diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
--- "a/last-checkpoint/trainer_state.json"
+++ "b/last-checkpoint/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.7506750675067507,
+  "epoch": 0.9998199819981998,
   "eval_steps": 695,
-  "global_step": 2085,
+  "global_step": 2777,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -14634,6 +14634,4850 @@
       "eval_samples_per_second": 45.959,
       "eval_steps_per_second": 11.495,
       "step": 2085
+    },
+    {
+      "epoch": 0.7510351035103511,
+      "grad_norm": 0.671138346195221,
+      "learning_rate": 3.0046836669235433e-05,
+      "loss": 4.6252,
+      "step": 2086
+    },
+    {
+      "epoch": 0.7513951395139514,
+      "grad_norm": 0.9334102869033813,
+      "learning_rate": 2.996455867635155e-05,
+      "loss": 4.5999,
+      "step": 2087
+    },
+    {
+      "epoch": 0.7517551755175518,
+      "grad_norm": 0.8894442319869995,
+      "learning_rate": 2.988237363301758e-05,
+      "loss": 5.0383,
+      "step": 2088
+    },
+    {
+      "epoch": 0.7521152115211521,
+      "grad_norm": 0.6898563504219055,
+      "learning_rate": 2.9800281648307794e-05,
+      "loss": 4.8665,
+      "step": 2089
+    },
+    {
+      "epoch": 0.7524752475247525,
+      "grad_norm": 0.6982426047325134,
+      "learning_rate": 2.9718282831172883e-05,
+      "loss": 4.8052,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7528352835283528,
+      "grad_norm": 0.6563032865524292,
+      "learning_rate": 2.9636377290439944e-05,
+      "loss": 4.4563,
+      "step": 2091
+    },
+    {
+      "epoch": 0.7531953195319532,
+      "grad_norm": 0.747215986251831,
+      "learning_rate": 2.9554565134812294e-05,
+      "loss": 5.0272,
+      "step": 2092
+    },
+    {
+      "epoch": 0.7535553555355535,
+      "grad_norm": 0.842694878578186,
+      "learning_rate": 2.9472846472869298e-05,
+      "loss": 5.1278,
+      "step": 2093
+    },
+    {
+      "epoch": 0.7539153915391539,
+      "grad_norm": 1.041561484336853,
+      "learning_rate": 2.9391221413066182e-05,
+      "loss": 5.2336,
+      "step": 2094
+    },
+    {
+      "epoch": 0.7542754275427542,
+      "grad_norm": 0.7632551193237305,
+      "learning_rate": 2.930969006373402e-05,
+      "loss": 4.7186,
+      "step": 2095
+    },
+    {
+      "epoch": 0.7546354635463547,
+      "grad_norm": 1.0100700855255127,
+      "learning_rate": 2.922825253307947e-05,
+      "loss": 5.0076,
+      "step": 2096
+    },
+    {
+      "epoch": 0.754995499549955,
+      "grad_norm": 0.5503267049789429,
+      "learning_rate": 2.9146908929184713e-05,
+      "loss": 4.6652,
+      "step": 2097
+    },
+    {
+      "epoch": 0.7553555355535554,
+      "grad_norm": 0.7919767498970032,
+      "learning_rate": 2.9065659360007247e-05,
+      "loss": 5.4393,
+      "step": 2098
+    },
+    {
+      "epoch": 0.7557155715571557,
+      "grad_norm": 1.0583096742630005,
+      "learning_rate": 2.898450393337977e-05,
+      "loss": 5.2044,
+      "step": 2099
+    },
+    {
+      "epoch": 0.7560756075607561,
+      "grad_norm": 1.1439197063446045,
+      "learning_rate": 2.8903442757010035e-05,
+      "loss": 5.2308,
+      "step": 2100
+    },
+    {
+      "epoch": 0.7564356435643564,
+      "grad_norm": 1.3537611961364746,
+      "learning_rate": 2.8822475938480764e-05,
+      "loss": 4.7218,
+      "step": 2101
+    },
+    {
+      "epoch": 0.7567956795679568,
+      "grad_norm": 0.6820851564407349,
+      "learning_rate": 2.874160358524931e-05,
+      "loss": 4.7861,
+      "step": 2102
+    },
+    {
+      "epoch": 0.7571557155715571,
+      "grad_norm": 1.5382546186447144,
+      "learning_rate": 2.8660825804647795e-05,
+      "loss": 4.6253,
+      "step": 2103
+    },
+    {
+      "epoch": 0.7575157515751575,
+      "grad_norm": 0.9953200221061707,
+      "learning_rate": 2.8580142703882796e-05,
+      "loss": 4.5608,
+      "step": 2104
+    },
+    {
+      "epoch": 0.7578757875787578,
+      "grad_norm": 0.7481911778450012,
+      "learning_rate": 2.8499554390035143e-05,
+      "loss": 4.9828,
+      "step": 2105
+    },
+    {
+      "epoch": 0.7582358235823582,
+      "grad_norm": 0.8973466753959656,
+      "learning_rate": 2.8419060970059974e-05,
+      "loss": 4.4942,
+      "step": 2106
+    },
+    {
+      "epoch": 0.7585958595859585,
+      "grad_norm": 0.6188588738441467,
+      "learning_rate": 2.8338662550786443e-05,
+      "loss": 4.917,
+      "step": 2107
+    },
+    {
+      "epoch": 0.758955895589559,
+      "grad_norm": 0.8116443157196045,
+      "learning_rate": 2.8258359238917665e-05,
+      "loss": 4.8898,
+      "step": 2108
+    },
+    {
+      "epoch": 0.7593159315931594,
+      "grad_norm": 0.996688187122345,
+      "learning_rate": 2.8178151141030406e-05,
+      "loss": 4.6581,
+      "step": 2109
+    },
+    {
+      "epoch": 0.7596759675967597,
+      "grad_norm": 1.0279045104980469,
+      "learning_rate": 2.8098038363575186e-05,
+      "loss": 4.8797,
+      "step": 2110
+    },
+    {
+      "epoch": 0.76003600360036,
+      "grad_norm": 0.7636517286300659,
+      "learning_rate": 2.8018021012875994e-05,
+      "loss": 4.4845,
+      "step": 2111
+    },
+    {
+      "epoch": 0.7603960396039604,
+      "grad_norm": 0.8215141892433167,
+      "learning_rate": 2.7938099195130153e-05,
+      "loss": 4.7628,
+      "step": 2112
+    },
+    {
+      "epoch": 0.7607560756075608,
+      "grad_norm": 1.067277431488037,
+      "learning_rate": 2.7858273016408197e-05,
+      "loss": 5.2078,
+      "step": 2113
+    },
+    {
+      "epoch": 0.7611161116111611,
+      "grad_norm": 0.7776371240615845,
+      "learning_rate": 2.7778542582653744e-05,
+      "loss": 5.1417,
+      "step": 2114
+    },
+    {
+      "epoch": 0.7614761476147615,
+      "grad_norm": 0.8139586448669434,
+      "learning_rate": 2.769890799968332e-05,
+      "loss": 4.8309,
+      "step": 2115
+    },
+    {
+      "epoch": 0.7618361836183618,
+      "grad_norm": 0.7961241006851196,
+      "learning_rate": 2.7619369373186288e-05,
+      "loss": 4.9364,
+      "step": 2116
+    },
+    {
+      "epoch": 0.7621962196219622,
+      "grad_norm": 0.6579238176345825,
+      "learning_rate": 2.753992680872457e-05,
+      "loss": 4.8429,
+      "step": 2117
+    },
+    {
+      "epoch": 0.7625562556255625,
+      "grad_norm": 0.6686379909515381,
+      "learning_rate": 2.746058041173266e-05,
+      "loss": 4.8558,
+      "step": 2118
+    },
+    {
+      "epoch": 0.762916291629163,
+      "grad_norm": 0.6807066202163696,
+      "learning_rate": 2.7381330287517426e-05,
+      "loss": 4.4459,
+      "step": 2119
+    },
+    {
+      "epoch": 0.7632763276327633,
+      "grad_norm": 0.6293361783027649,
+      "learning_rate": 2.7302176541257986e-05,
+      "loss": 4.898,
+      "step": 2120
+    },
+    {
+      "epoch": 0.7636363636363637,
+      "grad_norm": 0.8723804950714111,
+      "learning_rate": 2.7223119278005438e-05,
+      "loss": 4.6484,
+      "step": 2121
+    },
+    {
+      "epoch": 0.763996399639964,
+      "grad_norm": 0.848070502281189,
+      "learning_rate": 2.7144158602682924e-05,
+      "loss": 5.2841,
+      "step": 2122
+    },
+    {
+      "epoch": 0.7643564356435644,
+      "grad_norm": 0.9075490832328796,
+      "learning_rate": 2.7065294620085424e-05,
+      "loss": 5.2137,
+      "step": 2123
+    },
+    {
+      "epoch": 0.7647164716471647,
+      "grad_norm": 1.0051218271255493,
+      "learning_rate": 2.6986527434879472e-05,
+      "loss": 5.1463,
+      "step": 2124
+    },
+    {
+      "epoch": 0.7650765076507651,
+      "grad_norm": 1.058005928993225,
+      "learning_rate": 2.6907857151603234e-05,
+      "loss": 5.2255,
+      "step": 2125
+    },
+    {
+      "epoch": 0.7654365436543654,
+      "grad_norm": 1.0294654369354248,
+      "learning_rate": 2.6829283874666233e-05,
+      "loss": 4.6688,
+      "step": 2126
+    },
+    {
+      "epoch": 0.7657965796579658,
+      "grad_norm": 0.8858208656311035,
+      "learning_rate": 2.6750807708349267e-05,
+      "loss": 4.6658,
+      "step": 2127
+    },
+    {
+      "epoch": 0.7661566156615661,
+      "grad_norm": 0.8013286590576172,
+      "learning_rate": 2.6672428756804225e-05,
+      "loss": 4.849,
+      "step": 2128
+    },
+    {
+      "epoch": 0.7665166516651665,
+      "grad_norm": 0.7389269471168518,
+      "learning_rate": 2.659414712405398e-05,
+      "loss": 4.4728,
+      "step": 2129
+    },
+    {
+      "epoch": 0.7668766876687669,
+      "grad_norm": 0.6583431363105774,
+      "learning_rate": 2.6515962913992275e-05,
+      "loss": 4.9133,
+      "step": 2130
+    },
+    {
+      "epoch": 0.7672367236723673,
+      "grad_norm": 0.6830927133560181,
+      "learning_rate": 2.643787623038354e-05,
+      "loss": 4.7863,
+      "step": 2131
+    },
+    {
+      "epoch": 0.7675967596759676,
+      "grad_norm": 0.6234518885612488,
+      "learning_rate": 2.6359887176862718e-05,
+      "loss": 4.6174,
+      "step": 2132
+    },
+    {
+      "epoch": 0.767956795679568,
+      "grad_norm": 0.7106865048408508,
+      "learning_rate": 2.6281995856935237e-05,
+      "loss": 4.7574,
+      "step": 2133
+    },
+    {
+      "epoch": 0.7683168316831683,
+      "grad_norm": 0.9051821231842041,
+      "learning_rate": 2.6204202373976818e-05,
+      "loss": 4.8563,
+      "step": 2134
+    },
+    {
+      "epoch": 0.7686768676867687,
+      "grad_norm": 0.6196146011352539,
+      "learning_rate": 2.6126506831233344e-05,
+      "loss": 4.5112,
+      "step": 2135
+    },
+    {
+      "epoch": 0.769036903690369,
+      "grad_norm": 0.6337679624557495,
+      "learning_rate": 2.6048909331820636e-05,
+      "loss": 4.6024,
+      "step": 2136
+    },
+    {
+      "epoch": 0.7693969396939694,
+      "grad_norm": 0.6263675689697266,
+      "learning_rate": 2.5971409978724458e-05,
+      "loss": 4.6405,
+      "step": 2137
+    },
+    {
+      "epoch": 0.7697569756975697,
+      "grad_norm": 1.0419749021530151,
+      "learning_rate": 2.5894008874800325e-05,
+      "loss": 4.7905,
+      "step": 2138
+    },
+    {
+      "epoch": 0.7701170117011701,
+      "grad_norm": 0.8129563927650452,
+      "learning_rate": 2.581670612277335e-05,
+      "loss": 4.4431,
+      "step": 2139
+    },
+    {
+      "epoch": 0.7704770477047704,
+      "grad_norm": 0.7132530212402344,
+      "learning_rate": 2.5739501825238053e-05,
+      "loss": 4.7698,
+      "step": 2140
+    },
+    {
+      "epoch": 0.7708370837083708,
+      "grad_norm": 0.6291266083717346,
+      "learning_rate": 2.566239608465838e-05,
+      "loss": 4.5729,
+      "step": 2141
+    },
+    {
+      "epoch": 0.7711971197119712,
+      "grad_norm": 0.5569373965263367,
+      "learning_rate": 2.558538900336741e-05,
+      "loss": 4.9298,
+      "step": 2142
+    },
+    {
+      "epoch": 0.7715571557155716,
+      "grad_norm": 0.72600257396698,
+      "learning_rate": 2.5508480683567315e-05,
+      "loss": 4.8837,
+      "step": 2143
+    },
+    {
+      "epoch": 0.771917191719172,
+      "grad_norm": 0.7014186978340149,
+      "learning_rate": 2.543167122732918e-05,
+      "loss": 4.6547,
+      "step": 2144
+    },
+    {
+      "epoch": 0.7722772277227723,
+      "grad_norm": 0.8487231731414795,
+      "learning_rate": 2.5354960736592883e-05,
+      "loss": 4.6923,
+      "step": 2145
+    },
+    {
+      "epoch": 0.7726372637263726,
+      "grad_norm": 0.6868178844451904,
+      "learning_rate": 2.5278349313166992e-05,
+      "loss": 4.6032,
+      "step": 2146
+    },
+    {
+      "epoch": 0.772997299729973,
+      "grad_norm": 0.739762544631958,
+      "learning_rate": 2.5201837058728505e-05,
+      "loss": 4.8253,
+      "step": 2147
+    },
+    {
+      "epoch": 0.7733573357335733,
+      "grad_norm": 1.2171458005905151,
+      "learning_rate": 2.512542407482289e-05,
+      "loss": 5.0895,
+      "step": 2148
+    },
+    {
+      "epoch": 0.7737173717371737,
+      "grad_norm": 1.3037205934524536,
+      "learning_rate": 2.504911046286382e-05,
+      "loss": 4.9572,
+      "step": 2149
+    },
+    {
+      "epoch": 0.774077407740774,
+      "grad_norm": 1.3282265663146973,
+      "learning_rate": 2.4972896324133144e-05,
+      "loss": 5.4213,
+      "step": 2150
+    },
+    {
+      "epoch": 0.7744374437443744,
+      "grad_norm": 1.392838716506958,
+      "learning_rate": 2.4896781759780585e-05,
+      "loss": 4.7398,
+      "step": 2151
+    },
+    {
+      "epoch": 0.7747974797479747,
+      "grad_norm": 0.6054560542106628,
+      "learning_rate": 2.4820766870823807e-05,
+      "loss": 4.7054,
+      "step": 2152
+    },
+    {
+      "epoch": 0.7751575157515752,
+      "grad_norm": 0.6699235439300537,
+      "learning_rate": 2.4744851758148156e-05,
+      "loss": 4.6727,
+      "step": 2153
+    },
+    {
+      "epoch": 0.7755175517551756,
+      "grad_norm": 0.5790958404541016,
+      "learning_rate": 2.4669036522506584e-05,
+      "loss": 4.8101,
+      "step": 2154
+    },
+    {
+      "epoch": 0.7758775877587759,
+      "grad_norm": 0.8083108067512512,
+      "learning_rate": 2.45933212645194e-05,
+      "loss": 4.5601,
+      "step": 2155
+    },
+    {
+      "epoch": 0.7762376237623763,
+      "grad_norm": 1.0053972005844116,
+      "learning_rate": 2.451770608467432e-05,
+      "loss": 4.6501,
+      "step": 2156
+    },
+    {
+      "epoch": 0.7765976597659766,
+      "grad_norm": 0.6478819847106934,
+      "learning_rate": 2.4442191083326195e-05,
+      "loss": 4.6527,
+      "step": 2157
+    },
+    {
+      "epoch": 0.776957695769577,
+      "grad_norm": 1.433661937713623,
+      "learning_rate": 2.4366776360696942e-05,
+      "loss": 5.0813,
+      "step": 2158
+    },
+    {
+      "epoch": 0.7773177317731773,
+      "grad_norm": 1.1521214246749878,
+      "learning_rate": 2.429146201687538e-05,
+      "loss": 4.5685,
+      "step": 2159
+    },
+    {
+      "epoch": 0.7776777677767777,
+      "grad_norm": 0.6991515159606934,
+      "learning_rate": 2.42162481518171e-05,
+      "loss": 4.7082,
+      "step": 2160
+    },
+    {
+      "epoch": 0.778037803780378,
+      "grad_norm": 0.8507077693939209,
+      "learning_rate": 2.414113486534434e-05,
+      "loss": 4.6635,
+      "step": 2161
+    },
+    {
+      "epoch": 0.7783978397839784,
+      "grad_norm": 0.6432574987411499,
+      "learning_rate": 2.4066122257145894e-05,
+      "loss": 4.5532,
+      "step": 2162
+    },
+    {
+      "epoch": 0.7787578757875787,
+      "grad_norm": 0.6845300793647766,
+      "learning_rate": 2.3991210426776855e-05,
+      "loss": 4.6837,
+      "step": 2163
+    },
+    {
+      "epoch": 0.7791179117911791,
+      "grad_norm": 0.743084728717804,
+      "learning_rate": 2.3916399473658623e-05,
+      "loss": 4.7629,
+      "step": 2164
+    },
+    {
+      "epoch": 0.7794779477947795,
+      "grad_norm": 0.7912582159042358,
+      "learning_rate": 2.3841689497078746e-05,
+      "loss": 4.8711,
+      "step": 2165
+    },
+    {
+      "epoch": 0.7798379837983799,
+      "grad_norm": 0.6151897311210632,
+      "learning_rate": 2.376708059619065e-05,
+      "loss": 4.5838,
+      "step": 2166
+    },
+    {
+      "epoch": 0.7801980198019802,
+      "grad_norm": 0.8373451828956604,
+      "learning_rate": 2.3692572870013718e-05,
+      "loss": 4.9176,
+      "step": 2167
+    },
+    {
+      "epoch": 0.7805580558055806,
+      "grad_norm": 0.5105034708976746,
+      "learning_rate": 2.361816641743303e-05,
+      "loss": 4.5368,
+      "step": 2168
+    },
+    {
+      "epoch": 0.7809180918091809,
+      "grad_norm": 0.6912795305252075,
+      "learning_rate": 2.354386133719927e-05,
+      "loss": 4.3363,
+      "step": 2169
+    },
+    {
+      "epoch": 0.7812781278127813,
+      "grad_norm": 0.8666832447052002,
+      "learning_rate": 2.3469657727928506e-05,
+      "loss": 4.5923,
+      "step": 2170
+    },
+    {
+      "epoch": 0.7816381638163816,
+      "grad_norm": 0.8544076085090637,
+      "learning_rate": 2.339555568810221e-05,
+      "loss": 4.9953,
+      "step": 2171
+    },
+    {
+      "epoch": 0.781998199819982,
+      "grad_norm": 1.0247130393981934,
+      "learning_rate": 2.3321555316067045e-05,
+      "loss": 5.4613,
+      "step": 2172
+    },
+    {
+      "epoch": 0.7823582358235823,
+      "grad_norm": 1.0253169536590576,
+      "learning_rate": 2.3247656710034737e-05,
+      "loss": 5.1228,
+      "step": 2173
+    },
+    {
+      "epoch": 0.7827182718271827,
+      "grad_norm": 1.0105392932891846,
+      "learning_rate": 2.3173859968081944e-05,
+      "loss": 4.7797,
+      "step": 2174
+    },
+    {
+      "epoch": 0.783078307830783,
+      "grad_norm": 0.8575915098190308,
+      "learning_rate": 2.3100165188150125e-05,
+      "loss": 5.0751,
+      "step": 2175
+    },
+    {
+      "epoch": 0.7834383438343835,
+      "grad_norm": 1.8292618989944458,
+      "learning_rate": 2.3026572468045437e-05,
+      "loss": 4.8729,
+      "step": 2176
+    },
+    {
+      "epoch": 0.7837983798379838,
+      "grad_norm": 0.7877120971679688,
+      "learning_rate": 2.295308190543859e-05,
+      "loss": 4.3804,
+      "step": 2177
+    },
+    {
+      "epoch": 0.7841584158415842,
+      "grad_norm": 2.0647926330566406,
+      "learning_rate": 2.287969359786466e-05,
+      "loss": 4.7416,
+      "step": 2178
+    },
+    {
+      "epoch": 0.7845184518451845,
+      "grad_norm": 1.065865159034729,
+      "learning_rate": 2.280640764272306e-05,
+      "loss": 4.7786,
+      "step": 2179
+    },
+    {
+      "epoch": 0.7848784878487849,
+      "grad_norm": 0.7052769064903259,
+      "learning_rate": 2.2733224137277366e-05,
+      "loss": 4.9935,
+      "step": 2180
+    },
+    {
+      "epoch": 0.7852385238523852,
+      "grad_norm": 0.8767135739326477,
+      "learning_rate": 2.266014317865519e-05,
+      "loss": 4.7496,
+      "step": 2181
+    },
+    {
+      "epoch": 0.7855985598559856,
+      "grad_norm": 0.5927574634552002,
+      "learning_rate": 2.2587164863847975e-05,
+      "loss": 5.0468,
+      "step": 2182
+    },
+    {
+      "epoch": 0.7859585958595859,
+      "grad_norm": 1.1241964101791382,
+      "learning_rate": 2.251428928971102e-05,
+      "loss": 4.837,
+      "step": 2183
+    },
+    {
+      "epoch": 0.7863186318631863,
+      "grad_norm": 0.5535566806793213,
+      "learning_rate": 2.244151655296327e-05,
+      "loss": 4.5108,
+      "step": 2184
+    },
+    {
+      "epoch": 0.7866786678667866,
+      "grad_norm": 0.8191620111465454,
+      "learning_rate": 2.236884675018709e-05,
+      "loss": 4.9146,
+      "step": 2185
+    },
+    {
+      "epoch": 0.787038703870387,
+      "grad_norm": 0.7227510809898376,
+      "learning_rate": 2.2296279977828337e-05,
+      "loss": 4.4775,
+      "step": 2186
+    },
+    {
+      "epoch": 0.7873987398739875,
+      "grad_norm": 0.5830511450767517,
+      "learning_rate": 2.222381633219608e-05,
+      "loss": 4.604,
+      "step": 2187
+    },
+    {
+      "epoch": 0.7877587758775878,
+      "grad_norm": 0.6540189981460571,
+      "learning_rate": 2.2151455909462538e-05,
+      "loss": 4.5976,
+      "step": 2188
+    },
+    {
+      "epoch": 0.7881188118811882,
+      "grad_norm": 0.6274272203445435,
+      "learning_rate": 2.2079198805662914e-05,
+      "loss": 4.5981,
+      "step": 2189
+    },
+    {
+      "epoch": 0.7884788478847885,
+      "grad_norm": 0.9411934018135071,
+      "learning_rate": 2.2007045116695313e-05,
+      "loss": 4.6338,
+      "step": 2190
+    },
+    {
+      "epoch": 0.7888388838883889,
+      "grad_norm": 0.640446662902832,
+      "learning_rate": 2.1934994938320584e-05,
+      "loss": 4.5651,
+      "step": 2191
+    },
+    {
+      "epoch": 0.7891989198919892,
+      "grad_norm": 0.6163681149482727,
+      "learning_rate": 2.1863048366162208e-05,
+      "loss": 4.4887,
+      "step": 2192
+    },
+    {
+      "epoch": 0.7895589558955896,
+      "grad_norm": 0.6504824757575989,
+      "learning_rate": 2.179120549570609e-05,
+      "loss": 4.84,
+      "step": 2193
+    },
+    {
+      "epoch": 0.7899189918991899,
+      "grad_norm": 0.6089352965354919,
+      "learning_rate": 2.1719466422300607e-05,
+      "loss": 4.8128,
+      "step": 2194
+    },
+    {
+      "epoch": 0.7902790279027903,
+      "grad_norm": 0.8295852541923523,
+      "learning_rate": 2.1647831241156302e-05,
+      "loss": 5.124,
+      "step": 2195
+    },
+    {
+      "epoch": 0.7906390639063906,
+      "grad_norm": 0.6523507237434387,
+      "learning_rate": 2.1576300047345932e-05,
+      "loss": 4.6459,
+      "step": 2196
+    },
+    {
+      "epoch": 0.790999099909991,
+      "grad_norm": 0.6584704518318176,
+      "learning_rate": 2.15048729358041e-05,
+      "loss": 4.6353,
+      "step": 2197
+    },
+    {
+      "epoch": 0.7913591359135913,
+      "grad_norm": 0.9848087430000305,
+      "learning_rate": 2.1433550001327373e-05,
+      "loss": 5.323,
+      "step": 2198
+    },
+    {
+      "epoch": 0.7917191719171918,
+      "grad_norm": 0.8073936104774475,
+      "learning_rate": 2.136233133857405e-05,
+      "loss": 5.1241,
+      "step": 2199
+    },
+    {
+      "epoch": 0.7920792079207921,
+      "grad_norm": 1.2630630731582642,
+      "learning_rate": 2.129121704206405e-05,
+      "loss": 5.2807,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7924392439243925,
+      "grad_norm": 1.061435580253601,
+      "learning_rate": 2.1220207206178688e-05,
+      "loss": 4.6137,
+      "step": 2201
+    },
+    {
+      "epoch": 0.7927992799279928,
+      "grad_norm": 0.7881227731704712,
+      "learning_rate": 2.114930192516076e-05,
+      "loss": 4.5962,
+      "step": 2202
+    },
+    {
+      "epoch": 0.7931593159315932,
+      "grad_norm": 0.7256401777267456,
+      "learning_rate": 2.107850129311426e-05,
+      "loss": 4.8192,
+      "step": 2203
+    },
+    {
+      "epoch": 0.7935193519351935,
+      "grad_norm": 0.9993703365325928,
+      "learning_rate": 2.1007805404004242e-05,
+      "loss": 5.1882,
+      "step": 2204
+    },
+    {
+      "epoch": 0.7938793879387939,
+      "grad_norm": 0.714908242225647,
+      "learning_rate": 2.09372143516568e-05,
+      "loss": 4.6632,
+      "step": 2205
+    },
+    {
+      "epoch": 0.7942394239423942,
+      "grad_norm": 0.9289857745170593,
+      "learning_rate": 2.0866728229758857e-05,
+      "loss": 4.8051,
+      "step": 2206
+    },
+    {
+      "epoch": 0.7945994599459946,
+      "grad_norm": 1.0063886642456055,
+      "learning_rate": 2.0796347131858186e-05,
+      "loss": 4.8422,
+      "step": 2207
+    },
+    {
+      "epoch": 0.7949594959495949,
+      "grad_norm": 0.7214351296424866,
+      "learning_rate": 2.072607115136298e-05,
+      "loss": 4.639,
+      "step": 2208
+    },
+    {
+      "epoch": 0.7953195319531953,
+      "grad_norm": 0.7175842523574829,
+      "learning_rate": 2.065590038154209e-05,
+      "loss": 5.0236,
+      "step": 2209
+    },
+    {
+      "epoch": 0.7956795679567957,
+      "grad_norm": 0.5457351207733154,
+      "learning_rate": 2.058583491552465e-05,
+      "loss": 4.5711,
+      "step": 2210
+    },
+    {
+      "epoch": 0.7960396039603961,
+      "grad_norm": 0.6841213703155518,
+      "learning_rate": 2.0515874846300077e-05,
+      "loss": 4.6929,
+      "step": 2211
+    },
+    {
+      "epoch": 0.7963996399639964,
+      "grad_norm": 0.8968706727027893,
+      "learning_rate": 2.044602026671786e-05,
+      "loss": 4.7611,
+      "step": 2212
+    },
+    {
+      "epoch": 0.7967596759675968,
+      "grad_norm": 0.6250977516174316,
+      "learning_rate": 2.0376271269487514e-05,
+      "loss": 4.3924,
+      "step": 2213
+    },
+    {
+      "epoch": 0.7971197119711971,
+      "grad_norm": 0.6848214864730835,
+      "learning_rate": 2.0306627947178446e-05,
+      "loss": 4.5518,
+      "step": 2214
+    },
+    {
+      "epoch": 0.7974797479747975,
+      "grad_norm": 0.5987040996551514,
+      "learning_rate": 2.0237090392219805e-05,
+      "loss": 4.9023,
+      "step": 2215
+    },
+    {
+      "epoch": 0.7978397839783978,
+      "grad_norm": 0.6589849591255188,
+      "learning_rate": 2.0167658696900317e-05,
+      "loss": 4.42,
+      "step": 2216
+    },
+    {
+      "epoch": 0.7981998199819982,
+      "grad_norm": 0.7822523713111877,
+      "learning_rate": 2.0098332953368272e-05,
+      "loss": 4.8455,
+      "step": 2217
+    },
+    {
+      "epoch": 0.7985598559855985,
+      "grad_norm": 0.8310371041297913,
+      "learning_rate": 2.0029113253631314e-05,
+      "loss": 5.1098,
+      "step": 2218
+    },
+    {
+      "epoch": 0.7989198919891989,
+      "grad_norm": 0.8362820148468018,
+      "learning_rate": 1.995999968955641e-05,
+      "loss": 4.5659,
+      "step": 2219
+    },
+    {
+      "epoch": 0.7992799279927992,
+      "grad_norm": 0.7935851812362671,
+      "learning_rate": 1.9890992352869543e-05,
+      "loss": 4.5946,
+      "step": 2220
+    },
+    {
+      "epoch": 0.7996399639963997,
+      "grad_norm": 0.6581932902336121,
+      "learning_rate": 1.9822091335155812e-05,
+      "loss": 5.0322,
+      "step": 2221
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6121277213096619,
+      "learning_rate": 1.9753296727859195e-05,
+      "loss": 4.5539,
+      "step": 2222
+    },
+    {
+      "epoch": 0.8003600360036004,
+      "grad_norm": 0.7371995449066162,
+      "learning_rate": 1.9684608622282417e-05,
+      "loss": 5.2824,
+      "step": 2223
+    },
+    {
+      "epoch": 0.8007200720072007,
+      "grad_norm": 0.825187087059021,
+      "learning_rate": 1.9616027109586887e-05,
+      "loss": 5.0987,
+      "step": 2224
+    },
+    {
+      "epoch": 0.8010801080108011,
+      "grad_norm": 1.2908596992492676,
+      "learning_rate": 1.9547552280792524e-05,
+      "loss": 5.2177,
+      "step": 2225
+    },
+    {
+      "epoch": 0.8014401440144014,
+      "grad_norm": 2.2421438694000244,
+      "learning_rate": 1.947918422677769e-05,
+      "loss": 4.7303,
+      "step": 2226
+    },
+    {
+      "epoch": 0.8018001800180018,
+      "grad_norm": 0.5044226050376892,
+      "learning_rate": 1.941092303827896e-05,
+      "loss": 4.6811,
+      "step": 2227
+    },
+    {
+      "epoch": 0.8021602160216021,
+      "grad_norm": 0.5311564207077026,
+      "learning_rate": 1.9342768805891178e-05,
+      "loss": 4.8866,
+      "step": 2228
+    },
+    {
+      "epoch": 0.8025202520252025,
+      "grad_norm": 0.8058616518974304,
+      "learning_rate": 1.927472162006717e-05,
+      "loss": 4.5184,
+      "step": 2229
+    },
+    {
+      "epoch": 0.8028802880288028,
+      "grad_norm": 0.690186619758606,
+      "learning_rate": 1.920678157111776e-05,
+      "loss": 4.657,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8032403240324032,
+      "grad_norm": 0.7219494581222534,
+      "learning_rate": 1.9138948749211472e-05,
+      "loss": 4.7445,
+      "step": 2231
+    },
+    {
+      "epoch": 0.8036003600360035,
+      "grad_norm": 0.6288260221481323,
+      "learning_rate": 1.9071223244374614e-05,
+      "loss": 4.9032,
+      "step": 2232
+    },
+    {
+      "epoch": 0.803960396039604,
+      "grad_norm": 0.7226089239120483,
+      "learning_rate": 1.9003605146491054e-05,
+      "loss": 5.1746,
+      "step": 2233
+    },
+    {
+      "epoch": 0.8043204320432044,
+      "grad_norm": 0.8098820447921753,
+      "learning_rate": 1.8936094545302095e-05,
+      "loss": 4.8424,
+      "step": 2234
+    },
+    {
+      "epoch": 0.8046804680468047,
+      "grad_norm": 0.7526129484176636,
+      "learning_rate": 1.8868691530406336e-05,
+      "loss": 4.5838,
+      "step": 2235
+    },
+    {
+      "epoch": 0.8050405040504051,
+      "grad_norm": 0.8739807605743408,
+      "learning_rate": 1.8801396191259645e-05,
+      "loss": 4.7443,
+      "step": 2236
+    },
+    {
+      "epoch": 0.8054005400540054,
+      "grad_norm": 0.6892088651657104,
+      "learning_rate": 1.8734208617174988e-05,
+      "loss": 4.9657,
+      "step": 2237
+    },
+    {
+      "epoch": 0.8057605760576058,
+      "grad_norm": 0.5015419125556946,
+      "learning_rate": 1.866712889732225e-05,
+      "loss": 4.7528,
+      "step": 2238
+    },
+    {
+      "epoch": 0.8061206120612061,
+      "grad_norm": 0.5005907416343689,
+      "learning_rate": 1.8600157120728244e-05,
+      "loss": 4.7449,
+      "step": 2239
+    },
+    {
+      "epoch": 0.8064806480648065,
+      "grad_norm": 0.526184618473053,
+      "learning_rate": 1.8533293376276472e-05,
+      "loss": 4.6524,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8068406840684068,
+      "grad_norm": 0.6554297804832458,
+      "learning_rate": 1.8466537752707068e-05,
+      "loss": 4.8402,
+      "step": 2241
+    },
+    {
+      "epoch": 0.8072007200720072,
+      "grad_norm": 0.6219531893730164,
+      "learning_rate": 1.839989033861673e-05,
+      "loss": 4.6053,
+      "step": 2242
+    },
+    {
+      "epoch": 0.8075607560756075,
+      "grad_norm": 0.6740393042564392,
+      "learning_rate": 1.8333351222458407e-05,
+      "loss": 4.385,
+      "step": 2243
+    },
+    {
+      "epoch": 0.807920792079208,
+      "grad_norm": 0.7463712096214294,
+      "learning_rate": 1.826692049254145e-05,
+      "loss": 4.8666,
+      "step": 2244
+    },
+    {
+      "epoch": 0.8082808280828083,
+      "grad_norm": 0.6941218376159668,
+      "learning_rate": 1.820059823703133e-05,
+      "loss": 4.6104,
+      "step": 2245
+    },
+    {
+      "epoch": 0.8086408640864087,
+      "grad_norm": 0.8766574859619141,
+      "learning_rate": 1.8134384543949478e-05,
+      "loss": 4.6754,
+      "step": 2246
+    },
+    {
+      "epoch": 0.809000900090009,
+      "grad_norm": 0.7981788516044617,
+      "learning_rate": 1.8068279501173335e-05,
+      "loss": 4.8805,
+      "step": 2247
+    },
+    {
+      "epoch": 0.8093609360936094,
+      "grad_norm": 1.02590811252594,
+      "learning_rate": 1.8002283196436097e-05,
+      "loss": 4.9571,
+      "step": 2248
+    },
+    {
+      "epoch": 0.8097209720972097,
+      "grad_norm": 1.0470219850540161,
+      "learning_rate": 1.7936395717326704e-05,
+      "loss": 5.117,
+      "step": 2249
+    },
+    {
+      "epoch": 0.8100810081008101,
+      "grad_norm": 1.3908179998397827,
+      "learning_rate": 1.787061715128956e-05,
+      "loss": 5.2106,
+      "step": 2250
+    },
+    {
+      "epoch": 0.8104410441044104,
+      "grad_norm": 0.970249354839325,
+      "learning_rate": 1.7804947585624588e-05,
+      "loss": 5.0495,
+      "step": 2251
+    },
+    {
+      "epoch": 0.8108010801080108,
+      "grad_norm": 0.7076825499534607,
+      "learning_rate": 1.773938710748706e-05,
+      "loss": 4.6217,
+      "step": 2252
+    },
+    {
+      "epoch": 0.8111611161116111,
+      "grad_norm": 0.657702624797821,
+      "learning_rate": 1.7673935803887453e-05,
+      "loss": 4.4113,
+      "step": 2253
+    },
+    {
+      "epoch": 0.8115211521152115,
+      "grad_norm": 0.6246639490127563,
+      "learning_rate": 1.760859376169133e-05,
+      "loss": 4.5696,
+      "step": 2254
+    },
+    {
+      "epoch": 0.8118811881188119,
+      "grad_norm": 0.5377760529518127,
+      "learning_rate": 1.754336106761927e-05,
+      "loss": 4.5045,
+      "step": 2255
+    },
+    {
+      "epoch": 0.8122412241224123,
+      "grad_norm": 0.5879418849945068,
+      "learning_rate": 1.7478237808246722e-05,
+      "loss": 5.1566,
+      "step": 2256
+    },
+    {
+      "epoch": 0.8126012601260126,
+      "grad_norm": 0.6851582527160645,
+      "learning_rate": 1.741322407000391e-05,
+      "loss": 4.776,
+      "step": 2257
+    },
+    {
+      "epoch": 0.812961296129613,
+      "grad_norm": 1.0152539014816284,
+      "learning_rate": 1.7348319939175637e-05,
+      "loss": 4.4992,
+      "step": 2258
+    },
+    {
+      "epoch": 0.8133213321332133,
+      "grad_norm": 0.6916372179985046,
+      "learning_rate": 1.7283525501901323e-05,
+      "loss": 4.1276,
+      "step": 2259
+    },
+    {
+      "epoch": 0.8136813681368137,
+      "grad_norm": 0.5402399897575378,
+      "learning_rate": 1.7218840844174754e-05,
+      "loss": 4.7419,
+      "step": 2260
+    },
+    {
+      "epoch": 0.814041404140414,
+      "grad_norm": 0.48339545726776123,
+      "learning_rate": 1.715426605184407e-05,
+      "loss": 4.3284,
+      "step": 2261
+    },
+    {
+      "epoch": 0.8144014401440144,
+      "grad_norm": 0.7083166241645813,
+      "learning_rate": 1.70898012106115e-05,
+      "loss": 4.5239,
+      "step": 2262
+    },
+    {
+      "epoch": 0.8147614761476147,
+      "grad_norm": 0.5646843314170837,
+      "learning_rate": 1.7025446406033453e-05,
+      "loss": 4.5845,
+      "step": 2263
+    },
+    {
+      "epoch": 0.8151215121512151,
+      "grad_norm": 0.7311326265335083,
+      "learning_rate": 1.696120172352025e-05,
+      "loss": 4.5815,
+      "step": 2264
+    },
+    {
+      "epoch": 0.8154815481548154,
+      "grad_norm": 0.743579626083374,
+      "learning_rate": 1.6897067248336095e-05,
+      "loss": 5.0977,
+      "step": 2265
+    },
+    {
+      "epoch": 0.8158415841584158,
+      "grad_norm": 0.6290472149848938,
+      "learning_rate": 1.683304306559884e-05,
+      "loss": 4.7559,
+      "step": 2266
+    },
+    {
+      "epoch": 0.8162016201620163,
+      "grad_norm": 0.5165727734565735,
+      "learning_rate": 1.676912926028007e-05,
+      "loss": 4.8021,
+      "step": 2267
+    },
+    {
+      "epoch": 0.8165616561656166,
+      "grad_norm": 0.5954656600952148,
+      "learning_rate": 1.6705325917204805e-05,
+      "loss": 4.7522,
+      "step": 2268
+    },
+    {
+      "epoch": 0.816921692169217,
+      "grad_norm": 0.6288970708847046,
+      "learning_rate": 1.66416331210515e-05,
+      "loss": 4.7196,
+      "step": 2269
+    },
+    {
+      "epoch": 0.8172817281728173,
+      "grad_norm": 0.6287972331047058,
+      "learning_rate": 1.6578050956351886e-05,
+      "loss": 5.1404,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8176417641764177,
+      "grad_norm": 0.8024221658706665,
+      "learning_rate": 1.6514579507490848e-05,
+      "loss": 5.0936,
+      "step": 2271
+    },
+    {
+      "epoch": 0.818001800180018,
+      "grad_norm": 0.5822760462760925,
+      "learning_rate": 1.6451218858706374e-05,
+      "loss": 4.9177,
+      "step": 2272
+    },
+    {
+      "epoch": 0.8183618361836184,
+      "grad_norm": 0.8307278752326965,
+      "learning_rate": 1.6387969094089316e-05,
+      "loss": 5.2499,
+      "step": 2273
+    },
+    {
+      "epoch": 0.8187218721872187,
+      "grad_norm": 0.9024346470832825,
+      "learning_rate": 1.632483029758345e-05,
+      "loss": 5.276,
+      "step": 2274
+    },
+    {
+      "epoch": 0.819081908190819,
+      "grad_norm": 1.2056093215942383,
+      "learning_rate": 1.626180255298525e-05,
+      "loss": 5.3459,
+      "step": 2275
+    },
+    {
+      "epoch": 0.8194419441944194,
+      "grad_norm": 1.9106166362762451,
+      "learning_rate": 1.619888594394382e-05,
+      "loss": 4.8601,
+      "step": 2276
+    },
+    {
+      "epoch": 0.8198019801980198,
+      "grad_norm": 0.6660627126693726,
+      "learning_rate": 1.6136080553960687e-05,
+      "loss": 4.7553,
+      "step": 2277
+    },
+    {
+      "epoch": 0.8201620162016202,
+      "grad_norm": 0.4384493827819824,
+      "learning_rate": 1.6073386466389872e-05,
+      "loss": 4.5587,
+      "step": 2278
+    },
+    {
+      "epoch": 0.8205220522052206,
+      "grad_norm": 0.5030012726783752,
+      "learning_rate": 1.601080376443763e-05,
+      "loss": 4.5943,
+      "step": 2279
+    },
+    {
+      "epoch": 0.8208820882088209,
+      "grad_norm": 0.5574566721916199,
+      "learning_rate": 1.5948332531162413e-05,
+      "loss": 4.5229,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8212421242124213,
+      "grad_norm": 0.519743025302887,
+      "learning_rate": 1.5885972849474672e-05,
+      "loss": 4.5851,
+      "step": 2281
+    },
+    {
+      "epoch": 0.8216021602160216,
+      "grad_norm": 0.5831529498100281,
+      "learning_rate": 1.5823724802136865e-05,
+      "loss": 4.7909,
+      "step": 2282
+    },
+    {
+      "epoch": 0.821962196219622,
+      "grad_norm": 0.8960453271865845,
+      "learning_rate": 1.576158847176329e-05,
+      "loss": 4.8842,
+      "step": 2283
+    },
+    {
+      "epoch": 0.8223222322232223,
+      "grad_norm": 0.6710848808288574,
+      "learning_rate": 1.5699563940819962e-05,
+      "loss": 4.8542,
+      "step": 2284
+    },
+    {
+      "epoch": 0.8226822682268227,
+      "grad_norm": 0.5260566473007202,
+      "learning_rate": 1.5637651291624523e-05,
+      "loss": 4.6999,
+      "step": 2285
+    },
+    {
+      "epoch": 0.823042304230423,
+      "grad_norm": 0.6472052335739136,
+      "learning_rate": 1.557585060634612e-05,
+      "loss": 4.5737,
+      "step": 2286
+    },
+    {
+      "epoch": 0.8234023402340234,
+      "grad_norm": 0.757953941822052,
+      "learning_rate": 1.5514161967005337e-05,
+      "loss": 4.4606,
+      "step": 2287
+    },
+    {
+      "epoch": 0.8237623762376237,
+      "grad_norm": 0.5936845541000366,
+      "learning_rate": 1.5452585455473977e-05,
+      "loss": 4.5589,
+      "step": 2288
+    },
+    {
+      "epoch": 0.8241224122412241,
+      "grad_norm": 0.6369442343711853,
+      "learning_rate": 1.539112115347511e-05,
+      "loss": 4.963,
+      "step": 2289
+    },
+    {
+      "epoch": 0.8244824482448245,
+      "grad_norm": 0.8031920194625854,
+      "learning_rate": 1.5329769142582827e-05,
+      "loss": 4.9157,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8248424842484249,
+      "grad_norm": 0.756523847579956,
+      "learning_rate": 1.526852950422226e-05,
+      "loss": 4.6052,
+      "step": 2291
+    },
+    {
+      "epoch": 0.8252025202520252,
+      "grad_norm": 0.6446996331214905,
+      "learning_rate": 1.5207402319669306e-05,
+      "loss": 4.7325,
+      "step": 2292
+    },
+    {
+      "epoch": 0.8255625562556256,
+      "grad_norm": 0.6594321727752686,
+      "learning_rate": 1.5146387670050687e-05,
+      "loss": 4.7268,
+      "step": 2293
+    },
+    {
+      "epoch": 0.8259225922592259,
+      "grad_norm": 0.6736615896224976,
+      "learning_rate": 1.5085485636343755e-05,
+      "loss": 5.0275,
+      "step": 2294
+    },
+    {
+      "epoch": 0.8262826282628263,
+      "grad_norm": 0.5317438244819641,
+      "learning_rate": 1.5024696299376407e-05,
+      "loss": 4.7624,
+      "step": 2295
+    },
+    {
+      "epoch": 0.8266426642664266,
+      "grad_norm": 0.6013367772102356,
+      "learning_rate": 1.4964019739826907e-05,
+      "loss": 4.5998,
+      "step": 2296
+    },
+    {
+      "epoch": 0.827002700270027,
+      "grad_norm": 0.5864752531051636,
+      "learning_rate": 1.4903456038223939e-05,
+      "loss": 4.7793,
+      "step": 2297
+    },
+    {
+      "epoch": 0.8273627362736273,
+      "grad_norm": 1.0099995136260986,
+      "learning_rate": 1.4843005274946365e-05,
+      "loss": 5.441,
+      "step": 2298
+    },
+    {
+      "epoch": 0.8277227722772277,
+      "grad_norm": 1.1725364923477173,
+      "learning_rate": 1.4782667530223126e-05,
+      "loss": 4.894,
+      "step": 2299
+    },
+    {
+      "epoch": 0.828082808280828,
+      "grad_norm": 1.4512721300125122,
+      "learning_rate": 1.4722442884133214e-05,
+      "loss": 5.316,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8284428442844285,
+      "grad_norm": 0.6787356734275818,
+      "learning_rate": 1.4662331416605501e-05,
+      "loss": 4.4722,
+      "step": 2301
+    },
+    {
+      "epoch": 0.8288028802880288,
+      "grad_norm": 1.0046409368515015,
+      "learning_rate": 1.4602333207418651e-05,
+      "loss": 4.455,
+      "step": 2302
+    },
+    {
+      "epoch": 0.8291629162916292,
+      "grad_norm": 0.7777742743492126,
+      "learning_rate": 1.454244833620102e-05,
+      "loss": 4.277,
+      "step": 2303
+    },
+    {
+      "epoch": 0.8295229522952295,
+      "grad_norm": 1.2001330852508545,
+      "learning_rate": 1.4482676882430502e-05,
+      "loss": 4.9538,
+      "step": 2304
+    },
+    {
+      "epoch": 0.8298829882988299,
+      "grad_norm": 0.6005185842514038,
+      "learning_rate": 1.4423018925434528e-05,
+      "loss": 4.5779,
+      "step": 2305
+    },
+    {
+      "epoch": 0.8302430243024302,
+      "grad_norm": 0.7251618504524231,
+      "learning_rate": 1.4363474544389877e-05,
+      "loss": 4.6006,
+      "step": 2306
+    },
+    {
+      "epoch": 0.8306030603060306,
+      "grad_norm": 0.9123652577400208,
+      "learning_rate": 1.4304043818322565e-05,
+      "loss": 4.927,
+      "step": 2307
+    },
+    {
+      "epoch": 0.8309630963096309,
+      "grad_norm": 0.9391204118728638,
+      "learning_rate": 1.424472682610779e-05,
+      "loss": 5.0999,
+      "step": 2308
+    },
+    {
+      "epoch": 0.8313231323132313,
+      "grad_norm": 0.6396461725234985,
+      "learning_rate": 1.4185523646469822e-05,
+      "loss": 4.552,
+      "step": 2309
+    },
+    {
+      "epoch": 0.8316831683168316,
+      "grad_norm": 0.5968081951141357,
+      "learning_rate": 1.4126434357981877e-05,
+      "loss": 4.6309,
+      "step": 2310
+    },
+    {
+      "epoch": 0.832043204320432,
+      "grad_norm": 0.8504002690315247,
+      "learning_rate": 1.4067459039065956e-05,
+      "loss": 4.995,
+      "step": 2311
+    },
+    {
+      "epoch": 0.8324032403240325,
+      "grad_norm": 0.9265114068984985,
+      "learning_rate": 1.4008597767992871e-05,
+      "loss": 4.8508,
+      "step": 2312
+    },
+    {
+      "epoch": 0.8327632763276328,
+      "grad_norm": 0.5217415690422058,
+      "learning_rate": 1.3949850622882054e-05,
+      "loss": 4.8427,
+      "step": 2313
+    },
+    {
+      "epoch": 0.8331233123312332,
+      "grad_norm": 0.8731891512870789,
+      "learning_rate": 1.3891217681701474e-05,
+      "loss": 4.713,
+      "step": 2314
+    },
+    {
+      "epoch": 0.8334833483348335,
+      "grad_norm": 0.602576732635498,
+      "learning_rate": 1.3832699022267515e-05,
+      "loss": 4.5357,
+      "step": 2315
+    },
+    {
+      "epoch": 0.8338433843384339,
+      "grad_norm": 0.6332751512527466,
+      "learning_rate": 1.3774294722244907e-05,
+      "loss": 4.6608,
+      "step": 2316
+    },
+    {
+      "epoch": 0.8342034203420342,
+      "grad_norm": 0.7543874979019165,
+      "learning_rate": 1.3716004859146592e-05,
+      "loss": 4.6207,
+      "step": 2317
+    },
+    {
+      "epoch": 0.8345634563456346,
+      "grad_norm": 0.888930082321167,
+      "learning_rate": 1.3657829510333654e-05,
+      "loss": 4.6348,
+      "step": 2318
+    },
+    {
+      "epoch": 0.8349234923492349,
+      "grad_norm": 0.8453302979469299,
+      "learning_rate": 1.3599768753015152e-05,
+      "loss": 4.7868,
+      "step": 2319
+    },
+    {
+      "epoch": 0.8352835283528353,
+      "grad_norm": 0.5163218379020691,
+      "learning_rate": 1.3541822664248094e-05,
+      "loss": 4.4345,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8356435643564356,
+      "grad_norm": 0.49913036823272705,
+      "learning_rate": 1.3483991320937306e-05,
+      "loss": 4.961,
+      "step": 2321
+    },
+    {
+      "epoch": 0.836003600360036,
+      "grad_norm": 1.1673535108566284,
+      "learning_rate": 1.3426274799835337e-05,
+      "loss": 4.8401,
+      "step": 2322
+    },
+    {
+      "epoch": 0.8363636363636363,
+      "grad_norm": 0.8345517516136169,
+      "learning_rate": 1.336867317754229e-05,
+      "loss": 5.0522,
+      "step": 2323
+    },
+    {
+      "epoch": 0.8367236723672368,
+      "grad_norm": 0.9774869680404663,
+      "learning_rate": 1.3311186530505838e-05,
+      "loss": 5.1665,
+      "step": 2324
+    },
+    {
+      "epoch": 0.8370837083708371,
+      "grad_norm": 1.130599856376648,
+      "learning_rate": 1.3253814935021026e-05,
+      "loss": 5.1565,
+      "step": 2325
+    },
+    {
+      "epoch": 0.8374437443744375,
+      "grad_norm": 0.9888586401939392,
+      "learning_rate": 1.3196558467230247e-05,
+      "loss": 4.9575,
+      "step": 2326
+    },
+    {
+      "epoch": 0.8378037803780378,
+      "grad_norm": 0.6156508326530457,
+      "learning_rate": 1.3139417203123027e-05,
+      "loss": 4.9267,
+      "step": 2327
+    },
+    {
+      "epoch": 0.8381638163816382,
+      "grad_norm": 0.6930103302001953,
+      "learning_rate": 1.3082391218536061e-05,
+      "loss": 4.7754,
+      "step": 2328
+    },
+    {
+      "epoch": 0.8385238523852385,
+      "grad_norm": 0.7321805357933044,
+      "learning_rate": 1.3025480589153005e-05,
+      "loss": 4.8377,
+      "step": 2329
+    },
+    {
+      "epoch": 0.8388838883888389,
+      "grad_norm": 0.7025576233863831,
+      "learning_rate": 1.2968685390504465e-05,
+      "loss": 4.8009,
+      "step": 2330
+    },
+    {
+      "epoch": 0.8392439243924392,
+      "grad_norm": 0.5497130155563354,
+      "learning_rate": 1.29120056979678e-05,
+      "loss": 4.4953,
+      "step": 2331
+    },
+    {
+      "epoch": 0.8396039603960396,
+      "grad_norm": 0.5606801509857178,
+      "learning_rate": 1.2855441586767113e-05,
+      "loss": 4.4783,
+      "step": 2332
+    },
+    {
+      "epoch": 0.8399639963996399,
+      "grad_norm": 0.5393441915512085,
+      "learning_rate": 1.2798993131973091e-05,
+      "loss": 4.4954,
+      "step": 2333
+    },
+    {
+      "epoch": 0.8403240324032403,
+      "grad_norm": 0.601349949836731,
+      "learning_rate": 1.2742660408502904e-05,
+      "loss": 4.6084,
+      "step": 2334
+    },
+    {
+      "epoch": 0.8406840684068407,
+      "grad_norm": 0.5767045617103577,
+      "learning_rate": 1.2686443491120149e-05,
+      "loss": 4.7144,
+      "step": 2335
+    },
+    {
+      "epoch": 0.8410441044104411,
+      "grad_norm": 0.7713471055030823,
+      "learning_rate": 1.263034245443473e-05,
+      "loss": 4.673,
+      "step": 2336
+    },
+    {
+      "epoch": 0.8414041404140414,
+      "grad_norm": 0.771685779094696,
+      "learning_rate": 1.2574357372902767e-05,
+      "loss": 4.9869,
+      "step": 2337
+    },
+    {
+      "epoch": 0.8417641764176418,
+      "grad_norm": 0.5168091654777527,
+      "learning_rate": 1.2518488320826449e-05,
+      "loss": 4.5899,
+      "step": 2338
+    },
+    {
+      "epoch": 0.8421242124212421,
+      "grad_norm": 1.225448727607727,
+      "learning_rate": 1.2462735372353996e-05,
+      "loss": 4.7007,
+      "step": 2339
+    },
+    {
+      "epoch": 0.8424842484248425,
+      "grad_norm": 0.5362923741340637,
+      "learning_rate": 1.2407098601479539e-05,
+      "loss": 4.696,
+      "step": 2340
+    },
+    {
+      "epoch": 0.8428442844284428,
+      "grad_norm": 0.6797324419021606,
+      "learning_rate": 1.2351578082043047e-05,
+      "loss": 4.791,
+      "step": 2341
+    },
+    {
+      "epoch": 0.8432043204320432,
+      "grad_norm": 1.1674048900604248,
+      "learning_rate": 1.2296173887730123e-05,
+      "loss": 4.9394,
+      "step": 2342
+    },
+    {
+      "epoch": 0.8435643564356435,
+      "grad_norm": 0.6934359073638916,
+      "learning_rate": 1.2240886092072068e-05,
+      "loss": 4.5077,
+      "step": 2343
+    },
+    {
+      "epoch": 0.8439243924392439,
+      "grad_norm": 0.5877205729484558,
+      "learning_rate": 1.2185714768445667e-05,
+      "loss": 4.5705,
+      "step": 2344
+    },
+    {
+      "epoch": 0.8442844284428442,
+      "grad_norm": 0.6093941926956177,
+      "learning_rate": 1.2130659990073146e-05,
+      "loss": 4.5415,
+      "step": 2345
+    },
+    {
+      "epoch": 0.8446444644464447,
+      "grad_norm": 0.7287856936454773,
+      "learning_rate": 1.2075721830021969e-05,
+      "loss": 5.0374,
+      "step": 2346
+    },
+    {
+      "epoch": 0.845004500450045,
+      "grad_norm": 0.8647234439849854,
+      "learning_rate": 1.2020900361204968e-05,
+      "loss": 4.6786,
+      "step": 2347
+    },
+    {
+      "epoch": 0.8453645364536454,
+      "grad_norm": 0.7940008044242859,
+      "learning_rate": 1.1966195656380031e-05,
+      "loss": 5.0893,
+      "step": 2348
+    },
+    {
+      "epoch": 0.8457245724572457,
+      "grad_norm": 1.0051583051681519,
+      "learning_rate": 1.1911607788150036e-05,
+      "loss": 5.203,
+      "step": 2349
+    },
+    {
+      "epoch": 0.8460846084608461,
+      "grad_norm": 0.9990129470825195,
+      "learning_rate": 1.1857136828962855e-05,
+      "loss": 5.0418,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8464446444644464,
+      "grad_norm": 1.3434467315673828,
+      "learning_rate": 1.1802782851111205e-05,
+      "loss": 5.1628,
+      "step": 2351
+    },
+    {
+      "epoch": 0.8468046804680468,
+      "grad_norm": 0.7505450248718262,
+      "learning_rate": 1.1748545926732535e-05,
+      "loss": 4.6661,
+      "step": 2352
+    },
+    {
+      "epoch": 0.8471647164716472,
+      "grad_norm": 1.0241285562515259,
+      "learning_rate": 1.169442612780891e-05,
+      "loss": 4.7007,
+      "step": 2353
+    },
+    {
+      "epoch": 0.8475247524752475,
+      "grad_norm": 1.4250359535217285,
+      "learning_rate": 1.1640423526166988e-05,
+      "loss": 4.5734,
+      "step": 2354
+    },
+    {
+      "epoch": 0.8478847884788479,
+      "grad_norm": 0.8431483507156372,
+      "learning_rate": 1.158653819347788e-05,
+      "loss": 4.6603,
+      "step": 2355
+    },
+    {
+      "epoch": 0.8482448244824482,
+      "grad_norm": 0.6706793904304504,
+      "learning_rate": 1.1532770201257082e-05,
+      "loss": 4.7574,
+      "step": 2356
+    },
+    {
+      "epoch": 0.8486048604860486,
+      "grad_norm": 0.6527566909790039,
+      "learning_rate": 1.1479119620864276e-05,
+      "loss": 4.6939,
+      "step": 2357
+    },
+    {
+      "epoch": 0.848964896489649,
+      "grad_norm": 0.846868634223938,
+      "learning_rate": 1.1425586523503395e-05,
+      "loss": 4.7602,
+      "step": 2358
+    },
+    {
+      "epoch": 0.8493249324932494,
+      "grad_norm": 0.8098002672195435,
+      "learning_rate": 1.1372170980222441e-05,
+      "loss": 4.9113,
+      "step": 2359
+    },
+    {
+      "epoch": 0.8496849684968497,
+      "grad_norm": 0.6968252062797546,
+      "learning_rate": 1.1318873061913405e-05,
+      "loss": 4.8477,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8500450045004501,
+      "grad_norm": 0.6603767275810242,
+      "learning_rate": 1.1265692839312092e-05,
+      "loss": 4.7987,
+      "step": 2361
+    },
+    {
+      "epoch": 0.8504050405040504,
+      "grad_norm": 0.9465769529342651,
+      "learning_rate": 1.1212630382998213e-05,
+      "loss": 4.5938,
+      "step": 2362
+    },
+    {
+      "epoch": 0.8507650765076508,
+      "grad_norm": 0.6345024704933167,
+      "learning_rate": 1.1159685763395111e-05,
+      "loss": 4.7153,
+      "step": 2363
+    },
+    {
+      "epoch": 0.8511251125112511,
+      "grad_norm": 0.7378025054931641,
+      "learning_rate": 1.1106859050769769e-05,
+      "loss": 4.5385,
+      "step": 2364
+    },
+    {
+      "epoch": 0.8514851485148515,
+      "grad_norm": 0.9905508160591125,
+      "learning_rate": 1.1054150315232681e-05,
+      "loss": 4.7941,
+      "step": 2365
+    },
+    {
+      "epoch": 0.8518451845184518,
+      "grad_norm": 0.8407430648803711,
+      "learning_rate": 1.1001559626737756e-05,
+      "loss": 4.7788,
+      "step": 2366
+    },
+    {
+      "epoch": 0.8522052205220522,
+      "grad_norm": 0.8498520255088806,
+      "learning_rate": 1.0949087055082252e-05,
+      "loss": 4.1181,
+      "step": 2367
+    },
+    {
+      "epoch": 0.8525652565256525,
+      "grad_norm": 0.9800708889961243,
+      "learning_rate": 1.089673266990663e-05,
+      "loss": 4.8319,
+      "step": 2368
+    },
+    {
+      "epoch": 0.852925292529253,
+      "grad_norm": 0.7625902891159058,
+      "learning_rate": 1.0844496540694515e-05,
+      "loss": 4.7761,
+      "step": 2369
+    },
+    {
+      "epoch": 0.8532853285328533,
+      "grad_norm": 0.638638436794281,
+      "learning_rate": 1.0792378736772612e-05,
+      "loss": 4.7487,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8536453645364537,
+      "grad_norm": 0.6259344220161438,
+      "learning_rate": 1.0740379327310569e-05,
+      "loss": 4.9022,
+      "step": 2371
+    },
+    {
+      "epoch": 0.854005400540054,
+      "grad_norm": 0.9138006567955017,
+      "learning_rate": 1.0688498381320855e-05,
+      "loss": 5.0246,
+      "step": 2372
+    },
+    {
+      "epoch": 0.8543654365436544,
+      "grad_norm": 0.9508568048477173,
+      "learning_rate": 1.0636735967658784e-05,
+      "loss": 4.9478,
+      "step": 2373
+    },
+    {
+      "epoch": 0.8547254725472547,
+      "grad_norm": 1.1043336391448975,
+      "learning_rate": 1.0585092155022336e-05,
+      "loss": 4.8974,
+      "step": 2374
+    },
+    {
+      "epoch": 0.8550855085508551,
+      "grad_norm": 1.3299425840377808,
+      "learning_rate": 1.0533567011952094e-05,
+      "loss": 5.1492,
+      "step": 2375
+    },
+    {
+      "epoch": 0.8554455445544554,
+      "grad_norm": 0.6157066822052002,
+      "learning_rate": 1.0482160606831093e-05,
+      "loss": 4.7135,
+      "step": 2376
+    },
+    {
+      "epoch": 0.8558055805580558,
+      "grad_norm": 1.1526126861572266,
+      "learning_rate": 1.0430873007884857e-05,
+      "loss": 4.8683,
+      "step": 2377
+    },
+    {
+      "epoch": 0.8561656165616561,
+      "grad_norm": 0.5212879180908203,
+      "learning_rate": 1.0379704283181179e-05,
+      "loss": 4.5955,
+      "step": 2378
+    },
+    {
+      "epoch": 0.8565256525652565,
+      "grad_norm": 0.8130112290382385,
+      "learning_rate": 1.0328654500630108e-05,
+      "loss": 4.7918,
+      "step": 2379
+    },
+    {
+      "epoch": 0.8568856885688569,
+      "grad_norm": 0.7200890183448792,
+      "learning_rate": 1.0277723727983845e-05,
+      "loss": 4.8406,
+      "step": 2380
+    },
+    {
+      "epoch": 0.8572457245724573,
+      "grad_norm": 0.6026584506034851,
+      "learning_rate": 1.0226912032836611e-05,
+      "loss": 4.7515,
+      "step": 2381
+    },
+    {
+      "epoch": 0.8576057605760576,
+      "grad_norm": 0.9684290885925293,
+      "learning_rate": 1.0176219482624616e-05,
+      "loss": 4.8093,
+      "step": 2382
+    },
+    {
+      "epoch": 0.857965796579658,
+      "grad_norm": 0.6427994966506958,
+      "learning_rate": 1.0125646144625955e-05,
+      "loss": 4.5308,
+      "step": 2383
+    },
+    {
+      "epoch": 0.8583258325832583,
+      "grad_norm": 1.033554196357727,
+      "learning_rate": 1.007519208596045e-05,
+      "loss": 4.8341,
+      "step": 2384
+    },
+    {
+      "epoch": 0.8586858685868587,
+      "grad_norm": 0.6669801473617554,
+      "learning_rate": 1.002485737358968e-05,
+      "loss": 4.8964,
+      "step": 2385
+    },
+    {
+      "epoch": 0.859045904590459,
+      "grad_norm": 0.6307418942451477,
+      "learning_rate": 9.974642074316798e-06,
+      "loss": 4.8266,
+      "step": 2386
+    },
+    {
+      "epoch": 0.8594059405940594,
+      "grad_norm": 0.6424444913864136,
+      "learning_rate": 9.924546254786493e-06,
+      "loss": 4.7471,
+      "step": 2387
+    },
+    {
+      "epoch": 0.8597659765976597,
+      "grad_norm": 0.8725467920303345,
+      "learning_rate": 9.874569981484861e-06,
+      "loss": 4.5142,
+      "step": 2388
+    },
+    {
+      "epoch": 0.8601260126012601,
+      "grad_norm": 1.1564704179763794,
+      "learning_rate": 9.824713320739342e-06,
+      "loss": 4.7016,
+      "step": 2389
+    },
+    {
+      "epoch": 0.8604860486048604,
+      "grad_norm": 0.7655138969421387,
+      "learning_rate": 9.774976338718677e-06,
+      "loss": 4.3319,
+      "step": 2390
+    },
+    {
+      "epoch": 0.8608460846084608,
+      "grad_norm": 0.7302666306495667,
+      "learning_rate": 9.725359101432674e-06,
+      "loss": 4.6624,
+      "step": 2391
+    },
+    {
+      "epoch": 0.8612061206120613,
+      "grad_norm": 0.7123817801475525,
+      "learning_rate": 9.675861674732312e-06,
+      "loss": 4.5181,
+      "step": 2392
+    },
+    {
+      "epoch": 0.8615661566156616,
+      "grad_norm": 0.5299736261367798,
+      "learning_rate": 9.62648412430951e-06,
+      "loss": 4.6567,
+      "step": 2393
+    },
+    {
+      "epoch": 0.861926192619262,
+      "grad_norm": 0.711216390132904,
+      "learning_rate": 9.577226515697124e-06,
+      "loss": 4.6318,
+      "step": 2394
+    },
+    {
+      "epoch": 0.8622862286228623,
+      "grad_norm": 0.7324408888816833,
+      "learning_rate": 9.528088914268784e-06,
+      "loss": 4.6413,
+      "step": 2395
+    },
+    {
+      "epoch": 0.8626462646264627,
+      "grad_norm": 0.6073545217514038,
+      "learning_rate": 9.479071385238892e-06,
+      "loss": 4.6813,
+      "step": 2396
+    },
+    {
+      "epoch": 0.863006300630063,
+      "grad_norm": 0.9429351687431335,
+      "learning_rate": 9.430173993662451e-06,
+      "loss": 4.7784,
+      "step": 2397
+    },
+    {
+      "epoch": 0.8633663366336634,
+      "grad_norm": 0.8551303148269653,
+      "learning_rate": 9.381396804435061e-06,
+      "loss": 5.4424,
+      "step": 2398
+    },
+    {
+      "epoch": 0.8637263726372637,
+      "grad_norm": 0.9081370234489441,
+      "learning_rate": 9.332739882292752e-06,
+      "loss": 5.1161,
+      "step": 2399
+    },
+    {
+      "epoch": 0.8640864086408641,
+      "grad_norm": 0.9314940571784973,
+      "learning_rate": 9.284203291811954e-06,
+      "loss": 5.0829,
+      "step": 2400
+    },
+    {
+      "epoch": 0.8644464446444644,
+      "grad_norm": 0.747048556804657,
+      "learning_rate": 9.23578709740942e-06,
+      "loss": 4.671,
+      "step": 2401
+    },
+    {
+      "epoch": 0.8648064806480648,
+      "grad_norm": 0.8901441097259521,
+      "learning_rate": 9.187491363342093e-06,
+      "loss": 4.7503,
+      "step": 2402
+    },
+    {
+      "epoch": 0.8651665166516652,
+      "grad_norm": 0.8733905553817749,
+      "learning_rate": 9.139316153707023e-06,
+      "loss": 4.1668,
+      "step": 2403
+    },
+    {
+      "epoch": 0.8655265526552656,
+      "grad_norm": 0.8293418288230896,
+      "learning_rate": 9.091261532441342e-06,
+      "loss": 4.9468,
+      "step": 2404
+    },
+    {
+      "epoch": 0.8658865886588659,
+      "grad_norm": 0.6938745379447937,
+      "learning_rate": 9.043327563322112e-06,
+      "loss": 4.8042,
+      "step": 2405
+    },
+    {
+      "epoch": 0.8662466246624663,
+      "grad_norm": 0.7350160479545593,
+      "learning_rate": 8.995514309966302e-06,
+      "loss": 5.0458,
+      "step": 2406
+    },
+    {
+      "epoch": 0.8666066606660666,
+      "grad_norm": 0.5033836960792542,
+      "learning_rate": 8.947821835830616e-06,
+      "loss": 4.7879,
+      "step": 2407
+    },
+    {
+      "epoch": 0.866966696669667,
+      "grad_norm": 0.9561224579811096,
+      "learning_rate": 8.900250204211514e-06,
+      "loss": 4.8389,
+      "step": 2408
+    },
+    {
+      "epoch": 0.8673267326732673,
+      "grad_norm": 0.8856688141822815,
+      "learning_rate": 8.852799478245032e-06,
+      "loss": 4.5677,
+      "step": 2409
+    },
+    {
+      "epoch": 0.8676867686768677,
+      "grad_norm": 0.48359963297843933,
+      "learning_rate": 8.80546972090679e-06,
+      "loss": 4.6154,
+      "step": 2410
+    },
+    {
+      "epoch": 0.868046804680468,
+      "grad_norm": 0.6133762001991272,
+      "learning_rate": 8.758260995011825e-06,
+      "loss": 4.818,
+      "step": 2411
+    },
+    {
+      "epoch": 0.8684068406840684,
+      "grad_norm": 0.7412658333778381,
+      "learning_rate": 8.711173363214553e-06,
+      "loss": 4.5966,
+      "step": 2412
+    },
+    {
+      "epoch": 0.8687668766876687,
+      "grad_norm": 0.681463897228241,
+      "learning_rate": 8.664206888008697e-06,
+      "loss": 4.5624,
+      "step": 2413
+    },
+    {
+      "epoch": 0.8691269126912692,
+      "grad_norm": 0.7318177223205566,
+      "learning_rate": 8.617361631727138e-06,
+      "loss": 4.5771,
+      "step": 2414
+    },
+    {
+      "epoch": 0.8694869486948695,
+      "grad_norm": 0.8274372220039368,
+      "learning_rate": 8.570637656541914e-06,
+      "loss": 4.9709,
+      "step": 2415
+    },
+    {
+      "epoch": 0.8698469846984699,
+      "grad_norm": 0.6884218454360962,
+      "learning_rate": 8.524035024464105e-06,
+      "loss": 4.9416,
+      "step": 2416
+    },
+    {
+      "epoch": 0.8702070207020702,
+      "grad_norm": 0.7390003800392151,
+      "learning_rate": 8.47755379734373e-06,
+      "loss": 4.7099,
+      "step": 2417
+    },
+    {
+      "epoch": 0.8705670567056706,
+      "grad_norm": 1.129050374031067,
+      "learning_rate": 8.431194036869672e-06,
+      "loss": 4.8879,
+      "step": 2418
+    },
+    {
+      "epoch": 0.8709270927092709,
+      "grad_norm": 0.7855664491653442,
+      "learning_rate": 8.384955804569627e-06,
+      "loss": 4.8775,
+      "step": 2419
+    },
+    {
+      "epoch": 0.8712871287128713,
+      "grad_norm": 0.5347578525543213,
+      "learning_rate": 8.338839161809997e-06,
+      "loss": 4.6191,
+      "step": 2420
+    },
+    {
+      "epoch": 0.8716471647164716,
+      "grad_norm": 0.754165768623352,
+      "learning_rate": 8.292844169795833e-06,
+      "loss": 4.5964,
+      "step": 2421
+    },
+    {
+      "epoch": 0.872007200720072,
+      "grad_norm": 0.7422668933868408,
+      "learning_rate": 8.24697088957066e-06,
+      "loss": 4.804,
+      "step": 2422
+    },
+    {
+      "epoch": 0.8723672367236723,
+      "grad_norm": 0.8895533680915833,
+      "learning_rate": 8.201219382016556e-06,
+      "loss": 5.1019,
+      "step": 2423
+    },
+    {
+      "epoch": 0.8727272727272727,
+      "grad_norm": 0.7388155460357666,
+      "learning_rate": 8.15558970785395e-06,
+      "loss": 4.7046,
+      "step": 2424
+    },
+    {
+      "epoch": 0.873087308730873,
+      "grad_norm": 1.3550125360488892,
+      "learning_rate": 8.110081927641566e-06,
+      "loss": 4.9702,
+      "step": 2425
+    },
+    {
+      "epoch": 0.8734473447344735,
+      "grad_norm": 2.8201375007629395,
+      "learning_rate": 8.064696101776358e-06,
+      "loss": 5.3071,
+      "step": 2426
+    },
+    {
+      "epoch": 0.8738073807380738,
+      "grad_norm": 1.0882468223571777,
+      "learning_rate": 8.019432290493457e-06,
+      "loss": 4.851,
+      "step": 2427
+    },
+    {
+      "epoch": 0.8741674167416742,
+      "grad_norm": 0.9948346018791199,
+      "learning_rate": 7.974290553866005e-06,
+      "loss": 5.0427,
+      "step": 2428
+    },
+    {
+      "epoch": 0.8745274527452745,
+      "grad_norm": 0.6691415309906006,
+      "learning_rate": 7.929270951805178e-06,
+      "loss": 4.9892,
+      "step": 2429
+    },
+    {
+      "epoch": 0.8748874887488749,
+      "grad_norm": 0.775093138217926,
+      "learning_rate": 7.884373544060009e-06,
+      "loss": 4.3908,
+      "step": 2430
+    },
+    {
+      "epoch": 0.8752475247524752,
+      "grad_norm": 0.6868644952774048,
+      "learning_rate": 7.839598390217396e-06,
+      "loss": 4.7946,
+      "step": 2431
+    },
+    {
+      "epoch": 0.8756075607560756,
+      "grad_norm": 0.6689639091491699,
+      "learning_rate": 7.794945549701993e-06,
+      "loss": 4.8674,
+      "step": 2432
+    },
+    {
+      "epoch": 0.875967596759676,
+      "grad_norm": 0.9124707579612732,
+      "learning_rate": 7.750415081776063e-06,
+      "loss": 4.9911,
+      "step": 2433
+    },
+    {
+      "epoch": 0.8763276327632763,
+      "grad_norm": 0.7038251161575317,
+      "learning_rate": 7.70600704553951e-06,
+      "loss": 4.8584,
+      "step": 2434
+    },
+    {
+      "epoch": 0.8766876687668766,
+      "grad_norm": 0.7156389951705933,
+      "learning_rate": 7.661721499929753e-06,
+      "loss": 4.3274,
+      "step": 2435
+    },
+    {
+      "epoch": 0.877047704770477,
+      "grad_norm": 0.8068670034408569,
+      "learning_rate": 7.6175585037216226e-06,
+      "loss": 4.5658,
+      "step": 2436
+    },
+    {
+      "epoch": 0.8774077407740775,
+      "grad_norm": 0.7935437560081482,
+      "learning_rate": 7.573518115527289e-06,
+      "loss": 4.9122,
+      "step": 2437
+    },
+    {
+      "epoch": 0.8777677767776778,
+      "grad_norm": 0.9261611700057983,
+      "learning_rate": 7.529600393796232e-06,
+      "loss": 4.8509,
+      "step": 2438
+    },
+    {
+      "epoch": 0.8781278127812782,
+      "grad_norm": 0.8355916738510132,
+      "learning_rate": 7.485805396815126e-06,
+      "loss": 4.3652,
+      "step": 2439
+    },
+    {
+      "epoch": 0.8784878487848785,
+      "grad_norm": 0.43560856580734253,
+      "learning_rate": 7.442133182707745e-06,
+      "loss": 4.5542,
+      "step": 2440
+    },
+    {
+      "epoch": 0.8788478847884789,
+      "grad_norm": 0.7524927258491516,
+      "learning_rate": 7.3985838094349444e-06,
+      "loss": 4.7226,
+      "step": 2441
+    },
+    {
+      "epoch": 0.8792079207920792,
+      "grad_norm": 0.7564715147018433,
+      "learning_rate": 7.355157334794516e-06,
+      "loss": 4.7208,
+      "step": 2442
+    },
+    {
+      "epoch": 0.8795679567956796,
+      "grad_norm": 0.9967451691627502,
+      "learning_rate": 7.3118538164211545e-06,
+      "loss": 5.2971,
+      "step": 2443
+    },
+    {
+      "epoch": 0.8799279927992799,
+      "grad_norm": 0.8346577286720276,
+      "learning_rate": 7.2686733117863784e-06,
+      "loss": 4.7256,
+      "step": 2444
+    },
+    {
+      "epoch": 0.8802880288028803,
+      "grad_norm": 0.638346254825592,
+      "learning_rate": 7.225615878198422e-06,
+      "loss": 4.8184,
+      "step": 2445
+    },
+    {
+      "epoch": 0.8806480648064806,
+      "grad_norm": 0.5529339909553528,
+      "learning_rate": 7.1826815728021965e-06,
+      "loss": 4.5564,
+      "step": 2446
+    },
+    {
+      "epoch": 0.881008100810081,
+      "grad_norm": 0.5807334184646606,
+      "learning_rate": 7.1398704525792e-06,
+      "loss": 4.8166,
+      "step": 2447
+    },
+    {
+      "epoch": 0.8813681368136813,
+      "grad_norm": 0.7800282835960388,
+      "learning_rate": 7.097182574347472e-06,
+      "loss": 5.1405,
+      "step": 2448
+    },
+    {
+      "epoch": 0.8817281728172818,
+      "grad_norm": 0.6698582768440247,
+      "learning_rate": 7.054617994761414e-06,
+      "loss": 4.907,
+      "step": 2449
+    },
+    {
+      "epoch": 0.8820882088208821,
+      "grad_norm": 1.5296711921691895,
+      "learning_rate": 7.012176770311862e-06,
+      "loss": 5.3242,
+      "step": 2450
+    },
+    {
+      "epoch": 0.8824482448244825,
+      "grad_norm": 1.2979846000671387,
+      "learning_rate": 6.969858957325904e-06,
+      "loss": 4.922,
+      "step": 2451
+    },
+    {
+      "epoch": 0.8828082808280828,
+      "grad_norm": 0.6412333250045776,
+      "learning_rate": 6.927664611966811e-06,
+      "loss": 5.0412,
+      "step": 2452
+    },
+    {
+      "epoch": 0.8831683168316832,
+      "grad_norm": 0.619648814201355,
+      "learning_rate": 6.8855937902340576e-06,
+      "loss": 4.507,
+      "step": 2453
+    },
+    {
+      "epoch": 0.8835283528352835,
+      "grad_norm": 0.5163532495498657,
+      "learning_rate": 6.843646547963123e-06,
+      "loss": 4.9747,
+      "step": 2454
+    },
+    {
+      "epoch": 0.8838883888388839,
+      "grad_norm": 0.7194183468818665,
+      "learning_rate": 6.801822940825509e-06,
+      "loss": 4.5837,
+      "step": 2455
+    },
+    {
+      "epoch": 0.8842484248424842,
+      "grad_norm": 0.8414213061332703,
+      "learning_rate": 6.760123024328624e-06,
+      "loss": 4.6327,
+      "step": 2456
+    },
+    {
+      "epoch": 0.8846084608460846,
+      "grad_norm": 0.684772253036499,
+      "learning_rate": 6.718546853815688e-06,
+      "loss": 4.9001,
+      "step": 2457
+    },
+    {
+      "epoch": 0.8849684968496849,
+      "grad_norm": 0.47863101959228516,
+      "learning_rate": 6.67709448446574e-06,
+      "loss": 4.8486,
+      "step": 2458
+    },
+    {
+      "epoch": 0.8853285328532853,
+      "grad_norm": 0.6075344681739807,
+      "learning_rate": 6.635765971293484e-06,
+      "loss": 4.9541,
+      "step": 2459
+    },
+    {
+      "epoch": 0.8856885688568857,
+      "grad_norm": 0.5354955196380615,
+      "learning_rate": 6.594561369149199e-06,
+      "loss": 4.6317,
+      "step": 2460
+    },
+    {
+      "epoch": 0.8860486048604861,
+      "grad_norm": 0.7306193113327026,
+      "learning_rate": 6.553480732718808e-06,
+      "loss": 4.524,
+      "step": 2461
+    },
+    {
+      "epoch": 0.8864086408640864,
+      "grad_norm": 0.6062951683998108,
+      "learning_rate": 6.512524116523633e-06,
+      "loss": 4.5702,
+      "step": 2462
+    },
+    {
+      "epoch": 0.8867686768676868,
+      "grad_norm": 0.7549055814743042,
+      "learning_rate": 6.4716915749204465e-06,
+      "loss": 4.6434,
+      "step": 2463
+    },
+    {
+      "epoch": 0.8871287128712871,
+      "grad_norm": 0.83303302526474,
+      "learning_rate": 6.4309831621013005e-06,
+      "loss": 4.7192,
+      "step": 2464
+    },
+    {
+      "epoch": 0.8874887488748875,
+      "grad_norm": 0.6464311480522156,
+      "learning_rate": 6.390398932093555e-06,
+      "loss": 4.7064,
+      "step": 2465
+    },
+    {
+      "epoch": 0.8878487848784878,
+      "grad_norm": 1.1855717897415161,
+      "learning_rate": 6.3499389387597254e-06,
+      "loss": 4.9074,
+      "step": 2466
+    },
+    {
+      "epoch": 0.8882088208820882,
+      "grad_norm": 0.8404142260551453,
+      "learning_rate": 6.30960323579749e-06,
+      "loss": 4.8212,
+      "step": 2467
+    },
+    {
+      "epoch": 0.8885688568856885,
+      "grad_norm": 0.5642232894897461,
+      "learning_rate": 6.269391876739495e-06,
+      "loss": 4.8076,
+      "step": 2468
+    },
+    {
+      "epoch": 0.8889288928892889,
+      "grad_norm": 0.8032687306404114,
+      "learning_rate": 6.229304914953405e-06,
+      "loss": 5.019,
+      "step": 2469
+    },
+    {
+      "epoch": 0.8892889288928892,
+      "grad_norm": 0.8474968671798706,
+      "learning_rate": 6.189342403641807e-06,
+      "loss": 5.0512,
+      "step": 2470
+    },
+    {
+      "epoch": 0.8896489648964897,
+      "grad_norm": 0.6336872577667236,
+      "learning_rate": 6.149504395842087e-06,
+      "loss": 4.6737,
+      "step": 2471
+    },
+    {
+      "epoch": 0.89000900090009,
+      "grad_norm": 0.704339325428009,
+      "learning_rate": 6.109790944426397e-06,
+      "loss": 4.5293,
+      "step": 2472
+    },
+    {
+      "epoch": 0.8903690369036904,
+      "grad_norm": 0.8684128522872925,
+      "learning_rate": 6.070202102101597e-06,
+      "loss": 4.7989,
+      "step": 2473
+    },
+    {
+      "epoch": 0.8907290729072908,
+      "grad_norm": 1.042490839958191,
+      "learning_rate": 6.030737921409169e-06,
+      "loss": 4.9338,
+      "step": 2474
+    },
+    {
+      "epoch": 0.8910891089108911,
+      "grad_norm": 1.1774296760559082,
+      "learning_rate": 5.9913984547250945e-06,
+      "loss": 5.2439,
+      "step": 2475
+    },
+    {
+      "epoch": 0.8914491449144915,
+      "grad_norm": 1.316601037979126,
+      "learning_rate": 5.95218375425991e-06,
+      "loss": 5.2065,
+      "step": 2476
+    },
+    {
+      "epoch": 0.8918091809180918,
+      "grad_norm": 0.9096778631210327,
+      "learning_rate": 5.913093872058528e-06,
+      "loss": 4.6322,
+      "step": 2477
+    },
+    {
+      "epoch": 0.8921692169216922,
+      "grad_norm": 0.6777582764625549,
+      "learning_rate": 5.874128860000216e-06,
+      "loss": 4.4147,
+      "step": 2478
+    },
+    {
+      "epoch": 0.8925292529252925,
+      "grad_norm": 0.5754499435424805,
+      "learning_rate": 5.835288769798486e-06,
+      "loss": 4.833,
+      "step": 2479
+    },
+    {
+      "epoch": 0.8928892889288929,
+      "grad_norm": 1.0769809484481812,
+      "learning_rate": 5.7965736530010916e-06,
+      "loss": 4.4729,
+      "step": 2480
+    },
+    {
+      "epoch": 0.8932493249324932,
+      "grad_norm": 0.6358700394630432,
+      "learning_rate": 5.757983560989921e-06,
+      "loss": 4.9246,
+      "step": 2481
+    },
+    {
+      "epoch": 0.8936093609360936,
+      "grad_norm": 1.1338918209075928,
+      "learning_rate": 5.719518544980929e-06,
+      "loss": 4.7706,
+      "step": 2482
+    },
+    {
+      "epoch": 0.893969396939694,
+      "grad_norm": 0.4643517732620239,
+      "learning_rate": 5.681178656024055e-06,
+      "loss": 4.7676,
+      "step": 2483
+    },
+    {
+      "epoch": 0.8943294329432944,
+      "grad_norm": 1.0510177612304688,
+      "learning_rate": 5.642963945003188e-06,
+      "loss": 4.6983,
+      "step": 2484
+    },
+    {
+      "epoch": 0.8946894689468947,
+      "grad_norm": 0.8604787588119507,
+      "learning_rate": 5.604874462636078e-06,
+      "loss": 4.4971,
+      "step": 2485
+    },
+    {
+      "epoch": 0.8950495049504951,
+      "grad_norm": 0.8846144080162048,
+      "learning_rate": 5.566910259474289e-06,
+      "loss": 4.4979,
+      "step": 2486
+    },
+    {
+      "epoch": 0.8954095409540954,
+      "grad_norm": 0.7206079959869385,
+      "learning_rate": 5.529071385903084e-06,
+      "loss": 4.8595,
+      "step": 2487
+    },
+    {
+      "epoch": 0.8957695769576958,
+      "grad_norm": 0.8029129505157471,
+      "learning_rate": 5.491357892141425e-06,
+      "loss": 4.871,
+      "step": 2488
+    },
+    {
+      "epoch": 0.8961296129612961,
+      "grad_norm": 0.5468530654907227,
+      "learning_rate": 5.453769828241872e-06,
+      "loss": 4.4908,
+      "step": 2489
+    },
+    {
+      "epoch": 0.8964896489648965,
+      "grad_norm": 1.086614727973938,
+      "learning_rate": 5.416307244090502e-06,
+      "loss": 4.8457,
+      "step": 2490
+    },
+    {
+      "epoch": 0.8968496849684968,
+      "grad_norm": 1.064418077468872,
+      "learning_rate": 5.378970189406829e-06,
+      "loss": 4.6813,
+      "step": 2491
+    },
+    {
+      "epoch": 0.8972097209720972,
+      "grad_norm": 0.5295194387435913,
+      "learning_rate": 5.341758713743828e-06,
+      "loss": 4.557,
+      "step": 2492
+    },
+    {
+      "epoch": 0.8975697569756975,
+      "grad_norm": 0.8219357132911682,
+      "learning_rate": 5.304672866487792e-06,
+      "loss": 4.8301,
+      "step": 2493
+    },
+    {
+      "epoch": 0.897929792979298,
+      "grad_norm": 0.9336304664611816,
+      "learning_rate": 5.267712696858229e-06,
+      "loss": 4.7836,
+      "step": 2494
+    },
+    {
+      "epoch": 0.8982898289828983,
+      "grad_norm": 0.8698000907897949,
+      "learning_rate": 5.230878253907912e-06,
+      "loss": 4.6849,
+      "step": 2495
+    },
+    {
+      "epoch": 0.8986498649864987,
+      "grad_norm": 0.6905087232589722,
+      "learning_rate": 5.194169586522734e-06,
+      "loss": 4.7967,
+      "step": 2496
+    },
+    {
+      "epoch": 0.899009900990099,
+      "grad_norm": 0.754138171672821,
+      "learning_rate": 5.157586743421672e-06,
+      "loss": 4.9295,
+      "step": 2497
+    },
+    {
+      "epoch": 0.8993699369936994,
+      "grad_norm": 1.8262755870819092,
+      "learning_rate": 5.121129773156663e-06,
+      "loss": 5.313,
+      "step": 2498
+    },
+    {
+      "epoch": 0.8997299729972997,
+      "grad_norm": 0.7745803594589233,
+      "learning_rate": 5.0847987241126385e-06,
+      "loss": 5.1595,
+      "step": 2499
+    },
+    {
+      "epoch": 0.9000900090009001,
+      "grad_norm": 1.3876433372497559,
+      "learning_rate": 5.0485936445074046e-06,
+      "loss": 5.2019,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9004500450045004,
+      "grad_norm": 1.133023738861084,
+      "learning_rate": 5.012514582391592e-06,
+      "loss": 4.5723,
+      "step": 2501
+    },
+    {
+      "epoch": 0.9008100810081008,
+      "grad_norm": 0.6465590000152588,
+      "learning_rate": 4.976561585648509e-06,
+      "loss": 4.7929,
+      "step": 2502
+    },
+    {
+      "epoch": 0.9011701170117011,
+      "grad_norm": 1.2847857475280762,
+      "learning_rate": 4.9407347019942544e-06,
+      "loss": 4.8718,
+      "step": 2503
+    },
+    {
+      "epoch": 0.9015301530153015,
+      "grad_norm": 0.569114089012146,
+      "learning_rate": 4.905033978977491e-06,
+      "loss": 4.4804,
+      "step": 2504
+    },
+    {
+      "epoch": 0.9018901890189019,
+      "grad_norm": 0.9793164134025574,
+      "learning_rate": 4.869459463979465e-06,
+      "loss": 4.986,
+      "step": 2505
+    },
+    {
+      "epoch": 0.9022502250225023,
+      "grad_norm": 0.5514426231384277,
+      "learning_rate": 4.8340112042139065e-06,
+      "loss": 4.9524,
+      "step": 2506
+    },
+    {
+      "epoch": 0.9026102610261026,
+      "grad_norm": 0.8211607336997986,
+      "learning_rate": 4.798689246727006e-06,
+      "loss": 4.8468,
+      "step": 2507
+    },
+    {
+      "epoch": 0.902970297029703,
+      "grad_norm": 0.5269903540611267,
+      "learning_rate": 4.7634936383973095e-06,
+      "loss": 4.8626,
+      "step": 2508
+    },
+    {
+      "epoch": 0.9033303330333033,
+      "grad_norm": 0.6444000005722046,
+      "learning_rate": 4.728424425935707e-06,
+      "loss": 4.551,
+      "step": 2509
+    },
+    {
+      "epoch": 0.9036903690369037,
+      "grad_norm": 1.075435757637024,
+      "learning_rate": 4.693481655885257e-06,
+      "loss": 4.8247,
+      "step": 2510
+    },
+    {
+      "epoch": 0.904050405040504,
+      "grad_norm": 1.0397629737854004,
+      "learning_rate": 4.658665374621307e-06,
+      "loss": 4.6963,
+      "step": 2511
+    },
+    {
+      "epoch": 0.9044104410441044,
+      "grad_norm": 0.6805405616760254,
+      "learning_rate": 4.623975628351273e-06,
+      "loss": 4.4516,
+      "step": 2512
+    },
+    {
+      "epoch": 0.9047704770477047,
+      "grad_norm": 0.7398169040679932,
+      "learning_rate": 4.58941246311464e-06,
+      "loss": 4.7951,
+      "step": 2513
+    },
+    {
+      "epoch": 0.9051305130513051,
+      "grad_norm": 0.6716864109039307,
+      "learning_rate": 4.554975924782912e-06,
+      "loss": 4.7471,
+      "step": 2514
+    },
+    {
+      "epoch": 0.9054905490549054,
+      "grad_norm": 0.6767914295196533,
+      "learning_rate": 4.520666059059531e-06,
+      "loss": 4.5634,
+      "step": 2515
+    },
+    {
+      "epoch": 0.9058505850585058,
+      "grad_norm": 0.7175542712211609,
+      "learning_rate": 4.486482911479839e-06,
+      "loss": 4.719,
+      "step": 2516
+    },
+    {
+      "epoch": 0.9062106210621063,
+      "grad_norm": 0.9069615602493286,
+      "learning_rate": 4.452426527410947e-06,
+      "loss": 5.5713,
+      "step": 2517
+    },
+    {
+      "epoch": 0.9065706570657066,
+      "grad_norm": 0.6263923048973083,
+      "learning_rate": 4.418496952051798e-06,
+      "loss": 4.7829,
+      "step": 2518
+    },
+    {
+      "epoch": 0.906930693069307,
+      "grad_norm": 0.7558562159538269,
+      "learning_rate": 4.384694230432984e-06,
+      "loss": 4.9266,
+      "step": 2519
+    },
+    {
+      "epoch": 0.9072907290729073,
+      "grad_norm": 0.6696991324424744,
+      "learning_rate": 4.351018407416763e-06,
+      "loss": 4.3571,
+      "step": 2520
+    },
+    {
+      "epoch": 0.9076507650765077,
+      "grad_norm": 0.6993823051452637,
+      "learning_rate": 4.317469527696983e-06,
+      "loss": 5.2419,
+      "step": 2521
+    },
+    {
+      "epoch": 0.908010801080108,
+      "grad_norm": 0.6072081923484802,
+      "learning_rate": 4.2840476357989825e-06,
+      "loss": 4.9883,
+      "step": 2522
+    },
+    {
+      "epoch": 0.9083708370837084,
+      "grad_norm": 0.8503673672676086,
+      "learning_rate": 4.250752776079614e-06,
+      "loss": 5.0176,
+      "step": 2523
+    },
+    {
+      "epoch": 0.9087308730873087,
+      "grad_norm": 0.9142279624938965,
+      "learning_rate": 4.217584992727108e-06,
+      "loss": 5.2182,
+      "step": 2524
+    },
+    {
+      "epoch": 0.9090909090909091,
+      "grad_norm": 1.77701735496521,
+      "learning_rate": 4.184544329761009e-06,
+      "loss": 5.2836,
+      "step": 2525
+    },
+    {
+      "epoch": 0.9094509450945094,
+      "grad_norm": 2.4094419479370117,
+      "learning_rate": 4.151630831032205e-06,
+      "loss": 4.853,
+      "step": 2526
+    },
+    {
+      "epoch": 0.9098109810981098,
+      "grad_norm": 0.5602378249168396,
+      "learning_rate": 4.118844540222788e-06,
+      "loss": 5.1699,
+      "step": 2527
+    },
+    {
+      "epoch": 0.9101710171017102,
+      "grad_norm": 0.7844763994216919,
+      "learning_rate": 4.0861855008460405e-06,
+      "loss": 4.8816,
+      "step": 2528
+    },
+    {
+      "epoch": 0.9105310531053106,
+      "grad_norm": 0.5660812258720398,
+      "learning_rate": 4.0536537562463225e-06,
+      "loss": 4.8106,
+      "step": 2529
+    },
+    {
+      "epoch": 0.9108910891089109,
+      "grad_norm": 0.5048322081565857,
+      "learning_rate": 4.021249349599077e-06,
+      "loss": 4.2835,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9112511251125113,
+      "grad_norm": 0.7268801927566528,
+      "learning_rate": 3.988972323910778e-06,
+      "loss": 4.59,
+      "step": 2531
+    },
+    {
+      "epoch": 0.9116111611161116,
+      "grad_norm": 0.7188135981559753,
+      "learning_rate": 3.95682272201876e-06,
+      "loss": 4.6502,
+      "step": 2532
+    },
+    {
+      "epoch": 0.911971197119712,
+      "grad_norm": 0.6325691342353821,
+      "learning_rate": 3.924800586591326e-06,
+      "loss": 4.7787,
+      "step": 2533
+    },
+    {
+      "epoch": 0.9123312331233123,
+      "grad_norm": 0.5503108501434326,
+      "learning_rate": 3.892905960127546e-06,
+      "loss": 4.6064,
+      "step": 2534
+    },
+    {
+      "epoch": 0.9126912691269127,
+      "grad_norm": 0.8401983380317688,
+      "learning_rate": 3.861138884957316e-06,
+      "loss": 4.3366,
+      "step": 2535
+    },
+    {
+      "epoch": 0.913051305130513,
+      "grad_norm": 0.521528422832489,
+      "learning_rate": 3.829499403241221e-06,
+      "loss": 4.7727,
+      "step": 2536
+    },
+    {
+      "epoch": 0.9134113411341134,
+      "grad_norm": 0.5778352618217468,
+      "learning_rate": 3.797987556970495e-06,
+      "loss": 4.9353,
+      "step": 2537
+    },
+    {
+      "epoch": 0.9137713771377137,
+      "grad_norm": 0.5739848613739014,
+      "learning_rate": 3.7666033879670048e-06,
+      "loss": 4.973,
+      "step": 2538
+    },
+    {
+      "epoch": 0.9141314131413142,
+      "grad_norm": 0.5352855324745178,
+      "learning_rate": 3.735346937883144e-06,
+      "loss": 4.511,
+      "step": 2539
+    },
+    {
+      "epoch": 0.9144914491449145,
+      "grad_norm": 0.7256152033805847,
+      "learning_rate": 3.7042182482018075e-06,
+      "loss": 4.5012,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9148514851485149,
+      "grad_norm": 0.8612756133079529,
+      "learning_rate": 3.6732173602363363e-06,
+      "loss": 4.6615,
+      "step": 2541
+    },
+    {
+      "epoch": 0.9152115211521152,
+      "grad_norm": 0.6105715036392212,
+      "learning_rate": 3.6423443151304526e-06,
+      "loss": 4.451,
+      "step": 2542
+    },
+    {
+      "epoch": 0.9155715571557156,
+      "grad_norm": 0.6826533079147339,
+      "learning_rate": 3.611599153858214e-06,
+      "loss": 4.6159,
+      "step": 2543
+    },
+    {
+      "epoch": 0.9159315931593159,
+      "grad_norm": 0.6430963277816772,
+      "learning_rate": 3.580981917223913e-06,
+      "loss": 4.7071,
+      "step": 2544
+    },
+    {
+      "epoch": 0.9162916291629163,
+      "grad_norm": 0.6871779561042786,
+      "learning_rate": 3.5504926458621246e-06,
+      "loss": 4.741,
+      "step": 2545
+    },
+    {
+      "epoch": 0.9166516651665166,
+      "grad_norm": 0.7385034561157227,
+      "learning_rate": 3.5201313802375456e-06,
+      "loss": 4.6154,
+      "step": 2546
+    },
+    {
+      "epoch": 0.917011701170117,
+      "grad_norm": 0.9472239017486572,
+      "learning_rate": 3.4898981606450333e-06,
+      "loss": 5.0247,
+      "step": 2547
+    },
+    {
+      "epoch": 0.9173717371737173,
+      "grad_norm": 0.7648311853408813,
+      "learning_rate": 3.4597930272094235e-06,
+      "loss": 5.1778,
+      "step": 2548
+    },
+    {
+      "epoch": 0.9177317731773177,
+      "grad_norm": 1.013818383216858,
+      "learning_rate": 3.4298160198856568e-06,
+      "loss": 4.9648,
+      "step": 2549
+    },
+    {
+      "epoch": 0.918091809180918,
+      "grad_norm": 1.1295243501663208,
+      "learning_rate": 3.3999671784585517e-06,
+      "loss": 5.2376,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9184518451845185,
+      "grad_norm": 3.160092830657959,
+      "learning_rate": 3.370246542542865e-06,
+      "loss": 4.6393,
+      "step": 2551
+    },
+    {
+      "epoch": 0.9188118811881189,
+      "grad_norm": 0.7330535054206848,
+      "learning_rate": 3.3406541515832003e-06,
+      "loss": 4.5272,
+      "step": 2552
+    },
+    {
+      "epoch": 0.9191719171917192,
+      "grad_norm": 0.9496917724609375,
+      "learning_rate": 3.311190044853951e-06,
+      "loss": 5.049,
+      "step": 2553
+    },
+    {
+      "epoch": 0.9195319531953196,
+      "grad_norm": 0.7707210779190063,
+      "learning_rate": 3.2818542614592497e-06,
+      "loss": 4.7573,
+      "step": 2554
+    },
+    {
+      "epoch": 0.9198919891989199,
+      "grad_norm": 0.673030436038971,
+      "learning_rate": 3.252646840332918e-06,
+      "loss": 4.6971,
+      "step": 2555
+    },
+    {
+      "epoch": 0.9202520252025203,
+      "grad_norm": 0.6012186408042908,
+      "learning_rate": 3.2235678202384267e-06,
+      "loss": 4.5512,
+      "step": 2556
+    },
+    {
+      "epoch": 0.9206120612061206,
+      "grad_norm": 1.0911415815353394,
+      "learning_rate": 3.1946172397688267e-06,
+      "loss": 4.3919,
+      "step": 2557
+    },
+    {
+      "epoch": 0.920972097209721,
+      "grad_norm": 0.7308849692344666,
+      "learning_rate": 3.1657951373467497e-06,
+      "loss": 4.8589,
+      "step": 2558
+    },
+    {
+      "epoch": 0.9213321332133213,
+      "grad_norm": 0.676698625087738,
+      "learning_rate": 3.1371015512242306e-06,
+      "loss": 4.8728,
+      "step": 2559
+    },
+    {
+      "epoch": 0.9216921692169217,
+      "grad_norm": 1.072251796722412,
+      "learning_rate": 3.1085365194828075e-06,
+      "loss": 5.1019,
+      "step": 2560
+    },
+    {
+      "epoch": 0.922052205220522,
+      "grad_norm": 0.6631917953491211,
+      "learning_rate": 3.0801000800333877e-06,
+      "loss": 4.6797,
+      "step": 2561
+    },
+    {
+      "epoch": 0.9224122412241225,
+      "grad_norm": 0.5610212683677673,
+      "learning_rate": 3.051792270616216e-06,
+      "loss": 4.6816,
+      "step": 2562
+    },
+    {
+      "epoch": 0.9227722772277228,
+      "grad_norm": 0.5413954257965088,
+      "learning_rate": 3.023613128800795e-06,
+      "loss": 4.6725,
+      "step": 2563
+    },
+    {
+      "epoch": 0.9231323132313232,
+      "grad_norm": 0.7505655288696289,
+      "learning_rate": 2.995562691985898e-06,
+      "loss": 4.7661,
+      "step": 2564
+    },
+    {
+      "epoch": 0.9234923492349235,
+      "grad_norm": 0.810063362121582,
+      "learning_rate": 2.9676409973994566e-06,
+      "loss": 4.773,
+      "step": 2565
+    },
+    {
+      "epoch": 0.9238523852385239,
+      "grad_norm": 1.0670089721679688,
+      "learning_rate": 2.939848082098562e-06,
+      "loss": 5.1925,
+      "step": 2566
+    },
+    {
+      "epoch": 0.9242124212421242,
+      "grad_norm": 0.6282406449317932,
+      "learning_rate": 2.912183982969385e-06,
+      "loss": 4.9323,
+      "step": 2567
+    },
+    {
+      "epoch": 0.9245724572457246,
+      "grad_norm": 0.9556277394294739,
+      "learning_rate": 2.8846487367271135e-06,
+      "loss": 4.8946,
+      "step": 2568
+    },
+    {
+      "epoch": 0.9249324932493249,
+      "grad_norm": 0.5398349761962891,
+      "learning_rate": 2.8572423799159586e-06,
+      "loss": 4.5384,
+      "step": 2569
+    },
+    {
+      "epoch": 0.9252925292529253,
+      "grad_norm": 0.6668215990066528,
+      "learning_rate": 2.8299649489090475e-06,
+      "loss": 4.6729,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9256525652565256,
+      "grad_norm": 0.6637030839920044,
+      "learning_rate": 2.802816479908399e-06,
+      "loss": 4.8677,
+      "step": 2571
+    },
+    {
+      "epoch": 0.926012601260126,
+      "grad_norm": 0.5411921143531799,
+      "learning_rate": 2.7757970089449024e-06,
+      "loss": 4.6855,
+      "step": 2572
+    },
+    {
+      "epoch": 0.9263726372637263,
+      "grad_norm": 0.5961670875549316,
+      "learning_rate": 2.748906571878207e-06,
+      "loss": 4.913,
+      "step": 2573
+    },
+    {
+      "epoch": 0.9267326732673268,
+      "grad_norm": 0.8857282400131226,
+      "learning_rate": 2.722145204396742e-06,
+      "loss": 5.2247,
+      "step": 2574
+    },
+    {
+      "epoch": 0.9270927092709271,
+      "grad_norm": 1.259244680404663,
+      "learning_rate": 2.6955129420176196e-06,
+      "loss": 5.4074,
+      "step": 2575
+    },
+    {
+      "epoch": 0.9274527452745275,
+      "grad_norm": 0.9549702405929565,
+      "learning_rate": 2.6690098200866098e-06,
+      "loss": 4.5297,
+      "step": 2576
+    },
+    {
+      "epoch": 0.9278127812781278,
+      "grad_norm": 0.7229859232902527,
+      "learning_rate": 2.6426358737781098e-06,
+      "loss": 4.9248,
+      "step": 2577
+    },
+    {
+      "epoch": 0.9281728172817282,
+      "grad_norm": 0.722059428691864,
+      "learning_rate": 2.6163911380950425e-06,
+      "loss": 4.8236,
+      "step": 2578
+    },
+    {
+      "epoch": 0.9285328532853285,
+      "grad_norm": 0.6506609916687012,
+      "learning_rate": 2.590275647868867e-06,
+      "loss": 4.4557,
+      "step": 2579
+    },
+    {
+      "epoch": 0.9288928892889289,
+      "grad_norm": 0.7389491200447083,
+      "learning_rate": 2.564289437759515e-06,
+      "loss": 4.7994,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9292529252925292,
+      "grad_norm": 0.5561854839324951,
+      "learning_rate": 2.53843254225532e-06,
+      "loss": 4.2714,
+      "step": 2581
+    },
+    {
+      "epoch": 0.9296129612961296,
+      "grad_norm": 0.7266308069229126,
+      "learning_rate": 2.5127049956730207e-06,
+      "loss": 4.3677,
+      "step": 2582
+    },
+    {
+      "epoch": 0.9299729972997299,
+      "grad_norm": 0.7052050828933716,
+      "learning_rate": 2.4871068321576596e-06,
+      "loss": 4.5698,
+      "step": 2583
+    },
+    {
+      "epoch": 0.9303330333033303,
+      "grad_norm": 0.7006728649139404,
+      "learning_rate": 2.4616380856825716e-06,
+      "loss": 4.7132,
+      "step": 2584
+    },
+    {
+      "epoch": 0.9306930693069307,
+      "grad_norm": 0.6953990459442139,
+      "learning_rate": 2.436298790049363e-06,
+      "loss": 4.6789,
+      "step": 2585
+    },
+    {
+      "epoch": 0.9310531053105311,
+      "grad_norm": 0.6005169749259949,
+      "learning_rate": 2.4110889788877656e-06,
+      "loss": 4.4247,
+      "step": 2586
+    },
+    {
+      "epoch": 0.9314131413141314,
+      "grad_norm": 0.5607119798660278,
+      "learning_rate": 2.3860086856557383e-06,
+      "loss": 4.5032,
+      "step": 2587
+    },
+    {
+      "epoch": 0.9317731773177318,
+      "grad_norm": 0.8107718825340271,
+      "learning_rate": 2.3610579436393e-06,
+      "loss": 4.9354,
+      "step": 2588
+    },
+    {
+      "epoch": 0.9321332133213321,
+      "grad_norm": 0.5992099642753601,
+      "learning_rate": 2.33623678595255e-06,
+      "loss": 4.5928,
+      "step": 2589
+    },
+    {
+      "epoch": 0.9324932493249325,
+      "grad_norm": 0.5759278535842896,
+      "learning_rate": 2.311545245537594e-06,
+      "loss": 4.8887,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9328532853285328,
+      "grad_norm": 0.7263954281806946,
+      "learning_rate": 2.286983355164529e-06,
+      "loss": 4.8844,
+      "step": 2591
+    },
+    {
+      "epoch": 0.9332133213321332,
+      "grad_norm": 0.6867969632148743,
+      "learning_rate": 2.2625511474313685e-06,
+      "loss": 4.5716,
+      "step": 2592
+    },
+    {
+      "epoch": 0.9335733573357335,
+      "grad_norm": 0.6144258379936218,
+      "learning_rate": 2.23824865476403e-06,
+      "loss": 4.7591,
+      "step": 2593
+    },
+    {
+      "epoch": 0.9339333933393339,
+      "grad_norm": 0.6983000040054321,
+      "learning_rate": 2.2140759094162467e-06,
+      "loss": 4.5595,
+      "step": 2594
+    },
+    {
+      "epoch": 0.9342934293429342,
+      "grad_norm": 0.7773663997650146,
+      "learning_rate": 2.1900329434695887e-06,
+      "loss": 4.6514,
+      "step": 2595
+    },
+    {
+      "epoch": 0.9346534653465347,
+      "grad_norm": 0.7720439434051514,
+      "learning_rate": 2.166119788833354e-06,
+      "loss": 4.9227,
+      "step": 2596
+    },
+    {
+      "epoch": 0.9350135013501351,
+      "grad_norm": 0.7285527586936951,
+      "learning_rate": 2.1423364772445887e-06,
+      "loss": 4.8956,
+      "step": 2597
+    },
+    {
+      "epoch": 0.9353735373537354,
+      "grad_norm": 1.0870779752731323,
+      "learning_rate": 2.118683040267999e-06,
+      "loss": 5.0033,
+      "step": 2598
+    },
+    {
+      "epoch": 0.9357335733573358,
+      "grad_norm": 0.965726912021637,
+      "learning_rate": 2.095159509295919e-06,
+      "loss": 5.2102,
+      "step": 2599
+    },
+    {
+      "epoch": 0.9360936093609361,
+      "grad_norm": 1.4614653587341309,
+      "learning_rate": 2.0717659155482738e-06,
+      "loss": 5.5101,
+      "step": 2600
+    },
+    {
+      "epoch": 0.9364536453645365,
+      "grad_norm": 0.5772082209587097,
+      "learning_rate": 2.0485022900725513e-06,
+      "loss": 4.786,
+      "step": 2601
+    },
+    {
+      "epoch": 0.9368136813681368,
+      "grad_norm": 0.7124701142311096,
+      "learning_rate": 2.025368663743743e-06,
+      "loss": 4.4701,
+      "step": 2602
+    },
+    {
+      "epoch": 0.9371737173717372,
+      "grad_norm": 0.7923992872238159,
+      "learning_rate": 2.002365067264289e-06,
+      "loss": 5.1817,
+      "step": 2603
+    },
+    {
+      "epoch": 0.9375337533753375,
+      "grad_norm": 0.6645485758781433,
+      "learning_rate": 1.9794915311641018e-06,
+      "loss": 4.5608,
+      "step": 2604
+    },
+    {
+      "epoch": 0.9378937893789379,
+      "grad_norm": 1.0128847360610962,
+      "learning_rate": 1.9567480858004306e-06,
+      "loss": 4.549,
+      "step": 2605
+    },
+    {
+      "epoch": 0.9382538253825382,
+      "grad_norm": 0.6514415144920349,
+      "learning_rate": 1.9341347613579087e-06,
+      "loss": 5.0031,
+      "step": 2606
+    },
+    {
+      "epoch": 0.9386138613861386,
+      "grad_norm": 0.9877171516418457,
+      "learning_rate": 1.91165158784844e-06,
+      "loss": 5.0519,
+      "step": 2607
+    },
+    {
+      "epoch": 0.938973897389739,
+      "grad_norm": 0.6819136738777161,
+      "learning_rate": 1.889298595111233e-06,
+      "loss": 4.6016,
+      "step": 2608
+    },
+    {
+      "epoch": 0.9393339333933394,
+      "grad_norm": 0.9286605715751648,
+      "learning_rate": 1.8670758128126909e-06,
+      "loss": 4.8841,
+      "step": 2609
+    },
+    {
+      "epoch": 0.9396939693969397,
+      "grad_norm": 0.5582537651062012,
+      "learning_rate": 1.844983270446432e-06,
+      "loss": 4.7466,
+      "step": 2610
+    },
+    {
+      "epoch": 0.9400540054005401,
+      "grad_norm": 0.9149574041366577,
+      "learning_rate": 1.8230209973331914e-06,
+      "loss": 4.5378,
+      "step": 2611
+    },
+    {
+      "epoch": 0.9404140414041404,
+      "grad_norm": 0.5852335691452026,
+      "learning_rate": 1.8011890226208527e-06,
+      "loss": 4.906,
+      "step": 2612
+    },
+    {
+      "epoch": 0.9407740774077408,
+      "grad_norm": 0.7403162717819214,
+      "learning_rate": 1.7794873752843277e-06,
+      "loss": 4.6849,
+      "step": 2613
+    },
+    {
+      "epoch": 0.9411341134113411,
+      "grad_norm": 0.7608280777931213,
+      "learning_rate": 1.7579160841256104e-06,
+      "loss": 4.6213,
+      "step": 2614
+    },
+    {
+      "epoch": 0.9414941494149415,
+      "grad_norm": 0.7691354751586914,
+      "learning_rate": 1.7364751777736332e-06,
+      "loss": 4.6725,
+      "step": 2615
+    },
+    {
+      "epoch": 0.9418541854185418,
+      "grad_norm": 0.7262532114982605,
+      "learning_rate": 1.7151646846843227e-06,
+      "loss": 4.9798,
+      "step": 2616
+    },
+    {
+      "epoch": 0.9422142214221422,
+      "grad_norm": 0.5878902077674866,
+      "learning_rate": 1.6939846331405108e-06,
+      "loss": 4.707,
+      "step": 2617
+    },
+    {
+      "epoch": 0.9425742574257425,
+      "grad_norm": 0.637844443321228,
+      "learning_rate": 1.6729350512519005e-06,
+      "loss": 4.9285,
+      "step": 2618
+    },
+    {
+      "epoch": 0.942934293429343,
+      "grad_norm": 0.6767174005508423,
+      "learning_rate": 1.6520159669550783e-06,
+      "loss": 4.8655,
+      "step": 2619
+    },
+    {
+      "epoch": 0.9432943294329433,
+      "grad_norm": 0.6745604872703552,
+      "learning_rate": 1.6312274080133804e-06,
+      "loss": 4.9893,
+      "step": 2620
+    },
+    {
+      "epoch": 0.9436543654365437,
+      "grad_norm": 0.6165941953659058,
+      "learning_rate": 1.6105694020169593e-06,
+      "loss": 4.676,
+      "step": 2621
+    },
+    {
+      "epoch": 0.944014401440144,
+      "grad_norm": 1.2223420143127441,
+      "learning_rate": 1.5900419763826614e-06,
+      "loss": 4.7306,
+      "step": 2622
+    },
+    {
+      "epoch": 0.9443744374437444,
+      "grad_norm": 0.6925899386405945,
+      "learning_rate": 1.5696451583540827e-06,
+      "loss": 5.2183,
+      "step": 2623
+    },
+    {
+      "epoch": 0.9447344734473447,
+      "grad_norm": 0.6742193102836609,
+      "learning_rate": 1.5493789750014031e-06,
+      "loss": 4.9525,
+      "step": 2624
+    },
+    {
+      "epoch": 0.9450945094509451,
+      "grad_norm": 1.818490982055664,
+      "learning_rate": 1.5292434532215072e-06,
+      "loss": 5.372,
+      "step": 2625
+    },
+    {
+      "epoch": 0.9454545454545454,
+      "grad_norm": 3.1944949626922607,
+      "learning_rate": 1.5092386197378183e-06,
+      "loss": 4.7497,
+      "step": 2626
+    },
+    {
+      "epoch": 0.9458145814581458,
+      "grad_norm": 0.5739015340805054,
+      "learning_rate": 1.489364501100332e-06,
+      "loss": 4.6389,
+      "step": 2627
+    },
+    {
+      "epoch": 0.9461746174617461,
+      "grad_norm": 0.8385379910469055,
+      "learning_rate": 1.4696211236855272e-06,
+      "loss": 4.8675,
+      "step": 2628
+    },
+    {
+      "epoch": 0.9465346534653465,
+      "grad_norm": 0.7263203859329224,
+      "learning_rate": 1.4500085136964326e-06,
+      "loss": 4.5676,
+      "step": 2629
+    },
+    {
+      "epoch": 0.946894689468947,
+      "grad_norm": 0.9063706398010254,
+      "learning_rate": 1.430526697162482e-06,
+      "loss": 4.4987,
+      "step": 2630
+    },
+    {
+      "epoch": 0.9472547254725473,
+      "grad_norm": 0.9547297954559326,
+      "learning_rate": 1.4111756999395154e-06,
+      "loss": 4.7062,
+      "step": 2631
+    },
+    {
+      "epoch": 0.9476147614761476,
+      "grad_norm": 0.7871283292770386,
+      "learning_rate": 1.3919555477097668e-06,
+      "loss": 4.6941,
+      "step": 2632
+    },
+    {
+      "epoch": 0.947974797479748,
+      "grad_norm": 0.8934873342514038,
+      "learning_rate": 1.3728662659818204e-06,
+      "loss": 4.5796,
+      "step": 2633
+    },
+    {
+      "epoch": 0.9483348334833483,
+      "grad_norm": 0.5550655126571655,
+      "learning_rate": 1.3539078800905659e-06,
+      "loss": 4.6309,
+      "step": 2634
+    },
+    {
+      "epoch": 0.9486948694869487,
+      "grad_norm": 0.8748136758804321,
+      "learning_rate": 1.3350804151971653e-06,
+      "loss": 4.7027,
+      "step": 2635
+    },
+    {
+      "epoch": 0.949054905490549,
+      "grad_norm": 0.6412554383277893,
+      "learning_rate": 1.3163838962890195e-06,
+      "loss": 4.7536,
+      "step": 2636
+    },
+    {
+      "epoch": 0.9494149414941494,
+      "grad_norm": 0.5575628280639648,
+      "learning_rate": 1.2978183481797801e-06,
+      "loss": 4.9255,
+      "step": 2637
+    },
+    {
+      "epoch": 0.9497749774977498,
+      "grad_norm": 0.6832976341247559,
+      "learning_rate": 1.2793837955092258e-06,
+      "loss": 4.8386,
+      "step": 2638
+    },
+    {
+      "epoch": 0.9501350135013501,
+      "grad_norm": 0.7369568347930908,
+      "learning_rate": 1.261080262743297e-06,
+      "loss": 4.8507,
+      "step": 2639
+    },
+    {
+      "epoch": 0.9504950495049505,
+      "grad_norm": 0.6423168778419495,
+      "learning_rate": 1.2429077741740736e-06,
+      "loss": 4.5613,
+      "step": 2640
+    },
+    {
+      "epoch": 0.9508550855085508,
+      "grad_norm": 0.6358969807624817,
+      "learning_rate": 1.2248663539196848e-06,
+      "loss": 4.5165,
+      "step": 2641
+    },
+    {
+      "epoch": 0.9512151215121513,
+      "grad_norm": 0.726514995098114,
+      "learning_rate": 1.2069560259243328e-06,
+      "loss": 5.0404,
+      "step": 2642
+    },
+    {
+      "epoch": 0.9515751575157516,
+      "grad_norm": 0.6469401121139526,
+      "learning_rate": 1.1891768139582037e-06,
+      "loss": 4.3748,
+      "step": 2643
+    },
+    {
+      "epoch": 0.951935193519352,
+      "grad_norm": 0.7196247577667236,
+      "learning_rate": 1.1715287416175113e-06,
+      "loss": 4.7141,
+      "step": 2644
+    },
+    {
+      "epoch": 0.9522952295229523,
+      "grad_norm": 0.6038268804550171,
+      "learning_rate": 1.1540118323243865e-06,
+      "loss": 4.5587,
+      "step": 2645
+    },
+    {
+      "epoch": 0.9526552655265527,
+      "grad_norm": 0.6786099076271057,
+      "learning_rate": 1.1366261093268992e-06,
+      "loss": 5.1968,
+      "step": 2646
+    },
+    {
+      "epoch": 0.953015301530153,
+      "grad_norm": 0.5878410339355469,
+      "learning_rate": 1.1193715956990258e-06,
+      "loss": 4.7435,
+      "step": 2647
+    },
+    {
+      "epoch": 0.9533753375337534,
+      "grad_norm": 0.8352496027946472,
+      "learning_rate": 1.1022483143405705e-06,
+      "loss": 4.9383,
+      "step": 2648
+    },
+    {
+      "epoch": 0.9537353735373537,
+      "grad_norm": 0.7436163425445557,
+      "learning_rate": 1.08525628797721e-06,
+      "loss": 4.9565,
+      "step": 2649
+    },
+    {
+      "epoch": 0.9540954095409541,
+      "grad_norm": 0.9482848644256592,
+      "learning_rate": 1.068395539160394e-06,
+      "loss": 4.9916,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9544554455445544,
+      "grad_norm": 1.374624252319336,
+      "learning_rate": 1.0516660902673448e-06,
+      "loss": 4.6837,
+      "step": 2651
+    },
+    {
+      "epoch": 0.9548154815481548,
+      "grad_norm": 0.5657400488853455,
+      "learning_rate": 1.035067963501024e-06,
+      "loss": 4.3842,
+      "step": 2652
+    },
+    {
+      "epoch": 0.9551755175517552,
+      "grad_norm": 0.640466570854187,
+      "learning_rate": 1.018601180890133e-06,
+      "loss": 4.8339,
+      "step": 2653
+    },
+    {
+      "epoch": 0.9555355535553556,
+      "grad_norm": 0.7675738334655762,
+      "learning_rate": 1.0022657642890231e-06,
+      "loss": 4.5358,
+      "step": 2654
+    },
+    {
+      "epoch": 0.9558955895589559,
+      "grad_norm": 0.6688444018363953,
+      "learning_rate": 9.86061735377708e-07,
+      "loss": 4.8008,
+      "step": 2655
+    },
+    {
+      "epoch": 0.9562556255625563,
+      "grad_norm": 0.7679070830345154,
+      "learning_rate": 9.699891156618402e-07,
+      "loss": 4.8804,
+      "step": 2656
+    },
+    {
+      "epoch": 0.9566156615661566,
+      "grad_norm": 0.7303805351257324,
+      "learning_rate": 9.540479264726676e-07,
+      "loss": 4.6286,
+      "step": 2657
+    },
+    {
+      "epoch": 0.956975697569757,
+      "grad_norm": 0.7357656955718994,
+      "learning_rate": 9.382381889669667e-07,
+      "loss": 4.7679,
+      "step": 2658
+    },
+    {
+      "epoch": 0.9573357335733573,
+      "grad_norm": 0.8778092265129089,
+      "learning_rate": 9.225599241271199e-07,
+      "loss": 4.8732,
+      "step": 2659
+    },
+    {
+      "epoch": 0.9576957695769577,
+      "grad_norm": 0.4600130021572113,
+      "learning_rate": 9.070131527609604e-07,
+      "loss": 4.8048,
+      "step": 2660
+    },
+    {
+      "epoch": 0.958055805580558,
+      "grad_norm": 0.6551803946495056,
+      "learning_rate": 8.9159789550185e-07,
+      "loss": 4.506,
+      "step": 2661
+    },
+    {
+      "epoch": 0.9584158415841584,
+      "grad_norm": 0.7119221687316895,
+      "learning_rate": 8.763141728085789e-07,
+      "loss": 4.5518,
+      "step": 2662
+    },
+    {
+      "epoch": 0.9587758775877587,
+      "grad_norm": 0.571420431137085,
+      "learning_rate": 8.611620049653879e-07,
+      "loss": 4.8277,
+      "step": 2663
+    },
+    {
+      "epoch": 0.9591359135913592,
+      "grad_norm": 0.9699096083641052,
+      "learning_rate": 8.461414120819133e-07,
+      "loss": 4.9504,
+      "step": 2664
+    },
+    {
+      "epoch": 0.9594959495949595,
+      "grad_norm": 1.1072616577148438,
+      "learning_rate": 8.312524140931644e-07,
+      "loss": 4.5514,
+      "step": 2665
+    },
+    {
+      "epoch": 0.9598559855985599,
+      "grad_norm": 0.5243213176727295,
+      "learning_rate": 8.16495030759501e-07,
+      "loss": 5.0715,
+      "step": 2666
+    },
+    {
+      "epoch": 0.9602160216021602,
+      "grad_norm": 1.0006332397460938,
+      "learning_rate": 8.018692816666118e-07,
+      "loss": 5.0062,
+      "step": 2667
+    },
+    {
+      "epoch": 0.9605760576057606,
+      "grad_norm": 0.7771649956703186,
+      "learning_rate": 7.873751862254696e-07,
+      "loss": 4.8082,
+      "step": 2668
+    },
+    {
+      "epoch": 0.9609360936093609,
+      "grad_norm": 0.6831532120704651,
+      "learning_rate": 7.730127636723539e-07,
+      "loss": 4.8668,
+      "step": 2669
+    },
+    {
+      "epoch": 0.9612961296129613,
+      "grad_norm": 0.9201724529266357,
+      "learning_rate": 7.587820330687389e-07,
+      "loss": 4.5572,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9616561656165616,
+      "grad_norm": 0.730021595954895,
+      "learning_rate": 7.446830133013616e-07,
+      "loss": 4.5861,
+      "step": 2671
+    },
+    {
+      "epoch": 0.962016201620162,
+      "grad_norm": 0.7570610642433167,
+      "learning_rate": 7.307157230821426e-07,
+      "loss": 5.0164,
+      "step": 2672
+    },
+    {
+      "epoch": 0.9623762376237623,
+      "grad_norm": 0.8982148170471191,
+      "learning_rate": 7.168801809481763e-07,
+      "loss": 5.0613,
+      "step": 2673
+    },
+    {
+      "epoch": 0.9627362736273627,
+      "grad_norm": 0.7683990001678467,
+      "learning_rate": 7.031764052616852e-07,
+      "loss": 4.9976,
+      "step": 2674
+    },
+    {
+      "epoch": 0.963096309630963,
+      "grad_norm": 1.3601149320602417,
+      "learning_rate": 6.896044142100433e-07,
+      "loss": 5.3859,
+      "step": 2675
+    },
+    {
+      "epoch": 0.9634563456345635,
+      "grad_norm": 0.7018367648124695,
+      "learning_rate": 6.761642258056978e-07,
+      "loss": 4.7254,
+      "step": 2676
+    },
+    {
+      "epoch": 0.9638163816381639,
+      "grad_norm": 0.8308577537536621,
+      "learning_rate": 6.628558578862021e-07,
+      "loss": 5.1009,
+      "step": 2677
+    },
+    {
+      "epoch": 0.9641764176417642,
+      "grad_norm": 0.6796321272850037,
+      "learning_rate": 6.496793281141056e-07,
+      "loss": 4.7258,
+      "step": 2678
+    },
+    {
+      "epoch": 0.9645364536453646,
+      "grad_norm": 0.8252047300338745,
+      "learning_rate": 6.366346539770529e-07,
+      "loss": 4.5518,
+      "step": 2679
+    },
+    {
+      "epoch": 0.9648964896489649,
+      "grad_norm": 0.7564746737480164,
+      "learning_rate": 6.237218527876399e-07,
+      "loss": 4.4544,
+      "step": 2680
+    },
+    {
+      "epoch": 0.9652565256525653,
+      "grad_norm": 0.49899938702583313,
+      "learning_rate": 6.109409416834688e-07,
+      "loss": 4.7035,
+      "step": 2681
+    },
+    {
+      "epoch": 0.9656165616561656,
+      "grad_norm": 0.8740639686584473,
+      "learning_rate": 5.982919376270823e-07,
+      "loss": 4.5884,
+      "step": 2682
+    },
+    {
+      "epoch": 0.965976597659766,
+      "grad_norm": 1.1928685903549194,
+      "learning_rate": 5.857748574059851e-07,
+      "loss": 4.7686,
+      "step": 2683
+    },
+    {
+      "epoch": 0.9663366336633663,
+      "grad_norm": 0.8761278986930847,
+      "learning_rate": 5.733897176325665e-07,
+      "loss": 4.9492,
+      "step": 2684
+    },
+    {
+      "epoch": 0.9666966696669667,
+      "grad_norm": 0.6431811451911926,
+      "learning_rate": 5.611365347441334e-07,
+      "loss": 4.8721,
+      "step": 2685
+    },
+    {
+      "epoch": 0.967056705670567,
+      "grad_norm": 0.7401660084724426,
+      "learning_rate": 5.49015325002833e-07,
+      "loss": 4.8349,
+      "step": 2686
+    },
+    {
+      "epoch": 0.9674167416741675,
+      "grad_norm": 0.7588707804679871,
+      "learning_rate": 5.370261044956971e-07,
+      "loss": 4.6324,
+      "step": 2687
+    },
+    {
+      "epoch": 0.9677767776777678,
+      "grad_norm": 0.5506744980812073,
+      "learning_rate": 5.25168889134553e-07,
+      "loss": 4.6068,
+      "step": 2688
+    },
+    {
+      "epoch": 0.9681368136813682,
+      "grad_norm": 0.6155896782875061,
+      "learning_rate": 5.134436946560572e-07,
+      "loss": 4.7839,
+      "step": 2689
+    },
+    {
+      "epoch": 0.9684968496849685,
+      "grad_norm": 0.7731359601020813,
+      "learning_rate": 5.018505366216175e-07,
+      "loss": 4.8457,
+      "step": 2690
+    },
+    {
+      "epoch": 0.9688568856885689,
+      "grad_norm": 0.6838685870170593,
+      "learning_rate": 4.903894304174372e-07,
+      "loss": 4.9202,
+      "step": 2691
+    },
+    {
+      "epoch": 0.9692169216921692,
+      "grad_norm": 0.7036203742027283,
+      "learning_rate": 4.790603912544489e-07,
+      "loss": 4.609,
+      "step": 2692
+    },
+    {
+      "epoch": 0.9695769576957696,
+      "grad_norm": 0.6218310594558716,
+      "learning_rate": 4.678634341683252e-07,
+      "loss": 4.7149,
+      "step": 2693
+    },
+    {
+      "epoch": 0.9699369936993699,
+      "grad_norm": 0.6616173982620239,
+      "learning_rate": 4.567985740194236e-07,
+      "loss": 4.5665,
+      "step": 2694
+    },
+    {
+      "epoch": 0.9702970297029703,
+      "grad_norm": 0.677174985408783,
+      "learning_rate": 4.458658254927972e-07,
+      "loss": 5.1044,
+      "step": 2695
+    },
+    {
+      "epoch": 0.9706570657065706,
+      "grad_norm": 0.7272735238075256,
+      "learning_rate": 4.3506520309813947e-07,
+      "loss": 5.0046,
+      "step": 2696
+    },
+    {
+      "epoch": 0.971017101710171,
+      "grad_norm": 0.9171267151832581,
+      "learning_rate": 4.2439672116982855e-07,
+      "loss": 5.1378,
+      "step": 2697
+    },
+    {
+      "epoch": 0.9713771377137714,
+      "grad_norm": 0.9265881180763245,
+      "learning_rate": 4.138603938668273e-07,
+      "loss": 5.1032,
+      "step": 2698
+    },
+    {
+      "epoch": 0.9717371737173718,
+      "grad_norm": 0.8648788332939148,
+      "learning_rate": 4.034562351727389e-07,
+      "loss": 5.3116,
+      "step": 2699
+    },
+    {
+      "epoch": 0.9720972097209721,
+      "grad_norm": 1.2427067756652832,
+      "learning_rate": 3.9318425889574017e-07,
+      "loss": 5.5403,
+      "step": 2700
+    },
+    {
+      "epoch": 0.9724572457245725,
+      "grad_norm": 1.1116799116134644,
+      "learning_rate": 3.8304447866857053e-07,
+      "loss": 5.1365,
+      "step": 2701
+    },
+    {
+      "epoch": 0.9728172817281728,
+      "grad_norm": 0.5051364898681641,
+      "learning_rate": 3.73036907948543e-07,
+      "loss": 4.3293,
+      "step": 2702
+    },
+    {
+      "epoch": 0.9731773177317732,
+      "grad_norm": 1.0161361694335938,
+      "learning_rate": 3.631615600174887e-07,
+      "loss": 4.4407,
+      "step": 2703
+    },
+    {
+      "epoch": 0.9735373537353735,
+      "grad_norm": 0.8723888993263245,
+      "learning_rate": 3.5341844798174594e-07,
+      "loss": 5.0663,
+      "step": 2704
+    },
+    {
+      "epoch": 0.9738973897389739,
+      "grad_norm": 0.6905868649482727,
+      "learning_rate": 3.4380758477219333e-07,
+      "loss": 4.9586,
+      "step": 2705
+    },
+    {
+      "epoch": 0.9742574257425742,
+      "grad_norm": 0.6786131262779236,
+      "learning_rate": 3.343289831441387e-07,
+      "loss": 4.4407,
+      "step": 2706
+    },
+    {
+      "epoch": 0.9746174617461746,
+      "grad_norm": 0.7154151201248169,
+      "learning_rate": 3.2498265567739717e-07,
+      "loss": 4.6941,
+      "step": 2707
+    },
+    {
+      "epoch": 0.9749774977497749,
+      "grad_norm": 0.7994034290313721,
+      "learning_rate": 3.1576861477621287e-07,
+      "loss": 4.857,
+      "step": 2708
+    },
+    {
+      "epoch": 0.9753375337533753,
+      "grad_norm": 0.6819135546684265,
+      "learning_rate": 3.0668687266925956e-07,
+      "loss": 5.0646,
+      "step": 2709
+    },
+    {
+      "epoch": 0.9756975697569757,
+      "grad_norm": 0.613059401512146,
+      "learning_rate": 2.977374414096401e-07,
+      "loss": 4.9047,
+      "step": 2710
+    },
+    {
+      "epoch": 0.9760576057605761,
+      "grad_norm": 0.9619209170341492,
+      "learning_rate": 2.889203328748424e-07,
+      "loss": 4.5585,
+      "step": 2711
+    },
+    {
+      "epoch": 0.9764176417641764,
+      "grad_norm": 0.5810762047767639,
+      "learning_rate": 2.8023555876673937e-07,
+      "loss": 4.8418,
+      "step": 2712
+    },
+    {
+      "epoch": 0.9767776777677768,
+      "grad_norm": 0.6123738884925842,
+      "learning_rate": 2.7168313061159964e-07,
+      "loss": 4.5524,
+      "step": 2713
+    },
+    {
+      "epoch": 0.9771377137713771,
+      "grad_norm": 0.6715987324714661,
+      "learning_rate": 2.6326305976001055e-07,
+      "loss": 4.9585,
+      "step": 2714
+    },
+    {
+      "epoch": 0.9774977497749775,
+      "grad_norm": 0.6126198768615723,
+      "learning_rate": 2.549753573869107e-07,
+      "loss": 4.6261,
+      "step": 2715
+    },
+    {
+      "epoch": 0.9778577857785778,
+      "grad_norm": 0.6247376203536987,
+      "learning_rate": 2.468200344915572e-07,
+      "loss": 4.7465,
+      "step": 2716
+    },
+    {
+      "epoch": 0.9782178217821782,
+      "grad_norm": 0.6758084297180176,
+      "learning_rate": 2.3879710189753656e-07,
+      "loss": 4.6146,
+      "step": 2717
+    },
+    {
+      "epoch": 0.9785778577857785,
+      "grad_norm": 0.8351864218711853,
+      "learning_rate": 2.3090657025270912e-07,
+      "loss": 4.9151,
+      "step": 2718
+    },
+    {
+      "epoch": 0.9789378937893789,
+      "grad_norm": 1.0861241817474365,
+      "learning_rate": 2.2314845002922025e-07,
+      "loss": 4.75,
+      "step": 2719
+    },
+    {
+      "epoch": 0.9792979297929792,
+      "grad_norm": 0.8706967830657959,
+      "learning_rate": 2.15522751523467e-07,
+      "loss": 5.033,
+      "step": 2720
+    },
+    {
+      "epoch": 0.9796579657965797,
+      "grad_norm": 0.8538120985031128,
+      "learning_rate": 2.080294848561426e-07,
+      "loss": 4.5049,
+      "step": 2721
+    },
+    {
+      "epoch": 0.9800180018001801,
+      "grad_norm": 0.567323625087738,
+      "learning_rate": 2.0066865997212525e-07,
+      "loss": 4.7168,
+      "step": 2722
+    },
+    {
+      "epoch": 0.9803780378037804,
+      "grad_norm": 0.8506109118461609,
+      "learning_rate": 1.9344028664056713e-07,
+      "loss": 5.0249,
+      "step": 2723
+    },
+    {
+      "epoch": 0.9807380738073808,
+      "grad_norm": 1.3589588403701782,
+      "learning_rate": 1.8634437445479435e-07,
+      "loss": 5.1037,
+      "step": 2724
+    },
+    {
+      "epoch": 0.9810981098109811,
+      "grad_norm": 1.3143948316574097,
+      "learning_rate": 1.7938093283236258e-07,
+      "loss": 5.555,
+      "step": 2725
+    },
+    {
+      "epoch": 0.9814581458145815,
+      "grad_norm": 1.7907097339630127,
+      "learning_rate": 1.7254997101500137e-07,
+      "loss": 4.8558,
+      "step": 2726
+    },
+    {
+      "epoch": 0.9818181818181818,
+      "grad_norm": 0.6720486283302307,
+      "learning_rate": 1.6585149806860324e-07,
+      "loss": 4.4563,
+      "step": 2727
+    },
+    {
+      "epoch": 0.9821782178217822,
+      "grad_norm": 0.6979610919952393,
+      "learning_rate": 1.5928552288326793e-07,
+      "loss": 4.5348,
+      "step": 2728
+    },
+    {
+      "epoch": 0.9825382538253825,
+      "grad_norm": 0.8894920349121094,
+      "learning_rate": 1.5285205417319149e-07,
+      "loss": 4.7034,
+      "step": 2729
+    },
+    {
+      "epoch": 0.9828982898289829,
+      "grad_norm": 0.6074763536453247,
+      "learning_rate": 1.4655110047675503e-07,
+      "loss": 4.5014,
+      "step": 2730
+    },
+    {
+      "epoch": 0.9832583258325832,
+      "grad_norm": 0.7769091129302979,
+      "learning_rate": 1.403826701564359e-07,
+      "loss": 4.7347,
+      "step": 2731
+    },
+    {
+      "epoch": 0.9836183618361836,
+      "grad_norm": 0.7279649376869202,
+      "learning_rate": 1.3434677139885222e-07,
+      "loss": 4.5844,
+      "step": 2732
+    },
+    {
+      "epoch": 0.983978397839784,
+      "grad_norm": 0.8169944882392883,
+      "learning_rate": 1.2844341221471824e-07,
+      "loss": 4.939,
+      "step": 2733
+    },
+    {
+      "epoch": 0.9843384338433844,
+      "grad_norm": 0.7737520337104797,
+      "learning_rate": 1.2267260043885564e-07,
+      "loss": 4.3103,
+      "step": 2734
+    },
+    {
+      "epoch": 0.9846984698469847,
+      "grad_norm": 0.7852435111999512,
+      "learning_rate": 1.170343437301491e-07,
+      "loss": 4.6278,
+      "step": 2735
+    },
+    {
+      "epoch": 0.9850585058505851,
+      "grad_norm": 0.5969253182411194,
+      "learning_rate": 1.1152864957157949e-07,
+      "loss": 4.5914,
+      "step": 2736
+    },
+    {
+      "epoch": 0.9854185418541854,
+      "grad_norm": 0.8359677791595459,
+      "learning_rate": 1.0615552527017958e-07,
+      "loss": 4.5289,
+      "step": 2737
+    },
+    {
+      "epoch": 0.9857785778577858,
+      "grad_norm": 0.8683612942695618,
+      "learning_rate": 1.0091497795706728e-07,
+      "loss": 4.5384,
+      "step": 2738
+    },
+    {
+      "epoch": 0.9861386138613861,
+      "grad_norm": 0.6547925472259521,
+      "learning_rate": 9.580701458736796e-08,
+      "loss": 4.7082,
+      "step": 2739
+    },
+    {
+      "epoch": 0.9864986498649865,
+      "grad_norm": 0.5458620190620422,
+      "learning_rate": 9.083164194025883e-08,
+      "loss": 4.6053,
+      "step": 2740
+    },
+    {
+      "epoch": 0.9868586858685868,
+      "grad_norm": 0.6568934917449951,
+      "learning_rate": 8.598886661895788e-08,
+      "loss": 4.5799,
+      "step": 2741
+    },
+    {
+      "epoch": 0.9872187218721872,
+      "grad_norm": 0.661688506603241,
+      "learning_rate": 8.127869505069053e-08,
+      "loss": 4.3995,
+      "step": 2742
+    },
+    {
+      "epoch": 0.9875787578757875,
+      "grad_norm": 0.7148826122283936,
+      "learning_rate": 7.670113348670071e-08,
+      "loss": 4.607,
+      "step": 2743
+    },
+    {
+      "epoch": 0.987938793879388,
+      "grad_norm": 0.5700961947441101,
+      "learning_rate": 7.225618800222877e-08,
+      "loss": 4.7385,
+      "step": 2744
+    },
+    {
+      "epoch": 0.9882988298829883,
+      "grad_norm": 0.6849833130836487,
+      "learning_rate": 6.794386449651135e-08,
+      "loss": 4.9545,
+      "step": 2745
+    },
+    {
+      "epoch": 0.9886588658865887,
+      "grad_norm": 0.6745620965957642,
+      "learning_rate": 6.376416869277036e-08,
+      "loss": 4.8148,
+      "step": 2746
+    },
+    {
+      "epoch": 0.989018901890189,
+      "grad_norm": 0.9582657814025879,
+      "learning_rate": 5.971710613821291e-08,
+      "loss": 5.3346,
+      "step": 2747
+    },
+    {
+      "epoch": 0.9893789378937894,
+      "grad_norm": 0.8358666300773621,
+      "learning_rate": 5.5802682204009194e-08,
+      "loss": 5.1755,
+      "step": 2748
+    },
+    {
+      "epoch": 0.9897389738973897,
+      "grad_norm": 0.9131016135215759,
+      "learning_rate": 5.2020902085303525e-08,
+      "loss": 5.2653,
+      "step": 2749
+    },
+    {
+      "epoch": 0.9900990099009901,
+      "grad_norm": 1.4401273727416992,
+      "learning_rate": 4.837177080119215e-08,
+      "loss": 5.5157,
+      "step": 2750
+    },
+    {
+      "epoch": 0.9904590459045904,
+      "grad_norm": 0.676723301410675,
+      "learning_rate": 4.485529319473436e-08,
+      "loss": 4.8633,
+      "step": 2751
+    },
+    {
+      "epoch": 0.9908190819081908,
+      "grad_norm": 0.7600442171096802,
+      "learning_rate": 4.147147393290807e-08,
+      "loss": 4.7974,
+      "step": 2752
+    },
+    {
+      "epoch": 0.9911791179117911,
+      "grad_norm": 1.2827167510986328,
+      "learning_rate": 3.8220317506654226e-08,
+      "loss": 4.8168,
+      "step": 2753
+    },
+    {
+      "epoch": 0.9915391539153915,
+      "grad_norm": 0.9814483523368835,
+      "learning_rate": 3.510182823083241e-08,
+      "loss": 4.736,
+      "step": 2754
+    },
+    {
+      "epoch": 0.991899189918992,
+      "grad_norm": 0.9813688397407532,
+      "learning_rate": 3.2116010244254144e-08,
+      "loss": 4.7668,
+      "step": 2755
+    },
+    {
+      "epoch": 0.9922592259225923,
+      "grad_norm": 0.7599554657936096,
+      "learning_rate": 2.9262867509605163e-08,
+      "loss": 4.7808,
+      "step": 2756
+    },
+    {
+      "epoch": 0.9926192619261927,
+      "grad_norm": 0.5308656692504883,
+      "learning_rate": 2.6542403813545334e-08,
+      "loss": 4.571,
+      "step": 2757
+    },
+    {
+      "epoch": 0.992979297929793,
+      "grad_norm": 0.5827195048332214,
+      "learning_rate": 2.3954622766597657e-08,
+      "loss": 4.7896,
+      "step": 2758
+    },
+    {
+      "epoch": 0.9933393339333934,
+      "grad_norm": 0.9100288152694702,
+      "learning_rate": 2.1499527803214846e-08,
+      "loss": 4.8101,
+      "step": 2759
+    },
+    {
+      "epoch": 0.9936993699369937,
+      "grad_norm": 1.0028470754623413,
+      "learning_rate": 1.9177122181757156e-08,
+      "loss": 4.6816,
+      "step": 2760
+    },
+    {
+      "epoch": 0.994059405940594,
+      "grad_norm": 0.7796440720558167,
+      "learning_rate": 1.698740898444795e-08,
+      "loss": 4.6857,
+      "step": 2761
+    },
+    {
+      "epoch": 0.9944194419441944,
+      "grad_norm": 0.6619350910186768,
+      "learning_rate": 1.4930391117451426e-08,
+      "loss": 4.8632,
+      "step": 2762
+    },
+    {
+      "epoch": 0.9947794779477948,
+      "grad_norm": 0.6579605937004089,
+      "learning_rate": 1.3006071310783797e-08,
+      "loss": 4.8485,
+      "step": 2763
+    },
+    {
+      "epoch": 0.9951395139513951,
+      "grad_norm": 0.7904688119888306,
+      "learning_rate": 1.1214452118368802e-08,
+      "loss": 4.6414,
+      "step": 2764
+    },
+    {
+      "epoch": 0.9954995499549955,
+      "grad_norm": 0.6213950514793396,
+      "learning_rate": 9.555535917993297e-09,
+      "loss": 4.56,
+      "step": 2765
+    },
+    {
+      "epoch": 0.9958595859585958,
+      "grad_norm": 0.5204569697380066,
+      "learning_rate": 8.029324911351666e-09,
+      "loss": 4.42,
+      "step": 2766
+    },
+    {
+      "epoch": 0.9962196219621963,
+      "grad_norm": 0.5820871591567993,
+      "learning_rate": 6.635821124001406e-09,
+      "loss": 4.7025,
+      "step": 2767
+    },
+    {
+      "epoch": 0.9965796579657966,
+      "grad_norm": 0.5928208231925964,
+      "learning_rate": 5.375026405352035e-09,
+      "loss": 5.1903,
+      "step": 2768
+    },
+    {
+      "epoch": 0.996939693969397,
+      "grad_norm": 0.5016415119171143,
+      "learning_rate": 4.246942428709488e-09,
+      "loss": 4.7524,
+      "step": 2769
+    },
+    {
+      "epoch": 0.9972997299729973,
+      "grad_norm": 0.650364339351654,
+      "learning_rate": 3.2515706912539245e-09,
+      "loss": 4.3618,
+      "step": 2770
+    },
+    {
+      "epoch": 0.9976597659765977,
+      "grad_norm": 1.0009171962738037,
+      "learning_rate": 2.388912514017516e-09,
+      "loss": 4.9389,
+      "step": 2771
+    },
+    {
+      "epoch": 0.998019801980198,
+      "grad_norm": 1.0807517766952515,
+      "learning_rate": 1.6589690418955528e-09,
+      "loss": 5.1394,
+      "step": 2772
+    },
+    {
+      "epoch": 0.9983798379837984,
+      "grad_norm": 0.9495770335197449,
+      "learning_rate": 1.0617412436464413e-09,
+      "loss": 5.1927,
+      "step": 2773
+    },
+    {
+      "epoch": 0.9987398739873987,
+      "grad_norm": 0.8993694186210632,
+      "learning_rate": 5.972299119250125e-10,
+      "loss": 5.2012,
+      "step": 2774
+    },
+    {
+      "epoch": 0.9990999099909991,
+      "grad_norm": 1.2877267599105835,
+      "learning_rate": 2.6543566319370275e-10,
+      "loss": 5.2814,
+      "step": 2775
+    },
+    {
+      "epoch": 0.9994599459945994,
+      "grad_norm": 0.5570570826530457,
+      "learning_rate": 6.63589378113727e-11,
+      "loss": 4.7664,
+      "step": 2776
+    },
+    {
+      "epoch": 0.9998199819981998,
+      "grad_norm": 0.6813939213752747,
+      "learning_rate": 0.0,
+      "loss": 4.551,
+      "step": 2777
     }
   ],
   "logging_steps": 1,
@@ -14648,12 +19492,12 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }
   },
-  "total_flos": 2264269974405120.0,
+  "total_flos": 3015768690130944.0,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null