AutoL2S-Plus-7b / trainer_state.json
amandaa's picture
Upload folder using huggingface_hub
d0b5a8b verified
Raw
History Blame Contribute Delete
202 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 50000,
"global_step": 2500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0008,
"grad_norm": 1.5249121521928364,
"learning_rate": 8e-10,
"loss": 0.3578,
"step": 2
},
{
"epoch": 0.0016,
"grad_norm": 3.04013114408118,
"learning_rate": 1.6e-09,
"loss": -0.3638,
"step": 4
},
{
"epoch": 0.0024,
"grad_norm": 14.399566409644445,
"learning_rate": 2.4e-09,
"loss": -0.5453,
"step": 6
},
{
"epoch": 0.0032,
"grad_norm": 1.058111653988024,
"learning_rate": 3.2e-09,
"loss": 0.4043,
"step": 8
},
{
"epoch": 0.004,
"grad_norm": 5.8550272702903285,
"learning_rate": 4e-09,
"loss": 0.2492,
"step": 10
},
{
"epoch": 0.0048,
"grad_norm": 0.8103035684471422,
"learning_rate": 4.8e-09,
"loss": 0.1155,
"step": 12
},
{
"epoch": 0.0056,
"grad_norm": 1.439010851325566,
"learning_rate": 5.6e-09,
"loss": 0.4965,
"step": 14
},
{
"epoch": 0.0064,
"grad_norm": 3.279695602874879,
"learning_rate": 6.4e-09,
"loss": 0.1119,
"step": 16
},
{
"epoch": 0.0072,
"grad_norm": 3.9544253242598595,
"learning_rate": 7.199999999999999e-09,
"loss": 0.2711,
"step": 18
},
{
"epoch": 0.008,
"grad_norm": 2.303298337789742,
"learning_rate": 8e-09,
"loss": -0.2384,
"step": 20
},
{
"epoch": 0.0088,
"grad_norm": 0.6956054922056536,
"learning_rate": 8.8e-09,
"loss": -0.0266,
"step": 22
},
{
"epoch": 0.0096,
"grad_norm": 1.6058336187090627,
"learning_rate": 9.6e-09,
"loss": 0.2667,
"step": 24
},
{
"epoch": 0.0104,
"grad_norm": 0.5873869381231204,
"learning_rate": 1.0399999999999999e-08,
"loss": 0.279,
"step": 26
},
{
"epoch": 0.0112,
"grad_norm": 1.3919968190642873,
"learning_rate": 1.12e-08,
"loss": 0.2859,
"step": 28
},
{
"epoch": 0.012,
"grad_norm": 0.9786688052322894,
"learning_rate": 1.1999999999999998e-08,
"loss": 0.2605,
"step": 30
},
{
"epoch": 0.0128,
"grad_norm": 0.997140725588297,
"learning_rate": 1.28e-08,
"loss": -0.0304,
"step": 32
},
{
"epoch": 0.0136,
"grad_norm": 0.712545871939318,
"learning_rate": 1.3600000000000001e-08,
"loss": 0.1921,
"step": 34
},
{
"epoch": 0.0144,
"grad_norm": 2.3171450622225698,
"learning_rate": 1.4399999999999998e-08,
"loss": 0.0005,
"step": 36
},
{
"epoch": 0.0152,
"grad_norm": 1.383748393656482,
"learning_rate": 1.52e-08,
"loss": 0.6168,
"step": 38
},
{
"epoch": 0.016,
"grad_norm": 2.2831084349959063,
"learning_rate": 1.6e-08,
"loss": -0.2458,
"step": 40
},
{
"epoch": 0.0168,
"grad_norm": 3.7540820685705554,
"learning_rate": 1.68e-08,
"loss": -0.5973,
"step": 42
},
{
"epoch": 0.0176,
"grad_norm": 1.581263276047286,
"learning_rate": 1.76e-08,
"loss": 0.2239,
"step": 44
},
{
"epoch": 0.0184,
"grad_norm": 2.233121763402274,
"learning_rate": 1.84e-08,
"loss": 0.0193,
"step": 46
},
{
"epoch": 0.0192,
"grad_norm": 0.8606914336336169,
"learning_rate": 1.92e-08,
"loss": -0.253,
"step": 48
},
{
"epoch": 0.02,
"grad_norm": 5.702797171498987,
"learning_rate": 2e-08,
"loss": 0.2654,
"step": 50
},
{
"epoch": 0.0208,
"grad_norm": 0.4089715021280977,
"learning_rate": 2.0799999999999998e-08,
"loss": 0.292,
"step": 52
},
{
"epoch": 0.0216,
"grad_norm": 0.8023217053247889,
"learning_rate": 2.1599999999999998e-08,
"loss": 0.46,
"step": 54
},
{
"epoch": 0.0224,
"grad_norm": 5.4143014554749245,
"learning_rate": 2.24e-08,
"loss": -0.1526,
"step": 56
},
{
"epoch": 0.0232,
"grad_norm": 0.6957451761861397,
"learning_rate": 2.32e-08,
"loss": 0.2491,
"step": 58
},
{
"epoch": 0.024,
"grad_norm": 3.6795198695452123,
"learning_rate": 2.3999999999999997e-08,
"loss": -0.0906,
"step": 60
},
{
"epoch": 0.0248,
"grad_norm": 1.7378271537499541,
"learning_rate": 2.4799999999999997e-08,
"loss": 0.7429,
"step": 62
},
{
"epoch": 0.0256,
"grad_norm": 3.6225251386805883,
"learning_rate": 2.56e-08,
"loss": 0.3386,
"step": 64
},
{
"epoch": 0.0264,
"grad_norm": 1.7773145305018176,
"learning_rate": 2.64e-08,
"loss": -0.2629,
"step": 66
},
{
"epoch": 0.0272,
"grad_norm": 0.7495628843105195,
"learning_rate": 2.7200000000000002e-08,
"loss": 0.1966,
"step": 68
},
{
"epoch": 0.028,
"grad_norm": 2.885359440654563,
"learning_rate": 2.8000000000000003e-08,
"loss": 0.0439,
"step": 70
},
{
"epoch": 0.0288,
"grad_norm": 1.015945747749439,
"learning_rate": 2.8799999999999996e-08,
"loss": -0.2864,
"step": 72
},
{
"epoch": 0.0296,
"grad_norm": 0.5182079375617211,
"learning_rate": 2.9599999999999997e-08,
"loss": -0.166,
"step": 74
},
{
"epoch": 0.0304,
"grad_norm": 0.9817246812987965,
"learning_rate": 3.04e-08,
"loss": 0.3014,
"step": 76
},
{
"epoch": 0.0312,
"grad_norm": 0.4557071988963493,
"learning_rate": 3.12e-08,
"loss": -0.1152,
"step": 78
},
{
"epoch": 0.032,
"grad_norm": 4.763328548185686,
"learning_rate": 3.2e-08,
"loss": -0.1538,
"step": 80
},
{
"epoch": 0.0328,
"grad_norm": 0.9349584077447443,
"learning_rate": 3.28e-08,
"loss": -0.0778,
"step": 82
},
{
"epoch": 0.0336,
"grad_norm": 4.258851934530662,
"learning_rate": 3.36e-08,
"loss": -0.1298,
"step": 84
},
{
"epoch": 0.0344,
"grad_norm": 1.3039223561516173,
"learning_rate": 3.44e-08,
"loss": 0.5712,
"step": 86
},
{
"epoch": 0.0352,
"grad_norm": 0.8670442729703873,
"learning_rate": 3.52e-08,
"loss": 0.0303,
"step": 88
},
{
"epoch": 0.036,
"grad_norm": 1.0441412548980031,
"learning_rate": 3.6e-08,
"loss": -0.1422,
"step": 90
},
{
"epoch": 0.0368,
"grad_norm": 3.5521797457066855,
"learning_rate": 3.68e-08,
"loss": -0.453,
"step": 92
},
{
"epoch": 0.0376,
"grad_norm": 0.7571538831950513,
"learning_rate": 3.76e-08,
"loss": 0.4032,
"step": 94
},
{
"epoch": 0.0384,
"grad_norm": 0.9420773865217862,
"learning_rate": 3.84e-08,
"loss": 0.4746,
"step": 96
},
{
"epoch": 0.0392,
"grad_norm": 0.7161321209562558,
"learning_rate": 3.92e-08,
"loss": 0.0431,
"step": 98
},
{
"epoch": 0.04,
"grad_norm": 0.7244176997474762,
"learning_rate": 4e-08,
"loss": 0.0103,
"step": 100
},
{
"epoch": 0.0408,
"grad_norm": 1.2508780434898261,
"learning_rate": 4.0799999999999995e-08,
"loss": -0.2087,
"step": 102
},
{
"epoch": 0.0416,
"grad_norm": 2.8150494985583787,
"learning_rate": 4.1599999999999995e-08,
"loss": -0.4386,
"step": 104
},
{
"epoch": 0.0424,
"grad_norm": 5.244508372577221,
"learning_rate": 4.2399999999999996e-08,
"loss": -0.0776,
"step": 106
},
{
"epoch": 0.0432,
"grad_norm": 1.0118362643085135,
"learning_rate": 4.3199999999999996e-08,
"loss": 0.2148,
"step": 108
},
{
"epoch": 0.044,
"grad_norm": 2.453561169568128,
"learning_rate": 4.4e-08,
"loss": -0.28,
"step": 110
},
{
"epoch": 0.0448,
"grad_norm": 5.3747338716221895,
"learning_rate": 4.48e-08,
"loss": 0.4104,
"step": 112
},
{
"epoch": 0.0456,
"grad_norm": 0.3632823515674558,
"learning_rate": 4.56e-08,
"loss": 0.1336,
"step": 114
},
{
"epoch": 0.0464,
"grad_norm": 2.2581925885780403,
"learning_rate": 4.64e-08,
"loss": 0.3285,
"step": 116
},
{
"epoch": 0.0472,
"grad_norm": 1.4956909565408185,
"learning_rate": 4.719999999999999e-08,
"loss": 0.0122,
"step": 118
},
{
"epoch": 0.048,
"grad_norm": 1.7769183707894964,
"learning_rate": 4.799999999999999e-08,
"loss": 0.5743,
"step": 120
},
{
"epoch": 0.0488,
"grad_norm": 1.2583349136163662,
"learning_rate": 4.8799999999999994e-08,
"loss": -0.0307,
"step": 122
},
{
"epoch": 0.0496,
"grad_norm": 0.7008190210050554,
"learning_rate": 4.9599999999999994e-08,
"loss": 0.0008,
"step": 124
},
{
"epoch": 0.0504,
"grad_norm": 8.020060442286011,
"learning_rate": 5.04e-08,
"loss": -0.1381,
"step": 126
},
{
"epoch": 0.0512,
"grad_norm": 3.2600734226930705,
"learning_rate": 5.12e-08,
"loss": -0.152,
"step": 128
},
{
"epoch": 0.052,
"grad_norm": 0.8910507406833589,
"learning_rate": 5.2e-08,
"loss": -0.4405,
"step": 130
},
{
"epoch": 0.0528,
"grad_norm": 0.7722355780927433,
"learning_rate": 5.28e-08,
"loss": 0.2391,
"step": 132
},
{
"epoch": 0.0536,
"grad_norm": 1.6289451664213528,
"learning_rate": 5.3600000000000004e-08,
"loss": -0.1101,
"step": 134
},
{
"epoch": 0.0544,
"grad_norm": 1.0056870022444548,
"learning_rate": 5.4400000000000004e-08,
"loss": 0.0633,
"step": 136
},
{
"epoch": 0.0552,
"grad_norm": 2.3308622414962508,
"learning_rate": 5.5200000000000005e-08,
"loss": 0.4368,
"step": 138
},
{
"epoch": 0.056,
"grad_norm": 0.7924323699418531,
"learning_rate": 5.6000000000000005e-08,
"loss": 0.5237,
"step": 140
},
{
"epoch": 0.0568,
"grad_norm": 3.8386273487570692,
"learning_rate": 5.679999999999999e-08,
"loss": -0.1094,
"step": 142
},
{
"epoch": 0.0576,
"grad_norm": 1.2380510876659898,
"learning_rate": 5.759999999999999e-08,
"loss": 0.1722,
"step": 144
},
{
"epoch": 0.0584,
"grad_norm": 1.9382814912362036,
"learning_rate": 5.8399999999999994e-08,
"loss": -0.1648,
"step": 146
},
{
"epoch": 0.0592,
"grad_norm": 1.3282095077859473,
"learning_rate": 5.9199999999999994e-08,
"loss": 0.0622,
"step": 148
},
{
"epoch": 0.06,
"grad_norm": 1.555494681939169,
"learning_rate": 6e-08,
"loss": -0.3911,
"step": 150
},
{
"epoch": 0.0608,
"grad_norm": 2.669943445564743,
"learning_rate": 6.08e-08,
"loss": -0.4855,
"step": 152
},
{
"epoch": 0.0616,
"grad_norm": 2.8935870289471177,
"learning_rate": 6.16e-08,
"loss": -0.5626,
"step": 154
},
{
"epoch": 0.0624,
"grad_norm": 4.067814139067205,
"learning_rate": 6.24e-08,
"loss": -0.6101,
"step": 156
},
{
"epoch": 0.0632,
"grad_norm": 7.45875665988826,
"learning_rate": 6.32e-08,
"loss": -0.8887,
"step": 158
},
{
"epoch": 0.064,
"grad_norm": 1.113875500199247,
"learning_rate": 6.4e-08,
"loss": -0.0743,
"step": 160
},
{
"epoch": 0.0648,
"grad_norm": 1.0811766104737213,
"learning_rate": 6.48e-08,
"loss": 0.4856,
"step": 162
},
{
"epoch": 0.0656,
"grad_norm": 3.6239338057499606,
"learning_rate": 6.56e-08,
"loss": -0.2057,
"step": 164
},
{
"epoch": 0.0664,
"grad_norm": 1.9376472037589891,
"learning_rate": 6.64e-08,
"loss": 0.0629,
"step": 166
},
{
"epoch": 0.0672,
"grad_norm": 0.8329527775863854,
"learning_rate": 6.72e-08,
"loss": 0.1928,
"step": 168
},
{
"epoch": 0.068,
"grad_norm": 3.7893101636338664,
"learning_rate": 6.8e-08,
"loss": 0.4232,
"step": 170
},
{
"epoch": 0.0688,
"grad_norm": 1.0070374331551073,
"learning_rate": 6.88e-08,
"loss": 0.1722,
"step": 172
},
{
"epoch": 0.0696,
"grad_norm": 0.9497901793517014,
"learning_rate": 6.959999999999999e-08,
"loss": -0.2653,
"step": 174
},
{
"epoch": 0.0704,
"grad_norm": 1.650849094611574,
"learning_rate": 7.04e-08,
"loss": -0.7653,
"step": 176
},
{
"epoch": 0.0712,
"grad_norm": 1.3515156957900079,
"learning_rate": 7.119999999999999e-08,
"loss": -0.8214,
"step": 178
},
{
"epoch": 0.072,
"grad_norm": 3.1324182478150067,
"learning_rate": 7.2e-08,
"loss": -0.7823,
"step": 180
},
{
"epoch": 0.0728,
"grad_norm": 1.4174012963850768,
"learning_rate": 7.279999999999999e-08,
"loss": 0.2644,
"step": 182
},
{
"epoch": 0.0736,
"grad_norm": 1.210359954223697,
"learning_rate": 7.36e-08,
"loss": 0.2054,
"step": 184
},
{
"epoch": 0.0744,
"grad_norm": 0.8136072967624918,
"learning_rate": 7.439999999999999e-08,
"loss": 0.2106,
"step": 186
},
{
"epoch": 0.0752,
"grad_norm": 2.585403352286107,
"learning_rate": 7.52e-08,
"loss": -0.4224,
"step": 188
},
{
"epoch": 0.076,
"grad_norm": 10.231551056457478,
"learning_rate": 7.599999999999999e-08,
"loss": 0.2142,
"step": 190
},
{
"epoch": 0.0768,
"grad_norm": 3.2173359330709483,
"learning_rate": 7.68e-08,
"loss": 0.6189,
"step": 192
},
{
"epoch": 0.0776,
"grad_norm": 2.124353538166263,
"learning_rate": 7.76e-08,
"loss": 0.2982,
"step": 194
},
{
"epoch": 0.0784,
"grad_norm": 0.552359523142279,
"learning_rate": 7.84e-08,
"loss": 0.143,
"step": 196
},
{
"epoch": 0.0792,
"grad_norm": 0.8478239264393724,
"learning_rate": 7.92e-08,
"loss": 0.2097,
"step": 198
},
{
"epoch": 0.08,
"grad_norm": 5.107721369380915,
"learning_rate": 8e-08,
"loss": 0.5133,
"step": 200
},
{
"epoch": 0.0808,
"grad_norm": 0.7042643328899001,
"learning_rate": 8.08e-08,
"loss": 0.299,
"step": 202
},
{
"epoch": 0.0816,
"grad_norm": 1.8155271031174458,
"learning_rate": 8.159999999999999e-08,
"loss": -0.2053,
"step": 204
},
{
"epoch": 0.0824,
"grad_norm": 9.910433671130228,
"learning_rate": 8.24e-08,
"loss": -0.6662,
"step": 206
},
{
"epoch": 0.0832,
"grad_norm": 2.785320281862689,
"learning_rate": 8.319999999999999e-08,
"loss": -0.1989,
"step": 208
},
{
"epoch": 0.084,
"grad_norm": 8.98969821762216,
"learning_rate": 8.4e-08,
"loss": 0.0962,
"step": 210
},
{
"epoch": 0.0848,
"grad_norm": 3.886512189538636,
"learning_rate": 8.479999999999999e-08,
"loss": 0.6847,
"step": 212
},
{
"epoch": 0.0856,
"grad_norm": 1.354000381017739,
"learning_rate": 8.56e-08,
"loss": 0.0522,
"step": 214
},
{
"epoch": 0.0864,
"grad_norm": 5.066941305120673,
"learning_rate": 8.639999999999999e-08,
"loss": -0.1425,
"step": 216
},
{
"epoch": 0.0872,
"grad_norm": 3.7051328665717844,
"learning_rate": 8.72e-08,
"loss": -0.7429,
"step": 218
},
{
"epoch": 0.088,
"grad_norm": 0.45345084267213714,
"learning_rate": 8.8e-08,
"loss": -0.0116,
"step": 220
},
{
"epoch": 0.0888,
"grad_norm": 3.441731395892135,
"learning_rate": 8.88e-08,
"loss": -0.448,
"step": 222
},
{
"epoch": 0.0896,
"grad_norm": 1.3984343399803179,
"learning_rate": 8.96e-08,
"loss": -0.4024,
"step": 224
},
{
"epoch": 0.0904,
"grad_norm": 3.182154434652412,
"learning_rate": 9.04e-08,
"loss": -0.3474,
"step": 226
},
{
"epoch": 0.0912,
"grad_norm": 6.415250120623118,
"learning_rate": 9.12e-08,
"loss": 0.4669,
"step": 228
},
{
"epoch": 0.092,
"grad_norm": 1.7120711032079754,
"learning_rate": 9.2e-08,
"loss": 0.0868,
"step": 230
},
{
"epoch": 0.0928,
"grad_norm": 0.5203104044416641,
"learning_rate": 9.28e-08,
"loss": 0.0763,
"step": 232
},
{
"epoch": 0.0936,
"grad_norm": 0.9813223999574935,
"learning_rate": 9.36e-08,
"loss": 0.1502,
"step": 234
},
{
"epoch": 0.0944,
"grad_norm": 1.7901780101296731,
"learning_rate": 9.439999999999999e-08,
"loss": 0.0934,
"step": 236
},
{
"epoch": 0.0952,
"grad_norm": 0.8817934891230337,
"learning_rate": 9.519999999999999e-08,
"loss": 0.2892,
"step": 238
},
{
"epoch": 0.096,
"grad_norm": 2.9378020793772595,
"learning_rate": 9.599999999999999e-08,
"loss": -0.0585,
"step": 240
},
{
"epoch": 0.0968,
"grad_norm": 1.568885492438894,
"learning_rate": 9.679999999999999e-08,
"loss": 0.0506,
"step": 242
},
{
"epoch": 0.0976,
"grad_norm": 8.732342707439889,
"learning_rate": 9.759999999999999e-08,
"loss": -0.7635,
"step": 244
},
{
"epoch": 0.0984,
"grad_norm": 6.993836328479647,
"learning_rate": 9.84e-08,
"loss": -0.6434,
"step": 246
},
{
"epoch": 0.0992,
"grad_norm": 0.7295566781785762,
"learning_rate": 9.919999999999999e-08,
"loss": 0.232,
"step": 248
},
{
"epoch": 0.1,
"grad_norm": 4.386003779293064,
"learning_rate": 1e-07,
"loss": -0.0654,
"step": 250
},
{
"epoch": 0.1008,
"grad_norm": 0.7038059549047322,
"learning_rate": 9.999980504497802e-08,
"loss": 0.6679,
"step": 252
},
{
"epoch": 0.1016,
"grad_norm": 2.9598822379536207,
"learning_rate": 9.99992201814324e-08,
"loss": 0.131,
"step": 254
},
{
"epoch": 0.1024,
"grad_norm": 0.715967277720422,
"learning_rate": 9.999824541392403e-08,
"loss": 0.2579,
"step": 256
},
{
"epoch": 0.1032,
"grad_norm": 3.2102671741061517,
"learning_rate": 9.999688075005433e-08,
"loss": 0.2389,
"step": 258
},
{
"epoch": 0.104,
"grad_norm": 6.929091165178379,
"learning_rate": 9.999512620046521e-08,
"loss": -0.7848,
"step": 260
},
{
"epoch": 0.1048,
"grad_norm": 2.1099354362146148,
"learning_rate": 9.999298177883901e-08,
"loss": 0.3355,
"step": 262
},
{
"epoch": 0.1056,
"grad_norm": 1.2644257110280757,
"learning_rate": 9.999044750189838e-08,
"loss": 0.0575,
"step": 264
},
{
"epoch": 0.1064,
"grad_norm": 1.996930248451988,
"learning_rate": 9.99875233894061e-08,
"loss": -0.2129,
"step": 266
},
{
"epoch": 0.1072,
"grad_norm": 1.5240927675226754,
"learning_rate": 9.9984209464165e-08,
"loss": -0.2003,
"step": 268
},
{
"epoch": 0.108,
"grad_norm": 4.774899294709947,
"learning_rate": 9.99805057520177e-08,
"loss": 0.638,
"step": 270
},
{
"epoch": 0.1088,
"grad_norm": 0.8387248005163368,
"learning_rate": 9.997641228184654e-08,
"loss": 0.3563,
"step": 272
},
{
"epoch": 0.1096,
"grad_norm": 1.208727180928816,
"learning_rate": 9.997192908557321e-08,
"loss": 0.069,
"step": 274
},
{
"epoch": 0.1104,
"grad_norm": 5.53379592764042,
"learning_rate": 9.996705619815856e-08,
"loss": -0.0175,
"step": 276
},
{
"epoch": 0.1112,
"grad_norm": 0.9868503521319727,
"learning_rate": 9.996179365760234e-08,
"loss": -0.291,
"step": 278
},
{
"epoch": 0.112,
"grad_norm": 1.9505786022160347,
"learning_rate": 9.995614150494291e-08,
"loss": -0.1812,
"step": 280
},
{
"epoch": 0.1128,
"grad_norm": 1.4087235500925508,
"learning_rate": 9.995009978425691e-08,
"loss": 0.3919,
"step": 282
},
{
"epoch": 0.1136,
"grad_norm": 0.9739855158507091,
"learning_rate": 9.994366854265884e-08,
"loss": 0.092,
"step": 284
},
{
"epoch": 0.1144,
"grad_norm": 0.8823957198854653,
"learning_rate": 9.993684783030087e-08,
"loss": 0.1602,
"step": 286
},
{
"epoch": 0.1152,
"grad_norm": 0.4284628297532619,
"learning_rate": 9.992963770037227e-08,
"loss": 0.0426,
"step": 288
},
{
"epoch": 0.116,
"grad_norm": 5.36183114688074,
"learning_rate": 9.992203820909905e-08,
"loss": -0.5808,
"step": 290
},
{
"epoch": 0.1168,
"grad_norm": 1.6702312862565898,
"learning_rate": 9.991404941574359e-08,
"loss": 0.6036,
"step": 292
},
{
"epoch": 0.1176,
"grad_norm": 4.073720902655535,
"learning_rate": 9.990567138260413e-08,
"loss": -0.2273,
"step": 294
},
{
"epoch": 0.1184,
"grad_norm": 0.9971658863201419,
"learning_rate": 9.989690417501422e-08,
"loss": -0.1263,
"step": 296
},
{
"epoch": 0.1192,
"grad_norm": 3.631953518488206,
"learning_rate": 9.988774786134233e-08,
"loss": 0.4006,
"step": 298
},
{
"epoch": 0.12,
"grad_norm": 0.7820782538072933,
"learning_rate": 9.98782025129912e-08,
"loss": -0.2066,
"step": 300
},
{
"epoch": 0.1208,
"grad_norm": 0.5878841242648716,
"learning_rate": 9.986826820439743e-08,
"loss": 0.0478,
"step": 302
},
{
"epoch": 0.1216,
"grad_norm": 4.974449133943707,
"learning_rate": 9.985794501303068e-08,
"loss": -0.1471,
"step": 304
},
{
"epoch": 0.1224,
"grad_norm": 0.601813755158989,
"learning_rate": 9.984723301939335e-08,
"loss": 0.7176,
"step": 306
},
{
"epoch": 0.1232,
"grad_norm": 1.730417084082021,
"learning_rate": 9.983613230701966e-08,
"loss": 0.2672,
"step": 308
},
{
"epoch": 0.124,
"grad_norm": 7.242082158593375,
"learning_rate": 9.982464296247522e-08,
"loss": -0.2698,
"step": 310
},
{
"epoch": 0.1248,
"grad_norm": 1.0223985096226282,
"learning_rate": 9.981276507535624e-08,
"loss": -0.0735,
"step": 312
},
{
"epoch": 0.1256,
"grad_norm": 0.8180401255353809,
"learning_rate": 9.980049873828886e-08,
"loss": 0.1891,
"step": 314
},
{
"epoch": 0.1264,
"grad_norm": 0.30041193539482725,
"learning_rate": 9.978784404692845e-08,
"loss": 0.2988,
"step": 316
},
{
"epoch": 0.1272,
"grad_norm": 1.1628849074237195,
"learning_rate": 9.977480109995883e-08,
"loss": 0.3059,
"step": 318
},
{
"epoch": 0.128,
"grad_norm": 0.4521879964000807,
"learning_rate": 9.976136999909155e-08,
"loss": -0.05,
"step": 320
},
{
"epoch": 0.1288,
"grad_norm": 1.5045100482252924,
"learning_rate": 9.974755084906502e-08,
"loss": 0.2253,
"step": 322
},
{
"epoch": 0.1296,
"grad_norm": 2.6671918830350028,
"learning_rate": 9.97333437576437e-08,
"loss": 0.4161,
"step": 324
},
{
"epoch": 0.1304,
"grad_norm": 1.3360264740430006,
"learning_rate": 9.97187488356174e-08,
"loss": 0.2868,
"step": 326
},
{
"epoch": 0.1312,
"grad_norm": 1.0280162385963292,
"learning_rate": 9.970376619680023e-08,
"loss": 0.2481,
"step": 328
},
{
"epoch": 0.132,
"grad_norm": 2.0832282668075215,
"learning_rate": 9.968839595802981e-08,
"loss": 0.0362,
"step": 330
},
{
"epoch": 0.1328,
"grad_norm": 2.0881240693312186,
"learning_rate": 9.967263823916638e-08,
"loss": 0.2028,
"step": 332
},
{
"epoch": 0.1336,
"grad_norm": 4.159751465077768,
"learning_rate": 9.965649316309176e-08,
"loss": -0.2774,
"step": 334
},
{
"epoch": 0.1344,
"grad_norm": 0.7533129835190434,
"learning_rate": 9.963996085570852e-08,
"loss": -0.1283,
"step": 336
},
{
"epoch": 0.1352,
"grad_norm": 2.2959735255682103,
"learning_rate": 9.962304144593891e-08,
"loss": 0.0126,
"step": 338
},
{
"epoch": 0.136,
"grad_norm": 2.197098452582562,
"learning_rate": 9.96057350657239e-08,
"loss": 0.4126,
"step": 340
},
{
"epoch": 0.1368,
"grad_norm": 1.58670105786256,
"learning_rate": 9.958804185002208e-08,
"loss": 0.0979,
"step": 342
},
{
"epoch": 0.1376,
"grad_norm": 0.8455424268179775,
"learning_rate": 9.956996193680873e-08,
"loss": 0.2987,
"step": 344
},
{
"epoch": 0.1384,
"grad_norm": 2.302950068704004,
"learning_rate": 9.955149546707462e-08,
"loss": -0.767,
"step": 346
},
{
"epoch": 0.1392,
"grad_norm": 7.511930834971542,
"learning_rate": 9.953264258482504e-08,
"loss": -0.0085,
"step": 348
},
{
"epoch": 0.14,
"grad_norm": 1.9232817968569562,
"learning_rate": 9.951340343707851e-08,
"loss": -0.3246,
"step": 350
},
{
"epoch": 0.1408,
"grad_norm": 2.6343239762516686,
"learning_rate": 9.949377817386579e-08,
"loss": 0.0478,
"step": 352
},
{
"epoch": 0.1416,
"grad_norm": 7.259506557578069,
"learning_rate": 9.94737669482286e-08,
"loss": 0.2815,
"step": 354
},
{
"epoch": 0.1424,
"grad_norm": 3.994459016722064,
"learning_rate": 9.945336991621852e-08,
"loss": 0.2718,
"step": 356
},
{
"epoch": 0.1432,
"grad_norm": 4.984507811555781,
"learning_rate": 9.94325872368957e-08,
"loss": 0.1226,
"step": 358
},
{
"epoch": 0.144,
"grad_norm": 1.79885150464052,
"learning_rate": 9.941141907232764e-08,
"loss": 0.2574,
"step": 360
},
{
"epoch": 0.1448,
"grad_norm": 1.3244125851707795,
"learning_rate": 9.938986558758794e-08,
"loss": 0.2669,
"step": 362
},
{
"epoch": 0.1456,
"grad_norm": 1.3154285100930394,
"learning_rate": 9.936792695075501e-08,
"loss": 0.2291,
"step": 364
},
{
"epoch": 0.1464,
"grad_norm": 6.029684432028888,
"learning_rate": 9.934560333291076e-08,
"loss": -0.6053,
"step": 366
},
{
"epoch": 0.1472,
"grad_norm": 0.5904631749204958,
"learning_rate": 9.932289490813921e-08,
"loss": 0.0027,
"step": 368
},
{
"epoch": 0.148,
"grad_norm": 2.339856125696244,
"learning_rate": 9.929980185352524e-08,
"loss": -0.0744,
"step": 370
},
{
"epoch": 0.1488,
"grad_norm": 0.688355527790139,
"learning_rate": 9.927632434915314e-08,
"loss": -0.1327,
"step": 372
},
{
"epoch": 0.1496,
"grad_norm": 3.7397017176068985,
"learning_rate": 9.925246257810518e-08,
"loss": 0.2688,
"step": 374
},
{
"epoch": 0.1504,
"grad_norm": 4.068851885945964,
"learning_rate": 9.922821672646027e-08,
"loss": 0.2873,
"step": 376
},
{
"epoch": 0.1512,
"grad_norm": 1.5558033071036992,
"learning_rate": 9.920358698329241e-08,
"loss": 0.3764,
"step": 378
},
{
"epoch": 0.152,
"grad_norm": 4.653760724189732,
"learning_rate": 9.91785735406693e-08,
"loss": 0.2148,
"step": 380
},
{
"epoch": 0.1528,
"grad_norm": 3.650076624124744,
"learning_rate": 9.915317659365076e-08,
"loss": 0.5768,
"step": 382
},
{
"epoch": 0.1536,
"grad_norm": 1.9826449037148115,
"learning_rate": 9.912739634028733e-08,
"loss": 0.222,
"step": 384
},
{
"epoch": 0.1544,
"grad_norm": 3.4468216333543062,
"learning_rate": 9.910123298161858e-08,
"loss": -0.1868,
"step": 386
},
{
"epoch": 0.1552,
"grad_norm": 3.165093593385751,
"learning_rate": 9.907468672167164e-08,
"loss": 0.091,
"step": 388
},
{
"epoch": 0.156,
"grad_norm": 1.7541440814162952,
"learning_rate": 9.904775776745957e-08,
"loss": -0.1529,
"step": 390
},
{
"epoch": 0.1568,
"grad_norm": 2.1449500638726633,
"learning_rate": 9.902044632897977e-08,
"loss": 0.4718,
"step": 392
},
{
"epoch": 0.1576,
"grad_norm": 1.2436286136089842,
"learning_rate": 9.899275261921234e-08,
"loss": 0.1954,
"step": 394
},
{
"epoch": 0.1584,
"grad_norm": 0.7997219065764383,
"learning_rate": 9.896467685411837e-08,
"loss": 0.175,
"step": 396
},
{
"epoch": 0.1592,
"grad_norm": 0.697589771763831,
"learning_rate": 9.89362192526383e-08,
"loss": 0.3507,
"step": 398
},
{
"epoch": 0.16,
"grad_norm": 3.5679392354742254,
"learning_rate": 9.890738003669028e-08,
"loss": 0.0395,
"step": 400
},
{
"epoch": 0.1608,
"grad_norm": 1.6853215016570104,
"learning_rate": 9.887815943116827e-08,
"loss": -0.5082,
"step": 402
},
{
"epoch": 0.1616,
"grad_norm": 6.85324036827588,
"learning_rate": 9.88485576639404e-08,
"loss": -0.1898,
"step": 404
},
{
"epoch": 0.1624,
"grad_norm": 0.7110369727557027,
"learning_rate": 9.881857496584725e-08,
"loss": 0.3305,
"step": 406
},
{
"epoch": 0.1632,
"grad_norm": 5.312906782551955,
"learning_rate": 9.878821157069988e-08,
"loss": -0.6801,
"step": 408
},
{
"epoch": 0.164,
"grad_norm": 1.6548236452861331,
"learning_rate": 9.875746771527815e-08,
"loss": 0.1536,
"step": 410
},
{
"epoch": 0.1648,
"grad_norm": 0.8245149921627257,
"learning_rate": 9.872634363932886e-08,
"loss": 0.2232,
"step": 412
},
{
"epoch": 0.1656,
"grad_norm": 3.3934339705566265,
"learning_rate": 9.869483958556375e-08,
"loss": 0.291,
"step": 414
},
{
"epoch": 0.1664,
"grad_norm": 1.4486896021406688,
"learning_rate": 9.86629557996578e-08,
"loss": -0.0683,
"step": 416
},
{
"epoch": 0.1672,
"grad_norm": 3.52336675846488,
"learning_rate": 9.863069253024718e-08,
"loss": -0.0894,
"step": 418
},
{
"epoch": 0.168,
"grad_norm": 1.1828263915820312,
"learning_rate": 9.859805002892732e-08,
"loss": 0.1527,
"step": 420
},
{
"epoch": 0.1688,
"grad_norm": 1.372569800294142,
"learning_rate": 9.856502855025099e-08,
"loss": 0.2844,
"step": 422
},
{
"epoch": 0.1696,
"grad_norm": 0.5902723465561974,
"learning_rate": 9.853162835172635e-08,
"loss": 0.1511,
"step": 424
},
{
"epoch": 0.1704,
"grad_norm": 8.507287499579553,
"learning_rate": 9.849784969381486e-08,
"loss": -0.3861,
"step": 426
},
{
"epoch": 0.1712,
"grad_norm": 3.9377780888287806,
"learning_rate": 9.846369283992926e-08,
"loss": -0.4743,
"step": 428
},
{
"epoch": 0.172,
"grad_norm": 6.138594217904337,
"learning_rate": 9.842915805643155e-08,
"loss": 0.0042,
"step": 430
},
{
"epoch": 0.1728,
"grad_norm": 1.0255687135099294,
"learning_rate": 9.839424561263092e-08,
"loss": -0.1807,
"step": 432
},
{
"epoch": 0.1736,
"grad_norm": 7.290670829010674,
"learning_rate": 9.835895578078164e-08,
"loss": -0.2322,
"step": 434
},
{
"epoch": 0.1744,
"grad_norm": 1.6945340966017044,
"learning_rate": 9.832328883608088e-08,
"loss": 0.2358,
"step": 436
},
{
"epoch": 0.1752,
"grad_norm": 1.1149664885651573,
"learning_rate": 9.828724505666664e-08,
"loss": 0.1624,
"step": 438
},
{
"epoch": 0.176,
"grad_norm": 0.33285825801011937,
"learning_rate": 9.825082472361557e-08,
"loss": 0.3141,
"step": 440
},
{
"epoch": 0.1768,
"grad_norm": 22.31453237037553,
"learning_rate": 9.821402812094073e-08,
"loss": -0.0172,
"step": 442
},
{
"epoch": 0.1776,
"grad_norm": 0.8704346962948754,
"learning_rate": 9.817685553558944e-08,
"loss": 0.1275,
"step": 444
},
{
"epoch": 0.1784,
"grad_norm": 1.2245015051060388,
"learning_rate": 9.813930725744094e-08,
"loss": 0.1762,
"step": 446
},
{
"epoch": 0.1792,
"grad_norm": 2.737100038252088,
"learning_rate": 9.810138357930428e-08,
"loss": 0.5655,
"step": 448
},
{
"epoch": 0.18,
"grad_norm": 1.284138044981843,
"learning_rate": 9.806308479691594e-08,
"loss": -0.7118,
"step": 450
},
{
"epoch": 0.1808,
"grad_norm": 3.123077229176161,
"learning_rate": 9.802441120893748e-08,
"loss": -0.2685,
"step": 452
},
{
"epoch": 0.1816,
"grad_norm": 0.875376956117242,
"learning_rate": 9.798536311695334e-08,
"loss": 0.2294,
"step": 454
},
{
"epoch": 0.1824,
"grad_norm": 7.85609200053166,
"learning_rate": 9.794594082546834e-08,
"loss": -0.4686,
"step": 456
},
{
"epoch": 0.1832,
"grad_norm": 0.9291930325028112,
"learning_rate": 9.790614464190548e-08,
"loss": -0.6784,
"step": 458
},
{
"epoch": 0.184,
"grad_norm": 0.9820076685954113,
"learning_rate": 9.786597487660335e-08,
"loss": -0.2107,
"step": 460
},
{
"epoch": 0.1848,
"grad_norm": 1.460713191289714,
"learning_rate": 9.782543184281388e-08,
"loss": 0.1244,
"step": 462
},
{
"epoch": 0.1856,
"grad_norm": 0.6179361492687564,
"learning_rate": 9.77845158566998e-08,
"loss": 0.1714,
"step": 464
},
{
"epoch": 0.1864,
"grad_norm": 0.4346256375122305,
"learning_rate": 9.774322723733215e-08,
"loss": 0.3512,
"step": 466
},
{
"epoch": 0.1872,
"grad_norm": 1.2715138565567314,
"learning_rate": 9.770156630668789e-08,
"loss": 0.1171,
"step": 468
},
{
"epoch": 0.188,
"grad_norm": 1.5946034157660203,
"learning_rate": 9.765953338964735e-08,
"loss": 0.3309,
"step": 470
},
{
"epoch": 0.1888,
"grad_norm": 0.7622483814283684,
"learning_rate": 9.761712881399163e-08,
"loss": 0.4465,
"step": 472
},
{
"epoch": 0.1896,
"grad_norm": 1.9649939425146115,
"learning_rate": 9.757435291040015e-08,
"loss": 0.2973,
"step": 474
},
{
"epoch": 0.1904,
"grad_norm": 0.9914397389532802,
"learning_rate": 9.7531206012448e-08,
"loss": 0.3274,
"step": 476
},
{
"epoch": 0.1912,
"grad_norm": 0.7243228964834183,
"learning_rate": 9.748768845660334e-08,
"loss": 0.2104,
"step": 478
},
{
"epoch": 0.192,
"grad_norm": 0.6054705552582188,
"learning_rate": 9.744380058222482e-08,
"loss": 0.0408,
"step": 480
},
{
"epoch": 0.1928,
"grad_norm": 1.1899869470335924,
"learning_rate": 9.739954273155891e-08,
"loss": 0.1459,
"step": 482
},
{
"epoch": 0.1936,
"grad_norm": 0.808307853178962,
"learning_rate": 9.735491524973722e-08,
"loss": 0.1125,
"step": 484
},
{
"epoch": 0.1944,
"grad_norm": 1.0815367893705388,
"learning_rate": 9.730991848477379e-08,
"loss": -0.1682,
"step": 486
},
{
"epoch": 0.1952,
"grad_norm": 1.1920456124546996,
"learning_rate": 9.726455278756247e-08,
"loss": 0.2205,
"step": 488
},
{
"epoch": 0.196,
"grad_norm": 1.7732089420749002,
"learning_rate": 9.721881851187405e-08,
"loss": 0.1552,
"step": 490
},
{
"epoch": 0.1968,
"grad_norm": 2.2419073705232715,
"learning_rate": 9.717271601435362e-08,
"loss": 0.2732,
"step": 492
},
{
"epoch": 0.1976,
"grad_norm": 0.7131704530453624,
"learning_rate": 9.71262456545177e-08,
"loss": 0.2964,
"step": 494
},
{
"epoch": 0.1984,
"grad_norm": 1.9224161883701323,
"learning_rate": 9.70794077947515e-08,
"loss": -0.1305,
"step": 496
},
{
"epoch": 0.1992,
"grad_norm": 2.420968387762715,
"learning_rate": 9.703220280030607e-08,
"loss": -0.1025,
"step": 498
},
{
"epoch": 0.2,
"grad_norm": 0.5917101352557091,
"learning_rate": 9.698463103929542e-08,
"loss": 0.4162,
"step": 500
},
{
"epoch": 0.2008,
"grad_norm": 0.37261372003965865,
"learning_rate": 9.693669288269371e-08,
"loss": -0.0685,
"step": 502
},
{
"epoch": 0.2016,
"grad_norm": 0.9143691951148681,
"learning_rate": 9.688838870433229e-08,
"loss": -0.2294,
"step": 504
},
{
"epoch": 0.2024,
"grad_norm": 0.9193655718296037,
"learning_rate": 9.683971888089688e-08,
"loss": 0.391,
"step": 506
},
{
"epoch": 0.2032,
"grad_norm": 7.494175919423912,
"learning_rate": 9.679068379192454e-08,
"loss": -0.3181,
"step": 508
},
{
"epoch": 0.204,
"grad_norm": 19.535752409376773,
"learning_rate": 9.67412838198007e-08,
"loss": -0.86,
"step": 510
},
{
"epoch": 0.2048,
"grad_norm": 1.1410695459720073,
"learning_rate": 9.669151934975633e-08,
"loss": 0.2809,
"step": 512
},
{
"epoch": 0.2056,
"grad_norm": 1.377010762791552,
"learning_rate": 9.664139076986472e-08,
"loss": 0.162,
"step": 514
},
{
"epoch": 0.2064,
"grad_norm": 0.557979664652959,
"learning_rate": 9.659089847103862e-08,
"loss": 0.33,
"step": 516
},
{
"epoch": 0.2072,
"grad_norm": 1.553002684885157,
"learning_rate": 9.654004284702711e-08,
"loss": 0.0052,
"step": 518
},
{
"epoch": 0.208,
"grad_norm": 0.889311576739284,
"learning_rate": 9.648882429441257e-08,
"loss": 0.1826,
"step": 520
},
{
"epoch": 0.2088,
"grad_norm": 3.4858180208508043,
"learning_rate": 9.643724321260756e-08,
"loss": -0.3282,
"step": 522
},
{
"epoch": 0.2096,
"grad_norm": 2.1719560909307174,
"learning_rate": 9.63853000038517e-08,
"loss": 0.4663,
"step": 524
},
{
"epoch": 0.2104,
"grad_norm": 1.084770634896945,
"learning_rate": 9.63329950732086e-08,
"loss": 0.1794,
"step": 526
},
{
"epoch": 0.2112,
"grad_norm": 1.570077846681942,
"learning_rate": 9.628032882856261e-08,
"loss": 0.4361,
"step": 528
},
{
"epoch": 0.212,
"grad_norm": 3.257604693080942,
"learning_rate": 9.622730168061565e-08,
"loss": 0.267,
"step": 530
},
{
"epoch": 0.2128,
"grad_norm": 1.2962127715499705,
"learning_rate": 9.61739140428841e-08,
"loss": -0.5759,
"step": 532
},
{
"epoch": 0.2136,
"grad_norm": 0.4783553609890881,
"learning_rate": 9.612016633169549e-08,
"loss": -0.4113,
"step": 534
},
{
"epoch": 0.2144,
"grad_norm": 1.1329977586766047,
"learning_rate": 9.606605896618526e-08,
"loss": -0.6282,
"step": 536
},
{
"epoch": 0.2152,
"grad_norm": 9.092699988961803,
"learning_rate": 9.601159236829351e-08,
"loss": 0.0028,
"step": 538
},
{
"epoch": 0.216,
"grad_norm": 1.7521340541068868,
"learning_rate": 9.595676696276171e-08,
"loss": 0.3847,
"step": 540
},
{
"epoch": 0.2168,
"grad_norm": 1.5211299849991022,
"learning_rate": 9.59015831771294e-08,
"loss": 0.1855,
"step": 542
},
{
"epoch": 0.2176,
"grad_norm": 2.3997915440768085,
"learning_rate": 9.584604144173082e-08,
"loss": 0.0154,
"step": 544
},
{
"epoch": 0.2184,
"grad_norm": 7.7867544210113655,
"learning_rate": 9.579014218969157e-08,
"loss": -0.2799,
"step": 546
},
{
"epoch": 0.2192,
"grad_norm": 2.246425515456882,
"learning_rate": 9.573388585692524e-08,
"loss": 0.0657,
"step": 548
},
{
"epoch": 0.22,
"grad_norm": 0.8920881325520396,
"learning_rate": 9.567727288213003e-08,
"loss": 0.0662,
"step": 550
},
{
"epoch": 0.2208,
"grad_norm": 2.069608229929687,
"learning_rate": 9.562030370678531e-08,
"loss": -0.352,
"step": 552
},
{
"epoch": 0.2216,
"grad_norm": 7.019347228888489,
"learning_rate": 9.556297877514811e-08,
"loss": 0.6677,
"step": 554
},
{
"epoch": 0.2224,
"grad_norm": 0.7140207308779009,
"learning_rate": 9.550529853424978e-08,
"loss": -0.1234,
"step": 556
},
{
"epoch": 0.2232,
"grad_norm": 1.739597248566285,
"learning_rate": 9.544726343389244e-08,
"loss": 0.176,
"step": 558
},
{
"epoch": 0.224,
"grad_norm": 3.759867753849374,
"learning_rate": 9.538887392664543e-08,
"loss": -0.4603,
"step": 560
},
{
"epoch": 0.2248,
"grad_norm": 0.4680240642658474,
"learning_rate": 9.533013046784188e-08,
"loss": -0.1854,
"step": 562
},
{
"epoch": 0.2256,
"grad_norm": 4.0223246024546455,
"learning_rate": 9.527103351557508e-08,
"loss": -0.5082,
"step": 564
},
{
"epoch": 0.2264,
"grad_norm": 0.4298188905176273,
"learning_rate": 9.521158353069493e-08,
"loss": 0.352,
"step": 566
},
{
"epoch": 0.2272,
"grad_norm": 6.510408037187395,
"learning_rate": 9.515178097680436e-08,
"loss": -0.1873,
"step": 568
},
{
"epoch": 0.228,
"grad_norm": 2.7431324947016487,
"learning_rate": 9.509162632025569e-08,
"loss": -0.2765,
"step": 570
},
{
"epoch": 0.2288,
"grad_norm": 1.7314378634262093,
"learning_rate": 9.503112003014701e-08,
"loss": 0.1129,
"step": 572
},
{
"epoch": 0.2296,
"grad_norm": 0.4152358895518038,
"learning_rate": 9.497026257831855e-08,
"loss": -0.1104,
"step": 574
},
{
"epoch": 0.2304,
"grad_norm": 3.0143102426871207,
"learning_rate": 9.490905443934891e-08,
"loss": 0.233,
"step": 576
},
{
"epoch": 0.2312,
"grad_norm": 1.254785319884344,
"learning_rate": 9.484749609055149e-08,
"loss": 0.0741,
"step": 578
},
{
"epoch": 0.232,
"grad_norm": 0.4516039050645454,
"learning_rate": 9.478558801197064e-08,
"loss": -0.5912,
"step": 580
},
{
"epoch": 0.2328,
"grad_norm": 2.0311852841582443,
"learning_rate": 9.4723330686378e-08,
"loss": 0.5177,
"step": 582
},
{
"epoch": 0.2336,
"grad_norm": 0.7527146505774129,
"learning_rate": 9.466072459926868e-08,
"loss": -0.2521,
"step": 584
},
{
"epoch": 0.2344,
"grad_norm": 1.7528339797546952,
"learning_rate": 9.459777023885753e-08,
"loss": 0.1981,
"step": 586
},
{
"epoch": 0.2352,
"grad_norm": 0.9484429211584349,
"learning_rate": 9.453446809607533e-08,
"loss": 0.0565,
"step": 588
},
{
"epoch": 0.236,
"grad_norm": 2.576119540140844,
"learning_rate": 9.447081866456487e-08,
"loss": -0.3846,
"step": 590
},
{
"epoch": 0.2368,
"grad_norm": 1.9661922778166474,
"learning_rate": 9.440682244067722e-08,
"loss": -0.1543,
"step": 592
},
{
"epoch": 0.2376,
"grad_norm": 1.0393258447068892,
"learning_rate": 9.434247992346779e-08,
"loss": 0.3713,
"step": 594
},
{
"epoch": 0.2384,
"grad_norm": 0.926942209886049,
"learning_rate": 9.427779161469245e-08,
"loss": 0.1346,
"step": 596
},
{
"epoch": 0.2392,
"grad_norm": 2.329000993582009,
"learning_rate": 9.421275801880362e-08,
"loss": -0.126,
"step": 598
},
{
"epoch": 0.24,
"grad_norm": 1.6875574527329094,
"learning_rate": 9.414737964294634e-08,
"loss": 0.3109,
"step": 600
},
{
"epoch": 0.2408,
"grad_norm": 0.5023669006234125,
"learning_rate": 9.408165699695434e-08,
"loss": -0.0132,
"step": 602
},
{
"epoch": 0.2416,
"grad_norm": 2.0881231177462456,
"learning_rate": 9.4015590593346e-08,
"loss": 0.5198,
"step": 604
},
{
"epoch": 0.2424,
"grad_norm": 1.0863869319130528,
"learning_rate": 9.394918094732042e-08,
"loss": 0.1587,
"step": 606
},
{
"epoch": 0.2432,
"grad_norm": 1.5582534983831973,
"learning_rate": 9.388242857675335e-08,
"loss": -0.3655,
"step": 608
},
{
"epoch": 0.244,
"grad_norm": 0.5461657587975756,
"learning_rate": 9.381533400219317e-08,
"loss": 0.0735,
"step": 610
},
{
"epoch": 0.2448,
"grad_norm": 2.4930252888255886,
"learning_rate": 9.374789774685689e-08,
"loss": -0.9575,
"step": 612
},
{
"epoch": 0.2456,
"grad_norm": 2.0194068377552132,
"learning_rate": 9.368012033662593e-08,
"loss": -0.2119,
"step": 614
},
{
"epoch": 0.2464,
"grad_norm": 1.3750808155144445,
"learning_rate": 9.361200230004218e-08,
"loss": -0.0461,
"step": 616
},
{
"epoch": 0.2472,
"grad_norm": 0.8051442888860127,
"learning_rate": 9.354354416830377e-08,
"loss": -0.2925,
"step": 618
},
{
"epoch": 0.248,
"grad_norm": 0.8432217688207168,
"learning_rate": 9.347474647526095e-08,
"loss": -0.2154,
"step": 620
},
{
"epoch": 0.2488,
"grad_norm": 0.6853710470578885,
"learning_rate": 9.340560975741196e-08,
"loss": 0.4846,
"step": 622
},
{
"epoch": 0.2496,
"grad_norm": 1.403223282772839,
"learning_rate": 9.333613455389882e-08,
"loss": 0.0296,
"step": 624
},
{
"epoch": 0.2504,
"grad_norm": 2.1131053251547693,
"learning_rate": 9.32663214065031e-08,
"loss": -0.2981,
"step": 626
},
{
"epoch": 0.2512,
"grad_norm": 0.7873380820908549,
"learning_rate": 9.319617085964175e-08,
"loss": 0.2121,
"step": 628
},
{
"epoch": 0.252,
"grad_norm": 0.9528682959478656,
"learning_rate": 9.312568346036287e-08,
"loss": 0.4284,
"step": 630
},
{
"epoch": 0.2528,
"grad_norm": 1.097234538369852,
"learning_rate": 9.30548597583413e-08,
"loss": 0.3954,
"step": 632
},
{
"epoch": 0.2536,
"grad_norm": 1.8851125035672085,
"learning_rate": 9.298370030587455e-08,
"loss": 0.4441,
"step": 634
},
{
"epoch": 0.2544,
"grad_norm": 2.1853617450331693,
"learning_rate": 9.291220565787828e-08,
"loss": 0.4437,
"step": 636
},
{
"epoch": 0.2552,
"grad_norm": 2.3936375731578625,
"learning_rate": 9.284037637188213e-08,
"loss": -0.4182,
"step": 638
},
{
"epoch": 0.256,
"grad_norm": 2.646325229764863,
"learning_rate": 9.276821300802533e-08,
"loss": 0.119,
"step": 640
},
{
"epoch": 0.2568,
"grad_norm": 0.9756272329795392,
"learning_rate": 9.269571612905225e-08,
"loss": -0.3106,
"step": 642
},
{
"epoch": 0.2576,
"grad_norm": 1.4379456308252396,
"learning_rate": 9.262288630030812e-08,
"loss": 0.3465,
"step": 644
},
{
"epoch": 0.2584,
"grad_norm": 0.6377811081598714,
"learning_rate": 9.254972408973459e-08,
"loss": 0.0878,
"step": 646
},
{
"epoch": 0.2592,
"grad_norm": 1.5771004258451025,
"learning_rate": 9.247623006786527e-08,
"loss": 0.4204,
"step": 648
},
{
"epoch": 0.26,
"grad_norm": 0.8176119979876094,
"learning_rate": 9.240240480782128e-08,
"loss": 0.3434,
"step": 650
},
{
"epoch": 0.2608,
"grad_norm": 1.1167330010140615,
"learning_rate": 9.232824888530688e-08,
"loss": -0.1252,
"step": 652
},
{
"epoch": 0.2616,
"grad_norm": 2.902711665458815,
"learning_rate": 9.225376287860483e-08,
"loss": 0.1784,
"step": 654
},
{
"epoch": 0.2624,
"grad_norm": 1.1021902043030998,
"learning_rate": 9.217894736857194e-08,
"loss": 0.077,
"step": 656
},
{
"epoch": 0.2632,
"grad_norm": 1.5828751888652974,
"learning_rate": 9.210380293863461e-08,
"loss": 0.6164,
"step": 658
},
{
"epoch": 0.264,
"grad_norm": 1.7579085274204935,
"learning_rate": 9.20283301747842e-08,
"loss": 0.2696,
"step": 660
},
{
"epoch": 0.2648,
"grad_norm": 0.7147277225164167,
"learning_rate": 9.19525296655725e-08,
"loss": 0.2439,
"step": 662
},
{
"epoch": 0.2656,
"grad_norm": 3.6763274266756794,
"learning_rate": 9.187640200210708e-08,
"loss": -0.6141,
"step": 664
},
{
"epoch": 0.2664,
"grad_norm": 0.8865442357420804,
"learning_rate": 9.179994777804675e-08,
"loss": -0.0577,
"step": 666
},
{
"epoch": 0.2672,
"grad_norm": 1.8760630726535639,
"learning_rate": 9.172316758959695e-08,
"loss": -0.3246,
"step": 668
},
{
"epoch": 0.268,
"grad_norm": 4.0499280919666685,
"learning_rate": 9.164606203550497e-08,
"loss": -0.3236,
"step": 670
},
{
"epoch": 0.2688,
"grad_norm": 0.2388213957572059,
"learning_rate": 9.156863171705543e-08,
"loss": 0.3181,
"step": 672
},
{
"epoch": 0.2696,
"grad_norm": 0.7858037287124114,
"learning_rate": 9.149087723806548e-08,
"loss": -0.3492,
"step": 674
},
{
"epoch": 0.2704,
"grad_norm": 0.7846789738735171,
"learning_rate": 9.141279920488021e-08,
"loss": -0.0031,
"step": 676
},
{
"epoch": 0.2712,
"grad_norm": 0.6079509641140145,
"learning_rate": 9.133439822636777e-08,
"loss": 0.5217,
"step": 678
},
{
"epoch": 0.272,
"grad_norm": 3.1157318757855785,
"learning_rate": 9.125567491391475e-08,
"loss": -0.2582,
"step": 680
},
{
"epoch": 0.2728,
"grad_norm": 2.8344713124178016,
"learning_rate": 9.117662988142136e-08,
"loss": -0.0037,
"step": 682
},
{
"epoch": 0.2736,
"grad_norm": 1.1048675461441084,
"learning_rate": 9.109726374529665e-08,
"loss": -0.3684,
"step": 684
},
{
"epoch": 0.2744,
"grad_norm": 0.6734941872876059,
"learning_rate": 9.101757712445368e-08,
"loss": -0.3566,
"step": 686
},
{
"epoch": 0.2752,
"grad_norm": 5.2809791174675205,
"learning_rate": 9.093757064030472e-08,
"loss": 0.4808,
"step": 688
},
{
"epoch": 0.276,
"grad_norm": 1.0801262787173656,
"learning_rate": 9.085724491675641e-08,
"loss": 0.2411,
"step": 690
},
{
"epoch": 0.2768,
"grad_norm": 2.006574190663329,
"learning_rate": 9.077660058020491e-08,
"loss": 0.4176,
"step": 692
},
{
"epoch": 0.2776,
"grad_norm": 1.2950834174730455,
"learning_rate": 9.06956382595309e-08,
"loss": 0.1336,
"step": 694
},
{
"epoch": 0.2784,
"grad_norm": 5.230890162016179,
"learning_rate": 9.061435858609485e-08,
"loss": -0.1621,
"step": 696
},
{
"epoch": 0.2792,
"grad_norm": 12.169821952332974,
"learning_rate": 9.053276219373199e-08,
"loss": 0.2004,
"step": 698
},
{
"epoch": 0.28,
"grad_norm": 3.660251173933071,
"learning_rate": 9.045084971874737e-08,
"loss": 0.2771,
"step": 700
},
{
"epoch": 0.2808,
"grad_norm": 1.163297828202275,
"learning_rate": 9.036862179991091e-08,
"loss": 0.2461,
"step": 702
},
{
"epoch": 0.2816,
"grad_norm": 1.0335361654882531,
"learning_rate": 9.028607907845246e-08,
"loss": 0.1717,
"step": 704
},
{
"epoch": 0.2824,
"grad_norm": 1.8229718995331785,
"learning_rate": 9.020322219805673e-08,
"loss": 0.2178,
"step": 706
},
{
"epoch": 0.2832,
"grad_norm": 2.2765609772885416,
"learning_rate": 9.012005180485833e-08,
"loss": 0.2202,
"step": 708
},
{
"epoch": 0.284,
"grad_norm": 0.4850755241158522,
"learning_rate": 9.003656854743666e-08,
"loss": 0.2864,
"step": 710
},
{
"epoch": 0.2848,
"grad_norm": 1.4341650344819474,
"learning_rate": 8.995277307681098e-08,
"loss": -0.6383,
"step": 712
},
{
"epoch": 0.2856,
"grad_norm": 12.279835227706903,
"learning_rate": 8.986866604643517e-08,
"loss": -0.7269,
"step": 714
},
{
"epoch": 0.2864,
"grad_norm": 3.9952094340159023,
"learning_rate": 8.978424811219276e-08,
"loss": -0.1366,
"step": 716
},
{
"epoch": 0.2872,
"grad_norm": 0.8124079844550998,
"learning_rate": 8.969951993239177e-08,
"loss": 0.1862,
"step": 718
},
{
"epoch": 0.288,
"grad_norm": 8.097449745501883,
"learning_rate": 8.961448216775953e-08,
"loss": -0.3673,
"step": 720
},
{
"epoch": 0.2888,
"grad_norm": 1.1732802960791322,
"learning_rate": 8.952913548143764e-08,
"loss": 0.4006,
"step": 722
},
{
"epoch": 0.2896,
"grad_norm": 1.5360714076560518,
"learning_rate": 8.94434805389767e-08,
"loss": 0.1462,
"step": 724
},
{
"epoch": 0.2904,
"grad_norm": 0.5114107508157508,
"learning_rate": 8.935751800833116e-08,
"loss": 0.0877,
"step": 726
},
{
"epoch": 0.2912,
"grad_norm": 1.427690720685384,
"learning_rate": 8.927124855985408e-08,
"loss": 0.1363,
"step": 728
},
{
"epoch": 0.292,
"grad_norm": 4.4627972614515565,
"learning_rate": 8.918467286629199e-08,
"loss": -0.3287,
"step": 730
},
{
"epoch": 0.2928,
"grad_norm": 3.3272363902338182,
"learning_rate": 8.90977916027795e-08,
"loss": -0.2832,
"step": 732
},
{
"epoch": 0.2936,
"grad_norm": 0.7746604883747309,
"learning_rate": 8.901060544683418e-08,
"loss": 0.0766,
"step": 734
},
{
"epoch": 0.2944,
"grad_norm": 1.5873072227936746,
"learning_rate": 8.892311507835117e-08,
"loss": 0.4345,
"step": 736
},
{
"epoch": 0.2952,
"grad_norm": 1.3768473445563956,
"learning_rate": 8.883532117959796e-08,
"loss": 0.0693,
"step": 738
},
{
"epoch": 0.296,
"grad_norm": 20.81939969162968,
"learning_rate": 8.874722443520898e-08,
"loss": -0.4809,
"step": 740
},
{
"epoch": 0.2968,
"grad_norm": 14.945778018331922,
"learning_rate": 8.865882553218036e-08,
"loss": -0.5521,
"step": 742
},
{
"epoch": 0.2976,
"grad_norm": 0.39785231657900016,
"learning_rate": 8.857012515986451e-08,
"loss": 0.1503,
"step": 744
},
{
"epoch": 0.2984,
"grad_norm": 1.2686422724522106,
"learning_rate": 8.848112400996473e-08,
"loss": 0.2367,
"step": 746
},
{
"epoch": 0.2992,
"grad_norm": 1.569638918420828,
"learning_rate": 8.839182277652987e-08,
"loss": -0.4723,
"step": 748
},
{
"epoch": 0.3,
"grad_norm": 1.1198134951928183,
"learning_rate": 8.83022221559489e-08,
"loss": -0.2171,
"step": 750
},
{
"epoch": 0.3008,
"grad_norm": 0.8104270584597341,
"learning_rate": 8.821232284694544e-08,
"loss": 0.0921,
"step": 752
},
{
"epoch": 0.3016,
"grad_norm": 2.9028381646787453,
"learning_rate": 8.812212555057239e-08,
"loss": -0.0545,
"step": 754
},
{
"epoch": 0.3024,
"grad_norm": 0.8899926905526622,
"learning_rate": 8.803163097020636e-08,
"loss": -0.0558,
"step": 756
},
{
"epoch": 0.3032,
"grad_norm": 1.7828842004216436,
"learning_rate": 8.794083981154227e-08,
"loss": 0.7208,
"step": 758
},
{
"epoch": 0.304,
"grad_norm": 2.9786430555101795,
"learning_rate": 8.784975278258782e-08,
"loss": -0.078,
"step": 760
},
{
"epoch": 0.3048,
"grad_norm": 2.8287310614940777,
"learning_rate": 8.775837059365795e-08,
"loss": 0.088,
"step": 762
},
{
"epoch": 0.3056,
"grad_norm": 0.8845002129698117,
"learning_rate": 8.766669395736934e-08,
"loss": -0.0491,
"step": 764
},
{
"epoch": 0.3064,
"grad_norm": 3.5268253668779934,
"learning_rate": 8.75747235886348e-08,
"loss": 0.0034,
"step": 766
},
{
"epoch": 0.3072,
"grad_norm": 1.7452615611959836,
"learning_rate": 8.748246020465775e-08,
"loss": 0.4788,
"step": 768
},
{
"epoch": 0.308,
"grad_norm": 13.507874280818228,
"learning_rate": 8.738990452492659e-08,
"loss": -0.1238,
"step": 770
},
{
"epoch": 0.3088,
"grad_norm": 3.4954667832405177,
"learning_rate": 8.729705727120911e-08,
"loss": 0.1762,
"step": 772
},
{
"epoch": 0.3096,
"grad_norm": 11.084239211030003,
"learning_rate": 8.720391916754682e-08,
"loss": -0.3814,
"step": 774
},
{
"epoch": 0.3104,
"grad_norm": 1.190008032651553,
"learning_rate": 8.711049094024941e-08,
"loss": 0.4016,
"step": 776
},
{
"epoch": 0.3112,
"grad_norm": 2.6524719701229373,
"learning_rate": 8.70167733178889e-08,
"loss": -0.4937,
"step": 778
},
{
"epoch": 0.312,
"grad_norm": 3.6563349073371434,
"learning_rate": 8.69227670312942e-08,
"loss": 0.2101,
"step": 780
},
{
"epoch": 0.3128,
"grad_norm": 5.589523245640799,
"learning_rate": 8.682847281354516e-08,
"loss": -0.5319,
"step": 782
},
{
"epoch": 0.3136,
"grad_norm": 0.9937104750252121,
"learning_rate": 8.673389139996707e-08,
"loss": 0.5273,
"step": 784
},
{
"epoch": 0.3144,
"grad_norm": 1.5490092245622005,
"learning_rate": 8.663902352812478e-08,
"loss": 0.4851,
"step": 786
},
{
"epoch": 0.3152,
"grad_norm": 0.9247150320866911,
"learning_rate": 8.654386993781701e-08,
"loss": -0.1605,
"step": 788
},
{
"epoch": 0.316,
"grad_norm": 0.8661553436999507,
"learning_rate": 8.644843137107057e-08,
"loss": 0.0903,
"step": 790
},
{
"epoch": 0.3168,
"grad_norm": 0.8966245324936362,
"learning_rate": 8.635270857213459e-08,
"loss": 0.2469,
"step": 792
},
{
"epoch": 0.3176,
"grad_norm": 6.075740058565734,
"learning_rate": 8.625670228747466e-08,
"loss": 0.3518,
"step": 794
},
{
"epoch": 0.3184,
"grad_norm": 5.768120136868427,
"learning_rate": 8.61604132657671e-08,
"loss": -0.1788,
"step": 796
},
{
"epoch": 0.3192,
"grad_norm": 8.014889998847247,
"learning_rate": 8.606384225789303e-08,
"loss": -0.5493,
"step": 798
},
{
"epoch": 0.32,
"grad_norm": 2.732855863760049,
"learning_rate": 8.596699001693255e-08,
"loss": -0.4401,
"step": 800
},
{
"epoch": 0.3208,
"grad_norm": 5.674762098665864,
"learning_rate": 8.586985729815893e-08,
"loss": -0.6151,
"step": 802
},
{
"epoch": 0.3216,
"grad_norm": 1.203107745655249,
"learning_rate": 8.577244485903259e-08,
"loss": 0.3376,
"step": 804
},
{
"epoch": 0.3224,
"grad_norm": 0.7743743600052015,
"learning_rate": 8.567475345919531e-08,
"loss": 0.2416,
"step": 806
},
{
"epoch": 0.3232,
"grad_norm": 3.2313530818726686,
"learning_rate": 8.557678386046428e-08,
"loss": -0.2523,
"step": 808
},
{
"epoch": 0.324,
"grad_norm": 3.2135557626127316,
"learning_rate": 8.547853682682604e-08,
"loss": -0.8818,
"step": 810
},
{
"epoch": 0.3248,
"grad_norm": 2.443067918618587,
"learning_rate": 8.538001312443076e-08,
"loss": -0.623,
"step": 812
},
{
"epoch": 0.3256,
"grad_norm": 6.5609892559503855,
"learning_rate": 8.528121352158604e-08,
"loss": -0.4258,
"step": 814
},
{
"epoch": 0.3264,
"grad_norm": 0.7000997613564107,
"learning_rate": 8.518213878875102e-08,
"loss": 0.1103,
"step": 816
},
{
"epoch": 0.3272,
"grad_norm": 6.1200733738526525,
"learning_rate": 8.508278969853036e-08,
"loss": 0.1271,
"step": 818
},
{
"epoch": 0.328,
"grad_norm": 2.6047004705005787,
"learning_rate": 8.498316702566827e-08,
"loss": 0.0993,
"step": 820
},
{
"epoch": 0.3288,
"grad_norm": 0.9677372851549512,
"learning_rate": 8.488327154704232e-08,
"loss": 0.0594,
"step": 822
},
{
"epoch": 0.3296,
"grad_norm": 3.4430495127951977,
"learning_rate": 8.478310404165754e-08,
"loss": -0.1936,
"step": 824
},
{
"epoch": 0.3304,
"grad_norm": 0.3601588143401079,
"learning_rate": 8.468266529064024e-08,
"loss": -0.1365,
"step": 826
},
{
"epoch": 0.3312,
"grad_norm": 1.292385094296757,
"learning_rate": 8.4581956077232e-08,
"loss": -0.3464,
"step": 828
},
{
"epoch": 0.332,
"grad_norm": 1.9117854973492814,
"learning_rate": 8.448097718678348e-08,
"loss": 0.4335,
"step": 830
},
{
"epoch": 0.3328,
"grad_norm": 1.8283142200028784,
"learning_rate": 8.437972940674837e-08,
"loss": -0.0491,
"step": 832
},
{
"epoch": 0.3336,
"grad_norm": 3.8320528125639104,
"learning_rate": 8.427821352667717e-08,
"loss": -0.4612,
"step": 834
},
{
"epoch": 0.3344,
"grad_norm": 1.8095888332666445,
"learning_rate": 8.417643033821113e-08,
"loss": 0.4844,
"step": 836
},
{
"epoch": 0.3352,
"grad_norm": 1.5804671191319755,
"learning_rate": 8.407438063507599e-08,
"loss": 0.1414,
"step": 838
},
{
"epoch": 0.336,
"grad_norm": 2.4402963026274174,
"learning_rate": 8.397206521307583e-08,
"loss": 0.2388,
"step": 840
},
{
"epoch": 0.3368,
"grad_norm": 1.4966413361633364,
"learning_rate": 8.386948487008686e-08,
"loss": 0.5727,
"step": 842
},
{
"epoch": 0.3376,
"grad_norm": 1.7284763532645568,
"learning_rate": 8.376664040605121e-08,
"loss": -0.3214,
"step": 844
},
{
"epoch": 0.3384,
"grad_norm": 0.7112316019800302,
"learning_rate": 8.366353262297068e-08,
"loss": 0.1026,
"step": 846
},
{
"epoch": 0.3392,
"grad_norm": 10.49080301872087,
"learning_rate": 8.356016232490046e-08,
"loss": -0.8355,
"step": 848
},
{
"epoch": 0.34,
"grad_norm": 0.7290221856294098,
"learning_rate": 8.34565303179429e-08,
"loss": -0.3258,
"step": 850
},
{
"epoch": 0.3408,
"grad_norm": 1.1703618452430726,
"learning_rate": 8.335263741024122e-08,
"loss": 0.6358,
"step": 852
},
{
"epoch": 0.3416,
"grad_norm": 2.850142298515102,
"learning_rate": 8.324848441197316e-08,
"loss": -0.1372,
"step": 854
},
{
"epoch": 0.3424,
"grad_norm": 0.973368655758617,
"learning_rate": 8.314407213534475e-08,
"loss": 0.9766,
"step": 856
},
{
"epoch": 0.3432,
"grad_norm": 1.3793864686515331,
"learning_rate": 8.303940139458388e-08,
"loss": 0.2899,
"step": 858
},
{
"epoch": 0.344,
"grad_norm": 1.6558863887284723,
"learning_rate": 8.293447300593401e-08,
"loss": -0.6861,
"step": 860
},
{
"epoch": 0.3448,
"grad_norm": 3.229334640432589,
"learning_rate": 8.282928778764781e-08,
"loss": 0.4103,
"step": 862
},
{
"epoch": 0.3456,
"grad_norm": 1.9022283387070154,
"learning_rate": 8.272384655998074e-08,
"loss": 0.3683,
"step": 864
},
{
"epoch": 0.3464,
"grad_norm": 3.861434626902648,
"learning_rate": 8.261815014518466e-08,
"loss": 0.1213,
"step": 866
},
{
"epoch": 0.3472,
"grad_norm": 0.3346123728991653,
"learning_rate": 8.251219936750143e-08,
"loss": 0.3042,
"step": 868
},
{
"epoch": 0.348,
"grad_norm": 1.804028815127322,
"learning_rate": 8.240599505315654e-08,
"loss": 0.3678,
"step": 870
},
{
"epoch": 0.3488,
"grad_norm": 2.22654583204003,
"learning_rate": 8.229953803035255e-08,
"loss": -0.5044,
"step": 872
},
{
"epoch": 0.3496,
"grad_norm": 2.8036340506880593,
"learning_rate": 8.219282912926268e-08,
"loss": -0.0045,
"step": 874
},
{
"epoch": 0.3504,
"grad_norm": 1.2472980474279702,
"learning_rate": 8.208586918202443e-08,
"loss": -0.4355,
"step": 876
},
{
"epoch": 0.3512,
"grad_norm": 2.0236633425024078,
"learning_rate": 8.19786590227329e-08,
"loss": -0.4585,
"step": 878
},
{
"epoch": 0.352,
"grad_norm": 2.553585628725583,
"learning_rate": 8.187119948743448e-08,
"loss": -0.4744,
"step": 880
},
{
"epoch": 0.3528,
"grad_norm": 4.526772208608427,
"learning_rate": 8.176349141412021e-08,
"loss": 0.0674,
"step": 882
},
{
"epoch": 0.3536,
"grad_norm": 3.0131476143455584,
"learning_rate": 8.165553564271928e-08,
"loss": -0.8816,
"step": 884
},
{
"epoch": 0.3544,
"grad_norm": 2.4303141324929434,
"learning_rate": 8.154733301509249e-08,
"loss": 0.0354,
"step": 886
},
{
"epoch": 0.3552,
"grad_norm": 1.4923649469015177,
"learning_rate": 8.143888437502565e-08,
"loss": 0.2033,
"step": 888
},
{
"epoch": 0.356,
"grad_norm": 1.0499738094965403,
"learning_rate": 8.133019056822303e-08,
"loss": -0.084,
"step": 890
},
{
"epoch": 0.3568,
"grad_norm": 2.306429104903375,
"learning_rate": 8.122125244230078e-08,
"loss": -0.0835,
"step": 892
},
{
"epoch": 0.3576,
"grad_norm": 9.447440348177143,
"learning_rate": 8.111207084678033e-08,
"loss": -0.1714,
"step": 894
},
{
"epoch": 0.3584,
"grad_norm": 3.3146862784828723,
"learning_rate": 8.100264663308162e-08,
"loss": 0.3046,
"step": 896
},
{
"epoch": 0.3592,
"grad_norm": 0.5247128917973009,
"learning_rate": 8.089298065451672e-08,
"loss": 0.0776,
"step": 898
},
{
"epoch": 0.36,
"grad_norm": 2.6116572800466438,
"learning_rate": 8.07830737662829e-08,
"loss": 0.1196,
"step": 900
},
{
"epoch": 0.3608,
"grad_norm": 1.7532572130157607,
"learning_rate": 8.06729268254562e-08,
"loss": 0.4049,
"step": 902
},
{
"epoch": 0.3616,
"grad_norm": 0.872383155375364,
"learning_rate": 8.056254069098459e-08,
"loss": 0.4134,
"step": 904
},
{
"epoch": 0.3624,
"grad_norm": 2.0280592050080077,
"learning_rate": 8.045191622368127e-08,
"loss": -0.1264,
"step": 906
},
{
"epoch": 0.3632,
"grad_norm": 7.2837279920656295,
"learning_rate": 8.034105428621811e-08,
"loss": -0.4519,
"step": 908
},
{
"epoch": 0.364,
"grad_norm": 2.010871912862232,
"learning_rate": 8.022995574311875e-08,
"loss": 0.0267,
"step": 910
},
{
"epoch": 0.3648,
"grad_norm": 2.18845588960694,
"learning_rate": 8.011862146075193e-08,
"loss": -0.5606,
"step": 912
},
{
"epoch": 0.3656,
"grad_norm": 10.672904965184774,
"learning_rate": 8.000705230732477e-08,
"loss": -0.3007,
"step": 914
},
{
"epoch": 0.3664,
"grad_norm": 16.229779007192096,
"learning_rate": 7.989524915287594e-08,
"loss": 0.1867,
"step": 916
},
{
"epoch": 0.3672,
"grad_norm": 1.321736634994464,
"learning_rate": 7.978321286926891e-08,
"loss": 0.1814,
"step": 918
},
{
"epoch": 0.368,
"grad_norm": 2.137899951816609,
"learning_rate": 7.967094433018507e-08,
"loss": 0.4335,
"step": 920
},
{
"epoch": 0.3688,
"grad_norm": 0.8963724642161506,
"learning_rate": 7.95584444111171e-08,
"loss": -0.0275,
"step": 922
},
{
"epoch": 0.3696,
"grad_norm": 0.784399520286717,
"learning_rate": 7.944571398936193e-08,
"loss": 0.0624,
"step": 924
},
{
"epoch": 0.3704,
"grad_norm": 3.366159505865075,
"learning_rate": 7.933275394401407e-08,
"loss": -0.1544,
"step": 926
},
{
"epoch": 0.3712,
"grad_norm": 0.8620601529573644,
"learning_rate": 7.92195651559586e-08,
"loss": -0.1152,
"step": 928
},
{
"epoch": 0.372,
"grad_norm": 4.160557824531761,
"learning_rate": 7.910614850786447e-08,
"loss": -0.1786,
"step": 930
},
{
"epoch": 0.3728,
"grad_norm": 1.1732726155263975,
"learning_rate": 7.899250488417746e-08,
"loss": -0.1146,
"step": 932
},
{
"epoch": 0.3736,
"grad_norm": 0.7626058693505505,
"learning_rate": 7.887863517111337e-08,
"loss": 0.0288,
"step": 934
},
{
"epoch": 0.3744,
"grad_norm": 1.2784721116511524,
"learning_rate": 7.876454025665113e-08,
"loss": 0.3692,
"step": 936
},
{
"epoch": 0.3752,
"grad_norm": 3.8432163391561636,
"learning_rate": 7.865022103052577e-08,
"loss": -0.6191,
"step": 938
},
{
"epoch": 0.376,
"grad_norm": 0.41200718458395047,
"learning_rate": 7.853567838422159e-08,
"loss": -0.2775,
"step": 940
},
{
"epoch": 0.3768,
"grad_norm": 3.210456122992894,
"learning_rate": 7.842091321096515e-08,
"loss": 0.04,
"step": 942
},
{
"epoch": 0.3776,
"grad_norm": 8.946861643817154,
"learning_rate": 7.830592640571832e-08,
"loss": 0.0712,
"step": 944
},
{
"epoch": 0.3784,
"grad_norm": 1.1380080448822825,
"learning_rate": 7.819071886517134e-08,
"loss": 0.2481,
"step": 946
},
{
"epoch": 0.3792,
"grad_norm": 0.903931154067042,
"learning_rate": 7.807529148773572e-08,
"loss": 0.5167,
"step": 948
},
{
"epoch": 0.38,
"grad_norm": 5.787579761684592,
"learning_rate": 7.795964517353733e-08,
"loss": -0.2009,
"step": 950
},
{
"epoch": 0.3808,
"grad_norm": 1.23258440991286,
"learning_rate": 7.78437808244094e-08,
"loss": 0.1332,
"step": 952
},
{
"epoch": 0.3816,
"grad_norm": 8.57781181693697,
"learning_rate": 7.772769934388537e-08,
"loss": -0.3453,
"step": 954
},
{
"epoch": 0.3824,
"grad_norm": 9.984455229867079,
"learning_rate": 7.761140163719193e-08,
"loss": -0.9802,
"step": 956
},
{
"epoch": 0.3832,
"grad_norm": 14.58764723753251,
"learning_rate": 7.749488861124199e-08,
"loss": -0.4665,
"step": 958
},
{
"epoch": 0.384,
"grad_norm": 3.4788077577478496,
"learning_rate": 7.737816117462751e-08,
"loss": -0.4591,
"step": 960
},
{
"epoch": 0.3848,
"grad_norm": 0.8756186903705511,
"learning_rate": 7.72612202376125e-08,
"loss": -0.2308,
"step": 962
},
{
"epoch": 0.3856,
"grad_norm": 0.21704219860580345,
"learning_rate": 7.714406671212587e-08,
"loss": 0.3277,
"step": 964
},
{
"epoch": 0.3864,
"grad_norm": 5.59613367286991,
"learning_rate": 7.702670151175435e-08,
"loss": -0.5679,
"step": 966
},
{
"epoch": 0.3872,
"grad_norm": 0.84568283496626,
"learning_rate": 7.690912555173535e-08,
"loss": 0.1284,
"step": 968
},
{
"epoch": 0.388,
"grad_norm": 4.19251355544568,
"learning_rate": 7.679133974894983e-08,
"loss": -0.4117,
"step": 970
},
{
"epoch": 0.3888,
"grad_norm": 1.087531497704038,
"learning_rate": 7.667334502191513e-08,
"loss": 0.1098,
"step": 972
},
{
"epoch": 0.3896,
"grad_norm": 1.603047898028564,
"learning_rate": 7.655514229077783e-08,
"loss": -0.4038,
"step": 974
},
{
"epoch": 0.3904,
"grad_norm": 2.333486581996855,
"learning_rate": 7.643673247730658e-08,
"loss": 0.0592,
"step": 976
},
{
"epoch": 0.3912,
"grad_norm": 8.69472674858354,
"learning_rate": 7.631811650488489e-08,
"loss": -0.7602,
"step": 978
},
{
"epoch": 0.392,
"grad_norm": 1.0210334223940398,
"learning_rate": 7.619929529850396e-08,
"loss": 0.2272,
"step": 980
},
{
"epoch": 0.3928,
"grad_norm": 1.048418990456133,
"learning_rate": 7.608026978475539e-08,
"loss": -0.4598,
"step": 982
},
{
"epoch": 0.3936,
"grad_norm": 0.9998711691112794,
"learning_rate": 7.596104089182406e-08,
"loss": 0.1215,
"step": 984
},
{
"epoch": 0.3944,
"grad_norm": 3.0306983649369843,
"learning_rate": 7.584160954948084e-08,
"loss": 0.0192,
"step": 986
},
{
"epoch": 0.3952,
"grad_norm": 10.037202598057146,
"learning_rate": 7.572197668907532e-08,
"loss": 0.0057,
"step": 988
},
{
"epoch": 0.396,
"grad_norm": 0.5591265397866508,
"learning_rate": 7.560214324352857e-08,
"loss": 0.1889,
"step": 990
},
{
"epoch": 0.3968,
"grad_norm": 2.1451327721737683,
"learning_rate": 7.548211014732589e-08,
"loss": -0.0682,
"step": 992
},
{
"epoch": 0.3976,
"grad_norm": 1.4947596521706652,
"learning_rate": 7.536187833650945e-08,
"loss": 0.0717,
"step": 994
},
{
"epoch": 0.3984,
"grad_norm": 0.7487011304939436,
"learning_rate": 7.524144874867109e-08,
"loss": -0.041,
"step": 996
},
{
"epoch": 0.3992,
"grad_norm": 2.8470129286226644,
"learning_rate": 7.51208223229449e-08,
"loss": -0.685,
"step": 998
},
{
"epoch": 0.4,
"grad_norm": 1.1509305577734978,
"learning_rate": 7.5e-08,
"loss": 0.4244,
"step": 1000
},
{
"epoch": 0.4008,
"grad_norm": 2.2471149932142804,
"learning_rate": 7.487898272203312e-08,
"loss": -0.2415,
"step": 1002
},
{
"epoch": 0.4016,
"grad_norm": 2.3336181887777396,
"learning_rate": 7.475777143276131e-08,
"loss": -0.1276,
"step": 1004
},
{
"epoch": 0.4024,
"grad_norm": 3.451215377583405,
"learning_rate": 7.463636707741458e-08,
"loss": 0.186,
"step": 1006
},
{
"epoch": 0.4032,
"grad_norm": 4.432546794380947,
"learning_rate": 7.451477060272842e-08,
"loss": -0.4754,
"step": 1008
},
{
"epoch": 0.404,
"grad_norm": 1.2726178274814892,
"learning_rate": 7.439298295693663e-08,
"loss": -0.1686,
"step": 1010
},
{
"epoch": 0.4048,
"grad_norm": 1.6595160080860332,
"learning_rate": 7.427100508976369e-08,
"loss": 0.4122,
"step": 1012
},
{
"epoch": 0.4056,
"grad_norm": 0.9527094988783708,
"learning_rate": 7.414883795241753e-08,
"loss": 0.0897,
"step": 1014
},
{
"epoch": 0.4064,
"grad_norm": 0.8197073391125492,
"learning_rate": 7.402648249758203e-08,
"loss": 0.0477,
"step": 1016
},
{
"epoch": 0.4072,
"grad_norm": 0.4730719086957203,
"learning_rate": 7.390393967940962e-08,
"loss": 0.0388,
"step": 1018
},
{
"epoch": 0.408,
"grad_norm": 0.6938425628800243,
"learning_rate": 7.378121045351376e-08,
"loss": -0.0765,
"step": 1020
},
{
"epoch": 0.4088,
"grad_norm": 2.4855748056225404,
"learning_rate": 7.365829577696165e-08,
"loss": -0.4727,
"step": 1022
},
{
"epoch": 0.4096,
"grad_norm": 1.2525228424015165,
"learning_rate": 7.353519660826664e-08,
"loss": -0.1318,
"step": 1024
},
{
"epoch": 0.4104,
"grad_norm": 0.6828130325148093,
"learning_rate": 7.341191390738072e-08,
"loss": 0.0043,
"step": 1026
},
{
"epoch": 0.4112,
"grad_norm": 2.016735168065313,
"learning_rate": 7.32884486356872e-08,
"loss": 0.3202,
"step": 1028
},
{
"epoch": 0.412,
"grad_norm": 4.271680320972975,
"learning_rate": 7.316480175599309e-08,
"loss": -0.2884,
"step": 1030
},
{
"epoch": 0.4128,
"grad_norm": 0.44940231823968296,
"learning_rate": 7.304097423252155e-08,
"loss": 0.279,
"step": 1032
},
{
"epoch": 0.4136,
"grad_norm": 1.7416239543161751,
"learning_rate": 7.291696703090449e-08,
"loss": 0.1598,
"step": 1034
},
{
"epoch": 0.4144,
"grad_norm": 8.770657113289268,
"learning_rate": 7.2792781118175e-08,
"loss": -0.5416,
"step": 1036
},
{
"epoch": 0.4152,
"grad_norm": 0.9351697205394859,
"learning_rate": 7.266841746275975e-08,
"loss": 0.1555,
"step": 1038
},
{
"epoch": 0.416,
"grad_norm": 2.5220066205160485,
"learning_rate": 7.254387703447154e-08,
"loss": 0.1865,
"step": 1040
},
{
"epoch": 0.4168,
"grad_norm": 2.209916897967486,
"learning_rate": 7.241916080450162e-08,
"loss": -0.263,
"step": 1042
},
{
"epoch": 0.4176,
"grad_norm": 5.469961942467383,
"learning_rate": 7.22942697454122e-08,
"loss": -0.2177,
"step": 1044
},
{
"epoch": 0.4184,
"grad_norm": 0.5272933331368399,
"learning_rate": 7.216920483112885e-08,
"loss": 0.0983,
"step": 1046
},
{
"epoch": 0.4192,
"grad_norm": 1.4143301998010451,
"learning_rate": 7.204396703693293e-08,
"loss": -0.2537,
"step": 1048
},
{
"epoch": 0.42,
"grad_norm": 0.4246102208293558,
"learning_rate": 7.191855733945387e-08,
"loss": 0.3528,
"step": 1050
},
{
"epoch": 0.4208,
"grad_norm": 0.5561666198760952,
"learning_rate": 7.17929767166617e-08,
"loss": 0.0547,
"step": 1052
},
{
"epoch": 0.4216,
"grad_norm": 0.8856453777963561,
"learning_rate": 7.166722614785936e-08,
"loss": -0.4613,
"step": 1054
},
{
"epoch": 0.4224,
"grad_norm": 3.8778489840855306,
"learning_rate": 7.154130661367503e-08,
"loss": -0.3592,
"step": 1056
},
{
"epoch": 0.4232,
"grad_norm": 1.111848344560476,
"learning_rate": 7.141521909605451e-08,
"loss": 0.3647,
"step": 1058
},
{
"epoch": 0.424,
"grad_norm": 0.8976496692377764,
"learning_rate": 7.128896457825363e-08,
"loss": -0.3394,
"step": 1060
},
{
"epoch": 0.4248,
"grad_norm": 1.9854864333684115,
"learning_rate": 7.116254404483048e-08,
"loss": 0.5498,
"step": 1062
},
{
"epoch": 0.4256,
"grad_norm": 0.518945106869338,
"learning_rate": 7.103595848163774e-08,
"loss": 0.1052,
"step": 1064
},
{
"epoch": 0.4264,
"grad_norm": 0.81312188171614,
"learning_rate": 7.090920887581506e-08,
"loss": 0.2085,
"step": 1066
},
{
"epoch": 0.4272,
"grad_norm": 2.542255401078519,
"learning_rate": 7.078229621578139e-08,
"loss": -0.401,
"step": 1068
},
{
"epoch": 0.428,
"grad_norm": 0.7969133711740632,
"learning_rate": 7.06552214912271e-08,
"loss": 0.2313,
"step": 1070
},
{
"epoch": 0.4288,
"grad_norm": 2.013951635316577,
"learning_rate": 7.05279856931064e-08,
"loss": -0.0206,
"step": 1072
},
{
"epoch": 0.4296,
"grad_norm": 3.233439286834546,
"learning_rate": 7.040058981362963e-08,
"loss": 0.0567,
"step": 1074
},
{
"epoch": 0.4304,
"grad_norm": 2.038925823472598,
"learning_rate": 7.027303484625546e-08,
"loss": 0.4609,
"step": 1076
},
{
"epoch": 0.4312,
"grad_norm": 5.3939877710558495,
"learning_rate": 7.014532178568313e-08,
"loss": -0.2891,
"step": 1078
},
{
"epoch": 0.432,
"grad_norm": 5.083293406395563,
"learning_rate": 7.001745162784475e-08,
"loss": -0.7984,
"step": 1080
},
{
"epoch": 0.4328,
"grad_norm": 1.6561122980559624,
"learning_rate": 6.988942536989749e-08,
"loss": -0.3287,
"step": 1082
},
{
"epoch": 0.4336,
"grad_norm": 1.4731754331723663,
"learning_rate": 6.976124401021582e-08,
"loss": -0.3449,
"step": 1084
},
{
"epoch": 0.4344,
"grad_norm": 2.130459655887093,
"learning_rate": 6.963290854838375e-08,
"loss": 0.1084,
"step": 1086
},
{
"epoch": 0.4352,
"grad_norm": 0.7379825159584281,
"learning_rate": 6.950441998518698e-08,
"loss": 0.1926,
"step": 1088
},
{
"epoch": 0.436,
"grad_norm": 5.9866066131944375,
"learning_rate": 6.937577932260514e-08,
"loss": -0.2981,
"step": 1090
},
{
"epoch": 0.4368,
"grad_norm": 1.6941594895354701,
"learning_rate": 6.924698756380397e-08,
"loss": -0.04,
"step": 1092
},
{
"epoch": 0.4376,
"grad_norm": 17.324712344427475,
"learning_rate": 6.911804571312744e-08,
"loss": 0.631,
"step": 1094
},
{
"epoch": 0.4384,
"grad_norm": 5.676574825229265,
"learning_rate": 6.898895477609006e-08,
"loss": 0.4486,
"step": 1096
},
{
"epoch": 0.4392,
"grad_norm": 2.188959065969351,
"learning_rate": 6.885971575936883e-08,
"loss": 0.1522,
"step": 1098
},
{
"epoch": 0.44,
"grad_norm": 1.205606460513578,
"learning_rate": 6.87303296707956e-08,
"loss": -0.1985,
"step": 1100
},
{
"epoch": 0.4408,
"grad_norm": 2.860447532152119,
"learning_rate": 6.860079751934908e-08,
"loss": 0.4336,
"step": 1102
},
{
"epoch": 0.4416,
"grad_norm": 0.6325136441386785,
"learning_rate": 6.847112031514696e-08,
"loss": 0.0757,
"step": 1104
},
{
"epoch": 0.4424,
"grad_norm": 1.2112086018815402,
"learning_rate": 6.83412990694382e-08,
"loss": -0.3603,
"step": 1106
},
{
"epoch": 0.4432,
"grad_norm": 5.835113794965243,
"learning_rate": 6.82113347945949e-08,
"loss": 0.031,
"step": 1108
},
{
"epoch": 0.444,
"grad_norm": 1.414086162153916,
"learning_rate": 6.808122850410461e-08,
"loss": 0.6686,
"step": 1110
},
{
"epoch": 0.4448,
"grad_norm": 1.9404063883884395,
"learning_rate": 6.79509812125623e-08,
"loss": -0.039,
"step": 1112
},
{
"epoch": 0.4456,
"grad_norm": 3.1144430060760655,
"learning_rate": 6.782059393566253e-08,
"loss": 0.1551,
"step": 1114
},
{
"epoch": 0.4464,
"grad_norm": 0.954886541069574,
"learning_rate": 6.769006769019147e-08,
"loss": 0.4238,
"step": 1116
},
{
"epoch": 0.4472,
"grad_norm": 2.417988928194323,
"learning_rate": 6.755940349401899e-08,
"loss": 0.4811,
"step": 1118
},
{
"epoch": 0.448,
"grad_norm": 0.7492639665364688,
"learning_rate": 6.742860236609076e-08,
"loss": 0.1234,
"step": 1120
},
{
"epoch": 0.4488,
"grad_norm": 1.3267895869858666,
"learning_rate": 6.729766532642024e-08,
"loss": -0.0093,
"step": 1122
},
{
"epoch": 0.4496,
"grad_norm": 3.1523703905692986,
"learning_rate": 6.716659339608076e-08,
"loss": -1.1553,
"step": 1124
},
{
"epoch": 0.4504,
"grad_norm": 6.3271911129401595,
"learning_rate": 6.70353875971976e-08,
"loss": 0.4817,
"step": 1126
},
{
"epoch": 0.4512,
"grad_norm": 3.458276951838705,
"learning_rate": 6.690404895293986e-08,
"loss": -0.106,
"step": 1128
},
{
"epoch": 0.452,
"grad_norm": 1.1983521344982837,
"learning_rate": 6.677257848751275e-08,
"loss": -0.1368,
"step": 1130
},
{
"epoch": 0.4528,
"grad_norm": 2.562911563697214,
"learning_rate": 6.664097722614933e-08,
"loss": -0.6246,
"step": 1132
},
{
"epoch": 0.4536,
"grad_norm": 2.292976959310392,
"learning_rate": 6.650924619510268e-08,
"loss": -0.3059,
"step": 1134
},
{
"epoch": 0.4544,
"grad_norm": 1.9146997642645225,
"learning_rate": 6.637738642163784e-08,
"loss": -0.4552,
"step": 1136
},
{
"epoch": 0.4552,
"grad_norm": 2.303506195238013,
"learning_rate": 6.624539893402382e-08,
"loss": -0.0411,
"step": 1138
},
{
"epoch": 0.456,
"grad_norm": 2.429304538352616,
"learning_rate": 6.611328476152556e-08,
"loss": -0.09,
"step": 1140
},
{
"epoch": 0.4568,
"grad_norm": 1.5901142036458538,
"learning_rate": 6.598104493439589e-08,
"loss": 0.366,
"step": 1142
},
{
"epoch": 0.4576,
"grad_norm": 2.2501212390984793,
"learning_rate": 6.58486804838676e-08,
"loss": -0.2237,
"step": 1144
},
{
"epoch": 0.4584,
"grad_norm": 3.1938054664306015,
"learning_rate": 6.57161924421452e-08,
"loss": -0.2891,
"step": 1146
},
{
"epoch": 0.4592,
"grad_norm": 0.8017365109031118,
"learning_rate": 6.558358184239709e-08,
"loss": -0.1112,
"step": 1148
},
{
"epoch": 0.46,
"grad_norm": 2.0186522257754107,
"learning_rate": 6.545084971874738e-08,
"loss": 0.1282,
"step": 1150
},
{
"epoch": 0.4608,
"grad_norm": 1.474938378888324,
"learning_rate": 6.531799710626778e-08,
"loss": 0.1798,
"step": 1152
},
{
"epoch": 0.4616,
"grad_norm": 7.871129967536662,
"learning_rate": 6.518502504096971e-08,
"loss": -0.8227,
"step": 1154
},
{
"epoch": 0.4624,
"grad_norm": 2.8902307300787933,
"learning_rate": 6.505193455979603e-08,
"loss": 0.2635,
"step": 1156
},
{
"epoch": 0.4632,
"grad_norm": 1.143186288398571,
"learning_rate": 6.491872670061302e-08,
"loss": -0.0355,
"step": 1158
},
{
"epoch": 0.464,
"grad_norm": 3.7075989676791417,
"learning_rate": 6.478540250220234e-08,
"loss": -0.0786,
"step": 1160
},
{
"epoch": 0.4648,
"grad_norm": 4.792785138183704,
"learning_rate": 6.465196300425286e-08,
"loss": -0.4437,
"step": 1162
},
{
"epoch": 0.4656,
"grad_norm": 0.8922724462770598,
"learning_rate": 6.451840924735264e-08,
"loss": -0.2213,
"step": 1164
},
{
"epoch": 0.4664,
"grad_norm": 2.491727325910274,
"learning_rate": 6.438474227298065e-08,
"loss": 0.0669,
"step": 1166
},
{
"epoch": 0.4672,
"grad_norm": 1.4732480640539267,
"learning_rate": 6.42509631234988e-08,
"loss": 0.2351,
"step": 1168
},
{
"epoch": 0.468,
"grad_norm": 8.137988516764139,
"learning_rate": 6.411707284214382e-08,
"loss": -0.3911,
"step": 1170
},
{
"epoch": 0.4688,
"grad_norm": 8.004252108406288,
"learning_rate": 6.398307247301899e-08,
"loss": -0.0938,
"step": 1172
},
{
"epoch": 0.4696,
"grad_norm": 1.9959364399427888,
"learning_rate": 6.384896306108611e-08,
"loss": -0.1873,
"step": 1174
},
{
"epoch": 0.4704,
"grad_norm": 10.098214183461678,
"learning_rate": 6.371474565215733e-08,
"loss": 0.0064,
"step": 1176
},
{
"epoch": 0.4712,
"grad_norm": 0.6662802639277847,
"learning_rate": 6.358042129288693e-08,
"loss": -0.1318,
"step": 1178
},
{
"epoch": 0.472,
"grad_norm": 3.1607119261285224,
"learning_rate": 6.344599103076328e-08,
"loss": 0.7315,
"step": 1180
},
{
"epoch": 0.4728,
"grad_norm": 2.020466390354618,
"learning_rate": 6.331145591410057e-08,
"loss": 0.2858,
"step": 1182
},
{
"epoch": 0.4736,
"grad_norm": 0.4928769684178318,
"learning_rate": 6.317681699203063e-08,
"loss": 0.2666,
"step": 1184
},
{
"epoch": 0.4744,
"grad_norm": 9.872920047662523,
"learning_rate": 6.304207531449485e-08,
"loss": 0.0105,
"step": 1186
},
{
"epoch": 0.4752,
"grad_norm": 8.16277185549274,
"learning_rate": 6.290723193223589e-08,
"loss": -0.5512,
"step": 1188
},
{
"epoch": 0.476,
"grad_norm": 2.8766477891050917,
"learning_rate": 6.277228789678953e-08,
"loss": 0.6091,
"step": 1190
},
{
"epoch": 0.4768,
"grad_norm": 0.6120893860982233,
"learning_rate": 6.263724426047647e-08,
"loss": 0.1113,
"step": 1192
},
{
"epoch": 0.4776,
"grad_norm": 2.867371220131585,
"learning_rate": 6.25021020763941e-08,
"loss": -0.1485,
"step": 1194
},
{
"epoch": 0.4784,
"grad_norm": 2.3805466052582953,
"learning_rate": 6.236686239840835e-08,
"loss": 0.1272,
"step": 1196
},
{
"epoch": 0.4792,
"grad_norm": 0.8751211655303213,
"learning_rate": 6.223152628114536e-08,
"loss": 0.1164,
"step": 1198
},
{
"epoch": 0.48,
"grad_norm": 3.7721597694707953,
"learning_rate": 6.209609477998338e-08,
"loss": 0.7313,
"step": 1200
},
{
"epoch": 0.4808,
"grad_norm": 3.419798472878702,
"learning_rate": 6.196056895104447e-08,
"loss": 0.013,
"step": 1202
},
{
"epoch": 0.4816,
"grad_norm": 4.687774153819853,
"learning_rate": 6.182494985118624e-08,
"loss": 0.23,
"step": 1204
},
{
"epoch": 0.4824,
"grad_norm": 2.125081697473343,
"learning_rate": 6.168923853799368e-08,
"loss": -0.0661,
"step": 1206
},
{
"epoch": 0.4832,
"grad_norm": 2.556521965745637,
"learning_rate": 6.15534360697709e-08,
"loss": -0.1983,
"step": 1208
},
{
"epoch": 0.484,
"grad_norm": 1.2329813510099201,
"learning_rate": 6.141754350553279e-08,
"loss": 0.2845,
"step": 1210
},
{
"epoch": 0.4848,
"grad_norm": 1.3362598649725015,
"learning_rate": 6.128156190499687e-08,
"loss": -0.2324,
"step": 1212
},
{
"epoch": 0.4856,
"grad_norm": 6.79709632065024,
"learning_rate": 6.114549232857502e-08,
"loss": -0.3551,
"step": 1214
},
{
"epoch": 0.4864,
"grad_norm": 1.2638697254758269,
"learning_rate": 6.100933583736507e-08,
"loss": 0.255,
"step": 1216
},
{
"epoch": 0.4872,
"grad_norm": 5.7754492883128385,
"learning_rate": 6.087309349314274e-08,
"loss": -0.6309,
"step": 1218
},
{
"epoch": 0.488,
"grad_norm": 0.9571850975922288,
"learning_rate": 6.073676635835316e-08,
"loss": -0.2826,
"step": 1220
},
{
"epoch": 0.4888,
"grad_norm": 1.4309851718621818,
"learning_rate": 6.060035549610274e-08,
"loss": 0.2769,
"step": 1222
},
{
"epoch": 0.4896,
"grad_norm": 1.0145754210048563,
"learning_rate": 6.046386197015075e-08,
"loss": -0.9362,
"step": 1224
},
{
"epoch": 0.4904,
"grad_norm": 6.626223220436085,
"learning_rate": 6.032728684490118e-08,
"loss": 0.4322,
"step": 1226
},
{
"epoch": 0.4912,
"grad_norm": 1.3984038449961043,
"learning_rate": 6.019063118539424e-08,
"loss": -0.4794,
"step": 1228
},
{
"epoch": 0.492,
"grad_norm": 0.5800974015428817,
"learning_rate": 6.005389605729824e-08,
"loss": -0.2262,
"step": 1230
},
{
"epoch": 0.4928,
"grad_norm": 1.6940574308328136,
"learning_rate": 5.991708252690116e-08,
"loss": 0.1601,
"step": 1232
},
{
"epoch": 0.4936,
"grad_norm": 1.18887607438921,
"learning_rate": 5.978019166110241e-08,
"loss": 0.3557,
"step": 1234
},
{
"epoch": 0.4944,
"grad_norm": 3.2297465206800453,
"learning_rate": 5.964322452740445e-08,
"loss": 0.4208,
"step": 1236
},
{
"epoch": 0.4952,
"grad_norm": 2.763317691111721,
"learning_rate": 5.950618219390451e-08,
"loss": -0.101,
"step": 1238
},
{
"epoch": 0.496,
"grad_norm": 1.9082022213348315,
"learning_rate": 5.936906572928624e-08,
"loss": 0.0438,
"step": 1240
},
{
"epoch": 0.4968,
"grad_norm": 0.7433758520431626,
"learning_rate": 5.923187620281135e-08,
"loss": 0.3283,
"step": 1242
},
{
"epoch": 0.4976,
"grad_norm": 2.7729120367779854,
"learning_rate": 5.909461468431134e-08,
"loss": -0.0354,
"step": 1244
},
{
"epoch": 0.4984,
"grad_norm": 20.0112850631017,
"learning_rate": 5.895728224417912e-08,
"loss": -1.2127,
"step": 1246
},
{
"epoch": 0.4992,
"grad_norm": 1.0654897055122667,
"learning_rate": 5.881987995336062e-08,
"loss": -0.0584,
"step": 1248
},
{
"epoch": 0.5,
"grad_norm": 2.7229608606775404,
"learning_rate": 5.868240888334653e-08,
"loss": -0.6298,
"step": 1250
},
{
"epoch": 0.5008,
"grad_norm": 4.006337091885257,
"learning_rate": 5.854487010616384e-08,
"loss": -0.3494,
"step": 1252
},
{
"epoch": 0.5016,
"grad_norm": 5.495686782120867,
"learning_rate": 5.840726469436757e-08,
"loss": 0.2171,
"step": 1254
},
{
"epoch": 0.5024,
"grad_norm": 1.1311488028436698,
"learning_rate": 5.826959372103239e-08,
"loss": 0.3292,
"step": 1256
},
{
"epoch": 0.5032,
"grad_norm": 3.065820295521943,
"learning_rate": 5.8131858259744184e-08,
"loss": -0.4639,
"step": 1258
},
{
"epoch": 0.504,
"grad_norm": 1.6251303612715686,
"learning_rate": 5.799405938459174e-08,
"loss": -0.1503,
"step": 1260
},
{
"epoch": 0.5048,
"grad_norm": 0.863499018143362,
"learning_rate": 5.7856198170158386e-08,
"loss": 0.2304,
"step": 1262
},
{
"epoch": 0.5056,
"grad_norm": 1.1532321430846115,
"learning_rate": 5.771827569151356e-08,
"loss": 0.459,
"step": 1264
},
{
"epoch": 0.5064,
"grad_norm": 1.7764896164807287,
"learning_rate": 5.758029302420445e-08,
"loss": -0.0018,
"step": 1266
},
{
"epoch": 0.5072,
"grad_norm": 2.650509930359115,
"learning_rate": 5.744225124424761e-08,
"loss": 0.4307,
"step": 1268
},
{
"epoch": 0.508,
"grad_norm": 18.546073788989784,
"learning_rate": 5.730415142812058e-08,
"loss": -0.3913,
"step": 1270
},
{
"epoch": 0.5088,
"grad_norm": 0.9501942078311948,
"learning_rate": 5.716599465275347e-08,
"loss": -0.1711,
"step": 1272
},
{
"epoch": 0.5096,
"grad_norm": 1.138324851030599,
"learning_rate": 5.702778199552054e-08,
"loss": 0.2208,
"step": 1274
},
{
"epoch": 0.5104,
"grad_norm": 6.125933607526262,
"learning_rate": 5.68895145342319e-08,
"loss": -0.6492,
"step": 1276
},
{
"epoch": 0.5112,
"grad_norm": 1.656529407336565,
"learning_rate": 5.6751193347124956e-08,
"loss": -0.855,
"step": 1278
},
{
"epoch": 0.512,
"grad_norm": 4.206513837985695,
"learning_rate": 5.6612819512856126e-08,
"loss": -0.3752,
"step": 1280
},
{
"epoch": 0.5128,
"grad_norm": 2.233239265114843,
"learning_rate": 5.647439411049234e-08,
"loss": -0.0278,
"step": 1282
},
{
"epoch": 0.5136,
"grad_norm": 6.244695059810562,
"learning_rate": 5.6335918219502735e-08,
"loss": -0.6849,
"step": 1284
},
{
"epoch": 0.5144,
"grad_norm": 5.213764846569251,
"learning_rate": 5.6197392919750087e-08,
"loss": 0.1645,
"step": 1286
},
{
"epoch": 0.5152,
"grad_norm": 1.6884839847809912,
"learning_rate": 5.605881929148253e-08,
"loss": 0.1449,
"step": 1288
},
{
"epoch": 0.516,
"grad_norm": 4.038680184005945,
"learning_rate": 5.592019841532506e-08,
"loss": -0.6065,
"step": 1290
},
{
"epoch": 0.5168,
"grad_norm": 0.9467192350608314,
"learning_rate": 5.578153137227108e-08,
"loss": 0.0454,
"step": 1292
},
{
"epoch": 0.5176,
"grad_norm": 1.9828901866201523,
"learning_rate": 5.564281924367408e-08,
"loss": 0.1477,
"step": 1294
},
{
"epoch": 0.5184,
"grad_norm": 0.7766424541672694,
"learning_rate": 5.5504063111239105e-08,
"loss": -0.1194,
"step": 1296
},
{
"epoch": 0.5192,
"grad_norm": 2.870243408385356,
"learning_rate": 5.536526405701433e-08,
"loss": -0.0621,
"step": 1298
},
{
"epoch": 0.52,
"grad_norm": 4.470988805451735,
"learning_rate": 5.5226423163382677e-08,
"loss": -0.2258,
"step": 1300
},
{
"epoch": 0.5208,
"grad_norm": 1.6523567908201209,
"learning_rate": 5.5087541513053315e-08,
"loss": 0.468,
"step": 1302
},
{
"epoch": 0.5216,
"grad_norm": 8.354761935132903,
"learning_rate": 5.494862018905325e-08,
"loss": -0.6695,
"step": 1304
},
{
"epoch": 0.5224,
"grad_norm": 1.0600003214166593,
"learning_rate": 5.480966027471888e-08,
"loss": 0.1322,
"step": 1306
},
{
"epoch": 0.5232,
"grad_norm": 1.447189011032088,
"learning_rate": 5.4670662853687534e-08,
"loss": 0.3698,
"step": 1308
},
{
"epoch": 0.524,
"grad_norm": 24.626486600138083,
"learning_rate": 5.4531629009889016e-08,
"loss": -0.9646,
"step": 1310
},
{
"epoch": 0.5248,
"grad_norm": 1.347665993514397,
"learning_rate": 5.4392559827537165e-08,
"loss": -0.3711,
"step": 1312
},
{
"epoch": 0.5256,
"grad_norm": 1.4919349101226638,
"learning_rate": 5.42534563911214e-08,
"loss": 0.0689,
"step": 1314
},
{
"epoch": 0.5264,
"grad_norm": 2.758769622371056,
"learning_rate": 5.4114319785398274e-08,
"loss": -0.8156,
"step": 1316
},
{
"epoch": 0.5272,
"grad_norm": 1.6889051720367219,
"learning_rate": 5.397515109538299e-08,
"loss": -0.0759,
"step": 1318
},
{
"epoch": 0.528,
"grad_norm": 1.1002825695875742,
"learning_rate": 5.383595140634093e-08,
"loss": -0.2669,
"step": 1320
},
{
"epoch": 0.5288,
"grad_norm": 2.659963420284105,
"learning_rate": 5.369672180377926e-08,
"loss": 0.5772,
"step": 1322
},
{
"epoch": 0.5296,
"grad_norm": 1.2855668675869305,
"learning_rate": 5.355746337343835e-08,
"loss": -0.5896,
"step": 1324
},
{
"epoch": 0.5304,
"grad_norm": 0.5835664795498647,
"learning_rate": 5.341817720128343e-08,
"loss": 0.0908,
"step": 1326
},
{
"epoch": 0.5312,
"grad_norm": 3.3942164266943045,
"learning_rate": 5.327886437349608e-08,
"loss": -0.2154,
"step": 1328
},
{
"epoch": 0.532,
"grad_norm": 3.1474290472029947,
"learning_rate": 5.313952597646567e-08,
"loss": -0.2349,
"step": 1330
},
{
"epoch": 0.5328,
"grad_norm": 2.2807680960100885,
"learning_rate": 5.300016309678104e-08,
"loss": -0.295,
"step": 1332
},
{
"epoch": 0.5336,
"grad_norm": 16.907562511619666,
"learning_rate": 5.286077682122191e-08,
"loss": -0.3385,
"step": 1334
},
{
"epoch": 0.5344,
"grad_norm": 1.5533490919001942,
"learning_rate": 5.272136823675045e-08,
"loss": 0.1332,
"step": 1336
},
{
"epoch": 0.5352,
"grad_norm": 8.807183927730819,
"learning_rate": 5.258193843050283e-08,
"loss": -0.1528,
"step": 1338
},
{
"epoch": 0.536,
"grad_norm": 1.4862400548123054,
"learning_rate": 5.2442488489780666e-08,
"loss": 0.1324,
"step": 1340
},
{
"epoch": 0.5368,
"grad_norm": 2.753596760613398,
"learning_rate": 5.230301950204261e-08,
"loss": 0.2668,
"step": 1342
},
{
"epoch": 0.5376,
"grad_norm": 1.3967208213678357,
"learning_rate": 5.216353255489585e-08,
"loss": 0.1049,
"step": 1344
},
{
"epoch": 0.5384,
"grad_norm": 3.5485496578024294,
"learning_rate": 5.2024028736087624e-08,
"loss": 0.3956,
"step": 1346
},
{
"epoch": 0.5392,
"grad_norm": 10.383666207219719,
"learning_rate": 5.188450913349674e-08,
"loss": -0.5953,
"step": 1348
},
{
"epoch": 0.54,
"grad_norm": 1.361423830248357,
"learning_rate": 5.1744974835125056e-08,
"loss": 0.3575,
"step": 1350
},
{
"epoch": 0.5408,
"grad_norm": 1.3762037764186998,
"learning_rate": 5.160542692908908e-08,
"loss": 0.0923,
"step": 1352
},
{
"epoch": 0.5416,
"grad_norm": 0.7036483104762083,
"learning_rate": 5.146586650361142e-08,
"loss": 0.0543,
"step": 1354
},
{
"epoch": 0.5424,
"grad_norm": 1.7035046490079102,
"learning_rate": 5.132629464701229e-08,
"loss": 0.594,
"step": 1356
},
{
"epoch": 0.5432,
"grad_norm": 0.6577494292524566,
"learning_rate": 5.11867124477011e-08,
"loss": -0.4076,
"step": 1358
},
{
"epoch": 0.544,
"grad_norm": 1.705082611585673,
"learning_rate": 5.104712099416785e-08,
"loss": 0.0251,
"step": 1360
},
{
"epoch": 0.5448,
"grad_norm": 3.689442768358194,
"learning_rate": 5.090752137497474e-08,
"loss": 0.3629,
"step": 1362
},
{
"epoch": 0.5456,
"grad_norm": 2.0580858641655766,
"learning_rate": 5.0767914678747645e-08,
"loss": -0.6431,
"step": 1364
},
{
"epoch": 0.5464,
"grad_norm": 4.358761098082887,
"learning_rate": 5.0628301994167634e-08,
"loss": -0.1779,
"step": 1366
},
{
"epoch": 0.5472,
"grad_norm": 2.246896501191125,
"learning_rate": 5.048868440996246e-08,
"loss": 0.1894,
"step": 1368
},
{
"epoch": 0.548,
"grad_norm": 2.3402721079385125,
"learning_rate": 5.034906301489807e-08,
"loss": 0.2495,
"step": 1370
},
{
"epoch": 0.5488,
"grad_norm": 1.6112023177078798,
"learning_rate": 5.02094388977702e-08,
"loss": 0.4752,
"step": 1372
},
{
"epoch": 0.5496,
"grad_norm": 1.0857915819905197,
"learning_rate": 5.0069813147395725e-08,
"loss": 0.0583,
"step": 1374
},
{
"epoch": 0.5504,
"grad_norm": 1.3169352802028227,
"learning_rate": 4.993018685260428e-08,
"loss": 0.1941,
"step": 1376
},
{
"epoch": 0.5512,
"grad_norm": 8.45210897957043,
"learning_rate": 4.979056110222981e-08,
"loss": 0.2704,
"step": 1378
},
{
"epoch": 0.552,
"grad_norm": 2.195741227779714,
"learning_rate": 4.9650936985101923e-08,
"loss": -0.2134,
"step": 1380
},
{
"epoch": 0.5528,
"grad_norm": 6.579583911683551,
"learning_rate": 4.9511315590037556e-08,
"loss": 0.4378,
"step": 1382
},
{
"epoch": 0.5536,
"grad_norm": 5.730838024543505,
"learning_rate": 4.937169800583236e-08,
"loss": -0.9895,
"step": 1384
},
{
"epoch": 0.5544,
"grad_norm": 2.261790288927768,
"learning_rate": 4.923208532125235e-08,
"loss": 0.3107,
"step": 1386
},
{
"epoch": 0.5552,
"grad_norm": 3.5203869215668147,
"learning_rate": 4.909247862502526e-08,
"loss": -0.0657,
"step": 1388
},
{
"epoch": 0.556,
"grad_norm": 1.1765877795788182,
"learning_rate": 4.895287900583215e-08,
"loss": 0.3294,
"step": 1390
},
{
"epoch": 0.5568,
"grad_norm": 6.610541172365496,
"learning_rate": 4.881328755229891e-08,
"loss": 0.8219,
"step": 1392
},
{
"epoch": 0.5576,
"grad_norm": 2.047741288861338,
"learning_rate": 4.867370535298769e-08,
"loss": -0.1988,
"step": 1394
},
{
"epoch": 0.5584,
"grad_norm": 1.5984989104657576,
"learning_rate": 4.8534133496388587e-08,
"loss": 0.1535,
"step": 1396
},
{
"epoch": 0.5592,
"grad_norm": 2.9232487652209476,
"learning_rate": 4.839457307091092e-08,
"loss": -0.6327,
"step": 1398
},
{
"epoch": 0.56,
"grad_norm": 6.6685003828240825,
"learning_rate": 4.8255025164874966e-08,
"loss": -0.2491,
"step": 1400
},
{
"epoch": 0.5608,
"grad_norm": 2.122107355750566,
"learning_rate": 4.8115490866503265e-08,
"loss": 0.3453,
"step": 1402
},
{
"epoch": 0.5616,
"grad_norm": 1.9370580784349283,
"learning_rate": 4.797597126391238e-08,
"loss": -0.1718,
"step": 1404
},
{
"epoch": 0.5624,
"grad_norm": 1.1675280554254117,
"learning_rate": 4.783646744510415e-08,
"loss": 0.3293,
"step": 1406
},
{
"epoch": 0.5632,
"grad_norm": 1.3803962338075701,
"learning_rate": 4.7696980497957376e-08,
"loss": 0.0989,
"step": 1408
},
{
"epoch": 0.564,
"grad_norm": 4.01537644874263,
"learning_rate": 4.7557511510219336e-08,
"loss": -0.3572,
"step": 1410
},
{
"epoch": 0.5648,
"grad_norm": 2.09156888701562,
"learning_rate": 4.741806156949717e-08,
"loss": -0.5134,
"step": 1412
},
{
"epoch": 0.5656,
"grad_norm": 1.1090034858220563,
"learning_rate": 4.727863176324955e-08,
"loss": -0.4258,
"step": 1414
},
{
"epoch": 0.5664,
"grad_norm": 1.561513006214317,
"learning_rate": 4.7139223178778094e-08,
"loss": 1.0116,
"step": 1416
},
{
"epoch": 0.5672,
"grad_norm": 2.3494087850012355,
"learning_rate": 4.6999836903218977e-08,
"loss": -0.3086,
"step": 1418
},
{
"epoch": 0.568,
"grad_norm": 3.018769501552223,
"learning_rate": 4.686047402353433e-08,
"loss": -0.2078,
"step": 1420
},
{
"epoch": 0.5688,
"grad_norm": 1.7120820897257356,
"learning_rate": 4.6721135626503934e-08,
"loss": -0.2402,
"step": 1422
},
{
"epoch": 0.5696,
"grad_norm": 3.19271086146731,
"learning_rate": 4.658182279871657e-08,
"loss": 0.1192,
"step": 1424
},
{
"epoch": 0.5704,
"grad_norm": 1.1452136953708933,
"learning_rate": 4.6442536626561665e-08,
"loss": -0.2687,
"step": 1426
},
{
"epoch": 0.5712,
"grad_norm": 0.8475452145849302,
"learning_rate": 4.630327819622075e-08,
"loss": 0.3452,
"step": 1428
},
{
"epoch": 0.572,
"grad_norm": 0.8234581334530525,
"learning_rate": 4.6164048593659066e-08,
"loss": -0.2424,
"step": 1430
},
{
"epoch": 0.5728,
"grad_norm": 2.4890263153905052,
"learning_rate": 4.6024848904617014e-08,
"loss": -0.9922,
"step": 1432
},
{
"epoch": 0.5736,
"grad_norm": 1.9063590684348184,
"learning_rate": 4.5885680214601715e-08,
"loss": 0.1718,
"step": 1434
},
{
"epoch": 0.5744,
"grad_norm": 1.8826789109483555,
"learning_rate": 4.57465436088786e-08,
"loss": 0.4459,
"step": 1436
},
{
"epoch": 0.5752,
"grad_norm": 2.2514983487036946,
"learning_rate": 4.5607440172462844e-08,
"loss": 0.0877,
"step": 1438
},
{
"epoch": 0.576,
"grad_norm": 4.746528942441705,
"learning_rate": 4.5468370990111e-08,
"loss": -0.3161,
"step": 1440
},
{
"epoch": 0.5768,
"grad_norm": 3.997914923728025,
"learning_rate": 4.5329337146312474e-08,
"loss": -0.3476,
"step": 1442
},
{
"epoch": 0.5776,
"grad_norm": 1.5223776751107094,
"learning_rate": 4.5190339725281135e-08,
"loss": 0.212,
"step": 1444
},
{
"epoch": 0.5784,
"grad_norm": 12.004095787528229,
"learning_rate": 4.505137981094675e-08,
"loss": 0.1348,
"step": 1446
},
{
"epoch": 0.5792,
"grad_norm": 2.6311737832063007,
"learning_rate": 4.491245848694669e-08,
"loss": -0.2547,
"step": 1448
},
{
"epoch": 0.58,
"grad_norm": 3.5026859584821617,
"learning_rate": 4.477357683661733e-08,
"loss": 0.4626,
"step": 1450
},
{
"epoch": 0.5808,
"grad_norm": 2.096013846352325,
"learning_rate": 4.463473594298566e-08,
"loss": 0.5435,
"step": 1452
},
{
"epoch": 0.5816,
"grad_norm": 0.870773168164334,
"learning_rate": 4.44959368887609e-08,
"loss": 0.1898,
"step": 1454
},
{
"epoch": 0.5824,
"grad_norm": 1.555370027932697,
"learning_rate": 4.435718075632591e-08,
"loss": 0.4965,
"step": 1456
},
{
"epoch": 0.5832,
"grad_norm": 1.3573063856986074,
"learning_rate": 4.421846862772893e-08,
"loss": 0.4763,
"step": 1458
},
{
"epoch": 0.584,
"grad_norm": 1.1211818311912025,
"learning_rate": 4.407980158467495e-08,
"loss": -0.1748,
"step": 1460
},
{
"epoch": 0.5848,
"grad_norm": 6.912907432379036,
"learning_rate": 4.3941180708517485e-08,
"loss": -0.269,
"step": 1462
},
{
"epoch": 0.5856,
"grad_norm": 8.017186323455453,
"learning_rate": 4.380260708024991e-08,
"loss": -0.0244,
"step": 1464
},
{
"epoch": 0.5864,
"grad_norm": 1.0752198808722395,
"learning_rate": 4.3664081780497274e-08,
"loss": 0.0872,
"step": 1466
},
{
"epoch": 0.5872,
"grad_norm": 2.4634723955812987,
"learning_rate": 4.352560588950766e-08,
"loss": 0.077,
"step": 1468
},
{
"epoch": 0.588,
"grad_norm": 2.981216805268154,
"learning_rate": 4.338718048714387e-08,
"loss": -0.8672,
"step": 1470
},
{
"epoch": 0.5888,
"grad_norm": 1.4875870951817871,
"learning_rate": 4.324880665287504e-08,
"loss": -0.6377,
"step": 1472
},
{
"epoch": 0.5896,
"grad_norm": 1.782623092796258,
"learning_rate": 4.31104854657681e-08,
"loss": -0.0933,
"step": 1474
},
{
"epoch": 0.5904,
"grad_norm": 0.8071293257332645,
"learning_rate": 4.2972218004479454e-08,
"loss": 0.352,
"step": 1476
},
{
"epoch": 0.5912,
"grad_norm": 0.769833398090707,
"learning_rate": 4.2834005347246527e-08,
"loss": 0.3886,
"step": 1478
},
{
"epoch": 0.592,
"grad_norm": 2.9776410320929094,
"learning_rate": 4.269584857187942e-08,
"loss": -0.2999,
"step": 1480
},
{
"epoch": 0.5928,
"grad_norm": 2.39213087733995,
"learning_rate": 4.2557748755752384e-08,
"loss": -0.0361,
"step": 1482
},
{
"epoch": 0.5936,
"grad_norm": 2.819987283948475,
"learning_rate": 4.2419706975795565e-08,
"loss": -0.4916,
"step": 1484
},
{
"epoch": 0.5944,
"grad_norm": 6.92777825106192,
"learning_rate": 4.228172430848644e-08,
"loss": -0.0633,
"step": 1486
},
{
"epoch": 0.5952,
"grad_norm": 1.4052822764386514,
"learning_rate": 4.214380182984163e-08,
"loss": 0.018,
"step": 1488
},
{
"epoch": 0.596,
"grad_norm": 3.8090615419941787,
"learning_rate": 4.200594061540826e-08,
"loss": 0.0335,
"step": 1490
},
{
"epoch": 0.5968,
"grad_norm": 1.6846374599762706,
"learning_rate": 4.186814174025582e-08,
"loss": -0.1758,
"step": 1492
},
{
"epoch": 0.5976,
"grad_norm": 1.072240401366043,
"learning_rate": 4.1730406278967615e-08,
"loss": 0.4449,
"step": 1494
},
{
"epoch": 0.5984,
"grad_norm": 10.451556251883762,
"learning_rate": 4.159273530563242e-08,
"loss": -0.4106,
"step": 1496
},
{
"epoch": 0.5992,
"grad_norm": 11.313841249734088,
"learning_rate": 4.145512989383617e-08,
"loss": -0.8206,
"step": 1498
},
{
"epoch": 0.6,
"grad_norm": 2.4207080559333343,
"learning_rate": 4.131759111665348e-08,
"loss": -0.2507,
"step": 1500
},
{
"epoch": 0.6008,
"grad_norm": 1.2087387002841612,
"learning_rate": 4.118012004663939e-08,
"loss": 0.1111,
"step": 1502
},
{
"epoch": 0.6016,
"grad_norm": 11.267790806667549,
"learning_rate": 4.1042717755820884e-08,
"loss": -0.74,
"step": 1504
},
{
"epoch": 0.6024,
"grad_norm": 1.578751129473022,
"learning_rate": 4.0905385315688665e-08,
"loss": -0.04,
"step": 1506
},
{
"epoch": 0.6032,
"grad_norm": 3.0840269712264425,
"learning_rate": 4.076812379718866e-08,
"loss": -0.031,
"step": 1508
},
{
"epoch": 0.604,
"grad_norm": 0.8881811712106842,
"learning_rate": 4.0630934270713755e-08,
"loss": 0.2177,
"step": 1510
},
{
"epoch": 0.6048,
"grad_norm": 0.4207527868210718,
"learning_rate": 4.04938178060955e-08,
"loss": -0.1827,
"step": 1512
},
{
"epoch": 0.6056,
"grad_norm": 2.082799290232294,
"learning_rate": 4.035677547259554e-08,
"loss": 0.3833,
"step": 1514
},
{
"epoch": 0.6064,
"grad_norm": 1.824044182709251,
"learning_rate": 4.0219808338897595e-08,
"loss": -0.3403,
"step": 1516
},
{
"epoch": 0.6072,
"grad_norm": 4.001405593088844,
"learning_rate": 4.008291747309884e-08,
"loss": -0.1168,
"step": 1518
},
{
"epoch": 0.608,
"grad_norm": 7.045037320579325,
"learning_rate": 3.9946103942701774e-08,
"loss": -0.7949,
"step": 1520
},
{
"epoch": 0.6088,
"grad_norm": 2.254675863007796,
"learning_rate": 3.980936881460576e-08,
"loss": 0.1668,
"step": 1522
},
{
"epoch": 0.6096,
"grad_norm": 1.212683615035007,
"learning_rate": 3.967271315509884e-08,
"loss": 0.3422,
"step": 1524
},
{
"epoch": 0.6104,
"grad_norm": 0.724005345416449,
"learning_rate": 3.9536138029849244e-08,
"loss": -0.3961,
"step": 1526
},
{
"epoch": 0.6112,
"grad_norm": 0.8678246145177999,
"learning_rate": 3.939964450389728e-08,
"loss": -0.1928,
"step": 1528
},
{
"epoch": 0.612,
"grad_norm": 2.191880510252536,
"learning_rate": 3.926323364164684e-08,
"loss": 0.6699,
"step": 1530
},
{
"epoch": 0.6128,
"grad_norm": 2.551242846475273,
"learning_rate": 3.912690650685726e-08,
"loss": 0.614,
"step": 1532
},
{
"epoch": 0.6136,
"grad_norm": 3.3125167161611127,
"learning_rate": 3.8990664162634925e-08,
"loss": 0.0522,
"step": 1534
},
{
"epoch": 0.6144,
"grad_norm": 1.2654840033977905,
"learning_rate": 3.8854507671424976e-08,
"loss": 0.2405,
"step": 1536
},
{
"epoch": 0.6152,
"grad_norm": 21.026384912783755,
"learning_rate": 3.8718438095003126e-08,
"loss": -0.8664,
"step": 1538
},
{
"epoch": 0.616,
"grad_norm": 2.0778940857004526,
"learning_rate": 3.858245649446721e-08,
"loss": 0.4077,
"step": 1540
},
{
"epoch": 0.6168,
"grad_norm": 1.6499639662053487,
"learning_rate": 3.8446563930229115e-08,
"loss": 0.5043,
"step": 1542
},
{
"epoch": 0.6176,
"grad_norm": 3.6786502697072185,
"learning_rate": 3.831076146200632e-08,
"loss": 0.498,
"step": 1544
},
{
"epoch": 0.6184,
"grad_norm": 5.321778159467506,
"learning_rate": 3.8175050148813775e-08,
"loss": -1.2537,
"step": 1546
},
{
"epoch": 0.6192,
"grad_norm": 2.1309052496847487,
"learning_rate": 3.8039431048955536e-08,
"loss": -0.221,
"step": 1548
},
{
"epoch": 0.62,
"grad_norm": 7.567379404853537,
"learning_rate": 3.7903905220016615e-08,
"loss": -0.3417,
"step": 1550
},
{
"epoch": 0.6208,
"grad_norm": 1.7798868128143455,
"learning_rate": 3.776847371885464e-08,
"loss": 0.0114,
"step": 1552
},
{
"epoch": 0.6216,
"grad_norm": 1.9507417381620744,
"learning_rate": 3.763313760159164e-08,
"loss": 0.3459,
"step": 1554
},
{
"epoch": 0.6224,
"grad_norm": 5.572740486405863,
"learning_rate": 3.749789792360589e-08,
"loss": -0.2116,
"step": 1556
},
{
"epoch": 0.6232,
"grad_norm": 2.7566473475035354,
"learning_rate": 3.7362755739523535e-08,
"loss": 0.1338,
"step": 1558
},
{
"epoch": 0.624,
"grad_norm": 1.580365481427382,
"learning_rate": 3.7227712103210477e-08,
"loss": 0.2273,
"step": 1560
},
{
"epoch": 0.6248,
"grad_norm": 1.3510627818996688,
"learning_rate": 3.709276806776412e-08,
"loss": 0.129,
"step": 1562
},
{
"epoch": 0.6256,
"grad_norm": 1.7556788740153348,
"learning_rate": 3.695792468550516e-08,
"loss": 0.1901,
"step": 1564
},
{
"epoch": 0.6264,
"grad_norm": 1.6324801438453431,
"learning_rate": 3.682318300796937e-08,
"loss": 0.1479,
"step": 1566
},
{
"epoch": 0.6272,
"grad_norm": 0.9547955178731168,
"learning_rate": 3.668854408589944e-08,
"loss": -0.7838,
"step": 1568
},
{
"epoch": 0.628,
"grad_norm": 2.8978051303938144,
"learning_rate": 3.6554008969236713e-08,
"loss": 0.6166,
"step": 1570
},
{
"epoch": 0.6288,
"grad_norm": 2.8743673901784264,
"learning_rate": 3.641957870711305e-08,
"loss": -0.436,
"step": 1572
},
{
"epoch": 0.6296,
"grad_norm": 2.503466521405713,
"learning_rate": 3.6285254347842674e-08,
"loss": 0.1542,
"step": 1574
},
{
"epoch": 0.6304,
"grad_norm": 3.2381022094414145,
"learning_rate": 3.615103693891388e-08,
"loss": 0.0577,
"step": 1576
},
{
"epoch": 0.6312,
"grad_norm": 1.8671448154056591,
"learning_rate": 3.601692752698101e-08,
"loss": 0.2242,
"step": 1578
},
{
"epoch": 0.632,
"grad_norm": 2.0522855223453926,
"learning_rate": 3.588292715785617e-08,
"loss": 0.4403,
"step": 1580
},
{
"epoch": 0.6328,
"grad_norm": 3.1330172236269536,
"learning_rate": 3.574903687650119e-08,
"loss": 0.0543,
"step": 1582
},
{
"epoch": 0.6336,
"grad_norm": 1.3225626787576295,
"learning_rate": 3.561525772701937e-08,
"loss": 0.4842,
"step": 1584
},
{
"epoch": 0.6344,
"grad_norm": 2.6414224390304817,
"learning_rate": 3.548159075264738e-08,
"loss": 0.2436,
"step": 1586
},
{
"epoch": 0.6352,
"grad_norm": 1.7396874661159625,
"learning_rate": 3.5348036995747135e-08,
"loss": 0.0321,
"step": 1588
},
{
"epoch": 0.636,
"grad_norm": 13.844859220247393,
"learning_rate": 3.5214597497797685e-08,
"loss": -0.244,
"step": 1590
},
{
"epoch": 0.6368,
"grad_norm": 4.157937629367746,
"learning_rate": 3.508127329938699e-08,
"loss": -0.6609,
"step": 1592
},
{
"epoch": 0.6376,
"grad_norm": 1.3790702379267534,
"learning_rate": 3.4948065440203976e-08,
"loss": -0.2822,
"step": 1594
},
{
"epoch": 0.6384,
"grad_norm": 1.5374452328347417,
"learning_rate": 3.481497495903029e-08,
"loss": 0.1825,
"step": 1596
},
{
"epoch": 0.6392,
"grad_norm": 3.5241161870744406,
"learning_rate": 3.4682002893732196e-08,
"loss": -0.2878,
"step": 1598
},
{
"epoch": 0.64,
"grad_norm": 3.8388338807250255,
"learning_rate": 3.4549150281252633e-08,
"loss": -0.2252,
"step": 1600
},
{
"epoch": 0.6408,
"grad_norm": 1.3127426977863277,
"learning_rate": 3.4416418157602904e-08,
"loss": 0.4729,
"step": 1602
},
{
"epoch": 0.6416,
"grad_norm": 2.3638705643343987,
"learning_rate": 3.428380755785481e-08,
"loss": -0.1709,
"step": 1604
},
{
"epoch": 0.6424,
"grad_norm": 1.8145811980021247,
"learning_rate": 3.415131951613241e-08,
"loss": -0.2825,
"step": 1606
},
{
"epoch": 0.6432,
"grad_norm": 2.4403191054850026,
"learning_rate": 3.40189550656041e-08,
"loss": -0.955,
"step": 1608
},
{
"epoch": 0.644,
"grad_norm": 2.397240952582479,
"learning_rate": 3.388671523847445e-08,
"loss": 0.099,
"step": 1610
},
{
"epoch": 0.6448,
"grad_norm": 1.6254363718545937,
"learning_rate": 3.3754601065976186e-08,
"loss": -0.12,
"step": 1612
},
{
"epoch": 0.6456,
"grad_norm": 2.786169171848472,
"learning_rate": 3.362261357836216e-08,
"loss": 0.0492,
"step": 1614
},
{
"epoch": 0.6464,
"grad_norm": 6.170904881710663,
"learning_rate": 3.349075380489731e-08,
"loss": -0.1465,
"step": 1616
},
{
"epoch": 0.6472,
"grad_norm": 1.0165447336675035,
"learning_rate": 3.335902277385067e-08,
"loss": -0.452,
"step": 1618
},
{
"epoch": 0.648,
"grad_norm": 1.7328957888882133,
"learning_rate": 3.3227421512487255e-08,
"loss": 0.0252,
"step": 1620
},
{
"epoch": 0.6488,
"grad_norm": 1.6122988397073823,
"learning_rate": 3.3095951047060146e-08,
"loss": 0.3178,
"step": 1622
},
{
"epoch": 0.6496,
"grad_norm": 3.0074140803632896,
"learning_rate": 3.2964612402802415e-08,
"loss": -0.3556,
"step": 1624
},
{
"epoch": 0.6504,
"grad_norm": 3.4523497524991344,
"learning_rate": 3.283340660391924e-08,
"loss": -0.1988,
"step": 1626
},
{
"epoch": 0.6512,
"grad_norm": 6.280464925298079,
"learning_rate": 3.270233467357976e-08,
"loss": -0.2276,
"step": 1628
},
{
"epoch": 0.652,
"grad_norm": 5.443046138761929,
"learning_rate": 3.2571397633909245e-08,
"loss": -0.2363,
"step": 1630
},
{
"epoch": 0.6528,
"grad_norm": 6.722447896835382,
"learning_rate": 3.2440596505981e-08,
"loss": -0.1321,
"step": 1632
},
{
"epoch": 0.6536,
"grad_norm": 1.7144203731513765,
"learning_rate": 3.230993230980852e-08,
"loss": 0.3665,
"step": 1634
},
{
"epoch": 0.6544,
"grad_norm": 4.593149677505712,
"learning_rate": 3.217940606433747e-08,
"loss": -0.4483,
"step": 1636
},
{
"epoch": 0.6552,
"grad_norm": 3.078404716919227,
"learning_rate": 3.2049018787437685e-08,
"loss": 0.2563,
"step": 1638
},
{
"epoch": 0.656,
"grad_norm": 0.4005810169418313,
"learning_rate": 3.191877149589539e-08,
"loss": -1.0674,
"step": 1640
},
{
"epoch": 0.6568,
"grad_norm": 3.253000576185386,
"learning_rate": 3.178866520540508e-08,
"loss": 0.2512,
"step": 1642
},
{
"epoch": 0.6576,
"grad_norm": 1.3405373696289142,
"learning_rate": 3.16587009305618e-08,
"loss": 0.2423,
"step": 1644
},
{
"epoch": 0.6584,
"grad_norm": 3.0107159967656805,
"learning_rate": 3.1528879684853026e-08,
"loss": 0.3202,
"step": 1646
},
{
"epoch": 0.6592,
"grad_norm": 1.133978959437793,
"learning_rate": 3.1399202480650944e-08,
"loss": -0.2006,
"step": 1648
},
{
"epoch": 0.66,
"grad_norm": 0.8674875540781807,
"learning_rate": 3.126967032920439e-08,
"loss": 0.2478,
"step": 1650
},
{
"epoch": 0.6608,
"grad_norm": 4.877486225676917,
"learning_rate": 3.114028424063118e-08,
"loss": 0.8045,
"step": 1652
},
{
"epoch": 0.6616,
"grad_norm": 6.668255903811672,
"learning_rate": 3.101104522390995e-08,
"loss": 0.0588,
"step": 1654
},
{
"epoch": 0.6624,
"grad_norm": 4.225599344016223,
"learning_rate": 3.088195428687253e-08,
"loss": -0.4538,
"step": 1656
},
{
"epoch": 0.6632,
"grad_norm": 3.2904442215649268,
"learning_rate": 3.075301243619603e-08,
"loss": -0.8012,
"step": 1658
},
{
"epoch": 0.664,
"grad_norm": 1.2097594409637307,
"learning_rate": 3.0624220677394854e-08,
"loss": 0.2129,
"step": 1660
},
{
"epoch": 0.6648,
"grad_norm": 1.2181710897816944,
"learning_rate": 3.0495580014813014e-08,
"loss": 0.1647,
"step": 1662
},
{
"epoch": 0.6656,
"grad_norm": 2.1392759816544125,
"learning_rate": 3.036709145161625e-08,
"loss": 0.0828,
"step": 1664
},
{
"epoch": 0.6664,
"grad_norm": 2.1928200746586746,
"learning_rate": 3.0238755989784186e-08,
"loss": 0.337,
"step": 1666
},
{
"epoch": 0.6672,
"grad_norm": 2.839822192787894,
"learning_rate": 3.0110574630102516e-08,
"loss": 0.2746,
"step": 1668
},
{
"epoch": 0.668,
"grad_norm": 2.3747566081007414,
"learning_rate": 2.998254837215526e-08,
"loss": 0.2913,
"step": 1670
},
{
"epoch": 0.6688,
"grad_norm": 1.0576318241526304,
"learning_rate": 2.985467821431687e-08,
"loss": 0.445,
"step": 1672
},
{
"epoch": 0.6696,
"grad_norm": 5.321277022454464,
"learning_rate": 2.9726965153744547e-08,
"loss": -0.7301,
"step": 1674
},
{
"epoch": 0.6704,
"grad_norm": 1.6420958574076474,
"learning_rate": 2.959941018637036e-08,
"loss": 0.3301,
"step": 1676
},
{
"epoch": 0.6712,
"grad_norm": 6.762985212922795,
"learning_rate": 2.94720143068936e-08,
"loss": -0.5514,
"step": 1678
},
{
"epoch": 0.672,
"grad_norm": 5.343972437244011,
"learning_rate": 2.9344778508772915e-08,
"loss": -0.7603,
"step": 1680
},
{
"epoch": 0.6728,
"grad_norm": 1.4173868345663152,
"learning_rate": 2.9217703784218607e-08,
"loss": -0.2702,
"step": 1682
},
{
"epoch": 0.6736,
"grad_norm": 2.408297080132785,
"learning_rate": 2.909079112418493e-08,
"loss": 0.239,
"step": 1684
},
{
"epoch": 0.6744,
"grad_norm": 2.526003439412088,
"learning_rate": 2.896404151836227e-08,
"loss": -0.1545,
"step": 1686
},
{
"epoch": 0.6752,
"grad_norm": 2.152154481783791,
"learning_rate": 2.8837455955169542e-08,
"loss": 0.2247,
"step": 1688
},
{
"epoch": 0.676,
"grad_norm": 4.841731380516718,
"learning_rate": 2.8711035421746366e-08,
"loss": -0.2435,
"step": 1690
},
{
"epoch": 0.6768,
"grad_norm": 1.4440262452483728,
"learning_rate": 2.8584780903945488e-08,
"loss": 0.1137,
"step": 1692
},
{
"epoch": 0.6776,
"grad_norm": 1.058358316472121,
"learning_rate": 2.8458693386324996e-08,
"loss": -0.0944,
"step": 1694
},
{
"epoch": 0.6784,
"grad_norm": 9.090390939902395,
"learning_rate": 2.833277385214064e-08,
"loss": 0.6503,
"step": 1696
},
{
"epoch": 0.6792,
"grad_norm": 7.788906709654453,
"learning_rate": 2.8207023283338298e-08,
"loss": 0.483,
"step": 1698
},
{
"epoch": 0.68,
"grad_norm": 2.497229006444577,
"learning_rate": 2.8081442660546124e-08,
"loss": -0.3163,
"step": 1700
},
{
"epoch": 0.6808,
"grad_norm": 1.245945203078446,
"learning_rate": 2.7956032963067077e-08,
"loss": 0.1814,
"step": 1702
},
{
"epoch": 0.6816,
"grad_norm": 1.5957075071574214,
"learning_rate": 2.7830795168871123e-08,
"loss": 0.1276,
"step": 1704
},
{
"epoch": 0.6824,
"grad_norm": 0.912369643362412,
"learning_rate": 2.77057302545878e-08,
"loss": 0.0809,
"step": 1706
},
{
"epoch": 0.6832,
"grad_norm": 12.398334532058389,
"learning_rate": 2.7580839195498396e-08,
"loss": 0.6571,
"step": 1708
},
{
"epoch": 0.684,
"grad_norm": 1.8705294347889838,
"learning_rate": 2.7456122965528473e-08,
"loss": -0.0012,
"step": 1710
},
{
"epoch": 0.6848,
"grad_norm": 8.43780230171603,
"learning_rate": 2.7331582537240237e-08,
"loss": 0.5914,
"step": 1712
},
{
"epoch": 0.6856,
"grad_norm": 6.045183963191954,
"learning_rate": 2.720721888182501e-08,
"loss": 0.2553,
"step": 1714
},
{
"epoch": 0.6864,
"grad_norm": 10.933471488223638,
"learning_rate": 2.7083032969095502e-08,
"loss": -0.1032,
"step": 1716
},
{
"epoch": 0.6872,
"grad_norm": 6.965055302019932,
"learning_rate": 2.695902576747846e-08,
"loss": -0.526,
"step": 1718
},
{
"epoch": 0.688,
"grad_norm": 8.512965363552777,
"learning_rate": 2.6835198244006924e-08,
"loss": -0.8081,
"step": 1720
},
{
"epoch": 0.6888,
"grad_norm": 1.192736035070821,
"learning_rate": 2.6711551364312785e-08,
"loss": -0.0847,
"step": 1722
},
{
"epoch": 0.6896,
"grad_norm": 1.6765610057391003,
"learning_rate": 2.6588086092619278e-08,
"loss": -0.27,
"step": 1724
},
{
"epoch": 0.6904,
"grad_norm": 1.2188199326627063,
"learning_rate": 2.646480339173337e-08,
"loss": 0.1273,
"step": 1726
},
{
"epoch": 0.6912,
"grad_norm": 1.8123646203452417,
"learning_rate": 2.6341704223038347e-08,
"loss": 0.2395,
"step": 1728
},
{
"epoch": 0.692,
"grad_norm": 4.3649576952135485,
"learning_rate": 2.6218789546486232e-08,
"loss": -0.7106,
"step": 1730
},
{
"epoch": 0.6928,
"grad_norm": 4.342633194501302,
"learning_rate": 2.609606032059039e-08,
"loss": -0.5549,
"step": 1732
},
{
"epoch": 0.6936,
"grad_norm": 1.2870887711942558,
"learning_rate": 2.5973517502417963e-08,
"loss": 0.0922,
"step": 1734
},
{
"epoch": 0.6944,
"grad_norm": 8.857731444056965,
"learning_rate": 2.5851162047582476e-08,
"loss": -1.2119,
"step": 1736
},
{
"epoch": 0.6952,
"grad_norm": 0.9514384344774223,
"learning_rate": 2.5728994910236302e-08,
"loss": 0.111,
"step": 1738
},
{
"epoch": 0.696,
"grad_norm": 11.6142955162987,
"learning_rate": 2.5607017043063355e-08,
"loss": 0.415,
"step": 1740
},
{
"epoch": 0.6968,
"grad_norm": 4.579542574808095,
"learning_rate": 2.548522939727156e-08,
"loss": 0.3388,
"step": 1742
},
{
"epoch": 0.6976,
"grad_norm": 1.4400831949708577,
"learning_rate": 2.5363632922585427e-08,
"loss": -0.0981,
"step": 1744
},
{
"epoch": 0.6984,
"grad_norm": 2.087009445621481,
"learning_rate": 2.5242228567238687e-08,
"loss": 0.2813,
"step": 1746
},
{
"epoch": 0.6992,
"grad_norm": 2.6328146915202986,
"learning_rate": 2.512101727796687e-08,
"loss": 0.1758,
"step": 1748
},
{
"epoch": 0.7,
"grad_norm": 7.056808040316399,
"learning_rate": 2.500000000000001e-08,
"loss": 0.3373,
"step": 1750
},
{
"epoch": 0.7008,
"grad_norm": 1.0437467398666396,
"learning_rate": 2.4879177677055098e-08,
"loss": 0.099,
"step": 1752
},
{
"epoch": 0.7016,
"grad_norm": 3.8414096686764676,
"learning_rate": 2.475855125132892e-08,
"loss": -0.2663,
"step": 1754
},
{
"epoch": 0.7024,
"grad_norm": 2.439424508673481,
"learning_rate": 2.463812166349054e-08,
"loss": 0.2117,
"step": 1756
},
{
"epoch": 0.7032,
"grad_norm": 2.387800089257116,
"learning_rate": 2.4517889852674113e-08,
"loss": -0.0708,
"step": 1758
},
{
"epoch": 0.704,
"grad_norm": 1.3000333539494446,
"learning_rate": 2.439785675647143e-08,
"loss": 0.1515,
"step": 1760
},
{
"epoch": 0.7048,
"grad_norm": 2.198671971396822,
"learning_rate": 2.4278023310924674e-08,
"loss": 0.065,
"step": 1762
},
{
"epoch": 0.7056,
"grad_norm": 17.120279372145863,
"learning_rate": 2.4158390450519155e-08,
"loss": -0.0036,
"step": 1764
},
{
"epoch": 0.7064,
"grad_norm": 2.324948949045274,
"learning_rate": 2.4038959108175926e-08,
"loss": 0.1964,
"step": 1766
},
{
"epoch": 0.7072,
"grad_norm": 3.109048925076997,
"learning_rate": 2.391973021524461e-08,
"loss": -0.0786,
"step": 1768
},
{
"epoch": 0.708,
"grad_norm": 1.653300372155837,
"learning_rate": 2.3800704701496048e-08,
"loss": -0.0222,
"step": 1770
},
{
"epoch": 0.7088,
"grad_norm": 1.2147631669738923,
"learning_rate": 2.3681883495115112e-08,
"loss": 0.6688,
"step": 1772
},
{
"epoch": 0.7096,
"grad_norm": 1.0924259194845596,
"learning_rate": 2.3563267522693414e-08,
"loss": -0.0123,
"step": 1774
},
{
"epoch": 0.7104,
"grad_norm": 0.4425561025986487,
"learning_rate": 2.3444857709222176e-08,
"loss": -0.065,
"step": 1776
},
{
"epoch": 0.7112,
"grad_norm": 0.7572768433247227,
"learning_rate": 2.332665497808487e-08,
"loss": -0.1034,
"step": 1778
},
{
"epoch": 0.712,
"grad_norm": 1.9702052124787697,
"learning_rate": 2.3208660251050153e-08,
"loss": 0.1152,
"step": 1780
},
{
"epoch": 0.7128,
"grad_norm": 2.562823426275877,
"learning_rate": 2.3090874448264637e-08,
"loss": -0.3823,
"step": 1782
},
{
"epoch": 0.7136,
"grad_norm": 2.7558169515582054,
"learning_rate": 2.2973298488245645e-08,
"loss": 0.1862,
"step": 1784
},
{
"epoch": 0.7144,
"grad_norm": 0.9205424687808403,
"learning_rate": 2.2855933287874134e-08,
"loss": -0.2344,
"step": 1786
},
{
"epoch": 0.7152,
"grad_norm": 4.049891733614343,
"learning_rate": 2.2738779762387494e-08,
"loss": 0.5732,
"step": 1788
},
{
"epoch": 0.716,
"grad_norm": 5.109297282246342,
"learning_rate": 2.2621838825372492e-08,
"loss": 0.1657,
"step": 1790
},
{
"epoch": 0.7168,
"grad_norm": 3.492531909450355,
"learning_rate": 2.2505111388758008e-08,
"loss": 0.5486,
"step": 1792
},
{
"epoch": 0.7176,
"grad_norm": 1.7475753446352464,
"learning_rate": 2.238859836280807e-08,
"loss": 0.4643,
"step": 1794
},
{
"epoch": 0.7184,
"grad_norm": 0.361925949755798,
"learning_rate": 2.2272300656114645e-08,
"loss": 0.1032,
"step": 1796
},
{
"epoch": 0.7192,
"grad_norm": 3.604668638859301,
"learning_rate": 2.215621917559062e-08,
"loss": 0.0287,
"step": 1798
},
{
"epoch": 0.72,
"grad_norm": 1.4617719449367537,
"learning_rate": 2.2040354826462664e-08,
"loss": 0.0297,
"step": 1800
},
{
"epoch": 0.7208,
"grad_norm": 3.620213951707984,
"learning_rate": 2.1924708512264277e-08,
"loss": 0.1951,
"step": 1802
},
{
"epoch": 0.7216,
"grad_norm": 0.7583235187586578,
"learning_rate": 2.180928113482866e-08,
"loss": -0.1489,
"step": 1804
},
{
"epoch": 0.7224,
"grad_norm": 2.119491033850218,
"learning_rate": 2.169407359428166e-08,
"loss": -0.2957,
"step": 1806
},
{
"epoch": 0.7232,
"grad_norm": 1.162852312754879,
"learning_rate": 2.1579086789034866e-08,
"loss": -0.2817,
"step": 1808
},
{
"epoch": 0.724,
"grad_norm": 1.166771156161991,
"learning_rate": 2.146432161577842e-08,
"loss": 0.1943,
"step": 1810
},
{
"epoch": 0.7248,
"grad_norm": 2.0229771272952326,
"learning_rate": 2.1349778969474248e-08,
"loss": -0.0322,
"step": 1812
},
{
"epoch": 0.7256,
"grad_norm": 0.5930882949036291,
"learning_rate": 2.123545974334887e-08,
"loss": 0.1233,
"step": 1814
},
{
"epoch": 0.7264,
"grad_norm": 5.29569204399418,
"learning_rate": 2.1121364828886627e-08,
"loss": -0.4993,
"step": 1816
},
{
"epoch": 0.7272,
"grad_norm": 2.929051905849708,
"learning_rate": 2.1007495115822537e-08,
"loss": 0.1715,
"step": 1818
},
{
"epoch": 0.728,
"grad_norm": 15.257361025978756,
"learning_rate": 2.0893851492135533e-08,
"loss": 0.2343,
"step": 1820
},
{
"epoch": 0.7288,
"grad_norm": 2.803107615477731,
"learning_rate": 2.07804348440414e-08,
"loss": 1.046,
"step": 1822
},
{
"epoch": 0.7296,
"grad_norm": 2.86449910527955,
"learning_rate": 2.0667246055985938e-08,
"loss": -0.0914,
"step": 1824
},
{
"epoch": 0.7304,
"grad_norm": 4.345282427711913,
"learning_rate": 2.0554286010638074e-08,
"loss": 0.1557,
"step": 1826
},
{
"epoch": 0.7312,
"grad_norm": 1.0680979217606996,
"learning_rate": 2.0441555588882897e-08,
"loss": 0.112,
"step": 1828
},
{
"epoch": 0.732,
"grad_norm": 22.35076669899949,
"learning_rate": 2.032905566981493e-08,
"loss": -0.7291,
"step": 1830
},
{
"epoch": 0.7328,
"grad_norm": 6.147459767039074,
"learning_rate": 2.02167871307311e-08,
"loss": 0.0367,
"step": 1832
},
{
"epoch": 0.7336,
"grad_norm": 0.9403821792893974,
"learning_rate": 2.0104750847124076e-08,
"loss": 0.1168,
"step": 1834
},
{
"epoch": 0.7344,
"grad_norm": 1.0648295993596697,
"learning_rate": 1.9992947692675227e-08,
"loss": -0.0064,
"step": 1836
},
{
"epoch": 0.7352,
"grad_norm": 6.834382123337444,
"learning_rate": 1.9881378539248077e-08,
"loss": -0.0209,
"step": 1838
},
{
"epoch": 0.736,
"grad_norm": 12.194299716833651,
"learning_rate": 1.977004425688126e-08,
"loss": -0.6629,
"step": 1840
},
{
"epoch": 0.7368,
"grad_norm": 1.6811402432981566,
"learning_rate": 1.9658945713781883e-08,
"loss": -0.3665,
"step": 1842
},
{
"epoch": 0.7376,
"grad_norm": 4.974628272901464,
"learning_rate": 1.9548083776318723e-08,
"loss": 0.3637,
"step": 1844
},
{
"epoch": 0.7384,
"grad_norm": 0.8026057709807365,
"learning_rate": 1.9437459309015425e-08,
"loss": -0.1781,
"step": 1846
},
{
"epoch": 0.7392,
"grad_norm": 0.9672604377243258,
"learning_rate": 1.93270731745438e-08,
"loss": 0.0373,
"step": 1848
},
{
"epoch": 0.74,
"grad_norm": 1.6981668123356704,
"learning_rate": 1.9216926233717085e-08,
"loss": -0.0772,
"step": 1850
},
{
"epoch": 0.7408,
"grad_norm": 4.5757533711944856,
"learning_rate": 1.910701934548329e-08,
"loss": 0.1371,
"step": 1852
},
{
"epoch": 0.7416,
"grad_norm": 6.385512164095145,
"learning_rate": 1.899735336691837e-08,
"loss": -0.3518,
"step": 1854
},
{
"epoch": 0.7424,
"grad_norm": 3.248106093810886,
"learning_rate": 1.8887929153219685e-08,
"loss": 0.2268,
"step": 1856
},
{
"epoch": 0.7432,
"grad_norm": 2.8252792830886575,
"learning_rate": 1.877874755769922e-08,
"loss": 0.1589,
"step": 1858
},
{
"epoch": 0.744,
"grad_norm": 5.264540846855634,
"learning_rate": 1.866980943177699e-08,
"loss": 0.0568,
"step": 1860
},
{
"epoch": 0.7448,
"grad_norm": 1.275428245866234,
"learning_rate": 1.8561115624974373e-08,
"loss": 0.1517,
"step": 1862
},
{
"epoch": 0.7456,
"grad_norm": 14.19208598391273,
"learning_rate": 1.8452666984907517e-08,
"loss": 0.2998,
"step": 1864
},
{
"epoch": 0.7464,
"grad_norm": 7.6814967219852655,
"learning_rate": 1.834446435728072e-08,
"loss": 0.4844,
"step": 1866
},
{
"epoch": 0.7472,
"grad_norm": 1.4193640099469234,
"learning_rate": 1.8236508585879777e-08,
"loss": 0.3173,
"step": 1868
},
{
"epoch": 0.748,
"grad_norm": 2.041627690513066,
"learning_rate": 1.8128800512565513e-08,
"loss": 0.3641,
"step": 1870
},
{
"epoch": 0.7488,
"grad_norm": 8.578127916608778,
"learning_rate": 1.80213409772671e-08,
"loss": -0.2812,
"step": 1872
},
{
"epoch": 0.7496,
"grad_norm": 9.164783253693107,
"learning_rate": 1.7914130817975594e-08,
"loss": -0.9407,
"step": 1874
},
{
"epoch": 0.7504,
"grad_norm": 3.716026385677171,
"learning_rate": 1.7807170870737316e-08,
"loss": -0.3692,
"step": 1876
},
{
"epoch": 0.7512,
"grad_norm": 1.6681777892033154,
"learning_rate": 1.770046196964747e-08,
"loss": 0.3245,
"step": 1878
},
{
"epoch": 0.752,
"grad_norm": 14.618819444480113,
"learning_rate": 1.7594004946843455e-08,
"loss": -0.1325,
"step": 1880
},
{
"epoch": 0.7528,
"grad_norm": 3.4870170103139935,
"learning_rate": 1.7487800632498543e-08,
"loss": 0.4117,
"step": 1882
},
{
"epoch": 0.7536,
"grad_norm": 2.1920712515819907,
"learning_rate": 1.7381849854815355e-08,
"loss": 0.4374,
"step": 1884
},
{
"epoch": 0.7544,
"grad_norm": 7.829348703682726,
"learning_rate": 1.7276153440019257e-08,
"loss": -0.3397,
"step": 1886
},
{
"epoch": 0.7552,
"grad_norm": 1.6655805866682127,
"learning_rate": 1.7170712212352185e-08,
"loss": -0.0982,
"step": 1888
},
{
"epoch": 0.756,
"grad_norm": 0.593137714722611,
"learning_rate": 1.7065526994065972e-08,
"loss": -0.1808,
"step": 1890
},
{
"epoch": 0.7568,
"grad_norm": 3.2668043762805024,
"learning_rate": 1.6960598605416114e-08,
"loss": -0.1903,
"step": 1892
},
{
"epoch": 0.7576,
"grad_norm": 5.3787240854624745,
"learning_rate": 1.6855927864655238e-08,
"loss": 0.0949,
"step": 1894
},
{
"epoch": 0.7584,
"grad_norm": 1.936247550747232,
"learning_rate": 1.6751515588026826e-08,
"loss": -0.3573,
"step": 1896
},
{
"epoch": 0.7592,
"grad_norm": 2.071377080812025,
"learning_rate": 1.6647362589758786e-08,
"loss": 0.1047,
"step": 1898
},
{
"epoch": 0.76,
"grad_norm": 6.789235393980865,
"learning_rate": 1.6543469682057103e-08,
"loss": 0.3172,
"step": 1900
},
{
"epoch": 0.7608,
"grad_norm": 5.474981449589401,
"learning_rate": 1.6439837675099537e-08,
"loss": -0.2288,
"step": 1902
},
{
"epoch": 0.7616,
"grad_norm": 3.199305250881023,
"learning_rate": 1.6336467377029305e-08,
"loss": -0.2315,
"step": 1904
},
{
"epoch": 0.7624,
"grad_norm": 2.482569440776884,
"learning_rate": 1.6233359593948776e-08,
"loss": -0.025,
"step": 1906
},
{
"epoch": 0.7632,
"grad_norm": 8.551827628823498,
"learning_rate": 1.613051512991314e-08,
"loss": -0.1971,
"step": 1908
},
{
"epoch": 0.764,
"grad_norm": 13.097052250880473,
"learning_rate": 1.6027934786924186e-08,
"loss": 0.166,
"step": 1910
},
{
"epoch": 0.7648,
"grad_norm": 1.4383545215228888,
"learning_rate": 1.5925619364924016e-08,
"loss": 0.3462,
"step": 1912
},
{
"epoch": 0.7656,
"grad_norm": 4.345257668964619,
"learning_rate": 1.5823569661788878e-08,
"loss": -0.4007,
"step": 1914
},
{
"epoch": 0.7664,
"grad_norm": 2.6275961887041532,
"learning_rate": 1.5721786473322823e-08,
"loss": 0.1353,
"step": 1916
},
{
"epoch": 0.7672,
"grad_norm": 1.1508314194699722,
"learning_rate": 1.5620270593251633e-08,
"loss": 0.171,
"step": 1918
},
{
"epoch": 0.768,
"grad_norm": 8.731880546216733,
"learning_rate": 1.551902281321651e-08,
"loss": 0.3329,
"step": 1920
},
{
"epoch": 0.7688,
"grad_norm": 1.9434870913531628,
"learning_rate": 1.5418043922768e-08,
"loss": 0.6301,
"step": 1922
},
{
"epoch": 0.7696,
"grad_norm": 0.9056332881493195,
"learning_rate": 1.5317334709359757e-08,
"loss": 0.1506,
"step": 1924
},
{
"epoch": 0.7704,
"grad_norm": 7.321492065454491,
"learning_rate": 1.5216895958342456e-08,
"loss": -0.0844,
"step": 1926
},
{
"epoch": 0.7712,
"grad_norm": 1.2676581635722497,
"learning_rate": 1.5116728452957683e-08,
"loss": 0.235,
"step": 1928
},
{
"epoch": 0.772,
"grad_norm": 6.770596765408979,
"learning_rate": 1.5016832974331723e-08,
"loss": -0.1452,
"step": 1930
},
{
"epoch": 0.7728,
"grad_norm": 0.774237776575187,
"learning_rate": 1.4917210301469628e-08,
"loss": -0.3171,
"step": 1932
},
{
"epoch": 0.7736,
"grad_norm": 2.358703253521802,
"learning_rate": 1.4817861211248994e-08,
"loss": -0.3403,
"step": 1934
},
{
"epoch": 0.7744,
"grad_norm": 2.091359277579041,
"learning_rate": 1.471878647841398e-08,
"loss": 0.3112,
"step": 1936
},
{
"epoch": 0.7752,
"grad_norm": 2.4844540414134952,
"learning_rate": 1.4619986875569246e-08,
"loss": 0.0934,
"step": 1938
},
{
"epoch": 0.776,
"grad_norm": 0.3358727391238236,
"learning_rate": 1.4521463173173964e-08,
"loss": -0.3601,
"step": 1940
},
{
"epoch": 0.7768,
"grad_norm": 2.8912433454858864,
"learning_rate": 1.4423216139535732e-08,
"loss": -0.2012,
"step": 1942
},
{
"epoch": 0.7776,
"grad_norm": 6.6239087581701135,
"learning_rate": 1.4325246540804669e-08,
"loss": -0.2919,
"step": 1944
},
{
"epoch": 0.7784,
"grad_norm": 1.874229260557601,
"learning_rate": 1.4227555140967401e-08,
"loss": 0.3964,
"step": 1946
},
{
"epoch": 0.7792,
"grad_norm": 3.4765126127575923,
"learning_rate": 1.4130142701841075e-08,
"loss": 0.0069,
"step": 1948
},
{
"epoch": 0.78,
"grad_norm": 0.9836616938537979,
"learning_rate": 1.4033009983067451e-08,
"loss": 0.1279,
"step": 1950
},
{
"epoch": 0.7808,
"grad_norm": 1.8450406813526212,
"learning_rate": 1.3936157742106974e-08,
"loss": -0.8598,
"step": 1952
},
{
"epoch": 0.7816,
"grad_norm": 1.1621228145442448,
"learning_rate": 1.3839586734232905e-08,
"loss": -0.3734,
"step": 1954
},
{
"epoch": 0.7824,
"grad_norm": 2.080888564116752,
"learning_rate": 1.3743297712525331e-08,
"loss": -0.2687,
"step": 1956
},
{
"epoch": 0.7832,
"grad_norm": 1.8724244373925105,
"learning_rate": 1.3647291427865415e-08,
"loss": -0.0416,
"step": 1958
},
{
"epoch": 0.784,
"grad_norm": 1.499292821299661,
"learning_rate": 1.3551568628929434e-08,
"loss": 0.211,
"step": 1960
},
{
"epoch": 0.7848,
"grad_norm": 2.0052361060709814,
"learning_rate": 1.3456130062183001e-08,
"loss": -0.0768,
"step": 1962
},
{
"epoch": 0.7856,
"grad_norm": 2.2666289421242523,
"learning_rate": 1.3360976471875224e-08,
"loss": 0.2792,
"step": 1964
},
{
"epoch": 0.7864,
"grad_norm": 2.723585168664323,
"learning_rate": 1.3266108600032928e-08,
"loss": -0.4011,
"step": 1966
},
{
"epoch": 0.7872,
"grad_norm": 1.7268778334192665,
"learning_rate": 1.317152718645484e-08,
"loss": 0.6589,
"step": 1968
},
{
"epoch": 0.788,
"grad_norm": 2.0199606733082254,
"learning_rate": 1.3077232968705804e-08,
"loss": 0.1737,
"step": 1970
},
{
"epoch": 0.7888,
"grad_norm": 4.804739415956699,
"learning_rate": 1.2983226682111093e-08,
"loss": -0.2499,
"step": 1972
},
{
"epoch": 0.7896,
"grad_norm": 2.4472702050288295,
"learning_rate": 1.2889509059750603e-08,
"loss": 0.0527,
"step": 1974
},
{
"epoch": 0.7904,
"grad_norm": 1.229092797828586,
"learning_rate": 1.2796080832453182e-08,
"loss": 0.0615,
"step": 1976
},
{
"epoch": 0.7912,
"grad_norm": 4.579918800198668,
"learning_rate": 1.2702942728790894e-08,
"loss": -0.3586,
"step": 1978
},
{
"epoch": 0.792,
"grad_norm": 0.901599843591529,
"learning_rate": 1.2610095475073413e-08,
"loss": 0.0544,
"step": 1980
},
{
"epoch": 0.7928,
"grad_norm": 1.4180499738504504,
"learning_rate": 1.2517539795342247e-08,
"loss": -0.3661,
"step": 1982
},
{
"epoch": 0.7936,
"grad_norm": 19.19099294718674,
"learning_rate": 1.24252764113652e-08,
"loss": 0.1257,
"step": 1984
},
{
"epoch": 0.7944,
"grad_norm": 2.976073756733255,
"learning_rate": 1.233330604263067e-08,
"loss": -0.5093,
"step": 1986
},
{
"epoch": 0.7952,
"grad_norm": 1.447954681140385,
"learning_rate": 1.2241629406342047e-08,
"loss": 0.4885,
"step": 1988
},
{
"epoch": 0.796,
"grad_norm": 5.02311224236895,
"learning_rate": 1.2150247217412186e-08,
"loss": 0.4777,
"step": 1990
},
{
"epoch": 0.7968,
"grad_norm": 1.3931834037300923,
"learning_rate": 1.2059160188457724e-08,
"loss": 0.1565,
"step": 1992
},
{
"epoch": 0.7976,
"grad_norm": 2.213190281008005,
"learning_rate": 1.196836902979364e-08,
"loss": -0.1925,
"step": 1994
},
{
"epoch": 0.7984,
"grad_norm": 3.692697293182129,
"learning_rate": 1.18778744494276e-08,
"loss": -0.3022,
"step": 1996
},
{
"epoch": 0.7992,
"grad_norm": 1.368876462797977,
"learning_rate": 1.1787677153054548e-08,
"loss": 0.3078,
"step": 1998
},
{
"epoch": 0.8,
"grad_norm": 10.774817453056974,
"learning_rate": 1.1697777844051105e-08,
"loss": 0.488,
"step": 2000
},
{
"epoch": 0.8008,
"grad_norm": 1.9545132619300545,
"learning_rate": 1.1608177223470139e-08,
"loss": 0.3029,
"step": 2002
},
{
"epoch": 0.8016,
"grad_norm": 1.806725462227303,
"learning_rate": 1.1518875990035277e-08,
"loss": -0.016,
"step": 2004
},
{
"epoch": 0.8024,
"grad_norm": 3.92077777933892,
"learning_rate": 1.142987484013549e-08,
"loss": 0.2598,
"step": 2006
},
{
"epoch": 0.8032,
"grad_norm": 1.1561570657627025,
"learning_rate": 1.1341174467819637e-08,
"loss": -0.216,
"step": 2008
},
{
"epoch": 0.804,
"grad_norm": 4.944528510946132,
"learning_rate": 1.1252775564791022e-08,
"loss": -0.1682,
"step": 2010
},
{
"epoch": 0.8048,
"grad_norm": 7.806682630259347,
"learning_rate": 1.1164678820402057e-08,
"loss": -0.2217,
"step": 2012
},
{
"epoch": 0.8056,
"grad_norm": 0.5906600478058831,
"learning_rate": 1.1076884921648832e-08,
"loss": -0.0115,
"step": 2014
},
{
"epoch": 0.8064,
"grad_norm": 1.5923039456841297,
"learning_rate": 1.0989394553165831e-08,
"loss": -0.4475,
"step": 2016
},
{
"epoch": 0.8072,
"grad_norm": 5.561765344754305,
"learning_rate": 1.0902208397220497e-08,
"loss": 0.2437,
"step": 2018
},
{
"epoch": 0.808,
"grad_norm": 0.8867433808948969,
"learning_rate": 1.0815327133708013e-08,
"loss": -0.0309,
"step": 2020
},
{
"epoch": 0.8088,
"grad_norm": 1.200813453561211,
"learning_rate": 1.0728751440145906e-08,
"loss": -0.3607,
"step": 2022
},
{
"epoch": 0.8096,
"grad_norm": 1.9415562238348962,
"learning_rate": 1.0642481991668839e-08,
"loss": -0.0327,
"step": 2024
},
{
"epoch": 0.8104,
"grad_norm": 1.5620529100957672,
"learning_rate": 1.05565194610233e-08,
"loss": 0.2341,
"step": 2026
},
{
"epoch": 0.8112,
"grad_norm": 2.647699103427376,
"learning_rate": 1.047086451856235e-08,
"loss": -0.2593,
"step": 2028
},
{
"epoch": 0.812,
"grad_norm": 2.5701556161139893,
"learning_rate": 1.038551783224047e-08,
"loss": 0.6179,
"step": 2030
},
{
"epoch": 0.8128,
"grad_norm": 0.8549953817365237,
"learning_rate": 1.030048006760823e-08,
"loss": -0.0904,
"step": 2032
},
{
"epoch": 0.8136,
"grad_norm": 7.129839478186939,
"learning_rate": 1.0215751887807228e-08,
"loss": 0.3566,
"step": 2034
},
{
"epoch": 0.8144,
"grad_norm": 3.6986430600685414,
"learning_rate": 1.0131333953564824e-08,
"loss": -1.0713,
"step": 2036
},
{
"epoch": 0.8152,
"grad_norm": 2.1872738933486096,
"learning_rate": 1.0047226923189023e-08,
"loss": 0.3407,
"step": 2038
},
{
"epoch": 0.816,
"grad_norm": 0.3128925347394902,
"learning_rate": 9.96343145256333e-09,
"loss": 0.0907,
"step": 2040
},
{
"epoch": 0.8168,
"grad_norm": 4.786453975612638,
"learning_rate": 9.87994819514168e-09,
"loss": -0.0753,
"step": 2042
},
{
"epoch": 0.8176,
"grad_norm": 14.82356796616775,
"learning_rate": 9.796777801943268e-09,
"loss": 0.0171,
"step": 2044
},
{
"epoch": 0.8184,
"grad_norm": 2.633420400987255,
"learning_rate": 9.71392092154753e-09,
"loss": 0.161,
"step": 2046
},
{
"epoch": 0.8192,
"grad_norm": 1.1007379261044739,
"learning_rate": 9.63137820008908e-09,
"loss": -0.364,
"step": 2048
},
{
"epoch": 0.82,
"grad_norm": 0.8467779244521326,
"learning_rate": 9.549150281252633e-09,
"loss": 0.1341,
"step": 2050
},
{
"epoch": 0.8208,
"grad_norm": 0.8244126140483381,
"learning_rate": 9.467237806268008e-09,
"loss": -0.2897,
"step": 2052
},
{
"epoch": 0.8216,
"grad_norm": 5.462830167075669,
"learning_rate": 9.385641413905138e-09,
"loss": 0.0616,
"step": 2054
},
{
"epoch": 0.8224,
"grad_norm": 0.9567013585298734,
"learning_rate": 9.304361740469101e-09,
"loss": 0.1821,
"step": 2056
},
{
"epoch": 0.8232,
"grad_norm": 2.639434982510996,
"learning_rate": 9.223399419795091e-09,
"loss": -0.0443,
"step": 2058
},
{
"epoch": 0.824,
"grad_norm": 5.980931285372222,
"learning_rate": 9.142755083243575e-09,
"loss": -0.5847,
"step": 2060
},
{
"epoch": 0.8248,
"grad_norm": 1.221477188934739,
"learning_rate": 9.062429359695279e-09,
"loss": -0.6101,
"step": 2062
},
{
"epoch": 0.8256,
"grad_norm": 2.592258170134965,
"learning_rate": 8.98242287554633e-09,
"loss": -0.6055,
"step": 2064
},
{
"epoch": 0.8264,
"grad_norm": 14.286819868582796,
"learning_rate": 8.902736254703347e-09,
"loss": 0.1833,
"step": 2066
},
{
"epoch": 0.8272,
"grad_norm": 0.6219523174946232,
"learning_rate": 8.823370118578627e-09,
"loss": 0.1085,
"step": 2068
},
{
"epoch": 0.828,
"grad_norm": 2.4747243856386367,
"learning_rate": 8.744325086085247e-09,
"loss": 0.1628,
"step": 2070
},
{
"epoch": 0.8288,
"grad_norm": 7.017050884644161,
"learning_rate": 8.665601773632225e-09,
"loss": -0.4075,
"step": 2072
},
{
"epoch": 0.8296,
"grad_norm": 3.0344111746964733,
"learning_rate": 8.587200795119792e-09,
"loss": 0.1941,
"step": 2074
},
{
"epoch": 0.8304,
"grad_norm": 3.63122732750374,
"learning_rate": 8.509122761934518e-09,
"loss": -0.6316,
"step": 2076
},
{
"epoch": 0.8312,
"grad_norm": 2.727294872192778,
"learning_rate": 8.431368282944584e-09,
"loss": -0.5869,
"step": 2078
},
{
"epoch": 0.832,
"grad_norm": 8.87803956255867,
"learning_rate": 8.353937964495027e-09,
"loss": -0.0651,
"step": 2080
},
{
"epoch": 0.8328,
"grad_norm": 23.769707582315764,
"learning_rate": 8.27683241040305e-09,
"loss": -0.5846,
"step": 2082
},
{
"epoch": 0.8336,
"grad_norm": 9.259758087731516,
"learning_rate": 8.20005222195323e-09,
"loss": 0.3543,
"step": 2084
},
{
"epoch": 0.8344,
"grad_norm": 0.9286906076444958,
"learning_rate": 8.123597997892918e-09,
"loss": -0.3555,
"step": 2086
},
{
"epoch": 0.8352,
"grad_norm": 2.541911515775242,
"learning_rate": 8.047470334427504e-09,
"loss": 0.5603,
"step": 2088
},
{
"epoch": 0.836,
"grad_norm": 1.6475676211765033,
"learning_rate": 7.971669825215788e-09,
"loss": 0.3738,
"step": 2090
},
{
"epoch": 0.8368,
"grad_norm": 2.3364186215000884,
"learning_rate": 7.89619706136539e-09,
"loss": -0.2587,
"step": 2092
},
{
"epoch": 0.8376,
"grad_norm": 6.851362368895333,
"learning_rate": 7.82105263142806e-09,
"loss": -0.234,
"step": 2094
},
{
"epoch": 0.8384,
"grad_norm": 3.042257259563013,
"learning_rate": 7.746237121395184e-09,
"loss": 0.0635,
"step": 2096
},
{
"epoch": 0.8392,
"grad_norm": 5.577355769208644,
"learning_rate": 7.671751114693104e-09,
"loss": 0.346,
"step": 2098
},
{
"epoch": 0.84,
"grad_norm": 16.068257540691818,
"learning_rate": 7.597595192178703e-09,
"loss": -0.234,
"step": 2100
},
{
"epoch": 0.8408,
"grad_norm": 1.4550019286994413,
"learning_rate": 7.523769932134739e-09,
"loss": 0.3045,
"step": 2102
},
{
"epoch": 0.8416,
"grad_norm": 11.674263008573776,
"learning_rate": 7.450275910265413e-09,
"loss": 0.0012,
"step": 2104
},
{
"epoch": 0.8424,
"grad_norm": 3.5922243417097186,
"learning_rate": 7.377113699691878e-09,
"loss": 0.0608,
"step": 2106
},
{
"epoch": 0.8432,
"grad_norm": 4.565894897934769,
"learning_rate": 7.3042838709477476e-09,
"loss": -0.3134,
"step": 2108
},
{
"epoch": 0.844,
"grad_norm": 1.0790754610715667,
"learning_rate": 7.23178699197467e-09,
"loss": -0.7712,
"step": 2110
},
{
"epoch": 0.8448,
"grad_norm": 2.5576412226752945,
"learning_rate": 7.159623628117856e-09,
"loss": -0.6058,
"step": 2112
},
{
"epoch": 0.8456,
"grad_norm": 4.462503611601651,
"learning_rate": 7.087794342121723e-09,
"loss": -0.0884,
"step": 2114
},
{
"epoch": 0.8464,
"grad_norm": 1.1927787940264971,
"learning_rate": 7.0162996941254495e-09,
"loss": 0.0271,
"step": 2116
},
{
"epoch": 0.8472,
"grad_norm": 1.3778837117436245,
"learning_rate": 6.945140241658687e-09,
"loss": 0.0777,
"step": 2118
},
{
"epoch": 0.848,
"grad_norm": 1.2376536517660377,
"learning_rate": 6.874316539637126e-09,
"loss": 0.2371,
"step": 2120
},
{
"epoch": 0.8488,
"grad_norm": 10.4099869029666,
"learning_rate": 6.803829140358236e-09,
"loss": 0.0905,
"step": 2122
},
{
"epoch": 0.8496,
"grad_norm": 2.0945162332974863,
"learning_rate": 6.7336785934969e-09,
"loss": -0.2163,
"step": 2124
},
{
"epoch": 0.8504,
"grad_norm": 1.084301127375519,
"learning_rate": 6.663865446101191e-09,
"loss": 0.3561,
"step": 2126
},
{
"epoch": 0.8512,
"grad_norm": 3.3122087703299163,
"learning_rate": 6.594390242588044e-09,
"loss": 0.654,
"step": 2128
},
{
"epoch": 0.852,
"grad_norm": 16.300685078686683,
"learning_rate": 6.525253524739049e-09,
"loss": -0.4144,
"step": 2130
},
{
"epoch": 0.8528,
"grad_norm": 1.9622079833970945,
"learning_rate": 6.456455831696234e-09,
"loss": 0.1632,
"step": 2132
},
{
"epoch": 0.8536,
"grad_norm": 1.1269237351058636,
"learning_rate": 6.3879976999578146e-09,
"loss": 0.0483,
"step": 2134
},
{
"epoch": 0.8544,
"grad_norm": 1.2343076260198358,
"learning_rate": 6.319879663374067e-09,
"loss": -0.173,
"step": 2136
},
{
"epoch": 0.8552,
"grad_norm": 6.5182378553408125,
"learning_rate": 6.252102253143121e-09,
"loss": 0.3039,
"step": 2138
},
{
"epoch": 0.856,
"grad_norm": 0.9652570226909292,
"learning_rate": 6.184665997806831e-09,
"loss": 0.3249,
"step": 2140
},
{
"epoch": 0.8568,
"grad_norm": 1.0862936513727042,
"learning_rate": 6.1175714232466534e-09,
"loss": -0.0158,
"step": 2142
},
{
"epoch": 0.8576,
"grad_norm": 4.249676407264014,
"learning_rate": 6.050819052679585e-09,
"loss": 0.1583,
"step": 2144
},
{
"epoch": 0.8584,
"grad_norm": 2.9048667267483523,
"learning_rate": 5.9844094066539896e-09,
"loss": -0.1592,
"step": 2146
},
{
"epoch": 0.8592,
"grad_norm": 1.7700629849804357,
"learning_rate": 5.918343003045656e-09,
"loss": 0.4752,
"step": 2148
},
{
"epoch": 0.86,
"grad_norm": 0.6095234167162628,
"learning_rate": 5.8526203570536504e-09,
"loss": 0.1604,
"step": 2150
},
{
"epoch": 0.8608,
"grad_norm": 2.2319947159315983,
"learning_rate": 5.787241981196383e-09,
"loss": -0.2655,
"step": 2152
},
{
"epoch": 0.8616,
"grad_norm": 2.385565400146363,
"learning_rate": 5.7222083853075585e-09,
"loss": 0.1889,
"step": 2154
},
{
"epoch": 0.8624,
"grad_norm": 1.2740430540187664,
"learning_rate": 5.657520076532207e-09,
"loss": -0.0907,
"step": 2156
},
{
"epoch": 0.8632,
"grad_norm": 0.8496723396985106,
"learning_rate": 5.593177559322776e-09,
"loss": -0.1871,
"step": 2158
},
{
"epoch": 0.864,
"grad_norm": 4.349013100883454,
"learning_rate": 5.529181335435124e-09,
"loss": -0.2457,
"step": 2160
},
{
"epoch": 0.8648,
"grad_norm": 2.624184256356982,
"learning_rate": 5.46553190392467e-09,
"loss": -0.0564,
"step": 2162
},
{
"epoch": 0.8656,
"grad_norm": 4.891384467420414,
"learning_rate": 5.402229761142463e-09,
"loss": 0.0126,
"step": 2164
},
{
"epoch": 0.8664,
"grad_norm": 1.0690277941852824,
"learning_rate": 5.33927540073133e-09,
"loss": 0.1096,
"step": 2166
},
{
"epoch": 0.8672,
"grad_norm": 1.0864935645910627,
"learning_rate": 5.276669313622012e-09,
"loss": 0.4332,
"step": 2168
},
{
"epoch": 0.868,
"grad_norm": 1.3794846000185592,
"learning_rate": 5.214411988029355e-09,
"loss": -0.2237,
"step": 2170
},
{
"epoch": 0.8688,
"grad_norm": 2.1597838851262567,
"learning_rate": 5.1525039094485025e-09,
"loss": -0.0318,
"step": 2172
},
{
"epoch": 0.8696,
"grad_norm": 8.604568980124318,
"learning_rate": 5.090945560651072e-09,
"loss": 0.7412,
"step": 2174
},
{
"epoch": 0.8704,
"grad_norm": 9.383895725071664,
"learning_rate": 5.029737421681446e-09,
"loss": 0.1443,
"step": 2176
},
{
"epoch": 0.8712,
"grad_norm": 1.4793862750428757,
"learning_rate": 4.968879969852985e-09,
"loss": -0.0098,
"step": 2178
},
{
"epoch": 0.872,
"grad_norm": 1.3610591443697106,
"learning_rate": 4.9083736797443155e-09,
"loss": 0.1644,
"step": 2180
},
{
"epoch": 0.8728,
"grad_norm": 3.385230548173844,
"learning_rate": 4.848219023195643e-09,
"loss": -0.344,
"step": 2182
},
{
"epoch": 0.8736,
"grad_norm": 4.705608399579721,
"learning_rate": 4.788416469305068e-09,
"loss": -0.4683,
"step": 2184
},
{
"epoch": 0.8744,
"grad_norm": 1.8452932381974783,
"learning_rate": 4.728966484424912e-09,
"loss": 0.6021,
"step": 2186
},
{
"epoch": 0.8752,
"grad_norm": 8.270747639931717,
"learning_rate": 4.669869532158116e-09,
"loss": -0.0376,
"step": 2188
},
{
"epoch": 0.876,
"grad_norm": 1.5872349158016041,
"learning_rate": 4.611126073354571e-09,
"loss": -0.1316,
"step": 2190
},
{
"epoch": 0.8768,
"grad_norm": 1.1707857137806994,
"learning_rate": 4.552736566107562e-09,
"loss": 0.1637,
"step": 2192
},
{
"epoch": 0.8776,
"grad_norm": 3.206746320008537,
"learning_rate": 4.494701465750217e-09,
"loss": 0.0884,
"step": 2194
},
{
"epoch": 0.8784,
"grad_norm": 0.43228851408832186,
"learning_rate": 4.437021224851889e-09,
"loss": -0.07,
"step": 2196
},
{
"epoch": 0.8792,
"grad_norm": 5.848134722891433,
"learning_rate": 4.379696293214696e-09,
"loss": 0.3384,
"step": 2198
},
{
"epoch": 0.88,
"grad_norm": 8.755920463976658,
"learning_rate": 4.322727117869951e-09,
"loss": -0.1251,
"step": 2200
},
{
"epoch": 0.8808,
"grad_norm": 3.618392493004251,
"learning_rate": 4.26611414307475e-09,
"loss": -0.5217,
"step": 2202
},
{
"epoch": 0.8816,
"grad_norm": 3.074281375800813,
"learning_rate": 4.209857810308437e-09,
"loss": -0.2288,
"step": 2204
},
{
"epoch": 0.8824,
"grad_norm": 2.4682659786410603,
"learning_rate": 4.1539585582691885e-09,
"loss": 0.4702,
"step": 2206
},
{
"epoch": 0.8832,
"grad_norm": 0.6352925864601222,
"learning_rate": 4.098416822870593e-09,
"loss": 0.1221,
"step": 2208
},
{
"epoch": 0.884,
"grad_norm": 18.89293797441348,
"learning_rate": 4.043233037238281e-09,
"loss": -0.1175,
"step": 2210
},
{
"epoch": 0.8848,
"grad_norm": 1.2424133172966803,
"learning_rate": 3.9884076317064805e-09,
"loss": 0.1494,
"step": 2212
},
{
"epoch": 0.8856,
"grad_norm": 3.3762433833453573,
"learning_rate": 3.933941033814736e-09,
"loss": 0.4943,
"step": 2214
},
{
"epoch": 0.8864,
"grad_norm": 2.941556993324549,
"learning_rate": 3.879833668304505e-09,
"loss": 0.2086,
"step": 2216
},
{
"epoch": 0.8872,
"grad_norm": 1.2160258295220527,
"learning_rate": 3.826085957115887e-09,
"loss": 0.2668,
"step": 2218
},
{
"epoch": 0.888,
"grad_norm": 1.9880657961417336,
"learning_rate": 3.772698319384349e-09,
"loss": 0.0593,
"step": 2220
},
{
"epoch": 0.8888,
"grad_norm": 4.467655660380731,
"learning_rate": 3.719671171437394e-09,
"loss": -1.063,
"step": 2222
},
{
"epoch": 0.8896,
"grad_norm": 3.831471428262152,
"learning_rate": 3.667004926791395e-09,
"loss": -0.4956,
"step": 2224
},
{
"epoch": 0.8904,
"grad_norm": 2.917151972577474,
"learning_rate": 3.614699996148285e-09,
"loss": -0.1451,
"step": 2226
},
{
"epoch": 0.8912,
"grad_norm": 0.614337892942797,
"learning_rate": 3.5627567873924514e-09,
"loss": -0.1331,
"step": 2228
},
{
"epoch": 0.892,
"grad_norm": 1.1621377172904874,
"learning_rate": 3.5111757055874323e-09,
"loss": 0.5733,
"step": 2230
},
{
"epoch": 0.8928,
"grad_norm": 1.6634098853629664,
"learning_rate": 3.4599571529728867e-09,
"loss": 0.044,
"step": 2232
},
{
"epoch": 0.8936,
"grad_norm": 2.3445443840858844,
"learning_rate": 3.4091015289613777e-09,
"loss": -0.0438,
"step": 2234
},
{
"epoch": 0.8944,
"grad_norm": 2.676333362752966,
"learning_rate": 3.3586092301352675e-09,
"loss": -0.2776,
"step": 2236
},
{
"epoch": 0.8952,
"grad_norm": 4.7697980831198405,
"learning_rate": 3.3084806502436613e-09,
"loss": -0.3659,
"step": 2238
},
{
"epoch": 0.896,
"grad_norm": 6.057959897509257,
"learning_rate": 3.2587161801992778e-09,
"loss": -0.1006,
"step": 2240
},
{
"epoch": 0.8968,
"grad_norm": 2.4224067497579043,
"learning_rate": 3.209316208075463e-09,
"loss": 0.4279,
"step": 2242
},
{
"epoch": 0.8976,
"grad_norm": 2.2956297867062876,
"learning_rate": 3.1602811191031085e-09,
"loss": -0.2779,
"step": 2244
},
{
"epoch": 0.8984,
"grad_norm": 2.676327435440919,
"learning_rate": 3.111611295667704e-09,
"loss": 0.1528,
"step": 2246
},
{
"epoch": 0.8992,
"grad_norm": 2.4298741519867852,
"learning_rate": 3.0633071173062965e-09,
"loss": 0.7115,
"step": 2248
},
{
"epoch": 0.9,
"grad_norm": 1.2550146008412881,
"learning_rate": 3.015368960704584e-09,
"loss": 0.2171,
"step": 2250
},
{
"epoch": 0.9008,
"grad_norm": 1.8243762323293855,
"learning_rate": 2.967797199693928e-09,
"loss": -0.7472,
"step": 2252
},
{
"epoch": 0.9016,
"grad_norm": 0.9210039784654722,
"learning_rate": 2.9205922052484953e-09,
"loss": -0.1888,
"step": 2254
},
{
"epoch": 0.9024,
"grad_norm": 0.8824442352985862,
"learning_rate": 2.873754345482299e-09,
"loss": -0.023,
"step": 2256
},
{
"epoch": 0.9032,
"grad_norm": 0.7664098550437426,
"learning_rate": 2.827283985646378e-09,
"loss": 0.1549,
"step": 2258
},
{
"epoch": 0.904,
"grad_norm": 4.9714840928816155,
"learning_rate": 2.78118148812595e-09,
"loss": -0.3595,
"step": 2260
},
{
"epoch": 0.9048,
"grad_norm": 8.654708151874246,
"learning_rate": 2.7354472124375303e-09,
"loss": -0.7394,
"step": 2262
},
{
"epoch": 0.9056,
"grad_norm": 4.601144208392627,
"learning_rate": 2.690081515226206e-09,
"loss": 0.2249,
"step": 2264
},
{
"epoch": 0.9064,
"grad_norm": 3.654256720460866,
"learning_rate": 2.6450847502627882e-09,
"loss": 0.3918,
"step": 2266
},
{
"epoch": 0.9072,
"grad_norm": 2.8319108774051296,
"learning_rate": 2.600457268441092e-09,
"loss": -0.6636,
"step": 2268
},
{
"epoch": 0.908,
"grad_norm": 1.0973890560354937,
"learning_rate": 2.5561994177751735e-09,
"loss": -0.0566,
"step": 2270
},
{
"epoch": 0.9088,
"grad_norm": 6.841431014654473,
"learning_rate": 2.5123115433966615e-09,
"loss": 0.4878,
"step": 2272
},
{
"epoch": 0.9096,
"grad_norm": 0.9320764615479279,
"learning_rate": 2.468793987551998e-09,
"loss": 0.1557,
"step": 2274
},
{
"epoch": 0.9104,
"grad_norm": 6.370923388386578,
"learning_rate": 2.425647089599836e-09,
"loss": -0.1932,
"step": 2276
},
{
"epoch": 0.9112,
"grad_norm": 4.183725642916174,
"learning_rate": 2.3828711860083674e-09,
"loss": 0.098,
"step": 2278
},
{
"epoch": 0.912,
"grad_norm": 4.068230796992909,
"learning_rate": 2.340466610352654e-09,
"loss": -0.3505,
"step": 2280
},
{
"epoch": 0.9128,
"grad_norm": 5.482675919877662,
"learning_rate": 2.2984336933121073e-09,
"loss": -0.7297,
"step": 2282
},
{
"epoch": 0.9136,
"grad_norm": 3.566681537761448,
"learning_rate": 2.2567727626678524e-09,
"loss": -0.8328,
"step": 2284
},
{
"epoch": 0.9144,
"grad_norm": 9.269477254720458,
"learning_rate": 2.215484143300206e-09,
"loss": 0.0203,
"step": 2286
},
{
"epoch": 0.9152,
"grad_norm": 0.7216914942401899,
"learning_rate": 2.174568157186102e-09,
"loss": -0.1531,
"step": 2288
},
{
"epoch": 0.916,
"grad_norm": 1.1979336525277888,
"learning_rate": 2.1340251233966378e-09,
"loss": 0.021,
"step": 2290
},
{
"epoch": 0.9168,
"grad_norm": 1.922137573312157,
"learning_rate": 2.0938553580945204e-09,
"loss": 0.0329,
"step": 2292
},
{
"epoch": 0.9176,
"grad_norm": 3.2309428357308225,
"learning_rate": 2.054059174531653e-09,
"loss": 0.2099,
"step": 2294
},
{
"epoch": 0.9184,
"grad_norm": 1.2013871635397861,
"learning_rate": 2.0146368830466664e-09,
"loss": -0.114,
"step": 2296
},
{
"epoch": 0.9192,
"grad_norm": 2.574041866507427,
"learning_rate": 1.97558879106251e-09,
"loss": -0.0334,
"step": 2298
},
{
"epoch": 0.92,
"grad_norm": 2.1091928529055455,
"learning_rate": 1.9369152030840553e-09,
"loss": -0.1678,
"step": 2300
},
{
"epoch": 0.9208,
"grad_norm": 1.3784969035589727,
"learning_rate": 1.8986164206957035e-09,
"loss": 0.0433,
"step": 2302
},
{
"epoch": 0.9216,
"grad_norm": 1.445665066799248,
"learning_rate": 1.8606927425590612e-09,
"loss": -0.3298,
"step": 2304
},
{
"epoch": 0.9224,
"grad_norm": 4.39979141208013,
"learning_rate": 1.8231444644105754e-09,
"loss": -0.2644,
"step": 2306
},
{
"epoch": 0.9232,
"grad_norm": 6.792092961183168,
"learning_rate": 1.7859718790592726e-09,
"loss": 0.2548,
"step": 2308
},
{
"epoch": 0.924,
"grad_norm": 8.882254523598194,
"learning_rate": 1.7491752763844291e-09,
"loss": -0.3878,
"step": 2310
},
{
"epoch": 0.9248,
"grad_norm": 0.3523571433732991,
"learning_rate": 1.7127549433333555e-09,
"loss": 0.1046,
"step": 2312
},
{
"epoch": 0.9256,
"grad_norm": 1.222101846793018,
"learning_rate": 1.67671116391912e-09,
"loss": 0.0512,
"step": 2314
},
{
"epoch": 0.9264,
"grad_norm": 1.0258685268101109,
"learning_rate": 1.641044219218357e-09,
"loss": -0.0404,
"step": 2316
},
{
"epoch": 0.9272,
"grad_norm": 0.8920962345335293,
"learning_rate": 1.6057543873690682e-09,
"loss": -0.505,
"step": 2318
},
{
"epoch": 0.928,
"grad_norm": 4.300808456044036,
"learning_rate": 1.570841943568446e-09,
"loss": -0.0736,
"step": 2320
},
{
"epoch": 0.9288,
"grad_norm": 1.7383370212962577,
"learning_rate": 1.5363071600707434e-09,
"loss": -0.2897,
"step": 2322
},
{
"epoch": 0.9296,
"grad_norm": 3.8514538101100926,
"learning_rate": 1.5021503061851348e-09,
"loss": 0.5155,
"step": 2324
},
{
"epoch": 0.9304,
"grad_norm": 9.611231070810984,
"learning_rate": 1.4683716482736363e-09,
"loss": -1.3277,
"step": 2326
},
{
"epoch": 0.9312,
"grad_norm": 15.037380136585435,
"learning_rate": 1.4349714497490006e-09,
"loss": -0.1934,
"step": 2328
},
{
"epoch": 0.932,
"grad_norm": 5.102302088022175,
"learning_rate": 1.401949971072691e-09,
"loss": -0.1637,
"step": 2330
},
{
"epoch": 0.9328,
"grad_norm": 3.2589093792296877,
"learning_rate": 1.369307469752823e-09,
"loss": -0.4259,
"step": 2332
},
{
"epoch": 0.9336,
"grad_norm": 0.5647745584410777,
"learning_rate": 1.337044200342191e-09,
"loss": 0.0052,
"step": 2334
},
{
"epoch": 0.9344,
"grad_norm": 2.8587476735191135,
"learning_rate": 1.3051604144362404e-09,
"loss": -0.1694,
"step": 2336
},
{
"epoch": 0.9352,
"grad_norm": 6.105887846076872,
"learning_rate": 1.2736563606711382e-09,
"loss": -0.093,
"step": 2338
},
{
"epoch": 0.936,
"grad_norm": 0.946233622816971,
"learning_rate": 1.2425322847218367e-09,
"loss": 0.1256,
"step": 2340
},
{
"epoch": 0.9368,
"grad_norm": 0.8448247771748791,
"learning_rate": 1.2117884293001256e-09,
"loss": 0.3261,
"step": 2342
},
{
"epoch": 0.9376,
"grad_norm": 1.1930610173772718,
"learning_rate": 1.181425034152761e-09,
"loss": 0.3508,
"step": 2344
},
{
"epoch": 0.9384,
"grad_norm": 1.0424606834190295,
"learning_rate": 1.1514423360595938e-09,
"loss": 0.0268,
"step": 2346
},
{
"epoch": 0.9392,
"grad_norm": 1.9104458999656608,
"learning_rate": 1.1218405688317445e-09,
"loss": 0.35,
"step": 2348
},
{
"epoch": 0.94,
"grad_norm": 2.7868033745601655,
"learning_rate": 1.0926199633097156e-09,
"loss": 0.3013,
"step": 2350
},
{
"epoch": 0.9408,
"grad_norm": 3.114424044546039,
"learning_rate": 1.063780747361681e-09,
"loss": 0.4124,
"step": 2352
},
{
"epoch": 0.9416,
"grad_norm": 0.9758662069822744,
"learning_rate": 1.0353231458816337e-09,
"loss": 0.0924,
"step": 2354
},
{
"epoch": 0.9424,
"grad_norm": 5.730640616704293,
"learning_rate": 1.007247380787657e-09,
"loss": -0.4145,
"step": 2356
},
{
"epoch": 0.9432,
"grad_norm": 1.2895873153425517,
"learning_rate": 9.795536710202169e-10,
"loss": 0.1583,
"step": 2358
},
{
"epoch": 0.944,
"grad_norm": 2.3790760214218243,
"learning_rate": 9.522422325404233e-10,
"loss": 0.008,
"step": 2360
},
{
"epoch": 0.9448,
"grad_norm": 1.3863491242757506,
"learning_rate": 9.253132783283546e-10,
"loss": 0.1381,
"step": 2362
},
{
"epoch": 0.9456,
"grad_norm": 11.01027091184072,
"learning_rate": 8.987670183814134e-10,
"loss": -0.4539,
"step": 2364
},
{
"epoch": 0.9464,
"grad_norm": 2.65682385350019,
"learning_rate": 8.726036597126618e-10,
"loss": 0.0709,
"step": 2366
},
{
"epoch": 0.9472,
"grad_norm": 2.744888591204184,
"learning_rate": 8.468234063492285e-10,
"loss": 0.4649,
"step": 2368
},
{
"epoch": 0.948,
"grad_norm": 1.105547898891178,
"learning_rate": 8.214264593307096e-10,
"loss": 0.3651,
"step": 2370
},
{
"epoch": 0.9488,
"grad_norm": 11.040756137143747,
"learning_rate": 7.964130167075922e-10,
"loss": -0.3469,
"step": 2372
},
{
"epoch": 0.9496,
"grad_norm": 3.7513293798622853,
"learning_rate": 7.717832735397334e-10,
"loss": 0.3173,
"step": 2374
},
{
"epoch": 0.9504,
"grad_norm": 8.41123333843685,
"learning_rate": 7.475374218948116e-10,
"loss": -0.2059,
"step": 2376
},
{
"epoch": 0.9512,
"grad_norm": 7.723511557670311,
"learning_rate": 7.236756508468611e-10,
"loss": -0.3296,
"step": 2378
},
{
"epoch": 0.952,
"grad_norm": 1.7434046344380822,
"learning_rate": 7.001981464747564e-10,
"loss": -0.2653,
"step": 2380
},
{
"epoch": 0.9528,
"grad_norm": 1.2597364465844136,
"learning_rate": 6.771050918607913e-10,
"loss": 0.0558,
"step": 2382
},
{
"epoch": 0.9536,
"grad_norm": 0.4546682108800604,
"learning_rate": 6.543966670892464e-10,
"loss": 0.1425,
"step": 2384
},
{
"epoch": 0.9544,
"grad_norm": 10.810141795864281,
"learning_rate": 6.320730492449799e-10,
"loss": 0.6105,
"step": 2386
},
{
"epoch": 0.9552,
"grad_norm": 6.814183252292936,
"learning_rate": 6.101344124120555e-10,
"loss": 0.0885,
"step": 2388
},
{
"epoch": 0.956,
"grad_norm": 1.8255386397646123,
"learning_rate": 5.885809276723608e-10,
"loss": 0.1283,
"step": 2390
},
{
"epoch": 0.9568,
"grad_norm": 4.4064865857808835,
"learning_rate": 5.674127631043024e-10,
"loss": -0.0873,
"step": 2392
},
{
"epoch": 0.9576,
"grad_norm": 5.893788093122514,
"learning_rate": 5.466300837814797e-10,
"loss": -0.2882,
"step": 2394
},
{
"epoch": 0.9584,
"grad_norm": 2.015069846194777,
"learning_rate": 5.262330517713964e-10,
"loss": 0.0329,
"step": 2396
},
{
"epoch": 0.9592,
"grad_norm": 4.38595870418589,
"learning_rate": 5.062218261342122e-10,
"loss": -0.4992,
"step": 2398
},
{
"epoch": 0.96,
"grad_norm": 1.7573595331693213,
"learning_rate": 4.865965629214819e-10,
"loss": 0.3999,
"step": 2400
},
{
"epoch": 0.9608,
"grad_norm": 2.8275795449554244,
"learning_rate": 4.673574151749571e-10,
"loss": 0.2888,
"step": 2402
},
{
"epoch": 0.9616,
"grad_norm": 2.782006626438932,
"learning_rate": 4.485045329253645e-10,
"loss": -0.3105,
"step": 2404
},
{
"epoch": 0.9624,
"grad_norm": 2.067801654579334,
"learning_rate": 4.3003806319127365e-10,
"loss": -0.0758,
"step": 2406
},
{
"epoch": 0.9632,
"grad_norm": 0.905276767266366,
"learning_rate": 4.119581499779201e-10,
"loss": 0.0992,
"step": 2408
},
{
"epoch": 0.964,
"grad_norm": 6.919526983565309,
"learning_rate": 3.9426493427611176e-10,
"loss": -0.2644,
"step": 2410
},
{
"epoch": 0.9648,
"grad_norm": 9.034362661975814,
"learning_rate": 3.769585540610798e-10,
"loss": -0.2986,
"step": 2412
},
{
"epoch": 0.9656,
"grad_norm": 3.14086560297048,
"learning_rate": 3.600391442914741e-10,
"loss": 0.7052,
"step": 2414
},
{
"epoch": 0.9664,
"grad_norm": 1.3381605274973811,
"learning_rate": 3.4350683690823055e-10,
"loss": 0.6932,
"step": 2416
},
{
"epoch": 0.9672,
"grad_norm": 48.51483665029063,
"learning_rate": 3.273617608336221e-10,
"loss": -1.0707,
"step": 2418
},
{
"epoch": 0.968,
"grad_norm": 3.755662155712364,
"learning_rate": 3.116040419701815e-10,
"loss": 0.1557,
"step": 2420
},
{
"epoch": 0.9688,
"grad_norm": 1.933882174975995,
"learning_rate": 2.962338031997691e-10,
"loss": -0.3932,
"step": 2422
},
{
"epoch": 0.9696,
"grad_norm": 3.292274309661597,
"learning_rate": 2.81251164382601e-10,
"loss": 0.2927,
"step": 2424
},
{
"epoch": 0.9704,
"grad_norm": 0.9395481380197626,
"learning_rate": 2.666562423562946e-10,
"loss": -0.0198,
"step": 2426
},
{
"epoch": 0.9712,
"grad_norm": 2.9117731605023116,
"learning_rate": 2.5244915093499133e-10,
"loss": 0.383,
"step": 2428
},
{
"epoch": 0.972,
"grad_norm": 0.7403534269223006,
"learning_rate": 2.3863000090844075e-10,
"loss": 0.2351,
"step": 2430
},
{
"epoch": 0.9728,
"grad_norm": 4.024365992664171,
"learning_rate": 2.2519890004115138e-10,
"loss": -0.7044,
"step": 2432
},
{
"epoch": 0.9736,
"grad_norm": 1.791122582491928,
"learning_rate": 2.1215595307154665e-10,
"loss": 0.3281,
"step": 2434
},
{
"epoch": 0.9744,
"grad_norm": 3.643294879721196,
"learning_rate": 1.9950126171114357e-10,
"loss": -0.3029,
"step": 2436
},
{
"epoch": 0.9752,
"grad_norm": 2.401813097073988,
"learning_rate": 1.872349246437699e-10,
"loss": -0.346,
"step": 2438
},
{
"epoch": 0.976,
"grad_norm": 1.161470321037417,
"learning_rate": 1.753570375247815e-10,
"loss": 0.0214,
"step": 2440
},
{
"epoch": 0.9768,
"grad_norm": 2.0536425465151313,
"learning_rate": 1.6386769298034064e-10,
"loss": 0.221,
"step": 2442
},
{
"epoch": 0.9776,
"grad_norm": 2.1285226467223963,
"learning_rate": 1.5276698060665004e-10,
"loss": -0.3179,
"step": 2444
},
{
"epoch": 0.9784,
"grad_norm": 2.1485854952235712,
"learning_rate": 1.4205498696930328e-10,
"loss": 0.1872,
"step": 2446
},
{
"epoch": 0.9792,
"grad_norm": 2.6297283339660535,
"learning_rate": 1.317317956025743e-10,
"loss": 0.3872,
"step": 2448
},
{
"epoch": 0.98,
"grad_norm": 1.0558228777462992,
"learning_rate": 1.2179748700879012e-10,
"loss": -0.3332,
"step": 2450
},
{
"epoch": 0.9808,
"grad_norm": 1.9479430382919645,
"learning_rate": 1.1225213865767024e-10,
"loss": -0.212,
"step": 2452
},
{
"epoch": 0.9816,
"grad_norm": 10.17784962779835,
"learning_rate": 1.0309582498577718e-10,
"loss": -0.6738,
"step": 2454
},
{
"epoch": 0.9824,
"grad_norm": 3.2715225558994083,
"learning_rate": 9.432861739586683e-11,
"loss": -0.2205,
"step": 2456
},
{
"epoch": 0.9832,
"grad_norm": 12.126713223030167,
"learning_rate": 8.595058425640011e-11,
"loss": -0.026,
"step": 2458
},
{
"epoch": 0.984,
"grad_norm": 2.5444346731980882,
"learning_rate": 7.79617909009489e-11,
"loss": 0.2714,
"step": 2460
},
{
"epoch": 0.9848,
"grad_norm": 3.2112718047857647,
"learning_rate": 7.036229962774087e-11,
"loss": -0.4355,
"step": 2462
},
{
"epoch": 0.9856,
"grad_norm": 2.4317779702655082,
"learning_rate": 6.315216969912662e-11,
"loss": -0.1779,
"step": 2464
},
{
"epoch": 0.9864,
"grad_norm": 1.0366427411833663,
"learning_rate": 5.633145734114664e-11,
"loss": 0.0066,
"step": 2466
},
{
"epoch": 0.9872,
"grad_norm": 9.102037371991102,
"learning_rate": 4.990021574309833e-11,
"loss": -0.5161,
"step": 2468
},
{
"epoch": 0.988,
"grad_norm": 4.114717937151245,
"learning_rate": 4.3858495057080834e-11,
"loss": -0.3626,
"step": 2470
},
{
"epoch": 0.9888,
"grad_norm": 7.563987228393988,
"learning_rate": 3.820634239765641e-11,
"loss": 0.3373,
"step": 2472
},
{
"epoch": 0.9896,
"grad_norm": 2.948617798085717,
"learning_rate": 3.294380184143963e-11,
"loss": -0.1291,
"step": 2474
},
{
"epoch": 0.9904,
"grad_norm": 3.8406470359549023,
"learning_rate": 2.8070914426786552e-11,
"loss": -0.2217,
"step": 2476
},
{
"epoch": 0.9912,
"grad_norm": 4.310729814484611,
"learning_rate": 2.3587718153444957e-11,
"loss": 0.6706,
"step": 2478
},
{
"epoch": 0.992,
"grad_norm": 3.119684178739935,
"learning_rate": 1.9494247982282384e-11,
"loss": 0.203,
"step": 2480
},
{
"epoch": 0.9928,
"grad_norm": 2.0436367262108153,
"learning_rate": 1.5790535835003004e-11,
"loss": 0.0699,
"step": 2482
},
{
"epoch": 0.9936,
"grad_norm": 1.9758044654494797,
"learning_rate": 1.2476610593892268e-11,
"loss": -0.2463,
"step": 2484
},
{
"epoch": 0.9944,
"grad_norm": 7.343691779754545,
"learning_rate": 9.552498101611517e-12,
"loss": -0.2488,
"step": 2486
},
{
"epoch": 0.9952,
"grad_norm": 0.6231400638141823,
"learning_rate": 7.0182211609814965e-12,
"loss": 0.6614,
"step": 2488
},
{
"epoch": 0.996,
"grad_norm": 2.775645305565287,
"learning_rate": 4.873799534788059e-12,
"loss": 0.1694,
"step": 2490
},
{
"epoch": 0.9968,
"grad_norm": 10.227058254229792,
"learning_rate": 3.119249945676694e-12,
"loss": -0.0563,
"step": 2492
},
{
"epoch": 0.9976,
"grad_norm": 2.487983479578251,
"learning_rate": 1.7545860759693443e-12,
"loss": 0.1768,
"step": 2494
},
{
"epoch": 0.9984,
"grad_norm": 0.8570358384984528,
"learning_rate": 7.798185675866875e-13,
"loss": 0.3369,
"step": 2496
},
{
"epoch": 0.9992,
"grad_norm": 5.047844589765427,
"learning_rate": 1.9495502197042212e-13,
"loss": -0.8027,
"step": 2498
},
{
"epoch": 1.0,
"grad_norm": 1.124613481673833,
"learning_rate": 0.0,
"loss": 0.1599,
"step": 2500
},
{
"epoch": 1.0,
"step": 2500,
"total_flos": 75625527656448.0,
"train_loss": -0.011922632639855147,
"train_runtime": 6151.6354,
"train_samples_per_second": 1.625,
"train_steps_per_second": 0.406
}
],
"logging_steps": 2,
"max_steps": 2500,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 75625527656448.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}