diff --git a/client_states_v9_NEURIPS_DISJOINT_Memonly_LORA_llava_lr2e-5_bs1_gradacc32_iter0_25_gradnorm_scenario18_new_10000_random0_0625_seed1/0_trainer_state.json b/client_states_v9_NEURIPS_DISJOINT_Memonly_LORA_llava_lr2e-5_bs1_gradacc32_iter0_25_gradnorm_scenario18_new_10000_random0_0625_seed1/0_trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..6c827283b44ca682462de986d8091e9d6e681cfa
--- /dev/null
+++ b/client_states_v9_NEURIPS_DISJOINT_Memonly_LORA_llava_lr2e-5_bs1_gradacc32_iter0_25_gradnorm_scenario18_new_10000_random0_0625_seed1/0_trainer_state.json
@@ -0,0 +1,8782 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 2500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008,
+      "grad_norm": 7.428676605224609,
+      "learning_rate": 2.357535430610912e-06,
+      "loss": 0.8135,
+      "step": 2
+    },
+    {
+      "epoch": 0.0016,
+      "grad_norm": 4.688574314117432,
+      "learning_rate": 2.3755748898855234e-06,
+      "loss": 0.4197,
+      "step": 4
+    },
+    {
+      "epoch": 0.0024,
+      "grad_norm": 6.85806941986084,
+      "learning_rate": 2.3936738059587174e-06,
+      "loss": 0.5889,
+      "step": 6
+    },
+    {
+      "epoch": 0.0032,
+      "grad_norm": 4.939311981201172,
+      "learning_rate": 2.411832037691545e-06,
+      "loss": 1.7777,
+      "step": 8
+    },
+    {
+      "epoch": 0.004,
+      "grad_norm": 6.581257343292236,
+      "learning_rate": 2.430049443482434e-06,
+      "loss": 0.7227,
+      "step": 10
+    },
+    {
+      "epoch": 0.0048,
+      "grad_norm": 5.2789812088012695,
+      "learning_rate": 2.448325881268406e-06,
+      "loss": 0.6827,
+      "step": 12
+    },
+    {
+      "epoch": 0.0056,
+      "grad_norm": 8.367552757263184,
+      "learning_rate": 2.4666612085261277e-06,
+      "loss": 0.7583,
+      "step": 14
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 6.309530735015869,
+      "learning_rate": 2.4850552822730346e-06,
+      "loss": 0.5772,
+      "step": 16
+    },
+    {
+      "epoch": 0.0072,
+      "grad_norm": 11.15611457824707,
+      "learning_rate": 2.503507959068455e-06,
+      "loss": 0.5898,
+      "step": 18
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 2.9229376316070557,
+      "learning_rate": 2.522019095014686e-06,
+      "loss": 0.1423,
+      "step": 20
+    },
+    {
+      "epoch": 0.0088,
+      "grad_norm": 13.48912525177002,
+      "learning_rate": 2.5405885457581814e-06,
+      "loss": 0.9523,
+      "step": 22
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 7.22756290435791,
+      "learning_rate": 2.5592161664906243e-06,
+      "loss": 0.329,
+      "step": 24
+    },
+    {
+      "epoch": 0.0104,
+      "grad_norm": 3.864868402481079,
+      "learning_rate": 2.5779018119501086e-06,
+      "loss": 0.3043,
+      "step": 26
+    },
+    {
+      "epoch": 0.0112,
+      "grad_norm": 7.621895790100098,
+      "learning_rate": 2.596645336422219e-06,
+      "loss": 0.3123,
+      "step": 28
+    },
+    {
+      "epoch": 0.012,
+      "grad_norm": 6.358210563659668,
+      "learning_rate": 2.615446593741161e-06,
+      "loss": 0.5717,
+      "step": 30
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 2.5341553688049316,
+      "learning_rate": 2.6343054372909648e-06,
+      "loss": 0.3356,
+      "step": 32
+    },
+    {
+      "epoch": 0.0136,
+      "grad_norm": 7.333450794219971,
+      "learning_rate": 2.6532217200065826e-06,
+      "loss": 0.4261,
+      "step": 34
+    },
+    {
+      "epoch": 0.0144,
+      "grad_norm": 9.422982215881348,
+      "learning_rate": 2.6721952943750396e-06,
+      "loss": 0.7553,
+      "step": 36
+    },
+    {
+      "epoch": 0.0152,
+      "grad_norm": 7.869431972503662,
+      "learning_rate": 2.691226012436604e-06,
+      "loss": 0.4286,
+      "step": 38
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 5.751649379730225,
+      "learning_rate": 2.7103137257858893e-06,
+      "loss": 0.5196,
+      "step": 40
+    },
+    {
+      "epoch": 0.0168,
+      "grad_norm": 9.487853050231934,
+      "learning_rate": 2.7294582855730733e-06,
+      "loss": 0.3418,
+      "step": 42
+    },
+    {
+      "epoch": 0.0176,
+      "grad_norm": 0.8265520334243774,
+      "learning_rate": 2.7486595425050566e-06,
+      "loss": 0.36,
+      "step": 44
+    },
+    {
+      "epoch": 0.0184,
+      "grad_norm": 6.352355003356934,
+      "learning_rate": 2.7679173468465813e-06,
+      "loss": 0.6908,
+      "step": 46
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 6.220467567443848,
+      "learning_rate": 2.7872315484213954e-06,
+      "loss": 0.5355,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.542949676513672,
+      "learning_rate": 2.8066019966134873e-06,
+      "loss": 0.3584,
+      "step": 50
+    },
+    {
+      "epoch": 0.0208,
+      "grad_norm": 12.532072067260742,
+      "learning_rate": 2.826028540368212e-06,
+      "loss": 0.7105,
+      "step": 52
+    },
+    {
+      "epoch": 0.0216,
+      "grad_norm": 22.506433486938477,
+      "learning_rate": 2.845511028193477e-06,
+      "loss": 1.2368,
+      "step": 54
+    },
+    {
+      "epoch": 0.0224,
+      "grad_norm": 5.032310485839844,
+      "learning_rate": 2.865049308160931e-06,
+      "loss": 0.8072,
+      "step": 56
+    },
+    {
+      "epoch": 0.0232,
+      "grad_norm": 5.905128479003906,
+      "learning_rate": 2.8846432279071533e-06,
+      "loss": 0.4286,
+      "step": 58
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 4.69681453704834,
+      "learning_rate": 2.9042926346347835e-06,
+      "loss": 0.3218,
+      "step": 60
+    },
+    {
+      "epoch": 0.0248,
+      "grad_norm": 7.986056804656982,
+      "learning_rate": 2.9239973751138397e-06,
+      "loss": 0.2838,
+      "step": 62
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 7.595277309417725,
+      "learning_rate": 2.943757295682783e-06,
+      "loss": 0.8026,
+      "step": 64
+    },
+    {
+      "epoch": 0.0264,
+      "grad_norm": 8.339045524597168,
+      "learning_rate": 2.9635722422497983e-06,
+      "loss": 0.3001,
+      "step": 66
+    },
+    {
+      "epoch": 0.0272,
+      "grad_norm": 2.1323695182800293,
+      "learning_rate": 2.983442060293926e-06,
+      "loss": 0.2492,
+      "step": 68
+    },
+    {
+      "epoch": 0.028,
+      "grad_norm": 7.9373908042907715,
+      "learning_rate": 3.003366594866345e-06,
+      "loss": 0.7103,
+      "step": 70
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 4.134149551391602,
+      "learning_rate": 3.0233456905915338e-06,
+      "loss": 0.7687,
+      "step": 72
+    },
+    {
+      "epoch": 0.0296,
+      "grad_norm": 7.5926666259765625,
+      "learning_rate": 3.0433791916684885e-06,
+      "loss": 0.5618,
+      "step": 74
+    },
+    {
+      "epoch": 0.0304,
+      "grad_norm": 7.338794231414795,
+      "learning_rate": 3.0634669418719453e-06,
+      "loss": 0.4229,
+      "step": 76
+    },
+    {
+      "epoch": 0.0312,
+      "grad_norm": 10.961363792419434,
+      "learning_rate": 3.0836087845535933e-06,
+      "loss": 0.7771,
+      "step": 78
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 7.643119812011719,
+      "learning_rate": 3.1038045626432945e-06,
+      "loss": 0.5454,
+      "step": 80
+    },
+    {
+      "epoch": 0.0328,
+      "grad_norm": 6.308757305145264,
+      "learning_rate": 3.1240541186503173e-06,
+      "loss": 0.337,
+      "step": 82
+    },
+    {
+      "epoch": 0.0336,
+      "grad_norm": 5.920470714569092,
+      "learning_rate": 3.1443572946645683e-06,
+      "loss": 0.6162,
+      "step": 84
+    },
+    {
+      "epoch": 0.0344,
+      "grad_norm": 4.9982733726501465,
+      "learning_rate": 3.164713932357776e-06,
+      "loss": 0.3885,
+      "step": 86
+    },
+    {
+      "epoch": 0.0352,
+      "grad_norm": 4.772004127502441,
+      "learning_rate": 3.1851238729848033e-06,
+      "loss": 0.4505,
+      "step": 88
+    },
+    {
+      "epoch": 0.036,
+      "grad_norm": 6.033514022827148,
+      "learning_rate": 3.205586957384834e-06,
+      "loss": 0.5492,
+      "step": 90
+    },
+    {
+      "epoch": 0.0368,
+      "grad_norm": 7.672890663146973,
+      "learning_rate": 3.2261030259826253e-06,
+      "loss": 0.5542,
+      "step": 92
+    },
+    {
+      "epoch": 0.0376,
+      "grad_norm": 4.560334205627441,
+      "learning_rate": 3.246671918789752e-06,
+      "loss": 0.4246,
+      "step": 94
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 8.157868385314941,
+      "learning_rate": 3.267293475405858e-06,
+      "loss": 0.5161,
+      "step": 96
+    },
+    {
+      "epoch": 0.0392,
+      "grad_norm": 9.780196189880371,
+      "learning_rate": 3.2879675350199004e-06,
+      "loss": 0.4372,
+      "step": 98
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.107337951660156,
+      "learning_rate": 3.3086939364114113e-06,
+      "loss": 0.7497,
+      "step": 100
+    },
+    {
+      "epoch": 0.0408,
+      "grad_norm": 5.913912296295166,
+      "learning_rate": 3.329472517951747e-06,
+      "loss": 0.4863,
+      "step": 102
+    },
+    {
+      "epoch": 0.0416,
+      "grad_norm": 5.233573913574219,
+      "learning_rate": 3.350303117605369e-06,
+      "loss": 0.3886,
+      "step": 104
+    },
+    {
+      "epoch": 0.0424,
+      "grad_norm": 7.944777488708496,
+      "learning_rate": 3.3711855729310503e-06,
+      "loss": 0.7644,
+      "step": 106
+    },
+    {
+      "epoch": 0.0432,
+      "grad_norm": 7.0277485847473145,
+      "learning_rate": 3.3921197210832235e-06,
+      "loss": 0.7131,
+      "step": 108
+    },
+    {
+      "epoch": 0.044,
+      "grad_norm": 25.346858978271484,
+      "learning_rate": 3.4131053988131947e-06,
+      "loss": 0.8946,
+      "step": 110
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 4.840317249298096,
+      "learning_rate": 3.434142442470434e-06,
+      "loss": 0.3861,
+      "step": 112
+    },
+    {
+      "epoch": 0.0456,
+      "grad_norm": 5.518627643585205,
+      "learning_rate": 3.455230688003849e-06,
+      "loss": 0.6604,
+      "step": 114
+    },
+    {
+      "epoch": 0.0464,
+      "grad_norm": 6.281485557556152,
+      "learning_rate": 3.476369970963065e-06,
+      "loss": 0.648,
+      "step": 116
+    },
+    {
+      "epoch": 0.0472,
+      "grad_norm": 13.487313270568848,
+      "learning_rate": 3.497560126499706e-06,
+      "loss": 0.7007,
+      "step": 118
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 2.665107250213623,
+      "learning_rate": 3.5188009893686836e-06,
+      "loss": 0.1257,
+      "step": 120
+    },
+    {
+      "epoch": 0.0488,
+      "grad_norm": 8.823912620544434,
+      "learning_rate": 3.5400923939294827e-06,
+      "loss": 0.6357,
+      "step": 122
+    },
+    {
+      "epoch": 0.0496,
+      "grad_norm": 3.015528678894043,
+      "learning_rate": 3.5614341741474667e-06,
+      "loss": 0.1395,
+      "step": 124
+    },
+    {
+      "epoch": 0.0504,
+      "grad_norm": 3.595741033554077,
+      "learning_rate": 3.5828261635951177e-06,
+      "loss": 0.3638,
+      "step": 126
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 4.580526828765869,
+      "learning_rate": 3.604268195453421e-06,
+      "loss": 0.5191,
+      "step": 128
+    },
+    {
+      "epoch": 0.052,
+      "grad_norm": 7.3755292892456055,
+      "learning_rate": 3.6257601025130893e-06,
+      "loss": 0.4993,
+      "step": 130
+    },
+    {
+      "epoch": 0.0528,
+      "grad_norm": 9.659078598022461,
+      "learning_rate": 3.647301717175955e-06,
+      "loss": 0.4007,
+      "step": 132
+    },
+    {
+      "epoch": 0.0536,
+      "grad_norm": 6.970526695251465,
+      "learning_rate": 3.66889287145614e-06,
+      "loss": 0.9278,
+      "step": 134
+    },
+    {
+      "epoch": 0.0544,
+      "grad_norm": 9.35775375366211,
+      "learning_rate": 3.6905333969814995e-06,
+      "loss": 0.7499,
+      "step": 136
+    },
+    {
+      "epoch": 0.0552,
+      "grad_norm": 8.061389923095703,
+      "learning_rate": 3.712223124994867e-06,
+      "loss": 0.3931,
+      "step": 138
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 7.523435592651367,
+      "learning_rate": 3.7339618863553885e-06,
+      "loss": 0.3639,
+      "step": 140
+    },
+    {
+      "epoch": 0.0568,
+      "grad_norm": 5.77680778503418,
+      "learning_rate": 3.755749511539848e-06,
+      "loss": 0.5739,
+      "step": 142
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 8.214181900024414,
+      "learning_rate": 3.7775858306439404e-06,
+      "loss": 0.4753,
+      "step": 144
+    },
+    {
+      "epoch": 0.0584,
+      "grad_norm": 6.254648685455322,
+      "learning_rate": 3.799470673383677e-06,
+      "loss": 0.3452,
+      "step": 146
+    },
+    {
+      "epoch": 0.0592,
+      "grad_norm": 8.179805755615234,
+      "learning_rate": 3.821403869096644e-06,
+      "loss": 0.4398,
+      "step": 148
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.415241718292236,
+      "learning_rate": 3.8433852467434175e-06,
+      "loss": 0.6625,
+      "step": 150
+    },
+    {
+      "epoch": 0.0608,
+      "grad_norm": 4.284675121307373,
+      "learning_rate": 3.865414634908756e-06,
+      "loss": 0.4488,
+      "step": 152
+    },
+    {
+      "epoch": 0.0616,
+      "grad_norm": 7.8885345458984375,
+      "learning_rate": 3.887491861803081e-06,
+      "loss": 0.6684,
+      "step": 154
+    },
+    {
+      "epoch": 0.0624,
+      "grad_norm": 5.000823497772217,
+      "learning_rate": 3.909616755263741e-06,
+      "loss": 0.5359,
+      "step": 156
+    },
+    {
+      "epoch": 0.0632,
+      "grad_norm": 3.3473854064941406,
+      "learning_rate": 3.9317891427563725e-06,
+      "loss": 0.403,
+      "step": 158
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 5.74239444732666,
+      "learning_rate": 3.954008851376244e-06,
+      "loss": 0.4741,
+      "step": 160
+    },
+    {
+      "epoch": 0.0648,
+      "grad_norm": 7.925304889678955,
+      "learning_rate": 3.976275707849619e-06,
+      "loss": 0.4604,
+      "step": 162
+    },
+    {
+      "epoch": 0.0656,
+      "grad_norm": 7.1892571449279785,
+      "learning_rate": 3.99858953853505e-06,
+      "loss": 0.547,
+      "step": 164
+    },
+    {
+      "epoch": 0.0664,
+      "grad_norm": 6.141909599304199,
+      "learning_rate": 4.0209501694248e-06,
+      "loss": 0.4658,
+      "step": 166
+    },
+    {
+      "epoch": 0.0672,
+      "grad_norm": 11.796825408935547,
+      "learning_rate": 4.043357426146209e-06,
+      "loss": 0.6537,
+      "step": 168
+    },
+    {
+      "epoch": 0.068,
+      "grad_norm": 3.107785940170288,
+      "learning_rate": 4.065811133962987e-06,
+      "loss": 0.3216,
+      "step": 170
+    },
+    {
+      "epoch": 0.0688,
+      "grad_norm": 6.113830089569092,
+      "learning_rate": 4.08831111777658e-06,
+      "loss": 0.6906,
+      "step": 172
+    },
+    {
+      "epoch": 0.0696,
+      "grad_norm": 8.406272888183594,
+      "learning_rate": 4.110857202127611e-06,
+      "loss": 0.6615,
+      "step": 174
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 6.46627140045166,
+      "learning_rate": 4.133449211197183e-06,
+      "loss": 0.5224,
+      "step": 176
+    },
+    {
+      "epoch": 0.0712,
+      "grad_norm": 4.773953914642334,
+      "learning_rate": 4.156086968808274e-06,
+      "loss": 0.3986,
+      "step": 178
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 2.648827314376831,
+      "learning_rate": 4.178770298427114e-06,
+      "loss": 0.3031,
+      "step": 180
+    },
+    {
+      "epoch": 0.0728,
+      "grad_norm": 5.903095722198486,
+      "learning_rate": 4.201499023164515e-06,
+      "loss": 0.4427,
+      "step": 182
+    },
+    {
+      "epoch": 0.0736,
+      "grad_norm": 11.300853729248047,
+      "learning_rate": 4.224272965777315e-06,
+      "loss": 0.8331,
+      "step": 184
+    },
+    {
+      "epoch": 0.0744,
+      "grad_norm": 4.325449466705322,
+      "learning_rate": 4.247091948669764e-06,
+      "loss": 0.3271,
+      "step": 186
+    },
+    {
+      "epoch": 0.0752,
+      "grad_norm": 8.037109375,
+      "learning_rate": 4.269955793894849e-06,
+      "loss": 0.6188,
+      "step": 188
+    },
+    {
+      "epoch": 0.076,
+      "grad_norm": 5.573611259460449,
+      "learning_rate": 4.292864323155684e-06,
+      "loss": 0.5308,
+      "step": 190
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 9.673019409179688,
+      "learning_rate": 4.3158173578069696e-06,
+      "loss": 0.5004,
+      "step": 192
+    },
+    {
+      "epoch": 0.0776,
+      "grad_norm": 2.4727535247802734,
+      "learning_rate": 4.338814718856333e-06,
+      "loss": 0.3358,
+      "step": 194
+    },
+    {
+      "epoch": 0.0784,
+      "grad_norm": 5.131119728088379,
+      "learning_rate": 4.3618562269657285e-06,
+      "loss": 0.2717,
+      "step": 196
+    },
+    {
+      "epoch": 0.0792,
+      "grad_norm": 18.657052993774414,
+      "learning_rate": 4.384941702452852e-06,
+      "loss": 0.8001,
+      "step": 198
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7666983604431152,
+      "learning_rate": 4.408070965292526e-06,
+      "loss": 0.4866,
+      "step": 200
+    },
+    {
+      "epoch": 0.0808,
+      "grad_norm": 3.064669609069824,
+      "learning_rate": 4.431243835118112e-06,
+      "loss": 0.3906,
+      "step": 202
+    },
+    {
+      "epoch": 0.0816,
+      "grad_norm": 12.827176094055176,
+      "learning_rate": 4.4544601312229185e-06,
+      "loss": 0.7817,
+      "step": 204
+    },
+    {
+      "epoch": 0.0824,
+      "grad_norm": 12.40218734741211,
+      "learning_rate": 4.477719672561602e-06,
+      "loss": 0.511,
+      "step": 206
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 7.720707416534424,
+      "learning_rate": 4.501022277751605e-06,
+      "loss": 0.396,
+      "step": 208
+    },
+    {
+      "epoch": 0.084,
+      "grad_norm": 5.669599533081055,
+      "learning_rate": 4.524367765074499e-06,
+      "loss": 0.3701,
+      "step": 210
+    },
+    {
+      "epoch": 0.0848,
+      "grad_norm": 14.275040626525879,
+      "learning_rate": 4.5477559524775e-06,
+      "loss": 0.9219,
+      "step": 212
+    },
+    {
+      "epoch": 0.0856,
+      "grad_norm": 5.309834957122803,
+      "learning_rate": 4.571186657574823e-06,
+      "loss": 0.5128,
+      "step": 214
+    },
+    {
+      "epoch": 0.0864,
+      "grad_norm": 8.355277061462402,
+      "learning_rate": 4.5946596976491254e-06,
+      "loss": 0.7039,
+      "step": 216
+    },
+    {
+      "epoch": 0.0872,
+      "grad_norm": 8.066947937011719,
+      "learning_rate": 4.618174889652924e-06,
+      "loss": 0.5883,
+      "step": 218
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 8.851380348205566,
+      "learning_rate": 4.6417320502100286e-06,
+      "loss": 0.6141,
+      "step": 220
+    },
+    {
+      "epoch": 0.0888,
+      "grad_norm": 6.146321773529053,
+      "learning_rate": 4.665330995616967e-06,
+      "loss": 0.4213,
+      "step": 222
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 2.3837244510650635,
+      "learning_rate": 4.688971541844424e-06,
+      "loss": 0.372,
+      "step": 224
+    },
+    {
+      "epoch": 0.0904,
+      "grad_norm": 15.606354713439941,
+      "learning_rate": 4.712653504538672e-06,
+      "loss": 0.6691,
+      "step": 226
+    },
+    {
+      "epoch": 0.0912,
+      "grad_norm": 5.409621238708496,
+      "learning_rate": 4.736376699023023e-06,
+      "loss": 0.3749,
+      "step": 228
+    },
+    {
+      "epoch": 0.092,
+      "grad_norm": 3.3378796577453613,
+      "learning_rate": 4.76014094029921e-06,
+      "loss": 0.2424,
+      "step": 230
+    },
+    {
+      "epoch": 0.0928,
+      "grad_norm": 6.939662933349609,
+      "learning_rate": 4.7839460430489216e-06,
+      "loss": 0.2881,
+      "step": 232
+    },
+    {
+      "epoch": 0.0936,
+      "grad_norm": 11.248266220092773,
+      "learning_rate": 4.807791821635185e-06,
+      "loss": 0.6406,
+      "step": 234
+    },
+    {
+      "epoch": 0.0944,
+      "grad_norm": 14.713454246520996,
+      "learning_rate": 4.831678090103828e-06,
+      "loss": 1.1709,
+      "step": 236
+    },
+    {
+      "epoch": 0.0952,
+      "grad_norm": 3.425539016723633,
+      "learning_rate": 4.855604662184931e-06,
+      "loss": 0.4371,
+      "step": 238
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 4.274609088897705,
+      "learning_rate": 4.8795713512942785e-06,
+      "loss": 0.2196,
+      "step": 240
+    },
+    {
+      "epoch": 0.0968,
+      "grad_norm": 5.718649387359619,
+      "learning_rate": 4.903577970534815e-06,
+      "loss": 0.3956,
+      "step": 242
+    },
+    {
+      "epoch": 0.0976,
+      "grad_norm": 5.639620780944824,
+      "learning_rate": 4.9276243326981e-06,
+      "loss": 0.4738,
+      "step": 244
+    },
+    {
+      "epoch": 0.0984,
+      "grad_norm": 6.417396545410156,
+      "learning_rate": 4.951710250265788e-06,
+      "loss": 0.7023,
+      "step": 246
+    },
+    {
+      "epoch": 0.0992,
+      "grad_norm": 5.298447608947754,
+      "learning_rate": 4.975835535411023e-06,
+      "loss": 0.5218,
+      "step": 248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.021275520324707,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.4748,
+      "step": 250
+    },
+    {
+      "epoch": 0.1008,
+      "grad_norm": 4.228523254394531,
+      "learning_rate": 5.024203455593375e-06,
+      "loss": 0.1665,
+      "step": 252
+    },
+    {
+      "epoch": 0.1016,
+      "grad_norm": 4.622741222381592,
+      "learning_rate": 5.048445713447734e-06,
+      "loss": 0.4185,
+      "step": 254
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 5.1597208976745605,
+      "learning_rate": 5.072726584517083e-06,
+      "loss": 0.336,
+      "step": 256
+    },
+    {
+      "epoch": 0.1032,
+      "grad_norm": 7.466769695281982,
+      "learning_rate": 5.097045879454308e-06,
+      "loss": 0.3572,
+      "step": 258
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 4.210982799530029,
+      "learning_rate": 5.1214034086126685e-06,
+      "loss": 0.4357,
+      "step": 260
+    },
+    {
+      "epoch": 0.1048,
+      "grad_norm": 3.1034891605377197,
+      "learning_rate": 5.145798982047253e-06,
+      "loss": 0.4363,
+      "step": 262
+    },
+    {
+      "epoch": 0.1056,
+      "grad_norm": 5.984959602355957,
+      "learning_rate": 5.170232409516483e-06,
+      "loss": 0.4626,
+      "step": 264
+    },
+    {
+      "epoch": 0.1064,
+      "grad_norm": 17.632976531982422,
+      "learning_rate": 5.194703500483597e-06,
+      "loss": 0.4285,
+      "step": 266
+    },
+    {
+      "epoch": 0.1072,
+      "grad_norm": 7.416657447814941,
+      "learning_rate": 5.219212064118082e-06,
+      "loss": 0.3839,
+      "step": 268
+    },
+    {
+      "epoch": 0.108,
+      "grad_norm": 9.656270980834961,
+      "learning_rate": 5.24375790929725e-06,
+      "loss": 0.4548,
+      "step": 270
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 50.1550178527832,
+      "learning_rate": 5.268340844607653e-06,
+      "loss": 4.0707,
+      "step": 272
+    },
+    {
+      "epoch": 0.1096,
+      "grad_norm": 9.233906745910645,
+      "learning_rate": 5.2929606783466735e-06,
+      "loss": 0.662,
+      "step": 274
+    },
+    {
+      "epoch": 0.1104,
+      "grad_norm": 7.082289218902588,
+      "learning_rate": 5.317617218523853e-06,
+      "loss": 0.6863,
+      "step": 276
+    },
+    {
+      "epoch": 0.1112,
+      "grad_norm": 7.848796367645264,
+      "learning_rate": 5.342310272862553e-06,
+      "loss": 0.6412,
+      "step": 278
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 3.84726881980896,
+      "learning_rate": 5.367039648801377e-06,
+      "loss": 0.3674,
+      "step": 280
+    },
+    {
+      "epoch": 0.1128,
+      "grad_norm": 8.89835262298584,
+      "learning_rate": 5.391805153495684e-06,
+      "loss": 0.5559,
+      "step": 282
+    },
+    {
+      "epoch": 0.1136,
+      "grad_norm": 8.333953857421875,
+      "learning_rate": 5.416606593819109e-06,
+      "loss": 0.7547,
+      "step": 284
+    },
+    {
+      "epoch": 0.1144,
+      "grad_norm": 5.093605995178223,
+      "learning_rate": 5.441443776365005e-06,
+      "loss": 0.4324,
+      "step": 286
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 5.2162017822265625,
+      "learning_rate": 5.466316507448053e-06,
+      "loss": 0.8642,
+      "step": 288
+    },
+    {
+      "epoch": 0.116,
+      "grad_norm": 10.420114517211914,
+      "learning_rate": 5.49122459310568e-06,
+      "loss": 0.6503,
+      "step": 290
+    },
+    {
+      "epoch": 0.1168,
+      "grad_norm": 6.556491851806641,
+      "learning_rate": 5.516167839099662e-06,
+      "loss": 0.6044,
+      "step": 292
+    },
+    {
+      "epoch": 0.1176,
+      "grad_norm": 12.74498176574707,
+      "learning_rate": 5.5411460509175605e-06,
+      "loss": 0.7149,
+      "step": 294
+    },
+    {
+      "epoch": 0.1184,
+      "grad_norm": 7.758618354797363,
+      "learning_rate": 5.5661590337742255e-06,
+      "loss": 0.4331,
+      "step": 296
+    },
+    {
+      "epoch": 0.1192,
+      "grad_norm": 6.872960567474365,
+      "learning_rate": 5.591206592613412e-06,
+      "loss": 0.5118,
+      "step": 298
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.9478302001953125,
+      "learning_rate": 5.616288532109221e-06,
+      "loss": 0.3771,
+      "step": 300
+    },
+    {
+      "epoch": 0.1208,
+      "grad_norm": 4.419280529022217,
+      "learning_rate": 5.641404656667652e-06,
+      "loss": 0.3816,
+      "step": 302
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 4.0377936363220215,
+      "learning_rate": 5.666554770428136e-06,
+      "loss": 0.3645,
+      "step": 304
+    },
+    {
+      "epoch": 0.1224,
+      "grad_norm": 7.112875938415527,
+      "learning_rate": 5.6917386772650015e-06,
+      "loss": 0.4655,
+      "step": 306
+    },
+    {
+      "epoch": 0.1232,
+      "grad_norm": 4.4771904945373535,
+      "learning_rate": 5.716956180789086e-06,
+      "loss": 0.4779,
+      "step": 308
+    },
+    {
+      "epoch": 0.124,
+      "grad_norm": 5.968297958374023,
+      "learning_rate": 5.74220708434926e-06,
+      "loss": 0.8389,
+      "step": 310
+    },
+    {
+      "epoch": 0.1248,
+      "grad_norm": 7.576871395111084,
+      "learning_rate": 5.767491191033909e-06,
+      "loss": 0.4136,
+      "step": 312
+    },
+    {
+      "epoch": 0.1256,
+      "grad_norm": 6.465977668762207,
+      "learning_rate": 5.7928083036724535e-06,
+      "loss": 0.3366,
+      "step": 314
+    },
+    {
+      "epoch": 0.1264,
+      "grad_norm": 8.43101692199707,
+      "learning_rate": 5.818158224836983e-06,
+      "loss": 0.4671,
+      "step": 316
+    },
+    {
+      "epoch": 0.1272,
+      "grad_norm": 6.674490928649902,
+      "learning_rate": 5.8435407568437194e-06,
+      "loss": 1.8329,
+      "step": 318
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 4.587431907653809,
+      "learning_rate": 5.868955701754577e-06,
+      "loss": 0.339,
+      "step": 320
+    },
+    {
+      "epoch": 0.1288,
+      "grad_norm": 12.540658950805664,
+      "learning_rate": 5.894402861378714e-06,
+      "loss": 0.5991,
+      "step": 322
+    },
+    {
+      "epoch": 0.1296,
+      "grad_norm": 6.9753875732421875,
+      "learning_rate": 5.919882037274065e-06,
+      "loss": 0.4035,
+      "step": 324
+    },
+    {
+      "epoch": 0.1304,
+      "grad_norm": 4.0211663246154785,
+      "learning_rate": 5.9453930307488985e-06,
+      "loss": 0.5661,
+      "step": 326
+    },
+    {
+      "epoch": 0.1312,
+      "grad_norm": 11.671126365661621,
+      "learning_rate": 5.970935642863362e-06,
+      "loss": 0.776,
+      "step": 328
+    },
+    {
+      "epoch": 0.132,
+      "grad_norm": 4.407910346984863,
+      "learning_rate": 5.996509674431038e-06,
+      "loss": 0.4211,
+      "step": 330
+    },
+    {
+      "epoch": 0.1328,
+      "grad_norm": 3.740440607070923,
+      "learning_rate": 6.022114926020505e-06,
+      "loss": 0.3472,
+      "step": 332
+    },
+    {
+      "epoch": 0.1336,
+      "grad_norm": 4.749978542327881,
+      "learning_rate": 6.047751197956836e-06,
+      "loss": 0.2956,
+      "step": 334
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 3.446544647216797,
+      "learning_rate": 6.0734182903232475e-06,
+      "loss": 0.6477,
+      "step": 336
+    },
+    {
+      "epoch": 0.1352,
+      "grad_norm": 8.398823738098145,
+      "learning_rate": 6.0991160029626e-06,
+      "loss": 0.6604,
+      "step": 338
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 3.5237298011779785,
+      "learning_rate": 6.124844135478966e-06,
+      "loss": 0.4482,
+      "step": 340
+    },
+    {
+      "epoch": 0.1368,
+      "grad_norm": 5.967245578765869,
+      "learning_rate": 6.1506024872392e-06,
+      "loss": 0.4079,
+      "step": 342
+    },
+    {
+      "epoch": 0.1376,
+      "grad_norm": 9.490840911865234,
+      "learning_rate": 6.176390857374501e-06,
+      "loss": 0.8531,
+      "step": 344
+    },
+    {
+      "epoch": 0.1384,
+      "grad_norm": 3.8971786499023438,
+      "learning_rate": 6.202209044781979e-06,
+      "loss": 0.4008,
+      "step": 346
+    },
+    {
+      "epoch": 0.1392,
+      "grad_norm": 4.254534721374512,
+      "learning_rate": 6.228056848126223e-06,
+      "loss": 0.6325,
+      "step": 348
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.16749906539917,
+      "learning_rate": 6.253934065840883e-06,
+      "loss": 0.3609,
+      "step": 350
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 7.321227073669434,
+      "learning_rate": 6.279840496130188e-06,
+      "loss": 0.3653,
+      "step": 352
+    },
+    {
+      "epoch": 0.1416,
+      "grad_norm": 7.62925386428833,
+      "learning_rate": 6.305775936970606e-06,
+      "loss": 0.8053,
+      "step": 354
+    },
+    {
+      "epoch": 0.1424,
+      "grad_norm": 4.928267002105713,
+      "learning_rate": 6.331740186112359e-06,
+      "loss": 0.5276,
+      "step": 356
+    },
+    {
+      "epoch": 0.1432,
+      "grad_norm": 2.6523029804229736,
+      "learning_rate": 6.357733041081015e-06,
+      "loss": 0.2917,
+      "step": 358
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 12.000823020935059,
+      "learning_rate": 6.383754299179072e-06,
+      "loss": 0.5779,
+      "step": 360
+    },
+    {
+      "epoch": 0.1448,
+      "grad_norm": 4.4840407371521,
+      "learning_rate": 6.409803757487532e-06,
+      "loss": 0.5981,
+      "step": 362
+    },
+    {
+      "epoch": 0.1456,
+      "grad_norm": 5.198264122009277,
+      "learning_rate": 6.435881212867485e-06,
+      "loss": 0.6076,
+      "step": 364
+    },
+    {
+      "epoch": 0.1464,
+      "grad_norm": 3.4151840209960938,
+      "learning_rate": 6.4619864619616975e-06,
+      "loss": 0.4114,
+      "step": 366
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 6.271514415740967,
+      "learning_rate": 6.48811930119619e-06,
+      "loss": 0.4758,
+      "step": 368
+    },
+    {
+      "epoch": 0.148,
+      "grad_norm": 17.935209274291992,
+      "learning_rate": 6.514279526781853e-06,
+      "loss": 0.7199,
+      "step": 370
+    },
+    {
+      "epoch": 0.1488,
+      "grad_norm": 14.56944465637207,
+      "learning_rate": 6.540466934715955e-06,
+      "loss": 0.6454,
+      "step": 372
+    },
+    {
+      "epoch": 0.1496,
+      "grad_norm": 4.990140438079834,
+      "learning_rate": 6.566681320783848e-06,
+      "loss": 0.2431,
+      "step": 374
+    },
+    {
+      "epoch": 0.1504,
+      "grad_norm": 6.392396450042725,
+      "learning_rate": 6.592922480560483e-06,
+      "loss": 0.3967,
+      "step": 376
+    },
+    {
+      "epoch": 0.1512,
+      "grad_norm": 6.542480945587158,
+      "learning_rate": 6.619190209412025e-06,
+      "loss": 0.577,
+      "step": 378
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 6.3435893058776855,
+      "learning_rate": 6.6454843024974465e-06,
+      "loss": 0.3547,
+      "step": 380
+    },
+    {
+      "epoch": 0.1528,
+      "grad_norm": 7.85693883895874,
+      "learning_rate": 6.671804554770128e-06,
+      "loss": 0.9327,
+      "step": 382
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 4.3096137046813965,
+      "learning_rate": 6.698150760979456e-06,
+      "loss": 0.6549,
+      "step": 384
+    },
+    {
+      "epoch": 0.1544,
+      "grad_norm": 5.939504146575928,
+      "learning_rate": 6.724522715672421e-06,
+      "loss": 0.518,
+      "step": 386
+    },
+    {
+      "epoch": 0.1552,
+      "grad_norm": 5.392618656158447,
+      "learning_rate": 6.750920213195242e-06,
+      "loss": 0.7825,
+      "step": 388
+    },
+    {
+      "epoch": 0.156,
+      "grad_norm": 9.22570514678955,
+      "learning_rate": 6.777343047694894e-06,
+      "loss": 0.681,
+      "step": 390
+    },
+    {
+      "epoch": 0.1568,
+      "grad_norm": 5.902850151062012,
+      "learning_rate": 6.803791013120824e-06,
+      "loss": 0.3366,
+      "step": 392
+    },
+    {
+      "epoch": 0.1576,
+      "grad_norm": 6.317704200744629,
+      "learning_rate": 6.8302639032264836e-06,
+      "loss": 0.6636,
+      "step": 394
+    },
+    {
+      "epoch": 0.1584,
+      "grad_norm": 2.739799737930298,
+      "learning_rate": 6.856761511570944e-06,
+      "loss": 0.3568,
+      "step": 396
+    },
+    {
+      "epoch": 0.1592,
+      "grad_norm": 10.551313400268555,
+      "learning_rate": 6.883283631520579e-06,
+      "loss": 0.5487,
+      "step": 398
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.61369800567627,
+      "learning_rate": 6.909830056250522e-06,
+      "loss": 0.5602,
+      "step": 400
+    },
+    {
+      "epoch": 0.1608,
+      "grad_norm": 8.547295570373535,
+      "learning_rate": 6.936400578746436e-06,
+      "loss": 0.6884,
+      "step": 402
+    },
+    {
+      "epoch": 0.1616,
+      "grad_norm": 16.45806312561035,
+      "learning_rate": 6.96299499180605e-06,
+      "loss": 0.7216,
+      "step": 404
+    },
+    {
+      "epoch": 0.1624,
+      "grad_norm": 3.695611000061035,
+      "learning_rate": 6.989613088040787e-06,
+      "loss": 0.3861,
+      "step": 406
+    },
+    {
+      "epoch": 0.1632,
+      "grad_norm": 6.287961959838867,
+      "learning_rate": 7.016254659877404e-06,
+      "loss": 0.4634,
+      "step": 408
+    },
+    {
+      "epoch": 0.164,
+      "grad_norm": 7.198334217071533,
+      "learning_rate": 7.042919499559539e-06,
+      "loss": 0.473,
+      "step": 410
+    },
+    {
+      "epoch": 0.1648,
+      "grad_norm": 8.600248336791992,
+      "learning_rate": 7.06960739914943e-06,
+      "loss": 0.6679,
+      "step": 412
+    },
+    {
+      "epoch": 0.1656,
+      "grad_norm": 3.197615146636963,
+      "learning_rate": 7.09631815052946e-06,
+      "loss": 0.3843,
+      "step": 414
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 3.0980639457702637,
+      "learning_rate": 7.123051545403873e-06,
+      "loss": 0.6683,
+      "step": 416
+    },
+    {
+      "epoch": 0.1672,
+      "grad_norm": 6.608954906463623,
+      "learning_rate": 7.1498073753002375e-06,
+      "loss": 0.7753,
+      "step": 418
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 7.584798812866211,
+      "learning_rate": 7.1765854315712325e-06,
+      "loss": 0.5333,
+      "step": 420
+    },
+    {
+      "epoch": 0.1688,
+      "grad_norm": 6.7867960929870605,
+      "learning_rate": 7.203385505396197e-06,
+      "loss": 0.4801,
+      "step": 422
+    },
+    {
+      "epoch": 0.1696,
+      "grad_norm": 2.2217888832092285,
+      "learning_rate": 7.230207387782771e-06,
+      "loss": 0.2836,
+      "step": 424
+    },
+    {
+      "epoch": 0.1704,
+      "grad_norm": 3.6246209144592285,
+      "learning_rate": 7.257050869568527e-06,
+      "loss": 0.4577,
+      "step": 426
+    },
+    {
+      "epoch": 0.1712,
+      "grad_norm": 3.1784732341766357,
+      "learning_rate": 7.28391574142262e-06,
+      "loss": 0.5712,
+      "step": 428
+    },
+    {
+      "epoch": 0.172,
+      "grad_norm": 2.8604462146759033,
+      "learning_rate": 7.3108017938473485e-06,
+      "loss": 0.3046,
+      "step": 430
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 3.685370922088623,
+      "learning_rate": 7.337708817179875e-06,
+      "loss": 0.6305,
+      "step": 432
+    },
+    {
+      "epoch": 0.1736,
+      "grad_norm": 7.055061340332031,
+      "learning_rate": 7.36463660159386e-06,
+      "loss": 0.5632,
+      "step": 434
+    },
+    {
+      "epoch": 0.1744,
+      "grad_norm": 8.530369758605957,
+      "learning_rate": 7.39158493710103e-06,
+      "loss": 0.6991,
+      "step": 436
+    },
+    {
+      "epoch": 0.1752,
+      "grad_norm": 3.476318597793579,
+      "learning_rate": 7.418553613552822e-06,
+      "loss": 0.4436,
+      "step": 438
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 2.979520559310913,
+      "learning_rate": 7.445542420642091e-06,
+      "loss": 0.3924,
+      "step": 440
+    },
+    {
+      "epoch": 0.1768,
+      "grad_norm": 6.099870681762695,
+      "learning_rate": 7.472551147904703e-06,
+      "loss": 0.4088,
+      "step": 442
+    },
+    {
+      "epoch": 0.1776,
+      "grad_norm": 8.147076606750488,
+      "learning_rate": 7.499579584721173e-06,
+      "loss": 0.4561,
+      "step": 444
+    },
+    {
+      "epoch": 0.1784,
+      "grad_norm": 6.481442451477051,
+      "learning_rate": 7.5266275203183395e-06,
+      "loss": 0.5383,
+      "step": 446
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 2.7729148864746094,
+      "learning_rate": 7.553694743770917e-06,
+      "loss": 0.3598,
+      "step": 448
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.662606239318848,
+      "learning_rate": 7.580781044003312e-06,
+      "loss": 0.6492,
+      "step": 450
+    },
+    {
+      "epoch": 0.1808,
+      "grad_norm": 7.1027374267578125,
+      "learning_rate": 7.607886209791095e-06,
+      "loss": 0.8726,
+      "step": 452
+    },
+    {
+      "epoch": 0.1816,
+      "grad_norm": 3.737102508544922,
+      "learning_rate": 7.635010029762755e-06,
+      "loss": 0.4625,
+      "step": 454
+    },
+    {
+      "epoch": 0.1824,
+      "grad_norm": 7.406997203826904,
+      "learning_rate": 7.662152292401265e-06,
+      "loss": 0.6214,
+      "step": 456
+    },
+    {
+      "epoch": 0.1832,
+      "grad_norm": 6.555530548095703,
+      "learning_rate": 7.689312786045822e-06,
+      "loss": 0.7548,
+      "step": 458
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 3.5676968097686768,
+      "learning_rate": 7.716491298893441e-06,
+      "loss": 0.3623,
+      "step": 460
+    },
+    {
+      "epoch": 0.1848,
+      "grad_norm": 5.894664287567139,
+      "learning_rate": 7.74368761900062e-06,
+      "loss": 0.5739,
+      "step": 462
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 4.684706687927246,
+      "learning_rate": 7.770901534284991e-06,
+      "loss": 0.6722,
+      "step": 464
+    },
+    {
+      "epoch": 0.1864,
+      "grad_norm": 8.258028984069824,
+      "learning_rate": 7.798132832526976e-06,
+      "loss": 0.4903,
+      "step": 466
+    },
+    {
+      "epoch": 0.1872,
+      "grad_norm": 11.913901329040527,
+      "learning_rate": 7.825381301371444e-06,
+      "loss": 0.3768,
+      "step": 468
+    },
+    {
+      "epoch": 0.188,
+      "grad_norm": 2.9942965507507324,
+      "learning_rate": 7.852646728329358e-06,
+      "loss": 0.4586,
+      "step": 470
+    },
+    {
+      "epoch": 0.1888,
+      "grad_norm": 9.535568237304688,
+      "learning_rate": 7.879928900779441e-06,
+      "loss": 0.6082,
+      "step": 472
+    },
+    {
+      "epoch": 0.1896,
+      "grad_norm": 11.259536743164062,
+      "learning_rate": 7.907227605969852e-06,
+      "loss": 0.859,
+      "step": 474
+    },
+    {
+      "epoch": 0.1904,
+      "grad_norm": 10.543761253356934,
+      "learning_rate": 7.934542631019767e-06,
+      "loss": 0.4658,
+      "step": 476
+    },
+    {
+      "epoch": 0.1912,
+      "grad_norm": 9.306256294250488,
+      "learning_rate": 7.961873762921153e-06,
+      "loss": 0.9107,
+      "step": 478
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 3.1275129318237305,
+      "learning_rate": 7.989220788540351e-06,
+      "loss": 0.4828,
+      "step": 480
+    },
+    {
+      "epoch": 0.1928,
+      "grad_norm": 10.207925796508789,
+      "learning_rate": 8.016583494619764e-06,
+      "loss": 0.3932,
+      "step": 482
+    },
+    {
+      "epoch": 0.1936,
+      "grad_norm": 3.0926074981689453,
+      "learning_rate": 8.043961667779511e-06,
+      "loss": 0.514,
+      "step": 484
+    },
+    {
+      "epoch": 0.1944,
+      "grad_norm": 6.525674343109131,
+      "learning_rate": 8.071355094519103e-06,
+      "loss": 0.3918,
+      "step": 486
+    },
+    {
+      "epoch": 0.1952,
+      "grad_norm": 2.066398859024048,
+      "learning_rate": 8.098763561219089e-06,
+      "loss": 0.3501,
+      "step": 488
+    },
+    {
+      "epoch": 0.196,
+      "grad_norm": 4.382544040679932,
+      "learning_rate": 8.126186854142744e-06,
+      "loss": 0.4171,
+      "step": 490
+    },
+    {
+      "epoch": 0.1968,
+      "grad_norm": 2.5938968658447266,
+      "learning_rate": 8.153624759437718e-06,
+      "loss": 0.3604,
+      "step": 492
+    },
+    {
+      "epoch": 0.1976,
+      "grad_norm": 5.127084255218506,
+      "learning_rate": 8.181077063137735e-06,
+      "loss": 0.4352,
+      "step": 494
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 2.662642478942871,
+      "learning_rate": 8.208543551164178e-06,
+      "loss": 0.4609,
+      "step": 496
+    },
+    {
+      "epoch": 0.1992,
+      "grad_norm": 6.415817737579346,
+      "learning_rate": 8.236024009327877e-06,
+      "loss": 0.6271,
+      "step": 498
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.131330490112305,
+      "learning_rate": 8.263518223330695e-06,
+      "loss": 0.5425,
+      "step": 500
+    },
+    {
+      "epoch": 0.2008,
+      "grad_norm": 4.420889377593994,
+      "learning_rate": 8.29102597876723e-06,
+      "loss": 0.5649,
+      "step": 502
+    },
+    {
+      "epoch": 0.2016,
+      "grad_norm": 5.390644073486328,
+      "learning_rate": 8.31854706112648e-06,
+      "loss": 0.6567,
+      "step": 504
+    },
+    {
+      "epoch": 0.2024,
+      "grad_norm": 7.901724338531494,
+      "learning_rate": 8.346081255793516e-06,
+      "loss": 0.4477,
+      "step": 506
+    },
+    {
+      "epoch": 0.2032,
+      "grad_norm": 3.4163174629211426,
+      "learning_rate": 8.373628348051156e-06,
+      "loss": 0.4377,
+      "step": 508
+    },
+    {
+      "epoch": 0.204,
+      "grad_norm": 5.113077163696289,
+      "learning_rate": 8.401188123081642e-06,
+      "loss": 0.3727,
+      "step": 510
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 2.9778358936309814,
+      "learning_rate": 8.428760365968329e-06,
+      "loss": 0.5046,
+      "step": 512
+    },
+    {
+      "epoch": 0.2056,
+      "grad_norm": 9.621271133422852,
+      "learning_rate": 8.456344861697293e-06,
+      "loss": 0.5682,
+      "step": 514
+    },
+    {
+      "epoch": 0.2064,
+      "grad_norm": 3.1639251708984375,
+      "learning_rate": 8.483941395159114e-06,
+      "loss": 0.4007,
+      "step": 516
+    },
+    {
+      "epoch": 0.2072,
+      "grad_norm": 14.570175170898438,
+      "learning_rate": 8.511549751150478e-06,
+      "loss": 0.6807,
+      "step": 518
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 4.703779220581055,
+      "learning_rate": 8.539169714375883e-06,
+      "loss": 0.362,
+      "step": 520
+    },
+    {
+      "epoch": 0.2088,
+      "grad_norm": 8.251890182495117,
+      "learning_rate": 8.566801069449304e-06,
+      "loss": 0.669,
+      "step": 522
+    },
+    {
+      "epoch": 0.2096,
+      "grad_norm": 7.932413101196289,
+      "learning_rate": 8.594443600895886e-06,
+      "loss": 0.5107,
+      "step": 524
+    },
+    {
+      "epoch": 0.2104,
+      "grad_norm": 3.570723533630371,
+      "learning_rate": 8.622097093153612e-06,
+      "loss": 0.3977,
+      "step": 526
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 7.7874979972839355,
+      "learning_rate": 8.649761330575e-06,
+      "loss": 0.5925,
+      "step": 528
+    },
+    {
+      "epoch": 0.212,
+      "grad_norm": 6.07938814163208,
+      "learning_rate": 8.677436097428766e-06,
+      "loss": 0.5776,
+      "step": 530
+    },
+    {
+      "epoch": 0.2128,
+      "grad_norm": 5.536543369293213,
+      "learning_rate": 8.705121177901537e-06,
+      "loss": 0.4131,
+      "step": 532
+    },
+    {
+      "epoch": 0.2136,
+      "grad_norm": 3.3589935302734375,
+      "learning_rate": 8.732816356099459e-06,
+      "loss": 0.4839,
+      "step": 534
+    },
+    {
+      "epoch": 0.2144,
+      "grad_norm": 2.7732083797454834,
+      "learning_rate": 8.760521416049986e-06,
+      "loss": 0.2898,
+      "step": 536
+    },
+    {
+      "epoch": 0.2152,
+      "grad_norm": 3.485111713409424,
+      "learning_rate": 8.788236141703477e-06,
+      "loss": 0.2139,
+      "step": 538
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 5.76864767074585,
+      "learning_rate": 8.81596031693499e-06,
+      "loss": 0.445,
+      "step": 540
+    },
+    {
+      "epoch": 0.2168,
+      "grad_norm": 3.4889819622039795,
+      "learning_rate": 8.84369372554578e-06,
+      "loss": 0.3836,
+      "step": 542
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 3.0347912311553955,
+      "learning_rate": 8.87143615126518e-06,
+      "loss": 0.2902,
+      "step": 544
+    },
+    {
+      "epoch": 0.2184,
+      "grad_norm": 8.731011390686035,
+      "learning_rate": 8.899187377752173e-06,
+      "loss": 0.4252,
+      "step": 546
+    },
+    {
+      "epoch": 0.2192,
+      "grad_norm": 2.755082130432129,
+      "learning_rate": 8.926947188597127e-06,
+      "loss": 0.3409,
+      "step": 548
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.3828043937683105,
+      "learning_rate": 8.954715367323473e-06,
+      "loss": 0.416,
+      "step": 550
+    },
+    {
+      "epoch": 0.2208,
+      "grad_norm": 2.2503442764282227,
+      "learning_rate": 8.982491697389344e-06,
+      "loss": 0.655,
+      "step": 552
+    },
+    {
+      "epoch": 0.2216,
+      "grad_norm": 6.455631732940674,
+      "learning_rate": 9.010275962189356e-06,
+      "loss": 0.5366,
+      "step": 554
+    },
+    {
+      "epoch": 0.2224,
+      "grad_norm": 4.30733060836792,
+      "learning_rate": 9.03806794505621e-06,
+      "loss": 0.5306,
+      "step": 556
+    },
+    {
+      "epoch": 0.2232,
+      "grad_norm": 10.166731834411621,
+      "learning_rate": 9.065867429262497e-06,
+      "loss": 1.083,
+      "step": 558
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 3.9657294750213623,
+      "learning_rate": 9.093674198022198e-06,
+      "loss": 0.675,
+      "step": 560
+    },
+    {
+      "epoch": 0.2248,
+      "grad_norm": 11.929780006408691,
+      "learning_rate": 9.121488034492567e-06,
+      "loss": 0.6316,
+      "step": 562
+    },
+    {
+      "epoch": 0.2256,
+      "grad_norm": 10.191093444824219,
+      "learning_rate": 9.149308721775717e-06,
+      "loss": 0.6091,
+      "step": 564
+    },
+    {
+      "epoch": 0.2264,
+      "grad_norm": 7.18539571762085,
+      "learning_rate": 9.177136042920338e-06,
+      "loss": 0.403,
+      "step": 566
+    },
+    {
+      "epoch": 0.2272,
+      "grad_norm": 3.656543016433716,
+      "learning_rate": 9.204969780923396e-06,
+      "loss": 0.6639,
+      "step": 568
+    },
+    {
+      "epoch": 0.228,
+      "grad_norm": 7.359060764312744,
+      "learning_rate": 9.232809718731822e-06,
+      "loss": 0.3566,
+      "step": 570
+    },
+    {
+      "epoch": 0.2288,
+      "grad_norm": 9.004820823669434,
+      "learning_rate": 9.26065563924414e-06,
+      "loss": 0.6622,
+      "step": 572
+    },
+    {
+      "epoch": 0.2296,
+      "grad_norm": 10.838543891906738,
+      "learning_rate": 9.288507325312319e-06,
+      "loss": 0.6195,
+      "step": 574
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 2.975687265396118,
+      "learning_rate": 9.316364559743298e-06,
+      "loss": 0.3825,
+      "step": 576
+    },
+    {
+      "epoch": 0.2312,
+      "grad_norm": 8.698026657104492,
+      "learning_rate": 9.344227125300788e-06,
+      "loss": 1.0027,
+      "step": 578
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 5.095306396484375,
+      "learning_rate": 9.372094804706867e-06,
+      "loss": 0.4516,
+      "step": 580
+    },
+    {
+      "epoch": 0.2328,
+      "grad_norm": 11.336697578430176,
+      "learning_rate": 9.39996738064379e-06,
+      "loss": 1.0363,
+      "step": 582
+    },
+    {
+      "epoch": 0.2336,
+      "grad_norm": 9.10135269165039,
+      "learning_rate": 9.427844635755615e-06,
+      "loss": 0.437,
+      "step": 584
+    },
+    {
+      "epoch": 0.2344,
+      "grad_norm": 5.346004962921143,
+      "learning_rate": 9.455726352649904e-06,
+      "loss": 0.476,
+      "step": 586
+    },
+    {
+      "epoch": 0.2352,
+      "grad_norm": 6.9864983558654785,
+      "learning_rate": 9.483612313899446e-06,
+      "loss": 0.6293,
+      "step": 588
+    },
+    {
+      "epoch": 0.236,
+      "grad_norm": 3.372296094894409,
+      "learning_rate": 9.511502302043859e-06,
+      "loss": 0.4772,
+      "step": 590
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 6.616199493408203,
+      "learning_rate": 9.539396099591469e-06,
+      "loss": 0.6176,
+      "step": 592
+    },
+    {
+      "epoch": 0.2376,
+      "grad_norm": 10.571331977844238,
+      "learning_rate": 9.567293489020816e-06,
+      "loss": 0.6476,
+      "step": 594
+    },
+    {
+      "epoch": 0.2384,
+      "grad_norm": 6.520980358123779,
+      "learning_rate": 9.595194252782461e-06,
+      "loss": 0.4728,
+      "step": 596
+    },
+    {
+      "epoch": 0.2392,
+      "grad_norm": 5.69945764541626,
+      "learning_rate": 9.623098173300656e-06,
+      "loss": 0.3048,
+      "step": 598
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8320627212524414,
+      "learning_rate": 9.651005032974991e-06,
+      "loss": 0.3585,
+      "step": 600
+    },
+    {
+      "epoch": 0.2408,
+      "grad_norm": 5.430441379547119,
+      "learning_rate": 9.678914614182184e-06,
+      "loss": 0.362,
+      "step": 602
+    },
+    {
+      "epoch": 0.2416,
+      "grad_norm": 6.453746795654297,
+      "learning_rate": 9.706826699277714e-06,
+      "loss": 0.5644,
+      "step": 604
+    },
+    {
+      "epoch": 0.2424,
+      "grad_norm": 5.222368240356445,
+      "learning_rate": 9.734741070597535e-06,
+      "loss": 0.4847,
+      "step": 606
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 8.284361839294434,
+      "learning_rate": 9.762657510459774e-06,
+      "loss": 0.3067,
+      "step": 608
+    },
+    {
+      "epoch": 0.244,
+      "grad_norm": 2.8420135974884033,
+      "learning_rate": 9.790575801166422e-06,
+      "loss": 0.4242,
+      "step": 610
+    },
+    {
+      "epoch": 0.2448,
+      "grad_norm": 3.083292245864868,
+      "learning_rate": 9.818495725005043e-06,
+      "loss": 1.6922,
+      "step": 612
+    },
+    {
+      "epoch": 0.2456,
+      "grad_norm": 7.784399509429932,
+      "learning_rate": 9.846417064250459e-06,
+      "loss": 0.4656,
+      "step": 614
+    },
+    {
+      "epoch": 0.2464,
+      "grad_norm": 2.823225736618042,
+      "learning_rate": 9.874339601166479e-06,
+      "loss": 0.6133,
+      "step": 616
+    },
+    {
+      "epoch": 0.2472,
+      "grad_norm": 2.745265483856201,
+      "learning_rate": 9.902263118007513e-06,
+      "loss": 0.4453,
+      "step": 618
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 6.4653191566467285,
+      "learning_rate": 9.930187397020385e-06,
+      "loss": 0.4293,
+      "step": 620
+    },
+    {
+      "epoch": 0.2488,
+      "grad_norm": 7.129982948303223,
+      "learning_rate": 9.95811222044596e-06,
+      "loss": 0.662,
+      "step": 622
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 3.3546640872955322,
+      "learning_rate": 9.986037370520855e-06,
+      "loss": 0.4132,
+      "step": 624
+    },
+    {
+      "epoch": 0.2504,
+      "grad_norm": 6.033866882324219,
+      "learning_rate": 1.0013962629479139e-05,
+      "loss": 0.4106,
+      "step": 626
+    },
+    {
+      "epoch": 0.2512,
+      "grad_norm": 2.7721498012542725,
+      "learning_rate": 1.0041887779554034e-05,
+      "loss": 0.435,
+      "step": 628
+    },
+    {
+      "epoch": 0.252,
+      "grad_norm": 4.638468265533447,
+      "learning_rate": 1.0069812602979607e-05,
+      "loss": 0.3248,
+      "step": 630
+    },
+    {
+      "epoch": 0.2528,
+      "grad_norm": 4.487112522125244,
+      "learning_rate": 1.0097736881992482e-05,
+      "loss": 0.2748,
+      "step": 632
+    },
+    {
+      "epoch": 0.2536,
+      "grad_norm": 3.1372745037078857,
+      "learning_rate": 1.0125660398833514e-05,
+      "loss": 0.5276,
+      "step": 634
+    },
+    {
+      "epoch": 0.2544,
+      "grad_norm": 3.534034013748169,
+      "learning_rate": 1.0153582935749533e-05,
+      "loss": 0.3561,
+      "step": 636
+    },
+    {
+      "epoch": 0.2552,
+      "grad_norm": 3.8285937309265137,
+      "learning_rate": 1.0181504274994952e-05,
+      "loss": 0.2581,
+      "step": 638
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 8.048152923583984,
+      "learning_rate": 1.0209424198833571e-05,
+      "loss": 0.6331,
+      "step": 640
+    },
+    {
+      "epoch": 0.2568,
+      "grad_norm": 14.114180564880371,
+      "learning_rate": 1.0237342489540218e-05,
+      "loss": 0.5025,
+      "step": 642
+    },
+    {
+      "epoch": 0.2576,
+      "grad_norm": 2.971632480621338,
+      "learning_rate": 1.0265258929402458e-05,
+      "loss": 0.3162,
+      "step": 644
+    },
+    {
+      "epoch": 0.2584,
+      "grad_norm": 3.681770086288452,
+      "learning_rate": 1.029317330072228e-05,
+      "loss": 0.555,
+      "step": 646
+    },
+    {
+      "epoch": 0.2592,
+      "grad_norm": 6.693310260772705,
+      "learning_rate": 1.0321085385817811e-05,
+      "loss": 0.4238,
+      "step": 648
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.489689350128174,
+      "learning_rate": 1.0348994967025004e-05,
+      "loss": 0.5693,
+      "step": 650
+    },
+    {
+      "epoch": 0.2608,
+      "grad_norm": 10.314934730529785,
+      "learning_rate": 1.0376901826699337e-05,
+      "loss": 0.444,
+      "step": 652
+    },
+    {
+      "epoch": 0.2616,
+      "grad_norm": 8.000083923339844,
+      "learning_rate": 1.0404805747217532e-05,
+      "loss": 1.1624,
+      "step": 654
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 3.0513830184936523,
+      "learning_rate": 1.0432706510979175e-05,
+      "loss": 0.3671,
+      "step": 656
+    },
+    {
+      "epoch": 0.2632,
+      "grad_norm": 5.537996768951416,
+      "learning_rate": 1.0460603900408526e-05,
+      "loss": 0.3633,
+      "step": 658
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 3.6096031665802,
+      "learning_rate": 1.0488497697956134e-05,
+      "loss": 0.3869,
+      "step": 660
+    },
+    {
+      "epoch": 0.2648,
+      "grad_norm": 3.872437000274658,
+      "learning_rate": 1.0516387686100549e-05,
+      "loss": 0.2569,
+      "step": 662
+    },
+    {
+      "epoch": 0.2656,
+      "grad_norm": 19.229639053344727,
+      "learning_rate": 1.054427364735009e-05,
+      "loss": 1.5103,
+      "step": 664
+    },
+    {
+      "epoch": 0.2664,
+      "grad_norm": 4.366555213928223,
+      "learning_rate": 1.0572155364244378e-05,
+      "loss": 0.4739,
+      "step": 666
+    },
+    {
+      "epoch": 0.2672,
+      "grad_norm": 12.398948669433594,
+      "learning_rate": 1.0600032619356203e-05,
+      "loss": 0.7889,
+      "step": 668
+    },
+    {
+      "epoch": 0.268,
+      "grad_norm": 14.852575302124023,
+      "learning_rate": 1.0627905195293127e-05,
+      "loss": 0.9629,
+      "step": 670
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 10.531519889831543,
+      "learning_rate": 1.0655772874699206e-05,
+      "loss": 0.8866,
+      "step": 672
+    },
+    {
+      "epoch": 0.2696,
+      "grad_norm": 9.607449531555176,
+      "learning_rate": 1.0683635440256694e-05,
+      "loss": 0.8125,
+      "step": 674
+    },
+    {
+      "epoch": 0.2704,
+      "grad_norm": 4.8872785568237305,
+      "learning_rate": 1.0711492674687674e-05,
+      "loss": 0.3938,
+      "step": 676
+    },
+    {
+      "epoch": 0.2712,
+      "grad_norm": 11.146080017089844,
+      "learning_rate": 1.0739344360755855e-05,
+      "loss": 0.8325,
+      "step": 678
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 4.164681434631348,
+      "learning_rate": 1.0767190281268171e-05,
+      "loss": 0.4736,
+      "step": 680
+    },
+    {
+      "epoch": 0.2728,
+      "grad_norm": 7.81128454208374,
+      "learning_rate": 1.07950302190766e-05,
+      "loss": 0.3557,
+      "step": 682
+    },
+    {
+      "epoch": 0.2736,
+      "grad_norm": 4.940060138702393,
+      "learning_rate": 1.0822863957079654e-05,
+      "loss": 0.2446,
+      "step": 684
+    },
+    {
+      "epoch": 0.2744,
+      "grad_norm": 3.599069118499756,
+      "learning_rate": 1.0850691278224277e-05,
+      "loss": 0.3247,
+      "step": 686
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 2.723473072052002,
+      "learning_rate": 1.0878511965507428e-05,
+      "loss": 0.4377,
+      "step": 688
+    },
+    {
+      "epoch": 0.276,
+      "grad_norm": 9.285993576049805,
+      "learning_rate": 1.0906325801977795e-05,
+      "loss": 0.5683,
+      "step": 690
+    },
+    {
+      "epoch": 0.2768,
+      "grad_norm": 6.284435272216797,
+      "learning_rate": 1.0934132570737497e-05,
+      "loss": 0.4374,
+      "step": 692
+    },
+    {
+      "epoch": 0.2776,
+      "grad_norm": 3.0146565437316895,
+      "learning_rate": 1.0961932054943785e-05,
+      "loss": 0.36,
+      "step": 694
+    },
+    {
+      "epoch": 0.2784,
+      "grad_norm": 8.176236152648926,
+      "learning_rate": 1.098972403781064e-05,
+      "loss": 0.7033,
+      "step": 696
+    },
+    {
+      "epoch": 0.2792,
+      "grad_norm": 7.648057460784912,
+      "learning_rate": 1.101750830261065e-05,
+      "loss": 0.4907,
+      "step": 698
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.528674602508545,
+      "learning_rate": 1.104528463267652e-05,
+      "loss": 0.489,
+      "step": 700
+    },
+    {
+      "epoch": 0.2808,
+      "grad_norm": 5.457291603088379,
+      "learning_rate": 1.1073052811402867e-05,
+      "loss": 0.8648,
+      "step": 702
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 7.5247344970703125,
+      "learning_rate": 1.1100812622247821e-05,
+      "loss": 0.3886,
+      "step": 704
+    },
+    {
+      "epoch": 0.2824,
+      "grad_norm": 9.832022666931152,
+      "learning_rate": 1.1128563848734815e-05,
+      "loss": 0.7117,
+      "step": 706
+    },
+    {
+      "epoch": 0.2832,
+      "grad_norm": 12.296405792236328,
+      "learning_rate": 1.1156306274454211e-05,
+      "loss": 0.6653,
+      "step": 708
+    },
+    {
+      "epoch": 0.284,
+      "grad_norm": 3.188476324081421,
+      "learning_rate": 1.1184039683065002e-05,
+      "loss": 0.4074,
+      "step": 710
+    },
+    {
+      "epoch": 0.2848,
+      "grad_norm": 8.455747604370117,
+      "learning_rate": 1.1211763858296516e-05,
+      "loss": 0.803,
+      "step": 712
+    },
+    {
+      "epoch": 0.2856,
+      "grad_norm": 2.5128846168518066,
+      "learning_rate": 1.1239478583950007e-05,
+      "loss": 0.6004,
+      "step": 714
+    },
+    {
+      "epoch": 0.2864,
+      "grad_norm": 6.20869255065918,
+      "learning_rate": 1.1267183643900534e-05,
+      "loss": 0.2269,
+      "step": 716
+    },
+    {
+      "epoch": 0.2872,
+      "grad_norm": 3.149703025817871,
+      "learning_rate": 1.1294878822098456e-05,
+      "loss": 0.3468,
+      "step": 718
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 3.046736717224121,
+      "learning_rate": 1.1322563902571227e-05,
+      "loss": 0.435,
+      "step": 720
+    },
+    {
+      "epoch": 0.2888,
+      "grad_norm": 3.9733293056488037,
+      "learning_rate": 1.1350238669424993e-05,
+      "loss": 0.336,
+      "step": 722
+    },
+    {
+      "epoch": 0.2896,
+      "grad_norm": 5.300455093383789,
+      "learning_rate": 1.137790290684638e-05,
+      "loss": 0.4059,
+      "step": 724
+    },
+    {
+      "epoch": 0.2904,
+      "grad_norm": 6.307999134063721,
+      "learning_rate": 1.1405556399104108e-05,
+      "loss": 0.5065,
+      "step": 726
+    },
+    {
+      "epoch": 0.2912,
+      "grad_norm": 7.641139030456543,
+      "learning_rate": 1.143319893055069e-05,
+      "loss": 0.8458,
+      "step": 728
+    },
+    {
+      "epoch": 0.292,
+      "grad_norm": 6.329631328582764,
+      "learning_rate": 1.1460830285624112e-05,
+      "loss": 0.3936,
+      "step": 730
+    },
+    {
+      "epoch": 0.2928,
+      "grad_norm": 3.9163966178894043,
+      "learning_rate": 1.1488450248849515e-05,
+      "loss": 0.2481,
+      "step": 732
+    },
+    {
+      "epoch": 0.2936,
+      "grad_norm": 4.952850341796875,
+      "learning_rate": 1.1516058604840881e-05,
+      "loss": 0.5636,
+      "step": 734
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 4.797139644622803,
+      "learning_rate": 1.15436551383027e-05,
+      "loss": 0.3305,
+      "step": 736
+    },
+    {
+      "epoch": 0.2952,
+      "grad_norm": 5.792111873626709,
+      "learning_rate": 1.1571239634031666e-05,
+      "loss": 0.3829,
+      "step": 738
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 2.61521053314209,
+      "learning_rate": 1.1598811876918352e-05,
+      "loss": 0.4863,
+      "step": 740
+    },
+    {
+      "epoch": 0.2968,
+      "grad_norm": 4.18222188949585,
+      "learning_rate": 1.1626371651948839e-05,
+      "loss": 0.6332,
+      "step": 742
+    },
+    {
+      "epoch": 0.2976,
+      "grad_norm": 4.946098327636719,
+      "learning_rate": 1.1653918744206478e-05,
+      "loss": 0.4458,
+      "step": 744
+    },
+    {
+      "epoch": 0.2984,
+      "grad_norm": 5.32734489440918,
+      "learning_rate": 1.1681452938873515e-05,
+      "loss": 0.3399,
+      "step": 746
+    },
+    {
+      "epoch": 0.2992,
+      "grad_norm": 4.0147600173950195,
+      "learning_rate": 1.1708974021232763e-05,
+      "loss": 0.4765,
+      "step": 748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.177061080932617,
+      "learning_rate": 1.1736481776669297e-05,
+      "loss": 0.3631,
+      "step": 750
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 5.71055793762207,
+      "learning_rate": 1.1763975990672116e-05,
+      "loss": 0.3434,
+      "step": 752
+    },
+    {
+      "epoch": 0.3016,
+      "grad_norm": 2.9367713928222656,
+      "learning_rate": 1.1791456448835815e-05,
+      "loss": 0.338,
+      "step": 754
+    },
+    {
+      "epoch": 0.3024,
+      "grad_norm": 5.17765474319458,
+      "learning_rate": 1.1818922936862258e-05,
+      "loss": 0.2791,
+      "step": 756
+    },
+    {
+      "epoch": 0.3032,
+      "grad_norm": 11.932939529418945,
+      "learning_rate": 1.1846375240562274e-05,
+      "loss": 1.1732,
+      "step": 758
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 3.952331304550171,
+      "learning_rate": 1.187381314585725e-05,
+      "loss": 0.3083,
+      "step": 760
+    },
+    {
+      "epoch": 0.3048,
+      "grad_norm": 2.7587730884552,
+      "learning_rate": 1.1901236438780906e-05,
+      "loss": 0.3518,
+      "step": 762
+    },
+    {
+      "epoch": 0.3056,
+      "grad_norm": 1.007660984992981,
+      "learning_rate": 1.192864490548089e-05,
+      "loss": 0.3737,
+      "step": 764
+    },
+    {
+      "epoch": 0.3064,
+      "grad_norm": 2.9483938217163086,
+      "learning_rate": 1.195603833222048e-05,
+      "loss": 0.2933,
+      "step": 766
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 2.8194210529327393,
+      "learning_rate": 1.198341650538023e-05,
+      "loss": 0.5113,
+      "step": 768
+    },
+    {
+      "epoch": 0.308,
+      "grad_norm": 6.031118869781494,
+      "learning_rate": 1.2010779211459642e-05,
+      "loss": 0.8163,
+      "step": 770
+    },
+    {
+      "epoch": 0.3088,
+      "grad_norm": 3.5361063480377197,
+      "learning_rate": 1.203812623707884e-05,
+      "loss": 0.7252,
+      "step": 772
+    },
+    {
+      "epoch": 0.3096,
+      "grad_norm": 3.8909547328948975,
+      "learning_rate": 1.2065457368980227e-05,
+      "loss": 0.4168,
+      "step": 774
+    },
+    {
+      "epoch": 0.3104,
+      "grad_norm": 3.7777271270751953,
+      "learning_rate": 1.2092772394030141e-05,
+      "loss": 0.9927,
+      "step": 776
+    },
+    {
+      "epoch": 0.3112,
+      "grad_norm": 4.374517440795898,
+      "learning_rate": 1.2120071099220552e-05,
+      "loss": 0.5356,
+      "step": 778
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 7.112154483795166,
+      "learning_rate": 1.2147353271670637e-05,
+      "loss": 0.4746,
+      "step": 780
+    },
+    {
+      "epoch": 0.3128,
+      "grad_norm": 4.3697967529296875,
+      "learning_rate": 1.217461869862855e-05,
+      "loss": 0.5526,
+      "step": 782
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 8.930327415466309,
+      "learning_rate": 1.2201867167473015e-05,
+      "loss": 0.377,
+      "step": 784
+    },
+    {
+      "epoch": 0.3144,
+      "grad_norm": 3.697787046432495,
+      "learning_rate": 1.2229098465715002e-05,
+      "loss": 0.5277,
+      "step": 786
+    },
+    {
+      "epoch": 0.3152,
+      "grad_norm": 6.874600887298584,
+      "learning_rate": 1.2256312380999373e-05,
+      "loss": 0.7206,
+      "step": 788
+    },
+    {
+      "epoch": 0.316,
+      "grad_norm": 3.165311813354492,
+      "learning_rate": 1.2283508701106552e-05,
+      "loss": 0.4289,
+      "step": 790
+    },
+    {
+      "epoch": 0.3168,
+      "grad_norm": 8.956424713134766,
+      "learning_rate": 1.2310687213954173e-05,
+      "loss": 0.5692,
+      "step": 792
+    },
+    {
+      "epoch": 0.3176,
+      "grad_norm": 3.313727855682373,
+      "learning_rate": 1.233784770759873e-05,
+      "loss": 0.4057,
+      "step": 794
+    },
+    {
+      "epoch": 0.3184,
+      "grad_norm": 7.44467830657959,
+      "learning_rate": 1.2364989970237238e-05,
+      "loss": 0.425,
+      "step": 796
+    },
+    {
+      "epoch": 0.3192,
+      "grad_norm": 2.7032885551452637,
+      "learning_rate": 1.23921137902089e-05,
+      "loss": 1.1685,
+      "step": 798
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 21.789257049560547,
+      "learning_rate": 1.241921895599668e-05,
+      "loss": 2.9202,
+      "step": 800
+    },
+    {
+      "epoch": 0.3208,
+      "grad_norm": 8.086333274841309,
+      "learning_rate": 1.2446305256229076e-05,
+      "loss": 0.4955,
+      "step": 802
+    },
+    {
+      "epoch": 0.3216,
+      "grad_norm": 3.5499794483184814,
+      "learning_rate": 1.2473372479681653e-05,
+      "loss": 0.5145,
+      "step": 804
+    },
+    {
+      "epoch": 0.3224,
+      "grad_norm": 7.596743583679199,
+      "learning_rate": 1.2500420415278822e-05,
+      "loss": 0.5452,
+      "step": 806
+    },
+    {
+      "epoch": 0.3232,
+      "grad_norm": 6.792609214782715,
+      "learning_rate": 1.2527448852095292e-05,
+      "loss": 0.4943,
+      "step": 808
+    },
+    {
+      "epoch": 0.324,
+      "grad_norm": 6.553757667541504,
+      "learning_rate": 1.2554457579357902e-05,
+      "loss": 0.6194,
+      "step": 810
+    },
+    {
+      "epoch": 0.3248,
+      "grad_norm": 7.149604797363281,
+      "learning_rate": 1.2581446386447171e-05,
+      "loss": 0.5983,
+      "step": 812
+    },
+    {
+      "epoch": 0.3256,
+      "grad_norm": 3.606539726257324,
+      "learning_rate": 1.2608415062898963e-05,
+      "loss": 0.5224,
+      "step": 814
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 2.280977725982666,
+      "learning_rate": 1.2635363398406133e-05,
+      "loss": 0.623,
+      "step": 816
+    },
+    {
+      "epoch": 0.3272,
+      "grad_norm": 10.369894981384277,
+      "learning_rate": 1.266229118282012e-05,
+      "loss": 0.8874,
+      "step": 818
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 4.448359489440918,
+      "learning_rate": 1.2689198206152644e-05,
+      "loss": 0.352,
+      "step": 820
+    },
+    {
+      "epoch": 0.3288,
+      "grad_norm": 2.4175426959991455,
+      "learning_rate": 1.2716084258577373e-05,
+      "loss": 0.4758,
+      "step": 822
+    },
+    {
+      "epoch": 0.3296,
+      "grad_norm": 5.202447891235352,
+      "learning_rate": 1.2742949130431468e-05,
+      "loss": 0.3762,
+      "step": 824
+    },
+    {
+      "epoch": 0.3304,
+      "grad_norm": 10.678263664245605,
+      "learning_rate": 1.2769792612217224e-05,
+      "loss": 0.7578,
+      "step": 826
+    },
+    {
+      "epoch": 0.3312,
+      "grad_norm": 10.530842781066895,
+      "learning_rate": 1.2796614494603795e-05,
+      "loss": 0.5892,
+      "step": 828
+    },
+    {
+      "epoch": 0.332,
+      "grad_norm": 2.2692458629608154,
+      "learning_rate": 1.282341456842876e-05,
+      "loss": 0.4299,
+      "step": 830
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 29.20014762878418,
+      "learning_rate": 1.2850192624699756e-05,
+      "loss": 2.1729,
+      "step": 832
+    },
+    {
+      "epoch": 0.3336,
+      "grad_norm": 4.95890998840332,
+      "learning_rate": 1.2876948454596122e-05,
+      "loss": 0.7695,
+      "step": 834
+    },
+    {
+      "epoch": 0.3344,
+      "grad_norm": 4.845802307128906,
+      "learning_rate": 1.2903681849470535e-05,
+      "loss": 0.5292,
+      "step": 836
+    },
+    {
+      "epoch": 0.3352,
+      "grad_norm": 8.932195663452148,
+      "learning_rate": 1.2930392600850565e-05,
+      "loss": 0.7145,
+      "step": 838
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 5.170701503753662,
+      "learning_rate": 1.2957080500440455e-05,
+      "loss": 0.6364,
+      "step": 840
+    },
+    {
+      "epoch": 0.3368,
+      "grad_norm": 10.944951057434082,
+      "learning_rate": 1.2983745340122589e-05,
+      "loss": 0.6809,
+      "step": 842
+    },
+    {
+      "epoch": 0.3376,
+      "grad_norm": 15.961918830871582,
+      "learning_rate": 1.3010386911959205e-05,
+      "loss": 0.6377,
+      "step": 844
+    },
+    {
+      "epoch": 0.3384,
+      "grad_norm": 7.693042755126953,
+      "learning_rate": 1.3037005008193944e-05,
+      "loss": 0.5823,
+      "step": 846
+    },
+    {
+      "epoch": 0.3392,
+      "grad_norm": 10.104445457458496,
+      "learning_rate": 1.3063599421253556e-05,
+      "loss": 0.4535,
+      "step": 848
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.348386764526367,
+      "learning_rate": 1.309016994374947e-05,
+      "loss": 0.5694,
+      "step": 850
+    },
+    {
+      "epoch": 0.3408,
+      "grad_norm": 1.5034079551696777,
+      "learning_rate": 1.3116716368479415e-05,
+      "loss": 0.3579,
+      "step": 852
+    },
+    {
+      "epoch": 0.3416,
+      "grad_norm": 6.123833179473877,
+      "learning_rate": 1.3143238488429049e-05,
+      "loss": 0.5469,
+      "step": 854
+    },
+    {
+      "epoch": 0.3424,
+      "grad_norm": 2.4656760692596436,
+      "learning_rate": 1.316973609677351e-05,
+      "loss": 0.3558,
+      "step": 856
+    },
+    {
+      "epoch": 0.3432,
+      "grad_norm": 5.257678985595703,
+      "learning_rate": 1.319620898687917e-05,
+      "loss": 0.5961,
+      "step": 858
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 2.2889106273651123,
+      "learning_rate": 1.32226569523051e-05,
+      "loss": 0.4341,
+      "step": 860
+    },
+    {
+      "epoch": 0.3448,
+      "grad_norm": 6.256611347198486,
+      "learning_rate": 1.324907978680475e-05,
+      "loss": 0.4521,
+      "step": 862
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 4.55164909362793,
+      "learning_rate": 1.3275477284327572e-05,
+      "loss": 0.4435,
+      "step": 864
+    },
+    {
+      "epoch": 0.3464,
+      "grad_norm": 3.1022279262542725,
+      "learning_rate": 1.3301849239020537e-05,
+      "loss": 0.6874,
+      "step": 866
+    },
+    {
+      "epoch": 0.3472,
+      "grad_norm": 2.245877981185913,
+      "learning_rate": 1.3328195445229865e-05,
+      "loss": 0.5371,
+      "step": 868
+    },
+    {
+      "epoch": 0.348,
+      "grad_norm": 7.837813377380371,
+      "learning_rate": 1.3354515697502548e-05,
+      "loss": 0.5843,
+      "step": 870
+    },
+    {
+      "epoch": 0.3488,
+      "grad_norm": 6.17982816696167,
+      "learning_rate": 1.338080979058797e-05,
+      "loss": 0.5286,
+      "step": 872
+    },
+    {
+      "epoch": 0.3496,
+      "grad_norm": 11.103405952453613,
+      "learning_rate": 1.340707751943951e-05,
+      "loss": 0.6753,
+      "step": 874
+    },
+    {
+      "epoch": 0.3504,
+      "grad_norm": 14.157888412475586,
+      "learning_rate": 1.3433318679216145e-05,
+      "loss": 0.5989,
+      "step": 876
+    },
+    {
+      "epoch": 0.3512,
+      "grad_norm": 4.318014621734619,
+      "learning_rate": 1.3459533065284039e-05,
+      "loss": 0.3676,
+      "step": 878
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 6.474916934967041,
+      "learning_rate": 1.348572047321814e-05,
+      "loss": 0.5159,
+      "step": 880
+    },
+    {
+      "epoch": 0.3528,
+      "grad_norm": 5.441918849945068,
+      "learning_rate": 1.3511880698803803e-05,
+      "loss": 0.8862,
+      "step": 882
+    },
+    {
+      "epoch": 0.3536,
+      "grad_norm": 6.940057277679443,
+      "learning_rate": 1.3538013538038296e-05,
+      "loss": 0.6308,
+      "step": 884
+    },
+    {
+      "epoch": 0.3544,
+      "grad_norm": 3.7186665534973145,
+      "learning_rate": 1.3564118787132507e-05,
+      "loss": 0.3685,
+      "step": 886
+    },
+    {
+      "epoch": 0.3552,
+      "grad_norm": 19.710111618041992,
+      "learning_rate": 1.3590196242512461e-05,
+      "loss": 1.2655,
+      "step": 888
+    },
+    {
+      "epoch": 0.356,
+      "grad_norm": 2.533984899520874,
+      "learning_rate": 1.361624570082092e-05,
+      "loss": 0.3519,
+      "step": 890
+    },
+    {
+      "epoch": 0.3568,
+      "grad_norm": 11.8716402053833,
+      "learning_rate": 1.364226695891898e-05,
+      "loss": 0.7045,
+      "step": 892
+    },
+    {
+      "epoch": 0.3576,
+      "grad_norm": 2.3505406379699707,
+      "learning_rate": 1.3668259813887637e-05,
+      "loss": 0.492,
+      "step": 894
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 4.525683403015137,
+      "learning_rate": 1.3694224063029386e-05,
+      "loss": 0.475,
+      "step": 896
+    },
+    {
+      "epoch": 0.3592,
+      "grad_norm": 7.988845348358154,
+      "learning_rate": 1.3720159503869806e-05,
+      "loss": 0.3732,
+      "step": 898
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.825455665588379,
+      "learning_rate": 1.374606593415911e-05,
+      "loss": 0.4227,
+      "step": 900
+    },
+    {
+      "epoch": 0.3608,
+      "grad_norm": 2.447044610977173,
+      "learning_rate": 1.377194315187377e-05,
+      "loss": 0.5101,
+      "step": 902
+    },
+    {
+      "epoch": 0.3616,
+      "grad_norm": 0.4026242792606354,
+      "learning_rate": 1.3797790955218014e-05,
+      "loss": 0.3221,
+      "step": 904
+    },
+    {
+      "epoch": 0.3624,
+      "grad_norm": 7.215357780456543,
+      "learning_rate": 1.3823609142625492e-05,
+      "loss": 0.8086,
+      "step": 906
+    },
+    {
+      "epoch": 0.3632,
+      "grad_norm": 3.070342779159546,
+      "learning_rate": 1.3849397512760793e-05,
+      "loss": 0.5274,
+      "step": 908
+    },
+    {
+      "epoch": 0.364,
+      "grad_norm": 4.354795455932617,
+      "learning_rate": 1.3875155864521027e-05,
+      "loss": 0.4501,
+      "step": 910
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 4.162818908691406,
+      "learning_rate": 1.3900883997037393e-05,
+      "loss": 0.2185,
+      "step": 912
+    },
+    {
+      "epoch": 0.3656,
+      "grad_norm": 1.5700749158859253,
+      "learning_rate": 1.3926581709676746e-05,
+      "loss": 0.288,
+      "step": 914
+    },
+    {
+      "epoch": 0.3664,
+      "grad_norm": 4.362253189086914,
+      "learning_rate": 1.3952248802043158e-05,
+      "loss": 0.3545,
+      "step": 916
+    },
+    {
+      "epoch": 0.3672,
+      "grad_norm": 5.418710231781006,
+      "learning_rate": 1.397788507397949e-05,
+      "loss": 0.4328,
+      "step": 918
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 6.52181339263916,
+      "learning_rate": 1.4003490325568956e-05,
+      "loss": 0.3501,
+      "step": 920
+    },
+    {
+      "epoch": 0.3688,
+      "grad_norm": 4.86818265914917,
+      "learning_rate": 1.4029064357136632e-05,
+      "loss": 0.4529,
+      "step": 922
+    },
+    {
+      "epoch": 0.3696,
+      "grad_norm": 6.881949424743652,
+      "learning_rate": 1.4054606969251096e-05,
+      "loss": 0.5725,
+      "step": 924
+    },
+    {
+      "epoch": 0.3704,
+      "grad_norm": 3.368731737136841,
+      "learning_rate": 1.4080117962725929e-05,
+      "loss": 0.4468,
+      "step": 926
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 3.010711908340454,
+      "learning_rate": 1.4105597138621281e-05,
+      "loss": 0.4431,
+      "step": 928
+    },
+    {
+      "epoch": 0.372,
+      "grad_norm": 8.769145011901855,
+      "learning_rate": 1.4131044298245416e-05,
+      "loss": 0.5087,
+      "step": 930
+    },
+    {
+      "epoch": 0.3728,
+      "grad_norm": 7.3292975425720215,
+      "learning_rate": 1.4156459243156275e-05,
+      "loss": 0.4583,
+      "step": 932
+    },
+    {
+      "epoch": 0.3736,
+      "grad_norm": 2.95186448097229,
+      "learning_rate": 1.418184177516301e-05,
+      "loss": 0.4141,
+      "step": 934
+    },
+    {
+      "epoch": 0.3744,
+      "grad_norm": 3.9545629024505615,
+      "learning_rate": 1.420719169632754e-05,
+      "loss": 0.3136,
+      "step": 936
+    },
+    {
+      "epoch": 0.3752,
+      "grad_norm": 8.807525634765625,
+      "learning_rate": 1.4232508808966085e-05,
+      "loss": 0.792,
+      "step": 938
+    },
+    {
+      "epoch": 0.376,
+      "grad_norm": 3.8343663215637207,
+      "learning_rate": 1.4257792915650735e-05,
+      "loss": 0.4076,
+      "step": 940
+    },
+    {
+      "epoch": 0.3768,
+      "grad_norm": 3.995673179626465,
+      "learning_rate": 1.4283043819210906e-05,
+      "loss": 0.5437,
+      "step": 942
+    },
+    {
+      "epoch": 0.3776,
+      "grad_norm": 5.492018222808838,
+      "learning_rate": 1.430826132273499e-05,
+      "loss": 0.4751,
+      "step": 944
+    },
+    {
+      "epoch": 0.3784,
+      "grad_norm": 6.859824180603027,
+      "learning_rate": 1.4333445229571857e-05,
+      "loss": 0.6557,
+      "step": 946
+    },
+    {
+      "epoch": 0.3792,
+      "grad_norm": 8.421869277954102,
+      "learning_rate": 1.4358595343332342e-05,
+      "loss": 0.7003,
+      "step": 948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.7867302894592285,
+      "learning_rate": 1.4383711467890772e-05,
+      "loss": 0.2215,
+      "step": 950
+    },
+    {
+      "epoch": 0.3808,
+      "grad_norm": 8.494775772094727,
+      "learning_rate": 1.4408793407386584e-05,
+      "loss": 0.5345,
+      "step": 952
+    },
+    {
+      "epoch": 0.3816,
+      "grad_norm": 3.8272619247436523,
+      "learning_rate": 1.4433840966225767e-05,
+      "loss": 0.5162,
+      "step": 954
+    },
+    {
+      "epoch": 0.3824,
+      "grad_norm": 6.869154930114746,
+      "learning_rate": 1.4458853949082434e-05,
+      "loss": 0.7501,
+      "step": 956
+    },
+    {
+      "epoch": 0.3832,
+      "grad_norm": 3.7528223991394043,
+      "learning_rate": 1.4483832160900332e-05,
+      "loss": 0.2814,
+      "step": 958
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 12.88850212097168,
+      "learning_rate": 1.4508775406894315e-05,
+      "loss": 0.6589,
+      "step": 960
+    },
+    {
+      "epoch": 0.3848,
+      "grad_norm": 5.3597588539123535,
+      "learning_rate": 1.4533683492551942e-05,
+      "loss": 0.1747,
+      "step": 962
+    },
+    {
+      "epoch": 0.3856,
+      "grad_norm": 4.822731971740723,
+      "learning_rate": 1.4558556223634988e-05,
+      "loss": 0.3057,
+      "step": 964
+    },
+    {
+      "epoch": 0.3864,
+      "grad_norm": 6.9705891609191895,
+      "learning_rate": 1.4583393406180886e-05,
+      "loss": 0.4439,
+      "step": 966
+    },
+    {
+      "epoch": 0.3872,
+      "grad_norm": 7.96857213973999,
+      "learning_rate": 1.460819484650431e-05,
+      "loss": 0.586,
+      "step": 968
+    },
+    {
+      "epoch": 0.388,
+      "grad_norm": 5.491257667541504,
+      "learning_rate": 1.4632960351198618e-05,
+      "loss": 0.5213,
+      "step": 970
+    },
+    {
+      "epoch": 0.3888,
+      "grad_norm": 7.354158878326416,
+      "learning_rate": 1.4657689727137441e-05,
+      "loss": 0.4731,
+      "step": 972
+    },
+    {
+      "epoch": 0.3896,
+      "grad_norm": 8.442399978637695,
+      "learning_rate": 1.468238278147614e-05,
+      "loss": 0.5359,
+      "step": 974
+    },
+    {
+      "epoch": 0.3904,
+      "grad_norm": 1.9848921298980713,
+      "learning_rate": 1.470703932165332e-05,
+      "loss": 0.341,
+      "step": 976
+    },
+    {
+      "epoch": 0.3912,
+      "grad_norm": 2.058201551437378,
+      "learning_rate": 1.4731659155392339e-05,
+      "loss": 0.3913,
+      "step": 978
+    },
+    {
+      "epoch": 0.392,
+      "grad_norm": 8.525607109069824,
+      "learning_rate": 1.4756242090702744e-05,
+      "loss": 0.5909,
+      "step": 980
+    },
+    {
+      "epoch": 0.3928,
+      "grad_norm": 2.9162418842315674,
+      "learning_rate": 1.4780787935881913e-05,
+      "loss": 0.3393,
+      "step": 982
+    },
+    {
+      "epoch": 0.3936,
+      "grad_norm": 2.995479106903076,
+      "learning_rate": 1.4805296499516397e-05,
+      "loss": 0.5719,
+      "step": 984
+    },
+    {
+      "epoch": 0.3944,
+      "grad_norm": 8.575796127319336,
+      "learning_rate": 1.482976759048351e-05,
+      "loss": 0.619,
+      "step": 986
+    },
+    {
+      "epoch": 0.3952,
+      "grad_norm": 2.8806512355804443,
+      "learning_rate": 1.485420101795274e-05,
+      "loss": 0.3831,
+      "step": 988
+    },
+    {
+      "epoch": 0.396,
+      "grad_norm": 6.711476802825928,
+      "learning_rate": 1.4878596591387327e-05,
+      "loss": 0.4079,
+      "step": 990
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 3.724522590637207,
+      "learning_rate": 1.4902954120545686e-05,
+      "loss": 0.7505,
+      "step": 992
+    },
+    {
+      "epoch": 0.3976,
+      "grad_norm": 3.5688815116882324,
+      "learning_rate": 1.4927273415482913e-05,
+      "loss": 0.2601,
+      "step": 994
+    },
+    {
+      "epoch": 0.3984,
+      "grad_norm": 5.003245830535889,
+      "learning_rate": 1.4951554286552261e-05,
+      "loss": 0.3731,
+      "step": 996
+    },
+    {
+      "epoch": 0.3992,
+      "grad_norm": 8.481842041015625,
+      "learning_rate": 1.4975796544406617e-05,
+      "loss": 0.7391,
+      "step": 998
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.910562038421631,
+      "learning_rate": 1.4999999999999992e-05,
+      "loss": 0.4004,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4008,
+      "grad_norm": 12.834502220153809,
+      "learning_rate": 1.502416446458897e-05,
+      "loss": 1.0613,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4016,
+      "grad_norm": 6.666214942932129,
+      "learning_rate": 1.5048289749734206e-05,
+      "loss": 0.594,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4024,
+      "grad_norm": 2.4326231479644775,
+      "learning_rate": 1.5072375667301895e-05,
+      "loss": 0.3435,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 2.314138650894165,
+      "learning_rate": 1.5096422029465178e-05,
+      "loss": 0.35,
+      "step": 1008
+    },
+    {
+      "epoch": 0.404,
+      "grad_norm": 2.2951853275299072,
+      "learning_rate": 1.5120428648705714e-05,
+      "loss": 0.2415,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4048,
+      "grad_norm": 1.675195574760437,
+      "learning_rate": 1.5144395337815064e-05,
+      "loss": 0.2919,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4056,
+      "grad_norm": 9.622923851013184,
+      "learning_rate": 1.5168321909896166e-05,
+      "loss": 0.558,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4064,
+      "grad_norm": 9.771827697753906,
+      "learning_rate": 1.5192208178364808e-05,
+      "loss": 0.667,
+      "step": 1016
+    },
+    {
+      "epoch": 0.4072,
+      "grad_norm": 3.1868371963500977,
+      "learning_rate": 1.521605395695107e-05,
+      "loss": 0.3962,
+      "step": 1018
+    },
+    {
+      "epoch": 0.408,
+      "grad_norm": 3.322626829147339,
+      "learning_rate": 1.5239859059700784e-05,
+      "loss": 0.3846,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4088,
+      "grad_norm": 3.9479050636291504,
+      "learning_rate": 1.526362330097697e-05,
+      "loss": 0.3751,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 2.961951971054077,
+      "learning_rate": 1.5287346495461322e-05,
+      "loss": 0.452,
+      "step": 1024
+    },
+    {
+      "epoch": 0.4104,
+      "grad_norm": 2.779170274734497,
+      "learning_rate": 1.531102845815557e-05,
+      "loss": 0.3655,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4112,
+      "grad_norm": 9.825654983520508,
+      "learning_rate": 1.5334669004383025e-05,
+      "loss": 0.3564,
+      "step": 1028
+    },
+    {
+      "epoch": 0.412,
+      "grad_norm": 6.186550617218018,
+      "learning_rate": 1.5358267949789968e-05,
+      "loss": 0.3361,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4128,
+      "grad_norm": 8.893967628479004,
+      "learning_rate": 1.5381825110347072e-05,
+      "loss": 0.5267,
+      "step": 1032
+    },
+    {
+      "epoch": 0.4136,
+      "grad_norm": 4.747354507446289,
+      "learning_rate": 1.540534030235087e-05,
+      "loss": 0.1701,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4144,
+      "grad_norm": 7.46083927154541,
+      "learning_rate": 1.542881334242517e-05,
+      "loss": 0.488,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4152,
+      "grad_norm": 5.576011657714844,
+      "learning_rate": 1.5452244047522493e-05,
+      "loss": 0.5621,
+      "step": 1038
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 3.370360851287842,
+      "learning_rate": 1.5475632234925495e-05,
+      "loss": 0.6699,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4168,
+      "grad_norm": 5.393393039703369,
+      "learning_rate": 1.5498977722248388e-05,
+      "loss": 0.487,
+      "step": 1042
+    },
+    {
+      "epoch": 0.4176,
+      "grad_norm": 3.2120022773742676,
+      "learning_rate": 1.552228032743839e-05,
+      "loss": 0.6391,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4184,
+      "grad_norm": 3.1691336631774902,
+      "learning_rate": 1.5545539868777075e-05,
+      "loss": 0.4345,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4192,
+      "grad_norm": 2.4558982849121094,
+      "learning_rate": 1.556875616488188e-05,
+      "loss": 0.5859,
+      "step": 1048
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.1330342292785645,
+      "learning_rate": 1.5591929034707468e-05,
+      "loss": 0.3224,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4208,
+      "grad_norm": 2.89819073677063,
+      "learning_rate": 1.5615058297547144e-05,
+      "loss": 0.4887,
+      "step": 1052
+    },
+    {
+      "epoch": 0.4216,
+      "grad_norm": 4.832616806030273,
+      "learning_rate": 1.5638143773034268e-05,
+      "loss": 0.4998,
+      "step": 1054
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 2.52421236038208,
+      "learning_rate": 1.5661185281143663e-05,
+      "loss": 0.367,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4232,
+      "grad_norm": 2.5860087871551514,
+      "learning_rate": 1.5684182642193024e-05,
+      "loss": 0.2247,
+      "step": 1058
+    },
+    {
+      "epoch": 0.424,
+      "grad_norm": 20.680755615234375,
+      "learning_rate": 1.5707135676844312e-05,
+      "loss": 1.0634,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4248,
+      "grad_norm": 2.4126100540161133,
+      "learning_rate": 1.5730044206105146e-05,
+      "loss": 0.2989,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4256,
+      "grad_norm": 6.855231761932373,
+      "learning_rate": 1.5752908051330232e-05,
+      "loss": 0.5926,
+      "step": 1064
+    },
+    {
+      "epoch": 0.4264,
+      "grad_norm": 11.962879180908203,
+      "learning_rate": 1.577572703422268e-05,
+      "loss": 0.5489,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4272,
+      "grad_norm": 5.586971759796143,
+      "learning_rate": 1.579850097683548e-05,
+      "loss": 0.5642,
+      "step": 1068
+    },
+    {
+      "epoch": 0.428,
+      "grad_norm": 9.240392684936523,
+      "learning_rate": 1.582122970157288e-05,
+      "loss": 0.5487,
+      "step": 1070
+    },
+    {
+      "epoch": 0.4288,
+      "grad_norm": 2.5346271991729736,
+      "learning_rate": 1.5843913031191722e-05,
+      "loss": 0.2458,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4296,
+      "grad_norm": 3.1922338008880615,
+      "learning_rate": 1.586655078880281e-05,
+      "loss": 0.3974,
+      "step": 1074
+    },
+    {
+      "epoch": 0.4304,
+      "grad_norm": 7.212031364440918,
+      "learning_rate": 1.5889142797872383e-05,
+      "loss": 0.4685,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4312,
+      "grad_norm": 2.082321882247925,
+      "learning_rate": 1.5911688882223415e-05,
+      "loss": 0.2191,
+      "step": 1078
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 5.929754257202148,
+      "learning_rate": 1.5934188866037007e-05,
+      "loss": 0.5255,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4328,
+      "grad_norm": 4.870511531829834,
+      "learning_rate": 1.5956642573853787e-05,
+      "loss": 0.5902,
+      "step": 1082
+    },
+    {
+      "epoch": 0.4336,
+      "grad_norm": 7.386419296264648,
+      "learning_rate": 1.5979049830575193e-05,
+      "loss": 0.346,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4344,
+      "grad_norm": 3.978145122528076,
+      "learning_rate": 1.6001410461464945e-05,
+      "loss": 0.6408,
+      "step": 1086
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 2.4360148906707764,
+      "learning_rate": 1.6023724292150377e-05,
+      "loss": 0.643,
+      "step": 1088
+    },
+    {
+      "epoch": 0.436,
+      "grad_norm": 2.7286767959594727,
+      "learning_rate": 1.604599114862375e-05,
+      "loss": 0.3048,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4368,
+      "grad_norm": 2.107506036758423,
+      "learning_rate": 1.606821085724362e-05,
+      "loss": 0.2779,
+      "step": 1092
+    },
+    {
+      "epoch": 0.4376,
+      "grad_norm": 8.017492294311523,
+      "learning_rate": 1.6090383244736253e-05,
+      "loss": 0.453,
+      "step": 1094
+    },
+    {
+      "epoch": 0.4384,
+      "grad_norm": 8.654541015625,
+      "learning_rate": 1.6112508138196912e-05,
+      "loss": 0.4822,
+      "step": 1096
+    },
+    {
+      "epoch": 0.4392,
+      "grad_norm": 4.79012393951416,
+      "learning_rate": 1.613458536509124e-05,
+      "loss": 0.4431,
+      "step": 1098
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.138698101043701,
+      "learning_rate": 1.615661475325658e-05,
+      "loss": 0.2924,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4408,
+      "grad_norm": 7.6140522956848145,
+      "learning_rate": 1.6178596130903352e-05,
+      "loss": 0.5564,
+      "step": 1102
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 5.842023849487305,
+      "learning_rate": 1.620052932661632e-05,
+      "loss": 0.2565,
+      "step": 1104
+    },
+    {
+      "epoch": 0.4424,
+      "grad_norm": 3.1433494091033936,
+      "learning_rate": 1.6222414169356056e-05,
+      "loss": 0.5082,
+      "step": 1106
+    },
+    {
+      "epoch": 0.4432,
+      "grad_norm": 2.6837446689605713,
+      "learning_rate": 1.6244250488460146e-05,
+      "loss": 0.4007,
+      "step": 1108
+    },
+    {
+      "epoch": 0.444,
+      "grad_norm": 3.3919453620910645,
+      "learning_rate": 1.6266038113644605e-05,
+      "loss": 0.4995,
+      "step": 1110
+    },
+    {
+      "epoch": 0.4448,
+      "grad_norm": 2.611570358276367,
+      "learning_rate": 1.6287776875005127e-05,
+      "loss": 0.3224,
+      "step": 1112
+    },
+    {
+      "epoch": 0.4456,
+      "grad_norm": 4.856591701507568,
+      "learning_rate": 1.6309466603018497e-05,
+      "loss": 0.4728,
+      "step": 1114
+    },
+    {
+      "epoch": 0.4464,
+      "grad_norm": 13.449359893798828,
+      "learning_rate": 1.6331107128543856e-05,
+      "loss": 0.5741,
+      "step": 1116
+    },
+    {
+      "epoch": 0.4472,
+      "grad_norm": 2.368253707885742,
+      "learning_rate": 1.635269828282404e-05,
+      "loss": 0.561,
+      "step": 1118
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 2.5609967708587646,
+      "learning_rate": 1.6374239897486905e-05,
+      "loss": 0.4176,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4488,
+      "grad_norm": 3.7215967178344727,
+      "learning_rate": 1.6395731804546575e-05,
+      "loss": 0.3077,
+      "step": 1122
+    },
+    {
+      "epoch": 0.4496,
+      "grad_norm": 7.401405334472656,
+      "learning_rate": 1.6417173836404878e-05,
+      "loss": 0.3221,
+      "step": 1124
+    },
+    {
+      "epoch": 0.4504,
+      "grad_norm": 2.0310912132263184,
+      "learning_rate": 1.643856582585253e-05,
+      "loss": 0.3205,
+      "step": 1126
+    },
+    {
+      "epoch": 0.4512,
+      "grad_norm": 9.05074405670166,
+      "learning_rate": 1.6459907606070513e-05,
+      "loss": 0.3382,
+      "step": 1128
+    },
+    {
+      "epoch": 0.452,
+      "grad_norm": 11.440814018249512,
+      "learning_rate": 1.6481199010631312e-05,
+      "loss": 0.4869,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4528,
+      "grad_norm": 2.9795913696289062,
+      "learning_rate": 1.650243987350029e-05,
+      "loss": 0.6705,
+      "step": 1132
+    },
+    {
+      "epoch": 0.4536,
+      "grad_norm": 4.947592258453369,
+      "learning_rate": 1.652363002903693e-05,
+      "loss": 0.2737,
+      "step": 1134
+    },
+    {
+      "epoch": 0.4544,
+      "grad_norm": 6.063323974609375,
+      "learning_rate": 1.6544769311996146e-05,
+      "loss": 0.3493,
+      "step": 1136
+    },
+    {
+      "epoch": 0.4552,
+      "grad_norm": 8.340121269226074,
+      "learning_rate": 1.656585755752956e-05,
+      "loss": 0.6582,
+      "step": 1138
+    },
+    {
+      "epoch": 0.456,
+      "grad_norm": 11.626544952392578,
+      "learning_rate": 1.65868946011868e-05,
+      "loss": 0.6541,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4568,
+      "grad_norm": 4.603865623474121,
+      "learning_rate": 1.660788027891677e-05,
+      "loss": 0.6873,
+      "step": 1142
+    },
+    {
+      "epoch": 0.4576,
+      "grad_norm": 7.3429155349731445,
+      "learning_rate": 1.6628814427068944e-05,
+      "loss": 0.6526,
+      "step": 1144
+    },
+    {
+      "epoch": 0.4584,
+      "grad_norm": 3.518334150314331,
+      "learning_rate": 1.6649696882394625e-05,
+      "loss": 0.3181,
+      "step": 1146
+    },
+    {
+      "epoch": 0.4592,
+      "grad_norm": 5.010440349578857,
+      "learning_rate": 1.667052748204825e-05,
+      "loss": 0.4269,
+      "step": 1148
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.749119758605957,
+      "learning_rate": 1.6691306063588583e-05,
+      "loss": 0.3674,
+      "step": 1150
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 7.220366954803467,
+      "learning_rate": 1.6712032464980094e-05,
+      "loss": 0.2894,
+      "step": 1152
+    },
+    {
+      "epoch": 0.4616,
+      "grad_norm": 2.52457857131958,
+      "learning_rate": 1.6732706524594138e-05,
+      "loss": 0.472,
+      "step": 1154
+    },
+    {
+      "epoch": 0.4624,
+      "grad_norm": 1.485608458518982,
+      "learning_rate": 1.6753328081210244e-05,
+      "loss": 0.1831,
+      "step": 1156
+    },
+    {
+      "epoch": 0.4632,
+      "grad_norm": 4.405038356781006,
+      "learning_rate": 1.6773896974017373e-05,
+      "loss": 0.5844,
+      "step": 1158
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 8.147605895996094,
+      "learning_rate": 1.679441304261516e-05,
+      "loss": 0.48,
+      "step": 1160
+    },
+    {
+      "epoch": 0.4648,
+      "grad_norm": 4.937817573547363,
+      "learning_rate": 1.681487612701519e-05,
+      "loss": 0.5161,
+      "step": 1162
+    },
+    {
+      "epoch": 0.4656,
+      "grad_norm": 5.370089530944824,
+      "learning_rate": 1.683528606764222e-05,
+      "loss": 0.5149,
+      "step": 1164
+    },
+    {
+      "epoch": 0.4664,
+      "grad_norm": 8.323260307312012,
+      "learning_rate": 1.6855642705335428e-05,
+      "loss": 0.8392,
+      "step": 1166
+    },
+    {
+      "epoch": 0.4672,
+      "grad_norm": 2.0059239864349365,
+      "learning_rate": 1.687594588134968e-05,
+      "loss": 0.3356,
+      "step": 1168
+    },
+    {
+      "epoch": 0.468,
+      "grad_norm": 4.255488872528076,
+      "learning_rate": 1.68961954373567e-05,
+      "loss": 0.4183,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4688,
+      "grad_norm": 7.48425817489624,
+      "learning_rate": 1.6916391215446403e-05,
+      "loss": 0.5642,
+      "step": 1172
+    },
+    {
+      "epoch": 0.4696,
+      "grad_norm": 6.062568664550781,
+      "learning_rate": 1.693653305812805e-05,
+      "loss": 0.4021,
+      "step": 1174
+    },
+    {
+      "epoch": 0.4704,
+      "grad_norm": 6.629954814910889,
+      "learning_rate": 1.6956620808331505e-05,
+      "loss": 2.9714,
+      "step": 1176
+    },
+    {
+      "epoch": 0.4712,
+      "grad_norm": 4.56139612197876,
+      "learning_rate": 1.697665430940846e-05,
+      "loss": 0.4547,
+      "step": 1178
+    },
+    {
+      "epoch": 0.472,
+      "grad_norm": 6.164916038513184,
+      "learning_rate": 1.699663340513365e-05,
+      "loss": 0.7382,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4728,
+      "grad_norm": 4.28443717956543,
+      "learning_rate": 1.7016557939706068e-05,
+      "loss": 0.5405,
+      "step": 1182
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 9.564210891723633,
+      "learning_rate": 1.7036427757750198e-05,
+      "loss": 0.641,
+      "step": 1184
+    },
+    {
+      "epoch": 0.4744,
+      "grad_norm": 4.702108860015869,
+      "learning_rate": 1.7056242704317212e-05,
+      "loss": 0.4587,
+      "step": 1186
+    },
+    {
+      "epoch": 0.4752,
+      "grad_norm": 6.514930248260498,
+      "learning_rate": 1.7076002624886156e-05,
+      "loss": 0.4304,
+      "step": 1188
+    },
+    {
+      "epoch": 0.476,
+      "grad_norm": 7.816126823425293,
+      "learning_rate": 1.709570736536521e-05,
+      "loss": 0.5538,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4768,
+      "grad_norm": 4.538508415222168,
+      "learning_rate": 1.7115356772092844e-05,
+      "loss": 0.3398,
+      "step": 1192
+    },
+    {
+      "epoch": 0.4776,
+      "grad_norm": 7.519460678100586,
+      "learning_rate": 1.7134950691839063e-05,
+      "loss": 0.5039,
+      "step": 1194
+    },
+    {
+      "epoch": 0.4784,
+      "grad_norm": 3.8620688915252686,
+      "learning_rate": 1.7154488971806518e-05,
+      "loss": 0.4347,
+      "step": 1196
+    },
+    {
+      "epoch": 0.4792,
+      "grad_norm": 2.2250876426696777,
+      "learning_rate": 1.7173971459631783e-05,
+      "loss": 0.4299,
+      "step": 1198
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6023005247116089,
+      "learning_rate": 1.7193398003386507e-05,
+      "loss": 0.3742,
+      "step": 1200
+    },
+    {
+      "epoch": 0.4808,
+      "grad_norm": 8.703466415405273,
+      "learning_rate": 1.7212768451578602e-05,
+      "loss": 0.5665,
+      "step": 1202
+    },
+    {
+      "epoch": 0.4816,
+      "grad_norm": 2.355034351348877,
+      "learning_rate": 1.7232082653153416e-05,
+      "loss": 0.3028,
+      "step": 1204
+    },
+    {
+      "epoch": 0.4824,
+      "grad_norm": 2.3913300037384033,
+      "learning_rate": 1.7251340457494937e-05,
+      "loss": 0.2936,
+      "step": 1206
+    },
+    {
+      "epoch": 0.4832,
+      "grad_norm": 2.8259389400482178,
+      "learning_rate": 1.7270541714426923e-05,
+      "loss": 0.2899,
+      "step": 1208
+    },
+    {
+      "epoch": 0.484,
+      "grad_norm": 11.542742729187012,
+      "learning_rate": 1.7289686274214106e-05,
+      "loss": 1.6536,
+      "step": 1210
+    },
+    {
+      "epoch": 0.4848,
+      "grad_norm": 8.168238639831543,
+      "learning_rate": 1.7308773987563393e-05,
+      "loss": 0.7052,
+      "step": 1212
+    },
+    {
+      "epoch": 0.4856,
+      "grad_norm": 6.536013603210449,
+      "learning_rate": 1.732780470562496e-05,
+      "loss": 0.7356,
+      "step": 1214
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 6.662909984588623,
+      "learning_rate": 1.7346778279993413e-05,
+      "loss": 0.7118,
+      "step": 1216
+    },
+    {
+      "epoch": 0.4872,
+      "grad_norm": 5.744428634643555,
+      "learning_rate": 1.736569456270903e-05,
+      "loss": 0.4607,
+      "step": 1218
+    },
+    {
+      "epoch": 0.488,
+      "grad_norm": 3.2977943420410156,
+      "learning_rate": 1.7384553406258836e-05,
+      "loss": 0.6452,
+      "step": 1220
+    },
+    {
+      "epoch": 0.4888,
+      "grad_norm": 7.264194011688232,
+      "learning_rate": 1.740335466357778e-05,
+      "loss": 0.4696,
+      "step": 1222
+    },
+    {
+      "epoch": 0.4896,
+      "grad_norm": 9.938541412353516,
+      "learning_rate": 1.7422098188049888e-05,
+      "loss": 0.6597,
+      "step": 1224
+    },
+    {
+      "epoch": 0.4904,
+      "grad_norm": 4.108887195587158,
+      "learning_rate": 1.7440783833509373e-05,
+      "loss": 0.2282,
+      "step": 1226
+    },
+    {
+      "epoch": 0.4912,
+      "grad_norm": 3.81720232963562,
+      "learning_rate": 1.7459411454241816e-05,
+      "loss": 0.3304,
+      "step": 1228
+    },
+    {
+      "epoch": 0.492,
+      "grad_norm": 8.632997512817383,
+      "learning_rate": 1.747798090498531e-05,
+      "loss": 0.4577,
+      "step": 1230
+    },
+    {
+      "epoch": 0.4928,
+      "grad_norm": 4.479432106018066,
+      "learning_rate": 1.749649204093154e-05,
+      "loss": 0.5919,
+      "step": 1232
+    },
+    {
+      "epoch": 0.4936,
+      "grad_norm": 7.987699031829834,
+      "learning_rate": 1.7514944717726962e-05,
+      "loss": 1.0532,
+      "step": 1234
+    },
+    {
+      "epoch": 0.4944,
+      "grad_norm": 3.5861763954162598,
+      "learning_rate": 1.753333879147387e-05,
+      "loss": 0.3997,
+      "step": 1236
+    },
+    {
+      "epoch": 0.4952,
+      "grad_norm": 8.19918155670166,
+      "learning_rate": 1.755167411873159e-05,
+      "loss": 0.5248,
+      "step": 1238
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 0.6573111414909363,
+      "learning_rate": 1.7569950556517563e-05,
+      "loss": 0.1645,
+      "step": 1240
+    },
+    {
+      "epoch": 0.4968,
+      "grad_norm": 3.7899794578552246,
+      "learning_rate": 1.758816796230845e-05,
+      "loss": 0.6108,
+      "step": 1242
+    },
+    {
+      "epoch": 0.4976,
+      "grad_norm": 6.53424596786499,
+      "learning_rate": 1.7606326194041278e-05,
+      "loss": 0.5721,
+      "step": 1244
+    },
+    {
+      "epoch": 0.4984,
+      "grad_norm": 6.82229471206665,
+      "learning_rate": 1.762442511011447e-05,
+      "loss": 0.6002,
+      "step": 1246
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 10.31006908416748,
+      "learning_rate": 1.7642464569389083e-05,
+      "loss": 0.7883,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.102840900421143,
+      "learning_rate": 1.766044443118977e-05,
+      "loss": 0.3058,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5008,
+      "grad_norm": 13.063787460327148,
+      "learning_rate": 1.767836455530598e-05,
+      "loss": 0.7589,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5016,
+      "grad_norm": 12.737040519714355,
+      "learning_rate": 1.7696224801992947e-05,
+      "loss": 0.879,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5024,
+      "grad_norm": 5.0701751708984375,
+      "learning_rate": 1.77140250319729e-05,
+      "loss": 0.6462,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5032,
+      "grad_norm": 6.491806507110596,
+      "learning_rate": 1.7731765106436073e-05,
+      "loss": 0.4422,
+      "step": 1258
+    },
+    {
+      "epoch": 0.504,
+      "grad_norm": 7.690762042999268,
+      "learning_rate": 1.7749444887041793e-05,
+      "loss": 0.6886,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5048,
+      "grad_norm": 2.4480361938476562,
+      "learning_rate": 1.776706423591959e-05,
+      "loss": 0.3819,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5056,
+      "grad_norm": 6.522246837615967,
+      "learning_rate": 1.778462301567023e-05,
+      "loss": 1.1178,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5064,
+      "grad_norm": 7.040022373199463,
+      "learning_rate": 1.7802121089366832e-05,
+      "loss": 0.6745,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5072,
+      "grad_norm": 5.24992036819458,
+      "learning_rate": 1.7819558320555895e-05,
+      "loss": 0.4028,
+      "step": 1268
+    },
+    {
+      "epoch": 0.508,
+      "grad_norm": 2.2948031425476074,
+      "learning_rate": 1.7836934573258392e-05,
+      "loss": 0.3027,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5088,
+      "grad_norm": 2.4775025844573975,
+      "learning_rate": 1.785424971197082e-05,
+      "loss": 0.3507,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5096,
+      "grad_norm": 3.7970142364501953,
+      "learning_rate": 1.7871503601666233e-05,
+      "loss": 0.591,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5104,
+      "grad_norm": 8.241109848022461,
+      "learning_rate": 1.7888696107795343e-05,
+      "loss": 0.6565,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5112,
+      "grad_norm": 8.135817527770996,
+      "learning_rate": 1.790582709628753e-05,
+      "loss": 0.6159,
+      "step": 1278
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 4.600522518157959,
+      "learning_rate": 1.7922896433551903e-05,
+      "loss": 0.3696,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5128,
+      "grad_norm": 4.542240142822266,
+      "learning_rate": 1.793990398647835e-05,
+      "loss": 0.4028,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5136,
+      "grad_norm": 2.846844434738159,
+      "learning_rate": 1.795684962243855e-05,
+      "loss": 0.3901,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5144,
+      "grad_norm": 2.5168137550354004,
+      "learning_rate": 1.7973733209287032e-05,
+      "loss": 0.3155,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5152,
+      "grad_norm": 7.11170768737793,
+      "learning_rate": 1.7990554615362193e-05,
+      "loss": 0.4178,
+      "step": 1288
+    },
+    {
+      "epoch": 0.516,
+      "grad_norm": 2.3307881355285645,
+      "learning_rate": 1.800731370948734e-05,
+      "loss": 0.4979,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5168,
+      "grad_norm": 14.980961799621582,
+      "learning_rate": 1.802401036097167e-05,
+      "loss": 0.6443,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5176,
+      "grad_norm": 2.6857290267944336,
+      "learning_rate": 1.804064443961135e-05,
+      "loss": 0.4083,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 5.365575790405273,
+      "learning_rate": 1.8057215815690494e-05,
+      "loss": 0.3637,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5192,
+      "grad_norm": 2.6065447330474854,
+      "learning_rate": 1.8073724359982184e-05,
+      "loss": 0.1999,
+      "step": 1298
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.7862114906311035,
+      "learning_rate": 1.809016994374947e-05,
+      "loss": 0.2711,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5208,
+      "grad_norm": 3.6769278049468994,
+      "learning_rate": 1.81065524387464e-05,
+      "loss": 0.3093,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5216,
+      "grad_norm": 20.27410888671875,
+      "learning_rate": 1.8122871717218968e-05,
+      "loss": 0.5956,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5224,
+      "grad_norm": 15.000691413879395,
+      "learning_rate": 1.8139127651906176e-05,
+      "loss": 0.7266,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5232,
+      "grad_norm": 8.50636100769043,
+      "learning_rate": 1.8155320116040976e-05,
+      "loss": 1.346,
+      "step": 1308
+    },
+    {
+      "epoch": 0.524,
+      "grad_norm": 2.9468233585357666,
+      "learning_rate": 1.817144898335129e-05,
+      "loss": 0.2572,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 12.458033561706543,
+      "learning_rate": 1.818751412806095e-05,
+      "loss": 0.5091,
+      "step": 1312
+    },
+    {
+      "epoch": 0.5256,
+      "grad_norm": 4.08944034576416,
+      "learning_rate": 1.8203515424890738e-05,
+      "loss": 0.3258,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5264,
+      "grad_norm": 6.3507819175720215,
+      "learning_rate": 1.8219452749059322e-05,
+      "loss": 0.6816,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5272,
+      "grad_norm": 3.709740161895752,
+      "learning_rate": 1.8235325976284276e-05,
+      "loss": 0.3486,
+      "step": 1318
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 2.184805154800415,
+      "learning_rate": 1.8251134982782952e-05,
+      "loss": 0.2956,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5288,
+      "grad_norm": 3.21848726272583,
+      "learning_rate": 1.826687964527355e-05,
+      "loss": 0.4955,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5296,
+      "grad_norm": 11.279875755310059,
+      "learning_rate": 1.828255984097604e-05,
+      "loss": 0.5082,
+      "step": 1324
+    },
+    {
+      "epoch": 0.5304,
+      "grad_norm": 4.378537178039551,
+      "learning_rate": 1.8298175447613093e-05,
+      "loss": 0.3311,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5312,
+      "grad_norm": 5.776856422424316,
+      "learning_rate": 1.8313726343411092e-05,
+      "loss": 0.3442,
+      "step": 1328
+    },
+    {
+      "epoch": 0.532,
+      "grad_norm": 2.2048215866088867,
+      "learning_rate": 1.8329212407101e-05,
+      "loss": 0.3682,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5328,
+      "grad_norm": 2.4243111610412598,
+      "learning_rate": 1.8344633517919394e-05,
+      "loss": 0.4139,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5336,
+      "grad_norm": 8.74450397491455,
+      "learning_rate": 1.8359989555609344e-05,
+      "loss": 0.4095,
+      "step": 1334
+    },
+    {
+      "epoch": 0.5344,
+      "grad_norm": 8.038501739501953,
+      "learning_rate": 1.8375280400421407e-05,
+      "loss": 0.7819,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5352,
+      "grad_norm": 7.184566974639893,
+      "learning_rate": 1.8390505933114503e-05,
+      "loss": 0.3759,
+      "step": 1338
+    },
+    {
+      "epoch": 0.536,
+      "grad_norm": 7.550333499908447,
+      "learning_rate": 1.8405666034956842e-05,
+      "loss": 1.3611,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5368,
+      "grad_norm": 9.062027931213379,
+      "learning_rate": 1.842076058772692e-05,
+      "loss": 0.6528,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 4.2818708419799805,
+      "learning_rate": 1.8435789473714384e-05,
+      "loss": 0.9378,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5384,
+      "grad_norm": 7.2454938888549805,
+      "learning_rate": 1.8450752575720964e-05,
+      "loss": 0.2129,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5392,
+      "grad_norm": 2.8972926139831543,
+      "learning_rate": 1.8465649777061384e-05,
+      "loss": 0.4785,
+      "step": 1348
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.637453556060791,
+      "learning_rate": 1.8480480961564266e-05,
+      "loss": 0.7323,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5408,
+      "grad_norm": 3.591602325439453,
+      "learning_rate": 1.8495246013573047e-05,
+      "loss": 0.4019,
+      "step": 1352
+    },
+    {
+      "epoch": 0.5416,
+      "grad_norm": 8.564355850219727,
+      "learning_rate": 1.850994481794691e-05,
+      "loss": 0.5674,
+      "step": 1354
+    },
+    {
+      "epoch": 0.5424,
+      "grad_norm": 11.854432106018066,
+      "learning_rate": 1.8524577260061628e-05,
+      "loss": 0.5302,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5432,
+      "grad_norm": 4.832673072814941,
+      "learning_rate": 1.8539143225810453e-05,
+      "loss": 0.397,
+      "step": 1358
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 6.154983043670654,
+      "learning_rate": 1.8553642601605066e-05,
+      "loss": 0.8572,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5448,
+      "grad_norm": 4.938594341278076,
+      "learning_rate": 1.856807527437643e-05,
+      "loss": 0.6868,
+      "step": 1362
+    },
+    {
+      "epoch": 0.5456,
+      "grad_norm": 5.232631206512451,
+      "learning_rate": 1.8582441131575658e-05,
+      "loss": 0.5783,
+      "step": 1364
+    },
+    {
+      "epoch": 0.5464,
+      "grad_norm": 5.804172039031982,
+      "learning_rate": 1.859674006117491e-05,
+      "loss": 0.3358,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5472,
+      "grad_norm": 2.756375789642334,
+      "learning_rate": 1.8610971951668268e-05,
+      "loss": 0.396,
+      "step": 1368
+    },
+    {
+      "epoch": 0.548,
+      "grad_norm": 1.8918120861053467,
+      "learning_rate": 1.862513669207257e-05,
+      "loss": 0.2675,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5488,
+      "grad_norm": 4.279423236846924,
+      "learning_rate": 1.8639234171928348e-05,
+      "loss": 0.2105,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5496,
+      "grad_norm": 14.454187393188477,
+      "learning_rate": 1.8653264281300612e-05,
+      "loss": 0.7139,
+      "step": 1374
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 8.212747573852539,
+      "learning_rate": 1.8667226910779767e-05,
+      "loss": 0.4805,
+      "step": 1376
+    },
+    {
+      "epoch": 0.5512,
+      "grad_norm": 1.839599847793579,
+      "learning_rate": 1.8681121951482393e-05,
+      "loss": 0.3243,
+      "step": 1378
+    },
+    {
+      "epoch": 0.552,
+      "grad_norm": 8.92973518371582,
+      "learning_rate": 1.869494929505219e-05,
+      "loss": 1.4886,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5528,
+      "grad_norm": 9.111851692199707,
+      "learning_rate": 1.870870883366075e-05,
+      "loss": 1.0078,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5536,
+      "grad_norm": 2.5603065490722656,
+      "learning_rate": 1.8722400460008434e-05,
+      "loss": 0.2356,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5544,
+      "grad_norm": 2.080918312072754,
+      "learning_rate": 1.8736024067325195e-05,
+      "loss": 0.3798,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5552,
+      "grad_norm": 2.699968099594116,
+      "learning_rate": 1.8749579549371373e-05,
+      "loss": 0.4191,
+      "step": 1388
+    },
+    {
+      "epoch": 0.556,
+      "grad_norm": 5.318063259124756,
+      "learning_rate": 1.876306680043863e-05,
+      "loss": 0.3578,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 3.0889317989349365,
+      "learning_rate": 1.8776485715350665e-05,
+      "loss": 0.3083,
+      "step": 1392
+    },
+    {
+      "epoch": 0.5576,
+      "grad_norm": 9.75284194946289,
+      "learning_rate": 1.878983618946409e-05,
+      "loss": 0.5724,
+      "step": 1394
+    },
+    {
+      "epoch": 0.5584,
+      "grad_norm": 2.678609848022461,
+      "learning_rate": 1.8803118118669203e-05,
+      "loss": 0.287,
+      "step": 1396
+    },
+    {
+      "epoch": 0.5592,
+      "grad_norm": 1.7660149335861206,
+      "learning_rate": 1.881633139939087e-05,
+      "loss": 0.2172,
+      "step": 1398
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.460742950439453,
+      "learning_rate": 1.882947592858927e-05,
+      "loss": 0.3789,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5608,
+      "grad_norm": 2.597334146499634,
+      "learning_rate": 1.884255160376072e-05,
+      "loss": 0.5009,
+      "step": 1402
+    },
+    {
+      "epoch": 0.5616,
+      "grad_norm": 2.7347497940063477,
+      "learning_rate": 1.885555832293849e-05,
+      "loss": 0.3827,
+      "step": 1404
+    },
+    {
+      "epoch": 0.5624,
+      "grad_norm": 11.216575622558594,
+      "learning_rate": 1.886849598469356e-05,
+      "loss": 1.3072,
+      "step": 1406
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 1.3659820556640625,
+      "learning_rate": 1.888136448813544e-05,
+      "loss": 0.2031,
+      "step": 1408
+    },
+    {
+      "epoch": 0.564,
+      "grad_norm": 6.435552597045898,
+      "learning_rate": 1.8894163732912972e-05,
+      "loss": 0.3993,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5648,
+      "grad_norm": 6.237542629241943,
+      "learning_rate": 1.890689361921506e-05,
+      "loss": 0.3356,
+      "step": 1412
+    },
+    {
+      "epoch": 0.5656,
+      "grad_norm": 7.920870780944824,
+      "learning_rate": 1.891955404777151e-05,
+      "loss": 0.4875,
+      "step": 1414
+    },
+    {
+      "epoch": 0.5664,
+      "grad_norm": 8.32780933380127,
+      "learning_rate": 1.893214491985374e-05,
+      "loss": 0.4395,
+      "step": 1416
+    },
+    {
+      "epoch": 0.5672,
+      "grad_norm": 5.75148868560791,
+      "learning_rate": 1.89446661372756e-05,
+      "loss": 0.384,
+      "step": 1418
+    },
+    {
+      "epoch": 0.568,
+      "grad_norm": 6.593601226806641,
+      "learning_rate": 1.895711760239413e-05,
+      "loss": 0.4153,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5688,
+      "grad_norm": 5.3204474449157715,
+      "learning_rate": 1.89694992181103e-05,
+      "loss": 0.5154,
+      "step": 1422
+    },
+    {
+      "epoch": 0.5696,
+      "grad_norm": 7.025198459625244,
+      "learning_rate": 1.8981810887869784e-05,
+      "loss": 0.3926,
+      "step": 1424
+    },
+    {
+      "epoch": 0.5704,
+      "grad_norm": 3.4113192558288574,
+      "learning_rate": 1.8994052515663708e-05,
+      "loss": 0.4192,
+      "step": 1426
+    },
+    {
+      "epoch": 0.5712,
+      "grad_norm": 2.7692418098449707,
+      "learning_rate": 1.90062240060294e-05,
+      "loss": 0.4204,
+      "step": 1428
+    },
+    {
+      "epoch": 0.572,
+      "grad_norm": 7.360057830810547,
+      "learning_rate": 1.9018325264051136e-05,
+      "loss": 0.3961,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5728,
+      "grad_norm": 13.831832885742188,
+      "learning_rate": 1.9030356195360868e-05,
+      "loss": 0.7024,
+      "step": 1432
+    },
+    {
+      "epoch": 0.5736,
+      "grad_norm": 3.6427500247955322,
+      "learning_rate": 1.904231670613899e-05,
+      "loss": 0.2966,
+      "step": 1434
+    },
+    {
+      "epoch": 0.5744,
+      "grad_norm": 6.612185478210449,
+      "learning_rate": 1.905420670311502e-05,
+      "loss": 0.6514,
+      "step": 1436
+    },
+    {
+      "epoch": 0.5752,
+      "grad_norm": 7.35181188583374,
+      "learning_rate": 1.906602609356838e-05,
+      "loss": 0.6473,
+      "step": 1438
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 7.588043689727783,
+      "learning_rate": 1.9077774785329078e-05,
+      "loss": 0.5851,
+      "step": 1440
+    },
+    {
+      "epoch": 0.5768,
+      "grad_norm": 5.37501859664917,
+      "learning_rate": 1.9089452686778487e-05,
+      "loss": 0.4444,
+      "step": 1442
+    },
+    {
+      "epoch": 0.5776,
+      "grad_norm": 2.9378836154937744,
+      "learning_rate": 1.9101059706849957e-05,
+      "loss": 0.5089,
+      "step": 1444
+    },
+    {
+      "epoch": 0.5784,
+      "grad_norm": 7.634711742401123,
+      "learning_rate": 1.911259575502962e-05,
+      "loss": 0.5481,
+      "step": 1446
+    },
+    {
+      "epoch": 0.5792,
+      "grad_norm": 3.4533517360687256,
+      "learning_rate": 1.912406074135706e-05,
+      "loss": 0.3859,
+      "step": 1448
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.150989055633545,
+      "learning_rate": 1.9135454576426006e-05,
+      "loss": 0.4401,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5808,
+      "grad_norm": 15.217089653015137,
+      "learning_rate": 1.9146777171385053e-05,
+      "loss": 0.7959,
+      "step": 1452
+    },
+    {
+      "epoch": 0.5816,
+      "grad_norm": 7.8969197273254395,
+      "learning_rate": 1.9158028437938316e-05,
+      "loss": 0.3802,
+      "step": 1454
+    },
+    {
+      "epoch": 0.5824,
+      "grad_norm": 5.025742530822754,
+      "learning_rate": 1.9169208288346168e-05,
+      "loss": 0.3686,
+      "step": 1456
+    },
+    {
+      "epoch": 0.5832,
+      "grad_norm": 12.718505859375,
+      "learning_rate": 1.9180316635425876e-05,
+      "loss": 0.3675,
+      "step": 1458
+    },
+    {
+      "epoch": 0.584,
+      "grad_norm": 5.548830032348633,
+      "learning_rate": 1.9191353392552346e-05,
+      "loss": 0.8477,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5848,
+      "grad_norm": 5.419297695159912,
+      "learning_rate": 1.9202318473658703e-05,
+      "loss": 0.521,
+      "step": 1462
+    },
+    {
+      "epoch": 0.5856,
+      "grad_norm": 2.6505277156829834,
+      "learning_rate": 1.9213211793237052e-05,
+      "loss": 0.4772,
+      "step": 1464
+    },
+    {
+      "epoch": 0.5864,
+      "grad_norm": 5.732573509216309,
+      "learning_rate": 1.92240332663391e-05,
+      "loss": 0.619,
+      "step": 1466
+    },
+    {
+      "epoch": 0.5872,
+      "grad_norm": 7.656153678894043,
+      "learning_rate": 1.923478280857682e-05,
+      "loss": 0.6467,
+      "step": 1468
+    },
+    {
+      "epoch": 0.588,
+      "grad_norm": 12.494869232177734,
+      "learning_rate": 1.924546033612313e-05,
+      "loss": 1.3374,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 2.8723483085632324,
+      "learning_rate": 1.9256065765712524e-05,
+      "loss": 0.4653,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5896,
+      "grad_norm": 2.027968168258667,
+      "learning_rate": 1.9266599014641724e-05,
+      "loss": 0.4778,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5904,
+      "grad_norm": 3.951679229736328,
+      "learning_rate": 1.927706000077034e-05,
+      "loss": 0.2609,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5912,
+      "grad_norm": 9.217889785766602,
+      "learning_rate": 1.9287448642521507e-05,
+      "loss": 0.893,
+      "step": 1478
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 11.40930461883545,
+      "learning_rate": 1.9297764858882516e-05,
+      "loss": 0.7237,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5928,
+      "grad_norm": 2.650052070617676,
+      "learning_rate": 1.9308008569405424e-05,
+      "loss": 0.4074,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5936,
+      "grad_norm": 4.919514179229736,
+      "learning_rate": 1.9318179694207722e-05,
+      "loss": 1.0755,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5944,
+      "grad_norm": 5.501877784729004,
+      "learning_rate": 1.9328278153972943e-05,
+      "loss": 0.36,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 3.0425186157226562,
+      "learning_rate": 1.9338303869951266e-05,
+      "loss": 0.3789,
+      "step": 1488
+    },
+    {
+      "epoch": 0.596,
+      "grad_norm": 0.8871821165084839,
+      "learning_rate": 1.934825676396015e-05,
+      "loss": 0.131,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5968,
+      "grad_norm": 6.4956464767456055,
+      "learning_rate": 1.935813675838491e-05,
+      "loss": 0.5635,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5976,
+      "grad_norm": 8.6665620803833,
+      "learning_rate": 1.9367943776179375e-05,
+      "loss": 0.5478,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5984,
+      "grad_norm": 13.301843643188477,
+      "learning_rate": 1.9377677740866457e-05,
+      "loss": 1.3023,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5992,
+      "grad_norm": 8.54133129119873,
+      "learning_rate": 1.9387338576538743e-05,
+      "loss": 0.8299,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.637697219848633,
+      "learning_rate": 1.9396926207859085e-05,
+      "loss": 0.592,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6008,
+      "grad_norm": 6.597886562347412,
+      "learning_rate": 1.9406440560061214e-05,
+      "loss": 0.3459,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 9.389493942260742,
+      "learning_rate": 1.9415881558950302e-05,
+      "loss": 0.8848,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6024,
+      "grad_norm": 4.483582973480225,
+      "learning_rate": 1.942524913090354e-05,
+      "loss": 0.2175,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6032,
+      "grad_norm": 2.89554762840271,
+      "learning_rate": 1.9434543202870723e-05,
+      "loss": 2.5025,
+      "step": 1508
+    },
+    {
+      "epoch": 0.604,
+      "grad_norm": 3.055518388748169,
+      "learning_rate": 1.9443763702374815e-05,
+      "loss": 0.5473,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6048,
+      "grad_norm": 1.5298120975494385,
+      "learning_rate": 1.9452910557512494e-05,
+      "loss": 0.3529,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6056,
+      "grad_norm": 2.2401487827301025,
+      "learning_rate": 1.9461983696954756e-05,
+      "loss": 0.3263,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6064,
+      "grad_norm": 6.360352993011475,
+      "learning_rate": 1.947098304994744e-05,
+      "loss": 0.5558,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6072,
+      "grad_norm": 2.6135237216949463,
+      "learning_rate": 1.9479908546311787e-05,
+      "loss": 0.4786,
+      "step": 1518
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 3.436858654022217,
+      "learning_rate": 1.9488760116444966e-05,
+      "loss": 0.4016,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6088,
+      "grad_norm": 5.723163604736328,
+      "learning_rate": 1.949753769132067e-05,
+      "loss": 0.5532,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6096,
+      "grad_norm": 5.951367378234863,
+      "learning_rate": 1.95062412024896e-05,
+      "loss": 0.6564,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6104,
+      "grad_norm": 6.127574920654297,
+      "learning_rate": 1.951487058208003e-05,
+      "loss": 0.5847,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6112,
+      "grad_norm": 0.4084920287132263,
+      "learning_rate": 1.952342576279833e-05,
+      "loss": 0.1559,
+      "step": 1528
+    },
+    {
+      "epoch": 0.612,
+      "grad_norm": 4.772994518280029,
+      "learning_rate": 1.953190667792947e-05,
+      "loss": 0.3836,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6128,
+      "grad_norm": 3.275306463241577,
+      "learning_rate": 1.9540313261337578e-05,
+      "loss": 0.4954,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6136,
+      "grad_norm": 17.405899047851562,
+      "learning_rate": 1.954864544746643e-05,
+      "loss": 0.9868,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 3.4976556301116943,
+      "learning_rate": 1.955690317133996e-05,
+      "loss": 0.5427,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6152,
+      "grad_norm": 4.741286277770996,
+      "learning_rate": 1.956508636856278e-05,
+      "loss": 0.5814,
+      "step": 1538
+    },
+    {
+      "epoch": 0.616,
+      "grad_norm": 9.094281196594238,
+      "learning_rate": 1.9573194975320672e-05,
+      "loss": 0.7027,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6168,
+      "grad_norm": 10.391242027282715,
+      "learning_rate": 1.95812289283811e-05,
+      "loss": 0.6991,
+      "step": 1542
+    },
+    {
+      "epoch": 0.6176,
+      "grad_norm": 4.268300533294678,
+      "learning_rate": 1.958918816509367e-05,
+      "loss": 0.5425,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6184,
+      "grad_norm": 2.675271511077881,
+      "learning_rate": 1.9597072623390668e-05,
+      "loss": 0.427,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6192,
+      "grad_norm": 4.287508010864258,
+      "learning_rate": 1.9604882241787496e-05,
+      "loss": 0.4181,
+      "step": 1548
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.404676914215088,
+      "learning_rate": 1.9612616959383187e-05,
+      "loss": 0.4235,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6208,
+      "grad_norm": 13.730271339416504,
+      "learning_rate": 1.9620276715860856e-05,
+      "loss": 0.8335,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6216,
+      "grad_norm": 4.315786361694336,
+      "learning_rate": 1.9627861451488187e-05,
+      "loss": 0.5621,
+      "step": 1554
+    },
+    {
+      "epoch": 0.6224,
+      "grad_norm": 7.495677947998047,
+      "learning_rate": 1.963537110711789e-05,
+      "loss": 0.4497,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6232,
+      "grad_norm": 3.7830357551574707,
+      "learning_rate": 1.964280562418815e-05,
+      "loss": 0.2931,
+      "step": 1558
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 11.285057067871094,
+      "learning_rate": 1.9650164944723116e-05,
+      "loss": 0.7102,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6248,
+      "grad_norm": 4.690367221832275,
+      "learning_rate": 1.9657449011333328e-05,
+      "loss": 0.3579,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6256,
+      "grad_norm": 5.907130718231201,
+      "learning_rate": 1.9664657767216176e-05,
+      "loss": 0.5712,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6264,
+      "grad_norm": 4.572456359863281,
+      "learning_rate": 1.967179115615633e-05,
+      "loss": 0.4964,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 6.378812313079834,
+      "learning_rate": 1.967884912252619e-05,
+      "loss": 0.4332,
+      "step": 1568
+    },
+    {
+      "epoch": 0.628,
+      "grad_norm": 4.807539463043213,
+      "learning_rate": 1.9685831611286312e-05,
+      "loss": 0.3473,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6288,
+      "grad_norm": 7.2378339767456055,
+      "learning_rate": 1.969273856798585e-05,
+      "loss": 0.4192,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6296,
+      "grad_norm": 8.278740882873535,
+      "learning_rate": 1.9699569938762972e-05,
+      "loss": 0.4816,
+      "step": 1574
+    },
+    {
+      "epoch": 0.6304,
+      "grad_norm": 8.5382719039917,
+      "learning_rate": 1.9706325670345276e-05,
+      "loss": 0.6862,
+      "step": 1576
+    },
+    {
+      "epoch": 0.6312,
+      "grad_norm": 3.933340549468994,
+      "learning_rate": 1.9713005710050203e-05,
+      "loss": 0.4656,
+      "step": 1578
+    },
+    {
+      "epoch": 0.632,
+      "grad_norm": 7.442902565002441,
+      "learning_rate": 1.9719610005785466e-05,
+      "loss": 0.4644,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6328,
+      "grad_norm": 10.707149505615234,
+      "learning_rate": 1.9726138506049434e-05,
+      "loss": 0.6171,
+      "step": 1582
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 2.4153261184692383,
+      "learning_rate": 1.9732591159931564e-05,
+      "loss": 0.4276,
+      "step": 1584
+    },
+    {
+      "epoch": 0.6344,
+      "grad_norm": 2.093482732772827,
+      "learning_rate": 1.9738967917112752e-05,
+      "loss": 0.3721,
+      "step": 1586
+    },
+    {
+      "epoch": 0.6352,
+      "grad_norm": 5.4622344970703125,
+      "learning_rate": 1.974526872786577e-05,
+      "loss": 0.4802,
+      "step": 1588
+    },
+    {
+      "epoch": 0.636,
+      "grad_norm": 5.805969715118408,
+      "learning_rate": 1.9751493543055634e-05,
+      "loss": 0.9636,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6368,
+      "grad_norm": 9.242039680480957,
+      "learning_rate": 1.9757642314139977e-05,
+      "loss": 0.8385,
+      "step": 1592
+    },
+    {
+      "epoch": 0.6376,
+      "grad_norm": 4.233711242675781,
+      "learning_rate": 1.976371499316945e-05,
+      "loss": 0.2892,
+      "step": 1594
+    },
+    {
+      "epoch": 0.6384,
+      "grad_norm": 8.57029914855957,
+      "learning_rate": 1.9769711532788083e-05,
+      "loss": 0.5883,
+      "step": 1596
+    },
+    {
+      "epoch": 0.6392,
+      "grad_norm": 3.897994041442871,
+      "learning_rate": 1.9775631886233655e-05,
+      "loss": 0.4078,
+      "step": 1598
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.17345929145813,
+      "learning_rate": 1.9781476007338054e-05,
+      "loss": 0.3577,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6408,
+      "grad_norm": 11.469186782836914,
+      "learning_rate": 1.978724385052766e-05,
+      "loss": 0.5497,
+      "step": 1602
+    },
+    {
+      "epoch": 0.6416,
+      "grad_norm": 1.5659518241882324,
+      "learning_rate": 1.9792935370823673e-05,
+      "loss": 0.2753,
+      "step": 1604
+    },
+    {
+      "epoch": 0.6424,
+      "grad_norm": 2.373509168624878,
+      "learning_rate": 1.979855052384247e-05,
+      "loss": 0.3695,
+      "step": 1606
+    },
+    {
+      "epoch": 0.6432,
+      "grad_norm": 2.043501138687134,
+      "learning_rate": 1.9804089265795956e-05,
+      "loss": 0.3488,
+      "step": 1608
+    },
+    {
+      "epoch": 0.644,
+      "grad_norm": 4.598385334014893,
+      "learning_rate": 1.9809551553491918e-05,
+      "loss": 0.6336,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6448,
+      "grad_norm": 2.1152210235595703,
+      "learning_rate": 1.981493734433433e-05,
+      "loss": 0.302,
+      "step": 1612
+    },
+    {
+      "epoch": 0.6456,
+      "grad_norm": 2.2232983112335205,
+      "learning_rate": 1.982024659632372e-05,
+      "loss": 0.3294,
+      "step": 1614
+    },
+    {
+      "epoch": 0.6464,
+      "grad_norm": 5.589328289031982,
+      "learning_rate": 1.9825479268057472e-05,
+      "loss": 0.2848,
+      "step": 1616
+    },
+    {
+      "epoch": 0.6472,
+      "grad_norm": 7.114500522613525,
+      "learning_rate": 1.9830635318730155e-05,
+      "loss": 0.7318,
+      "step": 1618
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 5.905538082122803,
+      "learning_rate": 1.9835714708133858e-05,
+      "loss": 0.3707,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6488,
+      "grad_norm": 2.066833734512329,
+      "learning_rate": 1.9840717396658483e-05,
+      "loss": 0.5473,
+      "step": 1622
+    },
+    {
+      "epoch": 0.6496,
+      "grad_norm": 3.5688998699188232,
+      "learning_rate": 1.9845643345292055e-05,
+      "loss": 0.3819,
+      "step": 1624
+    },
+    {
+      "epoch": 0.6504,
+      "grad_norm": 1.7835557460784912,
+      "learning_rate": 1.9850492515621038e-05,
+      "loss": 0.309,
+      "step": 1626
+    },
+    {
+      "epoch": 0.6512,
+      "grad_norm": 2.2958924770355225,
+      "learning_rate": 1.985526486983063e-05,
+      "loss": 0.2285,
+      "step": 1628
+    },
+    {
+      "epoch": 0.652,
+      "grad_norm": 3.196260452270508,
+      "learning_rate": 1.985996037070505e-05,
+      "loss": 0.5863,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 8.522262573242188,
+      "learning_rate": 1.9864578981627844e-05,
+      "loss": 0.8679,
+      "step": 1632
+    },
+    {
+      "epoch": 0.6536,
+      "grad_norm": 3.0343809127807617,
+      "learning_rate": 1.9869120666582153e-05,
+      "loss": 0.5473,
+      "step": 1634
+    },
+    {
+      "epoch": 0.6544,
+      "grad_norm": 2.9875917434692383,
+      "learning_rate": 1.9873585390151003e-05,
+      "loss": 0.402,
+      "step": 1636
+    },
+    {
+      "epoch": 0.6552,
+      "grad_norm": 10.342791557312012,
+      "learning_rate": 1.987797311751759e-05,
+      "loss": 0.4833,
+      "step": 1638
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 6.520310401916504,
+      "learning_rate": 1.9882283814465528e-05,
+      "loss": 0.3338,
+      "step": 1640
+    },
+    {
+      "epoch": 0.6568,
+      "grad_norm": 2.931133508682251,
+      "learning_rate": 1.988651744737914e-05,
+      "loss": 0.2669,
+      "step": 1642
+    },
+    {
+      "epoch": 0.6576,
+      "grad_norm": 11.377016067504883,
+      "learning_rate": 1.9890673983243704e-05,
+      "loss": 0.6862,
+      "step": 1644
+    },
+    {
+      "epoch": 0.6584,
+      "grad_norm": 6.504305362701416,
+      "learning_rate": 1.9894753389645723e-05,
+      "loss": 0.329,
+      "step": 1646
+    },
+    {
+      "epoch": 0.6592,
+      "grad_norm": 2.494025230407715,
+      "learning_rate": 1.9898755634773155e-05,
+      "loss": 0.3779,
+      "step": 1648
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.32198429107666,
+      "learning_rate": 1.9902680687415704e-05,
+      "loss": 0.5162,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6608,
+      "grad_norm": 4.862305641174316,
+      "learning_rate": 1.9906528516965014e-05,
+      "loss": 0.6947,
+      "step": 1652
+    },
+    {
+      "epoch": 0.6616,
+      "grad_norm": 12.738851547241211,
+      "learning_rate": 1.9910299093414926e-05,
+      "loss": 0.4528,
+      "step": 1654
+    },
+    {
+      "epoch": 0.6624,
+      "grad_norm": 2.390115737915039,
+      "learning_rate": 1.9913992387361744e-05,
+      "loss": 0.4201,
+      "step": 1656
+    },
+    {
+      "epoch": 0.6632,
+      "grad_norm": 7.348924160003662,
+      "learning_rate": 1.9917608370004414e-05,
+      "loss": 0.4181,
+      "step": 1658
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 3.2150723934173584,
+      "learning_rate": 1.9921147013144782e-05,
+      "loss": 0.3665,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6648,
+      "grad_norm": 4.1383819580078125,
+      "learning_rate": 1.9924608289187786e-05,
+      "loss": 0.4743,
+      "step": 1662
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 6.19020938873291,
+      "learning_rate": 1.9927992171141707e-05,
+      "loss": 0.6406,
+      "step": 1664
+    },
+    {
+      "epoch": 0.6664,
+      "grad_norm": 2.8534698486328125,
+      "learning_rate": 1.9931298632618355e-05,
+      "loss": 0.2139,
+      "step": 1666
+    },
+    {
+      "epoch": 0.6672,
+      "grad_norm": 6.954126834869385,
+      "learning_rate": 1.9934527647833276e-05,
+      "loss": 0.4103,
+      "step": 1668
+    },
+    {
+      "epoch": 0.668,
+      "grad_norm": 1.7877683639526367,
+      "learning_rate": 1.9937679191605964e-05,
+      "loss": 0.4074,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6688,
+      "grad_norm": 3.242607593536377,
+      "learning_rate": 1.9940753239360047e-05,
+      "loss": 0.3618,
+      "step": 1672
+    },
+    {
+      "epoch": 0.6696,
+      "grad_norm": 3.0980658531188965,
+      "learning_rate": 1.994374976712348e-05,
+      "loss": 0.4763,
+      "step": 1674
+    },
+    {
+      "epoch": 0.6704,
+      "grad_norm": 4.143883228302002,
+      "learning_rate": 1.994666875152874e-05,
+      "loss": 0.5616,
+      "step": 1676
+    },
+    {
+      "epoch": 0.6712,
+      "grad_norm": 2.5318260192871094,
+      "learning_rate": 1.9949510169813003e-05,
+      "loss": 0.2323,
+      "step": 1678
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 12.222033500671387,
+      "learning_rate": 1.9952273999818312e-05,
+      "loss": 0.4916,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6728,
+      "grad_norm": 3.9567325115203857,
+      "learning_rate": 1.995496021999177e-05,
+      "loss": 0.4506,
+      "step": 1682
+    },
+    {
+      "epoch": 0.6736,
+      "grad_norm": 2.8990304470062256,
+      "learning_rate": 1.9957568809385693e-05,
+      "loss": 0.4624,
+      "step": 1684
+    },
+    {
+      "epoch": 0.6744,
+      "grad_norm": 3.471292018890381,
+      "learning_rate": 1.9960099747657774e-05,
+      "loss": 0.4444,
+      "step": 1686
+    },
+    {
+      "epoch": 0.6752,
+      "grad_norm": 3.2368650436401367,
+      "learning_rate": 1.996255301507125e-05,
+      "loss": 0.3354,
+      "step": 1688
+    },
+    {
+      "epoch": 0.676,
+      "grad_norm": 4.808667182922363,
+      "learning_rate": 1.9964928592495046e-05,
+      "loss": 0.8105,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6768,
+      "grad_norm": 2.451111316680908,
+      "learning_rate": 1.9967226461403934e-05,
+      "loss": 0.337,
+      "step": 1692
+    },
+    {
+      "epoch": 0.6776,
+      "grad_norm": 11.671585083007812,
+      "learning_rate": 1.996944660387867e-05,
+      "loss": 0.4399,
+      "step": 1694
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 7.658277988433838,
+      "learning_rate": 1.997158900260614e-05,
+      "loss": 0.4076,
+      "step": 1696
+    },
+    {
+      "epoch": 0.6792,
+      "grad_norm": 5.0221686363220215,
+      "learning_rate": 1.9973653640879486e-05,
+      "loss": 0.4134,
+      "step": 1698
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.082271575927734,
+      "learning_rate": 1.9975640502598246e-05,
+      "loss": 0.5284,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6808,
+      "grad_norm": 3.120457649230957,
+      "learning_rate": 1.997754957226847e-05,
+      "loss": 0.5992,
+      "step": 1702
+    },
+    {
+      "epoch": 0.6816,
+      "grad_norm": 1.900876760482788,
+      "learning_rate": 1.9979380835002846e-05,
+      "loss": 0.3063,
+      "step": 1704
+    },
+    {
+      "epoch": 0.6824,
+      "grad_norm": 9.405780792236328,
+      "learning_rate": 1.9981134276520828e-05,
+      "loss": 0.3785,
+      "step": 1706
+    },
+    {
+      "epoch": 0.6832,
+      "grad_norm": 2.7329249382019043,
+      "learning_rate": 1.998280988314872e-05,
+      "loss": 0.2797,
+      "step": 1708
+    },
+    {
+      "epoch": 0.684,
+      "grad_norm": 2.579680919647217,
+      "learning_rate": 1.998440764181981e-05,
+      "loss": 0.471,
+      "step": 1710
+    },
+    {
+      "epoch": 0.6848,
+      "grad_norm": 5.701201915740967,
+      "learning_rate": 1.9985927540074453e-05,
+      "loss": 0.3976,
+      "step": 1712
+    },
+    {
+      "epoch": 0.6856,
+      "grad_norm": 7.745370864868164,
+      "learning_rate": 1.998736956606018e-05,
+      "loss": 0.3947,
+      "step": 1714
+    },
+    {
+      "epoch": 0.6864,
+      "grad_norm": 8.640480995178223,
+      "learning_rate": 1.9988733708531772e-05,
+      "loss": 1.0942,
+      "step": 1716
+    },
+    {
+      "epoch": 0.6872,
+      "grad_norm": 8.300025939941406,
+      "learning_rate": 1.9990019956851384e-05,
+      "loss": 0.7977,
+      "step": 1718
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 5.9511895179748535,
+      "learning_rate": 1.9991228300988586e-05,
+      "loss": 0.4706,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6888,
+      "grad_norm": 5.14940881729126,
+      "learning_rate": 1.999235873152047e-05,
+      "loss": 0.2438,
+      "step": 1722
+    },
+    {
+      "epoch": 0.6896,
+      "grad_norm": 6.6042985916137695,
+      "learning_rate": 1.9993411239631713e-05,
+      "loss": 0.3952,
+      "step": 1724
+    },
+    {
+      "epoch": 0.6904,
+      "grad_norm": 6.875572681427002,
+      "learning_rate": 1.9994385817114644e-05,
+      "loss": 0.4717,
+      "step": 1726
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 2.505627155303955,
+      "learning_rate": 1.9995282456369313e-05,
+      "loss": 0.3459,
+      "step": 1728
+    },
+    {
+      "epoch": 0.692,
+      "grad_norm": 6.113884925842285,
+      "learning_rate": 1.9996101150403543e-05,
+      "loss": 0.5111,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6928,
+      "grad_norm": 2.1594738960266113,
+      "learning_rate": 1.9996841892833e-05,
+      "loss": 0.2743,
+      "step": 1732
+    },
+    {
+      "epoch": 0.6936,
+      "grad_norm": 20.72479820251465,
+      "learning_rate": 1.9997504677881224e-05,
+      "loss": 0.9828,
+      "step": 1734
+    },
+    {
+      "epoch": 0.6944,
+      "grad_norm": 7.1836700439453125,
+      "learning_rate": 1.999808950037968e-05,
+      "loss": 0.493,
+      "step": 1736
+    },
+    {
+      "epoch": 0.6952,
+      "grad_norm": 5.67575740814209,
+      "learning_rate": 1.9998596355767805e-05,
+      "loss": 0.3416,
+      "step": 1738
+    },
+    {
+      "epoch": 0.696,
+      "grad_norm": 9.29202651977539,
+      "learning_rate": 1.9999025240093045e-05,
+      "loss": 0.5197,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6968,
+      "grad_norm": 5.713569164276123,
+      "learning_rate": 1.9999376150010868e-05,
+      "loss": 0.4243,
+      "step": 1742
+    },
+    {
+      "epoch": 0.6976,
+      "grad_norm": 3.9301981925964355,
+      "learning_rate": 1.9999649082784807e-05,
+      "loss": 0.3626,
+      "step": 1744
+    },
+    {
+      "epoch": 0.6984,
+      "grad_norm": 3.975416898727417,
+      "learning_rate": 1.9999844036286483e-05,
+      "loss": 0.8586,
+      "step": 1746
+    },
+    {
+      "epoch": 0.6992,
+      "grad_norm": 6.84246826171875,
+      "learning_rate": 1.9999961008995607e-05,
+      "loss": 0.7227,
+      "step": 1748
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 20.004087448120117,
+      "learning_rate": 2e-05,
+      "loss": 0.9191,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7008,
+      "grad_norm": 5.791635990142822,
+      "learning_rate": 1.9999961008995607e-05,
+      "loss": 0.4023,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7016,
+      "grad_norm": 5.369904041290283,
+      "learning_rate": 1.9999844036286483e-05,
+      "loss": 0.3795,
+      "step": 1754
+    },
+    {
+      "epoch": 0.7024,
+      "grad_norm": 3.682668924331665,
+      "learning_rate": 1.9999649082784807e-05,
+      "loss": 0.5381,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7032,
+      "grad_norm": 2.7093493938446045,
+      "learning_rate": 1.9999376150010868e-05,
+      "loss": 0.6459,
+      "step": 1758
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 2.522735357284546,
+      "learning_rate": 1.9999025240093045e-05,
+      "loss": 0.5787,
+      "step": 1760
+    },
+    {
+      "epoch": 0.7048,
+      "grad_norm": 8.20150375366211,
+      "learning_rate": 1.9998596355767805e-05,
+      "loss": 0.5729,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7056,
+      "grad_norm": 8.837404251098633,
+      "learning_rate": 1.999808950037968e-05,
+      "loss": 0.5134,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7064,
+      "grad_norm": 5.532891273498535,
+      "learning_rate": 1.9997504677881224e-05,
+      "loss": 0.7412,
+      "step": 1766
+    },
+    {
+      "epoch": 0.7072,
+      "grad_norm": 5.736209869384766,
+      "learning_rate": 1.9996841892833e-05,
+      "loss": 0.5258,
+      "step": 1768
+    },
+    {
+      "epoch": 0.708,
+      "grad_norm": 4.132336616516113,
+      "learning_rate": 1.9996101150403547e-05,
+      "loss": 0.3544,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7088,
+      "grad_norm": 4.415733337402344,
+      "learning_rate": 1.9995282456369313e-05,
+      "loss": 0.4621,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7096,
+      "grad_norm": 4.9113264083862305,
+      "learning_rate": 1.9994385817114644e-05,
+      "loss": 0.5411,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 3.3670785427093506,
+      "learning_rate": 1.9993411239631713e-05,
+      "loss": 0.3433,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7112,
+      "grad_norm": 3.183927059173584,
+      "learning_rate": 1.999235873152047e-05,
+      "loss": 0.4211,
+      "step": 1778
+    },
+    {
+      "epoch": 0.712,
+      "grad_norm": 8.622447967529297,
+      "learning_rate": 1.9991228300988586e-05,
+      "loss": 1.7236,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7128,
+      "grad_norm": 3.417908191680908,
+      "learning_rate": 1.9990019956851384e-05,
+      "loss": 0.2192,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7136,
+      "grad_norm": 15.327376365661621,
+      "learning_rate": 1.9988733708531772e-05,
+      "loss": 0.949,
+      "step": 1784
+    },
+    {
+      "epoch": 0.7144,
+      "grad_norm": 6.369883060455322,
+      "learning_rate": 1.998736956606018e-05,
+      "loss": 0.5447,
+      "step": 1786
+    },
+    {
+      "epoch": 0.7152,
+      "grad_norm": 5.282643795013428,
+      "learning_rate": 1.9985927540074453e-05,
+      "loss": 0.2633,
+      "step": 1788
+    },
+    {
+      "epoch": 0.716,
+      "grad_norm": 1.8992022275924683,
+      "learning_rate": 1.9984407641819812e-05,
+      "loss": 0.3532,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 9.116549491882324,
+      "learning_rate": 1.998280988314872e-05,
+      "loss": 0.5309,
+      "step": 1792
+    },
+    {
+      "epoch": 0.7176,
+      "grad_norm": 4.018207550048828,
+      "learning_rate": 1.9981134276520828e-05,
+      "loss": 0.6177,
+      "step": 1794
+    },
+    {
+      "epoch": 0.7184,
+      "grad_norm": 3.2702534198760986,
+      "learning_rate": 1.9979380835002846e-05,
+      "loss": 0.5697,
+      "step": 1796
+    },
+    {
+      "epoch": 0.7192,
+      "grad_norm": 13.79327392578125,
+      "learning_rate": 1.9977549572268467e-05,
+      "loss": 0.86,
+      "step": 1798
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421355724334717,
+      "learning_rate": 1.9975640502598246e-05,
+      "loss": 0.3692,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7208,
+      "grad_norm": 6.39658260345459,
+      "learning_rate": 1.9973653640879486e-05,
+      "loss": 0.5341,
+      "step": 1802
+    },
+    {
+      "epoch": 0.7216,
+      "grad_norm": 12.072526931762695,
+      "learning_rate": 1.997158900260614e-05,
+      "loss": 0.843,
+      "step": 1804
+    },
+    {
+      "epoch": 0.7224,
+      "grad_norm": 6.202824115753174,
+      "learning_rate": 1.9969446603878673e-05,
+      "loss": 0.4508,
+      "step": 1806
+    },
+    {
+      "epoch": 0.7232,
+      "grad_norm": 12.091536521911621,
+      "learning_rate": 1.9967226461403934e-05,
+      "loss": 0.6631,
+      "step": 1808
+    },
+    {
+      "epoch": 0.724,
+      "grad_norm": 3.4513208866119385,
+      "learning_rate": 1.9964928592495046e-05,
+      "loss": 0.3141,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7248,
+      "grad_norm": 11.12137508392334,
+      "learning_rate": 1.996255301507125e-05,
+      "loss": 0.539,
+      "step": 1812
+    },
+    {
+      "epoch": 0.7256,
+      "grad_norm": 2.0206141471862793,
+      "learning_rate": 1.9960099747657774e-05,
+      "loss": 0.658,
+      "step": 1814
+    },
+    {
+      "epoch": 0.7264,
+      "grad_norm": 7.8277106285095215,
+      "learning_rate": 1.9957568809385693e-05,
+      "loss": 0.5622,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7272,
+      "grad_norm": 4.81088924407959,
+      "learning_rate": 1.995496021999177e-05,
+      "loss": 0.3566,
+      "step": 1818
+    },
+    {
+      "epoch": 0.728,
+      "grad_norm": 4.36074161529541,
+      "learning_rate": 1.9952273999818312e-05,
+      "loss": 0.4191,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7288,
+      "grad_norm": 4.0613226890563965,
+      "learning_rate": 1.9949510169813006e-05,
+      "loss": 0.4074,
+      "step": 1822
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 5.445612907409668,
+      "learning_rate": 1.9946668751528745e-05,
+      "loss": 0.3167,
+      "step": 1824
+    },
+    {
+      "epoch": 0.7304,
+      "grad_norm": 2.8114616870880127,
+      "learning_rate": 1.994374976712348e-05,
+      "loss": 0.6147,
+      "step": 1826
+    },
+    {
+      "epoch": 0.7312,
+      "grad_norm": 2.5986952781677246,
+      "learning_rate": 1.9940753239360047e-05,
+      "loss": 0.5259,
+      "step": 1828
+    },
+    {
+      "epoch": 0.732,
+      "grad_norm": 3.6003918647766113,
+      "learning_rate": 1.9937679191605964e-05,
+      "loss": 0.3197,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7328,
+      "grad_norm": 3.8639371395111084,
+      "learning_rate": 1.993452764783328e-05,
+      "loss": 0.21,
+      "step": 1832
+    },
+    {
+      "epoch": 0.7336,
+      "grad_norm": 2.496325969696045,
+      "learning_rate": 1.9931298632618352e-05,
+      "loss": 0.3632,
+      "step": 1834
+    },
+    {
+      "epoch": 0.7344,
+      "grad_norm": 2.928487777709961,
+      "learning_rate": 1.9927992171141707e-05,
+      "loss": 0.3398,
+      "step": 1836
+    },
+    {
+      "epoch": 0.7352,
+      "grad_norm": 5.671520709991455,
+      "learning_rate": 1.9924608289187786e-05,
+      "loss": 0.5193,
+      "step": 1838
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 3.602909803390503,
+      "learning_rate": 1.9921147013144782e-05,
+      "loss": 0.1562,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7368,
+      "grad_norm": 2.749011993408203,
+      "learning_rate": 1.9917608370004417e-05,
+      "loss": 0.315,
+      "step": 1842
+    },
+    {
+      "epoch": 0.7376,
+      "grad_norm": 4.273372173309326,
+      "learning_rate": 1.9913992387361747e-05,
+      "loss": 0.3395,
+      "step": 1844
+    },
+    {
+      "epoch": 0.7384,
+      "grad_norm": 5.967968463897705,
+      "learning_rate": 1.9910299093414932e-05,
+      "loss": 0.4194,
+      "step": 1846
+    },
+    {
+      "epoch": 0.7392,
+      "grad_norm": 2.5091896057128906,
+      "learning_rate": 1.990652851696501e-05,
+      "loss": 0.5031,
+      "step": 1848
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.949019432067871,
+      "learning_rate": 1.9902680687415704e-05,
+      "loss": 1.0475,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7408,
+      "grad_norm": 2.438105821609497,
+      "learning_rate": 1.9898755634773155e-05,
+      "loss": 0.3475,
+      "step": 1852
+    },
+    {
+      "epoch": 0.7416,
+      "grad_norm": 8.848106384277344,
+      "learning_rate": 1.9894753389645723e-05,
+      "loss": 0.4821,
+      "step": 1854
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 5.018491744995117,
+      "learning_rate": 1.9890673983243704e-05,
+      "loss": 0.3251,
+      "step": 1856
+    },
+    {
+      "epoch": 0.7432,
+      "grad_norm": 5.742210388183594,
+      "learning_rate": 1.9886517447379143e-05,
+      "loss": 0.2919,
+      "step": 1858
+    },
+    {
+      "epoch": 0.744,
+      "grad_norm": 10.081647872924805,
+      "learning_rate": 1.988228381446553e-05,
+      "loss": 0.5546,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7448,
+      "grad_norm": 4.09962272644043,
+      "learning_rate": 1.987797311751759e-05,
+      "loss": 0.3861,
+      "step": 1862
+    },
+    {
+      "epoch": 0.7456,
+      "grad_norm": 3.1583993434906006,
+      "learning_rate": 1.9873585390151007e-05,
+      "loss": 0.3128,
+      "step": 1864
+    },
+    {
+      "epoch": 0.7464,
+      "grad_norm": 1.916212797164917,
+      "learning_rate": 1.9869120666582153e-05,
+      "loss": 0.3608,
+      "step": 1866
+    },
+    {
+      "epoch": 0.7472,
+      "grad_norm": 8.458840370178223,
+      "learning_rate": 1.9864578981627844e-05,
+      "loss": 0.7873,
+      "step": 1868
+    },
+    {
+      "epoch": 0.748,
+      "grad_norm": 6.865233898162842,
+      "learning_rate": 1.985996037070505e-05,
+      "loss": 0.3612,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 10.543816566467285,
+      "learning_rate": 1.985526486983063e-05,
+      "loss": 0.698,
+      "step": 1872
+    },
+    {
+      "epoch": 0.7496,
+      "grad_norm": 6.476944446563721,
+      "learning_rate": 1.9850492515621038e-05,
+      "loss": 0.5256,
+      "step": 1874
+    },
+    {
+      "epoch": 0.7504,
+      "grad_norm": 8.187792778015137,
+      "learning_rate": 1.9845643345292055e-05,
+      "loss": 0.4788,
+      "step": 1876
+    },
+    {
+      "epoch": 0.7512,
+      "grad_norm": 1.9630470275878906,
+      "learning_rate": 1.9840717396658483e-05,
+      "loss": 0.2404,
+      "step": 1878
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 8.175183296203613,
+      "learning_rate": 1.983571470813386e-05,
+      "loss": 0.699,
+      "step": 1880
+    },
+    {
+      "epoch": 0.7528,
+      "grad_norm": 12.990069389343262,
+      "learning_rate": 1.983063531873016e-05,
+      "loss": 1.1128,
+      "step": 1882
+    },
+    {
+      "epoch": 0.7536,
+      "grad_norm": 5.050752639770508,
+      "learning_rate": 1.982547926805747e-05,
+      "loss": 0.7573,
+      "step": 1884
+    },
+    {
+      "epoch": 0.7544,
+      "grad_norm": 6.0260515213012695,
+      "learning_rate": 1.9820246596323724e-05,
+      "loss": 0.3665,
+      "step": 1886
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 14.117484092712402,
+      "learning_rate": 1.981493734433433e-05,
+      "loss": 0.8667,
+      "step": 1888
+    },
+    {
+      "epoch": 0.756,
+      "grad_norm": 11.667659759521484,
+      "learning_rate": 1.9809551553491918e-05,
+      "loss": 0.6158,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7568,
+      "grad_norm": 5.7208380699157715,
+      "learning_rate": 1.9804089265795963e-05,
+      "loss": 0.5358,
+      "step": 1892
+    },
+    {
+      "epoch": 0.7576,
+      "grad_norm": 5.21250057220459,
+      "learning_rate": 1.979855052384247e-05,
+      "loss": 0.4207,
+      "step": 1894
+    },
+    {
+      "epoch": 0.7584,
+      "grad_norm": 3.6927437782287598,
+      "learning_rate": 1.979293537082368e-05,
+      "loss": 0.6051,
+      "step": 1896
+    },
+    {
+      "epoch": 0.7592,
+      "grad_norm": 4.236647605895996,
+      "learning_rate": 1.9787243850527663e-05,
+      "loss": 0.3979,
+      "step": 1898
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.474388122558594,
+      "learning_rate": 1.978147600733806e-05,
+      "loss": 0.6603,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7608,
+      "grad_norm": 3.9843273162841797,
+      "learning_rate": 1.977563188623365e-05,
+      "loss": 0.4227,
+      "step": 1902
+    },
+    {
+      "epoch": 0.7616,
+      "grad_norm": 2.3406167030334473,
+      "learning_rate": 1.9769711532788086e-05,
+      "loss": 0.2071,
+      "step": 1904
+    },
+    {
+      "epoch": 0.7624,
+      "grad_norm": 8.789372444152832,
+      "learning_rate": 1.9763714993169448e-05,
+      "loss": 0.5955,
+      "step": 1906
+    },
+    {
+      "epoch": 0.7632,
+      "grad_norm": 4.878142833709717,
+      "learning_rate": 1.9757642314139977e-05,
+      "loss": 0.3273,
+      "step": 1908
+    },
+    {
+      "epoch": 0.764,
+      "grad_norm": 3.703111410140991,
+      "learning_rate": 1.9751493543055638e-05,
+      "loss": 0.2455,
+      "step": 1910
+    },
+    {
+      "epoch": 0.7648,
+      "grad_norm": 1.3227205276489258,
+      "learning_rate": 1.9745268727865774e-05,
+      "loss": 0.1464,
+      "step": 1912
+    },
+    {
+      "epoch": 0.7656,
+      "grad_norm": 2.471626043319702,
+      "learning_rate": 1.973896791711276e-05,
+      "loss": 0.4114,
+      "step": 1914
+    },
+    {
+      "epoch": 0.7664,
+      "grad_norm": 6.342803955078125,
+      "learning_rate": 1.9732591159931567e-05,
+      "loss": 0.3901,
+      "step": 1916
+    },
+    {
+      "epoch": 0.7672,
+      "grad_norm": 3.0534446239471436,
+      "learning_rate": 1.972613850604944e-05,
+      "loss": 0.6209,
+      "step": 1918
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 6.819042682647705,
+      "learning_rate": 1.9719610005785463e-05,
+      "loss": 0.4309,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7688,
+      "grad_norm": 7.792922496795654,
+      "learning_rate": 1.9713005710050206e-05,
+      "loss": 0.4873,
+      "step": 1922
+    },
+    {
+      "epoch": 0.7696,
+      "grad_norm": 5.196712493896484,
+      "learning_rate": 1.9706325670345276e-05,
+      "loss": 0.3977,
+      "step": 1924
+    },
+    {
+      "epoch": 0.7704,
+      "grad_norm": 5.214966297149658,
+      "learning_rate": 1.9699569938762975e-05,
+      "loss": 0.4326,
+      "step": 1926
+    },
+    {
+      "epoch": 0.7712,
+      "grad_norm": 8.276177406311035,
+      "learning_rate": 1.969273856798586e-05,
+      "loss": 0.446,
+      "step": 1928
+    },
+    {
+      "epoch": 0.772,
+      "grad_norm": 3.4883368015289307,
+      "learning_rate": 1.9685831611286312e-05,
+      "loss": 0.3672,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7728,
+      "grad_norm": 1.9015693664550781,
+      "learning_rate": 1.9678849122526195e-05,
+      "loss": 0.3345,
+      "step": 1932
+    },
+    {
+      "epoch": 0.7736,
+      "grad_norm": 8.110586166381836,
+      "learning_rate": 1.967179115615633e-05,
+      "loss": 0.4249,
+      "step": 1934
+    },
+    {
+      "epoch": 0.7744,
+      "grad_norm": 4.227611064910889,
+      "learning_rate": 1.966465776721618e-05,
+      "loss": 0.6143,
+      "step": 1936
+    },
+    {
+      "epoch": 0.7752,
+      "grad_norm": 2.341872215270996,
+      "learning_rate": 1.9657449011333328e-05,
+      "loss": 0.5497,
+      "step": 1938
+    },
+    {
+      "epoch": 0.776,
+      "grad_norm": 2.357506513595581,
+      "learning_rate": 1.965016494472312e-05,
+      "loss": 0.2292,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7768,
+      "grad_norm": 3.373274087905884,
+      "learning_rate": 1.964280562418815e-05,
+      "loss": 0.2772,
+      "step": 1942
+    },
+    {
+      "epoch": 0.7776,
+      "grad_norm": 27.803871154785156,
+      "learning_rate": 1.963537110711789e-05,
+      "loss": 0.8697,
+      "step": 1944
+    },
+    {
+      "epoch": 0.7784,
+      "grad_norm": 4.842161655426025,
+      "learning_rate": 1.9627861451488194e-05,
+      "loss": 0.4879,
+      "step": 1946
+    },
+    {
+      "epoch": 0.7792,
+      "grad_norm": 8.67030143737793,
+      "learning_rate": 1.962027671586086e-05,
+      "loss": 0.5492,
+      "step": 1948
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.379067897796631,
+      "learning_rate": 1.9612616959383194e-05,
+      "loss": 0.3028,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 3.8166677951812744,
+      "learning_rate": 1.96048822417875e-05,
+      "loss": 0.2783,
+      "step": 1952
+    },
+    {
+      "epoch": 0.7816,
+      "grad_norm": 3.9429714679718018,
+      "learning_rate": 1.9597072623390668e-05,
+      "loss": 0.4552,
+      "step": 1954
+    },
+    {
+      "epoch": 0.7824,
+      "grad_norm": 2.631047487258911,
+      "learning_rate": 1.9589188165093666e-05,
+      "loss": 0.3029,
+      "step": 1956
+    },
+    {
+      "epoch": 0.7832,
+      "grad_norm": 2.698028087615967,
+      "learning_rate": 1.95812289283811e-05,
+      "loss": 0.1427,
+      "step": 1958
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 6.577787399291992,
+      "learning_rate": 1.9573194975320672e-05,
+      "loss": 0.565,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7848,
+      "grad_norm": 2.450500249862671,
+      "learning_rate": 1.9565086368562784e-05,
+      "loss": 0.4542,
+      "step": 1962
+    },
+    {
+      "epoch": 0.7856,
+      "grad_norm": 10.015090942382812,
+      "learning_rate": 1.9556903171339966e-05,
+      "loss": 0.3882,
+      "step": 1964
+    },
+    {
+      "epoch": 0.7864,
+      "grad_norm": 4.287123680114746,
+      "learning_rate": 1.954864544746643e-05,
+      "loss": 0.2544,
+      "step": 1966
+    },
+    {
+      "epoch": 0.7872,
+      "grad_norm": 3.965257406234741,
+      "learning_rate": 1.9540313261337585e-05,
+      "loss": 0.4182,
+      "step": 1968
+    },
+    {
+      "epoch": 0.788,
+      "grad_norm": 5.50913143157959,
+      "learning_rate": 1.9531906677929472e-05,
+      "loss": 0.3617,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7888,
+      "grad_norm": 3.2805440425872803,
+      "learning_rate": 1.9523425762798335e-05,
+      "loss": 0.204,
+      "step": 1972
+    },
+    {
+      "epoch": 0.7896,
+      "grad_norm": 10.288555145263672,
+      "learning_rate": 1.9514870582080035e-05,
+      "loss": 0.6548,
+      "step": 1974
+    },
+    {
+      "epoch": 0.7904,
+      "grad_norm": 5.323610782623291,
+      "learning_rate": 1.95062412024896e-05,
+      "loss": 1.676,
+      "step": 1976
+    },
+    {
+      "epoch": 0.7912,
+      "grad_norm": 12.464410781860352,
+      "learning_rate": 1.9497537691320667e-05,
+      "loss": 0.7239,
+      "step": 1978
+    },
+    {
+      "epoch": 0.792,
+      "grad_norm": 3.5932133197784424,
+      "learning_rate": 1.948876011644497e-05,
+      "loss": 0.3674,
+      "step": 1980
+    },
+    {
+      "epoch": 0.7928,
+      "grad_norm": 3.760422945022583,
+      "learning_rate": 1.9479908546311787e-05,
+      "loss": 0.3465,
+      "step": 1982
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 5.273140907287598,
+      "learning_rate": 1.9470983049947443e-05,
+      "loss": 0.5971,
+      "step": 1984
+    },
+    {
+      "epoch": 0.7944,
+      "grad_norm": 2.991767168045044,
+      "learning_rate": 1.9461983696954767e-05,
+      "loss": 0.27,
+      "step": 1986
+    },
+    {
+      "epoch": 0.7952,
+      "grad_norm": 2.138972759246826,
+      "learning_rate": 1.9452910557512494e-05,
+      "loss": 0.2681,
+      "step": 1988
+    },
+    {
+      "epoch": 0.796,
+      "grad_norm": 8.971880912780762,
+      "learning_rate": 1.9443763702374818e-05,
+      "loss": 0.6808,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7968,
+      "grad_norm": 2.9185991287231445,
+      "learning_rate": 1.9434543202870726e-05,
+      "loss": 0.6934,
+      "step": 1992
+    },
+    {
+      "epoch": 0.7976,
+      "grad_norm": 4.297379493713379,
+      "learning_rate": 1.9425249130903544e-05,
+      "loss": 0.3042,
+      "step": 1994
+    },
+    {
+      "epoch": 0.7984,
+      "grad_norm": 3.2408552169799805,
+      "learning_rate": 1.94158815589503e-05,
+      "loss": 0.5953,
+      "step": 1996
+    },
+    {
+      "epoch": 0.7992,
+      "grad_norm": 8.040743827819824,
+      "learning_rate": 1.940644056006122e-05,
+      "loss": 0.8196,
+      "step": 1998
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.990960597991943,
+      "learning_rate": 1.939692620785909e-05,
+      "loss": 0.4612,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8008,
+      "grad_norm": 6.383898735046387,
+      "learning_rate": 1.9387338576538746e-05,
+      "loss": 0.4105,
+      "step": 2002
+    },
+    {
+      "epoch": 0.8016,
+      "grad_norm": 6.999117851257324,
+      "learning_rate": 1.9377677740866464e-05,
+      "loss": 0.2251,
+      "step": 2004
+    },
+    {
+      "epoch": 0.8024,
+      "grad_norm": 7.917981147766113,
+      "learning_rate": 1.936794377617938e-05,
+      "loss": 0.7872,
+      "step": 2006
+    },
+    {
+      "epoch": 0.8032,
+      "grad_norm": 5.2557501792907715,
+      "learning_rate": 1.9358136758384917e-05,
+      "loss": 0.5837,
+      "step": 2008
+    },
+    {
+      "epoch": 0.804,
+      "grad_norm": 4.809940338134766,
+      "learning_rate": 1.9348256763960146e-05,
+      "loss": 0.6812,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8048,
+      "grad_norm": 2.842546224594116,
+      "learning_rate": 1.9338303869951273e-05,
+      "loss": 0.3981,
+      "step": 2012
+    },
+    {
+      "epoch": 0.8056,
+      "grad_norm": 3.4043211936950684,
+      "learning_rate": 1.9328278153972943e-05,
+      "loss": 0.2863,
+      "step": 2014
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 2.5935723781585693,
+      "learning_rate": 1.931817969420773e-05,
+      "loss": 0.353,
+      "step": 2016
+    },
+    {
+      "epoch": 0.8072,
+      "grad_norm": 4.572518825531006,
+      "learning_rate": 1.930800856940543e-05,
+      "loss": 0.3879,
+      "step": 2018
+    },
+    {
+      "epoch": 0.808,
+      "grad_norm": 4.014364242553711,
+      "learning_rate": 1.929776485888252e-05,
+      "loss": 0.5561,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8088,
+      "grad_norm": 4.8947906494140625,
+      "learning_rate": 1.9287448642521517e-05,
+      "loss": 0.734,
+      "step": 2022
+    },
+    {
+      "epoch": 0.8096,
+      "grad_norm": 1.952275037765503,
+      "learning_rate": 1.9277060000770342e-05,
+      "loss": 0.5817,
+      "step": 2024
+    },
+    {
+      "epoch": 0.8104,
+      "grad_norm": 4.268755912780762,
+      "learning_rate": 1.9266599014641727e-05,
+      "loss": 0.5069,
+      "step": 2026
+    },
+    {
+      "epoch": 0.8112,
+      "grad_norm": 3.921480655670166,
+      "learning_rate": 1.925606576571252e-05,
+      "loss": 0.5792,
+      "step": 2028
+    },
+    {
+      "epoch": 0.812,
+      "grad_norm": 12.108016967773438,
+      "learning_rate": 1.9245460336123136e-05,
+      "loss": 0.6826,
+      "step": 2030
+    },
+    {
+      "epoch": 0.8128,
+      "grad_norm": 3.33408522605896,
+      "learning_rate": 1.923478280857682e-05,
+      "loss": 0.464,
+      "step": 2032
+    },
+    {
+      "epoch": 0.8136,
+      "grad_norm": 1.6179202795028687,
+      "learning_rate": 1.9224033266339103e-05,
+      "loss": 0.2836,
+      "step": 2034
+    },
+    {
+      "epoch": 0.8144,
+      "grad_norm": 12.353996276855469,
+      "learning_rate": 1.9213211793237066e-05,
+      "loss": 0.7539,
+      "step": 2036
+    },
+    {
+      "epoch": 0.8152,
+      "grad_norm": 6.888246536254883,
+      "learning_rate": 1.9202318473658707e-05,
+      "loss": 0.5177,
+      "step": 2038
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 4.788189888000488,
+      "learning_rate": 1.919135339255235e-05,
+      "loss": 0.6794,
+      "step": 2040
+    },
+    {
+      "epoch": 0.8168,
+      "grad_norm": 3.879225492477417,
+      "learning_rate": 1.918031663542588e-05,
+      "loss": 0.2968,
+      "step": 2042
+    },
+    {
+      "epoch": 0.8176,
+      "grad_norm": 2.4591026306152344,
+      "learning_rate": 1.916920828834617e-05,
+      "loss": 0.3379,
+      "step": 2044
+    },
+    {
+      "epoch": 0.8184,
+      "grad_norm": 3.039889335632324,
+      "learning_rate": 1.9158028437938313e-05,
+      "loss": 0.4206,
+      "step": 2046
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 9.00166130065918,
+      "learning_rate": 1.9146777171385057e-05,
+      "loss": 0.7119,
+      "step": 2048
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 16.83409309387207,
+      "learning_rate": 1.913545457642601e-05,
+      "loss": 0.9266,
+      "step": 2050
+    },
+    {
+      "epoch": 0.8208,
+      "grad_norm": 8.59817123413086,
+      "learning_rate": 1.9124060741357065e-05,
+      "loss": 0.4125,
+      "step": 2052
+    },
+    {
+      "epoch": 0.8216,
+      "grad_norm": 4.9218525886535645,
+      "learning_rate": 1.911259575502963e-05,
+      "loss": 0.8163,
+      "step": 2054
+    },
+    {
+      "epoch": 0.8224,
+      "grad_norm": 7.36513090133667,
+      "learning_rate": 1.910105970684996e-05,
+      "loss": 0.4727,
+      "step": 2056
+    },
+    {
+      "epoch": 0.8232,
+      "grad_norm": 2.194617986679077,
+      "learning_rate": 1.908945268677849e-05,
+      "loss": 0.8017,
+      "step": 2058
+    },
+    {
+      "epoch": 0.824,
+      "grad_norm": 4.550289154052734,
+      "learning_rate": 1.9077774785329085e-05,
+      "loss": 0.3739,
+      "step": 2060
+    },
+    {
+      "epoch": 0.8248,
+      "grad_norm": 3.4990861415863037,
+      "learning_rate": 1.9066026093568383e-05,
+      "loss": 0.3745,
+      "step": 2062
+    },
+    {
+      "epoch": 0.8256,
+      "grad_norm": 8.7343111038208,
+      "learning_rate": 1.9054206703115013e-05,
+      "loss": 0.8979,
+      "step": 2064
+    },
+    {
+      "epoch": 0.8264,
+      "grad_norm": 2.711123466491699,
+      "learning_rate": 1.9042316706138994e-05,
+      "loss": 0.3762,
+      "step": 2066
+    },
+    {
+      "epoch": 0.8272,
+      "grad_norm": 1.9720872640609741,
+      "learning_rate": 1.903035619536087e-05,
+      "loss": 0.4065,
+      "step": 2068
+    },
+    {
+      "epoch": 0.828,
+      "grad_norm": 4.787201881408691,
+      "learning_rate": 1.901832526405114e-05,
+      "loss": 0.4545,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8288,
+      "grad_norm": 4.116015434265137,
+      "learning_rate": 1.9006224006029414e-05,
+      "loss": 0.3861,
+      "step": 2072
+    },
+    {
+      "epoch": 0.8296,
+      "grad_norm": 4.5069451332092285,
+      "learning_rate": 1.899405251566371e-05,
+      "loss": 0.5742,
+      "step": 2074
+    },
+    {
+      "epoch": 0.8304,
+      "grad_norm": 5.005687713623047,
+      "learning_rate": 1.8981810887869797e-05,
+      "loss": 0.5577,
+      "step": 2076
+    },
+    {
+      "epoch": 0.8312,
+      "grad_norm": 5.526610374450684,
+      "learning_rate": 1.8969499218110302e-05,
+      "loss": 0.4934,
+      "step": 2078
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 2.5126290321350098,
+      "learning_rate": 1.8957117602394133e-05,
+      "loss": 0.3066,
+      "step": 2080
+    },
+    {
+      "epoch": 0.8328,
+      "grad_norm": 9.842791557312012,
+      "learning_rate": 1.8944666137275596e-05,
+      "loss": 0.6265,
+      "step": 2082
+    },
+    {
+      "epoch": 0.8336,
+      "grad_norm": 13.449177742004395,
+      "learning_rate": 1.8932144919853744e-05,
+      "loss": 0.7168,
+      "step": 2084
+    },
+    {
+      "epoch": 0.8344,
+      "grad_norm": 4.880086421966553,
+      "learning_rate": 1.8919554047771508e-05,
+      "loss": 0.3834,
+      "step": 2086
+    },
+    {
+      "epoch": 0.8352,
+      "grad_norm": 5.4077019691467285,
+      "learning_rate": 1.890689361921507e-05,
+      "loss": 0.3614,
+      "step": 2088
+    },
+    {
+      "epoch": 0.836,
+      "grad_norm": 2.8501031398773193,
+      "learning_rate": 1.8894163732912986e-05,
+      "loss": 0.4492,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8368,
+      "grad_norm": 5.37775993347168,
+      "learning_rate": 1.8881364488135445e-05,
+      "loss": 0.4278,
+      "step": 2092
+    },
+    {
+      "epoch": 0.8376,
+      "grad_norm": 2.3796370029449463,
+      "learning_rate": 1.886849598469357e-05,
+      "loss": 0.3203,
+      "step": 2094
+    },
+    {
+      "epoch": 0.8384,
+      "grad_norm": 5.196391582489014,
+      "learning_rate": 1.8855558322938492e-05,
+      "loss": 0.3425,
+      "step": 2096
+    },
+    {
+      "epoch": 0.8392,
+      "grad_norm": 5.295304775238037,
+      "learning_rate": 1.8842551603760725e-05,
+      "loss": 0.5215,
+      "step": 2098
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.4182562828063965,
+      "learning_rate": 1.8829475928589265e-05,
+      "loss": 0.4456,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8408,
+      "grad_norm": 3.1044559478759766,
+      "learning_rate": 1.8816331399390874e-05,
+      "loss": 0.3528,
+      "step": 2102
+    },
+    {
+      "epoch": 0.8416,
+      "grad_norm": 2.9102487564086914,
+      "learning_rate": 1.88031181186692e-05,
+      "loss": 0.34,
+      "step": 2104
+    },
+    {
+      "epoch": 0.8424,
+      "grad_norm": 6.911310195922852,
+      "learning_rate": 1.8789836189464092e-05,
+      "loss": 0.5755,
+      "step": 2106
+    },
+    {
+      "epoch": 0.8432,
+      "grad_norm": 5.777722358703613,
+      "learning_rate": 1.877648571535068e-05,
+      "loss": 0.8788,
+      "step": 2108
+    },
+    {
+      "epoch": 0.844,
+      "grad_norm": 5.194482326507568,
+      "learning_rate": 1.8763066800438638e-05,
+      "loss": 0.6766,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 10.82712459564209,
+      "learning_rate": 1.8749579549371387e-05,
+      "loss": 0.9228,
+      "step": 2112
+    },
+    {
+      "epoch": 0.8456,
+      "grad_norm": 4.030005931854248,
+      "learning_rate": 1.8736024067325188e-05,
+      "loss": 0.4271,
+      "step": 2114
+    },
+    {
+      "epoch": 0.8464,
+      "grad_norm": 5.070348262786865,
+      "learning_rate": 1.8722400460008437e-05,
+      "loss": 0.5131,
+      "step": 2116
+    },
+    {
+      "epoch": 0.8472,
+      "grad_norm": 3.887665033340454,
+      "learning_rate": 1.8708708833660748e-05,
+      "loss": 2.6075,
+      "step": 2118
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 4.675127983093262,
+      "learning_rate": 1.8694949295052198e-05,
+      "loss": 0.7133,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8488,
+      "grad_norm": 4.268080711364746,
+      "learning_rate": 1.868112195148239e-05,
+      "loss": 0.5845,
+      "step": 2122
+    },
+    {
+      "epoch": 0.8496,
+      "grad_norm": 3.9360458850860596,
+      "learning_rate": 1.866722691077977e-05,
+      "loss": 0.4167,
+      "step": 2124
+    },
+    {
+      "epoch": 0.8504,
+      "grad_norm": 20.160083770751953,
+      "learning_rate": 1.8653264281300626e-05,
+      "loss": 0.9318,
+      "step": 2126
+    },
+    {
+      "epoch": 0.8512,
+      "grad_norm": 4.777723789215088,
+      "learning_rate": 1.8639234171928355e-05,
+      "loss": 0.4258,
+      "step": 2128
+    },
+    {
+      "epoch": 0.852,
+      "grad_norm": 3.6808910369873047,
+      "learning_rate": 1.8625136692072587e-05,
+      "loss": 0.4202,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8528,
+      "grad_norm": 8.434818267822266,
+      "learning_rate": 1.8610971951668265e-05,
+      "loss": 0.7677,
+      "step": 2132
+    },
+    {
+      "epoch": 0.8536,
+      "grad_norm": 8.012714385986328,
+      "learning_rate": 1.8596740061174912e-05,
+      "loss": 0.5678,
+      "step": 2134
+    },
+    {
+      "epoch": 0.8544,
+      "grad_norm": 9.045832633972168,
+      "learning_rate": 1.858244113157566e-05,
+      "loss": 0.3242,
+      "step": 2136
+    },
+    {
+      "epoch": 0.8552,
+      "grad_norm": 6.075767993927002,
+      "learning_rate": 1.8568075274376432e-05,
+      "loss": 0.5078,
+      "step": 2138
+    },
+    {
+      "epoch": 0.856,
+      "grad_norm": 3.5789854526519775,
+      "learning_rate": 1.8553642601605083e-05,
+      "loss": 0.3583,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8568,
+      "grad_norm": 3.273469924926758,
+      "learning_rate": 1.8539143225810457e-05,
+      "loss": 0.162,
+      "step": 2142
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 6.589645862579346,
+      "learning_rate": 1.852457726006163e-05,
+      "loss": 0.7334,
+      "step": 2144
+    },
+    {
+      "epoch": 0.8584,
+      "grad_norm": 6.523824691772461,
+      "learning_rate": 1.8509944817946917e-05,
+      "loss": 0.4509,
+      "step": 2146
+    },
+    {
+      "epoch": 0.8592,
+      "grad_norm": 4.779366970062256,
+      "learning_rate": 1.8495246013573064e-05,
+      "loss": 0.2784,
+      "step": 2148
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8005824089050293,
+      "learning_rate": 1.848048096156426e-05,
+      "loss": 0.3085,
+      "step": 2150
+    },
+    {
+      "epoch": 0.8608,
+      "grad_norm": 4.007396221160889,
+      "learning_rate": 1.8465649777061387e-05,
+      "loss": 0.6944,
+      "step": 2152
+    },
+    {
+      "epoch": 0.8616,
+      "grad_norm": 7.373164653778076,
+      "learning_rate": 1.8450752575720967e-05,
+      "loss": 0.4137,
+      "step": 2154
+    },
+    {
+      "epoch": 0.8624,
+      "grad_norm": 2.44792103767395,
+      "learning_rate": 1.843578947371439e-05,
+      "loss": 0.494,
+      "step": 2156
+    },
+    {
+      "epoch": 0.8632,
+      "grad_norm": 5.441621780395508,
+      "learning_rate": 1.8420760587726935e-05,
+      "loss": 1.3811,
+      "step": 2158
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 14.928495407104492,
+      "learning_rate": 1.8405666034956846e-05,
+      "loss": 0.6873,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8648,
+      "grad_norm": 19.143177032470703,
+      "learning_rate": 1.8390505933114507e-05,
+      "loss": 0.9384,
+      "step": 2162
+    },
+    {
+      "epoch": 0.8656,
+      "grad_norm": 2.426013708114624,
+      "learning_rate": 1.8375280400421414e-05,
+      "loss": 0.4808,
+      "step": 2164
+    },
+    {
+      "epoch": 0.8664,
+      "grad_norm": 4.136850833892822,
+      "learning_rate": 1.8359989555609365e-05,
+      "loss": 0.32,
+      "step": 2166
+    },
+    {
+      "epoch": 0.8672,
+      "grad_norm": 7.991090774536133,
+      "learning_rate": 1.834463351791939e-05,
+      "loss": 0.7678,
+      "step": 2168
+    },
+    {
+      "epoch": 0.868,
+      "grad_norm": 4.490537643432617,
+      "learning_rate": 1.8329212407101006e-05,
+      "loss": 0.5311,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8688,
+      "grad_norm": 7.613470554351807,
+      "learning_rate": 1.8313726343411085e-05,
+      "loss": 0.4121,
+      "step": 2172
+    },
+    {
+      "epoch": 0.8696,
+      "grad_norm": 5.197305679321289,
+      "learning_rate": 1.82981754476131e-05,
+      "loss": 0.1904,
+      "step": 2174
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 9.690563201904297,
+      "learning_rate": 1.8282559840976053e-05,
+      "loss": 0.4502,
+      "step": 2176
+    },
+    {
+      "epoch": 0.8712,
+      "grad_norm": 2.421522855758667,
+      "learning_rate": 1.8266879645273557e-05,
+      "loss": 0.2495,
+      "step": 2178
+    },
+    {
+      "epoch": 0.872,
+      "grad_norm": 7.8181328773498535,
+      "learning_rate": 1.8251134982782966e-05,
+      "loss": 0.2981,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8728,
+      "grad_norm": 1.5006805658340454,
+      "learning_rate": 1.823532597628428e-05,
+      "loss": 0.3555,
+      "step": 2182
+    },
+    {
+      "epoch": 0.8736,
+      "grad_norm": 3.146165609359741,
+      "learning_rate": 1.8219452749059336e-05,
+      "loss": 0.3498,
+      "step": 2184
+    },
+    {
+      "epoch": 0.8744,
+      "grad_norm": 2.798628807067871,
+      "learning_rate": 1.8203515424890734e-05,
+      "loss": 0.5941,
+      "step": 2186
+    },
+    {
+      "epoch": 0.8752,
+      "grad_norm": 7.344174861907959,
+      "learning_rate": 1.8187514128060956e-05,
+      "loss": 0.8017,
+      "step": 2188
+    },
+    {
+      "epoch": 0.876,
+      "grad_norm": 1.4742686748504639,
+      "learning_rate": 1.8171448983351284e-05,
+      "loss": 0.2445,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8768,
+      "grad_norm": 3.6886837482452393,
+      "learning_rate": 1.8155320116040983e-05,
+      "loss": 0.1941,
+      "step": 2192
+    },
+    {
+      "epoch": 0.8776,
+      "grad_norm": 10.147557258605957,
+      "learning_rate": 1.8139127651906193e-05,
+      "loss": 0.7069,
+      "step": 2194
+    },
+    {
+      "epoch": 0.8784,
+      "grad_norm": 2.233205556869507,
+      "learning_rate": 1.8122871717218974e-05,
+      "loss": 0.438,
+      "step": 2196
+    },
+    {
+      "epoch": 0.8792,
+      "grad_norm": 5.252257823944092,
+      "learning_rate": 1.8106552438746413e-05,
+      "loss": 0.7512,
+      "step": 2198
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.280484199523926,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 1.0446,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8808,
+      "grad_norm": 8.377373695373535,
+      "learning_rate": 1.807372435998219e-05,
+      "loss": 0.4684,
+      "step": 2202
+    },
+    {
+      "epoch": 0.8816,
+      "grad_norm": 4.66433048248291,
+      "learning_rate": 1.8057215815690487e-05,
+      "loss": 0.3269,
+      "step": 2204
+    },
+    {
+      "epoch": 0.8824,
+      "grad_norm": 2.217611312866211,
+      "learning_rate": 1.8040644439611355e-05,
+      "loss": 0.5373,
+      "step": 2206
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 5.229759216308594,
+      "learning_rate": 1.8024010360971665e-05,
+      "loss": 0.3007,
+      "step": 2208
+    },
+    {
+      "epoch": 0.884,
+      "grad_norm": 5.907240390777588,
+      "learning_rate": 1.8007313709487345e-05,
+      "loss": 0.4164,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8848,
+      "grad_norm": 6.11362886428833,
+      "learning_rate": 1.7990554615362207e-05,
+      "loss": 0.2909,
+      "step": 2212
+    },
+    {
+      "epoch": 0.8856,
+      "grad_norm": 2.218219757080078,
+      "learning_rate": 1.7973733209287036e-05,
+      "loss": 0.3034,
+      "step": 2214
+    },
+    {
+      "epoch": 0.8864,
+      "grad_norm": 2.290370464324951,
+      "learning_rate": 1.7956849622438568e-05,
+      "loss": 0.2474,
+      "step": 2216
+    },
+    {
+      "epoch": 0.8872,
+      "grad_norm": 3.0256340503692627,
+      "learning_rate": 1.7939903986478357e-05,
+      "loss": 0.5,
+      "step": 2218
+    },
+    {
+      "epoch": 0.888,
+      "grad_norm": 7.33586311340332,
+      "learning_rate": 1.7922896433551913e-05,
+      "loss": 0.7149,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8888,
+      "grad_norm": 2.595695734024048,
+      "learning_rate": 1.7905827096287525e-05,
+      "loss": 0.24,
+      "step": 2222
+    },
+    {
+      "epoch": 0.8896,
+      "grad_norm": 4.993026256561279,
+      "learning_rate": 1.7888696107795347e-05,
+      "loss": 0.4084,
+      "step": 2224
+    },
+    {
+      "epoch": 0.8904,
+      "grad_norm": 13.632013320922852,
+      "learning_rate": 1.787150360166623e-05,
+      "loss": 0.5648,
+      "step": 2226
+    },
+    {
+      "epoch": 0.8912,
+      "grad_norm": 4.771726608276367,
+      "learning_rate": 1.7854249711970826e-05,
+      "loss": 0.3186,
+      "step": 2228
+    },
+    {
+      "epoch": 0.892,
+      "grad_norm": 1.7299787998199463,
+      "learning_rate": 1.783693457325841e-05,
+      "loss": 0.1849,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8928,
+      "grad_norm": 5.569767951965332,
+      "learning_rate": 1.7819558320555902e-05,
+      "loss": 0.5359,
+      "step": 2232
+    },
+    {
+      "epoch": 0.8936,
+      "grad_norm": 2.1846365928649902,
+      "learning_rate": 1.780212108936685e-05,
+      "loss": 0.2436,
+      "step": 2234
+    },
+    {
+      "epoch": 0.8944,
+      "grad_norm": 5.118314266204834,
+      "learning_rate": 1.7784623015670237e-05,
+      "loss": 0.4241,
+      "step": 2236
+    },
+    {
+      "epoch": 0.8952,
+      "grad_norm": 8.755192756652832,
+      "learning_rate": 1.7767064235919594e-05,
+      "loss": 0.5717,
+      "step": 2238
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 8.251045227050781,
+      "learning_rate": 1.77494448870418e-05,
+      "loss": 0.6775,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8968,
+      "grad_norm": 2.6442363262176514,
+      "learning_rate": 1.773176510643608e-05,
+      "loss": 0.3537,
+      "step": 2242
+    },
+    {
+      "epoch": 0.8976,
+      "grad_norm": 2.8393073081970215,
+      "learning_rate": 1.7714025031972894e-05,
+      "loss": 0.3711,
+      "step": 2244
+    },
+    {
+      "epoch": 0.8984,
+      "grad_norm": 4.766880035400391,
+      "learning_rate": 1.769622480199295e-05,
+      "loss": 0.3779,
+      "step": 2246
+    },
+    {
+      "epoch": 0.8992,
+      "grad_norm": 10.22143840789795,
+      "learning_rate": 1.7678364555305982e-05,
+      "loss": 0.5814,
+      "step": 2248
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.3465576171875,
+      "learning_rate": 1.7660444431189777e-05,
+      "loss": 0.4841,
+      "step": 2250
+    },
+    {
+      "epoch": 0.9008,
+      "grad_norm": 2.9356963634490967,
+      "learning_rate": 1.76424645693891e-05,
+      "loss": 0.4369,
+      "step": 2252
+    },
+    {
+      "epoch": 0.9016,
+      "grad_norm": 6.478868007659912,
+      "learning_rate": 1.762442511011448e-05,
+      "loss": 0.6658,
+      "step": 2254
+    },
+    {
+      "epoch": 0.9024,
+      "grad_norm": 4.515495777130127,
+      "learning_rate": 1.7606326194041285e-05,
+      "loss": 0.2635,
+      "step": 2256
+    },
+    {
+      "epoch": 0.9032,
+      "grad_norm": 3.4764370918273926,
+      "learning_rate": 1.7588167962308458e-05,
+      "loss": 0.3208,
+      "step": 2258
+    },
+    {
+      "epoch": 0.904,
+      "grad_norm": 4.561454772949219,
+      "learning_rate": 1.756995055651757e-05,
+      "loss": 0.5108,
+      "step": 2260
+    },
+    {
+      "epoch": 0.9048,
+      "grad_norm": 4.654769420623779,
+      "learning_rate": 1.7551674118731585e-05,
+      "loss": 0.3701,
+      "step": 2262
+    },
+    {
+      "epoch": 0.9056,
+      "grad_norm": 10.341426849365234,
+      "learning_rate": 1.7533338791473875e-05,
+      "loss": 0.6947,
+      "step": 2264
+    },
+    {
+      "epoch": 0.9064,
+      "grad_norm": 7.0824809074401855,
+      "learning_rate": 1.751494471772697e-05,
+      "loss": 0.4752,
+      "step": 2266
+    },
+    {
+      "epoch": 0.9072,
+      "grad_norm": 6.04065465927124,
+      "learning_rate": 1.7496492040931548e-05,
+      "loss": 0.3313,
+      "step": 2268
+    },
+    {
+      "epoch": 0.908,
+      "grad_norm": 3.093017578125,
+      "learning_rate": 1.747798090498533e-05,
+      "loss": 0.2161,
+      "step": 2270
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 3.0049962997436523,
+      "learning_rate": 1.745941145424182e-05,
+      "loss": 0.6934,
+      "step": 2272
+    },
+    {
+      "epoch": 0.9096,
+      "grad_norm": 2.6591107845306396,
+      "learning_rate": 1.744078383350938e-05,
+      "loss": 0.3383,
+      "step": 2274
+    },
+    {
+      "epoch": 0.9104,
+      "grad_norm": 9.022825241088867,
+      "learning_rate": 1.7422098188049885e-05,
+      "loss": 0.5028,
+      "step": 2276
+    },
+    {
+      "epoch": 0.9112,
+      "grad_norm": 7.651523590087891,
+      "learning_rate": 1.7403354663577782e-05,
+      "loss": 0.8057,
+      "step": 2278
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 5.598511219024658,
+      "learning_rate": 1.738455340625883e-05,
+      "loss": 0.7146,
+      "step": 2280
+    },
+    {
+      "epoch": 0.9128,
+      "grad_norm": 1.7681468725204468,
+      "learning_rate": 1.7365694562709038e-05,
+      "loss": 0.5604,
+      "step": 2282
+    },
+    {
+      "epoch": 0.9136,
+      "grad_norm": 4.347034454345703,
+      "learning_rate": 1.7346778279993433e-05,
+      "loss": 0.354,
+      "step": 2284
+    },
+    {
+      "epoch": 0.9144,
+      "grad_norm": 4.9559149742126465,
+      "learning_rate": 1.7327804705624962e-05,
+      "loss": 0.404,
+      "step": 2286
+    },
+    {
+      "epoch": 0.9152,
+      "grad_norm": 2.3942646980285645,
+      "learning_rate": 1.730877398756341e-05,
+      "loss": 0.6259,
+      "step": 2288
+    },
+    {
+      "epoch": 0.916,
+      "grad_norm": 6.6393351554870605,
+      "learning_rate": 1.7289686274214113e-05,
+      "loss": 0.3538,
+      "step": 2290
+    },
+    {
+      "epoch": 0.9168,
+      "grad_norm": 3.790372610092163,
+      "learning_rate": 1.727054171442693e-05,
+      "loss": 0.3411,
+      "step": 2292
+    },
+    {
+      "epoch": 0.9176,
+      "grad_norm": 4.417578220367432,
+      "learning_rate": 1.7251340457494934e-05,
+      "loss": 0.3735,
+      "step": 2294
+    },
+    {
+      "epoch": 0.9184,
+      "grad_norm": 2.050532579421997,
+      "learning_rate": 1.7232082653153422e-05,
+      "loss": 0.2166,
+      "step": 2296
+    },
+    {
+      "epoch": 0.9192,
+      "grad_norm": 4.665719032287598,
+      "learning_rate": 1.7212768451578595e-05,
+      "loss": 0.2797,
+      "step": 2298
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.98673415184021,
+      "learning_rate": 1.7193398003386517e-05,
+      "loss": 0.3617,
+      "step": 2300
+    },
+    {
+      "epoch": 0.9208,
+      "grad_norm": 3.666666030883789,
+      "learning_rate": 1.7173971459631803e-05,
+      "loss": 0.5871,
+      "step": 2302
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 4.225686550140381,
+      "learning_rate": 1.7154488971806525e-05,
+      "loss": 0.3627,
+      "step": 2304
+    },
+    {
+      "epoch": 0.9224,
+      "grad_norm": 2.3536264896392822,
+      "learning_rate": 1.713495069183907e-05,
+      "loss": 0.3145,
+      "step": 2306
+    },
+    {
+      "epoch": 0.9232,
+      "grad_norm": 2.9840025901794434,
+      "learning_rate": 1.7115356772092847e-05,
+      "loss": 0.5521,
+      "step": 2308
+    },
+    {
+      "epoch": 0.924,
+      "grad_norm": 3.0804269313812256,
+      "learning_rate": 1.709570736536522e-05,
+      "loss": 0.4238,
+      "step": 2310
+    },
+    {
+      "epoch": 0.9248,
+      "grad_norm": 2.455007791519165,
+      "learning_rate": 1.7076002624886152e-05,
+      "loss": 0.2472,
+      "step": 2312
+    },
+    {
+      "epoch": 0.9256,
+      "grad_norm": 7.137457847595215,
+      "learning_rate": 1.705624270431722e-05,
+      "loss": 0.6622,
+      "step": 2314
+    },
+    {
+      "epoch": 0.9264,
+      "grad_norm": 1.5967859029769897,
+      "learning_rate": 1.70364277577502e-05,
+      "loss": 0.1157,
+      "step": 2316
+    },
+    {
+      "epoch": 0.9272,
+      "grad_norm": 4.234429836273193,
+      "learning_rate": 1.7016557939706078e-05,
+      "loss": 0.3578,
+      "step": 2318
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 4.573144435882568,
+      "learning_rate": 1.6996633405133673e-05,
+      "loss": 0.288,
+      "step": 2320
+    },
+    {
+      "epoch": 0.9288,
+      "grad_norm": 10.306396484375,
+      "learning_rate": 1.6976654309408468e-05,
+      "loss": 0.6993,
+      "step": 2322
+    },
+    {
+      "epoch": 0.9296,
+      "grad_norm": 10.781827926635742,
+      "learning_rate": 1.6956620808331515e-05,
+      "loss": 0.735,
+      "step": 2324
+    },
+    {
+      "epoch": 0.9304,
+      "grad_norm": 3.1664340496063232,
+      "learning_rate": 1.6936533058128042e-05,
+      "loss": 0.4528,
+      "step": 2326
+    },
+    {
+      "epoch": 0.9312,
+      "grad_norm": 7.0673298835754395,
+      "learning_rate": 1.691639121544641e-05,
+      "loss": 0.3242,
+      "step": 2328
+    },
+    {
+      "epoch": 0.932,
+      "grad_norm": 8.26816463470459,
+      "learning_rate": 1.6896195437356696e-05,
+      "loss": 0.6315,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9328,
+      "grad_norm": 4.574904441833496,
+      "learning_rate": 1.6875945881349686e-05,
+      "loss": 0.4652,
+      "step": 2332
+    },
+    {
+      "epoch": 0.9336,
+      "grad_norm": 4.3775506019592285,
+      "learning_rate": 1.6855642705335435e-05,
+      "loss": 0.4898,
+      "step": 2334
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 4.384169578552246,
+      "learning_rate": 1.6835286067642228e-05,
+      "loss": 0.7099,
+      "step": 2336
+    },
+    {
+      "epoch": 0.9352,
+      "grad_norm": 4.223312854766846,
+      "learning_rate": 1.681487612701521e-05,
+      "loss": 0.5002,
+      "step": 2338
+    },
+    {
+      "epoch": 0.936,
+      "grad_norm": 5.141642093658447,
+      "learning_rate": 1.6794413042615168e-05,
+      "loss": 0.4201,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9368,
+      "grad_norm": 4.047835350036621,
+      "learning_rate": 1.677389697401739e-05,
+      "loss": 0.578,
+      "step": 2342
+    },
+    {
+      "epoch": 0.9376,
+      "grad_norm": 2.148810863494873,
+      "learning_rate": 1.675332808121025e-05,
+      "loss": 0.2867,
+      "step": 2344
+    },
+    {
+      "epoch": 0.9384,
+      "grad_norm": 6.967923164367676,
+      "learning_rate": 1.6732706524594145e-05,
+      "loss": 0.8902,
+      "step": 2346
+    },
+    {
+      "epoch": 0.9392,
+      "grad_norm": 9.913086891174316,
+      "learning_rate": 1.671203246498009e-05,
+      "loss": 0.4933,
+      "step": 2348
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1417829990386963,
+      "learning_rate": 1.6691306063588593e-05,
+      "loss": 0.3506,
+      "step": 2350
+    },
+    {
+      "epoch": 0.9408,
+      "grad_norm": 14.28138256072998,
+      "learning_rate": 1.6670527482048242e-05,
+      "loss": 1.1406,
+      "step": 2352
+    },
+    {
+      "epoch": 0.9416,
+      "grad_norm": 2.6398777961730957,
+      "learning_rate": 1.6649696882394635e-05,
+      "loss": 0.2026,
+      "step": 2354
+    },
+    {
+      "epoch": 0.9424,
+      "grad_norm": 3.2635467052459717,
+      "learning_rate": 1.6628814427068968e-05,
+      "loss": 0.7803,
+      "step": 2356
+    },
+    {
+      "epoch": 0.9432,
+      "grad_norm": 2.1623849868774414,
+      "learning_rate": 1.6607880278916778e-05,
+      "loss": 0.2551,
+      "step": 2358
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 2.4618899822235107,
+      "learning_rate": 1.6586894601186824e-05,
+      "loss": 0.4137,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9448,
+      "grad_norm": 2.293994903564453,
+      "learning_rate": 1.656585755752957e-05,
+      "loss": 0.2418,
+      "step": 2362
+    },
+    {
+      "epoch": 0.9456,
+      "grad_norm": 2.1502673625946045,
+      "learning_rate": 1.6544769311996153e-05,
+      "loss": 0.3008,
+      "step": 2364
+    },
+    {
+      "epoch": 0.9464,
+      "grad_norm": 2.998347520828247,
+      "learning_rate": 1.6523630029036924e-05,
+      "loss": 0.4707,
+      "step": 2366
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 5.802120208740234,
+      "learning_rate": 1.6502439873500294e-05,
+      "loss": 0.4726,
+      "step": 2368
+    },
+    {
+      "epoch": 0.948,
+      "grad_norm": 11.09598159790039,
+      "learning_rate": 1.6481199010631305e-05,
+      "loss": 0.5139,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9488,
+      "grad_norm": 5.688537120819092,
+      "learning_rate": 1.645990760607052e-05,
+      "loss": 0.3321,
+      "step": 2372
+    },
+    {
+      "epoch": 0.9496,
+      "grad_norm": 1.9983233213424683,
+      "learning_rate": 1.643856582585255e-05,
+      "loss": 0.394,
+      "step": 2374
+    },
+    {
+      "epoch": 0.9504,
+      "grad_norm": 6.963872909545898,
+      "learning_rate": 1.641717383640488e-05,
+      "loss": 0.3209,
+      "step": 2376
+    },
+    {
+      "epoch": 0.9512,
+      "grad_norm": 2.2653205394744873,
+      "learning_rate": 1.6395731804546596e-05,
+      "loss": 0.1529,
+      "step": 2378
+    },
+    {
+      "epoch": 0.952,
+      "grad_norm": 3.29524564743042,
+      "learning_rate": 1.63742398974869e-05,
+      "loss": 0.3794,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9528,
+      "grad_norm": 2.852288007736206,
+      "learning_rate": 1.6352698282824045e-05,
+      "loss": 0.2458,
+      "step": 2382
+    },
+    {
+      "epoch": 0.9536,
+      "grad_norm": 2.9293675422668457,
+      "learning_rate": 1.633110712854385e-05,
+      "loss": 0.3333,
+      "step": 2384
+    },
+    {
+      "epoch": 0.9544,
+      "grad_norm": 10.415287017822266,
+      "learning_rate": 1.6309466603018504e-05,
+      "loss": 0.5087,
+      "step": 2386
+    },
+    {
+      "epoch": 0.9552,
+      "grad_norm": 1.27485191822052,
+      "learning_rate": 1.6287776875005148e-05,
+      "loss": 0.2743,
+      "step": 2388
+    },
+    {
+      "epoch": 0.956,
+      "grad_norm": 2.6245243549346924,
+      "learning_rate": 1.6266038113644612e-05,
+      "loss": 0.516,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9568,
+      "grad_norm": 4.452570915222168,
+      "learning_rate": 1.624425048846017e-05,
+      "loss": 0.4682,
+      "step": 2392
+    },
+    {
+      "epoch": 0.9576,
+      "grad_norm": 2.626039743423462,
+      "learning_rate": 1.6222414169356063e-05,
+      "loss": 0.2863,
+      "step": 2394
+    },
+    {
+      "epoch": 0.9584,
+      "grad_norm": 2.5585098266601562,
+      "learning_rate": 1.6200529326616343e-05,
+      "loss": 0.4538,
+      "step": 2396
+    },
+    {
+      "epoch": 0.9592,
+      "grad_norm": 5.5422539710998535,
+      "learning_rate": 1.6178596130903345e-05,
+      "loss": 0.4687,
+      "step": 2398
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.677401542663574,
+      "learning_rate": 1.6156614753256587e-05,
+      "loss": 0.2982,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9608,
+      "grad_norm": 3.8012046813964844,
+      "learning_rate": 1.613458536509123e-05,
+      "loss": 0.2614,
+      "step": 2402
+    },
+    {
+      "epoch": 0.9616,
+      "grad_norm": 2.2905969619750977,
+      "learning_rate": 1.6112508138196922e-05,
+      "loss": 0.7023,
+      "step": 2404
+    },
+    {
+      "epoch": 0.9624,
+      "grad_norm": 2.1353418827056885,
+      "learning_rate": 1.6090383244736277e-05,
+      "loss": 0.423,
+      "step": 2406
+    },
+    {
+      "epoch": 0.9632,
+      "grad_norm": 1.8343054056167603,
+      "learning_rate": 1.606821085724363e-05,
+      "loss": 0.4421,
+      "step": 2408
+    },
+    {
+      "epoch": 0.964,
+      "grad_norm": 2.907493829727173,
+      "learning_rate": 1.6045991148623756e-05,
+      "loss": 0.3788,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9648,
+      "grad_norm": 7.708912372589111,
+      "learning_rate": 1.602372429215038e-05,
+      "loss": 0.556,
+      "step": 2412
+    },
+    {
+      "epoch": 0.9656,
+      "grad_norm": 2.3259804248809814,
+      "learning_rate": 1.600141046146497e-05,
+      "loss": 0.2895,
+      "step": 2414
+    },
+    {
+      "epoch": 0.9664,
+      "grad_norm": 4.717739105224609,
+      "learning_rate": 1.597904983057519e-05,
+      "loss": 0.8317,
+      "step": 2416
+    },
+    {
+      "epoch": 0.9672,
+      "grad_norm": 3.4694042205810547,
+      "learning_rate": 1.5956642573853794e-05,
+      "loss": 0.2237,
+      "step": 2418
+    },
+    {
+      "epoch": 0.968,
+      "grad_norm": 6.2816033363342285,
+      "learning_rate": 1.5934188866037014e-05,
+      "loss": 0.2499,
+      "step": 2420
+    },
+    {
+      "epoch": 0.9688,
+      "grad_norm": 3.9102213382720947,
+      "learning_rate": 1.591168888222342e-05,
+      "loss": 0.4237,
+      "step": 2422
+    },
+    {
+      "epoch": 0.9696,
+      "grad_norm": 9.37143611907959,
+      "learning_rate": 1.5889142797872407e-05,
+      "loss": 0.5436,
+      "step": 2424
+    },
+    {
+      "epoch": 0.9704,
+      "grad_norm": 2.8910293579101562,
+      "learning_rate": 1.5866550788802818e-05,
+      "loss": 0.3519,
+      "step": 2426
+    },
+    {
+      "epoch": 0.9712,
+      "grad_norm": 3.6150565147399902,
+      "learning_rate": 1.584391303119173e-05,
+      "loss": 0.1368,
+      "step": 2428
+    },
+    {
+      "epoch": 0.972,
+      "grad_norm": 7.551123142242432,
+      "learning_rate": 1.582122970157289e-05,
+      "loss": 0.4774,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 8.42343807220459,
+      "learning_rate": 1.5798500976835503e-05,
+      "loss": 0.7228,
+      "step": 2432
+    },
+    {
+      "epoch": 0.9736,
+      "grad_norm": 5.844367027282715,
+      "learning_rate": 1.577572703422267e-05,
+      "loss": 0.3905,
+      "step": 2434
+    },
+    {
+      "epoch": 0.9744,
+      "grad_norm": 4.236822605133057,
+      "learning_rate": 1.575290805133024e-05,
+      "loss": 0.5367,
+      "step": 2436
+    },
+    {
+      "epoch": 0.9752,
+      "grad_norm": 5.707516670227051,
+      "learning_rate": 1.5730044206105156e-05,
+      "loss": 1.0149,
+      "step": 2438
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 11.354463577270508,
+      "learning_rate": 1.570713567684432e-05,
+      "loss": 0.8809,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9768,
+      "grad_norm": 4.4314446449279785,
+      "learning_rate": 1.5684182642193047e-05,
+      "loss": 0.7045,
+      "step": 2442
+    },
+    {
+      "epoch": 0.9776,
+      "grad_norm": 2.725847005844116,
+      "learning_rate": 1.566118528114367e-05,
+      "loss": 0.2321,
+      "step": 2444
+    },
+    {
+      "epoch": 0.9784,
+      "grad_norm": 3.7199370861053467,
+      "learning_rate": 1.563814377303429e-05,
+      "loss": 0.4067,
+      "step": 2446
+    },
+    {
+      "epoch": 0.9792,
+      "grad_norm": 3.164414167404175,
+      "learning_rate": 1.561505829754715e-05,
+      "loss": 0.3257,
+      "step": 2448
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3418188095092773,
+      "learning_rate": 1.5591929034707475e-05,
+      "loss": 0.4852,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9808,
+      "grad_norm": 2.900113582611084,
+      "learning_rate": 1.5568756164881874e-05,
+      "loss": 0.3542,
+      "step": 2452
+    },
+    {
+      "epoch": 0.9816,
+      "grad_norm": 1.9147764444351196,
+      "learning_rate": 1.5545539868777085e-05,
+      "loss": 0.2348,
+      "step": 2454
+    },
+    {
+      "epoch": 0.9824,
+      "grad_norm": 4.320040702819824,
+      "learning_rate": 1.5522280327438384e-05,
+      "loss": 0.5722,
+      "step": 2456
+    },
+    {
+      "epoch": 0.9832,
+      "grad_norm": 2.6387248039245605,
+      "learning_rate": 1.5498977722248398e-05,
+      "loss": 0.3449,
+      "step": 2458
+    },
+    {
+      "epoch": 0.984,
+      "grad_norm": 1.3307088613510132,
+      "learning_rate": 1.547563223492552e-05,
+      "loss": 0.3057,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9848,
+      "grad_norm": 3.4244158267974854,
+      "learning_rate": 1.5452244047522504e-05,
+      "loss": 0.4097,
+      "step": 2462
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 3.6758108139038086,
+      "learning_rate": 1.5428813342425194e-05,
+      "loss": 0.4778,
+      "step": 2464
+    },
+    {
+      "epoch": 0.9864,
+      "grad_norm": 26.934717178344727,
+      "learning_rate": 1.5405340302350876e-05,
+      "loss": 1.4515,
+      "step": 2466
+    },
+    {
+      "epoch": 0.9872,
+      "grad_norm": 5.977929592132568,
+      "learning_rate": 1.538182511034708e-05,
+      "loss": 0.4473,
+      "step": 2468
+    },
+    {
+      "epoch": 0.988,
+      "grad_norm": 1.2147676944732666,
+      "learning_rate": 1.535826794978996e-05,
+      "loss": 0.3552,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9888,
+      "grad_norm": 5.006214618682861,
+      "learning_rate": 1.5334669004383036e-05,
+      "loss": 0.9009,
+      "step": 2472
+    },
+    {
+      "epoch": 0.9896,
+      "grad_norm": 8.950480461120605,
+      "learning_rate": 1.5311028458155564e-05,
+      "loss": 0.9034,
+      "step": 2474
+    },
+    {
+      "epoch": 0.9904,
+      "grad_norm": 3.063087224960327,
+      "learning_rate": 1.528734649546133e-05,
+      "loss": 0.4858,
+      "step": 2476
+    },
+    {
+      "epoch": 0.9912,
+      "grad_norm": 9.52176284790039,
+      "learning_rate": 1.5263623300976997e-05,
+      "loss": 0.4892,
+      "step": 2478
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 1.4161584377288818,
+      "learning_rate": 1.5239859059700792e-05,
+      "loss": 0.2002,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9928,
+      "grad_norm": 5.208809852600098,
+      "learning_rate": 1.5216053956951096e-05,
+      "loss": 0.5075,
+      "step": 2482
+    },
+    {
+      "epoch": 0.9936,
+      "grad_norm": 5.600852966308594,
+      "learning_rate": 1.5192208178364819e-05,
+      "loss": 0.5732,
+      "step": 2484
+    },
+    {
+      "epoch": 0.9944,
+      "grad_norm": 3.0066847801208496,
+      "learning_rate": 1.5168321909896176e-05,
+      "loss": 0.331,
+      "step": 2486
+    },
+    {
+      "epoch": 0.9952,
+      "grad_norm": 3.0223422050476074,
+      "learning_rate": 1.5144395337815057e-05,
+      "loss": 0.2142,
+      "step": 2488
+    },
+    {
+      "epoch": 0.996,
+      "grad_norm": 2.803382635116577,
+      "learning_rate": 1.5120428648705722e-05,
+      "loss": 0.2565,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9968,
+      "grad_norm": 7.8674445152282715,
+      "learning_rate": 1.5096422029465171e-05,
+      "loss": 0.6374,
+      "step": 2492
+    },
+    {
+      "epoch": 0.9976,
+      "grad_norm": 5.142702579498291,
+      "learning_rate": 1.5072375667301904e-05,
+      "loss": 0.414,
+      "step": 2494
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 5.461601734161377,
+      "learning_rate": 1.5048289749734231e-05,
+      "loss": 0.4925,
+      "step": 2496
+    },
+    {
+      "epoch": 0.9992,
+      "grad_norm": 7.165497303009033,
+      "learning_rate": 1.502416446458898e-05,
+      "loss": 0.496,
+      "step": 2498
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.7376537322998047,
+      "learning_rate": 1.5000000000000014e-05,
+      "loss": 0.3629,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0,
+      "step": 2500,
+      "total_flos": 1.0020465781768192e+16,
+      "train_loss": 0.5193727528691292,
+      "train_runtime": 17799.0301,
+      "train_samples_per_second": 2.247,
+      "train_steps_per_second": 0.14
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 2500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {},
+  "total_flos": 1.0020465781768192e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/client_states_v9_NEURIPS_DISJOINT_Memonly_LORA_llava_lr2e-5_bs1_gradacc32_iter0_25_gradnorm_scenario18_new_10000_random0_0625_seed1/server_model_round0.pth b/client_states_v9_NEURIPS_DISJOINT_Memonly_LORA_llava_lr2e-5_bs1_gradacc32_iter0_25_gradnorm_scenario18_new_10000_random0_0625_seed1/server_model_round0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..0f5bc49911711b3fc42970112cdf1d9bb1d4315a
--- /dev/null
+++ b/client_states_v9_NEURIPS_DISJOINT_Memonly_LORA_llava_lr2e-5_bs1_gradacc32_iter0_25_gradnorm_scenario18_new_10000_random0_0625_seed1/server_model_round0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0b639778d4e1f7aae966d0b330009ecbda922dd8869a6613e0eda2954c0915b
+size 639793378
diff --git a/client_states_v9_NEURIPS_DISJOINT_Memonly_LORA_llava_lr2e-5_bs1_gradacc32_iter0_25_gradnorm_scenario18_new_10000_random0_0625_seed1/server_model_round1.pth b/client_states_v9_NEURIPS_DISJOINT_Memonly_LORA_llava_lr2e-5_bs1_gradacc32_iter0_25_gradnorm_scenario18_new_10000_random0_0625_seed1/server_model_round1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..b511825b8b9e3b1dac247f297e0874c999b89776
--- /dev/null
+++ b/client_states_v9_NEURIPS_DISJOINT_Memonly_LORA_llava_lr2e-5_bs1_gradacc32_iter0_25_gradnorm_scenario18_new_10000_random0_0625_seed1/server_model_round1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e08f705eab7c3e575cee2a2e93a8f7e06e7514bce97e6eb4353b7448c6472262
+size 639793378
diff --git a/client_states_v9_NEURIPS_DISJOINT_Memonly_LORA_llava_lr2e-5_bs1_gradacc32_iter0_25_gradnorm_scenario18_new_10000_random0_0625_seed1/server_model_round2.pth b/client_states_v9_NEURIPS_DISJOINT_Memonly_LORA_llava_lr2e-5_bs1_gradacc32_iter0_25_gradnorm_scenario18_new_10000_random0_0625_seed1/server_model_round2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..9a820d6dc33485218ef15ad29b849dcc17e3924c
--- /dev/null
+++ b/client_states_v9_NEURIPS_DISJOINT_Memonly_LORA_llava_lr2e-5_bs1_gradacc32_iter0_25_gradnorm_scenario18_new_10000_random0_0625_seed1/server_model_round2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e535c21196e01af3ba8bd5b2d6cad8060cb4f93528c909e6610c444f3678e52b
+size 639793378
diff --git a/client_states_v9_NEURIPS_DISJOINT_Memonly_LORA_llava_lr2e-5_bs1_gradacc32_iter0_25_gradnorm_scenario18_new_10000_random0_0625_seed1/server_model_round3.pth b/client_states_v9_NEURIPS_DISJOINT_Memonly_LORA_llava_lr2e-5_bs1_gradacc32_iter0_25_gradnorm_scenario18_new_10000_random0_0625_seed1/server_model_round3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..f4ac6d9874913fa17651737eacc2204b1e2ee761
--- /dev/null
+++ b/client_states_v9_NEURIPS_DISJOINT_Memonly_LORA_llava_lr2e-5_bs1_gradacc32_iter0_25_gradnorm_scenario18_new_10000_random0_0625_seed1/server_model_round3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bde55ce76a5fd1470d6214ecfa14613869e50b34f417538b17eae2e496c9f7c6
+size 639793378