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": 0.9973115921020508,
+      "learning_rate": 5.454545454545455e-07,
+      "loss": 2.5618793964385986,
+      "step": 2
+    },
+    {
+      "epoch": 0.01092896174863388,
+      "grad_norm": 1.3840559720993042,
+      "learning_rate": 1.6363636363636363e-06,
+      "loss": 1.9354113340377808,
+      "step": 4
+    },
+    {
+      "epoch": 0.01639344262295082,
+      "grad_norm": 2.0087080001831055,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 1.8657492399215698,
+      "step": 6
+    },
+    {
+      "epoch": 0.02185792349726776,
+      "grad_norm": 0.9125573635101318,
+      "learning_rate": 3.818181818181818e-06,
+      "loss": 1.7251954078674316,
+      "step": 8
+    },
+    {
+      "epoch": 0.0273224043715847,
+      "grad_norm": 0.5344090461730957,
+      "learning_rate": 4.90909090909091e-06,
+      "loss": 1.6777604818344116,
+      "step": 10
+    },
+    {
+      "epoch": 0.03278688524590164,
+      "grad_norm": 0.604454755783081,
+      "learning_rate": 6e-06,
+      "loss": 1.5453071594238281,
+      "step": 12
+    },
+    {
+      "epoch": 0.03825136612021858,
+      "grad_norm": 0.7670559883117676,
+      "learning_rate": 7.090909090909091e-06,
+      "loss": 1.4508954286575317,
+      "step": 14
+    },
+    {
+      "epoch": 0.04371584699453552,
+      "grad_norm": 0.4149356782436371,
+      "learning_rate": 8.181818181818181e-06,
+      "loss": 1.1633912324905396,
+      "step": 16
+    },
+    {
+      "epoch": 0.04918032786885246,
+      "grad_norm": 0.7089836001396179,
+      "learning_rate": 9.272727272727273e-06,
+      "loss": 1.164050579071045,
+      "step": 18
+    },
+    {
+      "epoch": 0.0546448087431694,
+      "grad_norm": 0.39713212847709656,
+      "learning_rate": 1.0363636363636364e-05,
+      "loss": 1.3879873752593994,
+      "step": 20
+    },
+    {
+      "epoch": 0.060109289617486336,
+      "grad_norm": 0.5368804335594177,
+      "learning_rate": 1.1454545454545455e-05,
+      "loss": 1.3274301290512085,
+      "step": 22
+    },
+    {
+      "epoch": 0.06557377049180328,
+      "grad_norm": 0.30250489711761475,
+      "learning_rate": 1.2545454545454545e-05,
+      "loss": 1.385084867477417,
+      "step": 24
+    },
+    {
+      "epoch": 0.07103825136612021,
+      "grad_norm": 0.9990394115447998,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 1.5793037414550781,
+      "step": 26
+    },
+    {
+      "epoch": 0.07650273224043716,
+      "grad_norm": 0.3947621285915375,
+      "learning_rate": 1.4727272727272728e-05,
+      "loss": 1.3561811447143555,
+      "step": 28
+    },
+    {
+      "epoch": 0.08196721311475409,
+      "grad_norm": 0.49388208985328674,
+      "learning_rate": 1.5818181818181818e-05,
+      "loss": 1.273718237876892,
+      "step": 30
+    },
+    {
+      "epoch": 0.08743169398907104,
+      "grad_norm": 0.32329627871513367,
+      "learning_rate": 1.6909090909090907e-05,
+      "loss": 1.3053655624389648,
+      "step": 32
+    },
+    {
+      "epoch": 0.09289617486338798,
+      "grad_norm": 0.718172013759613,
+      "learning_rate": 1.8e-05,
+      "loss": 1.0134309530258179,
+      "step": 34
+    },
+    {
+      "epoch": 0.09836065573770492,
+      "grad_norm": 0.27196696400642395,
+      "learning_rate": 1.909090909090909e-05,
+      "loss": 1.1276304721832275,
+      "step": 36
+    },
+    {
+      "epoch": 0.10382513661202186,
+      "grad_norm": 0.4729017913341522,
+      "learning_rate": 2.0181818181818183e-05,
+      "loss": 1.3713783025741577,
+      "step": 38
+    },
+    {
+      "epoch": 0.1092896174863388,
+      "grad_norm": 0.337189644575119,
+      "learning_rate": 2.1272727272727273e-05,
+      "loss": 1.3217278718948364,
+      "step": 40
+    },
+    {
+      "epoch": 0.11475409836065574,
+      "grad_norm": 0.8202743530273438,
+      "learning_rate": 2.2363636363636366e-05,
+      "loss": 1.0108166933059692,
+      "step": 42
+    },
+    {
+      "epoch": 0.12021857923497267,
+      "grad_norm": 0.3670739233493805,
+      "learning_rate": 2.3454545454545456e-05,
+      "loss": 0.988700807094574,
+      "step": 44
+    },
+    {
+      "epoch": 0.12568306010928962,
+      "grad_norm": 0.3262913227081299,
+      "learning_rate": 2.454545454545455e-05,
+      "loss": 1.3549995422363281,
+      "step": 46
+    },
+    {
+      "epoch": 0.13114754098360656,
+      "grad_norm": 0.3616645336151123,
+      "learning_rate": 2.5636363636363635e-05,
+      "loss": 1.3315670490264893,
+      "step": 48
+    },
+    {
+      "epoch": 0.1366120218579235,
+      "grad_norm": 0.3672925531864166,
+      "learning_rate": 2.6727272727272728e-05,
+      "loss": 1.3015525341033936,
+      "step": 50
+    },
+    {
+      "epoch": 0.14207650273224043,
+      "grad_norm": 0.42094483971595764,
+      "learning_rate": 2.7818181818181818e-05,
+      "loss": 1.2684640884399414,
+      "step": 52
+    },
+    {
+      "epoch": 0.14754098360655737,
+      "grad_norm": 2.618896722793579,
+      "learning_rate": 2.890909090909091e-05,
+      "loss": 0.8703136444091797,
+      "step": 54
+    },
+    {
+      "epoch": 0.15300546448087432,
+      "grad_norm": 0.3619275391101837,
+      "learning_rate": 3e-05,
+      "loss": 1.4717642068862915,
+      "step": 56
+    },
+    {
+      "epoch": 0.15846994535519127,
+      "grad_norm": 0.3370303213596344,
+      "learning_rate": 2.9997491688899256e-05,
+      "loss": 1.451904535293579,
+      "step": 58
+    },
+    {
+      "epoch": 0.16393442622950818,
+      "grad_norm": 0.35009288787841797,
+      "learning_rate": 2.998996768768956e-05,
+      "loss": 1.3383069038391113,
+      "step": 60
+    },
+    {
+      "epoch": 0.16939890710382513,
+      "grad_norm": 0.4444884955883026,
+      "learning_rate": 2.9977430792302124e-05,
+      "loss": 1.5286966562271118,
+      "step": 62
+    },
+    {
+      "epoch": 0.17486338797814208,
+      "grad_norm": 0.7974561452865601,
+      "learning_rate": 2.9959885661467903e-05,
+      "loss": 1.2279592752456665,
+      "step": 64
+    },
+    {
+      "epoch": 0.18032786885245902,
+      "grad_norm": 0.3576641082763672,
+      "learning_rate": 2.993733881498636e-05,
+      "loss": 1.3408796787261963,
+      "step": 66
+    },
+    {
+      "epoch": 0.18579234972677597,
+      "grad_norm": 0.2858317494392395,
+      "learning_rate": 2.9909798631302736e-05,
+      "loss": 1.3054747581481934,
+      "step": 68
+    },
+    {
+      "epoch": 0.1912568306010929,
+      "grad_norm": 0.6741045117378235,
+      "learning_rate": 2.987727534439457e-05,
+      "loss": 1.3402438163757324,
+      "step": 70
+    },
+    {
+      "epoch": 0.19672131147540983,
+      "grad_norm": 4.621718406677246,
+      "learning_rate": 2.983978103996877e-05,
+      "loss": 1.189091444015503,
+      "step": 72
+    },
+    {
+      "epoch": 0.20218579234972678,
+      "grad_norm": 0.3123322129249573,
+      "learning_rate": 2.9797329650970525e-05,
+      "loss": 1.1878712177276611,
+      "step": 74
+    },
+    {
+      "epoch": 0.20765027322404372,
+      "grad_norm": 0.34585925936698914,
+      "learning_rate": 2.974993695240579e-05,
+      "loss": 1.38385808467865,
+      "step": 76
+    },
+    {
+      "epoch": 0.21311475409836064,
+      "grad_norm": 2.415900468826294,
+      "learning_rate": 2.9697620555479297e-05,
+      "loss": 0.7725762128829956,
+      "step": 78
+    },
+    {
+      "epoch": 0.2185792349726776,
+      "grad_norm": 0.48625677824020386,
+      "learning_rate": 2.9640399901050182e-05,
+      "loss": 0.8228914737701416,
+      "step": 80
+    },
+    {
+      "epoch": 0.22404371584699453,
+      "grad_norm": 0.5314300060272217,
+      "learning_rate": 2.9578296252407734e-05,
+      "loss": 1.5916510820388794,
+      "step": 82
+    },
+    {
+      "epoch": 0.22950819672131148,
+      "grad_norm": 0.7009768486022949,
+      "learning_rate": 2.9511332687369917e-05,
+      "loss": 0.8212544918060303,
+      "step": 84
+    },
+    {
+      "epoch": 0.23497267759562843,
+      "grad_norm": 0.2534927427768707,
+      "learning_rate": 2.9439534089707624e-05,
+      "loss": 1.3987239599227905,
+      "step": 86
+    },
+    {
+      "epoch": 0.24043715846994534,
+      "grad_norm": 0.3312060534954071,
+      "learning_rate": 2.9362927139897832e-05,
+      "loss": 1.3285937309265137,
+      "step": 88
+    },
+    {
+      "epoch": 0.2459016393442623,
+      "grad_norm": 0.9822537899017334,
+      "learning_rate": 2.9281540305209068e-05,
+      "loss": 0.9706141352653503,
+      "step": 90
+    },
+    {
+      "epoch": 0.25136612021857924,
+      "grad_norm": 0.6968846321105957,
+      "learning_rate": 2.919540382912294e-05,
+      "loss": 1.4107171297073364,
+      "step": 92
+    },
+    {
+      "epoch": 0.2568306010928962,
+      "grad_norm": 0.4691411256790161,
+      "learning_rate": 2.9104549720095634e-05,
+      "loss": 1.2292362451553345,
+      "step": 94
+    },
+    {
+      "epoch": 0.26229508196721313,
+      "grad_norm": 0.8305396437644958,
+      "learning_rate": 2.9009011739663467e-05,
+      "loss": 1.6303619146347046,
+      "step": 96
+    },
+    {
+      "epoch": 0.2677595628415301,
+      "grad_norm": 0.4047655761241913,
+      "learning_rate": 2.8908825389897094e-05,
+      "loss": 1.355845332145691,
+      "step": 98
+    },
+    {
+      "epoch": 0.273224043715847,
+      "grad_norm": 0.2581457793712616,
+      "learning_rate": 2.8804027900208843e-05,
+      "loss": 1.1548889875411987,
+      "step": 100
+    },
+    {
+      "epoch": 0.2786885245901639,
+      "grad_norm": 0.41052544116973877,
+      "learning_rate": 2.8694658213518226e-05,
+      "loss": 1.3467285633087158,
+      "step": 102
+    },
+    {
+      "epoch": 0.28415300546448086,
+      "grad_norm": 0.6879177689552307,
+      "learning_rate": 2.8580756971780686e-05,
+      "loss": 1.0806087255477905,
+      "step": 104
+    },
+    {
+      "epoch": 0.2896174863387978,
+      "grad_norm": 0.27630922198295593,
+      "learning_rate": 2.846236650088497e-05,
+      "loss": 1.3213335275650024,
+      "step": 106
+    },
+    {
+      "epoch": 0.29508196721311475,
+      "grad_norm": 0.2548673152923584,
+      "learning_rate": 2.833953079492476e-05,
+      "loss": 1.4522829055786133,
+      "step": 108
+    },
+    {
+      "epoch": 0.3005464480874317,
+      "grad_norm": 0.6334999203681946,
+      "learning_rate": 2.82122954998504e-05,
+      "loss": 1.3047692775726318,
+      "step": 110
+    },
+    {
+      "epoch": 0.30601092896174864,
+      "grad_norm": 0.2597161531448364,
+      "learning_rate": 2.808070789650679e-05,
+      "loss": 1.3145617246627808,
+      "step": 112
+    },
+    {
+      "epoch": 0.3114754098360656,
+      "grad_norm": 0.5719535946846008,
+      "learning_rate": 2.7944816883063727e-05,
+      "loss": 1.4522080421447754,
+      "step": 114
+    },
+    {
+      "epoch": 0.31693989071038253,
+      "grad_norm": 0.3497866690158844,
+      "learning_rate": 2.7804672956845295e-05,
+      "loss": 1.2907500267028809,
+      "step": 116
+    },
+    {
+      "epoch": 0.3224043715846995,
+      "grad_norm": 0.5994348526000977,
+      "learning_rate": 2.766032819556495e-05,
+      "loss": 1.1124215126037598,
+      "step": 118
+    },
+    {
+      "epoch": 0.32786885245901637,
+      "grad_norm": 3.034389019012451,
+      "learning_rate": 2.7511836237973366e-05,
+      "loss": 1.3056573867797852,
+      "step": 120
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.41870880126953125,
+      "learning_rate": 2.735925226392618e-05,
+      "loss": 1.3432774543762207,
+      "step": 122
+    },
+    {
+      "epoch": 0.33879781420765026,
+      "grad_norm": 0.3683644235134125,
+      "learning_rate": 2.7202632973879086e-05,
+      "loss": 1.0805763006210327,
+      "step": 124
+    },
+    {
+      "epoch": 0.3442622950819672,
+      "grad_norm": 0.34786108136177063,
+      "learning_rate": 2.7042036567817838e-05,
+      "loss": 1.1195154190063477,
+      "step": 126
+    },
+    {
+      "epoch": 0.34972677595628415,
+      "grad_norm": 1.6386656761169434,
+      "learning_rate": 2.6877522723631036e-05,
+      "loss": 1.3379027843475342,
+      "step": 128
+    },
+    {
+      "epoch": 0.3551912568306011,
+      "grad_norm": 0.6847893595695496,
+      "learning_rate": 2.6709152574933727e-05,
+      "loss": 1.282078742980957,
+      "step": 130
+    },
+    {
+      "epoch": 0.36065573770491804,
+      "grad_norm": 0.2398746758699417,
+      "learning_rate": 2.6536988688350067e-05,
+      "loss": 1.2325066328048706,
+      "step": 132
+    },
+    {
+      "epoch": 0.366120218579235,
+      "grad_norm": 1.0792237520217896,
+      "learning_rate": 2.6361095040263437e-05,
+      "loss": 1.1317086219787598,
+      "step": 134
+    },
+    {
+      "epoch": 0.37158469945355194,
+      "grad_norm": 1.1165977716445923,
+      "learning_rate": 2.618153699304274e-05,
+      "loss": 1.0417739152908325,
+      "step": 136
+    },
+    {
+      "epoch": 0.3770491803278688,
+      "grad_norm": 0.18367436528205872,
+      "learning_rate": 2.599838127075361e-05,
+      "loss": 1.01582932472229,
+      "step": 138
+    },
+    {
+      "epoch": 0.3825136612021858,
+      "grad_norm": 0.4686935842037201,
+      "learning_rate": 2.5811695934363666e-05,
+      "loss": 0.6732921004295349,
+      "step": 140
+    },
+    {
+      "epoch": 0.3879781420765027,
+      "grad_norm": 0.2868647277355194,
+      "learning_rate": 2.5621550356450914e-05,
+      "loss": 1.253921389579773,
+      "step": 142
+    },
+    {
+      "epoch": 0.39344262295081966,
+      "grad_norm": 0.22526639699935913,
+      "learning_rate": 2.5428015195424825e-05,
+      "loss": 1.2781373262405396,
+      "step": 144
+    },
+    {
+      "epoch": 0.3989071038251366,
+      "grad_norm": 0.2401130348443985,
+      "learning_rate": 2.5231162369269498e-05,
+      "loss": 1.240103006362915,
+      "step": 146
+    },
+    {
+      "epoch": 0.40437158469945356,
+      "grad_norm": 0.34567099809646606,
+      "learning_rate": 2.503106502881889e-05,
+      "loss": 1.2351547479629517,
+      "step": 148
+    },
+    {
+      "epoch": 0.4098360655737705,
+      "grad_norm": 0.38193589448928833,
+      "learning_rate": 2.4827797530573762e-05,
+      "loss": 1.1751936674118042,
+      "step": 150
+    },
+    {
+      "epoch": 0.41530054644808745,
+      "grad_norm": 0.2586861848831177,
+      "learning_rate": 2.4621435409070757e-05,
+      "loss": 1.2725920677185059,
+      "step": 152
+    },
+    {
+      "epoch": 0.4207650273224044,
+      "grad_norm": 0.3629944920539856,
+      "learning_rate": 2.4412055348813602e-05,
+      "loss": 1.2570769786834717,
+      "step": 154
+    },
+    {
+      "epoch": 0.4262295081967213,
+      "grad_norm": 0.4387210011482239,
+      "learning_rate": 2.4199735155777017e-05,
+      "loss": 1.3026829957962036,
+      "step": 156
+    },
+    {
+      "epoch": 0.43169398907103823,
+      "grad_norm": 0.7413071393966675,
+      "learning_rate": 2.3984553728493914e-05,
+      "loss": 1.008730173110962,
+      "step": 158
+    },
+    {
+      "epoch": 0.4371584699453552,
+      "grad_norm": 0.2501511573791504,
+      "learning_rate": 2.3766591028736547e-05,
+      "loss": 1.2349580526351929,
+      "step": 160
+    },
+    {
+      "epoch": 0.4426229508196721,
+      "grad_norm": 1.2075088024139404,
+      "learning_rate": 2.3545928051802588e-05,
+      "loss": 0.7488419413566589,
+      "step": 162
+    },
+    {
+      "epoch": 0.44808743169398907,
+      "grad_norm": 0.5644823312759399,
+      "learning_rate": 2.332264679641717e-05,
+      "loss": 1.6321135759353638,
+      "step": 164
+    },
+    {
+      "epoch": 0.453551912568306,
+      "grad_norm": 1.0617799758911133,
+      "learning_rate": 2.3096830234261996e-05,
+      "loss": 0.7753019332885742,
+      "step": 166
+    },
+    {
+      "epoch": 0.45901639344262296,
+      "grad_norm": 0.28146958351135254,
+      "learning_rate": 2.2868562279142912e-05,
+      "loss": 1.1729671955108643,
+      "step": 168
+    },
+    {
+      "epoch": 0.4644808743169399,
+      "grad_norm": 0.29739731550216675,
+      "learning_rate": 2.2637927755807458e-05,
+      "loss": 1.268522024154663,
+      "step": 170
+    },
+    {
+      "epoch": 0.46994535519125685,
+      "grad_norm": 1.04466712474823,
+      "learning_rate": 2.2405012368423786e-05,
+      "loss": 1.330442190170288,
+      "step": 172
+    },
+    {
+      "epoch": 0.47540983606557374,
+      "grad_norm": 0.46707776188850403,
+      "learning_rate": 2.2169902668732893e-05,
+      "loss": 1.4469293355941772,
+      "step": 174
+    },
+    {
+      "epoch": 0.4808743169398907,
+      "grad_norm": 0.4159034192562103,
+      "learning_rate": 2.193268602388583e-05,
+      "loss": 1.2424243688583374,
+      "step": 176
+    },
+    {
+      "epoch": 0.48633879781420764,
+      "grad_norm": 0.5740254521369934,
+      "learning_rate": 2.1693450583977953e-05,
+      "loss": 1.2044943571090698,
+      "step": 178
+    },
+    {
+      "epoch": 0.4918032786885246,
+      "grad_norm": 0.35157784819602966,
+      "learning_rate": 2.1452285249292147e-05,
+      "loss": 1.2033029794692993,
+      "step": 180
+    },
+    {
+      "epoch": 0.4972677595628415,
+      "grad_norm": 0.5813264846801758,
+      "learning_rate": 2.12092796372634e-05,
+      "loss": 1.2133545875549316,
+      "step": 182
+    },
+    {
+      "epoch": 0.5027322404371585,
+      "grad_norm": 1.1736332178115845,
+      "learning_rate": 2.096452404917679e-05,
+      "loss": 1.3634682893753052,
+      "step": 184
+    },
+    {
+      "epoch": 0.5081967213114754,
+      "grad_norm": 0.48712271451950073,
+      "learning_rate": 2.0718109436611348e-05,
+      "loss": 1.239789605140686,
+      "step": 186
+    },
+    {
+      "epoch": 0.5136612021857924,
+      "grad_norm": 0.7002618908882141,
+      "learning_rate": 2.0470127367642345e-05,
+      "loss": 1.2088130712509155,
+      "step": 188
+    },
+    {
+      "epoch": 0.5191256830601093,
+      "grad_norm": 0.3209463059902191,
+      "learning_rate": 2.022066999281444e-05,
+      "loss": 1.2543022632598877,
+      "step": 190
+    },
+    {
+      "epoch": 0.5245901639344263,
+      "grad_norm": 0.5514739155769348,
+      "learning_rate": 1.9969830010898358e-05,
+      "loss": 1.2285394668579102,
+      "step": 192
+    },
+    {
+      "epoch": 0.5300546448087432,
+      "grad_norm": 0.41652482748031616,
+      "learning_rate": 1.9717700634443903e-05,
+      "loss": 1.2374595403671265,
+      "step": 194
+    },
+    {
+      "epoch": 0.5355191256830601,
+      "grad_norm": 0.3611076772212982,
+      "learning_rate": 1.9464375555142e-05,
+      "loss": 1.288169503211975,
+      "step": 196
+    },
+    {
+      "epoch": 0.5409836065573771,
+      "grad_norm": 0.3940355181694031,
+      "learning_rate": 1.9209948909008734e-05,
+      "loss": 1.1237924098968506,
+      "step": 198
+    },
+    {
+      "epoch": 0.546448087431694,
+      "grad_norm": 0.3027671277523041,
+      "learning_rate": 1.8954515241404218e-05,
+      "loss": 1.2617192268371582,
+      "step": 200
+    },
+    {
+      "epoch": 0.5519125683060109,
+      "grad_norm": 0.3746492266654968,
+      "learning_rate": 1.8698169471899414e-05,
+      "loss": 1.1039533615112305,
+      "step": 202
+    },
+    {
+      "epoch": 0.5573770491803278,
+      "grad_norm": 0.8580636382102966,
+      "learning_rate": 1.8441006859003842e-05,
+      "loss": 0.9195830821990967,
+      "step": 204
+    },
+    {
+      "epoch": 0.5628415300546448,
+      "grad_norm": 0.33917945623397827,
+      "learning_rate": 1.818312296476737e-05,
+      "loss": 1.270058274269104,
+      "step": 206
+    },
+    {
+      "epoch": 0.5683060109289617,
+      "grad_norm": 0.3017415404319763,
+      "learning_rate": 1.792461361926921e-05,
+      "loss": 1.0368424654006958,
+      "step": 208
+    },
+    {
+      "epoch": 0.5737704918032787,
+      "grad_norm": 0.3447563052177429,
+      "learning_rate": 1.766557488500727e-05,
+      "loss": 1.216840386390686,
+      "step": 210
+    },
+    {
+      "epoch": 0.5792349726775956,
+      "grad_norm": 0.3713569641113281,
+      "learning_rate": 1.7406103021201212e-05,
+      "loss": 1.5115100145339966,
+      "step": 212
+    },
+    {
+      "epoch": 0.5846994535519126,
+      "grad_norm": 0.19028295576572418,
+      "learning_rate": 1.7146294448022335e-05,
+      "loss": 1.2231677770614624,
+      "step": 214
+    },
+    {
+      "epoch": 0.5901639344262295,
+      "grad_norm": 0.9744280576705933,
+      "learning_rate": 1.688624571076371e-05,
+      "loss": 1.5730527639389038,
+      "step": 216
+    },
+    {
+      "epoch": 0.5956284153005464,
+      "grad_norm": 0.2510036528110504,
+      "learning_rate": 1.6626053443963762e-05,
+      "loss": 1.2018402814865112,
+      "step": 218
+    },
+    {
+      "epoch": 0.6010928961748634,
+      "grad_norm": 0.26169928908348083,
+      "learning_rate": 1.636581433549674e-05,
+      "loss": 1.2398658990859985,
+      "step": 220
+    },
+    {
+      "epoch": 0.6065573770491803,
+      "grad_norm": 0.34778648614883423,
+      "learning_rate": 1.610562509064332e-05,
+      "loss": 1.107602596282959,
+      "step": 222
+    },
+    {
+      "epoch": 0.6120218579234973,
+      "grad_norm": 0.5183984637260437,
+      "learning_rate": 1.5845582396154786e-05,
+      "loss": 1.040785551071167,
+      "step": 224
+    },
+    {
+      "epoch": 0.6174863387978142,
+      "grad_norm": 0.21732346713542938,
+      "learning_rate": 1.5585782884324064e-05,
+      "loss": 1.2164437770843506,
+      "step": 226
+    },
+    {
+      "epoch": 0.6229508196721312,
+      "grad_norm": 0.8948683142662048,
+      "learning_rate": 1.5326323097077015e-05,
+      "loss": 1.229734182357788,
+      "step": 228
+    },
+    {
+      "epoch": 0.6284153005464481,
+      "grad_norm": 0.2247624546289444,
+      "learning_rate": 1.5067299450097261e-05,
+      "loss": 0.7190737128257751,
+      "step": 230
+    },
+    {
+      "epoch": 0.6338797814207651,
+      "grad_norm": 0.305253267288208,
+      "learning_rate": 1.4808808196998006e-05,
+      "loss": 1.22504460811615,
+      "step": 232
+    },
+    {
+      "epoch": 0.639344262295082,
+      "grad_norm": 1.4114190340042114,
+      "learning_rate": 1.4550945393554004e-05,
+      "loss": 0.8749437928199768,
+      "step": 234
+    },
+    {
+      "epoch": 0.644808743169399,
+      "grad_norm": 1.5399497747421265,
+      "learning_rate": 1.4293806862007085e-05,
+      "loss": 0.9087380766868591,
+      "step": 236
+    },
+    {
+      "epoch": 0.6502732240437158,
+      "grad_norm": 0.6150895357131958,
+      "learning_rate": 1.4037488155458448e-05,
+      "loss": 1.303091287612915,
+      "step": 238
+    },
+    {
+      "epoch": 0.6557377049180327,
+      "grad_norm": 0.5875135064125061,
+      "learning_rate": 1.3782084522360981e-05,
+      "loss": 1.1951395273208618,
+      "step": 240
+    },
+    {
+      "epoch": 0.6612021857923497,
+      "grad_norm": 0.7042756676673889,
+      "learning_rate": 1.3527690871124762e-05,
+      "loss": 1.1392256021499634,
+      "step": 242
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.628507673740387,
+      "learning_rate": 1.3274401734848958e-05,
+      "loss": 1.0318241119384766,
+      "step": 244
+    },
+    {
+      "epoch": 0.6721311475409836,
+      "grad_norm": 0.24752730131149292,
+      "learning_rate": 1.3022311236193156e-05,
+      "loss": 1.1875934600830078,
+      "step": 246
+    },
+    {
+      "epoch": 0.6775956284153005,
+      "grad_norm": 0.5493347644805908,
+      "learning_rate": 1.2771513052401236e-05,
+      "loss": 1.2356622219085693,
+      "step": 248
+    },
+    {
+      "epoch": 0.6830601092896175,
+      "grad_norm": 0.3261685371398926,
+      "learning_rate": 1.2522100380490744e-05,
+      "loss": 1.2128081321716309,
+      "step": 250
+    },
+    {
+      "epoch": 0.6885245901639344,
+      "grad_norm": 0.4376692771911621,
+      "learning_rate": 1.2274165902620732e-05,
+      "loss": 1.1218929290771484,
+      "step": 252
+    },
+    {
+      "epoch": 0.6939890710382514,
+      "grad_norm": 0.335615336894989,
+      "learning_rate": 1.2027801751650918e-05,
+      "loss": 1.1654932498931885,
+      "step": 254
+    },
+    {
+      "epoch": 0.6994535519125683,
+      "grad_norm": 0.4503217041492462,
+      "learning_rate": 1.1783099476904972e-05,
+      "loss": 1.1567964553833008,
+      "step": 256
+    },
+    {
+      "epoch": 0.7049180327868853,
+      "grad_norm": 0.3116782605648041,
+      "learning_rate": 1.1540150010150599e-05,
+      "loss": 1.0923233032226562,
+      "step": 258
+    },
+    {
+      "epoch": 0.7103825136612022,
+      "grad_norm": 0.2738410234451294,
+      "learning_rate": 1.1299043631809205e-05,
+      "loss": 1.159534215927124,
+      "step": 260
+    },
+    {
+      "epoch": 0.7158469945355191,
+      "grad_norm": 0.8347621560096741,
+      "learning_rate": 1.1059869937407486e-05,
+      "loss": 1.2569533586502075,
+      "step": 262
+    },
+    {
+      "epoch": 0.7213114754098361,
+      "grad_norm": 0.2674843370914459,
+      "learning_rate": 1.082271780428362e-05,
+      "loss": 1.1951040029525757,
+      "step": 264
+    },
+    {
+      "epoch": 0.726775956284153,
+      "grad_norm": 0.3185964524745941,
+      "learning_rate": 1.0587675358560278e-05,
+      "loss": 1.0627436637878418,
+      "step": 266
+    },
+    {
+      "epoch": 0.73224043715847,
+      "grad_norm": 0.5771196484565735,
+      "learning_rate": 1.0354829942396837e-05,
+      "loss": 1.1797475814819336,
+      "step": 268
+    },
+    {
+      "epoch": 0.7377049180327869,
+      "grad_norm": 0.2694435119628906,
+      "learning_rate": 1.012426808153287e-05,
+      "loss": 1.1862552165985107,
+      "step": 270
+    },
+    {
+      "epoch": 0.7431693989071039,
+      "grad_norm": 0.3423052132129669,
+      "learning_rate": 9.896075453135039e-06,
+      "loss": 1.1492632627487183,
+      "step": 272
+    },
+    {
+      "epoch": 0.7486338797814208,
+      "grad_norm": 0.38316285610198975,
+      "learning_rate": 9.67033685395934e-06,
+      "loss": 1.167414903640747,
+      "step": 274
+    },
+    {
+      "epoch": 0.7540983606557377,
+      "grad_norm": 0.3846551477909088,
+      "learning_rate": 9.447136168840466e-06,
+      "loss": 1.2184109687805176,
+      "step": 276
+    },
+    {
+      "epoch": 0.7595628415300546,
+      "grad_norm": 0.44415712356567383,
+      "learning_rate": 9.226556339520069e-06,
+      "loss": 1.1867570877075195,
+      "step": 278
+    },
+    {
+      "epoch": 0.7650273224043715,
+      "grad_norm": 0.8969565033912659,
+      "learning_rate": 9.008679333825478e-06,
+      "loss": 0.8917623162269592,
+      "step": 280
+    },
+    {
+      "epoch": 0.7704918032786885,
+      "grad_norm": 0.28591468930244446,
+      "learning_rate": 8.793586115210326e-06,
+      "loss": 1.188754677772522,
+      "step": 282
+    },
+    {
+      "epoch": 0.7759562841530054,
+      "grad_norm": 0.3005872666835785,
+      "learning_rate": 8.581356612668382e-06,
+      "loss": 1.173933744430542,
+      "step": 284
+    },
+    {
+      "epoch": 0.7814207650273224,
+      "grad_norm": 0.2967031002044678,
+      "learning_rate": 8.372069691031804e-06,
+      "loss": 1.1837879419326782,
+      "step": 286
+    },
+    {
+      "epoch": 0.7868852459016393,
+      "grad_norm": 0.2758060395717621,
+      "learning_rate": 8.165803121664869e-06,
+      "loss": 1.1517661809921265,
+      "step": 288
+    },
+    {
+      "epoch": 0.7923497267759563,
+      "grad_norm": 0.29872316122055054,
+      "learning_rate": 7.962633553563965e-06,
+      "loss": 1.1168982982635498,
+      "step": 290
+    },
+    {
+      "epoch": 0.7978142076502732,
+      "grad_norm": 0.3943071961402893,
+      "learning_rate": 7.762636484874723e-06,
+      "loss": 1.1435798406600952,
+      "step": 292
+    },
+    {
+      "epoch": 0.8032786885245902,
+      "grad_norm": 0.4497228264808655,
+      "learning_rate": 7.565886234836767e-06,
+      "loss": 1.1566179990768433,
+      "step": 294
+    },
+    {
+      "epoch": 0.8087431693989071,
+      "grad_norm": 0.354144424200058,
+      "learning_rate": 7.3724559161665876e-06,
+      "loss": 1.5012860298156738,
+      "step": 296
+    },
+    {
+      "epoch": 0.8142076502732241,
+      "grad_norm": 1.309401273727417,
+      "learning_rate": 7.182417407888703e-06,
+      "loss": 0.9079286456108093,
+      "step": 298
+    },
+    {
+      "epoch": 0.819672131147541,
+      "grad_norm": 0.3585984706878662,
+      "learning_rate": 6.995841328625321e-06,
+      "loss": 1.5736669301986694,
+      "step": 300
+    },
+    {
+      "epoch": 0.825136612021858,
+      "grad_norm": 0.5723401308059692,
+      "learning_rate": 6.812797010354325e-06,
+      "loss": 1.192858099937439,
+      "step": 302
+    },
+    {
+      "epoch": 0.8306010928961749,
+      "grad_norm": 2.9767942428588867,
+      "learning_rate": 6.63335247264542e-06,
+      "loss": 0.7880658507347107,
+      "step": 304
+    },
+    {
+      "epoch": 0.8360655737704918,
+      "grad_norm": 1.3081823587417603,
+      "learning_rate": 6.457574397383919e-06,
+      "loss": 1.5271462202072144,
+      "step": 306
+    },
+    {
+      "epoch": 0.8415300546448088,
+      "grad_norm": 0.34140118956565857,
+      "learning_rate": 6.285528103991665e-06,
+      "loss": 0.8412893414497375,
+      "step": 308
+    },
+    {
+      "epoch": 0.8469945355191257,
+      "grad_norm": 0.3129923939704895,
+      "learning_rate": 6.117277525154225e-06,
+      "loss": 1.20363187789917,
+      "step": 310
+    },
+    {
+      "epoch": 0.8524590163934426,
+      "grad_norm": 0.383305162191391,
+      "learning_rate": 5.952885183063397e-06,
+      "loss": 1.1824288368225098,
+      "step": 312
+    },
+    {
+      "epoch": 0.8579234972677595,
+      "grad_norm": 0.3238297402858734,
+      "learning_rate": 5.792412166183841e-06,
+      "loss": 1.1946678161621094,
+      "step": 314
+    },
+    {
+      "epoch": 0.8633879781420765,
+      "grad_norm": 2.138742208480835,
+      "learning_rate": 5.635918106552546e-06,
+      "loss": 1.201963186264038,
+      "step": 316
+    },
+    {
+      "epoch": 0.8688524590163934,
+      "grad_norm": 0.3680674433708191,
+      "learning_rate": 5.483461157619428e-06,
+      "loss": 1.141302227973938,
+      "step": 318
+    },
+    {
+      "epoch": 0.8743169398907104,
+      "grad_norm": 0.2648313641548157,
+      "learning_rate": 5.335097972637441e-06,
+      "loss": 1.3266459703445435,
+      "step": 320
+    },
+    {
+      "epoch": 0.8797814207650273,
+      "grad_norm": 0.2844366431236267,
+      "learning_rate": 5.1908836836101135e-06,
+      "loss": 1.123349666595459,
+      "step": 322
+    },
+    {
+      "epoch": 0.8852459016393442,
+      "grad_norm": 0.5806925296783447,
+      "learning_rate": 5.050871880804414e-06,
+      "loss": 0.5679168701171875,
+      "step": 324
+    },
+    {
+      "epoch": 0.8907103825136612,
+      "grad_norm": 0.2976682484149933,
+      "learning_rate": 4.915114592836521e-06,
+      "loss": 1.1642491817474365,
+      "step": 326
+    },
+    {
+      "epoch": 0.8961748633879781,
+      "grad_norm": 1.0322918891906738,
+      "learning_rate": 4.783662267337909e-06,
+      "loss": 1.0334808826446533,
+      "step": 328
+    },
+    {
+      "epoch": 0.9016393442622951,
+      "grad_norm": 0.48543456196784973,
+      "learning_rate": 4.656563752208907e-06,
+      "loss": 1.186019778251648,
+      "step": 330
+    },
+    {
+      "epoch": 0.907103825136612,
+      "grad_norm": 0.8842418789863586,
+      "learning_rate": 4.533866277466767e-06,
+      "loss": 1.249828815460205,
+      "step": 332
+    },
+    {
+      "epoch": 0.912568306010929,
+      "grad_norm": 0.45925524830818176,
+      "learning_rate": 4.415615437694876e-06,
+      "loss": 1.0486047267913818,
+      "step": 334
+    },
+    {
+      "epoch": 0.9180327868852459,
+      "grad_norm": 0.49386680126190186,
+      "learning_rate": 4.3018551750997694e-06,
+      "loss": 1.14175283908844,
+      "step": 336
+    },
+    {
+      "epoch": 0.9234972677595629,
+      "grad_norm": 0.4923037886619568,
+      "learning_rate": 4.192627763182111e-06,
+      "loss": 1.1671501398086548,
+      "step": 338
+    },
+    {
+      "epoch": 0.9289617486338798,
+      "grad_norm": 0.31538236141204834,
+      "learning_rate": 4.087973791027797e-06,
+      "loss": 1.1591607332229614,
+      "step": 340
+    },
+    {
+      "epoch": 0.9344262295081968,
+      "grad_norm": 0.31360623240470886,
+      "learning_rate": 3.987932148224993e-06,
+      "loss": 1.1940336227416992,
+      "step": 342
+    },
+    {
+      "epoch": 0.9398907103825137,
+      "grad_norm": 1.0204191207885742,
+      "learning_rate": 3.8925400104126834e-06,
+      "loss": 1.1359769105911255,
+      "step": 344
+    },
+    {
+      "epoch": 0.9453551912568307,
+      "grad_norm": 0.6455877423286438,
+      "learning_rate": 3.8018328254661618e-06,
+      "loss": 1.5209176540374756,
+      "step": 346
+    },
+    {
+      "epoch": 0.9508196721311475,
+      "grad_norm": 0.6131592392921448,
+      "learning_rate": 3.715844300324527e-06,
+      "loss": 0.8120126724243164,
+      "step": 348
+    },
+    {
+      "epoch": 0.9562841530054644,
+      "grad_norm": 0.3298521339893341,
+      "learning_rate": 3.6346063884651327e-06,
+      "loss": 0.8144136071205139,
+      "step": 350
+    },
+    {
+      "epoch": 0.9617486338797814,
+      "grad_norm": 0.35196903347969055,
+      "learning_rate": 3.558149278029624e-06,
+      "loss": 1.1486433744430542,
+      "step": 352
+    },
+    {
+      "epoch": 0.9672131147540983,
+      "grad_norm": 1.0103577375411987,
+      "learning_rate": 3.4865013806059817e-06,
+      "loss": 0.8401246666908264,
+      "step": 354
+    },
+    {
+      "epoch": 0.9726775956284153,
+      "grad_norm": 0.42854589223861694,
+      "learning_rate": 3.419689320670712e-06,
+      "loss": 1.042449951171875,
+      "step": 356
+    },
+    {
+      "epoch": 0.9781420765027322,
+      "grad_norm": 0.37466299533843994,
+      "learning_rate": 3.35773792569517e-06,
+      "loss": 0.7528899312019348,
+      "step": 358
+    },
+    {
+      "epoch": 0.9836065573770492,
+      "grad_norm": 0.2885702848434448,
+      "learning_rate": 3.300670216919602e-06,
+      "loss": 1.155380129814148,
+      "step": 360
+    },
+    {
+      "epoch": 0.9890710382513661,
+      "grad_norm": 0.33391743898391724,
+      "learning_rate": 3.2485074007984468e-06,
+      "loss": 0.9418554306030273,
+      "step": 362
+    },
+    {
+      "epoch": 0.994535519125683,
+      "grad_norm": 0.2690609097480774,
+      "learning_rate": 3.2012688611199566e-06,
+      "loss": 1.1393744945526123,
+      "step": 364
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4000090956687927,
+      "learning_rate": 3.158972151803165e-06,
+      "loss": 1.1187925338745117,
+      "step": 366
+    },
+    {
+      "epoch": 1.005464480874317,
+      "grad_norm": 0.40801742672920227,
+      "learning_rate": 3.1216329903748095e-06,
+      "loss": 0.981923520565033,
+      "step": 368
+    },
+    {
+      "epoch": 1.010928961748634,
+      "grad_norm": 0.39118003845214844,
+      "learning_rate": 3.089265252128686e-06,
+      "loss": 1.0384995937347412,
+      "step": 370
+    },
+    {
+      "epoch": 1.0163934426229508,
+      "grad_norm": 0.3082128167152405,
+      "learning_rate": 3.061880964969555e-06,
+      "loss": 0.9732365608215332,
+      "step": 372
+    },
+    {
+      "epoch": 1.0218579234972678,
+      "grad_norm": 0.41411662101745605,
+      "learning_rate": 3.039490304943562e-06,
+      "loss": 1.0013678073883057,
+      "step": 374
+    },
+    {
+      "epoch": 1.0273224043715847,
+      "grad_norm": 0.6949254870414734,
+      "learning_rate": 3.022101592456795e-06,
+      "loss": 0.848197877407074,
+      "step": 376
+    },
+    {
+      "epoch": 1.0327868852459017,
+      "grad_norm": 0.4030425250530243,
+      "learning_rate": 3.0097212891834095e-06,
+      "loss": 0.886638879776001,
+      "step": 378
+    },
+    {
+      "epoch": 1.0382513661202186,
+      "grad_norm": 0.37187013030052185,
+      "learning_rate": 3.0023539956644634e-06,
+      "loss": 0.60066819190979,
+      "step": 380
+    },
+    {
+      "epoch": 1.0437158469945356,
+      "grad_norm": 0.7314932942390442,
+      "learning_rate": 3.0000024495983428e-06,
+      "loss": 0.9156309962272644,
+      "step": 382
+    },
+    {
+      "epoch": 1.0491803278688525,
+      "grad_norm": 0.4109429121017456,
+      "learning_rate": 3.002667524823434e-06,
+      "loss": 0.9761273860931396,
+      "step": 384
+    },
+    {
+      "epoch": 1.0546448087431695,
+      "grad_norm": 0.7147516012191772,
+      "learning_rate": 3.010348230993402e-06,
+      "loss": 0.6097055077552795,
+      "step": 386
+    },
+    {
+      "epoch": 1.0601092896174864,
+      "grad_norm": 0.5150865316390991,
+      "learning_rate": 3.0230417139451987e-06,
+      "loss": 0.5354875326156616,
+      "step": 388
+    },
+    {
+      "epoch": 1.0655737704918034,
+      "grad_norm": 0.5118747353553772,
+      "learning_rate": 3.0407432567596883e-06,
+      "loss": 1.1215546131134033,
+      "step": 390
+    },
+    {
+      "epoch": 1.0710382513661203,
+      "grad_norm": 0.3243095874786377,
+      "learning_rate": 3.0634462815144474e-06,
+      "loss": 0.8853414058685303,
+      "step": 392
+    },
+    {
+      "epoch": 1.0765027322404372,
+      "grad_norm": 1.0932362079620361,
+      "learning_rate": 3.0911423517281404e-06,
+      "loss": 0.5655322670936584,
+      "step": 394
+    },
+    {
+      "epoch": 1.0819672131147542,
+      "grad_norm": 0.5749414563179016,
+      "learning_rate": 3.1238211754955294e-06,
+      "loss": 0.8961263298988342,
+      "step": 396
+    },
+    {
+      "epoch": 1.0874316939890711,
+      "grad_norm": 0.4210751950740814,
+      "learning_rate": 3.161470609311961e-06,
+      "loss": 0.9246914386749268,
+      "step": 398
+    },
+    {
+      "epoch": 1.092896174863388,
+      "grad_norm": 0.4931108057498932,
+      "learning_rate": 3.2040766625859115e-06,
+      "loss": 1.0521095991134644,
+      "step": 400
+    },
+    {
+      "epoch": 1.098360655737705,
+      "grad_norm": 3.348501205444336,
+      "learning_rate": 3.2516235028379157e-06,
+      "loss": 0.9077386260032654,
+      "step": 402
+    },
+    {
+      "epoch": 1.1038251366120218,
+      "grad_norm": 0.3238441050052643,
+      "learning_rate": 3.304093461583944e-06,
+      "loss": 0.970809817314148,
+      "step": 404
+    },
+    {
+      "epoch": 1.1092896174863387,
+      "grad_norm": 1.0212552547454834,
+      "learning_rate": 3.3614670409010353e-06,
+      "loss": 1.094781756401062,
+      "step": 406
+    },
+    {
+      "epoch": 1.1147540983606556,
+      "grad_norm": 0.5059043765068054,
+      "learning_rate": 3.4237229206727602e-06,
+      "loss": 0.643450140953064,
+      "step": 408
+    },
+    {
+      "epoch": 1.1202185792349726,
+      "grad_norm": 0.34191587567329407,
+      "learning_rate": 3.490837966511817e-06,
+      "loss": 1.1581542491912842,
+      "step": 410
+    },
+    {
+      "epoch": 1.1256830601092895,
+      "grad_norm": 1.0587300062179565,
+      "learning_rate": 3.5627872383567937e-06,
+      "loss": 0.7080036997795105,
+      "step": 412
+    },
+    {
+      "epoch": 1.1311475409836065,
+      "grad_norm": 0.2536488175392151,
+      "learning_rate": 3.6395439997399494e-06,
+      "loss": 0.6488229632377625,
+      "step": 414
+    },
+    {
+      "epoch": 1.1366120218579234,
+      "grad_norm": 0.8057353496551514,
+      "learning_rate": 3.721079727722522e-06,
+      "loss": 0.6402009129524231,
+      "step": 416
+    },
+    {
+      "epoch": 1.1420765027322404,
+      "grad_norm": 0.29355666041374207,
+      "learning_rate": 3.8073641234939055e-06,
+      "loss": 0.9499251842498779,
+      "step": 418
+    },
+    {
+      "epoch": 1.1475409836065573,
+      "grad_norm": 0.3825172185897827,
+      "learning_rate": 3.898365123630732e-06,
+      "loss": 0.8552459478378296,
+      "step": 420
+    },
+    {
+      "epoch": 1.1530054644808743,
+      "grad_norm": 0.3960343897342682,
+      "learning_rate": 3.994048912011692e-06,
+      "loss": 0.9106036424636841,
+      "step": 422
+    },
+    {
+      "epoch": 1.1584699453551912,
+      "grad_norm": 1.2985180616378784,
+      "learning_rate": 4.094379932383666e-06,
+      "loss": 0.7228477597236633,
+      "step": 424
+    },
+    {
+      "epoch": 1.1639344262295082,
+      "grad_norm": 0.29699158668518066,
+      "learning_rate": 4.199320901574489e-06,
+      "loss": 0.6496458053588867,
+      "step": 426
+    },
+    {
+      "epoch": 1.169398907103825,
+      "grad_norm": 0.3441046476364136,
+      "learning_rate": 4.3088328233474185e-06,
+      "loss": 1.1220229864120483,
+      "step": 428
+    },
+    {
+      "epoch": 1.174863387978142,
+      "grad_norm": 0.38102617859840393,
+      "learning_rate": 4.422875002892234e-06,
+      "loss": 0.9543809294700623,
+      "step": 430
+    },
+    {
+      "epoch": 1.180327868852459,
+      "grad_norm": 1.1369144916534424,
+      "learning_rate": 4.54140506194747e-06,
+      "loss": 0.8527680039405823,
+      "step": 432
+    },
+    {
+      "epoch": 1.185792349726776,
+      "grad_norm": 3.073438882827759,
+      "learning_rate": 4.664378954548241e-06,
+      "loss": 0.7841230630874634,
+      "step": 434
+    },
+    {
+      "epoch": 1.1912568306010929,
+      "grad_norm": 0.28754952549934387,
+      "learning_rate": 4.791750983393832e-06,
+      "loss": 0.9805347919464111,
+      "step": 436
+    },
+    {
+      "epoch": 1.1967213114754098,
+      "grad_norm": 0.23602163791656494,
+      "learning_rate": 4.9234738168288466e-06,
+      "loss": 0.8651551008224487,
+      "step": 438
+    },
+    {
+      "epoch": 1.2021857923497268,
+      "grad_norm": 0.3787078857421875,
+      "learning_rate": 5.059498506431758e-06,
+      "loss": 0.9147478342056274,
+      "step": 440
+    },
+    {
+      "epoch": 1.2076502732240437,
+      "grad_norm": 0.513213038444519,
+      "learning_rate": 5.199774505204206e-06,
+      "loss": 0.9759296774864197,
+      "step": 442
+    },
+    {
+      "epoch": 1.2131147540983607,
+      "grad_norm": 0.38409918546676636,
+      "learning_rate": 5.344249686354357e-06,
+      "loss": 0.9177553653717041,
+      "step": 444
+    },
+    {
+      "epoch": 1.2185792349726776,
+      "grad_norm": 0.296101450920105,
+      "learning_rate": 5.492870362667299e-06,
+      "loss": 0.5355038046836853,
+      "step": 446
+    },
+    {
+      "epoch": 1.2240437158469946,
+      "grad_norm": 0.6465612053871155,
+      "learning_rate": 5.645581306455302e-06,
+      "loss": 0.8347874879837036,
+      "step": 448
+    },
+    {
+      "epoch": 1.2295081967213115,
+      "grad_norm": 0.378653347492218,
+      "learning_rate": 5.802325770080506e-06,
+      "loss": 0.9033167362213135,
+      "step": 450
+    },
+    {
+      "epoch": 1.2349726775956285,
+      "grad_norm": 0.48362499475479126,
+      "learning_rate": 5.96304550704246e-06,
+      "loss": 0.971287727355957,
+      "step": 452
+    },
+    {
+      "epoch": 1.2404371584699454,
+      "grad_norm": 0.3260844647884369,
+      "learning_rate": 6.127680793622588e-06,
+      "loss": 0.27598556876182556,
+      "step": 454
+    },
+    {
+      "epoch": 1.2459016393442623,
+      "grad_norm": 4.79307746887207,
+      "learning_rate": 6.296170451077657e-06,
+      "loss": 0.7720631957054138,
+      "step": 456
+    },
+    {
+      "epoch": 1.2513661202185793,
+      "grad_norm": 0.5071384906768799,
+      "learning_rate": 6.468451868373856e-06,
+      "loss": 0.6973821520805359,
+      "step": 458
+    },
+    {
+      "epoch": 1.2568306010928962,
+      "grad_norm": 0.4039640426635742,
+      "learning_rate": 6.6444610254532e-06,
+      "loss": 0.9609543085098267,
+      "step": 460
+    },
+    {
+      "epoch": 1.2622950819672132,
+      "grad_norm": 0.5991283059120178,
+      "learning_rate": 6.824132517023449e-06,
+      "loss": 0.6427564024925232,
+      "step": 462
+    },
+    {
+      "epoch": 1.2677595628415301,
+      "grad_norm": 1.2359015941619873,
+      "learning_rate": 7.007399576862872e-06,
+      "loss": 1.003535509109497,
+      "step": 464
+    },
+    {
+      "epoch": 1.273224043715847,
+      "grad_norm": 0.6143922209739685,
+      "learning_rate": 7.1941941026306275e-06,
+      "loss": 0.9678149819374084,
+      "step": 466
+    },
+    {
+      "epoch": 1.278688524590164,
+      "grad_norm": 0.19856023788452148,
+      "learning_rate": 7.3844466811737555e-06,
+      "loss": 0.8354513645172119,
+      "step": 468
+    },
+    {
+      "epoch": 1.2841530054644807,
+      "grad_norm": 0.4483174979686737,
+      "learning_rate": 7.578086614321175e-06,
+      "loss": 1.0250217914581299,
+      "step": 470
+    },
+    {
+      "epoch": 1.289617486338798,
+      "grad_norm": 0.2900082468986511,
+      "learning_rate": 7.775041945155295e-06,
+      "loss": 1.206943154335022,
+      "step": 472
+    },
+    {
+      "epoch": 1.2950819672131146,
+      "grad_norm": 0.25206395983695984,
+      "learning_rate": 7.975239484751258e-06,
+      "loss": 0.8465134501457214,
+      "step": 474
+    },
+    {
+      "epoch": 1.3005464480874318,
+      "grad_norm": 0.44423073530197144,
+      "learning_rate": 8.178604839374125e-06,
+      "loss": 0.4682804346084595,
+      "step": 476
+    },
+    {
+      "epoch": 1.3060109289617485,
+      "grad_norm": 0.40745142102241516,
+      "learning_rate": 8.385062438123673e-06,
+      "loss": 1.0709359645843506,
+      "step": 478
+    },
+    {
+      "epoch": 1.3114754098360657,
+      "grad_norm": 0.5735993385314941,
+      "learning_rate": 8.594535561016661e-06,
+      "loss": 1.0265499353408813,
+      "step": 480
+    },
+    {
+      "epoch": 1.3169398907103824,
+      "grad_norm": 0.25811538100242615,
+      "learning_rate": 8.806946367496155e-06,
+      "loss": 0.9305548667907715,
+      "step": 482
+    },
+    {
+      "epoch": 1.3224043715846996,
+      "grad_norm": 0.5877478122711182,
+      "learning_rate": 9.02221592535712e-06,
+      "loss": 0.6250959634780884,
+      "step": 484
+    },
+    {
+      "epoch": 1.3278688524590163,
+      "grad_norm": 0.31082847714424133,
+      "learning_rate": 9.240264240077859e-06,
+      "loss": 0.8635732531547546,
+      "step": 486
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.4074012041091919,
+      "learning_rate": 9.461010284546016e-06,
+      "loss": 1.0225374698638916,
+      "step": 488
+    },
+    {
+      "epoch": 1.3387978142076502,
+      "grad_norm": 0.7163453102111816,
+      "learning_rate": 9.684372029168438e-06,
+      "loss": 0.6194128394126892,
+      "step": 490
+    },
+    {
+      "epoch": 1.3442622950819672,
+      "grad_norm": 0.3905765414237976,
+      "learning_rate": 9.91026647235348e-06,
+      "loss": 1.0774719715118408,
+      "step": 492
+    },
+    {
+      "epoch": 1.349726775956284,
+      "grad_norm": 0.6559311151504517,
+      "learning_rate": 1.0138609671354586e-05,
+      "loss": 0.877537727355957,
+      "step": 494
+    },
+    {
+      "epoch": 1.355191256830601,
+      "grad_norm": 0.4489327371120453,
+      "learning_rate": 1.0369316773463458e-05,
+      "loss": 0.9511969685554504,
+      "step": 496
+    },
+    {
+      "epoch": 1.360655737704918,
+      "grad_norm": 0.5031542778015137,
+      "learning_rate": 1.0602302047541566e-05,
+      "loss": 0.83577960729599,
+      "step": 498
+    },
+    {
+      "epoch": 1.366120218579235,
+      "grad_norm": 0.40769895911216736,
+      "learning_rate": 1.083747891587788e-05,
+      "loss": 1.0597424507141113,
+      "step": 500
+    },
+    {
+      "epoch": 1.3715846994535519,
+      "grad_norm": 0.4040791690349579,
+      "learning_rate": 1.1074759986361392e-05,
+      "loss": 0.9861813187599182,
+      "step": 502
+    },
+    {
+      "epoch": 1.3770491803278688,
+      "grad_norm": 0.4598991870880127,
+      "learning_rate": 1.1314057084956073e-05,
+      "loss": 0.5206344723701477,
+      "step": 504
+    },
+    {
+      "epoch": 1.3825136612021858,
+      "grad_norm": 0.2866649627685547,
+      "learning_rate": 1.1555281288466553e-05,
+      "loss": 0.7465002536773682,
+      "step": 506
+    },
+    {
+      "epoch": 1.3879781420765027,
+      "grad_norm": 0.230438232421875,
+      "learning_rate": 1.1798342957582084e-05,
+      "loss": 0.9311132431030273,
+      "step": 508
+    },
+    {
+      "epoch": 1.3934426229508197,
+      "grad_norm": 0.28892356157302856,
+      "learning_rate": 1.2043151770186725e-05,
+      "loss": 0.964007556438446,
+      "step": 510
+    },
+    {
+      "epoch": 1.3989071038251366,
+      "grad_norm": 0.2942189872264862,
+      "learning_rate": 1.2289616754923078e-05,
+      "loss": 1.0333213806152344,
+      "step": 512
+    },
+    {
+      "epoch": 1.4043715846994536,
+      "grad_norm": 0.29071956872940063,
+      "learning_rate": 1.253764632499752e-05,
+      "loss": 0.6722708344459534,
+      "step": 514
+    },
+    {
+      "epoch": 1.4098360655737705,
+      "grad_norm": 0.2511497437953949,
+      "learning_rate": 1.2787148312213901e-05,
+      "loss": 0.7008714079856873,
+      "step": 516
+    },
+    {
+      "epoch": 1.4153005464480874,
+      "grad_norm": 0.30298787355422974,
+      "learning_rate": 1.3038030001223439e-05,
+      "loss": 0.9937577843666077,
+      "step": 518
+    },
+    {
+      "epoch": 1.4207650273224044,
+      "grad_norm": 0.8531264066696167,
+      "learning_rate": 1.3290198163977933e-05,
+      "loss": 0.8048759698867798,
+      "step": 520
+    },
+    {
+      "epoch": 1.4262295081967213,
+      "grad_norm": 0.24621592462062836,
+      "learning_rate": 1.3543559094373372e-05,
+      "loss": 1.039415955543518,
+      "step": 522
+    },
+    {
+      "epoch": 1.4316939890710383,
+      "grad_norm": 0.6622775793075562,
+      "learning_rate": 1.3798018643071386e-05,
+      "loss": 0.996492862701416,
+      "step": 524
+    },
+    {
+      "epoch": 1.4371584699453552,
+      "grad_norm": 0.8804835677146912,
+      "learning_rate": 1.4053482252485178e-05,
+      "loss": 0.9746305346488953,
+      "step": 526
+    },
+    {
+      "epoch": 1.4426229508196722,
+      "grad_norm": 0.5442449450492859,
+      "learning_rate": 1.4309854991917388e-05,
+      "loss": 0.7142194509506226,
+      "step": 528
+    },
+    {
+      "epoch": 1.4480874316939891,
+      "grad_norm": 0.4819408059120178,
+      "learning_rate": 1.4567041592836413e-05,
+      "loss": 0.7606571316719055,
+      "step": 530
+    },
+    {
+      "epoch": 1.453551912568306,
+      "grad_norm": 0.6655029654502869,
+      "learning_rate": 1.48249464842784e-05,
+      "loss": 0.7583669424057007,
+      "step": 532
+    },
+    {
+      "epoch": 1.459016393442623,
+      "grad_norm": 0.30279040336608887,
+      "learning_rate": 1.508347382836153e-05,
+      "loss": 1.036142349243164,
+      "step": 534
+    },
+    {
+      "epoch": 1.46448087431694,
+      "grad_norm": 0.6906257271766663,
+      "learning_rate": 1.534252755589961e-05,
+      "loss": 0.963323712348938,
+      "step": 536
+    },
+    {
+      "epoch": 1.469945355191257,
+      "grad_norm": 0.3248192369937897,
+      "learning_rate": 1.5602011402101432e-05,
+      "loss": 1.0202207565307617,
+      "step": 538
+    },
+    {
+      "epoch": 1.4754098360655736,
+      "grad_norm": 0.5597788691520691,
+      "learning_rate": 1.5861828942343037e-05,
+      "loss": 1.0704936981201172,
+      "step": 540
+    },
+    {
+      "epoch": 1.4808743169398908,
+      "grad_norm": 0.3918614089488983,
+      "learning_rate": 1.612188362799917e-05,
+      "loss": 0.9736133217811584,
+      "step": 542
+    },
+    {
+      "epoch": 1.4863387978142075,
+      "grad_norm": 0.2836878299713135,
+      "learning_rate": 1.6382078822320964e-05,
+      "loss": 0.8202962875366211,
+      "step": 544
+    },
+    {
+      "epoch": 1.4918032786885247,
+      "grad_norm": 1.0198228359222412,
+      "learning_rate": 1.6642317836346324e-05,
+      "loss": 1.0485470294952393,
+      "step": 546
+    },
+    {
+      "epoch": 1.4972677595628414,
+      "grad_norm": 0.48356863856315613,
+      "learning_rate": 1.6902503964829644e-05,
+      "loss": 0.4609794318675995,
+      "step": 548
+    },
+    {
+      "epoch": 1.5027322404371586,
+      "grad_norm": 0.30593419075012207,
+      "learning_rate": 1.7162540522177685e-05,
+      "loss": 0.9458433985710144,
+      "step": 550
+    },
+    {
+      "epoch": 1.5081967213114753,
+      "grad_norm": 0.22826838493347168,
+      "learning_rate": 1.7422330878378113e-05,
+      "loss": 0.9808536171913147,
+      "step": 552
+    },
+    {
+      "epoch": 1.5136612021857925,
+      "grad_norm": 0.24512672424316406,
+      "learning_rate": 1.7681778494907298e-05,
+      "loss": 1.0151616334915161,
+      "step": 554
+    },
+    {
+      "epoch": 1.5191256830601092,
+      "grad_norm": 0.26424238085746765,
+      "learning_rate": 1.794078696060429e-05,
+      "loss": 1.1077044010162354,
+      "step": 556
+    },
+    {
+      "epoch": 1.5245901639344264,
+      "grad_norm": 0.2831886410713196,
+      "learning_rate": 1.819926002749727e-05,
+      "loss": 0.777061939239502,
+      "step": 558
+    },
+    {
+      "epoch": 1.530054644808743,
+      "grad_norm": 0.46841201186180115,
+      "learning_rate": 1.84571016465695e-05,
+      "loss": 0.5536693334579468,
+      "step": 560
+    },
+    {
+      "epoch": 1.5355191256830603,
+      "grad_norm": 0.40938982367515564,
+      "learning_rate": 1.8714216003451295e-05,
+      "loss": 0.966119110584259,
+      "step": 562
+    },
+    {
+      "epoch": 1.540983606557377,
+      "grad_norm": 0.24322885274887085,
+      "learning_rate": 1.8970507554024827e-05,
+      "loss": 0.8420946002006531,
+      "step": 564
+    },
+    {
+      "epoch": 1.5464480874316942,
+      "grad_norm": 0.31624189019203186,
+      "learning_rate": 1.922588105992838e-05,
+      "loss": 0.9916543364524841,
+      "step": 566
+    },
+    {
+      "epoch": 1.5519125683060109,
+      "grad_norm": 0.8164126873016357,
+      "learning_rate": 1.9480241623947206e-05,
+      "loss": 0.8523349761962891,
+      "step": 568
+    },
+    {
+      "epoch": 1.5573770491803278,
+      "grad_norm": 0.3358438014984131,
+      "learning_rate": 1.9733494725277413e-05,
+      "loss": 1.1619439125061035,
+      "step": 570
+    },
+    {
+      "epoch": 1.5628415300546448,
+      "grad_norm": 0.2056051790714264,
+      "learning_rate": 1.998554625465005e-05,
+      "loss": 0.9285228848457336,
+      "step": 572
+    },
+    {
+      "epoch": 1.5683060109289617,
+      "grad_norm": 0.28299570083618164,
+      "learning_rate": 2.0236302549302293e-05,
+      "loss": 1.0506218671798706,
+      "step": 574
+    },
+    {
+      "epoch": 1.5737704918032787,
+      "grad_norm": 0.7570331692695618,
+      "learning_rate": 2.0485670427782644e-05,
+      "loss": 0.8511103391647339,
+      "step": 576
+    },
+    {
+      "epoch": 1.5792349726775956,
+      "grad_norm": 0.34417569637298584,
+      "learning_rate": 2.073355722457739e-05,
+      "loss": 1.080989956855774,
+      "step": 578
+    },
+    {
+      "epoch": 1.5846994535519126,
+      "grad_norm": 0.7556970715522766,
+      "learning_rate": 2.0979870824545165e-05,
+      "loss": 0.5906158089637756,
+      "step": 580
+    },
+    {
+      "epoch": 1.5901639344262295,
+      "grad_norm": 0.2745168209075928,
+      "learning_rate": 2.1224519697147145e-05,
+      "loss": 0.7238953113555908,
+      "step": 582
+    },
+    {
+      "epoch": 1.5956284153005464,
+      "grad_norm": 0.4340955913066864,
+      "learning_rate": 2.1467412930459936e-05,
+      "loss": 0.6840563416481018,
+      "step": 584
+    },
+    {
+      "epoch": 1.6010928961748634,
+      "grad_norm": 0.2478933185338974,
+      "learning_rate": 2.1708460264958595e-05,
+      "loss": 1.0408371686935425,
+      "step": 586
+    },
+    {
+      "epoch": 1.6065573770491803,
+      "grad_norm": 0.2583276033401489,
+      "learning_rate": 2.194757212705718e-05,
+      "loss": 1.0583165884017944,
+      "step": 588
+    },
+    {
+      "epoch": 1.6120218579234973,
+      "grad_norm": 0.31770971417427063,
+      "learning_rate": 2.2184659662394522e-05,
+      "loss": 0.8676682710647583,
+      "step": 590
+    },
+    {
+      "epoch": 1.6174863387978142,
+      "grad_norm": 0.1893642693758011,
+      "learning_rate": 2.24196347688526e-05,
+      "loss": 1.0388176441192627,
+      "step": 592
+    },
+    {
+      "epoch": 1.6229508196721312,
+      "grad_norm": 1.554314374923706,
+      "learning_rate": 2.265241012929541e-05,
+      "loss": 0.8767275214195251,
+      "step": 594
+    },
+    {
+      "epoch": 1.6284153005464481,
+      "grad_norm": 0.3525921404361725,
+      "learning_rate": 2.28828992440162e-05,
+      "loss": 0.40555280447006226,
+      "step": 596
+    },
+    {
+      "epoch": 1.633879781420765,
+      "grad_norm": 0.24459926784038544,
+      "learning_rate": 2.3111016462880873e-05,
+      "loss": 1.2219781875610352,
+      "step": 598
+    },
+    {
+      "epoch": 1.639344262295082,
+      "grad_norm": 0.24944064021110535,
+      "learning_rate": 2.333667701715578e-05,
+      "loss": 1.16285240650177,
+      "step": 600
+    },
+    {
+      "epoch": 1.644808743169399,
+      "grad_norm": 0.21063530445098877,
+      "learning_rate": 2.3559797051007815e-05,
+      "loss": 1.0140953063964844,
+      "step": 602
+    },
+    {
+      "epoch": 1.650273224043716,
+      "grad_norm": 0.23877571523189545,
+      "learning_rate": 2.3780293652665477e-05,
+      "loss": 1.088266134262085,
+      "step": 604
+    },
+    {
+      "epoch": 1.6557377049180326,
+      "grad_norm": 0.33986350893974304,
+      "learning_rate": 2.399808488522895e-05,
+      "loss": 1.1801600456237793,
+      "step": 606
+    },
+    {
+      "epoch": 1.6612021857923498,
+      "grad_norm": 0.26012417674064636,
+      "learning_rate": 2.4213089817118078e-05,
+      "loss": 0.8795619606971741,
+      "step": 608
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.2159816473722458,
+      "learning_rate": 2.4425228552146573e-05,
+      "loss": 1.19132661819458,
+      "step": 610
+    },
+    {
+      "epoch": 1.6721311475409837,
+      "grad_norm": 1.038149118423462,
+      "learning_rate": 2.4634422259211614e-05,
+      "loss": 0.7393221259117126,
+      "step": 612
+    },
+    {
+      "epoch": 1.6775956284153004,
+      "grad_norm": 0.21248742938041687,
+      "learning_rate": 2.4840593201587626e-05,
+      "loss": 1.010488748550415,
+      "step": 614
+    },
+    {
+      "epoch": 1.6830601092896176,
+      "grad_norm": 0.21196357905864716,
+      "learning_rate": 2.5043664765813377e-05,
+      "loss": 1.0944840908050537,
+      "step": 616
+    },
+    {
+      "epoch": 1.6885245901639343,
+      "grad_norm": 0.35981178283691406,
+      "learning_rate": 2.524356149016163e-05,
+      "loss": 0.5883587598800659,
+      "step": 618
+    },
+    {
+      "epoch": 1.6939890710382515,
+      "grad_norm": 0.3057361841201782,
+      "learning_rate": 2.544020909268085e-05,
+      "loss": 1.1578078269958496,
+      "step": 620
+    },
+    {
+      "epoch": 1.6994535519125682,
+      "grad_norm": 0.5280173420906067,
+      "learning_rate": 2.5633534498798598e-05,
+      "loss": 1.2335985898971558,
+      "step": 622
+    },
+    {
+      "epoch": 1.7049180327868854,
+      "grad_norm": 0.1886816918849945,
+      "learning_rate": 2.5823465868475985e-05,
+      "loss": 0.8703290820121765,
+      "step": 624
+    },
+    {
+      "epoch": 1.710382513661202,
+      "grad_norm": 0.16325658559799194,
+      "learning_rate": 2.60099326229037e-05,
+      "loss": 0.7182884216308594,
+      "step": 626
+    },
+    {
+      "epoch": 1.7158469945355193,
+      "grad_norm": 0.17951571941375732,
+      "learning_rate": 2.619286547072914e-05,
+      "loss": 1.076783299446106,
+      "step": 628
+    },
+    {
+      "epoch": 1.721311475409836,
+      "grad_norm": 0.49342671036720276,
+      "learning_rate": 2.6372196433805214e-05,
+      "loss": 0.7743050456047058,
+      "step": 630
+    },
+    {
+      "epoch": 1.7267759562841531,
+      "grad_norm": 0.3581479787826538,
+      "learning_rate": 2.654785887245112e-05,
+      "loss": 1.046736478805542,
+      "step": 632
+    },
+    {
+      "epoch": 1.7322404371584699,
+      "grad_norm": 0.19438432157039642,
+      "learning_rate": 2.671978751021577e-05,
+      "loss": 1.142443060874939,
+      "step": 634
+    },
+    {
+      "epoch": 1.737704918032787,
+      "grad_norm": 0.20565520226955414,
+      "learning_rate": 2.6887918458134622e-05,
+      "loss": 1.021065354347229,
+      "step": 636
+    },
+    {
+      "epoch": 1.7431693989071038,
+      "grad_norm": 0.16628266870975494,
+      "learning_rate": 2.705218923847093e-05,
+      "loss": 1.1118640899658203,
+      "step": 638
+    },
+    {
+      "epoch": 1.748633879781421,
+      "grad_norm": 0.35274869203567505,
+      "learning_rate": 2.7212538807932576e-05,
+      "loss": 1.0129691362380981,
+      "step": 640
+    },
+    {
+      "epoch": 1.7540983606557377,
+      "grad_norm": 0.19920802116394043,
+      "learning_rate": 2.7368907580355843e-05,
+      "loss": 1.0399836301803589,
+      "step": 642
+    },
+    {
+      "epoch": 1.7595628415300546,
+      "grad_norm": 0.2887803018093109,
+      "learning_rate": 2.7521237448847734e-05,
+      "loss": 1.0195159912109375,
+      "step": 644
+    },
+    {
+      "epoch": 1.7650273224043715,
+      "grad_norm": 0.2489539533853531,
+      "learning_rate": 2.766947180737861e-05,
+      "loss": 0.6208384037017822,
+      "step": 646
+    },
+    {
+      "epoch": 1.7704918032786885,
+      "grad_norm": 0.2961595058441162,
+      "learning_rate": 2.781355557181706e-05,
+      "loss": 1.0299938917160034,
+      "step": 648
+    },
+    {
+      "epoch": 1.7759562841530054,
+      "grad_norm": 0.2027096301317215,
+      "learning_rate": 2.7953435200399262e-05,
+      "loss": 1.0568764209747314,
+      "step": 650
+    },
+    {
+      "epoch": 1.7814207650273224,
+      "grad_norm": 0.19219909608364105,
+      "learning_rate": 2.8089058713625194e-05,
+      "loss": 1.142011284828186,
+      "step": 652
+    },
+    {
+      "epoch": 1.7868852459016393,
+      "grad_norm": 0.17235226929187775,
+      "learning_rate": 2.8220375713574307e-05,
+      "loss": 1.0625348091125488,
+      "step": 654
+    },
+    {
+      "epoch": 1.7923497267759563,
+      "grad_norm": 1.209734559059143,
+      "learning_rate": 2.8347337402633456e-05,
+      "loss": 1.0341904163360596,
+      "step": 656
+    },
+    {
+      "epoch": 1.7978142076502732,
+      "grad_norm": 0.4596608579158783,
+      "learning_rate": 2.846989660163019e-05,
+      "loss": 1.0296437740325928,
+      "step": 658
+    },
+    {
+      "epoch": 1.8032786885245902,
+      "grad_norm": 0.36770763993263245,
+      "learning_rate": 2.858800776736461e-05,
+      "loss": 1.1404849290847778,
+      "step": 660
+    },
+    {
+      "epoch": 1.8087431693989071,
+      "grad_norm": 0.2144036591053009,
+      "learning_rate": 2.87016270095333e-05,
+      "loss": 1.0944205522537231,
+      "step": 662
+    },
+    {
+      "epoch": 1.814207650273224,
+      "grad_norm": 0.2577725350856781,
+      "learning_rate": 2.8810712107039e-05,
+      "loss": 1.1880568265914917,
+      "step": 664
+    },
+    {
+      "epoch": 1.819672131147541,
+      "grad_norm": 1.9875679016113281,
+      "learning_rate": 2.8915222523680082e-05,
+      "loss": 1.3935034275054932,
+      "step": 666
+    },
+    {
+      "epoch": 1.825136612021858,
+      "grad_norm": 0.17387168109416962,
+      "learning_rate": 2.9015119423213857e-05,
+      "loss": 1.1480551958084106,
+      "step": 668
+    },
+    {
+      "epoch": 1.830601092896175,
+      "grad_norm": 0.2092132717370987,
+      "learning_rate": 2.9110365683788173e-05,
+      "loss": 0.7491433620452881,
+      "step": 670
+    },
+    {
+      "epoch": 1.8360655737704918,
+      "grad_norm": 0.49205076694488525,
+      "learning_rate": 2.9200925911735956e-05,
+      "loss": 1.3666801452636719,
+      "step": 672
+    },
+    {
+      "epoch": 1.8415300546448088,
+      "grad_norm": 0.21502411365509033,
+      "learning_rate": 2.9286766454727563e-05,
+      "loss": 1.0823509693145752,
+      "step": 674
+    },
+    {
+      "epoch": 1.8469945355191257,
+      "grad_norm": 0.19282038509845734,
+      "learning_rate": 2.9367855414276073e-05,
+      "loss": 1.1538946628570557,
+      "step": 676
+    },
+    {
+      "epoch": 1.8524590163934427,
+      "grad_norm": 0.1708725392818451,
+      "learning_rate": 2.9444162657590747e-05,
+      "loss": 0.36654239892959595,
+      "step": 678
+    },
+    {
+      "epoch": 1.8579234972677594,
+      "grad_norm": 0.18482287228107452,
+      "learning_rate": 2.951565982877447e-05,
+      "loss": 1.1071395874023438,
+      "step": 680
+    },
+    {
+      "epoch": 1.8633879781420766,
+      "grad_norm": 0.1782202124595642,
+      "learning_rate": 2.9582320359360864e-05,
+      "loss": 1.1017282009124756,
+      "step": 682
+    },
+    {
+      "epoch": 1.8688524590163933,
+      "grad_norm": 0.251668244600296,
+      "learning_rate": 2.9644119478187126e-05,
+      "loss": 1.1279935836791992,
+      "step": 684
+    },
+    {
+      "epoch": 1.8743169398907105,
+      "grad_norm": 9.041706085205078,
+      "learning_rate": 2.9701034220599074e-05,
+      "loss": 1.0460294485092163,
+      "step": 686
+    },
+    {
+      "epoch": 1.8797814207650272,
+      "grad_norm": 0.18143460154533386,
+      "learning_rate": 2.975304343698483e-05,
+      "loss": 1.1729539632797241,
+      "step": 688
+    },
+    {
+      "epoch": 1.8852459016393444,
+      "grad_norm": 0.19067686796188354,
+      "learning_rate": 2.980012780063404e-05,
+      "loss": 1.1014466285705566,
+      "step": 690
+    },
+    {
+      "epoch": 1.890710382513661,
+      "grad_norm": 0.23826494812965393,
+      "learning_rate": 2.9842269814919755e-05,
+      "loss": 0.8912120461463928,
+      "step": 692
+    },
+    {
+      "epoch": 1.8961748633879782,
+      "grad_norm": 0.23750869929790497,
+      "learning_rate": 2.9879453819800156e-05,
+      "loss": 1.1922415494918823,
+      "step": 694
+    },
+    {
+      "epoch": 1.901639344262295,
+      "grad_norm": 0.18918921053409576,
+      "learning_rate": 2.991166599763788e-05,
+      "loss": 1.1193846464157104,
+      "step": 696
+    },
+    {
+      "epoch": 1.9071038251366121,
+      "grad_norm": 0.3542817533016205,
+      "learning_rate": 2.993889437833466e-05,
+      "loss": 1.0432852506637573,
+      "step": 698
+    },
+    {
+      "epoch": 1.9125683060109289,
+      "grad_norm": 0.17179521918296814,
+      "learning_rate": 2.9961128843779457e-05,
+      "loss": 1.1211903095245361,
+      "step": 700
+    },
+    {
+      "epoch": 1.918032786885246,
+      "grad_norm": 0.2006503790616989,
+      "learning_rate": 2.9978361131608348e-05,
+      "loss": 1.0682426691055298,
+      "step": 702
+    },
+    {
+      "epoch": 1.9234972677595628,
+      "grad_norm": 0.204038605093956,
+      "learning_rate": 2.999058483827483e-05,
+      "loss": 1.1585360765457153,
+      "step": 704
+    },
+    {
+      "epoch": 1.92896174863388,
+      "grad_norm": 0.28496795892715454,
+      "learning_rate": 2.9997795421429404e-05,
+      "loss": 1.1962625980377197,
+      "step": 706
+    },
+    {
+      "epoch": 1.9344262295081966,
+      "grad_norm": 0.18176399171352386,
+      "learning_rate": 2.9999990201607516e-05,
+      "loss": 1.1941583156585693,
+      "step": 708
+    },
+    {
+      "epoch": 1.9398907103825138,
+      "grad_norm": 0.2823626399040222,
+      "learning_rate": 2.999716836322524e-05,
+      "loss": 0.9332343339920044,
+      "step": 710
+    },
+    {
+      "epoch": 1.9453551912568305,
+      "grad_norm": 0.19465626776218414,
+      "learning_rate": 2.9989330954882366e-05,
+      "loss": 0.6781972050666809,
+      "step": 712
+    },
+    {
+      "epoch": 1.9508196721311475,
+      "grad_norm": 0.45259955525398254,
+      "learning_rate": 2.9976480888972708e-05,
+      "loss": 1.159096598625183,
+      "step": 714
+    },
+    {
+      "epoch": 1.9562841530054644,
+      "grad_norm": 0.23192590475082397,
+      "learning_rate": 2.9958622940601907e-05,
+      "loss": 0.8674803376197815,
+      "step": 716
+    },
+    {
+      "epoch": 1.9617486338797814,
+      "grad_norm": 0.8216494917869568,
+      "learning_rate": 2.9935763745812935e-05,
+      "loss": 1.116053581237793,
+      "step": 718
+    },
+    {
+      "epoch": 1.9672131147540983,
+      "grad_norm": 0.38614794611930847,
+      "learning_rate": 2.990791179912017e-05,
+      "loss": 1.1189167499542236,
+      "step": 720
+    },
+    {
+      "epoch": 1.9726775956284153,
+      "grad_norm": 0.20481540262699127,
+      "learning_rate": 2.9875077450352817e-05,
+      "loss": 1.139620065689087,
+      "step": 722
+    },
+    {
+      "epoch": 1.9781420765027322,
+      "grad_norm": 0.17928683757781982,
+      "learning_rate": 2.9837272900808863e-05,
+      "loss": 1.0577901601791382,
+      "step": 724
+    },
+    {
+      "epoch": 1.9836065573770492,
+      "grad_norm": 0.38278234004974365,
+      "learning_rate": 2.9794512198721092e-05,
+      "loss": 1.0994060039520264,
+      "step": 726
+    },
+    {
+      "epoch": 1.989071038251366,
+      "grad_norm": 0.39787301421165466,
+      "learning_rate": 2.9746811234036736e-05,
+      "loss": 1.0890076160430908,
+      "step": 728
+    },
+    {
+      "epoch": 1.994535519125683,
+      "grad_norm": 0.3581365942955017,
+      "learning_rate": 2.9694187732512702e-05,
+      "loss": 1.0212537050247192,
+      "step": 730
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.21458348631858826,
+      "learning_rate": 2.96366612491287e-05,
+      "loss": 1.1633673906326294,
+      "step": 732
+    },
+    {
+      "epoch": 2.0054644808743167,
+      "grad_norm": 0.1837082952260971,
+      "learning_rate": 2.9574253160820573e-05,
+      "loss": 0.7306780219078064,
+      "step": 734
+    },
+    {
+      "epoch": 2.010928961748634,
+      "grad_norm": 0.2729966342449188,
+      "learning_rate": 2.9506986658536562e-05,
+      "loss": 0.7522571682929993,
+      "step": 736
+    },
+    {
+      "epoch": 2.0163934426229506,
+      "grad_norm": 0.34518009424209595,
+      "learning_rate": 2.9434886738619537e-05,
+      "loss": 0.8162837624549866,
+      "step": 738
+    },
+    {
+      "epoch": 2.021857923497268,
+      "grad_norm": 0.31829017400741577,
+      "learning_rate": 2.9357980193518312e-05,
+      "loss": 0.8290322422981262,
+      "step": 740
+    },
+    {
+      "epoch": 2.0273224043715845,
+      "grad_norm": 0.9189549684524536,
+      "learning_rate": 2.927629560183153e-05,
+      "loss": 0.7381850481033325,
+      "step": 742
+    },
+    {
+      "epoch": 2.0327868852459017,
+      "grad_norm": 0.4974863529205322,
+      "learning_rate": 2.91898633176878e-05,
+      "loss": 0.9779095649719238,
+      "step": 744
+    },
+    {
+      "epoch": 2.0382513661202184,
+      "grad_norm": 0.22083698213100433,
+      "learning_rate": 2.909871545946603e-05,
+      "loss": 0.7365862131118774,
+      "step": 746
+    },
+    {
+      "epoch": 2.0437158469945356,
+      "grad_norm": 0.2540753483772278,
+      "learning_rate": 2.9002885897860252e-05,
+      "loss": 0.9322983026504517,
+      "step": 748
+    },
+    {
+      "epoch": 2.0491803278688523,
+      "grad_norm": 0.41510656476020813,
+      "learning_rate": 2.8902410243293152e-05,
+      "loss": 0.60332190990448,
+      "step": 750
+    },
+    {
+      "epoch": 2.0546448087431695,
+      "grad_norm": 0.34592926502227783,
+      "learning_rate": 2.8797325832683208e-05,
+      "loss": 0.523179829120636,
+      "step": 752
+    },
+    {
+      "epoch": 2.060109289617486,
+      "grad_norm": 0.8202958703041077,
+      "learning_rate": 2.868767171557021e-05,
+      "loss": 0.6999484896659851,
+      "step": 754
+    },
+    {
+      "epoch": 2.0655737704918034,
+      "grad_norm": 0.5787069201469421,
+      "learning_rate": 2.8573488639604418e-05,
+      "loss": 1.0504649877548218,
+      "step": 756
+    },
+    {
+      "epoch": 2.07103825136612,
+      "grad_norm": 0.36719974875450134,
+      "learning_rate": 2.845481903540464e-05,
+      "loss": 1.0751094818115234,
+      "step": 758
+    },
+    {
+      "epoch": 2.0765027322404372,
+      "grad_norm": 0.242353156208992,
+      "learning_rate": 2.8331707000790954e-05,
+      "loss": 0.8690615296363831,
+      "step": 760
+    },
+    {
+      "epoch": 2.081967213114754,
+      "grad_norm": 0.3211243152618408,
+      "learning_rate": 2.820419828439788e-05,
+      "loss": 0.9909636974334717,
+      "step": 762
+    },
+    {
+      "epoch": 2.087431693989071,
+      "grad_norm": 0.19196948409080505,
+      "learning_rate": 2.8072340268674133e-05,
+      "loss": 0.8517364263534546,
+      "step": 764
+    },
+    {
+      "epoch": 2.092896174863388,
+      "grad_norm": 0.1766076534986496,
+      "learning_rate": 2.793618195227521e-05,
+      "loss": 1.0775508880615234,
+      "step": 766
+    },
+    {
+      "epoch": 2.098360655737705,
+      "grad_norm": 0.5555011630058289,
+      "learning_rate": 2.779577393185539e-05,
+      "loss": 0.7964032888412476,
+      "step": 768
+    },
+    {
+      "epoch": 2.1038251366120218,
+      "grad_norm": 0.24965371191501617,
+      "learning_rate": 2.765116838326597e-05,
+      "loss": 0.90714031457901,
+      "step": 770
+    },
+    {
+      "epoch": 2.109289617486339,
+      "grad_norm": 0.38016757369041443,
+      "learning_rate": 2.750241904216663e-05,
+      "loss": 0.9129889607429504,
+      "step": 772
+    },
+    {
+      "epoch": 2.1147540983606556,
+      "grad_norm": 0.21674519777297974,
+      "learning_rate": 2.7349581184057144e-05,
+      "loss": 0.7319467663764954,
+      "step": 774
+    },
+    {
+      "epoch": 2.120218579234973,
+      "grad_norm": 0.28484922647476196,
+      "learning_rate": 2.719271160373693e-05,
+      "loss": 0.9362223744392395,
+      "step": 776
+    },
+    {
+      "epoch": 2.1256830601092895,
+      "grad_norm": 0.3728918731212616,
+      "learning_rate": 2.703186859420002e-05,
+      "loss": 0.7095181941986084,
+      "step": 778
+    },
+    {
+      "epoch": 2.1311475409836067,
+      "grad_norm": 0.2245863527059555,
+      "learning_rate": 2.6867111924973283e-05,
+      "loss": 0.7740026116371155,
+      "step": 780
+    },
+    {
+      "epoch": 2.1366120218579234,
+      "grad_norm": 0.1715814769268036,
+      "learning_rate": 2.6698502819905935e-05,
+      "loss": 0.8628339171409607,
+      "step": 782
+    },
+    {
+      "epoch": 2.1420765027322406,
+      "grad_norm": 0.1785881519317627,
+      "learning_rate": 2.652610393441872e-05,
+      "loss": 1.0129051208496094,
+      "step": 784
+    },
+    {
+      "epoch": 2.1475409836065573,
+      "grad_norm": 0.16147422790527344,
+      "learning_rate": 2.6349979332220992e-05,
+      "loss": 0.6580853462219238,
+      "step": 786
+    },
+    {
+      "epoch": 2.1530054644808745,
+      "grad_norm": 0.6977578401565552,
+      "learning_rate": 2.6170194461504586e-05,
+      "loss": 0.8348934054374695,
+      "step": 788
+    },
+    {
+      "epoch": 2.158469945355191,
+      "grad_norm": 0.2910391390323639,
+      "learning_rate": 2.5986816130623133e-05,
+      "loss": 1.0212098360061646,
+      "step": 790
+    },
+    {
+      "epoch": 2.1639344262295084,
+      "grad_norm": 0.23879674077033997,
+      "learning_rate": 2.579991248326594e-05,
+      "loss": 0.6800191402435303,
+      "step": 792
+    },
+    {
+      "epoch": 2.169398907103825,
+      "grad_norm": 0.7448290586471558,
+      "learning_rate": 2.560955297313575e-05,
+      "loss": 0.8883236646652222,
+      "step": 794
+    },
+    {
+      "epoch": 2.1748633879781423,
+      "grad_norm": 0.2625773251056671,
+      "learning_rate": 2.5415808338139595e-05,
+      "loss": 0.7563661336898804,
+      "step": 796
+    },
+    {
+      "epoch": 2.180327868852459,
+      "grad_norm": 0.21875064074993134,
+      "learning_rate": 2.5218750574102465e-05,
+      "loss": 0.8323014974594116,
+      "step": 798
+    },
+    {
+      "epoch": 2.185792349726776,
+      "grad_norm": 0.535250723361969,
+      "learning_rate": 2.5018452908013522e-05,
+      "loss": 0.9481061697006226,
+      "step": 800
+    },
+    {
+      "epoch": 2.191256830601093,
+      "grad_norm": 0.3556676506996155,
+      "learning_rate": 2.48149897708149e-05,
+      "loss": 0.8436076641082764,
+      "step": 802
+    },
+    {
+      "epoch": 2.19672131147541,
+      "grad_norm": 0.23821324110031128,
+      "learning_rate": 2.4608436769743e-05,
+      "loss": 0.8622503876686096,
+      "step": 804
+    },
+    {
+      "epoch": 2.202185792349727,
+      "grad_norm": 0.2570849657058716,
+      "learning_rate": 2.4398870660232684e-05,
+      "loss": 0.8588972091674805,
+      "step": 806
+    },
+    {
+      "epoch": 2.2076502732240435,
+      "grad_norm": 2.3321070671081543,
+      "learning_rate": 2.418636931739491e-05,
+      "loss": 0.48147663474082947,
+      "step": 808
+    },
+    {
+      "epoch": 2.2131147540983607,
+      "grad_norm": 0.26416775584220886,
+      "learning_rate": 2.3971011707078125e-05,
+      "loss": 0.6585652232170105,
+      "step": 810
+    },
+    {
+      "epoch": 2.2185792349726774,
+      "grad_norm": 0.21810820698738098,
+      "learning_rate": 2.3752877856524532e-05,
+      "loss": 0.8585912585258484,
+      "step": 812
+    },
+    {
+      "epoch": 2.2240437158469946,
+      "grad_norm": 0.2282879799604416,
+      "learning_rate": 2.353204882463168e-05,
+      "loss": 0.6812423467636108,
+      "step": 814
+    },
+    {
+      "epoch": 2.2295081967213113,
+      "grad_norm": 0.4175258278846741,
+      "learning_rate": 2.330860667183101e-05,
+      "loss": 0.7331743240356445,
+      "step": 816
+    },
+    {
+      "epoch": 2.2349726775956285,
+      "grad_norm": 0.27887681126594543,
+      "learning_rate": 2.308263442959396e-05,
+      "loss": 0.9462811350822449,
+      "step": 818
+    },
+    {
+      "epoch": 2.240437158469945,
+      "grad_norm": 0.24890534579753876,
+      "learning_rate": 2.2854216069577376e-05,
+      "loss": 0.9054920673370361,
+      "step": 820
+    },
+    {
+      "epoch": 2.2459016393442623,
+      "grad_norm": 0.6545159816741943,
+      "learning_rate": 2.2623436472419476e-05,
+      "loss": 1.0144051313400269,
+      "step": 822
+    },
+    {
+      "epoch": 2.251366120218579,
+      "grad_norm": 0.48057249188423157,
+      "learning_rate": 2.2390381396198102e-05,
+      "loss": 0.7360069751739502,
+      "step": 824
+    },
+    {
+      "epoch": 2.2568306010928962,
+      "grad_norm": 2.75382399559021,
+      "learning_rate": 2.2155137444562842e-05,
+      "loss": 0.6045563220977783,
+      "step": 826
+    },
+    {
+      "epoch": 2.262295081967213,
+      "grad_norm": 0.6408010125160217,
+      "learning_rate": 2.191779203455302e-05,
+      "loss": 0.5113135576248169,
+      "step": 828
+    },
+    {
+      "epoch": 2.26775956284153,
+      "grad_norm": 0.23403555154800415,
+      "learning_rate": 2.1678433364113297e-05,
+      "loss": 0.8575177192687988,
+      "step": 830
+    },
+    {
+      "epoch": 2.273224043715847,
+      "grad_norm": 0.49736300110816956,
+      "learning_rate": 2.1437150379319245e-05,
+      "loss": 0.5503892302513123,
+      "step": 832
+    },
+    {
+      "epoch": 2.278688524590164,
+      "grad_norm": 0.48317691683769226,
+      "learning_rate": 2.1194032741324823e-05,
+      "loss": 0.6569101810455322,
+      "step": 834
+    },
+    {
+      "epoch": 2.2841530054644807,
+      "grad_norm": 0.5925542712211609,
+      "learning_rate": 2.0949170793044142e-05,
+      "loss": 0.9202378988265991,
+      "step": 836
+    },
+    {
+      "epoch": 2.289617486338798,
+      "grad_norm": 0.41072195768356323,
+      "learning_rate": 2.070265552557985e-05,
+      "loss": 0.9517123699188232,
+      "step": 838
+    },
+    {
+      "epoch": 2.2950819672131146,
+      "grad_norm": 0.22094306349754333,
+      "learning_rate": 2.0454578544410758e-05,
+      "loss": 0.7035161852836609,
+      "step": 840
+    },
+    {
+      "epoch": 2.300546448087432,
+      "grad_norm": 0.29024970531463623,
+      "learning_rate": 2.0205032035351043e-05,
+      "loss": 0.8569685220718384,
+      "step": 842
+    },
+    {
+      "epoch": 2.3060109289617485,
+      "grad_norm": 0.19582630693912506,
+      "learning_rate": 1.9954108730293875e-05,
+      "loss": 0.975737452507019,
+      "step": 844
+    },
+    {
+      "epoch": 2.3114754098360657,
+      "grad_norm": 0.37361687421798706,
+      "learning_rate": 1.9701901872752047e-05,
+      "loss": 0.5019787549972534,
+      "step": 846
+    },
+    {
+      "epoch": 2.3169398907103824,
+      "grad_norm": 0.1886816918849945,
+      "learning_rate": 1.9448505183208607e-05,
+      "loss": 0.9394426345825195,
+      "step": 848
+    },
+    {
+      "epoch": 2.3224043715846996,
+      "grad_norm": 0.6417620182037354,
+      "learning_rate": 1.919401282429013e-05,
+      "loss": 0.8482524156570435,
+      "step": 850
+    },
+    {
+      "epoch": 2.3278688524590163,
+      "grad_norm": 0.4325665831565857,
+      "learning_rate": 1.893851936577567e-05,
+      "loss": 0.8246769905090332,
+      "step": 852
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.23205913603305817,
+      "learning_rate": 1.868211974945461e-05,
+      "loss": 0.9343006610870361,
+      "step": 854
+    },
+    {
+      "epoch": 2.33879781420765,
+      "grad_norm": 0.2003387063741684,
+      "learning_rate": 1.842490925384604e-05,
+      "loss": 0.9182968735694885,
+      "step": 856
+    },
+    {
+      "epoch": 2.3442622950819674,
+      "grad_norm": 0.2010021060705185,
+      "learning_rate": 1.816698345879313e-05,
+      "loss": 0.6269927620887756,
+      "step": 858
+    },
+    {
+      "epoch": 2.349726775956284,
+      "grad_norm": 0.21301890909671783,
+      "learning_rate": 1.790843820994548e-05,
+      "loss": 0.7751089334487915,
+      "step": 860
+    },
+    {
+      "epoch": 2.3551912568306013,
+      "grad_norm": 0.20366328954696655,
+      "learning_rate": 1.7649369583142763e-05,
+      "loss": 0.897426962852478,
+      "step": 862
+    },
+    {
+      "epoch": 2.360655737704918,
+      "grad_norm": 0.21869035065174103,
+      "learning_rate": 1.738987384871274e-05,
+      "loss": 0.6120696067810059,
+      "step": 864
+    },
+    {
+      "epoch": 2.366120218579235,
+      "grad_norm": 0.5016121864318848,
+      "learning_rate": 1.7130047435697118e-05,
+      "loss": 1.1808212995529175,
+      "step": 866
+    },
+    {
+      "epoch": 2.371584699453552,
+      "grad_norm": 0.20122063159942627,
+      "learning_rate": 1.6869986896018226e-05,
+      "loss": 0.8688129782676697,
+      "step": 868
+    },
+    {
+      "epoch": 2.3770491803278686,
+      "grad_norm": 0.18649625778198242,
+      "learning_rate": 1.66097888686003e-05,
+      "loss": 0.7019689083099365,
+      "step": 870
+    },
+    {
+      "epoch": 2.3825136612021858,
+      "grad_norm": 0.2085297703742981,
+      "learning_rate": 1.6349550043458252e-05,
+      "loss": 0.7158989310264587,
+      "step": 872
+    },
+    {
+      "epoch": 2.387978142076503,
+      "grad_norm": 0.26028934121131897,
+      "learning_rate": 1.608936712576749e-05,
+      "loss": 0.8539477586746216,
+      "step": 874
+    },
+    {
+      "epoch": 2.3934426229508197,
+      "grad_norm": 0.5020867586135864,
+      "learning_rate": 1.582933679992809e-05,
+      "loss": 0.7443297505378723,
+      "step": 876
+    },
+    {
+      "epoch": 2.3989071038251364,
+      "grad_norm": 0.2335505485534668,
+      "learning_rate": 1.556955569363678e-05,
+      "loss": 0.760352611541748,
+      "step": 878
+    },
+    {
+      "epoch": 2.4043715846994536,
+      "grad_norm": 0.2085113674402237,
+      "learning_rate": 1.531012034197988e-05,
+      "loss": 0.9256618022918701,
+      "step": 880
+    },
+    {
+      "epoch": 2.4098360655737707,
+      "grad_norm": 0.23211164772510529,
+      "learning_rate": 1.5051127151560745e-05,
+      "loss": 0.4930667281150818,
+      "step": 882
+    },
+    {
+      "epoch": 2.4153005464480874,
+      "grad_norm": 0.2237941026687622,
+      "learning_rate": 1.4792672364674816e-05,
+      "loss": 0.7172381281852722,
+      "step": 884
+    },
+    {
+      "epoch": 2.420765027322404,
+      "grad_norm": 0.21497218310832977,
+      "learning_rate": 1.4534852023545968e-05,
+      "loss": 0.823451578617096,
+      "step": 886
+    },
+    {
+      "epoch": 2.4262295081967213,
+      "grad_norm": 0.23108679056167603,
+      "learning_rate": 1.4277761934636963e-05,
+      "loss": 0.6176611185073853,
+      "step": 888
+    },
+    {
+      "epoch": 2.431693989071038,
+      "grad_norm": 0.2470959722995758,
+      "learning_rate": 1.4021497633047664e-05,
+      "loss": 0.509010374546051,
+      "step": 890
+    },
+    {
+      "epoch": 2.4371584699453552,
+      "grad_norm": 0.21755096316337585,
+      "learning_rate": 1.3766154347013933e-05,
+      "loss": 0.7690192461013794,
+      "step": 892
+    },
+    {
+      "epoch": 2.442622950819672,
+      "grad_norm": 0.19507208466529846,
+      "learning_rate": 1.3511826962520809e-05,
+      "loss": 0.4858570992946625,
+      "step": 894
+    },
+    {
+      "epoch": 2.448087431693989,
+      "grad_norm": 0.19376493990421295,
+      "learning_rate": 1.3258609988042627e-05,
+      "loss": 0.7765929698944092,
+      "step": 896
+    },
+    {
+      "epoch": 2.453551912568306,
+      "grad_norm": 0.15309995412826538,
+      "learning_rate": 1.300659751942353e-05,
+      "loss": 0.7383701205253601,
+      "step": 898
+    },
+    {
+      "epoch": 2.459016393442623,
+      "grad_norm": 0.20970426499843597,
+      "learning_rate": 1.2755883204911305e-05,
+      "loss": 0.8502997159957886,
+      "step": 900
+    },
+    {
+      "epoch": 2.4644808743169397,
+      "grad_norm": 0.20087599754333496,
+      "learning_rate": 1.2506560210357541e-05,
+      "loss": 0.8558724522590637,
+      "step": 902
+    },
+    {
+      "epoch": 2.469945355191257,
+      "grad_norm": 0.15894219279289246,
+      "learning_rate": 1.225872118459706e-05,
+      "loss": 0.6488800048828125,
+      "step": 904
+    },
+    {
+      "epoch": 2.4754098360655736,
+      "grad_norm": 0.3884161412715912,
+      "learning_rate": 1.2012458225019375e-05,
+      "loss": 0.9488789439201355,
+      "step": 906
+    },
+    {
+      "epoch": 2.480874316939891,
+      "grad_norm": 0.1707250028848648,
+      "learning_rate": 1.176786284334528e-05,
+      "loss": 0.5347706079483032,
+      "step": 908
+    },
+    {
+      "epoch": 2.4863387978142075,
+      "grad_norm": 0.11856268346309662,
+      "learning_rate": 1.1525025931620855e-05,
+      "loss": 0.6109844446182251,
+      "step": 910
+    },
+    {
+      "epoch": 2.4918032786885247,
+      "grad_norm": 0.16962800920009613,
+      "learning_rate": 1.1284037728441877e-05,
+      "loss": 0.44916611909866333,
+      "step": 912
+    },
+    {
+      "epoch": 2.4972677595628414,
+      "grad_norm": 0.1900777965784073,
+      "learning_rate": 1.1044987785420924e-05,
+      "loss": 0.8547566533088684,
+      "step": 914
+    },
+    {
+      "epoch": 2.5027322404371586,
+      "grad_norm": 0.20876799523830414,
+      "learning_rate": 1.0807964933909975e-05,
+      "loss": 0.7574434876441956,
+      "step": 916
+    },
+    {
+      "epoch": 2.5081967213114753,
+      "grad_norm": 0.21594154834747314,
+      "learning_rate": 1.0573057251990443e-05,
+      "loss": 0.7612369060516357,
+      "step": 918
+    },
+    {
+      "epoch": 2.5136612021857925,
+      "grad_norm": 0.17668958008289337,
+      "learning_rate": 1.0340352031743256e-05,
+      "loss": 0.7049843668937683,
+      "step": 920
+    },
+    {
+      "epoch": 2.519125683060109,
+      "grad_norm": 0.22894662618637085,
+      "learning_rate": 1.010993574681095e-05,
+      "loss": 0.754566490650177,
+      "step": 922
+    },
+    {
+      "epoch": 2.5245901639344264,
+      "grad_norm": 0.3702336847782135,
+      "learning_rate": 9.881894020263938e-06,
+      "loss": 0.48762306571006775,
+      "step": 924
+    },
+    {
+      "epoch": 2.530054644808743,
+      "grad_norm": 0.30566778779029846,
+      "learning_rate": 9.656311592782831e-06,
+      "loss": 0.5707772374153137,
+      "step": 926
+    },
+    {
+      "epoch": 2.5355191256830603,
+      "grad_norm": 0.2742585837841034,
+      "learning_rate": 9.433272291168689e-06,
+      "loss": 0.6204325556755066,
+      "step": 928
+    },
+    {
+      "epoch": 2.540983606557377,
+      "grad_norm": 0.29101449251174927,
+      "learning_rate": 9.212858997192744e-06,
+      "loss": 1.0284860134124756,
+      "step": 930
+    },
+    {
+      "epoch": 2.546448087431694,
+      "grad_norm": 0.2771863341331482,
+      "learning_rate": 8.995153616797544e-06,
+      "loss": 0.6103931069374084,
+      "step": 932
+    },
+    {
+      "epoch": 2.551912568306011,
+      "grad_norm": 0.20665957033634186,
+      "learning_rate": 8.78023704966047e-06,
+      "loss": 0.7804996371269226,
+      "step": 934
+    },
+    {
+      "epoch": 2.557377049180328,
+      "grad_norm": 0.5313201546669006,
+      "learning_rate": 8.568189159131336e-06,
+      "loss": 0.6838847398757935,
+      "step": 936
+    },
+    {
+      "epoch": 2.5628415300546448,
+      "grad_norm": 0.8041807413101196,
+      "learning_rate": 8.359088742554941e-06,
+      "loss": 0.5385434031486511,
+      "step": 938
+    },
+    {
+      "epoch": 2.5683060109289615,
+      "grad_norm": 0.26967042684555054,
+      "learning_rate": 8.15301350198999e-06,
+      "loss": 0.7957769632339478,
+      "step": 940
+    },
+    {
+      "epoch": 2.5737704918032787,
+      "grad_norm": 1.993246078491211,
+      "learning_rate": 7.950040015334789e-06,
+      "loss": 0.5354985594749451,
+      "step": 942
+    },
+    {
+      "epoch": 2.579234972677596,
+      "grad_norm": 0.19206559658050537,
+      "learning_rate": 7.750243707870748e-06,
+      "loss": 0.8785912394523621,
+      "step": 944
+    },
+    {
+      "epoch": 2.5846994535519126,
+      "grad_norm": 0.2833126187324524,
+      "learning_rate": 7.553698824234314e-06,
+      "loss": 0.25215014815330505,
+      "step": 946
+    },
+    {
+      "epoch": 2.5901639344262293,
+      "grad_norm": 0.1929856687784195,
+      "learning_rate": 7.360478400827475e-06,
+      "loss": 0.6651497483253479,
+      "step": 948
+    },
+    {
+      "epoch": 2.5956284153005464,
+      "grad_norm": 0.20867817103862762,
+      "learning_rate": 7.170654238677331e-06,
+      "loss": 0.6793198585510254,
+      "step": 950
+    },
+    {
+      "epoch": 2.6010928961748636,
+      "grad_norm": 0.32787761092185974,
+      "learning_rate": 6.984296876754711e-06,
+      "loss": 0.737807035446167,
+      "step": 952
+    },
+    {
+      "epoch": 2.6065573770491803,
+      "grad_norm": 1.5362244844436646,
+      "learning_rate": 6.801475565761783e-06,
+      "loss": 0.4789329171180725,
+      "step": 954
+    },
+    {
+      "epoch": 2.612021857923497,
+      "grad_norm": 0.14147597551345825,
+      "learning_rate": 6.622258242398371e-06,
+      "loss": 0.24011307954788208,
+      "step": 956
+    },
+    {
+      "epoch": 2.6174863387978142,
+      "grad_norm": 0.17159943282604218,
+      "learning_rate": 6.4467115041165855e-06,
+      "loss": 0.5085139274597168,
+      "step": 958
+    },
+    {
+      "epoch": 2.6229508196721314,
+      "grad_norm": 0.17652657628059387,
+      "learning_rate": 6.2749005843730336e-06,
+      "loss": 0.6372896432876587,
+      "step": 960
+    },
+    {
+      "epoch": 2.628415300546448,
+      "grad_norm": 0.8388151526451111,
+      "learning_rate": 6.106889328388064e-06,
+      "loss": 0.6449273824691772,
+      "step": 962
+    },
+    {
+      "epoch": 2.633879781420765,
+      "grad_norm": 0.25947555899620056,
+      "learning_rate": 5.942740169420701e-06,
+      "loss": 1.000860333442688,
+      "step": 964
+    },
+    {
+      "epoch": 2.639344262295082,
+      "grad_norm": 0.15862314403057098,
+      "learning_rate": 5.7825141055683895e-06,
+      "loss": 0.6797659993171692,
+      "step": 966
+    },
+    {
+      "epoch": 2.644808743169399,
+      "grad_norm": 0.2170058637857437,
+      "learning_rate": 5.62627067709992e-06,
+      "loss": 0.7449020743370056,
+      "step": 968
+    },
+    {
+      "epoch": 2.650273224043716,
+      "grad_norm": 0.5447753071784973,
+      "learning_rate": 5.474067944330285e-06,
+      "loss": 0.7707789540290833,
+      "step": 970
+    },
+    {
+      "epoch": 2.6557377049180326,
+      "grad_norm": 0.2564184069633484,
+      "learning_rate": 5.325962466045282e-06,
+      "loss": 1.0045585632324219,
+      "step": 972
+    },
+    {
+      "epoch": 2.66120218579235,
+      "grad_norm": 0.15954433381557465,
+      "learning_rate": 5.18200927848421e-06,
+      "loss": 0.5719258785247803,
+      "step": 974
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.15210166573524475,
+      "learning_rate": 5.042261874888308e-06,
+      "loss": 0.7093988060951233,
+      "step": 976
+    },
+    {
+      "epoch": 2.6721311475409837,
+      "grad_norm": 0.13644114136695862,
+      "learning_rate": 4.906772185622572e-06,
+      "loss": 0.7509814500808716,
+      "step": 978
+    },
+    {
+      "epoch": 2.6775956284153004,
+      "grad_norm": 0.15326020121574402,
+      "learning_rate": 4.775590558878368e-06,
+      "loss": 0.6545107960700989,
+      "step": 980
+    },
+    {
+      "epoch": 2.6830601092896176,
+      "grad_norm": 0.17753270268440247,
+      "learning_rate": 4.648765741963903e-06,
+      "loss": 0.8449227213859558,
+      "step": 982
+    },
+    {
+      "epoch": 2.6885245901639343,
+      "grad_norm": 0.3523927927017212,
+      "learning_rate": 4.526344863189724e-06,
+      "loss": 0.6510394811630249,
+      "step": 984
+    },
+    {
+      "epoch": 2.6939890710382515,
+      "grad_norm": 0.2865069508552551,
+      "learning_rate": 4.408373414355714e-06,
+      "loss": 0.9356023669242859,
+      "step": 986
+    },
+    {
+      "epoch": 2.699453551912568,
+      "grad_norm": 0.1558021456003189,
+      "learning_rate": 4.29489523384628e-06,
+      "loss": 0.8081143498420715,
+      "step": 988
+    },
+    {
+      "epoch": 2.7049180327868854,
+      "grad_norm": 0.2865149974822998,
+      "learning_rate": 4.185952490339899e-06,
+      "loss": 0.6692199110984802,
+      "step": 990
+    },
+    {
+      "epoch": 2.710382513661202,
+      "grad_norm": 0.19687670469284058,
+      "learning_rate": 4.081585667139231e-06,
+      "loss": 0.5164180397987366,
+      "step": 992
+    },
+    {
+      "epoch": 2.7158469945355193,
+      "grad_norm": 0.18878105282783508,
+      "learning_rate": 3.981833547127413e-06,
+      "loss": 0.811371922492981,
+      "step": 994
+    },
+    {
+      "epoch": 2.721311475409836,
+      "grad_norm": 0.16217704117298126,
+      "learning_rate": 3.886733198356298e-06,
+      "loss": 0.8148671388626099,
+      "step": 996
+    },
+    {
+      "epoch": 2.726775956284153,
+      "grad_norm": 0.15565072000026703,
+      "learning_rate": 3.7963199602718717e-06,
+      "loss": 0.6758864521980286,
+      "step": 998
+    },
+    {
+      "epoch": 2.73224043715847,
+      "grad_norm": 0.6546015739440918,
+      "learning_rate": 3.7106274305821034e-06,
+      "loss": 0.9574772119522095,
+      "step": 1000
+    },
+    {
+      "epoch": 2.737704918032787,
+      "grad_norm": 0.18078891932964325,
+      "learning_rate": 3.6296874527719515e-06,
+      "loss": 0.8926464915275574,
+      "step": 1002
+    },
+    {
+      "epoch": 2.7431693989071038,
+      "grad_norm": 0.22077669203281403,
+      "learning_rate": 3.553530104270281e-06,
+      "loss": 0.6632700562477112,
+      "step": 1004
+    },
+    {
+      "epoch": 2.748633879781421,
+      "grad_norm": 0.381185919046402,
+      "learning_rate": 3.4821836852730384e-06,
+      "loss": 0.35226285457611084,
+      "step": 1006
+    },
+    {
+      "epoch": 2.7540983606557377,
+      "grad_norm": 0.18125414848327637,
+      "learning_rate": 3.41567470822686e-06,
+      "loss": 0.9504106044769287,
+      "step": 1008
+    },
+    {
+      "epoch": 2.7595628415300544,
+      "grad_norm": 0.25593459606170654,
+      "learning_rate": 3.354027887976989e-06,
+      "loss": 0.9168705344200134,
+      "step": 1010
+    },
+    {
+      "epoch": 2.7650273224043715,
+      "grad_norm": 0.6389570832252502,
+      "learning_rate": 3.297266132583221e-06,
+      "loss": 0.6682818531990051,
+      "step": 1012
+    },
+    {
+      "epoch": 2.7704918032786887,
+      "grad_norm": 0.32402268052101135,
+      "learning_rate": 3.245410534807195e-06,
+      "loss": 0.9942286610603333,
+      "step": 1014
+    },
+    {
+      "epoch": 2.7759562841530054,
+      "grad_norm": 0.20764689147472382,
+      "learning_rate": 3.1984803642743314e-06,
+      "loss": 0.6048266291618347,
+      "step": 1016
+    },
+    {
+      "epoch": 2.781420765027322,
+      "grad_norm": 0.1485147923231125,
+      "learning_rate": 3.1564930603131777e-06,
+      "loss": 0.8052763342857361,
+      "step": 1018
+    },
+    {
+      "epoch": 2.7868852459016393,
+      "grad_norm": 0.38492631912231445,
+      "learning_rate": 3.1194642254749395e-06,
+      "loss": 0.6152138113975525,
+      "step": 1020
+    },
+    {
+      "epoch": 2.7923497267759565,
+      "grad_norm": 0.19493553042411804,
+      "learning_rate": 3.0874076197355317e-06,
+      "loss": 0.8494120836257935,
+      "step": 1022
+    },
+    {
+      "epoch": 2.797814207650273,
+      "grad_norm": 0.4454888701438904,
+      "learning_rate": 3.0603351553823717e-06,
+      "loss": 0.4298951029777527,
+      "step": 1024
+    },
+    {
+      "epoch": 2.80327868852459,
+      "grad_norm": 0.19092072546482086,
+      "learning_rate": 3.038256892587734e-06,
+      "loss": 0.8518891930580139,
+      "step": 1026
+    },
+    {
+      "epoch": 2.808743169398907,
+      "grad_norm": 0.6577255725860596,
+      "learning_rate": 3.0211810356703803e-06,
+      "loss": 0.7430834174156189,
+      "step": 1028
+    },
+    {
+      "epoch": 2.8142076502732243,
+      "grad_norm": 0.1505049765110016,
+      "learning_rate": 3.0091139300468266e-06,
+      "loss": 0.6131287813186646,
+      "step": 1030
+    },
+    {
+      "epoch": 2.819672131147541,
+      "grad_norm": 0.1536058932542801,
+      "learning_rate": 3.0020600598733656e-06,
+      "loss": 0.6610476970672607,
+      "step": 1032
+    },
+    {
+      "epoch": 2.8251366120218577,
+      "grad_norm": 1.4271342754364014,
+      "learning_rate": 3.000022046379753e-06,
+      "loss": 0.7583919763565063,
+      "step": 1034
+    },
+    {
+      "epoch": 2.830601092896175,
+      "grad_norm": 0.24797333776950836,
+      "learning_rate": 3.0030006468951557e-06,
+      "loss": 0.9907567501068115,
+      "step": 1036
+    },
+    {
+      "epoch": 2.836065573770492,
+      "grad_norm": 0.18040958046913147,
+      "learning_rate": 3.0109947545667246e-06,
+      "loss": 0.7043365240097046,
+      "step": 1038
+    },
+    {
+      "epoch": 2.841530054644809,
+      "grad_norm": 0.38979649543762207,
+      "learning_rate": 3.024001398770901e-06,
+      "loss": 0.7790292501449585,
+      "step": 1040
+    },
+    {
+      "epoch": 2.8469945355191255,
+      "grad_norm": 0.1695755124092102,
+      "learning_rate": 3.042015746217308e-06,
+      "loss": 0.7965599298477173,
+      "step": 1042
+    },
+    {
+      "epoch": 2.8524590163934427,
+      "grad_norm": 0.16300812363624573,
+      "learning_rate": 3.0650311027448116e-06,
+      "loss": 0.7110670804977417,
+      "step": 1044
+    },
+    {
+      "epoch": 2.8579234972677594,
+      "grad_norm": 0.18763573467731476,
+      "learning_rate": 3.0930389158090754e-06,
+      "loss": 0.9114327430725098,
+      "step": 1046
+    },
+    {
+      "epoch": 2.8633879781420766,
+      "grad_norm": 0.7911413311958313,
+      "learning_rate": 3.1260287776607025e-06,
+      "loss": 0.3735189735889435,
+      "step": 1048
+    },
+    {
+      "epoch": 2.8688524590163933,
+      "grad_norm": 0.20250028371810913,
+      "learning_rate": 3.163988429212773e-06,
+      "loss": 0.7022408246994019,
+      "step": 1050
+    },
+    {
+      "epoch": 2.8743169398907105,
+      "grad_norm": 0.21102067828178406,
+      "learning_rate": 3.206903764596349e-06,
+      "loss": 0.6459388732910156,
+      "step": 1052
+    },
+    {
+      "epoch": 2.879781420765027,
+      "grad_norm": 0.21323570609092712,
+      "learning_rate": 3.254758836402225e-06,
+      "loss": 0.9086355566978455,
+      "step": 1054
+    },
+    {
+      "epoch": 2.8852459016393444,
+      "grad_norm": 1.73580002784729,
+      "learning_rate": 3.3075358616070144e-06,
+      "loss": 0.7384663820266724,
+      "step": 1056
+    },
+    {
+      "epoch": 2.890710382513661,
+      "grad_norm": 1.4918919801712036,
+      "learning_rate": 3.365215228181358e-06,
+      "loss": 0.6115441918373108,
+      "step": 1058
+    },
+    {
+      "epoch": 2.8961748633879782,
+      "grad_norm": 0.22330603003501892,
+      "learning_rate": 3.4277755023777795e-06,
+      "loss": 0.7794143557548523,
+      "step": 1060
+    },
+    {
+      "epoch": 2.901639344262295,
+      "grad_norm": 0.19227337837219238,
+      "learning_rate": 3.495193436695504e-06,
+      "loss": 0.7703116536140442,
+      "step": 1062
+    },
+    {
+      "epoch": 2.907103825136612,
+      "grad_norm": 0.3918929398059845,
+      "learning_rate": 3.567443978519267e-06,
+      "loss": 0.876200795173645,
+      "step": 1064
+    },
+    {
+      "epoch": 2.912568306010929,
+      "grad_norm": 0.18335311114788055,
+      "learning_rate": 3.6445002794288992e-06,
+      "loss": 0.431761234998703,
+      "step": 1066
+    },
+    {
+      "epoch": 2.918032786885246,
+      "grad_norm": 0.227519690990448,
+      "learning_rate": 3.7263337051762718e-06,
+      "loss": 0.5974105000495911,
+      "step": 1068
+    },
+    {
+      "epoch": 2.9234972677595628,
+      "grad_norm": 0.43164753913879395,
+      "learning_rate": 3.8129138463257943e-06,
+      "loss": 0.6901212930679321,
+      "step": 1070
+    },
+    {
+      "epoch": 2.92896174863388,
+      "grad_norm": 0.22106696665287018,
+      "learning_rate": 3.904208529554625e-06,
+      "loss": 0.9134948253631592,
+      "step": 1072
+    },
+    {
+      "epoch": 2.9344262295081966,
+      "grad_norm": 0.17417727410793304,
+      "learning_rate": 4.000183829608332e-06,
+      "loss": 0.49246451258659363,
+      "step": 1074
+    },
+    {
+      "epoch": 2.939890710382514,
+      "grad_norm": 1.0001611709594727,
+      "learning_rate": 4.100804081907595e-06,
+      "loss": 0.5228325128555298,
+      "step": 1076
+    },
+    {
+      "epoch": 2.9453551912568305,
+      "grad_norm": 0.38090279698371887,
+      "learning_rate": 4.206031895801176e-06,
+      "loss": 0.5014840364456177,
+      "step": 1078
+    },
+    {
+      "epoch": 2.9508196721311473,
+      "grad_norm": 0.2635762691497803,
+      "learning_rate": 4.315828168460367e-06,
+      "loss": 0.9488551020622253,
+      "step": 1080
+    },
+    {
+      "epoch": 2.9562841530054644,
+      "grad_norm": 0.23455342650413513,
+      "learning_rate": 4.430152099409704e-06,
+      "loss": 0.5974031686782837,
+      "step": 1082
+    },
+    {
+      "epoch": 2.9617486338797816,
+      "grad_norm": 0.306372731924057,
+      "learning_rate": 4.548961205688424e-06,
+      "loss": 0.748331606388092,
+      "step": 1084
+    },
+    {
+      "epoch": 2.9672131147540983,
+      "grad_norm": 0.1417221575975418,
+      "learning_rate": 4.672211337637246e-06,
+      "loss": 0.52958744764328,
+      "step": 1086
+    },
+    {
+      "epoch": 2.972677595628415,
+      "grad_norm": 0.18579013645648956,
+      "learning_rate": 4.7998566953044445e-06,
+      "loss": 0.995449960231781,
+      "step": 1088
+    },
+    {
+      "epoch": 2.978142076502732,
+      "grad_norm": 0.31965434551239014,
+      "learning_rate": 4.931849845465193e-06,
+      "loss": 0.767081081867218,
+      "step": 1090
+    },
+    {
+      "epoch": 2.9836065573770494,
+      "grad_norm": 0.15356819331645966,
+      "learning_rate": 5.06814173924782e-06,
+      "loss": 0.6917383670806885,
+      "step": 1092
+    },
+    {
+      "epoch": 2.989071038251366,
+      "grad_norm": 0.17739875614643097,
+      "learning_rate": 5.208681730360458e-06,
+      "loss": 0.7700910568237305,
+      "step": 1094
+    },
+    {
+      "epoch": 2.994535519125683,
+      "grad_norm": 0.5607284903526306,
+      "learning_rate": 5.3534175939112694e-06,
+      "loss": 0.5115733742713928,
+      "step": 1096
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.22396568953990936,
+      "learning_rate": 5.50229554581536e-06,
+      "loss": 0.8061965107917786,
+      "step": 1098
+    },
+    {
+      "epoch": 3.0,
+      "step": 1098,
+      "total_flos": 4.957143256761631e+18,
+      "train_loss": 0.9666990600322765,
+      "train_runtime": 11531.2166,
+      "train_samples_per_second": 5.713,
+      "train_steps_per_second": 0.095
+    }
+  ],
+  "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": 4.957143256761631e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}