diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,3886 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1098,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00546448087431694,
+      "grad_norm": 1.0751630067825317,
+      "learning_rate": 5.454545454545455e-07,
+      "loss": 2.654850959777832,
+      "step": 2
+    },
+    {
+      "epoch": 0.01092896174863388,
+      "grad_norm": 0.6832827925682068,
+      "learning_rate": 1.6363636363636363e-06,
+      "loss": 1.9340370893478394,
+      "step": 4
+    },
+    {
+      "epoch": 0.01639344262295082,
+      "grad_norm": 0.44048377871513367,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 1.8661857843399048,
+      "step": 6
+    },
+    {
+      "epoch": 0.02185792349726776,
+      "grad_norm": 0.6479131579399109,
+      "learning_rate": 3.818181818181818e-06,
+      "loss": 1.7092058658599854,
+      "step": 8
+    },
+    {
+      "epoch": 0.0273224043715847,
+      "grad_norm": 0.4293830096721649,
+      "learning_rate": 4.90909090909091e-06,
+      "loss": 1.6882967948913574,
+      "step": 10
+    },
+    {
+      "epoch": 0.03278688524590164,
+      "grad_norm": 0.36752375960350037,
+      "learning_rate": 6e-06,
+      "loss": 1.5482172966003418,
+      "step": 12
+    },
+    {
+      "epoch": 0.03825136612021858,
+      "grad_norm": 0.45134979486465454,
+      "learning_rate": 7.090909090909091e-06,
+      "loss": 1.4477657079696655,
+      "step": 14
+    },
+    {
+      "epoch": 0.04371584699453552,
+      "grad_norm": 1.1161218881607056,
+      "learning_rate": 8.181818181818181e-06,
+      "loss": 1.2040513753890991,
+      "step": 16
+    },
+    {
+      "epoch": 0.04918032786885246,
+      "grad_norm": 0.4559285044670105,
+      "learning_rate": 9.272727272727273e-06,
+      "loss": 1.1814911365509033,
+      "step": 18
+    },
+    {
+      "epoch": 0.0546448087431694,
+      "grad_norm": 0.3014647364616394,
+      "learning_rate": 1.0363636363636364e-05,
+      "loss": 1.4038846492767334,
+      "step": 20
+    },
+    {
+      "epoch": 0.060109289617486336,
+      "grad_norm": 0.2782735228538513,
+      "learning_rate": 1.1454545454545455e-05,
+      "loss": 1.331674575805664,
+      "step": 22
+    },
+    {
+      "epoch": 0.06557377049180328,
+      "grad_norm": 0.2593174874782562,
+      "learning_rate": 1.2545454545454545e-05,
+      "loss": 1.39420485496521,
+      "step": 24
+    },
+    {
+      "epoch": 0.07103825136612021,
+      "grad_norm": 0.7720677256584167,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 1.5995798110961914,
+      "step": 26
+    },
+    {
+      "epoch": 0.07650273224043716,
+      "grad_norm": 0.2457510083913803,
+      "learning_rate": 1.4727272727272728e-05,
+      "loss": 1.3586376905441284,
+      "step": 28
+    },
+    {
+      "epoch": 0.08196721311475409,
+      "grad_norm": 0.333055704832077,
+      "learning_rate": 1.5818181818181818e-05,
+      "loss": 1.2947840690612793,
+      "step": 30
+    },
+    {
+      "epoch": 0.08743169398907104,
+      "grad_norm": 0.3230854570865631,
+      "learning_rate": 1.6909090909090907e-05,
+      "loss": 1.2872313261032104,
+      "step": 32
+    },
+    {
+      "epoch": 0.09289617486338798,
+      "grad_norm": 0.3659067153930664,
+      "learning_rate": 1.8e-05,
+      "loss": 1.013819694519043,
+      "step": 34
+    },
+    {
+      "epoch": 0.09836065573770492,
+      "grad_norm": 0.19767461717128754,
+      "learning_rate": 1.909090909090909e-05,
+      "loss": 1.1190152168273926,
+      "step": 36
+    },
+    {
+      "epoch": 0.10382513661202186,
+      "grad_norm": 0.3944986164569855,
+      "learning_rate": 2.0181818181818183e-05,
+      "loss": 1.3628164529800415,
+      "step": 38
+    },
+    {
+      "epoch": 0.1092896174863388,
+      "grad_norm": 0.5476788878440857,
+      "learning_rate": 2.1272727272727273e-05,
+      "loss": 1.3132535219192505,
+      "step": 40
+    },
+    {
+      "epoch": 0.11475409836065574,
+      "grad_norm": 0.6100736260414124,
+      "learning_rate": 2.2363636363636366e-05,
+      "loss": 1.0027183294296265,
+      "step": 42
+    },
+    {
+      "epoch": 0.12021857923497267,
+      "grad_norm": 0.32569247484207153,
+      "learning_rate": 2.3454545454545456e-05,
+      "loss": 1.01920747756958,
+      "step": 44
+    },
+    {
+      "epoch": 0.12568306010928962,
+      "grad_norm": 0.2778260111808777,
+      "learning_rate": 2.454545454545455e-05,
+      "loss": 1.3417850732803345,
+      "step": 46
+    },
+    {
+      "epoch": 0.13114754098360656,
+      "grad_norm": 0.2733558416366577,
+      "learning_rate": 2.5636363636363635e-05,
+      "loss": 1.32207453250885,
+      "step": 48
+    },
+    {
+      "epoch": 0.1366120218579235,
+      "grad_norm": 0.28408652544021606,
+      "learning_rate": 2.6727272727272728e-05,
+      "loss": 1.2943894863128662,
+      "step": 50
+    },
+    {
+      "epoch": 0.14207650273224043,
+      "grad_norm": 0.2375349998474121,
+      "learning_rate": 2.7818181818181818e-05,
+      "loss": 1.2506234645843506,
+      "step": 52
+    },
+    {
+      "epoch": 0.14754098360655737,
+      "grad_norm": 1.0531994104385376,
+      "learning_rate": 2.890909090909091e-05,
+      "loss": 0.8835932612419128,
+      "step": 54
+    },
+    {
+      "epoch": 0.15300546448087432,
+      "grad_norm": 0.3646933436393738,
+      "learning_rate": 3e-05,
+      "loss": 1.4430650472640991,
+      "step": 56
+    },
+    {
+      "epoch": 0.15846994535519127,
+      "grad_norm": 0.44717085361480713,
+      "learning_rate": 2.9997491688899256e-05,
+      "loss": 1.4402170181274414,
+      "step": 58
+    },
+    {
+      "epoch": 0.16393442622950818,
+      "grad_norm": 0.25877314805984497,
+      "learning_rate": 2.998996768768956e-05,
+      "loss": 1.3341705799102783,
+      "step": 60
+    },
+    {
+      "epoch": 0.16939890710382513,
+      "grad_norm": 0.3694762587547302,
+      "learning_rate": 2.9977430792302124e-05,
+      "loss": 1.5157015323638916,
+      "step": 62
+    },
+    {
+      "epoch": 0.17486338797814208,
+      "grad_norm": 0.7582504749298096,
+      "learning_rate": 2.9959885661467903e-05,
+      "loss": 1.2413986921310425,
+      "step": 64
+    },
+    {
+      "epoch": 0.18032786885245902,
+      "grad_norm": 0.7258033752441406,
+      "learning_rate": 2.993733881498636e-05,
+      "loss": 1.328322410583496,
+      "step": 66
+    },
+    {
+      "epoch": 0.18579234972677597,
+      "grad_norm": 0.4196822941303253,
+      "learning_rate": 2.9909798631302736e-05,
+      "loss": 1.3006751537322998,
+      "step": 68
+    },
+    {
+      "epoch": 0.1912568306010929,
+      "grad_norm": 0.8249898552894592,
+      "learning_rate": 2.987727534439457e-05,
+      "loss": 1.3277437686920166,
+      "step": 70
+    },
+    {
+      "epoch": 0.19672131147540983,
+      "grad_norm": 0.4432520568370819,
+      "learning_rate": 2.983978103996877e-05,
+      "loss": 1.1078237295150757,
+      "step": 72
+    },
+    {
+      "epoch": 0.20218579234972678,
+      "grad_norm": 0.557328999042511,
+      "learning_rate": 2.9797329650970525e-05,
+      "loss": 1.1530853509902954,
+      "step": 74
+    },
+    {
+      "epoch": 0.20765027322404372,
+      "grad_norm": 0.9870507121086121,
+      "learning_rate": 2.974993695240579e-05,
+      "loss": 1.3453693389892578,
+      "step": 76
+    },
+    {
+      "epoch": 0.21311475409836064,
+      "grad_norm": 1.0352654457092285,
+      "learning_rate": 2.9697620555479297e-05,
+      "loss": 0.6454078555107117,
+      "step": 78
+    },
+    {
+      "epoch": 0.2185792349726776,
+      "grad_norm": 0.39752116799354553,
+      "learning_rate": 2.9640399901050182e-05,
+      "loss": 0.6325056552886963,
+      "step": 80
+    },
+    {
+      "epoch": 0.22404371584699453,
+      "grad_norm": 0.3930615782737732,
+      "learning_rate": 2.9578296252407734e-05,
+      "loss": 1.3635072708129883,
+      "step": 82
+    },
+    {
+      "epoch": 0.22950819672131148,
+      "grad_norm": 0.3616217374801636,
+      "learning_rate": 2.9511332687369917e-05,
+      "loss": 0.8247113227844238,
+      "step": 84
+    },
+    {
+      "epoch": 0.23497267759562843,
+      "grad_norm": 0.6772981286048889,
+      "learning_rate": 2.9439534089707624e-05,
+      "loss": 1.3125399351119995,
+      "step": 86
+    },
+    {
+      "epoch": 0.24043715846994534,
+      "grad_norm": 0.3286456763744354,
+      "learning_rate": 2.9362927139897832e-05,
+      "loss": 1.272586464881897,
+      "step": 88
+    },
+    {
+      "epoch": 0.2459016393442623,
+      "grad_norm": 0.8590133190155029,
+      "learning_rate": 2.9281540305209068e-05,
+      "loss": 1.0977256298065186,
+      "step": 90
+    },
+    {
+      "epoch": 0.25136612021857924,
+      "grad_norm": 0.780782163143158,
+      "learning_rate": 2.919540382912294e-05,
+      "loss": 1.323669195175171,
+      "step": 92
+    },
+    {
+      "epoch": 0.2568306010928962,
+      "grad_norm": 0.2833038866519928,
+      "learning_rate": 2.9104549720095634e-05,
+      "loss": 1.1723302602767944,
+      "step": 94
+    },
+    {
+      "epoch": 0.26229508196721313,
+      "grad_norm": 0.4925417900085449,
+      "learning_rate": 2.9009011739663467e-05,
+      "loss": 1.5599980354309082,
+      "step": 96
+    },
+    {
+      "epoch": 0.2677595628415301,
+      "grad_norm": 0.5641136169433594,
+      "learning_rate": 2.8908825389897094e-05,
+      "loss": 1.2302381992340088,
+      "step": 98
+    },
+    {
+      "epoch": 0.273224043715847,
+      "grad_norm": 0.3122362792491913,
+      "learning_rate": 2.8804027900208843e-05,
+      "loss": 1.1082335710525513,
+      "step": 100
+    },
+    {
+      "epoch": 0.2786885245901639,
+      "grad_norm": 0.3353098928928375,
+      "learning_rate": 2.8694658213518226e-05,
+      "loss": 1.2942110300064087,
+      "step": 102
+    },
+    {
+      "epoch": 0.28415300546448086,
+      "grad_norm": 0.933445155620575,
+      "learning_rate": 2.8580756971780686e-05,
+      "loss": 0.9880049824714661,
+      "step": 104
+    },
+    {
+      "epoch": 0.2896174863387978,
+      "grad_norm": 0.27208322286605835,
+      "learning_rate": 2.846236650088497e-05,
+      "loss": 1.2753208875656128,
+      "step": 106
+    },
+    {
+      "epoch": 0.29508196721311475,
+      "grad_norm": 0.2629159390926361,
+      "learning_rate": 2.833953079492476e-05,
+      "loss": 1.410078525543213,
+      "step": 108
+    },
+    {
+      "epoch": 0.3005464480874317,
+      "grad_norm": 0.756142795085907,
+      "learning_rate": 2.82122954998504e-05,
+      "loss": 1.249933123588562,
+      "step": 110
+    },
+    {
+      "epoch": 0.30601092896174864,
+      "grad_norm": 0.3108830451965332,
+      "learning_rate": 2.808070789650679e-05,
+      "loss": 1.2787212133407593,
+      "step": 112
+    },
+    {
+      "epoch": 0.3114754098360656,
+      "grad_norm": 0.8976105451583862,
+      "learning_rate": 2.7944816883063727e-05,
+      "loss": 1.3811240196228027,
+      "step": 114
+    },
+    {
+      "epoch": 0.31693989071038253,
+      "grad_norm": 0.48146891593933105,
+      "learning_rate": 2.7804672956845295e-05,
+      "loss": 1.2635105848312378,
+      "step": 116
+    },
+    {
+      "epoch": 0.3224043715846995,
+      "grad_norm": 0.3662451505661011,
+      "learning_rate": 2.766032819556495e-05,
+      "loss": 1.0110238790512085,
+      "step": 118
+    },
+    {
+      "epoch": 0.32786885245901637,
+      "grad_norm": 1.0501397848129272,
+      "learning_rate": 2.7511836237973366e-05,
+      "loss": 1.2500156164169312,
+      "step": 120
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 1.6555485725402832,
+      "learning_rate": 2.735925226392618e-05,
+      "loss": 1.3619412183761597,
+      "step": 122
+    },
+    {
+      "epoch": 0.33879781420765026,
+      "grad_norm": 0.3158622682094574,
+      "learning_rate": 2.7202632973879086e-05,
+      "loss": 1.0549049377441406,
+      "step": 124
+    },
+    {
+      "epoch": 0.3442622950819672,
+      "grad_norm": 1.4328356981277466,
+      "learning_rate": 2.7042036567817838e-05,
+      "loss": 1.0704551935195923,
+      "step": 126
+    },
+    {
+      "epoch": 0.34972677595628415,
+      "grad_norm": 0.3114672005176544,
+      "learning_rate": 2.6877522723631036e-05,
+      "loss": 1.324272632598877,
+      "step": 128
+    },
+    {
+      "epoch": 0.3551912568306011,
+      "grad_norm": 0.41898471117019653,
+      "learning_rate": 2.6709152574933727e-05,
+      "loss": 1.2493348121643066,
+      "step": 130
+    },
+    {
+      "epoch": 0.36065573770491804,
+      "grad_norm": 0.28673115372657776,
+      "learning_rate": 2.6536988688350067e-05,
+      "loss": 1.1967706680297852,
+      "step": 132
+    },
+    {
+      "epoch": 0.366120218579235,
+      "grad_norm": 1.0675979852676392,
+      "learning_rate": 2.6361095040263437e-05,
+      "loss": 1.0794986486434937,
+      "step": 134
+    },
+    {
+      "epoch": 0.37158469945355194,
+      "grad_norm": 0.4790109097957611,
+      "learning_rate": 2.618153699304274e-05,
+      "loss": 1.0123281478881836,
+      "step": 136
+    },
+    {
+      "epoch": 0.3770491803278688,
+      "grad_norm": 0.3467015326023102,
+      "learning_rate": 2.599838127075361e-05,
+      "loss": 0.9970463514328003,
+      "step": 138
+    },
+    {
+      "epoch": 0.3825136612021858,
+      "grad_norm": 0.2335558384656906,
+      "learning_rate": 2.5811695934363666e-05,
+      "loss": 0.640708863735199,
+      "step": 140
+    },
+    {
+      "epoch": 0.3879781420765027,
+      "grad_norm": 0.4442676603794098,
+      "learning_rate": 2.5621550356450914e-05,
+      "loss": 1.2411315441131592,
+      "step": 142
+    },
+    {
+      "epoch": 0.39344262295081966,
+      "grad_norm": 0.2910548150539398,
+      "learning_rate": 2.5428015195424825e-05,
+      "loss": 1.2713288068771362,
+      "step": 144
+    },
+    {
+      "epoch": 0.3989071038251366,
+      "grad_norm": 0.38326823711395264,
+      "learning_rate": 2.5231162369269498e-05,
+      "loss": 1.225158929824829,
+      "step": 146
+    },
+    {
+      "epoch": 0.40437158469945356,
+      "grad_norm": 0.9609261751174927,
+      "learning_rate": 2.503106502881889e-05,
+      "loss": 1.216736078262329,
+      "step": 148
+    },
+    {
+      "epoch": 0.4098360655737705,
+      "grad_norm": 0.4641711115837097,
+      "learning_rate": 2.4827797530573762e-05,
+      "loss": 1.1474337577819824,
+      "step": 150
+    },
+    {
+      "epoch": 0.41530054644808745,
+      "grad_norm": 0.2951766550540924,
+      "learning_rate": 2.4621435409070757e-05,
+      "loss": 1.2647650241851807,
+      "step": 152
+    },
+    {
+      "epoch": 0.4207650273224044,
+      "grad_norm": 0.32328906655311584,
+      "learning_rate": 2.4412055348813602e-05,
+      "loss": 1.2341688871383667,
+      "step": 154
+    },
+    {
+      "epoch": 0.4262295081967213,
+      "grad_norm": 0.4336468279361725,
+      "learning_rate": 2.4199735155777017e-05,
+      "loss": 1.2835301160812378,
+      "step": 156
+    },
+    {
+      "epoch": 0.43169398907103823,
+      "grad_norm": 0.4230714738368988,
+      "learning_rate": 2.3984553728493914e-05,
+      "loss": 1.0034432411193848,
+      "step": 158
+    },
+    {
+      "epoch": 0.4371584699453552,
+      "grad_norm": 0.2932489812374115,
+      "learning_rate": 2.3766591028736547e-05,
+      "loss": 1.213523268699646,
+      "step": 160
+    },
+    {
+      "epoch": 0.4426229508196721,
+      "grad_norm": 3.7618532180786133,
+      "learning_rate": 2.3545928051802588e-05,
+      "loss": 0.6747933030128479,
+      "step": 162
+    },
+    {
+      "epoch": 0.44808743169398907,
+      "grad_norm": 0.7841958403587341,
+      "learning_rate": 2.332264679641717e-05,
+      "loss": 1.5993993282318115,
+      "step": 164
+    },
+    {
+      "epoch": 0.453551912568306,
+      "grad_norm": 0.647459864616394,
+      "learning_rate": 2.3096830234261996e-05,
+      "loss": 0.753265917301178,
+      "step": 166
+    },
+    {
+      "epoch": 0.45901639344262296,
+      "grad_norm": 0.5001090168952942,
+      "learning_rate": 2.2868562279142912e-05,
+      "loss": 1.1555891036987305,
+      "step": 168
+    },
+    {
+      "epoch": 0.4644808743169399,
+      "grad_norm": 0.2678913176059723,
+      "learning_rate": 2.2637927755807458e-05,
+      "loss": 1.246796727180481,
+      "step": 170
+    },
+    {
+      "epoch": 0.46994535519125685,
+      "grad_norm": 0.417643666267395,
+      "learning_rate": 2.2405012368423786e-05,
+      "loss": 1.3203487396240234,
+      "step": 172
+    },
+    {
+      "epoch": 0.47540983606557374,
+      "grad_norm": 0.5594274401664734,
+      "learning_rate": 2.2169902668732893e-05,
+      "loss": 1.417712926864624,
+      "step": 174
+    },
+    {
+      "epoch": 0.4808743169398907,
+      "grad_norm": 0.31678417325019836,
+      "learning_rate": 2.193268602388583e-05,
+      "loss": 1.2238776683807373,
+      "step": 176
+    },
+    {
+      "epoch": 0.48633879781420764,
+      "grad_norm": 0.6878252625465393,
+      "learning_rate": 2.1693450583977953e-05,
+      "loss": 1.1790485382080078,
+      "step": 178
+    },
+    {
+      "epoch": 0.4918032786885246,
+      "grad_norm": 0.28679513931274414,
+      "learning_rate": 2.1452285249292147e-05,
+      "loss": 1.1895920038223267,
+      "step": 180
+    },
+    {
+      "epoch": 0.4972677595628415,
+      "grad_norm": 0.322842538356781,
+      "learning_rate": 2.12092796372634e-05,
+      "loss": 1.174275517463684,
+      "step": 182
+    },
+    {
+      "epoch": 0.5027322404371585,
+      "grad_norm": 0.6668521761894226,
+      "learning_rate": 2.096452404917679e-05,
+      "loss": 1.3329179286956787,
+      "step": 184
+    },
+    {
+      "epoch": 0.5081967213114754,
+      "grad_norm": 0.7745230197906494,
+      "learning_rate": 2.0718109436611348e-05,
+      "loss": 1.1880346536636353,
+      "step": 186
+    },
+    {
+      "epoch": 0.5136612021857924,
+      "grad_norm": 0.5202271938323975,
+      "learning_rate": 2.0470127367642345e-05,
+      "loss": 1.1781209707260132,
+      "step": 188
+    },
+    {
+      "epoch": 0.5191256830601093,
+      "grad_norm": 0.5798245668411255,
+      "learning_rate": 2.022066999281444e-05,
+      "loss": 1.233837366104126,
+      "step": 190
+    },
+    {
+      "epoch": 0.5245901639344263,
+      "grad_norm": 0.5479984879493713,
+      "learning_rate": 1.9969830010898358e-05,
+      "loss": 1.2089662551879883,
+      "step": 192
+    },
+    {
+      "epoch": 0.5300546448087432,
+      "grad_norm": 0.23682758212089539,
+      "learning_rate": 1.9717700634443903e-05,
+      "loss": 1.218327522277832,
+      "step": 194
+    },
+    {
+      "epoch": 0.5355191256830601,
+      "grad_norm": 0.3199704587459564,
+      "learning_rate": 1.9464375555142e-05,
+      "loss": 1.2755553722381592,
+      "step": 196
+    },
+    {
+      "epoch": 0.5409836065573771,
+      "grad_norm": 0.2734091877937317,
+      "learning_rate": 1.9209948909008734e-05,
+      "loss": 1.1226749420166016,
+      "step": 198
+    },
+    {
+      "epoch": 0.546448087431694,
+      "grad_norm": 0.25196999311447144,
+      "learning_rate": 1.8954515241404218e-05,
+      "loss": 1.2352243661880493,
+      "step": 200
+    },
+    {
+      "epoch": 0.5519125683060109,
+      "grad_norm": 0.28729572892189026,
+      "learning_rate": 1.8698169471899414e-05,
+      "loss": 1.0766348838806152,
+      "step": 202
+    },
+    {
+      "epoch": 0.5573770491803278,
+      "grad_norm": 0.5765514373779297,
+      "learning_rate": 1.8441006859003842e-05,
+      "loss": 0.8550919890403748,
+      "step": 204
+    },
+    {
+      "epoch": 0.5628415300546448,
+      "grad_norm": 0.5975605845451355,
+      "learning_rate": 1.818312296476737e-05,
+      "loss": 1.239629864692688,
+      "step": 206
+    },
+    {
+      "epoch": 0.5683060109289617,
+      "grad_norm": 0.23489779233932495,
+      "learning_rate": 1.792461361926921e-05,
+      "loss": 1.2550042867660522,
+      "step": 208
+    },
+    {
+      "epoch": 0.5737704918032787,
+      "grad_norm": 0.4763418138027191,
+      "learning_rate": 1.766557488500727e-05,
+      "loss": 1.2072700262069702,
+      "step": 210
+    },
+    {
+      "epoch": 0.5792349726775956,
+      "grad_norm": 0.23921915888786316,
+      "learning_rate": 1.7406103021201212e-05,
+      "loss": 1.5028775930404663,
+      "step": 212
+    },
+    {
+      "epoch": 0.5846994535519126,
+      "grad_norm": 0.42051005363464355,
+      "learning_rate": 1.7146294448022335e-05,
+      "loss": 1.2044477462768555,
+      "step": 214
+    },
+    {
+      "epoch": 0.5901639344262295,
+      "grad_norm": 1.0494626760482788,
+      "learning_rate": 1.688624571076371e-05,
+      "loss": 1.560670018196106,
+      "step": 216
+    },
+    {
+      "epoch": 0.5956284153005464,
+      "grad_norm": 0.4011470377445221,
+      "learning_rate": 1.6626053443963762e-05,
+      "loss": 1.1750210523605347,
+      "step": 218
+    },
+    {
+      "epoch": 0.6010928961748634,
+      "grad_norm": 0.4405505657196045,
+      "learning_rate": 1.636581433549674e-05,
+      "loss": 1.2384079694747925,
+      "step": 220
+    },
+    {
+      "epoch": 0.6065573770491803,
+      "grad_norm": 0.86158287525177,
+      "learning_rate": 1.610562509064332e-05,
+      "loss": 1.091644048690796,
+      "step": 222
+    },
+    {
+      "epoch": 0.6120218579234973,
+      "grad_norm": 0.40252819657325745,
+      "learning_rate": 1.5845582396154786e-05,
+      "loss": 1.0258910655975342,
+      "step": 224
+    },
+    {
+      "epoch": 0.6174863387978142,
+      "grad_norm": 0.2338150292634964,
+      "learning_rate": 1.5585782884324064e-05,
+      "loss": 1.2002568244934082,
+      "step": 226
+    },
+    {
+      "epoch": 0.6229508196721312,
+      "grad_norm": 0.5918247699737549,
+      "learning_rate": 1.5326323097077015e-05,
+      "loss": 1.2207123041152954,
+      "step": 228
+    },
+    {
+      "epoch": 0.6284153005464481,
+      "grad_norm": 0.31474488973617554,
+      "learning_rate": 1.5067299450097261e-05,
+      "loss": 0.70670485496521,
+      "step": 230
+    },
+    {
+      "epoch": 0.6338797814207651,
+      "grad_norm": 0.32660529017448425,
+      "learning_rate": 1.4808808196998006e-05,
+      "loss": 1.2120585441589355,
+      "step": 232
+    },
+    {
+      "epoch": 0.639344262295082,
+      "grad_norm": 1.468921422958374,
+      "learning_rate": 1.4550945393554004e-05,
+      "loss": 0.9596046209335327,
+      "step": 234
+    },
+    {
+      "epoch": 0.644808743169399,
+      "grad_norm": 2.728471279144287,
+      "learning_rate": 1.4293806862007085e-05,
+      "loss": 0.903359055519104,
+      "step": 236
+    },
+    {
+      "epoch": 0.6502732240437158,
+      "grad_norm": 0.6711524724960327,
+      "learning_rate": 1.4037488155458448e-05,
+      "loss": 1.2866835594177246,
+      "step": 238
+    },
+    {
+      "epoch": 0.6557377049180327,
+      "grad_norm": 0.40566977858543396,
+      "learning_rate": 1.3782084522360981e-05,
+      "loss": 1.20192289352417,
+      "step": 240
+    },
+    {
+      "epoch": 0.6612021857923497,
+      "grad_norm": 0.3911570608615875,
+      "learning_rate": 1.3527690871124762e-05,
+      "loss": 1.1558232307434082,
+      "step": 242
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.5365505218505859,
+      "learning_rate": 1.3274401734848958e-05,
+      "loss": 1.029914379119873,
+      "step": 244
+    },
+    {
+      "epoch": 0.6721311475409836,
+      "grad_norm": 0.2497815638780594,
+      "learning_rate": 1.3022311236193156e-05,
+      "loss": 1.1662428379058838,
+      "step": 246
+    },
+    {
+      "epoch": 0.6775956284153005,
+      "grad_norm": 0.4294073283672333,
+      "learning_rate": 1.2771513052401236e-05,
+      "loss": 1.2328400611877441,
+      "step": 248
+    },
+    {
+      "epoch": 0.6830601092896175,
+      "grad_norm": 0.352763831615448,
+      "learning_rate": 1.2522100380490744e-05,
+      "loss": 1.1996644735336304,
+      "step": 250
+    },
+    {
+      "epoch": 0.6885245901639344,
+      "grad_norm": 0.46486788988113403,
+      "learning_rate": 1.2274165902620732e-05,
+      "loss": 1.1101069450378418,
+      "step": 252
+    },
+    {
+      "epoch": 0.6939890710382514,
+      "grad_norm": 0.24696502089500427,
+      "learning_rate": 1.2027801751650918e-05,
+      "loss": 1.1543952226638794,
+      "step": 254
+    },
+    {
+      "epoch": 0.6994535519125683,
+      "grad_norm": 0.25500428676605225,
+      "learning_rate": 1.1783099476904972e-05,
+      "loss": 1.148316740989685,
+      "step": 256
+    },
+    {
+      "epoch": 0.7049180327868853,
+      "grad_norm": 0.37820199131965637,
+      "learning_rate": 1.1540150010150599e-05,
+      "loss": 1.079501748085022,
+      "step": 258
+    },
+    {
+      "epoch": 0.7103825136612022,
+      "grad_norm": 0.48190441727638245,
+      "learning_rate": 1.1299043631809205e-05,
+      "loss": 1.1459242105484009,
+      "step": 260
+    },
+    {
+      "epoch": 0.7158469945355191,
+      "grad_norm": 0.5813128352165222,
+      "learning_rate": 1.1059869937407486e-05,
+      "loss": 1.240357756614685,
+      "step": 262
+    },
+    {
+      "epoch": 0.7213114754098361,
+      "grad_norm": 0.2543342113494873,
+      "learning_rate": 1.082271780428362e-05,
+      "loss": 1.1864430904388428,
+      "step": 264
+    },
+    {
+      "epoch": 0.726775956284153,
+      "grad_norm": 0.26917213201522827,
+      "learning_rate": 1.0587675358560278e-05,
+      "loss": 1.0468827486038208,
+      "step": 266
+    },
+    {
+      "epoch": 0.73224043715847,
+      "grad_norm": 0.4517075717449188,
+      "learning_rate": 1.0354829942396837e-05,
+      "loss": 1.1757278442382812,
+      "step": 268
+    },
+    {
+      "epoch": 0.7377049180327869,
+      "grad_norm": 0.23133859038352966,
+      "learning_rate": 1.012426808153287e-05,
+      "loss": 1.178215742111206,
+      "step": 270
+    },
+    {
+      "epoch": 0.7431693989071039,
+      "grad_norm": 0.3878965377807617,
+      "learning_rate": 9.896075453135039e-06,
+      "loss": 1.1388592720031738,
+      "step": 272
+    },
+    {
+      "epoch": 0.7486338797814208,
+      "grad_norm": 0.3238541781902313,
+      "learning_rate": 9.67033685395934e-06,
+      "loss": 1.161685585975647,
+      "step": 274
+    },
+    {
+      "epoch": 0.7540983606557377,
+      "grad_norm": 0.27236634492874146,
+      "learning_rate": 9.447136168840466e-06,
+      "loss": 1.2053990364074707,
+      "step": 276
+    },
+    {
+      "epoch": 0.7595628415300546,
+      "grad_norm": 0.3311232924461365,
+      "learning_rate": 9.226556339520069e-06,
+      "loss": 1.1811869144439697,
+      "step": 278
+    },
+    {
+      "epoch": 0.7650273224043715,
+      "grad_norm": 0.824040412902832,
+      "learning_rate": 9.008679333825478e-06,
+      "loss": 0.9021319150924683,
+      "step": 280
+    },
+    {
+      "epoch": 0.7704918032786885,
+      "grad_norm": 0.3380263149738312,
+      "learning_rate": 8.793586115210326e-06,
+      "loss": 1.1787052154541016,
+      "step": 282
+    },
+    {
+      "epoch": 0.7759562841530054,
+      "grad_norm": 0.408220112323761,
+      "learning_rate": 8.581356612668382e-06,
+      "loss": 1.1672601699829102,
+      "step": 284
+    },
+    {
+      "epoch": 0.7814207650273224,
+      "grad_norm": 0.35263848304748535,
+      "learning_rate": 8.372069691031804e-06,
+      "loss": 1.1768161058425903,
+      "step": 286
+    },
+    {
+      "epoch": 0.7868852459016393,
+      "grad_norm": 0.37962621450424194,
+      "learning_rate": 8.165803121664869e-06,
+      "loss": 1.146876335144043,
+      "step": 288
+    },
+    {
+      "epoch": 0.7923497267759563,
+      "grad_norm": 0.27549535036087036,
+      "learning_rate": 7.962633553563965e-06,
+      "loss": 1.1030545234680176,
+      "step": 290
+    },
+    {
+      "epoch": 0.7978142076502732,
+      "grad_norm": 0.591926097869873,
+      "learning_rate": 7.762636484874723e-06,
+      "loss": 1.1360045671463013,
+      "step": 292
+    },
+    {
+      "epoch": 0.8032786885245902,
+      "grad_norm": 0.26219072937965393,
+      "learning_rate": 7.565886234836767e-06,
+      "loss": 1.1480509042739868,
+      "step": 294
+    },
+    {
+      "epoch": 0.8087431693989071,
+      "grad_norm": 0.4182840585708618,
+      "learning_rate": 7.3724559161665876e-06,
+      "loss": 1.494627594947815,
+      "step": 296
+    },
+    {
+      "epoch": 0.8142076502732241,
+      "grad_norm": 0.7810959219932556,
+      "learning_rate": 7.182417407888703e-06,
+      "loss": 0.8853966593742371,
+      "step": 298
+    },
+    {
+      "epoch": 0.819672131147541,
+      "grad_norm": 0.3140762448310852,
+      "learning_rate": 6.995841328625321e-06,
+      "loss": 1.5603413581848145,
+      "step": 300
+    },
+    {
+      "epoch": 0.825136612021858,
+      "grad_norm": 0.3591131269931793,
+      "learning_rate": 6.812797010354325e-06,
+      "loss": 1.1755470037460327,
+      "step": 302
+    },
+    {
+      "epoch": 0.8306010928961749,
+      "grad_norm": 2.338329315185547,
+      "learning_rate": 6.63335247264542e-06,
+      "loss": 0.7928791046142578,
+      "step": 304
+    },
+    {
+      "epoch": 0.8360655737704918,
+      "grad_norm": 0.6885122060775757,
+      "learning_rate": 6.457574397383919e-06,
+      "loss": 1.514641284942627,
+      "step": 306
+    },
+    {
+      "epoch": 0.8415300546448088,
+      "grad_norm": 0.3260650336742401,
+      "learning_rate": 6.285528103991665e-06,
+      "loss": 0.8320769667625427,
+      "step": 308
+    },
+    {
+      "epoch": 0.8469945355191257,
+      "grad_norm": 0.39165613055229187,
+      "learning_rate": 6.117277525154225e-06,
+      "loss": 1.1644271612167358,
+      "step": 310
+    },
+    {
+      "epoch": 0.8524590163934426,
+      "grad_norm": 0.34331265091896057,
+      "learning_rate": 5.952885183063397e-06,
+      "loss": 1.17978835105896,
+      "step": 312
+    },
+    {
+      "epoch": 0.8579234972677595,
+      "grad_norm": 0.3723102807998657,
+      "learning_rate": 5.792412166183841e-06,
+      "loss": 1.1927130222320557,
+      "step": 314
+    },
+    {
+      "epoch": 0.8633879781420765,
+      "grad_norm": 0.5044635534286499,
+      "learning_rate": 5.635918106552546e-06,
+      "loss": 1.1958658695220947,
+      "step": 316
+    },
+    {
+      "epoch": 0.8688524590163934,
+      "grad_norm": 0.5813291668891907,
+      "learning_rate": 5.483461157619428e-06,
+      "loss": 1.0818508863449097,
+      "step": 318
+    },
+    {
+      "epoch": 0.8743169398907104,
+      "grad_norm": 0.23069122433662415,
+      "learning_rate": 5.335097972637441e-06,
+      "loss": 1.2942509651184082,
+      "step": 320
+    },
+    {
+      "epoch": 0.8797814207650273,
+      "grad_norm": 0.31037190556526184,
+      "learning_rate": 5.1908836836101135e-06,
+      "loss": 1.1336814165115356,
+      "step": 322
+    },
+    {
+      "epoch": 0.8852459016393442,
+      "grad_norm": 0.43608394265174866,
+      "learning_rate": 5.050871880804414e-06,
+      "loss": 0.5545326471328735,
+      "step": 324
+    },
+    {
+      "epoch": 0.8907103825136612,
+      "grad_norm": 0.40804681181907654,
+      "learning_rate": 4.915114592836521e-06,
+      "loss": 1.1569054126739502,
+      "step": 326
+    },
+    {
+      "epoch": 0.8961748633879781,
+      "grad_norm": 0.6417044997215271,
+      "learning_rate": 4.783662267337909e-06,
+      "loss": 1.0306603908538818,
+      "step": 328
+    },
+    {
+      "epoch": 0.9016393442622951,
+      "grad_norm": 0.3709650933742523,
+      "learning_rate": 4.656563752208907e-06,
+      "loss": 1.184291958808899,
+      "step": 330
+    },
+    {
+      "epoch": 0.907103825136612,
+      "grad_norm": 0.5156346559524536,
+      "learning_rate": 4.533866277466767e-06,
+      "loss": 1.239979863166809,
+      "step": 332
+    },
+    {
+      "epoch": 0.912568306010929,
+      "grad_norm": 0.32417598366737366,
+      "learning_rate": 4.415615437694876e-06,
+      "loss": 1.0216443538665771,
+      "step": 334
+    },
+    {
+      "epoch": 0.9180327868852459,
+      "grad_norm": 0.31912070512771606,
+      "learning_rate": 4.3018551750997694e-06,
+      "loss": 1.132981777191162,
+      "step": 336
+    },
+    {
+      "epoch": 0.9234972677595629,
+      "grad_norm": 0.6049280762672424,
+      "learning_rate": 4.192627763182111e-06,
+      "loss": 1.1525002717971802,
+      "step": 338
+    },
+    {
+      "epoch": 0.9289617486338798,
+      "grad_norm": 0.398091584444046,
+      "learning_rate": 4.087973791027797e-06,
+      "loss": 1.151615858078003,
+      "step": 340
+    },
+    {
+      "epoch": 0.9344262295081968,
+      "grad_norm": 0.7425937652587891,
+      "learning_rate": 3.987932148224993e-06,
+      "loss": 1.1880524158477783,
+      "step": 342
+    },
+    {
+      "epoch": 0.9398907103825137,
+      "grad_norm": 0.9477114677429199,
+      "learning_rate": 3.8925400104126834e-06,
+      "loss": 1.1307789087295532,
+      "step": 344
+    },
+    {
+      "epoch": 0.9453551912568307,
+      "grad_norm": 0.3224574327468872,
+      "learning_rate": 3.8018328254661618e-06,
+      "loss": 1.508508324623108,
+      "step": 346
+    },
+    {
+      "epoch": 0.9508196721311475,
+      "grad_norm": 0.42406103014945984,
+      "learning_rate": 3.715844300324527e-06,
+      "loss": 0.8069607615470886,
+      "step": 348
+    },
+    {
+      "epoch": 0.9562841530054644,
+      "grad_norm": 0.2576044201850891,
+      "learning_rate": 3.6346063884651327e-06,
+      "loss": 0.806373655796051,
+      "step": 350
+    },
+    {
+      "epoch": 0.9617486338797814,
+      "grad_norm": 0.5339590907096863,
+      "learning_rate": 3.558149278029624e-06,
+      "loss": 1.1425732374191284,
+      "step": 352
+    },
+    {
+      "epoch": 0.9672131147540983,
+      "grad_norm": 0.2579115331172943,
+      "learning_rate": 3.4865013806059817e-06,
+      "loss": 0.8368316292762756,
+      "step": 354
+    },
+    {
+      "epoch": 0.9726775956284153,
+      "grad_norm": 0.4226749539375305,
+      "learning_rate": 3.419689320670712e-06,
+      "loss": 1.0304514169692993,
+      "step": 356
+    },
+    {
+      "epoch": 0.9781420765027322,
+      "grad_norm": 0.23506596684455872,
+      "learning_rate": 3.35773792569517e-06,
+      "loss": 0.720623254776001,
+      "step": 358
+    },
+    {
+      "epoch": 0.9836065573770492,
+      "grad_norm": 0.3969164490699768,
+      "learning_rate": 3.300670216919602e-06,
+      "loss": 1.1500842571258545,
+      "step": 360
+    },
+    {
+      "epoch": 0.9890710382513661,
+      "grad_norm": 0.6647183895111084,
+      "learning_rate": 3.2485074007984468e-06,
+      "loss": 0.9328727126121521,
+      "step": 362
+    },
+    {
+      "epoch": 0.994535519125683,
+      "grad_norm": 0.2883590757846832,
+      "learning_rate": 3.2012688611199566e-06,
+      "loss": 1.1320202350616455,
+      "step": 364
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3654276728630066,
+      "learning_rate": 3.158972151803165e-06,
+      "loss": 1.1198816299438477,
+      "step": 366
+    },
+    {
+      "epoch": 1.005464480874317,
+      "grad_norm": 0.3304823040962219,
+      "learning_rate": 3.1216329903748095e-06,
+      "loss": 0.9562312364578247,
+      "step": 368
+    },
+    {
+      "epoch": 1.010928961748634,
+      "grad_norm": 0.43168309330940247,
+      "learning_rate": 3.089265252128686e-06,
+      "loss": 1.0244945287704468,
+      "step": 370
+    },
+    {
+      "epoch": 1.0163934426229508,
+      "grad_norm": 0.25342443585395813,
+      "learning_rate": 3.061880964969555e-06,
+      "loss": 0.9778792858123779,
+      "step": 372
+    },
+    {
+      "epoch": 1.0218579234972678,
+      "grad_norm": 0.35910850763320923,
+      "learning_rate": 3.039490304943562e-06,
+      "loss": 0.9393002390861511,
+      "step": 374
+    },
+    {
+      "epoch": 1.0273224043715847,
+      "grad_norm": 0.4086434245109558,
+      "learning_rate": 3.022101592456795e-06,
+      "loss": 0.8444104790687561,
+      "step": 376
+    },
+    {
+      "epoch": 1.0327868852459017,
+      "grad_norm": 0.4661693871021271,
+      "learning_rate": 3.0097212891834095e-06,
+      "loss": 0.9321051836013794,
+      "step": 378
+    },
+    {
+      "epoch": 1.0382513661202186,
+      "grad_norm": 0.37025684118270874,
+      "learning_rate": 3.0023539956644634e-06,
+      "loss": 0.6185029745101929,
+      "step": 380
+    },
+    {
+      "epoch": 1.0437158469945356,
+      "grad_norm": 0.4136655628681183,
+      "learning_rate": 3.0000024495983428e-06,
+      "loss": 0.8996444940567017,
+      "step": 382
+    },
+    {
+      "epoch": 1.0491803278688525,
+      "grad_norm": 0.2609977126121521,
+      "learning_rate": 3.002667524823434e-06,
+      "loss": 0.951973021030426,
+      "step": 384
+    },
+    {
+      "epoch": 1.0546448087431695,
+      "grad_norm": 1.5468955039978027,
+      "learning_rate": 3.010348230993402e-06,
+      "loss": 0.6188804507255554,
+      "step": 386
+    },
+    {
+      "epoch": 1.0601092896174864,
+      "grad_norm": 0.5185752511024475,
+      "learning_rate": 3.0230417139451987e-06,
+      "loss": 0.5328562259674072,
+      "step": 388
+    },
+    {
+      "epoch": 1.0655737704918034,
+      "grad_norm": 0.5812699794769287,
+      "learning_rate": 3.0407432567596883e-06,
+      "loss": 1.1187443733215332,
+      "step": 390
+    },
+    {
+      "epoch": 1.0710382513661203,
+      "grad_norm": 0.32687872648239136,
+      "learning_rate": 3.0634462815144474e-06,
+      "loss": 0.8662930727005005,
+      "step": 392
+    },
+    {
+      "epoch": 1.0765027322404372,
+      "grad_norm": 0.4097544550895691,
+      "learning_rate": 3.0911423517281404e-06,
+      "loss": 0.5184633135795593,
+      "step": 394
+    },
+    {
+      "epoch": 1.0819672131147542,
+      "grad_norm": 0.33863401412963867,
+      "learning_rate": 3.1238211754955294e-06,
+      "loss": 0.9240728616714478,
+      "step": 396
+    },
+    {
+      "epoch": 1.0874316939890711,
+      "grad_norm": 0.28052061796188354,
+      "learning_rate": 3.161470609311961e-06,
+      "loss": 0.9106302857398987,
+      "step": 398
+    },
+    {
+      "epoch": 1.092896174863388,
+      "grad_norm": 0.4579223394393921,
+      "learning_rate": 3.2040766625859115e-06,
+      "loss": 1.0799837112426758,
+      "step": 400
+    },
+    {
+      "epoch": 1.098360655737705,
+      "grad_norm": 0.930515706539154,
+      "learning_rate": 3.2516235028379157e-06,
+      "loss": 0.9288961291313171,
+      "step": 402
+    },
+    {
+      "epoch": 1.1038251366120218,
+      "grad_norm": 0.37287023663520813,
+      "learning_rate": 3.304093461583944e-06,
+      "loss": 0.9670571684837341,
+      "step": 404
+    },
+    {
+      "epoch": 1.1092896174863387,
+      "grad_norm": 1.240659475326538,
+      "learning_rate": 3.3614670409010353e-06,
+      "loss": 1.0186374187469482,
+      "step": 406
+    },
+    {
+      "epoch": 1.1147540983606556,
+      "grad_norm": 0.8452418446540833,
+      "learning_rate": 3.4237229206727602e-06,
+      "loss": 0.6442115306854248,
+      "step": 408
+    },
+    {
+      "epoch": 1.1202185792349726,
+      "grad_norm": 0.36524245142936707,
+      "learning_rate": 3.490837966511817e-06,
+      "loss": 1.2035061120986938,
+      "step": 410
+    },
+    {
+      "epoch": 1.1256830601092895,
+      "grad_norm": 0.6859813928604126,
+      "learning_rate": 3.5627872383567937e-06,
+      "loss": 0.6509969830513,
+      "step": 412
+    },
+    {
+      "epoch": 1.1311475409836065,
+      "grad_norm": 0.28654637932777405,
+      "learning_rate": 3.6395439997399494e-06,
+      "loss": 0.6548624634742737,
+      "step": 414
+    },
+    {
+      "epoch": 1.1366120218579234,
+      "grad_norm": 0.6482637524604797,
+      "learning_rate": 3.721079727722522e-06,
+      "loss": 0.6231327056884766,
+      "step": 416
+    },
+    {
+      "epoch": 1.1420765027322404,
+      "grad_norm": 0.40834060311317444,
+      "learning_rate": 3.8073641234939055e-06,
+      "loss": 0.9411674737930298,
+      "step": 418
+    },
+    {
+      "epoch": 1.1475409836065573,
+      "grad_norm": 0.42818281054496765,
+      "learning_rate": 3.898365123630732e-06,
+      "loss": 0.7944842576980591,
+      "step": 420
+    },
+    {
+      "epoch": 1.1530054644808743,
+      "grad_norm": 0.29905006289482117,
+      "learning_rate": 3.994048912011692e-06,
+      "loss": 0.9097725749015808,
+      "step": 422
+    },
+    {
+      "epoch": 1.1584699453551912,
+      "grad_norm": 0.789088249206543,
+      "learning_rate": 4.094379932383666e-06,
+      "loss": 0.7131223678588867,
+      "step": 424
+    },
+    {
+      "epoch": 1.1639344262295082,
+      "grad_norm": 0.3939225375652313,
+      "learning_rate": 4.199320901574489e-06,
+      "loss": 0.6140289306640625,
+      "step": 426
+    },
+    {
+      "epoch": 1.169398907103825,
+      "grad_norm": 0.4326651096343994,
+      "learning_rate": 4.3088328233474185e-06,
+      "loss": 1.078157901763916,
+      "step": 428
+    },
+    {
+      "epoch": 1.174863387978142,
+      "grad_norm": 0.30610015988349915,
+      "learning_rate": 4.422875002892234e-06,
+      "loss": 0.9334272742271423,
+      "step": 430
+    },
+    {
+      "epoch": 1.180327868852459,
+      "grad_norm": 0.34937551617622375,
+      "learning_rate": 4.54140506194747e-06,
+      "loss": 0.9314782023429871,
+      "step": 432
+    },
+    {
+      "epoch": 1.185792349726776,
+      "grad_norm": 0.8679988384246826,
+      "learning_rate": 4.664378954548241e-06,
+      "loss": 0.7978003025054932,
+      "step": 434
+    },
+    {
+      "epoch": 1.1912568306010929,
+      "grad_norm": 0.3360673785209656,
+      "learning_rate": 4.791750983393832e-06,
+      "loss": 0.9832219481468201,
+      "step": 436
+    },
+    {
+      "epoch": 1.1967213114754098,
+      "grad_norm": 0.264480859041214,
+      "learning_rate": 4.9234738168288466e-06,
+      "loss": 0.8762679696083069,
+      "step": 438
+    },
+    {
+      "epoch": 1.2021857923497268,
+      "grad_norm": 0.39207419753074646,
+      "learning_rate": 5.059498506431758e-06,
+      "loss": 0.7775551080703735,
+      "step": 440
+    },
+    {
+      "epoch": 1.2076502732240437,
+      "grad_norm": 0.2239266186952591,
+      "learning_rate": 5.199774505204206e-06,
+      "loss": 0.93987637758255,
+      "step": 442
+    },
+    {
+      "epoch": 1.2131147540983607,
+      "grad_norm": 0.33733996748924255,
+      "learning_rate": 5.344249686354357e-06,
+      "loss": 0.9468815922737122,
+      "step": 444
+    },
+    {
+      "epoch": 1.2185792349726776,
+      "grad_norm": 0.47637975215911865,
+      "learning_rate": 5.492870362667299e-06,
+      "loss": 0.5318366289138794,
+      "step": 446
+    },
+    {
+      "epoch": 1.2240437158469946,
+      "grad_norm": 0.26395881175994873,
+      "learning_rate": 5.645581306455302e-06,
+      "loss": 0.8348568081855774,
+      "step": 448
+    },
+    {
+      "epoch": 1.2295081967213115,
+      "grad_norm": 0.5106167197227478,
+      "learning_rate": 5.802325770080506e-06,
+      "loss": 0.9297827482223511,
+      "step": 450
+    },
+    {
+      "epoch": 1.2349726775956285,
+      "grad_norm": 0.47476786375045776,
+      "learning_rate": 5.96304550704246e-06,
+      "loss": 0.988614022731781,
+      "step": 452
+    },
+    {
+      "epoch": 1.2404371584699454,
+      "grad_norm": 0.5340549349784851,
+      "learning_rate": 6.127680793622588e-06,
+      "loss": 0.42125749588012695,
+      "step": 454
+    },
+    {
+      "epoch": 1.2459016393442623,
+      "grad_norm": 0.9180600643157959,
+      "learning_rate": 6.296170451077657e-06,
+      "loss": 0.7597475647926331,
+      "step": 456
+    },
+    {
+      "epoch": 1.2513661202185793,
+      "grad_norm": 0.46278026700019836,
+      "learning_rate": 6.468451868373856e-06,
+      "loss": 0.6985284686088562,
+      "step": 458
+    },
+    {
+      "epoch": 1.2568306010928962,
+      "grad_norm": 0.27935847640037537,
+      "learning_rate": 6.6444610254532e-06,
+      "loss": 0.9660181999206543,
+      "step": 460
+    },
+    {
+      "epoch": 1.2622950819672132,
+      "grad_norm": 0.8495526909828186,
+      "learning_rate": 6.824132517023449e-06,
+      "loss": 0.7250743508338928,
+      "step": 462
+    },
+    {
+      "epoch": 1.2677595628415301,
+      "grad_norm": 7.655890941619873,
+      "learning_rate": 7.007399576862872e-06,
+      "loss": 0.9311206936836243,
+      "step": 464
+    },
+    {
+      "epoch": 1.273224043715847,
+      "grad_norm": 0.7682108283042908,
+      "learning_rate": 7.1941941026306275e-06,
+      "loss": 0.976463258266449,
+      "step": 466
+    },
+    {
+      "epoch": 1.278688524590164,
+      "grad_norm": 0.31356796622276306,
+      "learning_rate": 7.3844466811737555e-06,
+      "loss": 0.9272637963294983,
+      "step": 468
+    },
+    {
+      "epoch": 1.2841530054644807,
+      "grad_norm": 0.304342657327652,
+      "learning_rate": 7.578086614321175e-06,
+      "loss": 0.9191130995750427,
+      "step": 470
+    },
+    {
+      "epoch": 1.289617486338798,
+      "grad_norm": 0.4537976086139679,
+      "learning_rate": 7.775041945155295e-06,
+      "loss": 1.185452938079834,
+      "step": 472
+    },
+    {
+      "epoch": 1.2950819672131146,
+      "grad_norm": 0.7000725269317627,
+      "learning_rate": 7.975239484751258e-06,
+      "loss": 0.859736979007721,
+      "step": 474
+    },
+    {
+      "epoch": 1.3005464480874318,
+      "grad_norm": 0.49727311730384827,
+      "learning_rate": 8.178604839374125e-06,
+      "loss": 0.5100428462028503,
+      "step": 476
+    },
+    {
+      "epoch": 1.3060109289617485,
+      "grad_norm": 0.3634011149406433,
+      "learning_rate": 8.385062438123673e-06,
+      "loss": 0.9873999953269958,
+      "step": 478
+    },
+    {
+      "epoch": 1.3114754098360657,
+      "grad_norm": 0.3196913003921509,
+      "learning_rate": 8.594535561016661e-06,
+      "loss": 1.0027223825454712,
+      "step": 480
+    },
+    {
+      "epoch": 1.3169398907103824,
+      "grad_norm": 0.275206595659256,
+      "learning_rate": 8.806946367496155e-06,
+      "loss": 0.8830161094665527,
+      "step": 482
+    },
+    {
+      "epoch": 1.3224043715846996,
+      "grad_norm": 0.7456058263778687,
+      "learning_rate": 9.02221592535712e-06,
+      "loss": 0.6142900586128235,
+      "step": 484
+    },
+    {
+      "epoch": 1.3278688524590163,
+      "grad_norm": 0.27207374572753906,
+      "learning_rate": 9.240264240077859e-06,
+      "loss": 0.8691650629043579,
+      "step": 486
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.31759360432624817,
+      "learning_rate": 9.461010284546016e-06,
+      "loss": 1.0310981273651123,
+      "step": 488
+    },
+    {
+      "epoch": 1.3387978142076502,
+      "grad_norm": 0.6895498037338257,
+      "learning_rate": 9.684372029168438e-06,
+      "loss": 0.660420298576355,
+      "step": 490
+    },
+    {
+      "epoch": 1.3442622950819672,
+      "grad_norm": 0.4173939824104309,
+      "learning_rate": 9.91026647235348e-06,
+      "loss": 1.0019088983535767,
+      "step": 492
+    },
+    {
+      "epoch": 1.349726775956284,
+      "grad_norm": 0.21376250684261322,
+      "learning_rate": 1.0138609671354586e-05,
+      "loss": 0.7606089115142822,
+      "step": 494
+    },
+    {
+      "epoch": 1.355191256830601,
+      "grad_norm": 0.6442067623138428,
+      "learning_rate": 1.0369316773463458e-05,
+      "loss": 0.9546438455581665,
+      "step": 496
+    },
+    {
+      "epoch": 1.360655737704918,
+      "grad_norm": 3.213313579559326,
+      "learning_rate": 1.0602302047541566e-05,
+      "loss": 0.9225825071334839,
+      "step": 498
+    },
+    {
+      "epoch": 1.366120218579235,
+      "grad_norm": 0.356067031621933,
+      "learning_rate": 1.083747891587788e-05,
+      "loss": 1.08867609500885,
+      "step": 500
+    },
+    {
+      "epoch": 1.3715846994535519,
+      "grad_norm": 0.46825549006462097,
+      "learning_rate": 1.1074759986361392e-05,
+      "loss": 0.9150928258895874,
+      "step": 502
+    },
+    {
+      "epoch": 1.3770491803278688,
+      "grad_norm": 0.5194010138511658,
+      "learning_rate": 1.1314057084956073e-05,
+      "loss": 0.5116444826126099,
+      "step": 504
+    },
+    {
+      "epoch": 1.3825136612021858,
+      "grad_norm": 0.7615500092506409,
+      "learning_rate": 1.1555281288466553e-05,
+      "loss": 0.7318437695503235,
+      "step": 506
+    },
+    {
+      "epoch": 1.3879781420765027,
+      "grad_norm": 0.32076653838157654,
+      "learning_rate": 1.1798342957582084e-05,
+      "loss": 0.9810932278633118,
+      "step": 508
+    },
+    {
+      "epoch": 1.3934426229508197,
+      "grad_norm": 0.26743048429489136,
+      "learning_rate": 1.2043151770186725e-05,
+      "loss": 1.0388559103012085,
+      "step": 510
+    },
+    {
+      "epoch": 1.3989071038251366,
+      "grad_norm": 0.34534981846809387,
+      "learning_rate": 1.2289616754923078e-05,
+      "loss": 0.8841665387153625,
+      "step": 512
+    },
+    {
+      "epoch": 1.4043715846994536,
+      "grad_norm": 0.16258741915225983,
+      "learning_rate": 1.253764632499752e-05,
+      "loss": 0.66373211145401,
+      "step": 514
+    },
+    {
+      "epoch": 1.4098360655737705,
+      "grad_norm": 0.3468623161315918,
+      "learning_rate": 1.2787148312213901e-05,
+      "loss": 0.7078993320465088,
+      "step": 516
+    },
+    {
+      "epoch": 1.4153005464480874,
+      "grad_norm": 0.2647278904914856,
+      "learning_rate": 1.3038030001223439e-05,
+      "loss": 0.9880443215370178,
+      "step": 518
+    },
+    {
+      "epoch": 1.4207650273224044,
+      "grad_norm": 0.8398441672325134,
+      "learning_rate": 1.3290198163977933e-05,
+      "loss": 0.6314539313316345,
+      "step": 520
+    },
+    {
+      "epoch": 1.4262295081967213,
+      "grad_norm": 0.3027450442314148,
+      "learning_rate": 1.3543559094373372e-05,
+      "loss": 1.0502456426620483,
+      "step": 522
+    },
+    {
+      "epoch": 1.4316939890710383,
+      "grad_norm": 0.7764866352081299,
+      "learning_rate": 1.3798018643071386e-05,
+      "loss": 0.9405328631401062,
+      "step": 524
+    },
+    {
+      "epoch": 1.4371584699453552,
+      "grad_norm": 0.37671011686325073,
+      "learning_rate": 1.4053482252485178e-05,
+      "loss": 0.9506468772888184,
+      "step": 526
+    },
+    {
+      "epoch": 1.4426229508196722,
+      "grad_norm": 0.20187242329120636,
+      "learning_rate": 1.4309854991917388e-05,
+      "loss": 0.6395024061203003,
+      "step": 528
+    },
+    {
+      "epoch": 1.4480874316939891,
+      "grad_norm": 0.4782966673374176,
+      "learning_rate": 1.4567041592836413e-05,
+      "loss": 0.748798131942749,
+      "step": 530
+    },
+    {
+      "epoch": 1.453551912568306,
+      "grad_norm": 3.2700560092926025,
+      "learning_rate": 1.48249464842784e-05,
+      "loss": 0.755515456199646,
+      "step": 532
+    },
+    {
+      "epoch": 1.459016393442623,
+      "grad_norm": 0.676388144493103,
+      "learning_rate": 1.508347382836153e-05,
+      "loss": 1.0103578567504883,
+      "step": 534
+    },
+    {
+      "epoch": 1.46448087431694,
+      "grad_norm": 0.30977484583854675,
+      "learning_rate": 1.534252755589961e-05,
+      "loss": 0.982090413570404,
+      "step": 536
+    },
+    {
+      "epoch": 1.469945355191257,
+      "grad_norm": 0.39242854714393616,
+      "learning_rate": 1.5602011402101432e-05,
+      "loss": 1.0859917402267456,
+      "step": 538
+    },
+    {
+      "epoch": 1.4754098360655736,
+      "grad_norm": 0.3935968279838562,
+      "learning_rate": 1.5861828942343037e-05,
+      "loss": 1.0562843084335327,
+      "step": 540
+    },
+    {
+      "epoch": 1.4808743169398908,
+      "grad_norm": 0.33550116419792175,
+      "learning_rate": 1.612188362799917e-05,
+      "loss": 0.9903753995895386,
+      "step": 542
+    },
+    {
+      "epoch": 1.4863387978142075,
+      "grad_norm": 0.4036963880062103,
+      "learning_rate": 1.6382078822320964e-05,
+      "loss": 0.8334465622901917,
+      "step": 544
+    },
+    {
+      "epoch": 1.4918032786885247,
+      "grad_norm": 0.6883772611618042,
+      "learning_rate": 1.6642317836346324e-05,
+      "loss": 1.1092900037765503,
+      "step": 546
+    },
+    {
+      "epoch": 1.4972677595628414,
+      "grad_norm": 3.7706735134124756,
+      "learning_rate": 1.6902503964829644e-05,
+      "loss": 0.4429691433906555,
+      "step": 548
+    },
+    {
+      "epoch": 1.5027322404371586,
+      "grad_norm": 1.0984396934509277,
+      "learning_rate": 1.7162540522177685e-05,
+      "loss": 1.0124402046203613,
+      "step": 550
+    },
+    {
+      "epoch": 1.5081967213114753,
+      "grad_norm": 0.457646906375885,
+      "learning_rate": 1.7422330878378113e-05,
+      "loss": 1.025429368019104,
+      "step": 552
+    },
+    {
+      "epoch": 1.5136612021857925,
+      "grad_norm": 0.2908318340778351,
+      "learning_rate": 1.7681778494907298e-05,
+      "loss": 0.9907000660896301,
+      "step": 554
+    },
+    {
+      "epoch": 1.5191256830601092,
+      "grad_norm": 0.4297858774662018,
+      "learning_rate": 1.794078696060429e-05,
+      "loss": 1.1015424728393555,
+      "step": 556
+    },
+    {
+      "epoch": 1.5245901639344264,
+      "grad_norm": 0.35660913586616516,
+      "learning_rate": 1.819926002749727e-05,
+      "loss": 0.8739935755729675,
+      "step": 558
+    },
+    {
+      "epoch": 1.530054644808743,
+      "grad_norm": 0.345431923866272,
+      "learning_rate": 1.84571016465695e-05,
+      "loss": 0.5142934322357178,
+      "step": 560
+    },
+    {
+      "epoch": 1.5355191256830603,
+      "grad_norm": 2.3219211101531982,
+      "learning_rate": 1.8714216003451295e-05,
+      "loss": 0.9209707975387573,
+      "step": 562
+    },
+    {
+      "epoch": 1.540983606557377,
+      "grad_norm": 1.1686280965805054,
+      "learning_rate": 1.8970507554024827e-05,
+      "loss": 0.8513541221618652,
+      "step": 564
+    },
+    {
+      "epoch": 1.5464480874316942,
+      "grad_norm": 0.2546832859516144,
+      "learning_rate": 1.922588105992838e-05,
+      "loss": 0.9810701012611389,
+      "step": 566
+    },
+    {
+      "epoch": 1.5519125683060109,
+      "grad_norm": 0.5964091420173645,
+      "learning_rate": 1.9480241623947206e-05,
+      "loss": 0.8227204084396362,
+      "step": 568
+    },
+    {
+      "epoch": 1.5573770491803278,
+      "grad_norm": 0.34018632769584656,
+      "learning_rate": 1.9733494725277413e-05,
+      "loss": 1.1703555583953857,
+      "step": 570
+    },
+    {
+      "epoch": 1.5628415300546448,
+      "grad_norm": 0.36660236120224,
+      "learning_rate": 1.998554625465005e-05,
+      "loss": 0.9431501626968384,
+      "step": 572
+    },
+    {
+      "epoch": 1.5683060109289617,
+      "grad_norm": 0.26774540543556213,
+      "learning_rate": 2.0236302549302293e-05,
+      "loss": 0.996977686882019,
+      "step": 574
+    },
+    {
+      "epoch": 1.5737704918032787,
+      "grad_norm": 0.28664132952690125,
+      "learning_rate": 2.0485670427782644e-05,
+      "loss": 0.7795593738555908,
+      "step": 576
+    },
+    {
+      "epoch": 1.5792349726775956,
+      "grad_norm": 0.3652798533439636,
+      "learning_rate": 2.073355722457739e-05,
+      "loss": 1.0711528062820435,
+      "step": 578
+    },
+    {
+      "epoch": 1.5846994535519126,
+      "grad_norm": 0.1629379242658615,
+      "learning_rate": 2.0979870824545165e-05,
+      "loss": 0.5316720008850098,
+      "step": 580
+    },
+    {
+      "epoch": 1.5901639344262295,
+      "grad_norm": 0.2500196099281311,
+      "learning_rate": 2.1224519697147145e-05,
+      "loss": 0.730338454246521,
+      "step": 582
+    },
+    {
+      "epoch": 1.5956284153005464,
+      "grad_norm": 0.3823454976081848,
+      "learning_rate": 2.1467412930459936e-05,
+      "loss": 0.6454403400421143,
+      "step": 584
+    },
+    {
+      "epoch": 1.6010928961748634,
+      "grad_norm": 0.796212375164032,
+      "learning_rate": 2.1708460264958595e-05,
+      "loss": 0.9703986048698425,
+      "step": 586
+    },
+    {
+      "epoch": 1.6065573770491803,
+      "grad_norm": 2.617865800857544,
+      "learning_rate": 2.194757212705718e-05,
+      "loss": 0.992373526096344,
+      "step": 588
+    },
+    {
+      "epoch": 1.6120218579234973,
+      "grad_norm": 0.4272755980491638,
+      "learning_rate": 2.2184659662394522e-05,
+      "loss": 0.9483923316001892,
+      "step": 590
+    },
+    {
+      "epoch": 1.6174863387978142,
+      "grad_norm": 0.9268650412559509,
+      "learning_rate": 2.24196347688526e-05,
+      "loss": 1.0497581958770752,
+      "step": 592
+    },
+    {
+      "epoch": 1.6229508196721312,
+      "grad_norm": 1.5188337564468384,
+      "learning_rate": 2.265241012929541e-05,
+      "loss": 0.8677475452423096,
+      "step": 594
+    },
+    {
+      "epoch": 1.6284153005464481,
+      "grad_norm": 0.49448922276496887,
+      "learning_rate": 2.28828992440162e-05,
+      "loss": 0.4782347083091736,
+      "step": 596
+    },
+    {
+      "epoch": 1.633879781420765,
+      "grad_norm": 0.259139746427536,
+      "learning_rate": 2.3111016462880873e-05,
+      "loss": 1.2155531644821167,
+      "step": 598
+    },
+    {
+      "epoch": 1.639344262295082,
+      "grad_norm": 0.24349214136600494,
+      "learning_rate": 2.333667701715578e-05,
+      "loss": 1.1115421056747437,
+      "step": 600
+    },
+    {
+      "epoch": 1.644808743169399,
+      "grad_norm": 0.36063164472579956,
+      "learning_rate": 2.3559797051007815e-05,
+      "loss": 1.0202444791793823,
+      "step": 602
+    },
+    {
+      "epoch": 1.650273224043716,
+      "grad_norm": 0.23508520424365997,
+      "learning_rate": 2.3780293652665477e-05,
+      "loss": 1.013314127922058,
+      "step": 604
+    },
+    {
+      "epoch": 1.6557377049180326,
+      "grad_norm": 0.3779641389846802,
+      "learning_rate": 2.399808488522895e-05,
+      "loss": 1.1837308406829834,
+      "step": 606
+    },
+    {
+      "epoch": 1.6612021857923498,
+      "grad_norm": 0.2397950142621994,
+      "learning_rate": 2.4213089817118078e-05,
+      "loss": 0.9537526965141296,
+      "step": 608
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.7148740887641907,
+      "learning_rate": 2.4425228552146573e-05,
+      "loss": 1.135209083557129,
+      "step": 610
+    },
+    {
+      "epoch": 1.6721311475409837,
+      "grad_norm": 2.617220401763916,
+      "learning_rate": 2.4634422259211614e-05,
+      "loss": 1.0198928117752075,
+      "step": 612
+    },
+    {
+      "epoch": 1.6775956284153004,
+      "grad_norm": 0.2389843910932541,
+      "learning_rate": 2.4840593201587626e-05,
+      "loss": 1.0087116956710815,
+      "step": 614
+    },
+    {
+      "epoch": 1.6830601092896176,
+      "grad_norm": 0.24383345246315002,
+      "learning_rate": 2.5043664765813377e-05,
+      "loss": 1.0644434690475464,
+      "step": 616
+    },
+    {
+      "epoch": 1.6885245901639343,
+      "grad_norm": 0.17243260145187378,
+      "learning_rate": 2.524356149016163e-05,
+      "loss": 0.5880939960479736,
+      "step": 618
+    },
+    {
+      "epoch": 1.6939890710382515,
+      "grad_norm": 0.32494592666625977,
+      "learning_rate": 2.544020909268085e-05,
+      "loss": 1.1145374774932861,
+      "step": 620
+    },
+    {
+      "epoch": 1.6994535519125682,
+      "grad_norm": 0.3530786633491516,
+      "learning_rate": 2.5633534498798598e-05,
+      "loss": 1.1676689386367798,
+      "step": 622
+    },
+    {
+      "epoch": 1.7049180327868854,
+      "grad_norm": 0.21238695085048676,
+      "learning_rate": 2.5823465868475985e-05,
+      "loss": 0.8080005049705505,
+      "step": 624
+    },
+    {
+      "epoch": 1.710382513661202,
+      "grad_norm": 0.22660064697265625,
+      "learning_rate": 2.60099326229037e-05,
+      "loss": 0.7797802686691284,
+      "step": 626
+    },
+    {
+      "epoch": 1.7158469945355193,
+      "grad_norm": 0.1813984215259552,
+      "learning_rate": 2.619286547072914e-05,
+      "loss": 1.1150728464126587,
+      "step": 628
+    },
+    {
+      "epoch": 1.721311475409836,
+      "grad_norm": 0.6166960597038269,
+      "learning_rate": 2.6372196433805214e-05,
+      "loss": 0.788729727268219,
+      "step": 630
+    },
+    {
+      "epoch": 1.7267759562841531,
+      "grad_norm": 0.21812567114830017,
+      "learning_rate": 2.654785887245112e-05,
+      "loss": 1.0968449115753174,
+      "step": 632
+    },
+    {
+      "epoch": 1.7322404371584699,
+      "grad_norm": 0.25673896074295044,
+      "learning_rate": 2.671978751021577e-05,
+      "loss": 1.1825098991394043,
+      "step": 634
+    },
+    {
+      "epoch": 1.737704918032787,
+      "grad_norm": 0.22896477580070496,
+      "learning_rate": 2.6887918458134622e-05,
+      "loss": 0.9839988350868225,
+      "step": 636
+    },
+    {
+      "epoch": 1.7431693989071038,
+      "grad_norm": 0.22252342104911804,
+      "learning_rate": 2.705218923847093e-05,
+      "loss": 1.037245750427246,
+      "step": 638
+    },
+    {
+      "epoch": 1.748633879781421,
+      "grad_norm": 0.724755048751831,
+      "learning_rate": 2.7212538807932576e-05,
+      "loss": 0.9484571814537048,
+      "step": 640
+    },
+    {
+      "epoch": 1.7540983606557377,
+      "grad_norm": 0.1993311047554016,
+      "learning_rate": 2.7368907580355843e-05,
+      "loss": 1.1303632259368896,
+      "step": 642
+    },
+    {
+      "epoch": 1.7595628415300546,
+      "grad_norm": 0.3169814944267273,
+      "learning_rate": 2.7521237448847734e-05,
+      "loss": 1.0198063850402832,
+      "step": 644
+    },
+    {
+      "epoch": 1.7650273224043715,
+      "grad_norm": 0.9802204370498657,
+      "learning_rate": 2.766947180737861e-05,
+      "loss": 0.7934456467628479,
+      "step": 646
+    },
+    {
+      "epoch": 1.7704918032786885,
+      "grad_norm": 0.22061200439929962,
+      "learning_rate": 2.781355557181706e-05,
+      "loss": 1.0378159284591675,
+      "step": 648
+    },
+    {
+      "epoch": 1.7759562841530054,
+      "grad_norm": 0.21627365052700043,
+      "learning_rate": 2.7953435200399262e-05,
+      "loss": 1.0567606687545776,
+      "step": 650
+    },
+    {
+      "epoch": 1.7814207650273224,
+      "grad_norm": 0.1991584300994873,
+      "learning_rate": 2.8089058713625194e-05,
+      "loss": 1.1661193370819092,
+      "step": 652
+    },
+    {
+      "epoch": 1.7868852459016393,
+      "grad_norm": 0.36394891142845154,
+      "learning_rate": 2.8220375713574307e-05,
+      "loss": 0.9987577795982361,
+      "step": 654
+    },
+    {
+      "epoch": 1.7923497267759563,
+      "grad_norm": 1.2923098802566528,
+      "learning_rate": 2.8347337402633456e-05,
+      "loss": 1.0449413061141968,
+      "step": 656
+    },
+    {
+      "epoch": 1.7978142076502732,
+      "grad_norm": 1.273789644241333,
+      "learning_rate": 2.846989660163019e-05,
+      "loss": 0.9262195825576782,
+      "step": 658
+    },
+    {
+      "epoch": 1.8032786885245902,
+      "grad_norm": 1.2179731130599976,
+      "learning_rate": 2.858800776736461e-05,
+      "loss": 1.1149400472640991,
+      "step": 660
+    },
+    {
+      "epoch": 1.8087431693989071,
+      "grad_norm": 0.2583407163619995,
+      "learning_rate": 2.87016270095333e-05,
+      "loss": 1.1010037660598755,
+      "step": 662
+    },
+    {
+      "epoch": 1.814207650273224,
+      "grad_norm": 0.29756709933280945,
+      "learning_rate": 2.8810712107039e-05,
+      "loss": 1.0471609830856323,
+      "step": 664
+    },
+    {
+      "epoch": 1.819672131147541,
+      "grad_norm": 0.24317026138305664,
+      "learning_rate": 2.8915222523680082e-05,
+      "loss": 1.5057708024978638,
+      "step": 666
+    },
+    {
+      "epoch": 1.825136612021858,
+      "grad_norm": 0.3088081479072571,
+      "learning_rate": 2.9015119423213857e-05,
+      "loss": 1.0954713821411133,
+      "step": 668
+    },
+    {
+      "epoch": 1.830601092896175,
+      "grad_norm": 0.613685667514801,
+      "learning_rate": 2.9110365683788173e-05,
+      "loss": 0.7805195450782776,
+      "step": 670
+    },
+    {
+      "epoch": 1.8360655737704918,
+      "grad_norm": 0.6538958549499512,
+      "learning_rate": 2.9200925911735956e-05,
+      "loss": 1.2720482349395752,
+      "step": 672
+    },
+    {
+      "epoch": 1.8415300546448088,
+      "grad_norm": 0.2265540212392807,
+      "learning_rate": 2.9286766454727563e-05,
+      "loss": 1.1147041320800781,
+      "step": 674
+    },
+    {
+      "epoch": 1.8469945355191257,
+      "grad_norm": 0.1871262490749359,
+      "learning_rate": 2.9367855414276073e-05,
+      "loss": 1.138785719871521,
+      "step": 676
+    },
+    {
+      "epoch": 1.8524590163934427,
+      "grad_norm": 0.47468239068984985,
+      "learning_rate": 2.9444162657590747e-05,
+      "loss": 0.32045072317123413,
+      "step": 678
+    },
+    {
+      "epoch": 1.8579234972677594,
+      "grad_norm": 0.19715717434883118,
+      "learning_rate": 2.951565982877447e-05,
+      "loss": 1.0369230508804321,
+      "step": 680
+    },
+    {
+      "epoch": 1.8633879781420766,
+      "grad_norm": 0.2893429398536682,
+      "learning_rate": 2.9582320359360864e-05,
+      "loss": 1.0589567422866821,
+      "step": 682
+    },
+    {
+      "epoch": 1.8688524590163933,
+      "grad_norm": 0.28792470693588257,
+      "learning_rate": 2.9644119478187126e-05,
+      "loss": 1.1283237934112549,
+      "step": 684
+    },
+    {
+      "epoch": 1.8743169398907105,
+      "grad_norm": 1.121538519859314,
+      "learning_rate": 2.9701034220599074e-05,
+      "loss": 0.8325943946838379,
+      "step": 686
+    },
+    {
+      "epoch": 1.8797814207650272,
+      "grad_norm": 0.2843276560306549,
+      "learning_rate": 2.975304343698483e-05,
+      "loss": 1.1573818922042847,
+      "step": 688
+    },
+    {
+      "epoch": 1.8852459016393444,
+      "grad_norm": 0.2129017412662506,
+      "learning_rate": 2.980012780063404e-05,
+      "loss": 1.1440218687057495,
+      "step": 690
+    },
+    {
+      "epoch": 1.890710382513661,
+      "grad_norm": 0.20859314501285553,
+      "learning_rate": 2.9842269814919755e-05,
+      "loss": 0.918772280216217,
+      "step": 692
+    },
+    {
+      "epoch": 1.8961748633879782,
+      "grad_norm": 2.0496702194213867,
+      "learning_rate": 2.9879453819800156e-05,
+      "loss": 1.1310737133026123,
+      "step": 694
+    },
+    {
+      "epoch": 1.901639344262295,
+      "grad_norm": 0.19168691337108612,
+      "learning_rate": 2.991166599763788e-05,
+      "loss": 1.150341510772705,
+      "step": 696
+    },
+    {
+      "epoch": 1.9071038251366121,
+      "grad_norm": 0.4647641181945801,
+      "learning_rate": 2.993889437833466e-05,
+      "loss": 1.0134280920028687,
+      "step": 698
+    },
+    {
+      "epoch": 1.9125683060109289,
+      "grad_norm": 0.22240066528320312,
+      "learning_rate": 2.9961128843779457e-05,
+      "loss": 1.093998670578003,
+      "step": 700
+    },
+    {
+      "epoch": 1.918032786885246,
+      "grad_norm": 0.21184808015823364,
+      "learning_rate": 2.9978361131608348e-05,
+      "loss": 1.0832879543304443,
+      "step": 702
+    },
+    {
+      "epoch": 1.9234972677595628,
+      "grad_norm": 0.2890876829624176,
+      "learning_rate": 2.999058483827483e-05,
+      "loss": 1.1648348569869995,
+      "step": 704
+    },
+    {
+      "epoch": 1.92896174863388,
+      "grad_norm": 0.23976802825927734,
+      "learning_rate": 2.9997795421429404e-05,
+      "loss": 1.1615331172943115,
+      "step": 706
+    },
+    {
+      "epoch": 1.9344262295081966,
+      "grad_norm": 0.21567755937576294,
+      "learning_rate": 2.9999990201607516e-05,
+      "loss": 1.2002637386322021,
+      "step": 708
+    },
+    {
+      "epoch": 1.9398907103825138,
+      "grad_norm": 0.30602723360061646,
+      "learning_rate": 2.999716836322524e-05,
+      "loss": 0.9376878142356873,
+      "step": 710
+    },
+    {
+      "epoch": 1.9453551912568305,
+      "grad_norm": 0.18824896216392517,
+      "learning_rate": 2.9989330954882366e-05,
+      "loss": 0.7084695100784302,
+      "step": 712
+    },
+    {
+      "epoch": 1.9508196721311475,
+      "grad_norm": 0.2943885028362274,
+      "learning_rate": 2.9976480888972708e-05,
+      "loss": 1.133286714553833,
+      "step": 714
+    },
+    {
+      "epoch": 1.9562841530054644,
+      "grad_norm": 0.2867864668369293,
+      "learning_rate": 2.9958622940601907e-05,
+      "loss": 0.9091066122055054,
+      "step": 716
+    },
+    {
+      "epoch": 1.9617486338797814,
+      "grad_norm": 0.22444675862789154,
+      "learning_rate": 2.9935763745812935e-05,
+      "loss": 1.1386990547180176,
+      "step": 718
+    },
+    {
+      "epoch": 1.9672131147540983,
+      "grad_norm": 0.27985692024230957,
+      "learning_rate": 2.990791179912017e-05,
+      "loss": 1.0977388620376587,
+      "step": 720
+    },
+    {
+      "epoch": 1.9726775956284153,
+      "grad_norm": 0.4131743907928467,
+      "learning_rate": 2.9875077450352817e-05,
+      "loss": 1.1935728788375854,
+      "step": 722
+    },
+    {
+      "epoch": 1.9781420765027322,
+      "grad_norm": 0.2299981564283371,
+      "learning_rate": 2.9837272900808863e-05,
+      "loss": 1.0896987915039062,
+      "step": 724
+    },
+    {
+      "epoch": 1.9836065573770492,
+      "grad_norm": 0.6982898712158203,
+      "learning_rate": 2.9794512198721092e-05,
+      "loss": 1.1530712842941284,
+      "step": 726
+    },
+    {
+      "epoch": 1.989071038251366,
+      "grad_norm": 0.4606316387653351,
+      "learning_rate": 2.9746811234036736e-05,
+      "loss": 1.090704083442688,
+      "step": 728
+    },
+    {
+      "epoch": 1.994535519125683,
+      "grad_norm": 0.3608991503715515,
+      "learning_rate": 2.9694187732512702e-05,
+      "loss": 1.044235110282898,
+      "step": 730
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.45757871866226196,
+      "learning_rate": 2.96366612491287e-05,
+      "loss": 1.1520291566848755,
+      "step": 732
+    },
+    {
+      "epoch": 2.0054644808743167,
+      "grad_norm": 0.30250194668769836,
+      "learning_rate": 2.9574253160820573e-05,
+      "loss": 0.7071864008903503,
+      "step": 734
+    },
+    {
+      "epoch": 2.010928961748634,
+      "grad_norm": 0.549813449382782,
+      "learning_rate": 2.9506986658536562e-05,
+      "loss": 0.7574299573898315,
+      "step": 736
+    },
+    {
+      "epoch": 2.0163934426229506,
+      "grad_norm": 0.4930074214935303,
+      "learning_rate": 2.9434886738619537e-05,
+      "loss": 0.7851999402046204,
+      "step": 738
+    },
+    {
+      "epoch": 2.021857923497268,
+      "grad_norm": 0.5250383615493774,
+      "learning_rate": 2.9357980193518312e-05,
+      "loss": 0.8171732425689697,
+      "step": 740
+    },
+    {
+      "epoch": 2.0273224043715845,
+      "grad_norm": 1.0003880262374878,
+      "learning_rate": 2.927629560183153e-05,
+      "loss": 0.6959396004676819,
+      "step": 742
+    },
+    {
+      "epoch": 2.0327868852459017,
+      "grad_norm": 1.62253737449646,
+      "learning_rate": 2.91898633176878e-05,
+      "loss": 0.9657958745956421,
+      "step": 744
+    },
+    {
+      "epoch": 2.0382513661202184,
+      "grad_norm": 0.3060198426246643,
+      "learning_rate": 2.909871545946603e-05,
+      "loss": 0.965573251247406,
+      "step": 746
+    },
+    {
+      "epoch": 2.0437158469945356,
+      "grad_norm": 0.36276838183403015,
+      "learning_rate": 2.9002885897860252e-05,
+      "loss": 0.971846878528595,
+      "step": 748
+    },
+    {
+      "epoch": 2.0491803278688523,
+      "grad_norm": 0.6612604260444641,
+      "learning_rate": 2.8902410243293152e-05,
+      "loss": 0.6588478088378906,
+      "step": 750
+    },
+    {
+      "epoch": 2.0546448087431695,
+      "grad_norm": 0.6954399943351746,
+      "learning_rate": 2.8797325832683208e-05,
+      "loss": 0.5652149319648743,
+      "step": 752
+    },
+    {
+      "epoch": 2.060109289617486,
+      "grad_norm": 0.6179106831550598,
+      "learning_rate": 2.868767171557021e-05,
+      "loss": 0.7592162489891052,
+      "step": 754
+    },
+    {
+      "epoch": 2.0655737704918034,
+      "grad_norm": 0.5607179999351501,
+      "learning_rate": 2.8573488639604418e-05,
+      "loss": 1.0012261867523193,
+      "step": 756
+    },
+    {
+      "epoch": 2.07103825136612,
+      "grad_norm": 0.3789028823375702,
+      "learning_rate": 2.845481903540464e-05,
+      "loss": 0.9981750249862671,
+      "step": 758
+    },
+    {
+      "epoch": 2.0765027322404372,
+      "grad_norm": 0.286051869392395,
+      "learning_rate": 2.8331707000790954e-05,
+      "loss": 0.8374607563018799,
+      "step": 760
+    },
+    {
+      "epoch": 2.081967213114754,
+      "grad_norm": 0.28156572580337524,
+      "learning_rate": 2.820419828439788e-05,
+      "loss": 0.9713491797447205,
+      "step": 762
+    },
+    {
+      "epoch": 2.087431693989071,
+      "grad_norm": 0.21401195228099823,
+      "learning_rate": 2.8072340268674133e-05,
+      "loss": 1.0316338539123535,
+      "step": 764
+    },
+    {
+      "epoch": 2.092896174863388,
+      "grad_norm": 0.3717721998691559,
+      "learning_rate": 2.793618195227521e-05,
+      "loss": 1.0349146127700806,
+      "step": 766
+    },
+    {
+      "epoch": 2.098360655737705,
+      "grad_norm": 1.2073692083358765,
+      "learning_rate": 2.779577393185539e-05,
+      "loss": 0.9263788461685181,
+      "step": 768
+    },
+    {
+      "epoch": 2.1038251366120218,
+      "grad_norm": 0.2607380747795105,
+      "learning_rate": 2.765116838326597e-05,
+      "loss": 0.9181485176086426,
+      "step": 770
+    },
+    {
+      "epoch": 2.109289617486339,
+      "grad_norm": 0.5935834646224976,
+      "learning_rate": 2.750241904216663e-05,
+      "loss": 0.8803791403770447,
+      "step": 772
+    },
+    {
+      "epoch": 2.1147540983606556,
+      "grad_norm": 0.22448551654815674,
+      "learning_rate": 2.7349581184057144e-05,
+      "loss": 0.7459222674369812,
+      "step": 774
+    },
+    {
+      "epoch": 2.120218579234973,
+      "grad_norm": 0.31990522146224976,
+      "learning_rate": 2.719271160373693e-05,
+      "loss": 0.9014583230018616,
+      "step": 776
+    },
+    {
+      "epoch": 2.1256830601092895,
+      "grad_norm": 1.4967182874679565,
+      "learning_rate": 2.703186859420002e-05,
+      "loss": 0.8457669615745544,
+      "step": 778
+    },
+    {
+      "epoch": 2.1311475409836067,
+      "grad_norm": 0.23060189187526703,
+      "learning_rate": 2.6867111924973283e-05,
+      "loss": 0.7118249535560608,
+      "step": 780
+    },
+    {
+      "epoch": 2.1366120218579234,
+      "grad_norm": 0.22133475542068481,
+      "learning_rate": 2.6698502819905935e-05,
+      "loss": 0.8637291193008423,
+      "step": 782
+    },
+    {
+      "epoch": 2.1420765027322406,
+      "grad_norm": 0.21762791275978088,
+      "learning_rate": 2.652610393441872e-05,
+      "loss": 0.9261735677719116,
+      "step": 784
+    },
+    {
+      "epoch": 2.1475409836065573,
+      "grad_norm": 0.22464466094970703,
+      "learning_rate": 2.6349979332220992e-05,
+      "loss": 0.6325854659080505,
+      "step": 786
+    },
+    {
+      "epoch": 2.1530054644808745,
+      "grad_norm": 1.0833162069320679,
+      "learning_rate": 2.6170194461504586e-05,
+      "loss": 0.7055314779281616,
+      "step": 788
+    },
+    {
+      "epoch": 2.158469945355191,
+      "grad_norm": 0.38726380467414856,
+      "learning_rate": 2.5986816130623133e-05,
+      "loss": 1.0715676546096802,
+      "step": 790
+    },
+    {
+      "epoch": 2.1639344262295084,
+      "grad_norm": 0.24775686860084534,
+      "learning_rate": 2.579991248326594e-05,
+      "loss": 0.6376872658729553,
+      "step": 792
+    },
+    {
+      "epoch": 2.169398907103825,
+      "grad_norm": 0.6708298921585083,
+      "learning_rate": 2.560955297313575e-05,
+      "loss": 0.7486907243728638,
+      "step": 794
+    },
+    {
+      "epoch": 2.1748633879781423,
+      "grad_norm": 0.2694443464279175,
+      "learning_rate": 2.5415808338139595e-05,
+      "loss": 0.8056567311286926,
+      "step": 796
+    },
+    {
+      "epoch": 2.180327868852459,
+      "grad_norm": 0.26330071687698364,
+      "learning_rate": 2.5218750574102465e-05,
+      "loss": 0.8525882363319397,
+      "step": 798
+    },
+    {
+      "epoch": 2.185792349726776,
+      "grad_norm": 1.952973484992981,
+      "learning_rate": 2.5018452908013522e-05,
+      "loss": 0.8876383304595947,
+      "step": 800
+    },
+    {
+      "epoch": 2.191256830601093,
+      "grad_norm": 0.3089567720890045,
+      "learning_rate": 2.48149897708149e-05,
+      "loss": 0.848046600818634,
+      "step": 802
+    },
+    {
+      "epoch": 2.19672131147541,
+      "grad_norm": 0.1766747534275055,
+      "learning_rate": 2.4608436769743e-05,
+      "loss": 0.8710022568702698,
+      "step": 804
+    },
+    {
+      "epoch": 2.202185792349727,
+      "grad_norm": 0.21580883860588074,
+      "learning_rate": 2.4398870660232684e-05,
+      "loss": 0.8753240704536438,
+      "step": 806
+    },
+    {
+      "epoch": 2.2076502732240435,
+      "grad_norm": 0.9658389091491699,
+      "learning_rate": 2.418636931739491e-05,
+      "loss": 0.5726881623268127,
+      "step": 808
+    },
+    {
+      "epoch": 2.2131147540983607,
+      "grad_norm": 0.22094398736953735,
+      "learning_rate": 2.3971011707078125e-05,
+      "loss": 0.6364350914955139,
+      "step": 810
+    },
+    {
+      "epoch": 2.2185792349726774,
+      "grad_norm": 0.19857634603977203,
+      "learning_rate": 2.3752877856524532e-05,
+      "loss": 0.8574658632278442,
+      "step": 812
+    },
+    {
+      "epoch": 2.2240437158469946,
+      "grad_norm": 0.20962177217006683,
+      "learning_rate": 2.353204882463168e-05,
+      "loss": 0.776292085647583,
+      "step": 814
+    },
+    {
+      "epoch": 2.2295081967213113,
+      "grad_norm": 0.23855309188365936,
+      "learning_rate": 2.330860667183101e-05,
+      "loss": 0.642566442489624,
+      "step": 816
+    },
+    {
+      "epoch": 2.2349726775956285,
+      "grad_norm": 0.24430891871452332,
+      "learning_rate": 2.308263442959396e-05,
+      "loss": 0.8425507545471191,
+      "step": 818
+    },
+    {
+      "epoch": 2.240437158469945,
+      "grad_norm": 0.6459073424339294,
+      "learning_rate": 2.2854216069577376e-05,
+      "loss": 0.9843753576278687,
+      "step": 820
+    },
+    {
+      "epoch": 2.2459016393442623,
+      "grad_norm": 0.46932026743888855,
+      "learning_rate": 2.2623436472419476e-05,
+      "loss": 0.9202935099601746,
+      "step": 822
+    },
+    {
+      "epoch": 2.251366120218579,
+      "grad_norm": 0.5952751636505127,
+      "learning_rate": 2.2390381396198102e-05,
+      "loss": 0.6778359413146973,
+      "step": 824
+    },
+    {
+      "epoch": 2.2568306010928962,
+      "grad_norm": 3.2729830741882324,
+      "learning_rate": 2.2155137444562842e-05,
+      "loss": 0.8709823489189148,
+      "step": 826
+    },
+    {
+      "epoch": 2.262295081967213,
+      "grad_norm": 0.560955822467804,
+      "learning_rate": 2.191779203455302e-05,
+      "loss": 0.6671867966651917,
+      "step": 828
+    },
+    {
+      "epoch": 2.26775956284153,
+      "grad_norm": 0.21046458184719086,
+      "learning_rate": 2.1678433364113297e-05,
+      "loss": 0.7721139788627625,
+      "step": 830
+    },
+    {
+      "epoch": 2.273224043715847,
+      "grad_norm": 2.744577646255493,
+      "learning_rate": 2.1437150379319245e-05,
+      "loss": 0.543931245803833,
+      "step": 832
+    },
+    {
+      "epoch": 2.278688524590164,
+      "grad_norm": 0.26252231001853943,
+      "learning_rate": 2.1194032741324823e-05,
+      "loss": 0.7887152433395386,
+      "step": 834
+    },
+    {
+      "epoch": 2.2841530054644807,
+      "grad_norm": 0.7648658156394958,
+      "learning_rate": 2.0949170793044142e-05,
+      "loss": 0.9888522624969482,
+      "step": 836
+    },
+    {
+      "epoch": 2.289617486338798,
+      "grad_norm": 0.8900395035743713,
+      "learning_rate": 2.070265552557985e-05,
+      "loss": 0.8767361640930176,
+      "step": 838
+    },
+    {
+      "epoch": 2.2950819672131146,
+      "grad_norm": 0.1943552941083908,
+      "learning_rate": 2.0454578544410758e-05,
+      "loss": 0.6835895776748657,
+      "step": 840
+    },
+    {
+      "epoch": 2.300546448087432,
+      "grad_norm": 0.16506916284561157,
+      "learning_rate": 2.0205032035351043e-05,
+      "loss": 0.918617844581604,
+      "step": 842
+    },
+    {
+      "epoch": 2.3060109289617485,
+      "grad_norm": 0.2222490906715393,
+      "learning_rate": 1.9954108730293875e-05,
+      "loss": 0.9415408968925476,
+      "step": 844
+    },
+    {
+      "epoch": 2.3114754098360657,
+      "grad_norm": 0.173303484916687,
+      "learning_rate": 1.9701901872752047e-05,
+      "loss": 0.4551407992839813,
+      "step": 846
+    },
+    {
+      "epoch": 2.3169398907103824,
+      "grad_norm": 0.23867173492908478,
+      "learning_rate": 1.9448505183208607e-05,
+      "loss": 0.9028974175453186,
+      "step": 848
+    },
+    {
+      "epoch": 2.3224043715846996,
+      "grad_norm": 0.23550698161125183,
+      "learning_rate": 1.919401282429013e-05,
+      "loss": 0.8313841223716736,
+      "step": 850
+    },
+    {
+      "epoch": 2.3278688524590163,
+      "grad_norm": 1.4692848920822144,
+      "learning_rate": 1.893851936577567e-05,
+      "loss": 0.7707965970039368,
+      "step": 852
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.29081717133522034,
+      "learning_rate": 1.868211974945461e-05,
+      "loss": 0.8993242979049683,
+      "step": 854
+    },
+    {
+      "epoch": 2.33879781420765,
+      "grad_norm": 0.21171832084655762,
+      "learning_rate": 1.842490925384604e-05,
+      "loss": 0.8694086074829102,
+      "step": 856
+    },
+    {
+      "epoch": 2.3442622950819674,
+      "grad_norm": 0.2612394392490387,
+      "learning_rate": 1.816698345879313e-05,
+      "loss": 0.5785654783248901,
+      "step": 858
+    },
+    {
+      "epoch": 2.349726775956284,
+      "grad_norm": 0.1895606517791748,
+      "learning_rate": 1.790843820994548e-05,
+      "loss": 0.6962811350822449,
+      "step": 860
+    },
+    {
+      "epoch": 2.3551912568306013,
+      "grad_norm": 0.19078537821769714,
+      "learning_rate": 1.7649369583142763e-05,
+      "loss": 0.8344950675964355,
+      "step": 862
+    },
+    {
+      "epoch": 2.360655737704918,
+      "grad_norm": 0.20540055632591248,
+      "learning_rate": 1.738987384871274e-05,
+      "loss": 0.5912719368934631,
+      "step": 864
+    },
+    {
+      "epoch": 2.366120218579235,
+      "grad_norm": 0.5205765962600708,
+      "learning_rate": 1.7130047435697118e-05,
+      "loss": 1.2037533521652222,
+      "step": 866
+    },
+    {
+      "epoch": 2.371584699453552,
+      "grad_norm": 0.16610278189182281,
+      "learning_rate": 1.6869986896018226e-05,
+      "loss": 0.8778015375137329,
+      "step": 868
+    },
+    {
+      "epoch": 2.3770491803278686,
+      "grad_norm": 0.21313391625881195,
+      "learning_rate": 1.66097888686003e-05,
+      "loss": 0.6815759539604187,
+      "step": 870
+    },
+    {
+      "epoch": 2.3825136612021858,
+      "grad_norm": 0.19833733141422272,
+      "learning_rate": 1.6349550043458252e-05,
+      "loss": 0.7003133893013,
+      "step": 872
+    },
+    {
+      "epoch": 2.387978142076503,
+      "grad_norm": 0.24588751792907715,
+      "learning_rate": 1.608936712576749e-05,
+      "loss": 0.8853883147239685,
+      "step": 874
+    },
+    {
+      "epoch": 2.3934426229508197,
+      "grad_norm": 1.4030362367630005,
+      "learning_rate": 1.582933679992809e-05,
+      "loss": 0.8193628191947937,
+      "step": 876
+    },
+    {
+      "epoch": 2.3989071038251364,
+      "grad_norm": 0.25144752860069275,
+      "learning_rate": 1.556955569363678e-05,
+      "loss": 0.7706636190414429,
+      "step": 878
+    },
+    {
+      "epoch": 2.4043715846994536,
+      "grad_norm": 0.20421741902828217,
+      "learning_rate": 1.531012034197988e-05,
+      "loss": 0.9738537073135376,
+      "step": 880
+    },
+    {
+      "epoch": 2.4098360655737707,
+      "grad_norm": 0.32910487055778503,
+      "learning_rate": 1.5051127151560745e-05,
+      "loss": 0.4443007707595825,
+      "step": 882
+    },
+    {
+      "epoch": 2.4153005464480874,
+      "grad_norm": 0.2254200130701065,
+      "learning_rate": 1.4792672364674816e-05,
+      "loss": 0.7611709237098694,
+      "step": 884
+    },
+    {
+      "epoch": 2.420765027322404,
+      "grad_norm": 0.2349507212638855,
+      "learning_rate": 1.4534852023545968e-05,
+      "loss": 0.7841254472732544,
+      "step": 886
+    },
+    {
+      "epoch": 2.4262295081967213,
+      "grad_norm": 0.5601933002471924,
+      "learning_rate": 1.4277761934636963e-05,
+      "loss": 0.631168007850647,
+      "step": 888
+    },
+    {
+      "epoch": 2.431693989071038,
+      "grad_norm": 0.3481084406375885,
+      "learning_rate": 1.4021497633047664e-05,
+      "loss": 0.5096313953399658,
+      "step": 890
+    },
+    {
+      "epoch": 2.4371584699453552,
+      "grad_norm": 0.5794140696525574,
+      "learning_rate": 1.3766154347013933e-05,
+      "loss": 0.8481501936912537,
+      "step": 892
+    },
+    {
+      "epoch": 2.442622950819672,
+      "grad_norm": 0.248020201921463,
+      "learning_rate": 1.3511826962520809e-05,
+      "loss": 0.4762151539325714,
+      "step": 894
+    },
+    {
+      "epoch": 2.448087431693989,
+      "grad_norm": 0.27063384652137756,
+      "learning_rate": 1.3258609988042627e-05,
+      "loss": 0.6739501357078552,
+      "step": 896
+    },
+    {
+      "epoch": 2.453551912568306,
+      "grad_norm": 0.2560213506221771,
+      "learning_rate": 1.300659751942353e-05,
+      "loss": 0.7505039572715759,
+      "step": 898
+    },
+    {
+      "epoch": 2.459016393442623,
+      "grad_norm": 0.27310049533843994,
+      "learning_rate": 1.2755883204911305e-05,
+      "loss": 0.8457780480384827,
+      "step": 900
+    },
+    {
+      "epoch": 2.4644808743169397,
+      "grad_norm": 0.22151170670986176,
+      "learning_rate": 1.2506560210357541e-05,
+      "loss": 0.7800691723823547,
+      "step": 902
+    },
+    {
+      "epoch": 2.469945355191257,
+      "grad_norm": 0.22379451990127563,
+      "learning_rate": 1.225872118459706e-05,
+      "loss": 0.74601811170578,
+      "step": 904
+    },
+    {
+      "epoch": 2.4754098360655736,
+      "grad_norm": 0.47037237882614136,
+      "learning_rate": 1.2012458225019375e-05,
+      "loss": 0.9208009839057922,
+      "step": 906
+    },
+    {
+      "epoch": 2.480874316939891,
+      "grad_norm": 0.14209216833114624,
+      "learning_rate": 1.176786284334528e-05,
+      "loss": 0.3645989000797272,
+      "step": 908
+    },
+    {
+      "epoch": 2.4863387978142075,
+      "grad_norm": 0.13509748876094818,
+      "learning_rate": 1.1525025931620855e-05,
+      "loss": 0.6787592172622681,
+      "step": 910
+    },
+    {
+      "epoch": 2.4918032786885247,
+      "grad_norm": 0.18052458763122559,
+      "learning_rate": 1.1284037728441877e-05,
+      "loss": 0.7986164093017578,
+      "step": 912
+    },
+    {
+      "epoch": 2.4972677595628414,
+      "grad_norm": 0.22115804255008698,
+      "learning_rate": 1.1044987785420924e-05,
+      "loss": 0.8646562099456787,
+      "step": 914
+    },
+    {
+      "epoch": 2.5027322404371586,
+      "grad_norm": 0.2403498888015747,
+      "learning_rate": 1.0807964933909975e-05,
+      "loss": 0.737191379070282,
+      "step": 916
+    },
+    {
+      "epoch": 2.5081967213114753,
+      "grad_norm": 0.29979637265205383,
+      "learning_rate": 1.0573057251990443e-05,
+      "loss": 0.8144265413284302,
+      "step": 918
+    },
+    {
+      "epoch": 2.5136612021857925,
+      "grad_norm": 0.18022961914539337,
+      "learning_rate": 1.0340352031743256e-05,
+      "loss": 0.8598709106445312,
+      "step": 920
+    },
+    {
+      "epoch": 2.519125683060109,
+      "grad_norm": 0.22041092813014984,
+      "learning_rate": 1.010993574681095e-05,
+      "loss": 0.7569805383682251,
+      "step": 922
+    },
+    {
+      "epoch": 2.5245901639344264,
+      "grad_norm": 0.2608581781387329,
+      "learning_rate": 9.881894020263938e-06,
+      "loss": 0.5048292875289917,
+      "step": 924
+    },
+    {
+      "epoch": 2.530054644808743,
+      "grad_norm": 0.3491852283477783,
+      "learning_rate": 9.656311592782831e-06,
+      "loss": 0.6451077461242676,
+      "step": 926
+    },
+    {
+      "epoch": 2.5355191256830603,
+      "grad_norm": 0.3480936288833618,
+      "learning_rate": 9.433272291168689e-06,
+      "loss": 0.6237382292747498,
+      "step": 928
+    },
+    {
+      "epoch": 2.540983606557377,
+      "grad_norm": 0.29056069254875183,
+      "learning_rate": 9.212858997192744e-06,
+      "loss": 0.93817138671875,
+      "step": 930
+    },
+    {
+      "epoch": 2.546448087431694,
+      "grad_norm": 0.3035399317741394,
+      "learning_rate": 8.995153616797544e-06,
+      "loss": 0.6400088667869568,
+      "step": 932
+    },
+    {
+      "epoch": 2.551912568306011,
+      "grad_norm": 0.3459092974662781,
+      "learning_rate": 8.78023704966047e-06,
+      "loss": 0.8244706988334656,
+      "step": 934
+    },
+    {
+      "epoch": 2.557377049180328,
+      "grad_norm": 0.38768166303634644,
+      "learning_rate": 8.568189159131336e-06,
+      "loss": 0.724708616733551,
+      "step": 936
+    },
+    {
+      "epoch": 2.5628415300546448,
+      "grad_norm": 1.209683895111084,
+      "learning_rate": 8.359088742554941e-06,
+      "loss": 0.5176321268081665,
+      "step": 938
+    },
+    {
+      "epoch": 2.5683060109289615,
+      "grad_norm": 0.33049488067626953,
+      "learning_rate": 8.15301350198999e-06,
+      "loss": 0.8478241562843323,
+      "step": 940
+    },
+    {
+      "epoch": 2.5737704918032787,
+      "grad_norm": 0.7910704612731934,
+      "learning_rate": 7.950040015334789e-06,
+      "loss": 0.5429540276527405,
+      "step": 942
+    },
+    {
+      "epoch": 2.579234972677596,
+      "grad_norm": 0.21583837270736694,
+      "learning_rate": 7.750243707870748e-06,
+      "loss": 0.8436353802680969,
+      "step": 944
+    },
+    {
+      "epoch": 2.5846994535519126,
+      "grad_norm": 0.3202984631061554,
+      "learning_rate": 7.553698824234314e-06,
+      "loss": 0.25999099016189575,
+      "step": 946
+    },
+    {
+      "epoch": 2.5901639344262293,
+      "grad_norm": 0.2738170623779297,
+      "learning_rate": 7.360478400827475e-06,
+      "loss": 0.6493714451789856,
+      "step": 948
+    },
+    {
+      "epoch": 2.5956284153005464,
+      "grad_norm": 0.19985666871070862,
+      "learning_rate": 7.170654238677331e-06,
+      "loss": 0.7574492692947388,
+      "step": 950
+    },
+    {
+      "epoch": 2.6010928961748636,
+      "grad_norm": 0.24671149253845215,
+      "learning_rate": 6.984296876754711e-06,
+      "loss": 0.8211013674736023,
+      "step": 952
+    },
+    {
+      "epoch": 2.6065573770491803,
+      "grad_norm": 0.1873648315668106,
+      "learning_rate": 6.801475565761783e-06,
+      "loss": 0.43293893337249756,
+      "step": 954
+    },
+    {
+      "epoch": 2.612021857923497,
+      "grad_norm": 0.17501504719257355,
+      "learning_rate": 6.622258242398371e-06,
+      "loss": 0.27505725622177124,
+      "step": 956
+    },
+    {
+      "epoch": 2.6174863387978142,
+      "grad_norm": 0.22481496632099152,
+      "learning_rate": 6.4467115041165855e-06,
+      "loss": 0.4883240759372711,
+      "step": 958
+    },
+    {
+      "epoch": 2.6229508196721314,
+      "grad_norm": 0.21473781764507294,
+      "learning_rate": 6.2749005843730336e-06,
+      "loss": 0.654322624206543,
+      "step": 960
+    },
+    {
+      "epoch": 2.628415300546448,
+      "grad_norm": 0.5412459969520569,
+      "learning_rate": 6.106889328388064e-06,
+      "loss": 0.6429625153541565,
+      "step": 962
+    },
+    {
+      "epoch": 2.633879781420765,
+      "grad_norm": 0.4350409209728241,
+      "learning_rate": 5.942740169420701e-06,
+      "loss": 0.9000084400177002,
+      "step": 964
+    },
+    {
+      "epoch": 2.639344262295082,
+      "grad_norm": 0.23633772134780884,
+      "learning_rate": 5.7825141055683895e-06,
+      "loss": 0.7249687314033508,
+      "step": 966
+    },
+    {
+      "epoch": 2.644808743169399,
+      "grad_norm": 0.31932181119918823,
+      "learning_rate": 5.62627067709992e-06,
+      "loss": 0.692674458026886,
+      "step": 968
+    },
+    {
+      "epoch": 2.650273224043716,
+      "grad_norm": 0.4318002462387085,
+      "learning_rate": 5.474067944330285e-06,
+      "loss": 0.7996351718902588,
+      "step": 970
+    },
+    {
+      "epoch": 2.6557377049180326,
+      "grad_norm": 0.23726297914981842,
+      "learning_rate": 5.325962466045282e-06,
+      "loss": 0.9792457818984985,
+      "step": 972
+    },
+    {
+      "epoch": 2.66120218579235,
+      "grad_norm": 0.19819673895835876,
+      "learning_rate": 5.18200927848421e-06,
+      "loss": 0.6123301982879639,
+      "step": 974
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.3812166452407837,
+      "learning_rate": 5.042261874888308e-06,
+      "loss": 0.7370781302452087,
+      "step": 976
+    },
+    {
+      "epoch": 2.6721311475409837,
+      "grad_norm": 0.17140597105026245,
+      "learning_rate": 4.906772185622572e-06,
+      "loss": 0.7503936886787415,
+      "step": 978
+    },
+    {
+      "epoch": 2.6775956284153004,
+      "grad_norm": 0.18407227098941803,
+      "learning_rate": 4.775590558878368e-06,
+      "loss": 0.7234741449356079,
+      "step": 980
+    },
+    {
+      "epoch": 2.6830601092896176,
+      "grad_norm": 0.21530817449092865,
+      "learning_rate": 4.648765741963903e-06,
+      "loss": 0.7761690616607666,
+      "step": 982
+    },
+    {
+      "epoch": 2.6885245901639343,
+      "grad_norm": 0.21703749895095825,
+      "learning_rate": 4.526344863189724e-06,
+      "loss": 0.6027669906616211,
+      "step": 984
+    },
+    {
+      "epoch": 2.6939890710382515,
+      "grad_norm": 0.26828664541244507,
+      "learning_rate": 4.408373414355714e-06,
+      "loss": 0.9321015477180481,
+      "step": 986
+    },
+    {
+      "epoch": 2.699453551912568,
+      "grad_norm": 0.20866970717906952,
+      "learning_rate": 4.29489523384628e-06,
+      "loss": 0.8720956444740295,
+      "step": 988
+    },
+    {
+      "epoch": 2.7049180327868854,
+      "grad_norm": 0.2858516573905945,
+      "learning_rate": 4.185952490339899e-06,
+      "loss": 0.7041909098625183,
+      "step": 990
+    },
+    {
+      "epoch": 2.710382513661202,
+      "grad_norm": 0.2261834740638733,
+      "learning_rate": 4.081585667139231e-06,
+      "loss": 0.45303869247436523,
+      "step": 992
+    },
+    {
+      "epoch": 2.7158469945355193,
+      "grad_norm": 0.1847374141216278,
+      "learning_rate": 3.981833547127413e-06,
+      "loss": 0.8106403946876526,
+      "step": 994
+    },
+    {
+      "epoch": 2.721311475409836,
+      "grad_norm": 0.20215702056884766,
+      "learning_rate": 3.886733198356298e-06,
+      "loss": 0.8668434619903564,
+      "step": 996
+    },
+    {
+      "epoch": 2.726775956284153,
+      "grad_norm": 0.15933895111083984,
+      "learning_rate": 3.7963199602718717e-06,
+      "loss": 0.6896766424179077,
+      "step": 998
+    },
+    {
+      "epoch": 2.73224043715847,
+      "grad_norm": 0.7066519260406494,
+      "learning_rate": 3.7106274305821034e-06,
+      "loss": 0.99676513671875,
+      "step": 1000
+    },
+    {
+      "epoch": 2.737704918032787,
+      "grad_norm": 0.24640890955924988,
+      "learning_rate": 3.6296874527719515e-06,
+      "loss": 0.7258309125900269,
+      "step": 1002
+    },
+    {
+      "epoch": 2.7431693989071038,
+      "grad_norm": 0.48880383372306824,
+      "learning_rate": 3.553530104270281e-06,
+      "loss": 0.6444882750511169,
+      "step": 1004
+    },
+    {
+      "epoch": 2.748633879781421,
+      "grad_norm": 0.7180402278900146,
+      "learning_rate": 3.4821836852730384e-06,
+      "loss": 0.30940499901771545,
+      "step": 1006
+    },
+    {
+      "epoch": 2.7540983606557377,
+      "grad_norm": 0.25939175486564636,
+      "learning_rate": 3.41567470822686e-06,
+      "loss": 0.9526011943817139,
+      "step": 1008
+    },
+    {
+      "epoch": 2.7595628415300544,
+      "grad_norm": 0.3068941533565521,
+      "learning_rate": 3.354027887976989e-06,
+      "loss": 0.9272940158843994,
+      "step": 1010
+    },
+    {
+      "epoch": 2.7650273224043715,
+      "grad_norm": 0.298452764749527,
+      "learning_rate": 3.297266132583221e-06,
+      "loss": 0.6733279228210449,
+      "step": 1012
+    },
+    {
+      "epoch": 2.7704918032786887,
+      "grad_norm": 0.253140389919281,
+      "learning_rate": 3.245410534807195e-06,
+      "loss": 0.9813092947006226,
+      "step": 1014
+    },
+    {
+      "epoch": 2.7759562841530054,
+      "grad_norm": 0.38031837344169617,
+      "learning_rate": 3.1984803642743314e-06,
+      "loss": 0.6480425596237183,
+      "step": 1016
+    },
+    {
+      "epoch": 2.781420765027322,
+      "grad_norm": 0.9425625205039978,
+      "learning_rate": 3.1564930603131777e-06,
+      "loss": 0.9104164838790894,
+      "step": 1018
+    },
+    {
+      "epoch": 2.7868852459016393,
+      "grad_norm": 0.33443892002105713,
+      "learning_rate": 3.1194642254749395e-06,
+      "loss": 0.5776557326316833,
+      "step": 1020
+    },
+    {
+      "epoch": 2.7923497267759565,
+      "grad_norm": 0.2240370362997055,
+      "learning_rate": 3.0874076197355317e-06,
+      "loss": 0.8249503970146179,
+      "step": 1022
+    },
+    {
+      "epoch": 2.797814207650273,
+      "grad_norm": 0.1996547132730484,
+      "learning_rate": 3.0603351553823717e-06,
+      "loss": 0.4351382553577423,
+      "step": 1024
+    },
+    {
+      "epoch": 2.80327868852459,
+      "grad_norm": 0.23715117573738098,
+      "learning_rate": 3.038256892587734e-06,
+      "loss": 0.9221212267875671,
+      "step": 1026
+    },
+    {
+      "epoch": 2.808743169398907,
+      "grad_norm": 0.6460992693901062,
+      "learning_rate": 3.0211810356703803e-06,
+      "loss": 0.6869730353355408,
+      "step": 1028
+    },
+    {
+      "epoch": 2.8142076502732243,
+      "grad_norm": 0.29128873348236084,
+      "learning_rate": 3.0091139300468266e-06,
+      "loss": 0.5977322459220886,
+      "step": 1030
+    },
+    {
+      "epoch": 2.819672131147541,
+      "grad_norm": 0.2111499011516571,
+      "learning_rate": 3.0020600598733656e-06,
+      "loss": 0.6241099834442139,
+      "step": 1032
+    },
+    {
+      "epoch": 2.8251366120218577,
+      "grad_norm": 0.3315550684928894,
+      "learning_rate": 3.000022046379753e-06,
+      "loss": 0.7141521573066711,
+      "step": 1034
+    },
+    {
+      "epoch": 2.830601092896175,
+      "grad_norm": 0.30943799018859863,
+      "learning_rate": 3.0030006468951557e-06,
+      "loss": 0.9645689725875854,
+      "step": 1036
+    },
+    {
+      "epoch": 2.836065573770492,
+      "grad_norm": 0.17847982048988342,
+      "learning_rate": 3.0109947545667246e-06,
+      "loss": 0.6744089126586914,
+      "step": 1038
+    },
+    {
+      "epoch": 2.841530054644809,
+      "grad_norm": 0.24691838026046753,
+      "learning_rate": 3.024001398770901e-06,
+      "loss": 0.801510751247406,
+      "step": 1040
+    },
+    {
+      "epoch": 2.8469945355191255,
+      "grad_norm": 0.21006713807582855,
+      "learning_rate": 3.042015746217308e-06,
+      "loss": 0.8129590749740601,
+      "step": 1042
+    },
+    {
+      "epoch": 2.8524590163934427,
+      "grad_norm": 0.26454004645347595,
+      "learning_rate": 3.0650311027448116e-06,
+      "loss": 0.6099104881286621,
+      "step": 1044
+    },
+    {
+      "epoch": 2.8579234972677594,
+      "grad_norm": 0.48515036702156067,
+      "learning_rate": 3.0930389158090754e-06,
+      "loss": 0.9171488881111145,
+      "step": 1046
+    },
+    {
+      "epoch": 2.8633879781420766,
+      "grad_norm": 0.14590629935264587,
+      "learning_rate": 3.1260287776607025e-06,
+      "loss": 0.3860487639904022,
+      "step": 1048
+    },
+    {
+      "epoch": 2.8688524590163933,
+      "grad_norm": 0.2847556471824646,
+      "learning_rate": 3.163988429212773e-06,
+      "loss": 0.7123141288757324,
+      "step": 1050
+    },
+    {
+      "epoch": 2.8743169398907105,
+      "grad_norm": 0.21968185901641846,
+      "learning_rate": 3.206903764596349e-06,
+      "loss": 0.7720679640769958,
+      "step": 1052
+    },
+    {
+      "epoch": 2.879781420765027,
+      "grad_norm": 0.24895037710666656,
+      "learning_rate": 3.254758836402225e-06,
+      "loss": 0.9379729628562927,
+      "step": 1054
+    },
+    {
+      "epoch": 2.8852459016393444,
+      "grad_norm": 1.1485183238983154,
+      "learning_rate": 3.3075358616070144e-06,
+      "loss": 0.6869798898696899,
+      "step": 1056
+    },
+    {
+      "epoch": 2.890710382513661,
+      "grad_norm": 2.465815305709839,
+      "learning_rate": 3.365215228181358e-06,
+      "loss": 0.594801664352417,
+      "step": 1058
+    },
+    {
+      "epoch": 2.8961748633879782,
+      "grad_norm": 0.19229429960250854,
+      "learning_rate": 3.4277755023777795e-06,
+      "loss": 0.7733983397483826,
+      "step": 1060
+    },
+    {
+      "epoch": 2.901639344262295,
+      "grad_norm": 0.20262494683265686,
+      "learning_rate": 3.495193436695504e-06,
+      "loss": 0.7548971176147461,
+      "step": 1062
+    },
+    {
+      "epoch": 2.907103825136612,
+      "grad_norm": 0.3187836706638336,
+      "learning_rate": 3.567443978519267e-06,
+      "loss": 0.7963505983352661,
+      "step": 1064
+    },
+    {
+      "epoch": 2.912568306010929,
+      "grad_norm": 0.22692307829856873,
+      "learning_rate": 3.6445002794288992e-06,
+      "loss": 0.4752350449562073,
+      "step": 1066
+    },
+    {
+      "epoch": 2.918032786885246,
+      "grad_norm": 0.2349584698677063,
+      "learning_rate": 3.7263337051762718e-06,
+      "loss": 0.649222731590271,
+      "step": 1068
+    },
+    {
+      "epoch": 2.9234972677595628,
+      "grad_norm": 0.5317696928977966,
+      "learning_rate": 3.8129138463257943e-06,
+      "loss": 0.7322983145713806,
+      "step": 1070
+    },
+    {
+      "epoch": 2.92896174863388,
+      "grad_norm": 0.24103935062885284,
+      "learning_rate": 3.904208529554625e-06,
+      "loss": 0.9223740100860596,
+      "step": 1072
+    },
+    {
+      "epoch": 2.9344262295081966,
+      "grad_norm": 0.2071618139743805,
+      "learning_rate": 4.000183829608332e-06,
+      "loss": 0.47986942529678345,
+      "step": 1074
+    },
+    {
+      "epoch": 2.939890710382514,
+      "grad_norm": 0.3855530917644501,
+      "learning_rate": 4.100804081907595e-06,
+      "loss": 0.5404314398765564,
+      "step": 1076
+    },
+    {
+      "epoch": 2.9453551912568305,
+      "grad_norm": 0.9493356347084045,
+      "learning_rate": 4.206031895801176e-06,
+      "loss": 0.547483503818512,
+      "step": 1078
+    },
+    {
+      "epoch": 2.9508196721311473,
+      "grad_norm": 0.3453940153121948,
+      "learning_rate": 4.315828168460367e-06,
+      "loss": 0.932254433631897,
+      "step": 1080
+    },
+    {
+      "epoch": 2.9562841530054644,
+      "grad_norm": 0.19494763016700745,
+      "learning_rate": 4.430152099409704e-06,
+      "loss": 0.5538932085037231,
+      "step": 1082
+    },
+    {
+      "epoch": 2.9617486338797816,
+      "grad_norm": 0.2678419053554535,
+      "learning_rate": 4.548961205688424e-06,
+      "loss": 0.7946881651878357,
+      "step": 1084
+    },
+    {
+      "epoch": 2.9672131147540983,
+      "grad_norm": 0.45159897208213806,
+      "learning_rate": 4.672211337637246e-06,
+      "loss": 0.5122095942497253,
+      "step": 1086
+    },
+    {
+      "epoch": 2.972677595628415,
+      "grad_norm": 0.3175777196884155,
+      "learning_rate": 4.7998566953044445e-06,
+      "loss": 1.1099841594696045,
+      "step": 1088
+    },
+    {
+      "epoch": 2.978142076502732,
+      "grad_norm": 0.21449358761310577,
+      "learning_rate": 4.931849845465193e-06,
+      "loss": 0.6492303013801575,
+      "step": 1090
+    },
+    {
+      "epoch": 2.9836065573770494,
+      "grad_norm": 0.25702965259552,
+      "learning_rate": 5.06814173924782e-06,
+      "loss": 0.7961921095848083,
+      "step": 1092
+    },
+    {
+      "epoch": 2.989071038251366,
+      "grad_norm": 0.22738364338874817,
+      "learning_rate": 5.208681730360458e-06,
+      "loss": 0.7180767059326172,
+      "step": 1094
+    },
+    {
+      "epoch": 2.994535519125683,
+      "grad_norm": 1.2310293912887573,
+      "learning_rate": 5.3534175939112694e-06,
+      "loss": 0.590616762638092,
+      "step": 1096
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.243174210190773,
+      "learning_rate": 5.50229554581536e-06,
+      "loss": 0.7194076180458069,
+      "step": 1098
+    },
+    {
+      "epoch": 3.0,
+      "step": 1098,
+      "total_flos": 4.931331991116186e+18,
+      "train_loss": 0.9604242436981375,
+      "train_runtime": 12050.7893,
+      "train_samples_per_second": 5.467,
+      "train_steps_per_second": 0.091
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 1098,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 99999,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.931331991116186e+18,
+  "train_batch_size": 3,
+  "trial_name": null,
+  "trial_params": null
+}