "loss": 1.3235,
+      "step": 6890
+    },
+    {
+      "epoch": 0.5295021511985248,
+      "grad_norm": 3.0474581718444824,
+      "learning_rate": 1.7881991395205903e-05,
+      "loss": 1.3269,
+      "step": 6892
+    },
+    {
+      "epoch": 0.5296558082360172,
+      "grad_norm": 2.956028938293457,
+      "learning_rate": 1.7881376767055933e-05,
+      "loss": 1.2897,
+      "step": 6894
+    },
+    {
+      "epoch": 0.5298094652735095,
+      "grad_norm": 3.143683433532715,
+      "learning_rate": 1.7880762138905962e-05,
+      "loss": 1.3913,
+      "step": 6896
+    },
+    {
+      "epoch": 0.5299631223110018,
+      "grad_norm": 2.942600727081299,
+      "learning_rate": 1.7880147510755995e-05,
+      "loss": 1.2993,
+      "step": 6898
+    },
+    {
+      "epoch": 0.5301167793484942,
+      "grad_norm": 2.7133402824401855,
+      "learning_rate": 1.7879532882606025e-05,
+      "loss": 1.2878,
+      "step": 6900
+    },
+    {
+      "epoch": 0.5302704363859865,
+      "grad_norm": 2.7319624423980713,
+      "learning_rate": 1.7878918254456055e-05,
+      "loss": 1.2711,
+      "step": 6902
+    },
+    {
+      "epoch": 0.5304240934234788,
+      "grad_norm": 3.0787789821624756,
+      "learning_rate": 1.7878303626306084e-05,
+      "loss": 1.4048,
+      "step": 6904
+    },
+    {
+      "epoch": 0.5305777504609711,
+      "grad_norm": 3.218059539794922,
+      "learning_rate": 1.7877688998156118e-05,
+      "loss": 1.346,
+      "step": 6906
+    },
+    {
+      "epoch": 0.5307314074984635,
+      "grad_norm": 3.2375054359436035,
+      "learning_rate": 1.7877074370006147e-05,
+      "loss": 1.3984,
+      "step": 6908
+    },
+    {
+      "epoch": 0.5308850645359557,
+      "grad_norm": 2.9513885974884033,
+      "learning_rate": 1.7876459741856177e-05,
+      "loss": 1.4062,
+      "step": 6910
+    },
+    {
+      "epoch": 0.5310387215734481,
+      "grad_norm": 3.1400296688079834,
+      "learning_rate": 1.787584511370621e-05,
+      "loss": 1.4292,
+      "step": 6912
+    },
+    {
+      "epoch": 0.5311923786109404,
+      "grad_norm": 3.114333152770996,
+      "learning_rate": 1.787523048555624e-05,
+      "loss": 1.5008,
+      "step": 6914
+    },
+    {
+      "epoch": 0.5313460356484327,
+      "grad_norm": 2.8262038230895996,
+      "learning_rate": 1.787461585740627e-05,
+      "loss": 1.4744,
+      "step": 6916
+    },
+    {
+      "epoch": 0.531499692685925,
+      "grad_norm": 3.2163949012756348,
+      "learning_rate": 1.7874001229256302e-05,
+      "loss": 1.3959,
+      "step": 6918
+    },
+    {
+      "epoch": 0.5316533497234174,
+      "grad_norm": 2.7985939979553223,
+      "learning_rate": 1.7873386601106332e-05,
+      "loss": 1.4022,
+      "step": 6920
+    },
+    {
+      "epoch": 0.5318070067609096,
+      "grad_norm": 2.731464147567749,
+      "learning_rate": 1.7872771972956362e-05,
+      "loss": 1.3438,
+      "step": 6922
+    },
+    {
+      "epoch": 0.531960663798402,
+      "grad_norm": 3.187049627304077,
+      "learning_rate": 1.7872157344806395e-05,
+      "loss": 1.4009,
+      "step": 6924
+    },
+    {
+      "epoch": 0.5321143208358943,
+      "grad_norm": 2.9336483478546143,
+      "learning_rate": 1.7871542716656425e-05,
+      "loss": 1.412,
+      "step": 6926
+    },
+    {
+      "epoch": 0.5322679778733866,
+      "grad_norm": 3.132073402404785,
+      "learning_rate": 1.7870928088506454e-05,
+      "loss": 1.3688,
+      "step": 6928
+    },
+    {
+      "epoch": 0.5324216349108789,
+      "grad_norm": 2.981800079345703,
+      "learning_rate": 1.7870313460356484e-05,
+      "loss": 1.3054,
+      "step": 6930
+    },
+    {
+      "epoch": 0.5325752919483713,
+      "grad_norm": 3.3410542011260986,
+      "learning_rate": 1.7869698832206517e-05,
+      "loss": 1.3939,
+      "step": 6932
+    },
+    {
+      "epoch": 0.5327289489858635,
+      "grad_norm": 2.8164236545562744,
+      "learning_rate": 1.7869084204056547e-05,
+      "loss": 1.3269,
+      "step": 6934
+    },
+    {
+      "epoch": 0.5328826060233559,
+      "grad_norm": 3.8438332080841064,
+      "learning_rate": 1.7868469575906576e-05,
+      "loss": 1.3612,
+      "step": 6936
+    },
+    {
+      "epoch": 0.5330362630608482,
+      "grad_norm": 2.7539196014404297,
+      "learning_rate": 1.786785494775661e-05,
+      "loss": 1.351,
+      "step": 6938
+    },
+    {
+      "epoch": 0.5331899200983405,
+      "grad_norm": 3.0499660968780518,
+      "learning_rate": 1.786724031960664e-05,
+      "loss": 1.4173,
+      "step": 6940
+    },
+    {
+      "epoch": 0.5333435771358328,
+      "grad_norm": 3.459529399871826,
+      "learning_rate": 1.786662569145667e-05,
+      "loss": 1.4837,
+      "step": 6942
+    },
+    {
+      "epoch": 0.5334972341733252,
+      "grad_norm": 3.2329859733581543,
+      "learning_rate": 1.7866011063306702e-05,
+      "loss": 1.4251,
+      "step": 6944
+    },
+    {
+      "epoch": 0.5336508912108174,
+      "grad_norm": 3.210688352584839,
+      "learning_rate": 1.786539643515673e-05,
+      "loss": 1.3004,
+      "step": 6946
+    },
+    {
+      "epoch": 0.5338045482483098,
+      "grad_norm": 3.564739227294922,
+      "learning_rate": 1.786478180700676e-05,
+      "loss": 1.443,
+      "step": 6948
+    },
+    {
+      "epoch": 0.5339582052858021,
+      "grad_norm": 2.8052964210510254,
+      "learning_rate": 1.7864167178856794e-05,
+      "loss": 1.2934,
+      "step": 6950
+    },
+    {
+      "epoch": 0.5341118623232944,
+      "grad_norm": 3.189671039581299,
+      "learning_rate": 1.7863552550706824e-05,
+      "loss": 1.3459,
+      "step": 6952
+    },
+    {
+      "epoch": 0.5342655193607867,
+      "grad_norm": 3.1696484088897705,
+      "learning_rate": 1.7862937922556857e-05,
+      "loss": 1.4234,
+      "step": 6954
+    },
+    {
+      "epoch": 0.5344191763982791,
+      "grad_norm": 2.973815679550171,
+      "learning_rate": 1.7862323294406883e-05,
+      "loss": 1.2077,
+      "step": 6956
+    },
+    {
+      "epoch": 0.5345728334357713,
+      "grad_norm": 3.508512258529663,
+      "learning_rate": 1.7861708666256916e-05,
+      "loss": 1.3832,
+      "step": 6958
+    },
+    {
+      "epoch": 0.5347264904732637,
+      "grad_norm": 2.8029847145080566,
+      "learning_rate": 1.7861094038106946e-05,
+      "loss": 1.3648,
+      "step": 6960
+    },
+    {
+      "epoch": 0.534880147510756,
+      "grad_norm": 3.183760404586792,
+      "learning_rate": 1.7860479409956976e-05,
+      "loss": 1.4012,
+      "step": 6962
+    },
+    {
+      "epoch": 0.5350338045482483,
+      "grad_norm": 3.181417942047119,
+      "learning_rate": 1.785986478180701e-05,
+      "loss": 1.3642,
+      "step": 6964
+    },
+    {
+      "epoch": 0.5351874615857406,
+      "grad_norm": 3.075376033782959,
+      "learning_rate": 1.785925015365704e-05,
+      "loss": 1.4268,
+      "step": 6966
+    },
+    {
+      "epoch": 0.535341118623233,
+      "grad_norm": 3.069875478744507,
+      "learning_rate": 1.7858635525507068e-05,
+      "loss": 1.4777,
+      "step": 6968
+    },
+    {
+      "epoch": 0.5354947756607252,
+      "grad_norm": 3.1056337356567383,
+      "learning_rate": 1.78580208973571e-05,
+      "loss": 1.4871,
+      "step": 6970
+    },
+    {
+      "epoch": 0.5356484326982176,
+      "grad_norm": 3.0267107486724854,
+      "learning_rate": 1.785740626920713e-05,
+      "loss": 1.4342,
+      "step": 6972
+    },
+    {
+      "epoch": 0.5358020897357099,
+      "grad_norm": 3.082707643508911,
+      "learning_rate": 1.7856791641057164e-05,
+      "loss": 1.4585,
+      "step": 6974
+    },
+    {
+      "epoch": 0.5359557467732022,
+      "grad_norm": 3.154895544052124,
+      "learning_rate": 1.7856177012907194e-05,
+      "loss": 1.4232,
+      "step": 6976
+    },
+    {
+      "epoch": 0.5361094038106945,
+      "grad_norm": 2.978407621383667,
+      "learning_rate": 1.7855562384757223e-05,
+      "loss": 1.3511,
+      "step": 6978
+    },
+    {
+      "epoch": 0.5362630608481869,
+      "grad_norm": 2.940761089324951,
+      "learning_rate": 1.7854947756607257e-05,
+      "loss": 1.372,
+      "step": 6980
+    },
+    {
+      "epoch": 0.5364167178856791,
+      "grad_norm": 2.9799039363861084,
+      "learning_rate": 1.7854333128457283e-05,
+      "loss": 1.5043,
+      "step": 6982
+    },
+    {
+      "epoch": 0.5365703749231715,
+      "grad_norm": 2.7917842864990234,
+      "learning_rate": 1.7853718500307316e-05,
+      "loss": 1.2237,
+      "step": 6984
+    },
+    {
+      "epoch": 0.5367240319606638,
+      "grad_norm": 2.8482396602630615,
+      "learning_rate": 1.7853103872157346e-05,
+      "loss": 1.4475,
+      "step": 6986
+    },
+    {
+      "epoch": 0.5368776889981561,
+      "grad_norm": 3.9226222038269043,
+      "learning_rate": 1.7852489244007375e-05,
+      "loss": 1.3577,
+      "step": 6988
+    },
+    {
+      "epoch": 0.5370313460356484,
+      "grad_norm": 2.828793525695801,
+      "learning_rate": 1.785187461585741e-05,
+      "loss": 1.3318,
+      "step": 6990
+    },
+    {
+      "epoch": 0.5371850030731408,
+      "grad_norm": 2.9199106693267822,
+      "learning_rate": 1.7851259987707438e-05,
+      "loss": 1.3467,
+      "step": 6992
+    },
+    {
+      "epoch": 0.537338660110633,
+      "grad_norm": 2.835475444793701,
+      "learning_rate": 1.785064535955747e-05,
+      "loss": 1.3634,
+      "step": 6994
+    },
+    {
+      "epoch": 0.5374923171481254,
+      "grad_norm": 3.4660840034484863,
+      "learning_rate": 1.78500307314075e-05,
+      "loss": 1.2317,
+      "step": 6996
+    },
+    {
+      "epoch": 0.5376459741856177,
+      "grad_norm": 4.105753421783447,
+      "learning_rate": 1.784941610325753e-05,
+      "loss": 1.3137,
+      "step": 6998
+    },
+    {
+      "epoch": 0.53779963122311,
+      "grad_norm": 2.8999927043914795,
+      "learning_rate": 1.7848801475107564e-05,
+      "loss": 1.2568,
+      "step": 7000
     }
   ],
   "logging_steps": 2,
@@ -19976,7 +24526,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 3.638521767175127e+19,
+  "total_flos": 4.46836006495191e+19,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null