PEFT
Safetensors
ans1 / trainer_state.json
chen
Upload folder using huggingface_hub
734e310 verified
Raw
History Blame Contribute Delete
79.6 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 455,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002201430930104568,
"grad_norm": 0.7037750482559204,
"learning_rate": 0.0,
"loss": 0.308,
"step": 1
},
{
"epoch": 0.004402861860209136,
"grad_norm": 1.0037487745285034,
"learning_rate": 1.4598540145985402e-06,
"loss": 0.2687,
"step": 2
},
{
"epoch": 0.006604292790313704,
"grad_norm": 2.8608241081237793,
"learning_rate": 2.9197080291970804e-06,
"loss": 0.391,
"step": 3
},
{
"epoch": 0.008805723720418272,
"grad_norm": 0.7382011413574219,
"learning_rate": 4.379562043795621e-06,
"loss": 0.2279,
"step": 4
},
{
"epoch": 0.01100715465052284,
"grad_norm": 2.2300755977630615,
"learning_rate": 5.839416058394161e-06,
"loss": 0.425,
"step": 5
},
{
"epoch": 0.013208585580627407,
"grad_norm": 1.0738085508346558,
"learning_rate": 7.2992700729927e-06,
"loss": 0.386,
"step": 6
},
{
"epoch": 0.015410016510731976,
"grad_norm": 1.1076860427856445,
"learning_rate": 8.759124087591241e-06,
"loss": 0.4133,
"step": 7
},
{
"epoch": 0.017611447440836543,
"grad_norm": 0.9959158897399902,
"learning_rate": 1.0218978102189781e-05,
"loss": 0.2856,
"step": 8
},
{
"epoch": 0.01981287837094111,
"grad_norm": 1.159873604774475,
"learning_rate": 1.1678832116788322e-05,
"loss": 0.302,
"step": 9
},
{
"epoch": 0.02201430930104568,
"grad_norm": 1.0309239625930786,
"learning_rate": 1.3138686131386862e-05,
"loss": 0.2593,
"step": 10
},
{
"epoch": 0.024215740231150248,
"grad_norm": 0.944840669631958,
"learning_rate": 1.45985401459854e-05,
"loss": 0.3306,
"step": 11
},
{
"epoch": 0.026417171161254815,
"grad_norm": 5.096871376037598,
"learning_rate": 1.605839416058394e-05,
"loss": 0.5597,
"step": 12
},
{
"epoch": 0.028618602091359385,
"grad_norm": 0.9964978694915771,
"learning_rate": 1.7518248175182482e-05,
"loss": 0.2529,
"step": 13
},
{
"epoch": 0.030820033021463952,
"grad_norm": 1.1680715084075928,
"learning_rate": 1.897810218978102e-05,
"loss": 0.2551,
"step": 14
},
{
"epoch": 0.03302146395156852,
"grad_norm": 1.3312640190124512,
"learning_rate": 2.0437956204379563e-05,
"loss": 0.2783,
"step": 15
},
{
"epoch": 0.035222894881673086,
"grad_norm": 1.0938522815704346,
"learning_rate": 2.1897810218978105e-05,
"loss": 0.2196,
"step": 16
},
{
"epoch": 0.03742432581177765,
"grad_norm": 0.844025731086731,
"learning_rate": 2.3357664233576643e-05,
"loss": 0.223,
"step": 17
},
{
"epoch": 0.03962575674188222,
"grad_norm": 1.0622016191482544,
"learning_rate": 2.4817518248175185e-05,
"loss": 0.4032,
"step": 18
},
{
"epoch": 0.041827187671986794,
"grad_norm": 1.267992377281189,
"learning_rate": 2.6277372262773724e-05,
"loss": 0.3885,
"step": 19
},
{
"epoch": 0.04402861860209136,
"grad_norm": 0.6925700902938843,
"learning_rate": 2.7737226277372262e-05,
"loss": 0.282,
"step": 20
},
{
"epoch": 0.04623004953219593,
"grad_norm": 0.49505650997161865,
"learning_rate": 2.91970802919708e-05,
"loss": 0.1894,
"step": 21
},
{
"epoch": 0.048431480462300495,
"grad_norm": 1.3833006620407104,
"learning_rate": 3.0656934306569346e-05,
"loss": 0.2406,
"step": 22
},
{
"epoch": 0.05063291139240506,
"grad_norm": 5.850524425506592,
"learning_rate": 3.211678832116788e-05,
"loss": 0.3697,
"step": 23
},
{
"epoch": 0.05283434232250963,
"grad_norm": 0.6721429824829102,
"learning_rate": 3.357664233576642e-05,
"loss": 0.1444,
"step": 24
},
{
"epoch": 0.0550357732526142,
"grad_norm": 1.379242181777954,
"learning_rate": 3.5036496350364965e-05,
"loss": 0.188,
"step": 25
},
{
"epoch": 0.05723720418271877,
"grad_norm": 1.0648667812347412,
"learning_rate": 3.649635036496351e-05,
"loss": 0.1628,
"step": 26
},
{
"epoch": 0.05943863511282334,
"grad_norm": 1.553063154220581,
"learning_rate": 3.795620437956204e-05,
"loss": 0.4074,
"step": 27
},
{
"epoch": 0.061640066042927905,
"grad_norm": 2.283454418182373,
"learning_rate": 3.941605839416059e-05,
"loss": 0.3169,
"step": 28
},
{
"epoch": 0.06384149697303247,
"grad_norm": 2.135236978530884,
"learning_rate": 4.0875912408759126e-05,
"loss": 0.3472,
"step": 29
},
{
"epoch": 0.06604292790313704,
"grad_norm": 7.040278911590576,
"learning_rate": 4.233576642335767e-05,
"loss": 0.2676,
"step": 30
},
{
"epoch": 0.0682443588332416,
"grad_norm": 1.1241406202316284,
"learning_rate": 4.379562043795621e-05,
"loss": 0.4671,
"step": 31
},
{
"epoch": 0.07044578976334617,
"grad_norm": 1.736489176750183,
"learning_rate": 4.5255474452554745e-05,
"loss": 0.2511,
"step": 32
},
{
"epoch": 0.07264722069345074,
"grad_norm": 1.144145131111145,
"learning_rate": 4.6715328467153287e-05,
"loss": 0.226,
"step": 33
},
{
"epoch": 0.0748486516235553,
"grad_norm": 0.8320156931877136,
"learning_rate": 4.817518248175183e-05,
"loss": 0.1128,
"step": 34
},
{
"epoch": 0.07705008255365987,
"grad_norm": 1.2651524543762207,
"learning_rate": 4.963503649635037e-05,
"loss": 0.1757,
"step": 35
},
{
"epoch": 0.07925151348376444,
"grad_norm": 0.9096120595932007,
"learning_rate": 5.109489051094891e-05,
"loss": 0.1775,
"step": 36
},
{
"epoch": 0.08145294441386902,
"grad_norm": 0.9276086091995239,
"learning_rate": 5.255474452554745e-05,
"loss": 0.1733,
"step": 37
},
{
"epoch": 0.08365437534397359,
"grad_norm": 0.9676797389984131,
"learning_rate": 5.401459854014599e-05,
"loss": 0.1706,
"step": 38
},
{
"epoch": 0.08585580627407816,
"grad_norm": 1.0610243082046509,
"learning_rate": 5.5474452554744524e-05,
"loss": 0.1549,
"step": 39
},
{
"epoch": 0.08805723720418272,
"grad_norm": 1.3586833477020264,
"learning_rate": 5.6934306569343066e-05,
"loss": 0.1767,
"step": 40
},
{
"epoch": 0.09025866813428729,
"grad_norm": 1.7117362022399902,
"learning_rate": 5.83941605839416e-05,
"loss": 0.2045,
"step": 41
},
{
"epoch": 0.09246009906439186,
"grad_norm": 1.0564417839050293,
"learning_rate": 5.985401459854016e-05,
"loss": 0.2057,
"step": 42
},
{
"epoch": 0.09466152999449642,
"grad_norm": 0.84938645362854,
"learning_rate": 6.131386861313869e-05,
"loss": 0.1064,
"step": 43
},
{
"epoch": 0.09686296092460099,
"grad_norm": 1.660563349723816,
"learning_rate": 6.277372262773723e-05,
"loss": 0.154,
"step": 44
},
{
"epoch": 0.09906439185470556,
"grad_norm": 2.250649929046631,
"learning_rate": 6.423357664233576e-05,
"loss": 0.29,
"step": 45
},
{
"epoch": 0.10126582278481013,
"grad_norm": 1.0987548828125,
"learning_rate": 6.56934306569343e-05,
"loss": 0.1161,
"step": 46
},
{
"epoch": 0.10346725371491469,
"grad_norm": 1.6264435052871704,
"learning_rate": 6.715328467153285e-05,
"loss": 0.3335,
"step": 47
},
{
"epoch": 0.10566868464501926,
"grad_norm": 1.36955726146698,
"learning_rate": 6.86131386861314e-05,
"loss": 0.1541,
"step": 48
},
{
"epoch": 0.10787011557512383,
"grad_norm": 1.5003987550735474,
"learning_rate": 7.007299270072993e-05,
"loss": 0.4296,
"step": 49
},
{
"epoch": 0.1100715465052284,
"grad_norm": 0.9643859267234802,
"learning_rate": 7.153284671532847e-05,
"loss": 0.1276,
"step": 50
},
{
"epoch": 0.11227297743533296,
"grad_norm": 0.8790437579154968,
"learning_rate": 7.299270072992701e-05,
"loss": 0.0857,
"step": 51
},
{
"epoch": 0.11447440836543754,
"grad_norm": 1.8895280361175537,
"learning_rate": 7.445255474452556e-05,
"loss": 0.3803,
"step": 52
},
{
"epoch": 0.11667583929554211,
"grad_norm": 0.6855382323265076,
"learning_rate": 7.591240875912408e-05,
"loss": 0.1926,
"step": 53
},
{
"epoch": 0.11887727022564668,
"grad_norm": 1.257523536682129,
"learning_rate": 7.737226277372264e-05,
"loss": 0.1345,
"step": 54
},
{
"epoch": 0.12107870115575124,
"grad_norm": 1.2312171459197998,
"learning_rate": 7.883211678832118e-05,
"loss": 0.3421,
"step": 55
},
{
"epoch": 0.12328013208585581,
"grad_norm": 1.0815132856369019,
"learning_rate": 8.029197080291971e-05,
"loss": 0.2859,
"step": 56
},
{
"epoch": 0.12548156301596036,
"grad_norm": 0.5844975709915161,
"learning_rate": 8.175182481751825e-05,
"loss": 0.1157,
"step": 57
},
{
"epoch": 0.12768299394606494,
"grad_norm": 1.176499605178833,
"learning_rate": 8.32116788321168e-05,
"loss": 0.2674,
"step": 58
},
{
"epoch": 0.1298844248761695,
"grad_norm": 0.8862079977989197,
"learning_rate": 8.467153284671534e-05,
"loss": 0.1591,
"step": 59
},
{
"epoch": 0.13208585580627408,
"grad_norm": 0.5914232134819031,
"learning_rate": 8.613138686131386e-05,
"loss": 0.0925,
"step": 60
},
{
"epoch": 0.13428728673637866,
"grad_norm": 0.5085222125053406,
"learning_rate": 8.759124087591242e-05,
"loss": 0.1197,
"step": 61
},
{
"epoch": 0.1364887176664832,
"grad_norm": 0.5677929520606995,
"learning_rate": 8.905109489051096e-05,
"loss": 0.1964,
"step": 62
},
{
"epoch": 0.1386901485965878,
"grad_norm": 0.7557106614112854,
"learning_rate": 9.051094890510949e-05,
"loss": 0.1455,
"step": 63
},
{
"epoch": 0.14089157952669235,
"grad_norm": 0.962935745716095,
"learning_rate": 9.197080291970803e-05,
"loss": 0.3851,
"step": 64
},
{
"epoch": 0.14309301045679693,
"grad_norm": 1.9366933107376099,
"learning_rate": 9.343065693430657e-05,
"loss": 0.2542,
"step": 65
},
{
"epoch": 0.14529444138690148,
"grad_norm": 1.4374990463256836,
"learning_rate": 9.489051094890511e-05,
"loss": 0.2277,
"step": 66
},
{
"epoch": 0.14749587231700606,
"grad_norm": 1.0540611743927002,
"learning_rate": 9.635036496350366e-05,
"loss": 0.2025,
"step": 67
},
{
"epoch": 0.1496973032471106,
"grad_norm": 1.1396342515945435,
"learning_rate": 9.78102189781022e-05,
"loss": 0.1476,
"step": 68
},
{
"epoch": 0.1518987341772152,
"grad_norm": 0.7722952365875244,
"learning_rate": 9.927007299270074e-05,
"loss": 0.0759,
"step": 69
},
{
"epoch": 0.15410016510731975,
"grad_norm": 1.0545134544372559,
"learning_rate": 0.00010072992700729928,
"loss": 0.1877,
"step": 70
},
{
"epoch": 0.15630159603742433,
"grad_norm": 0.7256941795349121,
"learning_rate": 0.00010218978102189782,
"loss": 0.2213,
"step": 71
},
{
"epoch": 0.15850302696752888,
"grad_norm": 0.8643118143081665,
"learning_rate": 0.00010364963503649637,
"loss": 0.1217,
"step": 72
},
{
"epoch": 0.16070445789763346,
"grad_norm": 1.3518829345703125,
"learning_rate": 0.0001051094890510949,
"loss": 0.1657,
"step": 73
},
{
"epoch": 0.16290588882773804,
"grad_norm": 1.2510859966278076,
"learning_rate": 0.00010656934306569344,
"loss": 0.3364,
"step": 74
},
{
"epoch": 0.1651073197578426,
"grad_norm": 0.7721868753433228,
"learning_rate": 0.00010802919708029198,
"loss": 0.1296,
"step": 75
},
{
"epoch": 0.16730875068794718,
"grad_norm": 1.2092798948287964,
"learning_rate": 0.00010948905109489052,
"loss": 0.183,
"step": 76
},
{
"epoch": 0.16951018161805173,
"grad_norm": 0.6521322727203369,
"learning_rate": 0.00011094890510948905,
"loss": 0.129,
"step": 77
},
{
"epoch": 0.1717116125481563,
"grad_norm": 1.9378713369369507,
"learning_rate": 0.00011240875912408759,
"loss": 0.2796,
"step": 78
},
{
"epoch": 0.17391304347826086,
"grad_norm": 1.120902180671692,
"learning_rate": 0.00011386861313868613,
"loss": 0.1491,
"step": 79
},
{
"epoch": 0.17611447440836545,
"grad_norm": 1.9726943969726562,
"learning_rate": 0.00011532846715328467,
"loss": 0.3284,
"step": 80
},
{
"epoch": 0.17831590533847,
"grad_norm": 0.799227237701416,
"learning_rate": 0.0001167883211678832,
"loss": 0.1668,
"step": 81
},
{
"epoch": 0.18051733626857458,
"grad_norm": 1.8196526765823364,
"learning_rate": 0.00011824817518248177,
"loss": 0.4528,
"step": 82
},
{
"epoch": 0.18271876719867913,
"grad_norm": 0.9238607883453369,
"learning_rate": 0.00011970802919708031,
"loss": 0.2334,
"step": 83
},
{
"epoch": 0.1849201981287837,
"grad_norm": 0.9001494646072388,
"learning_rate": 0.00012116788321167884,
"loss": 0.1732,
"step": 84
},
{
"epoch": 0.18712162905888827,
"grad_norm": 0.7415959239006042,
"learning_rate": 0.00012262773722627738,
"loss": 0.0985,
"step": 85
},
{
"epoch": 0.18932305998899285,
"grad_norm": 0.9463704824447632,
"learning_rate": 0.0001240875912408759,
"loss": 0.151,
"step": 86
},
{
"epoch": 0.1915244909190974,
"grad_norm": 0.735754132270813,
"learning_rate": 0.00012554744525547447,
"loss": 0.095,
"step": 87
},
{
"epoch": 0.19372592184920198,
"grad_norm": 0.9355161786079407,
"learning_rate": 0.000127007299270073,
"loss": 0.2339,
"step": 88
},
{
"epoch": 0.19592735277930656,
"grad_norm": 1.4002137184143066,
"learning_rate": 0.00012846715328467152,
"loss": 0.2057,
"step": 89
},
{
"epoch": 0.19812878370941112,
"grad_norm": 1.038177251815796,
"learning_rate": 0.00012992700729927008,
"loss": 0.1742,
"step": 90
},
{
"epoch": 0.2003302146395157,
"grad_norm": 0.9062851071357727,
"learning_rate": 0.0001313868613138686,
"loss": 0.134,
"step": 91
},
{
"epoch": 0.20253164556962025,
"grad_norm": 1.0446114540100098,
"learning_rate": 0.00013284671532846716,
"loss": 0.1967,
"step": 92
},
{
"epoch": 0.20473307649972483,
"grad_norm": 1.0713139772415161,
"learning_rate": 0.0001343065693430657,
"loss": 0.2289,
"step": 93
},
{
"epoch": 0.20693450742982938,
"grad_norm": 0.8900029063224792,
"learning_rate": 0.00013576642335766422,
"loss": 0.1469,
"step": 94
},
{
"epoch": 0.20913593835993396,
"grad_norm": 0.615839958190918,
"learning_rate": 0.0001372262773722628,
"loss": 0.2123,
"step": 95
},
{
"epoch": 0.21133736929003852,
"grad_norm": 0.9755569696426392,
"learning_rate": 0.00013868613138686133,
"loss": 0.1719,
"step": 96
},
{
"epoch": 0.2135388002201431,
"grad_norm": 0.7173188328742981,
"learning_rate": 0.00014014598540145986,
"loss": 0.2311,
"step": 97
},
{
"epoch": 0.21574023115024765,
"grad_norm": 2.7052557468414307,
"learning_rate": 0.00014160583941605841,
"loss": 0.2516,
"step": 98
},
{
"epoch": 0.21794166208035223,
"grad_norm": 1.7500773668289185,
"learning_rate": 0.00014306569343065694,
"loss": 0.2488,
"step": 99
},
{
"epoch": 0.2201430930104568,
"grad_norm": 1.7758418321609497,
"learning_rate": 0.00014452554744525547,
"loss": 0.4666,
"step": 100
},
{
"epoch": 0.22234452394056137,
"grad_norm": 1.0867117643356323,
"learning_rate": 0.00014598540145985403,
"loss": 0.3035,
"step": 101
},
{
"epoch": 0.22454595487066592,
"grad_norm": 1.184135913848877,
"learning_rate": 0.00014744525547445256,
"loss": 0.175,
"step": 102
},
{
"epoch": 0.2267473858007705,
"grad_norm": 0.8942990303039551,
"learning_rate": 0.0001489051094890511,
"loss": 0.1176,
"step": 103
},
{
"epoch": 0.22894881673087508,
"grad_norm": 1.145623803138733,
"learning_rate": 0.00015036496350364964,
"loss": 0.1915,
"step": 104
},
{
"epoch": 0.23115024766097964,
"grad_norm": 1.5253881216049194,
"learning_rate": 0.00015182481751824817,
"loss": 0.3467,
"step": 105
},
{
"epoch": 0.23335167859108422,
"grad_norm": 1.0943421125411987,
"learning_rate": 0.00015328467153284672,
"loss": 0.2571,
"step": 106
},
{
"epoch": 0.23555310952118877,
"grad_norm": 1.0738698244094849,
"learning_rate": 0.00015474452554744528,
"loss": 0.1974,
"step": 107
},
{
"epoch": 0.23775454045129335,
"grad_norm": 0.8418899178504944,
"learning_rate": 0.0001562043795620438,
"loss": 0.1585,
"step": 108
},
{
"epoch": 0.2399559713813979,
"grad_norm": 2.6217808723449707,
"learning_rate": 0.00015766423357664236,
"loss": 0.4359,
"step": 109
},
{
"epoch": 0.24215740231150248,
"grad_norm": 0.8705646395683289,
"learning_rate": 0.0001591240875912409,
"loss": 0.2176,
"step": 110
},
{
"epoch": 0.24435883324160704,
"grad_norm": 0.8313264846801758,
"learning_rate": 0.00016058394160583942,
"loss": 0.1573,
"step": 111
},
{
"epoch": 0.24656026417171162,
"grad_norm": 1.1230180263519287,
"learning_rate": 0.00016204379562043797,
"loss": 0.2631,
"step": 112
},
{
"epoch": 0.24876169510181617,
"grad_norm": 1.0018892288208008,
"learning_rate": 0.0001635036496350365,
"loss": 0.1711,
"step": 113
},
{
"epoch": 0.2509631260319207,
"grad_norm": 0.5920968055725098,
"learning_rate": 0.00016496350364963503,
"loss": 0.1189,
"step": 114
},
{
"epoch": 0.25316455696202533,
"grad_norm": 1.4356651306152344,
"learning_rate": 0.0001664233576642336,
"loss": 0.2634,
"step": 115
},
{
"epoch": 0.2553659878921299,
"grad_norm": 1.3919761180877686,
"learning_rate": 0.00016788321167883211,
"loss": 0.2505,
"step": 116
},
{
"epoch": 0.25756741882223444,
"grad_norm": 0.9886398911476135,
"learning_rate": 0.00016934306569343067,
"loss": 0.1908,
"step": 117
},
{
"epoch": 0.259768849752339,
"grad_norm": 0.9664818644523621,
"learning_rate": 0.0001708029197080292,
"loss": 0.1563,
"step": 118
},
{
"epoch": 0.2619702806824436,
"grad_norm": 0.9564926624298096,
"learning_rate": 0.00017226277372262773,
"loss": 0.1542,
"step": 119
},
{
"epoch": 0.26417171161254815,
"grad_norm": 0.613195538520813,
"learning_rate": 0.00017372262773722628,
"loss": 0.0748,
"step": 120
},
{
"epoch": 0.2663731425426527,
"grad_norm": 1.116240382194519,
"learning_rate": 0.00017518248175182484,
"loss": 0.2469,
"step": 121
},
{
"epoch": 0.2685745734727573,
"grad_norm": 0.6556792259216309,
"learning_rate": 0.00017664233576642337,
"loss": 0.1449,
"step": 122
},
{
"epoch": 0.27077600440286187,
"grad_norm": 1.6238877773284912,
"learning_rate": 0.00017810218978102192,
"loss": 0.2584,
"step": 123
},
{
"epoch": 0.2729774353329664,
"grad_norm": 1.42080819606781,
"learning_rate": 0.00017956204379562045,
"loss": 0.3169,
"step": 124
},
{
"epoch": 0.275178866263071,
"grad_norm": 0.721822202205658,
"learning_rate": 0.00018102189781021898,
"loss": 0.1188,
"step": 125
},
{
"epoch": 0.2773802971931756,
"grad_norm": 1.00337553024292,
"learning_rate": 0.00018248175182481753,
"loss": 0.2328,
"step": 126
},
{
"epoch": 0.27958172812328014,
"grad_norm": 1.1204198598861694,
"learning_rate": 0.00018394160583941606,
"loss": 0.191,
"step": 127
},
{
"epoch": 0.2817831590533847,
"grad_norm": 0.7608072757720947,
"learning_rate": 0.0001854014598540146,
"loss": 0.1397,
"step": 128
},
{
"epoch": 0.28398458998348924,
"grad_norm": 0.9406429529190063,
"learning_rate": 0.00018686131386861315,
"loss": 0.3071,
"step": 129
},
{
"epoch": 0.28618602091359385,
"grad_norm": 0.8172856569290161,
"learning_rate": 0.00018832116788321167,
"loss": 0.1572,
"step": 130
},
{
"epoch": 0.2883874518436984,
"grad_norm": 0.7377148866653442,
"learning_rate": 0.00018978102189781023,
"loss": 0.2005,
"step": 131
},
{
"epoch": 0.29058888277380296,
"grad_norm": 1.500381588935852,
"learning_rate": 0.00019124087591240879,
"loss": 0.2424,
"step": 132
},
{
"epoch": 0.29279031370390757,
"grad_norm": 1.5137453079223633,
"learning_rate": 0.00019270072992700731,
"loss": 0.1776,
"step": 133
},
{
"epoch": 0.2949917446340121,
"grad_norm": 1.6952030658721924,
"learning_rate": 0.00019416058394160587,
"loss": 0.1664,
"step": 134
},
{
"epoch": 0.2971931755641167,
"grad_norm": 1.0178738832473755,
"learning_rate": 0.0001956204379562044,
"loss": 0.1929,
"step": 135
},
{
"epoch": 0.2993946064942212,
"grad_norm": 2.562960624694824,
"learning_rate": 0.00019708029197080293,
"loss": 0.2782,
"step": 136
},
{
"epoch": 0.30159603742432584,
"grad_norm": 1.6147712469100952,
"learning_rate": 0.00019854014598540148,
"loss": 0.2905,
"step": 137
},
{
"epoch": 0.3037974683544304,
"grad_norm": 1.1269806623458862,
"learning_rate": 0.0002,
"loss": 0.2205,
"step": 138
},
{
"epoch": 0.30599889928453494,
"grad_norm": 1.709218978881836,
"learning_rate": 0.00019999967115038762,
"loss": 0.203,
"step": 139
},
{
"epoch": 0.3082003302146395,
"grad_norm": 1.1186352968215942,
"learning_rate": 0.00019999868460371332,
"loss": 0.2359,
"step": 140
},
{
"epoch": 0.3104017611447441,
"grad_norm": 2.0815446376800537,
"learning_rate": 0.00019999704036646558,
"loss": 0.7654,
"step": 141
},
{
"epoch": 0.31260319207484866,
"grad_norm": 1.123903751373291,
"learning_rate": 0.00019999473844945856,
"loss": 0.1531,
"step": 142
},
{
"epoch": 0.3148046230049532,
"grad_norm": 1.1775410175323486,
"learning_rate": 0.00019999177886783194,
"loss": 0.2142,
"step": 143
},
{
"epoch": 0.31700605393505776,
"grad_norm": 1.0821821689605713,
"learning_rate": 0.00019998816164105087,
"loss": 0.194,
"step": 144
},
{
"epoch": 0.31920748486516237,
"grad_norm": 1.846646785736084,
"learning_rate": 0.00019998388679290583,
"loss": 0.421,
"step": 145
},
{
"epoch": 0.3214089157952669,
"grad_norm": 0.9930664300918579,
"learning_rate": 0.00019997895435151246,
"loss": 0.2811,
"step": 146
},
{
"epoch": 0.3236103467253715,
"grad_norm": 0.9030304551124573,
"learning_rate": 0.00019997336434931137,
"loss": 0.1815,
"step": 147
},
{
"epoch": 0.3258117776554761,
"grad_norm": 0.6974136829376221,
"learning_rate": 0.000199967116823068,
"loss": 0.1949,
"step": 148
},
{
"epoch": 0.32801320858558064,
"grad_norm": 0.8216617107391357,
"learning_rate": 0.00019996021181387224,
"loss": 0.2127,
"step": 149
},
{
"epoch": 0.3302146395156852,
"grad_norm": 1.0203906297683716,
"learning_rate": 0.00019995264936713828,
"loss": 0.2603,
"step": 150
},
{
"epoch": 0.33241607044578975,
"grad_norm": 1.0693780183792114,
"learning_rate": 0.00019994442953260433,
"loss": 0.258,
"step": 151
},
{
"epoch": 0.33461750137589436,
"grad_norm": 1.3966225385665894,
"learning_rate": 0.00019993555236433213,
"loss": 0.2827,
"step": 152
},
{
"epoch": 0.3368189323059989,
"grad_norm": 0.5770395994186401,
"learning_rate": 0.00019992601792070679,
"loss": 0.1481,
"step": 153
},
{
"epoch": 0.33902036323610346,
"grad_norm": 1.5406584739685059,
"learning_rate": 0.00019991582626443618,
"loss": 0.1926,
"step": 154
},
{
"epoch": 0.341221794166208,
"grad_norm": 1.5380587577819824,
"learning_rate": 0.00019990497746255086,
"loss": 0.2582,
"step": 155
},
{
"epoch": 0.3434232250963126,
"grad_norm": 0.7996829748153687,
"learning_rate": 0.00019989347158640323,
"loss": 0.2201,
"step": 156
},
{
"epoch": 0.3456246560264172,
"grad_norm": 1.2687666416168213,
"learning_rate": 0.00019988130871166737,
"loss": 0.2027,
"step": 157
},
{
"epoch": 0.34782608695652173,
"grad_norm": 0.9393393993377686,
"learning_rate": 0.00019986848891833845,
"loss": 0.2079,
"step": 158
},
{
"epoch": 0.3500275178866263,
"grad_norm": 1.0372052192687988,
"learning_rate": 0.0001998550122907321,
"loss": 0.1452,
"step": 159
},
{
"epoch": 0.3522289488167309,
"grad_norm": 1.1515039205551147,
"learning_rate": 0.00019984087891748402,
"loss": 0.2363,
"step": 160
},
{
"epoch": 0.35443037974683544,
"grad_norm": 2.2971878051757812,
"learning_rate": 0.0001998260888915493,
"loss": 0.5014,
"step": 161
},
{
"epoch": 0.35663181067694,
"grad_norm": 1.187995433807373,
"learning_rate": 0.00019981064231020182,
"loss": 0.2729,
"step": 162
},
{
"epoch": 0.3588332416070446,
"grad_norm": 1.2573753595352173,
"learning_rate": 0.00019979453927503364,
"loss": 0.2384,
"step": 163
},
{
"epoch": 0.36103467253714916,
"grad_norm": 1.0439549684524536,
"learning_rate": 0.0001997777798919543,
"loss": 0.2956,
"step": 164
},
{
"epoch": 0.3632361034672537,
"grad_norm": 1.4184736013412476,
"learning_rate": 0.00019976036427119006,
"loss": 0.2207,
"step": 165
},
{
"epoch": 0.36543753439735827,
"grad_norm": 1.2747489213943481,
"learning_rate": 0.00019974229252728342,
"loss": 0.2199,
"step": 166
},
{
"epoch": 0.3676389653274629,
"grad_norm": 0.513881266117096,
"learning_rate": 0.00019972356477909205,
"loss": 0.113,
"step": 167
},
{
"epoch": 0.3698403962575674,
"grad_norm": 1.0329495668411255,
"learning_rate": 0.0001997041811497882,
"loss": 0.2586,
"step": 168
},
{
"epoch": 0.372041827187672,
"grad_norm": 0.6814864277839661,
"learning_rate": 0.00019968414176685788,
"loss": 0.1375,
"step": 169
},
{
"epoch": 0.37424325811777653,
"grad_norm": 1.2967740297317505,
"learning_rate": 0.0001996634467620999,
"loss": 0.201,
"step": 170
},
{
"epoch": 0.37644468904788114,
"grad_norm": 2.24857759475708,
"learning_rate": 0.00019964209627162526,
"loss": 0.2661,
"step": 171
},
{
"epoch": 0.3786461199779857,
"grad_norm": 1.3992667198181152,
"learning_rate": 0.00019962009043585584,
"loss": 0.0903,
"step": 172
},
{
"epoch": 0.38084755090809025,
"grad_norm": 2.2193541526794434,
"learning_rate": 0.00019959742939952392,
"loss": 0.2786,
"step": 173
},
{
"epoch": 0.3830489818381948,
"grad_norm": 1.2357624769210815,
"learning_rate": 0.00019957411331167094,
"loss": 0.3981,
"step": 174
},
{
"epoch": 0.3852504127682994,
"grad_norm": 0.8800520300865173,
"learning_rate": 0.00019955014232564663,
"loss": 0.1451,
"step": 175
},
{
"epoch": 0.38745184369840396,
"grad_norm": 0.7195889949798584,
"learning_rate": 0.000199525516599108,
"loss": 0.0959,
"step": 176
},
{
"epoch": 0.3896532746285085,
"grad_norm": 2.2375988960266113,
"learning_rate": 0.00019950023629401823,
"loss": 0.3118,
"step": 177
},
{
"epoch": 0.3918547055586131,
"grad_norm": 1.8316059112548828,
"learning_rate": 0.00019947430157664576,
"loss": 0.2448,
"step": 178
},
{
"epoch": 0.3940561364887177,
"grad_norm": 1.2774708271026611,
"learning_rate": 0.00019944771261756292,
"loss": 0.1959,
"step": 179
},
{
"epoch": 0.39625756741882223,
"grad_norm": 1.0134649276733398,
"learning_rate": 0.00019942046959164515,
"loss": 0.2163,
"step": 180
},
{
"epoch": 0.3984589983489268,
"grad_norm": 1.4235607385635376,
"learning_rate": 0.00019939257267806963,
"loss": 0.3726,
"step": 181
},
{
"epoch": 0.4006604292790314,
"grad_norm": 0.4046984910964966,
"learning_rate": 0.00019936402206031413,
"loss": 0.0713,
"step": 182
},
{
"epoch": 0.40286186020913595,
"grad_norm": 1.2674638032913208,
"learning_rate": 0.00019933481792615583,
"loss": 0.2618,
"step": 183
},
{
"epoch": 0.4050632911392405,
"grad_norm": 1.141114592552185,
"learning_rate": 0.00019930496046767008,
"loss": 0.3159,
"step": 184
},
{
"epoch": 0.40726472206934505,
"grad_norm": 0.9752854704856873,
"learning_rate": 0.00019927444988122918,
"loss": 0.3026,
"step": 185
},
{
"epoch": 0.40946615299944966,
"grad_norm": 1.7940340042114258,
"learning_rate": 0.00019924328636750102,
"loss": 0.3452,
"step": 186
},
{
"epoch": 0.4116675839295542,
"grad_norm": 1.4885001182556152,
"learning_rate": 0.0001992114701314478,
"loss": 0.2027,
"step": 187
},
{
"epoch": 0.41386901485965877,
"grad_norm": 1.3337723016738892,
"learning_rate": 0.0001991790013823246,
"loss": 0.2417,
"step": 188
},
{
"epoch": 0.4160704457897633,
"grad_norm": 1.3942782878875732,
"learning_rate": 0.00019914588033367823,
"loss": 0.3172,
"step": 189
},
{
"epoch": 0.41827187671986793,
"grad_norm": 0.6452035307884216,
"learning_rate": 0.00019911210720334547,
"loss": 0.1793,
"step": 190
},
{
"epoch": 0.4204733076499725,
"grad_norm": 1.146669864654541,
"learning_rate": 0.000199077682213452,
"loss": 0.2297,
"step": 191
},
{
"epoch": 0.42267473858007704,
"grad_norm": 1.0184296369552612,
"learning_rate": 0.0001990426055904107,
"loss": 0.1563,
"step": 192
},
{
"epoch": 0.42487616951018164,
"grad_norm": 1.3531533479690552,
"learning_rate": 0.0001990068775649202,
"loss": 0.194,
"step": 193
},
{
"epoch": 0.4270776004402862,
"grad_norm": 1.0374277830123901,
"learning_rate": 0.0001989704983719635,
"loss": 0.1652,
"step": 194
},
{
"epoch": 0.42927903137039075,
"grad_norm": 0.8380899429321289,
"learning_rate": 0.00019893346825080625,
"loss": 0.195,
"step": 195
},
{
"epoch": 0.4314804623004953,
"grad_norm": 1.239431619644165,
"learning_rate": 0.00019889578744499532,
"loss": 0.2463,
"step": 196
},
{
"epoch": 0.4336818932305999,
"grad_norm": 1.1668486595153809,
"learning_rate": 0.000198857456202357,
"loss": 0.237,
"step": 197
},
{
"epoch": 0.43588332416070447,
"grad_norm": 1.1140515804290771,
"learning_rate": 0.00019881847477499557,
"loss": 0.2848,
"step": 198
},
{
"epoch": 0.438084755090809,
"grad_norm": 1.3733386993408203,
"learning_rate": 0.00019877884341929164,
"loss": 0.2022,
"step": 199
},
{
"epoch": 0.4402861860209136,
"grad_norm": 1.389896273612976,
"learning_rate": 0.00019873856239590034,
"loss": 0.267,
"step": 200
},
{
"epoch": 0.4424876169510182,
"grad_norm": 1.0052469968795776,
"learning_rate": 0.00019869763196974957,
"loss": 0.2277,
"step": 201
},
{
"epoch": 0.44468904788112273,
"grad_norm": 1.034943699836731,
"learning_rate": 0.00019865605241003848,
"loss": 0.1446,
"step": 202
},
{
"epoch": 0.4468904788112273,
"grad_norm": 0.5385359525680542,
"learning_rate": 0.0001986138239902355,
"loss": 0.0705,
"step": 203
},
{
"epoch": 0.44909190974133184,
"grad_norm": 0.9618946313858032,
"learning_rate": 0.00019857094698807662,
"loss": 0.2932,
"step": 204
},
{
"epoch": 0.45129334067143645,
"grad_norm": 1.1805518865585327,
"learning_rate": 0.00019852742168556355,
"loss": 0.2388,
"step": 205
},
{
"epoch": 0.453494771601541,
"grad_norm": 1.1751959323883057,
"learning_rate": 0.00019848324836896185,
"loss": 0.2859,
"step": 206
},
{
"epoch": 0.45569620253164556,
"grad_norm": 1.0479226112365723,
"learning_rate": 0.00019843842732879914,
"loss": 0.2722,
"step": 207
},
{
"epoch": 0.45789763346175016,
"grad_norm": 1.0439608097076416,
"learning_rate": 0.00019839295885986296,
"loss": 0.1869,
"step": 208
},
{
"epoch": 0.4600990643918547,
"grad_norm": 1.0632809400558472,
"learning_rate": 0.00019834684326119916,
"loss": 0.2354,
"step": 209
},
{
"epoch": 0.46230049532195927,
"grad_norm": 0.8700536489486694,
"learning_rate": 0.00019830008083610964,
"loss": 0.1734,
"step": 210
},
{
"epoch": 0.4645019262520638,
"grad_norm": 1.8080542087554932,
"learning_rate": 0.0001982526718921505,
"loss": 0.411,
"step": 211
},
{
"epoch": 0.46670335718216843,
"grad_norm": 1.38103449344635,
"learning_rate": 0.00019820461674113,
"loss": 0.2625,
"step": 212
},
{
"epoch": 0.468904788112273,
"grad_norm": 1.205725073814392,
"learning_rate": 0.00019815591569910654,
"loss": 0.3933,
"step": 213
},
{
"epoch": 0.47110621904237754,
"grad_norm": 2.5588293075561523,
"learning_rate": 0.00019810656908638644,
"loss": 0.3953,
"step": 214
},
{
"epoch": 0.4733076499724821,
"grad_norm": 1.0547009706497192,
"learning_rate": 0.00019805657722752202,
"loss": 0.1817,
"step": 215
},
{
"epoch": 0.4755090809025867,
"grad_norm": 1.2319700717926025,
"learning_rate": 0.00019800594045130932,
"loss": 0.2348,
"step": 216
},
{
"epoch": 0.47771051183269125,
"grad_norm": 1.148016333580017,
"learning_rate": 0.00019795465909078607,
"loss": 0.201,
"step": 217
},
{
"epoch": 0.4799119427627958,
"grad_norm": 1.0850389003753662,
"learning_rate": 0.0001979027334832293,
"loss": 0.2078,
"step": 218
},
{
"epoch": 0.48211337369290036,
"grad_norm": 1.154354214668274,
"learning_rate": 0.0001978501639701534,
"loss": 0.3007,
"step": 219
},
{
"epoch": 0.48431480462300497,
"grad_norm": 1.7488455772399902,
"learning_rate": 0.00019779695089730765,
"loss": 0.2684,
"step": 220
},
{
"epoch": 0.4865162355531095,
"grad_norm": 2.215618848800659,
"learning_rate": 0.00019774309461467399,
"loss": 0.1546,
"step": 221
},
{
"epoch": 0.4887176664832141,
"grad_norm": 1.2282485961914062,
"learning_rate": 0.00019768859547646478,
"loss": 0.2675,
"step": 222
},
{
"epoch": 0.4909190974133187,
"grad_norm": 0.7270124554634094,
"learning_rate": 0.00019763345384112043,
"loss": 0.1491,
"step": 223
},
{
"epoch": 0.49312052834342324,
"grad_norm": 1.3781054019927979,
"learning_rate": 0.00019757767007130706,
"loss": 0.3164,
"step": 224
},
{
"epoch": 0.4953219592735278,
"grad_norm": 1.1072008609771729,
"learning_rate": 0.00019752124453391404,
"loss": 0.2054,
"step": 225
},
{
"epoch": 0.49752339020363234,
"grad_norm": 1.453762412071228,
"learning_rate": 0.00019746417760005178,
"loss": 0.1959,
"step": 226
},
{
"epoch": 0.49972482113373695,
"grad_norm": 0.9458669424057007,
"learning_rate": 0.00019740646964504902,
"loss": 0.195,
"step": 227
},
{
"epoch": 0.5019262520638414,
"grad_norm": 0.9705711603164673,
"learning_rate": 0.00019734812104845047,
"loss": 0.1799,
"step": 228
},
{
"epoch": 0.5041276829939461,
"grad_norm": 1.1919350624084473,
"learning_rate": 0.0001972891321940145,
"loss": 0.1922,
"step": 229
},
{
"epoch": 0.5063291139240507,
"grad_norm": 1.200522541999817,
"learning_rate": 0.0001972295034697103,
"loss": 0.2551,
"step": 230
},
{
"epoch": 0.5085305448541552,
"grad_norm": 1.3816964626312256,
"learning_rate": 0.00019716923526771552,
"loss": 0.3134,
"step": 231
},
{
"epoch": 0.5107319757842598,
"grad_norm": 0.6731604933738708,
"learning_rate": 0.00019710832798441365,
"loss": 0.1294,
"step": 232
},
{
"epoch": 0.5129334067143644,
"grad_norm": 0.9393294453620911,
"learning_rate": 0.0001970467820203915,
"loss": 0.1441,
"step": 233
},
{
"epoch": 0.5151348376444689,
"grad_norm": 0.743293821811676,
"learning_rate": 0.00019698459778043627,
"loss": 0.1028,
"step": 234
},
{
"epoch": 0.5173362685745735,
"grad_norm": 1.600050449371338,
"learning_rate": 0.00019692177567353332,
"loss": 0.2226,
"step": 235
},
{
"epoch": 0.519537699504678,
"grad_norm": 1.7203608751296997,
"learning_rate": 0.0001968583161128631,
"loss": 0.3033,
"step": 236
},
{
"epoch": 0.5217391304347826,
"grad_norm": 1.3202494382858276,
"learning_rate": 0.00019679421951579874,
"loss": 0.0937,
"step": 237
},
{
"epoch": 0.5239405613648872,
"grad_norm": 1.847307801246643,
"learning_rate": 0.00019672948630390294,
"loss": 0.2683,
"step": 238
},
{
"epoch": 0.5261419922949917,
"grad_norm": 1.322928786277771,
"learning_rate": 0.00019666411690292564,
"loss": 0.2265,
"step": 239
},
{
"epoch": 0.5283434232250963,
"grad_norm": 1.9180538654327393,
"learning_rate": 0.0001965981117428008,
"loss": 0.2164,
"step": 240
},
{
"epoch": 0.5305448541552009,
"grad_norm": 2.2924304008483887,
"learning_rate": 0.00019653147125764392,
"loss": 0.2647,
"step": 241
},
{
"epoch": 0.5327462850853054,
"grad_norm": 0.9613167643547058,
"learning_rate": 0.0001964641958857489,
"loss": 0.1628,
"step": 242
},
{
"epoch": 0.53494771601541,
"grad_norm": 2.0845701694488525,
"learning_rate": 0.00019639628606958533,
"loss": 0.3281,
"step": 243
},
{
"epoch": 0.5371491469455146,
"grad_norm": 1.326499104499817,
"learning_rate": 0.0001963277422557956,
"loss": 0.1656,
"step": 244
},
{
"epoch": 0.5393505778756191,
"grad_norm": 0.9228304028511047,
"learning_rate": 0.00019625856489519184,
"loss": 0.19,
"step": 245
},
{
"epoch": 0.5415520088057237,
"grad_norm": 1.0690181255340576,
"learning_rate": 0.00019618875444275293,
"loss": 0.2124,
"step": 246
},
{
"epoch": 0.5437534397358282,
"grad_norm": 1.492824673652649,
"learning_rate": 0.00019611831135762174,
"loss": 0.1966,
"step": 247
},
{
"epoch": 0.5459548706659328,
"grad_norm": 1.7507821321487427,
"learning_rate": 0.00019604723610310194,
"loss": 0.4592,
"step": 248
},
{
"epoch": 0.5481563015960375,
"grad_norm": 1.523306965827942,
"learning_rate": 0.00019597552914665483,
"loss": 0.3597,
"step": 249
},
{
"epoch": 0.550357732526142,
"grad_norm": 1.8270065784454346,
"learning_rate": 0.0001959031909598966,
"loss": 0.314,
"step": 250
},
{
"epoch": 0.5525591634562466,
"grad_norm": 0.5709218382835388,
"learning_rate": 0.00019583022201859483,
"loss": 0.1128,
"step": 251
},
{
"epoch": 0.5547605943863512,
"grad_norm": 1.1685845851898193,
"learning_rate": 0.00019575662280266578,
"loss": 0.2384,
"step": 252
},
{
"epoch": 0.5569620253164557,
"grad_norm": 0.5500782132148743,
"learning_rate": 0.00019568239379617088,
"loss": 0.1463,
"step": 253
},
{
"epoch": 0.5591634562465603,
"grad_norm": 1.9069963693618774,
"learning_rate": 0.00019560753548731374,
"loss": 0.5131,
"step": 254
},
{
"epoch": 0.5613648871766649,
"grad_norm": 1.0922200679779053,
"learning_rate": 0.0001955320483684369,
"loss": 0.2108,
"step": 255
},
{
"epoch": 0.5635663181067694,
"grad_norm": 1.1500283479690552,
"learning_rate": 0.0001954559329360185,
"loss": 0.2229,
"step": 256
},
{
"epoch": 0.565767749036874,
"grad_norm": 1.501747488975525,
"learning_rate": 0.0001953791896906692,
"loss": 0.1476,
"step": 257
},
{
"epoch": 0.5679691799669785,
"grad_norm": 0.5918743014335632,
"learning_rate": 0.00019530181913712872,
"loss": 0.1434,
"step": 258
},
{
"epoch": 0.5701706108970831,
"grad_norm": 0.9154205322265625,
"learning_rate": 0.00019522382178426258,
"loss": 0.2219,
"step": 259
},
{
"epoch": 0.5723720418271877,
"grad_norm": 1.260572910308838,
"learning_rate": 0.00019514519814505877,
"loss": 0.3396,
"step": 260
},
{
"epoch": 0.5745734727572922,
"grad_norm": 0.9173365235328674,
"learning_rate": 0.00019506594873662433,
"loss": 0.1369,
"step": 261
},
{
"epoch": 0.5767749036873968,
"grad_norm": 1.759440302848816,
"learning_rate": 0.00019498607408018203,
"loss": 0.7267,
"step": 262
},
{
"epoch": 0.5789763346175014,
"grad_norm": 1.300220251083374,
"learning_rate": 0.00019490557470106686,
"loss": 0.3193,
"step": 263
},
{
"epoch": 0.5811777655476059,
"grad_norm": 1.3596434593200684,
"learning_rate": 0.00019482445112872264,
"loss": 0.3193,
"step": 264
},
{
"epoch": 0.5833791964777105,
"grad_norm": 0.9206734299659729,
"learning_rate": 0.00019474270389669844,
"loss": 0.173,
"step": 265
},
{
"epoch": 0.5855806274078151,
"grad_norm": 1.1209437847137451,
"learning_rate": 0.0001946603335426452,
"loss": 0.1857,
"step": 266
},
{
"epoch": 0.5877820583379196,
"grad_norm": 0.7680370211601257,
"learning_rate": 0.00019457734060831204,
"loss": 0.2384,
"step": 267
},
{
"epoch": 0.5899834892680242,
"grad_norm": 1.1541163921356201,
"learning_rate": 0.00019449372563954293,
"loss": 0.2907,
"step": 268
},
{
"epoch": 0.5921849201981287,
"grad_norm": 0.7708389759063721,
"learning_rate": 0.00019440948918627283,
"loss": 0.1215,
"step": 269
},
{
"epoch": 0.5943863511282333,
"grad_norm": 0.7209833860397339,
"learning_rate": 0.00019432463180252423,
"loss": 0.2057,
"step": 270
},
{
"epoch": 0.596587782058338,
"grad_norm": 1.0058317184448242,
"learning_rate": 0.0001942391540464035,
"loss": 0.316,
"step": 271
},
{
"epoch": 0.5987892129884425,
"grad_norm": 0.912278950214386,
"learning_rate": 0.00019415305648009715,
"loss": 0.2154,
"step": 272
},
{
"epoch": 0.6009906439185471,
"grad_norm": 0.9229905009269714,
"learning_rate": 0.00019406633966986828,
"loss": 0.1988,
"step": 273
},
{
"epoch": 0.6031920748486517,
"grad_norm": 0.9416797161102295,
"learning_rate": 0.00019397900418605257,
"loss": 0.2454,
"step": 274
},
{
"epoch": 0.6053935057787562,
"grad_norm": 0.7916048765182495,
"learning_rate": 0.0001938910506030549,
"loss": 0.1595,
"step": 275
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.7961133718490601,
"learning_rate": 0.00019380247949934534,
"loss": 0.3149,
"step": 276
},
{
"epoch": 0.6097963676389653,
"grad_norm": 2.367426633834839,
"learning_rate": 0.00019371329145745524,
"loss": 0.4522,
"step": 277
},
{
"epoch": 0.6119977985690699,
"grad_norm": 1.6261329650878906,
"learning_rate": 0.00019362348706397373,
"loss": 0.3005,
"step": 278
},
{
"epoch": 0.6141992294991745,
"grad_norm": 0.9070632457733154,
"learning_rate": 0.0001935330669095436,
"loss": 0.2101,
"step": 279
},
{
"epoch": 0.616400660429279,
"grad_norm": 1.2766557931900024,
"learning_rate": 0.0001934420315888575,
"loss": 0.2502,
"step": 280
},
{
"epoch": 0.6186020913593836,
"grad_norm": 1.20702064037323,
"learning_rate": 0.00019335038170065403,
"loss": 0.2563,
"step": 281
},
{
"epoch": 0.6208035222894882,
"grad_norm": 1.1549800634384155,
"learning_rate": 0.0001932581178477138,
"loss": 0.1833,
"step": 282
},
{
"epoch": 0.6230049532195927,
"grad_norm": 1.2302500009536743,
"learning_rate": 0.0001931652406368554,
"loss": 0.226,
"step": 283
},
{
"epoch": 0.6252063841496973,
"grad_norm": 1.1015033721923828,
"learning_rate": 0.00019307175067893165,
"loss": 0.3038,
"step": 284
},
{
"epoch": 0.6274078150798019,
"grad_norm": 1.1137440204620361,
"learning_rate": 0.00019297764858882514,
"loss": 0.2073,
"step": 285
},
{
"epoch": 0.6296092460099064,
"grad_norm": 1.1460460424423218,
"learning_rate": 0.00019288293498544469,
"loss": 0.4567,
"step": 286
},
{
"epoch": 0.631810676940011,
"grad_norm": 1.245483636856079,
"learning_rate": 0.00019278761049172088,
"loss": 0.3153,
"step": 287
},
{
"epoch": 0.6340121078701155,
"grad_norm": 0.587056040763855,
"learning_rate": 0.0001926916757346022,
"loss": 0.0915,
"step": 288
},
{
"epoch": 0.6362135388002201,
"grad_norm": 0.8686281442642212,
"learning_rate": 0.00019259513134505077,
"loss": 0.1085,
"step": 289
},
{
"epoch": 0.6384149697303247,
"grad_norm": 1.245270848274231,
"learning_rate": 0.0001924979779580383,
"loss": 0.2146,
"step": 290
},
{
"epoch": 0.6406164006604292,
"grad_norm": 0.9802015423774719,
"learning_rate": 0.00019240021621254186,
"loss": 0.3619,
"step": 291
},
{
"epoch": 0.6428178315905339,
"grad_norm": 1.0316482782363892,
"learning_rate": 0.00019230184675153976,
"loss": 0.5368,
"step": 292
},
{
"epoch": 0.6450192625206385,
"grad_norm": 1.010973334312439,
"learning_rate": 0.00019220287022200707,
"loss": 0.1715,
"step": 293
},
{
"epoch": 0.647220693450743,
"grad_norm": 1.4453409910202026,
"learning_rate": 0.00019210328727491173,
"loss": 0.1925,
"step": 294
},
{
"epoch": 0.6494221243808476,
"grad_norm": 1.1811559200286865,
"learning_rate": 0.00019200309856520997,
"loss": 0.2017,
"step": 295
},
{
"epoch": 0.6516235553109522,
"grad_norm": 0.8928061127662659,
"learning_rate": 0.00019190230475184223,
"loss": 0.2713,
"step": 296
},
{
"epoch": 0.6538249862410567,
"grad_norm": 1.332140326499939,
"learning_rate": 0.00019180090649772855,
"loss": 0.2451,
"step": 297
},
{
"epoch": 0.6560264171711613,
"grad_norm": 1.1139875650405884,
"learning_rate": 0.00019169890446976454,
"loss": 0.2006,
"step": 298
},
{
"epoch": 0.6582278481012658,
"grad_norm": 1.582215666770935,
"learning_rate": 0.00019159629933881666,
"loss": 0.2862,
"step": 299
},
{
"epoch": 0.6604292790313704,
"grad_norm": 0.647504448890686,
"learning_rate": 0.00019149309177971815,
"loss": 0.0764,
"step": 300
},
{
"epoch": 0.662630709961475,
"grad_norm": 2.3032169342041016,
"learning_rate": 0.00019138928247126424,
"loss": 0.3415,
"step": 301
},
{
"epoch": 0.6648321408915795,
"grad_norm": 1.4954179525375366,
"learning_rate": 0.00019128487209620803,
"loss": 0.3911,
"step": 302
},
{
"epoch": 0.6670335718216841,
"grad_norm": 1.450543999671936,
"learning_rate": 0.0001911798613412557,
"loss": 0.2685,
"step": 303
},
{
"epoch": 0.6692350027517887,
"grad_norm": 1.65657639503479,
"learning_rate": 0.0001910742508970622,
"loss": 0.343,
"step": 304
},
{
"epoch": 0.6714364336818932,
"grad_norm": 2.165513515472412,
"learning_rate": 0.00019096804145822655,
"loss": 0.3681,
"step": 305
},
{
"epoch": 0.6736378646119978,
"grad_norm": 1.0061649084091187,
"learning_rate": 0.00019086123372328746,
"loss": 0.1908,
"step": 306
},
{
"epoch": 0.6758392955421023,
"grad_norm": 0.8672468066215515,
"learning_rate": 0.00019075382839471857,
"loss": 0.167,
"step": 307
},
{
"epoch": 0.6780407264722069,
"grad_norm": 1.4202598333358765,
"learning_rate": 0.0001906458261789238,
"loss": 0.1973,
"step": 308
},
{
"epoch": 0.6802421574023115,
"grad_norm": 1.4729435443878174,
"learning_rate": 0.00019053722778623305,
"loss": 0.3278,
"step": 309
},
{
"epoch": 0.682443588332416,
"grad_norm": 1.2635149955749512,
"learning_rate": 0.000190428033930897,
"loss": 0.2677,
"step": 310
},
{
"epoch": 0.6846450192625206,
"grad_norm": 1.1371662616729736,
"learning_rate": 0.0001903182453310828,
"loss": 0.3166,
"step": 311
},
{
"epoch": 0.6868464501926252,
"grad_norm": 0.8570672869682312,
"learning_rate": 0.0001902078627088692,
"loss": 0.2779,
"step": 312
},
{
"epoch": 0.6890478811227297,
"grad_norm": 0.9639473557472229,
"learning_rate": 0.0001900968867902419,
"loss": 0.1461,
"step": 313
},
{
"epoch": 0.6912493120528344,
"grad_norm": 1.0219823122024536,
"learning_rate": 0.00018998531830508869,
"loss": 0.259,
"step": 314
},
{
"epoch": 0.693450742982939,
"grad_norm": 0.8559497594833374,
"learning_rate": 0.00018987315798719452,
"loss": 0.2074,
"step": 315
},
{
"epoch": 0.6956521739130435,
"grad_norm": 1.8177236318588257,
"learning_rate": 0.000189760406574237,
"loss": 0.3885,
"step": 316
},
{
"epoch": 0.6978536048431481,
"grad_norm": 1.7845747470855713,
"learning_rate": 0.0001896470648077813,
"loss": 0.3579,
"step": 317
},
{
"epoch": 0.7000550357732526,
"grad_norm": 0.8975507020950317,
"learning_rate": 0.0001895331334332753,
"loss": 0.2452,
"step": 318
},
{
"epoch": 0.7022564667033572,
"grad_norm": 0.9428125023841858,
"learning_rate": 0.00018941861320004484,
"loss": 0.2423,
"step": 319
},
{
"epoch": 0.7044578976334618,
"grad_norm": 1.159013032913208,
"learning_rate": 0.00018930350486128856,
"loss": 0.4047,
"step": 320
},
{
"epoch": 0.7066593285635663,
"grad_norm": 0.8494213223457336,
"learning_rate": 0.00018918780917407307,
"loss": 0.2341,
"step": 321
},
{
"epoch": 0.7088607594936709,
"grad_norm": 1.068804144859314,
"learning_rate": 0.0001890715268993281,
"loss": 0.3007,
"step": 322
},
{
"epoch": 0.7110621904237755,
"grad_norm": 0.43914756178855896,
"learning_rate": 0.0001889546588018412,
"loss": 0.0936,
"step": 323
},
{
"epoch": 0.71326362135388,
"grad_norm": 0.7797726392745972,
"learning_rate": 0.00018883720565025295,
"loss": 0.2057,
"step": 324
},
{
"epoch": 0.7154650522839846,
"grad_norm": 0.7800988554954529,
"learning_rate": 0.00018871916821705184,
"loss": 0.2073,
"step": 325
},
{
"epoch": 0.7176664832140892,
"grad_norm": 0.5828583240509033,
"learning_rate": 0.00018860054727856915,
"loss": 0.1515,
"step": 326
},
{
"epoch": 0.7198679141441937,
"grad_norm": 1.471644401550293,
"learning_rate": 0.00018848134361497385,
"loss": 0.3854,
"step": 327
},
{
"epoch": 0.7220693450742983,
"grad_norm": 1.2705450057983398,
"learning_rate": 0.00018836155801026753,
"loss": 0.2625,
"step": 328
},
{
"epoch": 0.7242707760044028,
"grad_norm": 1.0784512758255005,
"learning_rate": 0.0001882411912522792,
"loss": 0.1387,
"step": 329
},
{
"epoch": 0.7264722069345074,
"grad_norm": 1.1987195014953613,
"learning_rate": 0.00018812024413266006,
"loss": 0.2613,
"step": 330
},
{
"epoch": 0.728673637864612,
"grad_norm": 0.9083921909332275,
"learning_rate": 0.0001879987174468784,
"loss": 0.1633,
"step": 331
},
{
"epoch": 0.7308750687947165,
"grad_norm": 1.1217319965362549,
"learning_rate": 0.00018787661199421427,
"loss": 0.158,
"step": 332
},
{
"epoch": 0.7330764997248211,
"grad_norm": 0.7454916834831238,
"learning_rate": 0.00018775392857775432,
"loss": 0.1587,
"step": 333
},
{
"epoch": 0.7352779306549257,
"grad_norm": 0.8220398426055908,
"learning_rate": 0.00018763066800438636,
"loss": 0.1583,
"step": 334
},
{
"epoch": 0.7374793615850302,
"grad_norm": 0.908816397190094,
"learning_rate": 0.00018750683108479432,
"loss": 0.2706,
"step": 335
},
{
"epoch": 0.7396807925151349,
"grad_norm": 1.1175532341003418,
"learning_rate": 0.0001873824186334526,
"loss": 0.241,
"step": 336
},
{
"epoch": 0.7418822234452394,
"grad_norm": 0.8782237768173218,
"learning_rate": 0.00018725743146862095,
"loss": 0.2187,
"step": 337
},
{
"epoch": 0.744083654375344,
"grad_norm": 1.36802339553833,
"learning_rate": 0.00018713187041233896,
"loss": 0.2422,
"step": 338
},
{
"epoch": 0.7462850853054486,
"grad_norm": 1.4567631483078003,
"learning_rate": 0.00018700573629042074,
"loss": 0.3323,
"step": 339
},
{
"epoch": 0.7484865162355531,
"grad_norm": 0.8530279994010925,
"learning_rate": 0.00018687902993244943,
"loss": 0.2511,
"step": 340
},
{
"epoch": 0.7506879471656577,
"grad_norm": 1.4750181436538696,
"learning_rate": 0.00018675175217177175,
"loss": 0.334,
"step": 341
},
{
"epoch": 0.7528893780957623,
"grad_norm": 0.5695929527282715,
"learning_rate": 0.0001866239038454926,
"loss": 0.0692,
"step": 342
},
{
"epoch": 0.7550908090258668,
"grad_norm": 1.2884031534194946,
"learning_rate": 0.00018649548579446936,
"loss": 0.3052,
"step": 343
},
{
"epoch": 0.7572922399559714,
"grad_norm": 1.2000869512557983,
"learning_rate": 0.0001863664988633066,
"loss": 0.1638,
"step": 344
},
{
"epoch": 0.759493670886076,
"grad_norm": 1.3409312963485718,
"learning_rate": 0.00018623694390035035,
"loss": 0.2515,
"step": 345
},
{
"epoch": 0.7616951018161805,
"grad_norm": 0.6578596234321594,
"learning_rate": 0.00018610682175768258,
"loss": 0.1144,
"step": 346
},
{
"epoch": 0.7638965327462851,
"grad_norm": 1.0106149911880493,
"learning_rate": 0.00018597613329111566,
"loss": 0.1943,
"step": 347
},
{
"epoch": 0.7660979636763896,
"grad_norm": 1.586403489112854,
"learning_rate": 0.00018584487936018661,
"loss": 0.1804,
"step": 348
},
{
"epoch": 0.7682993946064942,
"grad_norm": 1.031691074371338,
"learning_rate": 0.0001857130608281515,
"loss": 0.419,
"step": 349
},
{
"epoch": 0.7705008255365988,
"grad_norm": 1.0916547775268555,
"learning_rate": 0.0001855806785619798,
"loss": 0.1373,
"step": 350
},
{
"epoch": 0.7727022564667033,
"grad_norm": 2.66514253616333,
"learning_rate": 0.00018544773343234862,
"loss": 0.2008,
"step": 351
},
{
"epoch": 0.7749036873968079,
"grad_norm": 1.0321251153945923,
"learning_rate": 0.00018531422631363706,
"loss": 0.168,
"step": 352
},
{
"epoch": 0.7771051183269125,
"grad_norm": 1.2961331605911255,
"learning_rate": 0.00018518015808392045,
"loss": 0.2361,
"step": 353
},
{
"epoch": 0.779306549257017,
"grad_norm": 0.719615638256073,
"learning_rate": 0.0001850455296249644,
"loss": 0.1776,
"step": 354
},
{
"epoch": 0.7815079801871216,
"grad_norm": 1.1901960372924805,
"learning_rate": 0.0001849103418222194,
"loss": 0.2842,
"step": 355
},
{
"epoch": 0.7837094111172263,
"grad_norm": 0.9724762439727783,
"learning_rate": 0.00018477459556481441,
"loss": 0.2397,
"step": 356
},
{
"epoch": 0.7859108420473307,
"grad_norm": 1.761529803276062,
"learning_rate": 0.0001846382917455516,
"loss": 0.313,
"step": 357
},
{
"epoch": 0.7881122729774354,
"grad_norm": 1.0816775560379028,
"learning_rate": 0.00018450143126090015,
"loss": 0.1472,
"step": 358
},
{
"epoch": 0.7903137039075399,
"grad_norm": 2.434422016143799,
"learning_rate": 0.00018436401501099035,
"loss": 0.177,
"step": 359
},
{
"epoch": 0.7925151348376445,
"grad_norm": 1.8141069412231445,
"learning_rate": 0.00018422604389960783,
"loss": 0.375,
"step": 360
},
{
"epoch": 0.7947165657677491,
"grad_norm": 1.1976251602172852,
"learning_rate": 0.00018408751883418754,
"loss": 0.2226,
"step": 361
},
{
"epoch": 0.7969179966978536,
"grad_norm": 1.4153374433517456,
"learning_rate": 0.00018394844072580773,
"loss": 0.3456,
"step": 362
},
{
"epoch": 0.7991194276279582,
"grad_norm": 1.5557993650436401,
"learning_rate": 0.00018380881048918405,
"loss": 0.2763,
"step": 363
},
{
"epoch": 0.8013208585580628,
"grad_norm": 1.2922636270523071,
"learning_rate": 0.00018366862904266356,
"loss": 0.3598,
"step": 364
},
{
"epoch": 0.8035222894881673,
"grad_norm": 1.2480388879776,
"learning_rate": 0.00018352789730821843,
"loss": 0.3873,
"step": 365
},
{
"epoch": 0.8057237204182719,
"grad_norm": 0.9546335339546204,
"learning_rate": 0.00018338661621144025,
"loss": 0.1378,
"step": 366
},
{
"epoch": 0.8079251513483764,
"grad_norm": 1.3006221055984497,
"learning_rate": 0.00018324478668153366,
"loss": 0.2566,
"step": 367
},
{
"epoch": 0.810126582278481,
"grad_norm": 1.2144070863723755,
"learning_rate": 0.00018310240965131041,
"loss": 0.2278,
"step": 368
},
{
"epoch": 0.8123280132085856,
"grad_norm": 0.8332474827766418,
"learning_rate": 0.00018295948605718314,
"loss": 0.2634,
"step": 369
},
{
"epoch": 0.8145294441386901,
"grad_norm": 0.7528755068778992,
"learning_rate": 0.00018281601683915917,
"loss": 0.2693,
"step": 370
},
{
"epoch": 0.8167308750687947,
"grad_norm": 1.0342864990234375,
"learning_rate": 0.00018267200294083447,
"loss": 0.2044,
"step": 371
},
{
"epoch": 0.8189323059988993,
"grad_norm": 0.8297427296638489,
"learning_rate": 0.00018252744530938736,
"loss": 0.1956,
"step": 372
},
{
"epoch": 0.8211337369290038,
"grad_norm": 1.282098412513733,
"learning_rate": 0.00018238234489557215,
"loss": 0.2616,
"step": 373
},
{
"epoch": 0.8233351678591084,
"grad_norm": 0.6476250886917114,
"learning_rate": 0.0001822367026537133,
"loss": 0.1147,
"step": 374
},
{
"epoch": 0.825536598789213,
"grad_norm": 0.6506133079528809,
"learning_rate": 0.00018209051954169855,
"loss": 0.1762,
"step": 375
},
{
"epoch": 0.8277380297193175,
"grad_norm": 0.926846444606781,
"learning_rate": 0.0001819437965209732,
"loss": 0.1866,
"step": 376
},
{
"epoch": 0.8299394606494221,
"grad_norm": 0.9617591500282288,
"learning_rate": 0.0001817965345565334,
"loss": 0.2098,
"step": 377
},
{
"epoch": 0.8321408915795266,
"grad_norm": 1.071401834487915,
"learning_rate": 0.00018164873461691986,
"loss": 0.2428,
"step": 378
},
{
"epoch": 0.8343423225096313,
"grad_norm": 0.654083788394928,
"learning_rate": 0.00018150039767421178,
"loss": 0.0716,
"step": 379
},
{
"epoch": 0.8365437534397359,
"grad_norm": 1.6268547773361206,
"learning_rate": 0.00018135152470401996,
"loss": 0.2735,
"step": 380
},
{
"epoch": 0.8387451843698404,
"grad_norm": 0.4768295884132385,
"learning_rate": 0.00018120211668548088,
"loss": 0.0739,
"step": 381
},
{
"epoch": 0.840946615299945,
"grad_norm": 1.295459508895874,
"learning_rate": 0.00018105217460124981,
"loss": 0.2342,
"step": 382
},
{
"epoch": 0.8431480462300496,
"grad_norm": 0.7535791397094727,
"learning_rate": 0.00018090169943749476,
"loss": 0.1549,
"step": 383
},
{
"epoch": 0.8453494771601541,
"grad_norm": 1.5335655212402344,
"learning_rate": 0.00018075069218388964,
"loss": 0.1621,
"step": 384
},
{
"epoch": 0.8475509080902587,
"grad_norm": 1.2511568069458008,
"learning_rate": 0.00018059915383360808,
"loss": 0.1517,
"step": 385
},
{
"epoch": 0.8497523390203633,
"grad_norm": 0.8873250484466553,
"learning_rate": 0.00018044708538331655,
"loss": 0.2602,
"step": 386
},
{
"epoch": 0.8519537699504678,
"grad_norm": 1.4157307147979736,
"learning_rate": 0.00018029448783316813,
"loss": 0.4575,
"step": 387
},
{
"epoch": 0.8541552008805724,
"grad_norm": 1.280274510383606,
"learning_rate": 0.00018014136218679567,
"loss": 0.1614,
"step": 388
},
{
"epoch": 0.8563566318106769,
"grad_norm": 0.8422673940658569,
"learning_rate": 0.0001799877094513054,
"loss": 0.1639,
"step": 389
},
{
"epoch": 0.8585580627407815,
"grad_norm": 0.9522169232368469,
"learning_rate": 0.00017983353063727016,
"loss": 0.1054,
"step": 390
},
{
"epoch": 0.8607594936708861,
"grad_norm": 1.752172589302063,
"learning_rate": 0.00017967882675872277,
"loss": 0.2285,
"step": 391
},
{
"epoch": 0.8629609246009906,
"grad_norm": 1.322982907295227,
"learning_rate": 0.00017952359883314955,
"loss": 0.1765,
"step": 392
},
{
"epoch": 0.8651623555310952,
"grad_norm": 0.9844683408737183,
"learning_rate": 0.00017936784788148328,
"loss": 0.1054,
"step": 393
},
{
"epoch": 0.8673637864611998,
"grad_norm": 1.8924535512924194,
"learning_rate": 0.00017921157492809672,
"loss": 0.2393,
"step": 394
},
{
"epoch": 0.8695652173913043,
"grad_norm": 1.1759653091430664,
"learning_rate": 0.000179054781000796,
"loss": 0.2491,
"step": 395
},
{
"epoch": 0.8717666483214089,
"grad_norm": 1.2846308946609497,
"learning_rate": 0.00017889746713081343,
"loss": 0.1594,
"step": 396
},
{
"epoch": 0.8739680792515134,
"grad_norm": 0.7170493602752686,
"learning_rate": 0.00017873963435280121,
"loss": 0.228,
"step": 397
},
{
"epoch": 0.876169510181618,
"grad_norm": 0.9288744926452637,
"learning_rate": 0.00017858128370482426,
"loss": 0.1143,
"step": 398
},
{
"epoch": 0.8783709411117226,
"grad_norm": 1.3167988061904907,
"learning_rate": 0.00017842241622835356,
"loss": 0.1729,
"step": 399
},
{
"epoch": 0.8805723720418271,
"grad_norm": 1.0657830238342285,
"learning_rate": 0.00017826303296825926,
"loss": 0.286,
"step": 400
},
{
"epoch": 0.8827738029719318,
"grad_norm": 0.8168554902076721,
"learning_rate": 0.0001781031349728039,
"loss": 0.2013,
"step": 401
},
{
"epoch": 0.8849752339020364,
"grad_norm": 0.882465660572052,
"learning_rate": 0.00017794272329363527,
"loss": 0.1545,
"step": 402
},
{
"epoch": 0.8871766648321409,
"grad_norm": 0.9866340160369873,
"learning_rate": 0.00017778179898577973,
"loss": 0.1833,
"step": 403
},
{
"epoch": 0.8893780957622455,
"grad_norm": 0.7578817009925842,
"learning_rate": 0.00017762036310763532,
"loss": 0.1541,
"step": 404
},
{
"epoch": 0.8915795266923501,
"grad_norm": 0.947085976600647,
"learning_rate": 0.00017745841672096442,
"loss": 0.2227,
"step": 405
},
{
"epoch": 0.8937809576224546,
"grad_norm": 1.0397213697433472,
"learning_rate": 0.0001772959608908873,
"loss": 0.1625,
"step": 406
},
{
"epoch": 0.8959823885525592,
"grad_norm": 1.140871286392212,
"learning_rate": 0.00017713299668587458,
"loss": 0.2139,
"step": 407
},
{
"epoch": 0.8981838194826637,
"grad_norm": 1.7579021453857422,
"learning_rate": 0.00017696952517774062,
"loss": 0.3054,
"step": 408
},
{
"epoch": 0.9003852504127683,
"grad_norm": 1.5178894996643066,
"learning_rate": 0.00017680554744163625,
"loss": 0.2656,
"step": 409
},
{
"epoch": 0.9025866813428729,
"grad_norm": 1.9655121564865112,
"learning_rate": 0.00017664106455604178,
"loss": 0.3215,
"step": 410
},
{
"epoch": 0.9047881122729774,
"grad_norm": 1.8185651302337646,
"learning_rate": 0.00017647607760275987,
"loss": 0.3021,
"step": 411
},
{
"epoch": 0.906989543203082,
"grad_norm": 0.8997313976287842,
"learning_rate": 0.0001763105876669084,
"loss": 0.2806,
"step": 412
},
{
"epoch": 0.9091909741331866,
"grad_norm": 1.726635456085205,
"learning_rate": 0.00017614459583691346,
"loss": 0.2853,
"step": 413
},
{
"epoch": 0.9113924050632911,
"grad_norm": 1.0895217657089233,
"learning_rate": 0.000175978103204502,
"loss": 0.28,
"step": 414
},
{
"epoch": 0.9135938359933957,
"grad_norm": 1.582118034362793,
"learning_rate": 0.00017581111086469476,
"loss": 0.2495,
"step": 415
},
{
"epoch": 0.9157952669235003,
"grad_norm": 1.6275440454483032,
"learning_rate": 0.00017564361991579904,
"loss": 0.2424,
"step": 416
},
{
"epoch": 0.9179966978536048,
"grad_norm": 1.0979431867599487,
"learning_rate": 0.00017547563145940158,
"loss": 0.1526,
"step": 417
},
{
"epoch": 0.9201981287837094,
"grad_norm": 0.8231645226478577,
"learning_rate": 0.00017530714660036112,
"loss": 0.1641,
"step": 418
},
{
"epoch": 0.9223995597138139,
"grad_norm": 0.8364385366439819,
"learning_rate": 0.00017513816644680125,
"loss": 0.2394,
"step": 419
},
{
"epoch": 0.9246009906439185,
"grad_norm": 0.9016224145889282,
"learning_rate": 0.00017496869211010315,
"loss": 0.2125,
"step": 420
},
{
"epoch": 0.9268024215740231,
"grad_norm": 1.0510677099227905,
"learning_rate": 0.00017479872470489824,
"loss": 0.1643,
"step": 421
},
{
"epoch": 0.9290038525041276,
"grad_norm": 0.7203924059867859,
"learning_rate": 0.0001746282653490608,
"loss": 0.1425,
"step": 422
},
{
"epoch": 0.9312052834342323,
"grad_norm": 1.231278657913208,
"learning_rate": 0.0001744573151637007,
"loss": 0.2488,
"step": 423
},
{
"epoch": 0.9334067143643369,
"grad_norm": 1.0261121988296509,
"learning_rate": 0.000174285875273156,
"loss": 0.2443,
"step": 424
},
{
"epoch": 0.9356081452944414,
"grad_norm": 1.9562700986862183,
"learning_rate": 0.0001741139468049855,
"loss": 0.1857,
"step": 425
},
{
"epoch": 0.937809576224546,
"grad_norm": 0.8829732537269592,
"learning_rate": 0.0001739415308899614,
"loss": 0.1966,
"step": 426
},
{
"epoch": 0.9400110071546505,
"grad_norm": 1.1366294622421265,
"learning_rate": 0.00017376862866206187,
"loss": 0.195,
"step": 427
},
{
"epoch": 0.9422124380847551,
"grad_norm": 1.005967617034912,
"learning_rate": 0.0001735952412584635,
"loss": 0.2883,
"step": 428
},
{
"epoch": 0.9444138690148597,
"grad_norm": 1.5798931121826172,
"learning_rate": 0.0001734213698195339,
"loss": 0.2963,
"step": 429
},
{
"epoch": 0.9466152999449642,
"grad_norm": 1.1144102811813354,
"learning_rate": 0.00017324701548882422,
"loss": 0.2546,
"step": 430
},
{
"epoch": 0.9488167308750688,
"grad_norm": 3.013399600982666,
"learning_rate": 0.00017307217941306145,
"loss": 0.3737,
"step": 431
},
{
"epoch": 0.9510181618051734,
"grad_norm": 1.1589728593826294,
"learning_rate": 0.00017289686274214118,
"loss": 0.124,
"step": 432
},
{
"epoch": 0.9532195927352779,
"grad_norm": 0.958777666091919,
"learning_rate": 0.00017272106662911973,
"loss": 0.1727,
"step": 433
},
{
"epoch": 0.9554210236653825,
"grad_norm": 1.4360767602920532,
"learning_rate": 0.00017254479223020684,
"loss": 0.2203,
"step": 434
},
{
"epoch": 0.9576224545954871,
"grad_norm": 1.6432348489761353,
"learning_rate": 0.00017236804070475787,
"loss": 0.16,
"step": 435
},
{
"epoch": 0.9598238855255916,
"grad_norm": 1.4297266006469727,
"learning_rate": 0.00017219081321526617,
"loss": 0.2383,
"step": 436
},
{
"epoch": 0.9620253164556962,
"grad_norm": 1.1295509338378906,
"learning_rate": 0.00017201311092735565,
"loss": 0.2701,
"step": 437
},
{
"epoch": 0.9642267473858007,
"grad_norm": 0.958570122718811,
"learning_rate": 0.00017183493500977278,
"loss": 0.1856,
"step": 438
},
{
"epoch": 0.9664281783159053,
"grad_norm": 0.8746889233589172,
"learning_rate": 0.00017165628663437922,
"loss": 0.2212,
"step": 439
},
{
"epoch": 0.9686296092460099,
"grad_norm": 1.3758814334869385,
"learning_rate": 0.000171477166976144,
"loss": 0.2748,
"step": 440
},
{
"epoch": 0.9708310401761144,
"grad_norm": 1.5859287977218628,
"learning_rate": 0.00017129757721313568,
"loss": 0.3003,
"step": 441
},
{
"epoch": 0.973032471106219,
"grad_norm": 0.7726479172706604,
"learning_rate": 0.00017111751852651473,
"loss": 0.2363,
"step": 442
},
{
"epoch": 0.9752339020363237,
"grad_norm": 1.1339430809020996,
"learning_rate": 0.0001709369921005258,
"loss": 0.2288,
"step": 443
},
{
"epoch": 0.9774353329664281,
"grad_norm": 1.4856394529342651,
"learning_rate": 0.0001707559991224897,
"loss": 0.4531,
"step": 444
},
{
"epoch": 0.9796367638965328,
"grad_norm": 1.0055327415466309,
"learning_rate": 0.0001705745407827959,
"loss": 0.2731,
"step": 445
},
{
"epoch": 0.9818381948266374,
"grad_norm": 0.8089772462844849,
"learning_rate": 0.0001703926182748945,
"loss": 0.2677,
"step": 446
},
{
"epoch": 0.9840396257567419,
"grad_norm": 0.907011866569519,
"learning_rate": 0.0001702102327952884,
"loss": 0.1726,
"step": 447
},
{
"epoch": 0.9862410566868465,
"grad_norm": 0.7658312916755676,
"learning_rate": 0.00017002738554352552,
"loss": 0.2949,
"step": 448
},
{
"epoch": 0.988442487616951,
"grad_norm": 1.8704255819320679,
"learning_rate": 0.00016984407772219077,
"loss": 0.1561,
"step": 449
},
{
"epoch": 0.9906439185470556,
"grad_norm": 0.9276111125946045,
"learning_rate": 0.00016966031053689828,
"loss": 0.2613,
"step": 450
},
{
"epoch": 0.9928453494771602,
"grad_norm": 1.162001132965088,
"learning_rate": 0.00016947608519628343,
"loss": 0.2439,
"step": 451
},
{
"epoch": 0.9950467804072647,
"grad_norm": 0.7540222406387329,
"learning_rate": 0.00016929140291199483,
"loss": 0.1343,
"step": 452
},
{
"epoch": 0.9972482113373693,
"grad_norm": 1.1655153036117554,
"learning_rate": 0.00016910626489868649,
"loss": 0.3202,
"step": 453
},
{
"epoch": 0.9994496422674739,
"grad_norm": 1.484333872795105,
"learning_rate": 0.00016892067237400958,
"loss": 0.2112,
"step": 454
},
{
"epoch": 1.0,
"grad_norm": 0.4983161985874176,
"learning_rate": 0.0001687346265586048,
"loss": 0.0172,
"step": 455
}
],
"logging_steps": 1,
"max_steps": 1362,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 4.2085012748443046e+17,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}