gr00t_initial_ft2 / trainer_state.json
hiroyukikaneko's picture
Upload trainer_state.json with huggingface_hub
ea1ad33 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 20.0,
"eval_steps": 500,
"global_step": 2000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.1,
"grad_norm": 8.593878746032715,
"learning_rate": 1e-05,
"loss": 1.4852,
"step": 10
},
{
"epoch": 0.2,
"grad_norm": 7.231457233428955,
"learning_rate": 2e-05,
"loss": 1.4207,
"step": 20
},
{
"epoch": 0.3,
"grad_norm": 3.319319248199463,
"learning_rate": 3e-05,
"loss": 1.0614,
"step": 30
},
{
"epoch": 0.4,
"grad_norm": 3.60310435295105,
"learning_rate": 4e-05,
"loss": 0.7256,
"step": 40
},
{
"epoch": 0.5,
"grad_norm": 15.761638641357422,
"learning_rate": 5e-05,
"loss": 0.5587,
"step": 50
},
{
"epoch": 0.6,
"grad_norm": 24.317516326904297,
"learning_rate": 6e-05,
"loss": 0.5146,
"step": 60
},
{
"epoch": 0.7,
"grad_norm": 4.110610008239746,
"learning_rate": 7e-05,
"loss": 0.4222,
"step": 70
},
{
"epoch": 0.8,
"grad_norm": 4.141885757446289,
"learning_rate": 8e-05,
"loss": 0.4355,
"step": 80
},
{
"epoch": 0.9,
"grad_norm": 3.7163727283477783,
"learning_rate": 9e-05,
"loss": 0.3939,
"step": 90
},
{
"epoch": 1.0,
"grad_norm": 8.189441680908203,
"learning_rate": 0.0001,
"loss": 0.3712,
"step": 100
},
{
"epoch": 1.1,
"grad_norm": 2.9945340156555176,
"learning_rate": 9.999316524962345e-05,
"loss": 0.3548,
"step": 110
},
{
"epoch": 1.2,
"grad_norm": 1.997642159461975,
"learning_rate": 9.997266286704631e-05,
"loss": 0.2898,
"step": 120
},
{
"epoch": 1.3,
"grad_norm": 1.9837431907653809,
"learning_rate": 9.993849845741524e-05,
"loss": 0.2528,
"step": 130
},
{
"epoch": 1.4,
"grad_norm": 2.294177532196045,
"learning_rate": 9.989068136093873e-05,
"loss": 0.2187,
"step": 140
},
{
"epoch": 1.5,
"grad_norm": 1.4083565473556519,
"learning_rate": 9.98292246503335e-05,
"loss": 0.1986,
"step": 150
},
{
"epoch": 1.6,
"grad_norm": 7.1001811027526855,
"learning_rate": 9.975414512725057e-05,
"loss": 0.1738,
"step": 160
},
{
"epoch": 1.7,
"grad_norm": 1.4144936800003052,
"learning_rate": 9.966546331768191e-05,
"loss": 0.1537,
"step": 170
},
{
"epoch": 1.8,
"grad_norm": 1.5691213607788086,
"learning_rate": 9.956320346634876e-05,
"loss": 0.1519,
"step": 180
},
{
"epoch": 1.9,
"grad_norm": 1.2897148132324219,
"learning_rate": 9.944739353007344e-05,
"loss": 0.1287,
"step": 190
},
{
"epoch": 2.0,
"grad_norm": 1.6297178268432617,
"learning_rate": 9.931806517013612e-05,
"loss": 0.1169,
"step": 200
},
{
"epoch": 2.1,
"grad_norm": 1.364025354385376,
"learning_rate": 9.917525374361912e-05,
"loss": 0.1246,
"step": 210
},
{
"epoch": 2.2,
"grad_norm": 1.02783203125,
"learning_rate": 9.901899829374047e-05,
"loss": 0.1205,
"step": 220
},
{
"epoch": 2.3,
"grad_norm": 1.1652231216430664,
"learning_rate": 9.884934153917997e-05,
"loss": 0.1123,
"step": 230
},
{
"epoch": 2.4,
"grad_norm": 1.3376855850219727,
"learning_rate": 9.86663298624003e-05,
"loss": 0.1073,
"step": 240
},
{
"epoch": 2.5,
"grad_norm": 0.9518831968307495,
"learning_rate": 9.847001329696653e-05,
"loss": 0.0983,
"step": 250
},
{
"epoch": 2.6,
"grad_norm": 1.0183520317077637,
"learning_rate": 9.826044551386744e-05,
"loss": 0.0994,
"step": 260
},
{
"epoch": 2.7,
"grad_norm": 1.088984727859497,
"learning_rate": 9.803768380684242e-05,
"loss": 0.1057,
"step": 270
},
{
"epoch": 2.8,
"grad_norm": 0.7311228513717651,
"learning_rate": 9.780178907671789e-05,
"loss": 0.0979,
"step": 280
},
{
"epoch": 2.9,
"grad_norm": 0.4584773778915405,
"learning_rate": 9.755282581475769e-05,
"loss": 0.1015,
"step": 290
},
{
"epoch": 3.0,
"grad_norm": 1.5885437726974487,
"learning_rate": 9.729086208503174e-05,
"loss": 0.0798,
"step": 300
},
{
"epoch": 3.1,
"grad_norm": 0.7659353017807007,
"learning_rate": 9.701596950580806e-05,
"loss": 0.0978,
"step": 310
},
{
"epoch": 3.2,
"grad_norm": 0.472594290971756,
"learning_rate": 9.672822322997305e-05,
"loss": 0.0849,
"step": 320
},
{
"epoch": 3.3,
"grad_norm": 0.9068973064422607,
"learning_rate": 9.642770192448536e-05,
"loss": 0.0917,
"step": 330
},
{
"epoch": 3.4,
"grad_norm": 0.7755366563796997,
"learning_rate": 9.611448774886924e-05,
"loss": 0.0843,
"step": 340
},
{
"epoch": 3.5,
"grad_norm": 1.4321980476379395,
"learning_rate": 9.578866633275288e-05,
"loss": 0.0813,
"step": 350
},
{
"epoch": 3.6,
"grad_norm": 0.7324996590614319,
"learning_rate": 9.545032675245813e-05,
"loss": 0.0822,
"step": 360
},
{
"epoch": 3.7,
"grad_norm": 0.5374558568000793,
"learning_rate": 9.509956150664796e-05,
"loss": 0.0877,
"step": 370
},
{
"epoch": 3.8,
"grad_norm": 0.7615280151367188,
"learning_rate": 9.473646649103818e-05,
"loss": 0.0745,
"step": 380
},
{
"epoch": 3.9,
"grad_norm": 1.008134365081787,
"learning_rate": 9.43611409721806e-05,
"loss": 0.0739,
"step": 390
},
{
"epoch": 4.0,
"grad_norm": 1.376315951347351,
"learning_rate": 9.397368756032445e-05,
"loss": 0.0716,
"step": 400
},
{
"epoch": 4.1,
"grad_norm": 1.1624388694763184,
"learning_rate": 9.357421218136386e-05,
"loss": 0.0787,
"step": 410
},
{
"epoch": 4.2,
"grad_norm": 0.5839523077011108,
"learning_rate": 9.316282404787871e-05,
"loss": 0.0786,
"step": 420
},
{
"epoch": 4.3,
"grad_norm": 0.9589359760284424,
"learning_rate": 9.273963562927695e-05,
"loss": 0.0735,
"step": 430
},
{
"epoch": 4.4,
"grad_norm": 1.2537989616394043,
"learning_rate": 9.230476262104677e-05,
"loss": 0.0787,
"step": 440
},
{
"epoch": 4.5,
"grad_norm": 0.9431202411651611,
"learning_rate": 9.185832391312644e-05,
"loss": 0.0772,
"step": 450
},
{
"epoch": 4.6,
"grad_norm": 0.689441978931427,
"learning_rate": 9.140044155740101e-05,
"loss": 0.0775,
"step": 460
},
{
"epoch": 4.7,
"grad_norm": 0.6225442290306091,
"learning_rate": 9.093124073433463e-05,
"loss": 0.0838,
"step": 470
},
{
"epoch": 4.8,
"grad_norm": 0.8038715720176697,
"learning_rate": 9.045084971874738e-05,
"loss": 0.0708,
"step": 480
},
{
"epoch": 4.9,
"grad_norm": 1.0997047424316406,
"learning_rate": 8.995939984474624e-05,
"loss": 0.0736,
"step": 490
},
{
"epoch": 5.0,
"grad_norm": 1.7090861797332764,
"learning_rate": 8.945702546981969e-05,
"loss": 0.0664,
"step": 500
},
{
"epoch": 5.1,
"grad_norm": 0.6094951629638672,
"learning_rate": 8.894386393810563e-05,
"loss": 0.0718,
"step": 510
},
{
"epoch": 5.2,
"grad_norm": 0.577790379524231,
"learning_rate": 8.842005554284296e-05,
"loss": 0.0625,
"step": 520
},
{
"epoch": 5.3,
"grad_norm": 0.7675192952156067,
"learning_rate": 8.788574348801675e-05,
"loss": 0.0637,
"step": 530
},
{
"epoch": 5.4,
"grad_norm": 0.5569060444831848,
"learning_rate": 8.73410738492077e-05,
"loss": 0.0667,
"step": 540
},
{
"epoch": 5.5,
"grad_norm": 0.6511795520782471,
"learning_rate": 8.678619553365659e-05,
"loss": 0.07,
"step": 550
},
{
"epoch": 5.6,
"grad_norm": 0.7688615322113037,
"learning_rate": 8.622126023955446e-05,
"loss": 0.0579,
"step": 560
},
{
"epoch": 5.7,
"grad_norm": 0.668289840221405,
"learning_rate": 8.564642241456986e-05,
"loss": 0.0599,
"step": 570
},
{
"epoch": 5.8,
"grad_norm": 0.8834435939788818,
"learning_rate": 8.506183921362443e-05,
"loss": 0.0639,
"step": 580
},
{
"epoch": 5.9,
"grad_norm": 0.727353572845459,
"learning_rate": 8.44676704559283e-05,
"loss": 0.0675,
"step": 590
},
{
"epoch": 6.0,
"grad_norm": 2.630605936050415,
"learning_rate": 8.386407858128706e-05,
"loss": 0.0699,
"step": 600
},
{
"epoch": 6.1,
"grad_norm": 1.2145522832870483,
"learning_rate": 8.32512286056924e-05,
"loss": 0.0724,
"step": 610
},
{
"epoch": 6.2,
"grad_norm": 1.142543911933899,
"learning_rate": 8.262928807620843e-05,
"loss": 0.0727,
"step": 620
},
{
"epoch": 6.3,
"grad_norm": 0.6511779427528381,
"learning_rate": 8.199842702516583e-05,
"loss": 0.0611,
"step": 630
},
{
"epoch": 6.4,
"grad_norm": 0.5131811499595642,
"learning_rate": 8.135881792367686e-05,
"loss": 0.0591,
"step": 640
},
{
"epoch": 6.5,
"grad_norm": 0.8611190915107727,
"learning_rate": 8.07106356344834e-05,
"loss": 0.0644,
"step": 650
},
{
"epoch": 6.6,
"grad_norm": 0.7479195594787598,
"learning_rate": 8.005405736415126e-05,
"loss": 0.0638,
"step": 660
},
{
"epoch": 6.7,
"grad_norm": 1.030855655670166,
"learning_rate": 7.938926261462366e-05,
"loss": 0.0644,
"step": 670
},
{
"epoch": 6.8,
"grad_norm": 0.9318546056747437,
"learning_rate": 7.871643313414718e-05,
"loss": 0.0698,
"step": 680
},
{
"epoch": 6.9,
"grad_norm": 0.8005062937736511,
"learning_rate": 7.803575286758364e-05,
"loss": 0.064,
"step": 690
},
{
"epoch": 7.0,
"grad_norm": 1.5523598194122314,
"learning_rate": 7.734740790612136e-05,
"loss": 0.0581,
"step": 700
},
{
"epoch": 7.1,
"grad_norm": 0.7943648099899292,
"learning_rate": 7.66515864363997e-05,
"loss": 0.0699,
"step": 710
},
{
"epoch": 7.2,
"grad_norm": 0.775920569896698,
"learning_rate": 7.594847868906076e-05,
"loss": 0.0643,
"step": 720
},
{
"epoch": 7.3,
"grad_norm": 0.7172490954399109,
"learning_rate": 7.52382768867422e-05,
"loss": 0.055,
"step": 730
},
{
"epoch": 7.4,
"grad_norm": 0.7170266509056091,
"learning_rate": 7.452117519152542e-05,
"loss": 0.0561,
"step": 740
},
{
"epoch": 7.5,
"grad_norm": 0.5932777523994446,
"learning_rate": 7.379736965185368e-05,
"loss": 0.0541,
"step": 750
},
{
"epoch": 7.6,
"grad_norm": 0.5876290202140808,
"learning_rate": 7.30670581489344e-05,
"loss": 0.0568,
"step": 760
},
{
"epoch": 7.7,
"grad_norm": 0.7758692502975464,
"learning_rate": 7.233044034264034e-05,
"loss": 0.055,
"step": 770
},
{
"epoch": 7.8,
"grad_norm": 0.6511589288711548,
"learning_rate": 7.158771761692464e-05,
"loss": 0.0581,
"step": 780
},
{
"epoch": 7.9,
"grad_norm": 0.777108371257782,
"learning_rate": 7.083909302476453e-05,
"loss": 0.0491,
"step": 790
},
{
"epoch": 8.0,
"grad_norm": 3.6956264972686768,
"learning_rate": 7.008477123264848e-05,
"loss": 0.057,
"step": 800
},
{
"epoch": 8.1,
"grad_norm": 0.69562828540802,
"learning_rate": 6.932495846462261e-05,
"loss": 0.0572,
"step": 810
},
{
"epoch": 8.2,
"grad_norm": 0.7596244812011719,
"learning_rate": 6.855986244591104e-05,
"loss": 0.0598,
"step": 820
},
{
"epoch": 8.3,
"grad_norm": 0.8167744278907776,
"learning_rate": 6.778969234612584e-05,
"loss": 0.0523,
"step": 830
},
{
"epoch": 8.4,
"grad_norm": 0.7772257328033447,
"learning_rate": 6.701465872208216e-05,
"loss": 0.0563,
"step": 840
},
{
"epoch": 8.5,
"grad_norm": 0.6254410147666931,
"learning_rate": 6.623497346023418e-05,
"loss": 0.0527,
"step": 850
},
{
"epoch": 8.6,
"grad_norm": 0.4882695972919464,
"learning_rate": 6.545084971874738e-05,
"loss": 0.0543,
"step": 860
},
{
"epoch": 8.7,
"grad_norm": 0.8698508739471436,
"learning_rate": 6.466250186922325e-05,
"loss": 0.0495,
"step": 870
},
{
"epoch": 8.8,
"grad_norm": 0.6745582818984985,
"learning_rate": 6.387014543809223e-05,
"loss": 0.0505,
"step": 880
},
{
"epoch": 8.9,
"grad_norm": 1.0861890316009521,
"learning_rate": 6.307399704769099e-05,
"loss": 0.044,
"step": 890
},
{
"epoch": 9.0,
"grad_norm": 0.43595385551452637,
"learning_rate": 6.227427435703997e-05,
"loss": 0.0416,
"step": 900
},
{
"epoch": 9.1,
"grad_norm": 1.1349962949752808,
"learning_rate": 6.147119600233758e-05,
"loss": 0.0576,
"step": 910
},
{
"epoch": 9.2,
"grad_norm": 0.8424226641654968,
"learning_rate": 6.066498153718735e-05,
"loss": 0.0478,
"step": 920
},
{
"epoch": 9.3,
"grad_norm": 1.0484619140625,
"learning_rate": 5.985585137257401e-05,
"loss": 0.0549,
"step": 930
},
{
"epoch": 9.4,
"grad_norm": 0.6300575137138367,
"learning_rate": 5.90440267166055e-05,
"loss": 0.0507,
"step": 940
},
{
"epoch": 9.5,
"grad_norm": 0.6857895851135254,
"learning_rate": 5.8229729514036705e-05,
"loss": 0.0463,
"step": 950
},
{
"epoch": 9.6,
"grad_norm": 0.6959238648414612,
"learning_rate": 5.74131823855921e-05,
"loss": 0.0529,
"step": 960
},
{
"epoch": 9.7,
"grad_norm": 0.6516974568367004,
"learning_rate": 5.6594608567103456e-05,
"loss": 0.0461,
"step": 970
},
{
"epoch": 9.8,
"grad_norm": 1.1341180801391602,
"learning_rate": 5.577423184847932e-05,
"loss": 0.0458,
"step": 980
},
{
"epoch": 9.9,
"grad_norm": 0.6836873292922974,
"learning_rate": 5.495227651252315e-05,
"loss": 0.0498,
"step": 990
},
{
"epoch": 10.0,
"grad_norm": 1.8491439819335938,
"learning_rate": 5.4128967273616625e-05,
"loss": 0.0459,
"step": 1000
},
{
"epoch": 10.1,
"grad_norm": 0.7286385297775269,
"learning_rate": 5.330452921628497e-05,
"loss": 0.0495,
"step": 1010
},
{
"epoch": 10.2,
"grad_norm": 0.9026761651039124,
"learning_rate": 5.247918773366112e-05,
"loss": 0.0439,
"step": 1020
},
{
"epoch": 10.3,
"grad_norm": 0.5952501893043518,
"learning_rate": 5.165316846586541e-05,
"loss": 0.0438,
"step": 1030
},
{
"epoch": 10.4,
"grad_norm": 0.7074761986732483,
"learning_rate": 5.0826697238317935e-05,
"loss": 0.0444,
"step": 1040
},
{
"epoch": 10.5,
"grad_norm": 0.7149840593338013,
"learning_rate": 5e-05,
"loss": 0.0464,
"step": 1050
},
{
"epoch": 10.6,
"grad_norm": 0.9573127627372742,
"learning_rate": 4.917330276168208e-05,
"loss": 0.0434,
"step": 1060
},
{
"epoch": 10.7,
"grad_norm": 0.3760146498680115,
"learning_rate": 4.834683153413459e-05,
"loss": 0.0536,
"step": 1070
},
{
"epoch": 10.8,
"grad_norm": 0.7248881459236145,
"learning_rate": 4.7520812266338885e-05,
"loss": 0.052,
"step": 1080
},
{
"epoch": 10.9,
"grad_norm": 0.7167366147041321,
"learning_rate": 4.669547078371504e-05,
"loss": 0.0511,
"step": 1090
},
{
"epoch": 11.0,
"grad_norm": 0.554724931716919,
"learning_rate": 4.5871032726383386e-05,
"loss": 0.0361,
"step": 1100
},
{
"epoch": 11.1,
"grad_norm": 0.8976012468338013,
"learning_rate": 4.504772348747687e-05,
"loss": 0.0446,
"step": 1110
},
{
"epoch": 11.2,
"grad_norm": 1.2764275074005127,
"learning_rate": 4.4225768151520694e-05,
"loss": 0.0477,
"step": 1120
},
{
"epoch": 11.3,
"grad_norm": 0.9568806886672974,
"learning_rate": 4.3405391432896555e-05,
"loss": 0.0489,
"step": 1130
},
{
"epoch": 11.4,
"grad_norm": 0.5885640978813171,
"learning_rate": 4.2586817614407895e-05,
"loss": 0.0534,
"step": 1140
},
{
"epoch": 11.5,
"grad_norm": 0.48282381892204285,
"learning_rate": 4.17702704859633e-05,
"loss": 0.0489,
"step": 1150
},
{
"epoch": 11.6,
"grad_norm": 0.6082215905189514,
"learning_rate": 4.095597328339452e-05,
"loss": 0.0455,
"step": 1160
},
{
"epoch": 11.7,
"grad_norm": 0.8037371635437012,
"learning_rate": 4.0144148627425993e-05,
"loss": 0.0454,
"step": 1170
},
{
"epoch": 11.8,
"grad_norm": 0.6086834669113159,
"learning_rate": 3.933501846281267e-05,
"loss": 0.0427,
"step": 1180
},
{
"epoch": 11.9,
"grad_norm": 0.6202229857444763,
"learning_rate": 3.852880399766243e-05,
"loss": 0.0434,
"step": 1190
},
{
"epoch": 12.0,
"grad_norm": 1.6280860900878906,
"learning_rate": 3.772572564296005e-05,
"loss": 0.0405,
"step": 1200
},
{
"epoch": 12.1,
"grad_norm": 0.7294513583183289,
"learning_rate": 3.6926002952309016e-05,
"loss": 0.0474,
"step": 1210
},
{
"epoch": 12.2,
"grad_norm": 1.0804405212402344,
"learning_rate": 3.612985456190778e-05,
"loss": 0.0455,
"step": 1220
},
{
"epoch": 12.3,
"grad_norm": 0.8139125108718872,
"learning_rate": 3.533749813077677e-05,
"loss": 0.0396,
"step": 1230
},
{
"epoch": 12.4,
"grad_norm": 0.6082015037536621,
"learning_rate": 3.4549150281252636e-05,
"loss": 0.0455,
"step": 1240
},
{
"epoch": 12.5,
"grad_norm": 0.5969621539115906,
"learning_rate": 3.3765026539765834e-05,
"loss": 0.0393,
"step": 1250
},
{
"epoch": 12.6,
"grad_norm": 0.6664546728134155,
"learning_rate": 3.298534127791785e-05,
"loss": 0.0448,
"step": 1260
},
{
"epoch": 12.7,
"grad_norm": 0.5217233300209045,
"learning_rate": 3.221030765387417e-05,
"loss": 0.0446,
"step": 1270
},
{
"epoch": 12.8,
"grad_norm": 0.7529652118682861,
"learning_rate": 3.144013755408895e-05,
"loss": 0.0426,
"step": 1280
},
{
"epoch": 12.9,
"grad_norm": 0.41002699732780457,
"learning_rate": 3.0675041535377405e-05,
"loss": 0.0438,
"step": 1290
},
{
"epoch": 13.0,
"grad_norm": 3.2146363258361816,
"learning_rate": 2.991522876735154e-05,
"loss": 0.0435,
"step": 1300
},
{
"epoch": 13.1,
"grad_norm": 0.6749954223632812,
"learning_rate": 2.916090697523549e-05,
"loss": 0.0468,
"step": 1310
},
{
"epoch": 13.2,
"grad_norm": 0.7005327343940735,
"learning_rate": 2.8412282383075363e-05,
"loss": 0.0357,
"step": 1320
},
{
"epoch": 13.3,
"grad_norm": 0.6995335817337036,
"learning_rate": 2.766955965735968e-05,
"loss": 0.0426,
"step": 1330
},
{
"epoch": 13.4,
"grad_norm": 0.6105914115905762,
"learning_rate": 2.693294185106562e-05,
"loss": 0.0406,
"step": 1340
},
{
"epoch": 13.5,
"grad_norm": 0.6400395035743713,
"learning_rate": 2.6202630348146324e-05,
"loss": 0.043,
"step": 1350
},
{
"epoch": 13.6,
"grad_norm": 0.8495157957077026,
"learning_rate": 2.547882480847461e-05,
"loss": 0.0403,
"step": 1360
},
{
"epoch": 13.7,
"grad_norm": 0.6124421954154968,
"learning_rate": 2.476172311325783e-05,
"loss": 0.0395,
"step": 1370
},
{
"epoch": 13.8,
"grad_norm": 0.9979932904243469,
"learning_rate": 2.405152131093926e-05,
"loss": 0.0386,
"step": 1380
},
{
"epoch": 13.9,
"grad_norm": 0.9076727628707886,
"learning_rate": 2.3348413563600325e-05,
"loss": 0.0391,
"step": 1390
},
{
"epoch": 14.0,
"grad_norm": 0.8010100722312927,
"learning_rate": 2.2652592093878666e-05,
"loss": 0.0327,
"step": 1400
},
{
"epoch": 14.1,
"grad_norm": 0.7930116653442383,
"learning_rate": 2.196424713241637e-05,
"loss": 0.0416,
"step": 1410
},
{
"epoch": 14.2,
"grad_norm": 0.4719603359699249,
"learning_rate": 2.128356686585282e-05,
"loss": 0.0339,
"step": 1420
},
{
"epoch": 14.3,
"grad_norm": 0.5977267622947693,
"learning_rate": 2.061073738537635e-05,
"loss": 0.0433,
"step": 1430
},
{
"epoch": 14.4,
"grad_norm": 0.924698531627655,
"learning_rate": 1.9945942635848748e-05,
"loss": 0.0436,
"step": 1440
},
{
"epoch": 14.5,
"grad_norm": 0.3880954384803772,
"learning_rate": 1.928936436551661e-05,
"loss": 0.0403,
"step": 1450
},
{
"epoch": 14.6,
"grad_norm": 0.4617973566055298,
"learning_rate": 1.8641182076323148e-05,
"loss": 0.0373,
"step": 1460
},
{
"epoch": 14.7,
"grad_norm": 0.666461169719696,
"learning_rate": 1.800157297483417e-05,
"loss": 0.0418,
"step": 1470
},
{
"epoch": 14.8,
"grad_norm": 0.4455702304840088,
"learning_rate": 1.7370711923791567e-05,
"loss": 0.0352,
"step": 1480
},
{
"epoch": 14.9,
"grad_norm": 1.0331430435180664,
"learning_rate": 1.6748771394307585e-05,
"loss": 0.05,
"step": 1490
},
{
"epoch": 15.0,
"grad_norm": 1.1350957155227661,
"learning_rate": 1.6135921418712956e-05,
"loss": 0.0365,
"step": 1500
},
{
"epoch": 15.1,
"grad_norm": 0.6353791952133179,
"learning_rate": 1.553232954407171e-05,
"loss": 0.0399,
"step": 1510
},
{
"epoch": 15.2,
"grad_norm": 0.5416865944862366,
"learning_rate": 1.4938160786375572e-05,
"loss": 0.0451,
"step": 1520
},
{
"epoch": 15.3,
"grad_norm": 0.864016056060791,
"learning_rate": 1.435357758543015e-05,
"loss": 0.0379,
"step": 1530
},
{
"epoch": 15.4,
"grad_norm": 0.4366832971572876,
"learning_rate": 1.3778739760445552e-05,
"loss": 0.0405,
"step": 1540
},
{
"epoch": 15.5,
"grad_norm": 0.5444951057434082,
"learning_rate": 1.3213804466343421e-05,
"loss": 0.0368,
"step": 1550
},
{
"epoch": 15.6,
"grad_norm": 1.8182613849639893,
"learning_rate": 1.2658926150792322e-05,
"loss": 0.0497,
"step": 1560
},
{
"epoch": 15.7,
"grad_norm": 0.5947335362434387,
"learning_rate": 1.2114256511983274e-05,
"loss": 0.0391,
"step": 1570
},
{
"epoch": 15.8,
"grad_norm": 0.794297993183136,
"learning_rate": 1.157994445715706e-05,
"loss": 0.0372,
"step": 1580
},
{
"epoch": 15.9,
"grad_norm": 0.6530230641365051,
"learning_rate": 1.1056136061894384e-05,
"loss": 0.0385,
"step": 1590
},
{
"epoch": 16.0,
"grad_norm": 2.78792405128479,
"learning_rate": 1.0542974530180327e-05,
"loss": 0.0485,
"step": 1600
},
{
"epoch": 16.1,
"grad_norm": 0.9636825323104858,
"learning_rate": 1.0040600155253765e-05,
"loss": 0.0371,
"step": 1610
},
{
"epoch": 16.2,
"grad_norm": 0.4774838984012604,
"learning_rate": 9.549150281252633e-06,
"loss": 0.0379,
"step": 1620
},
{
"epoch": 16.3,
"grad_norm": 0.6014675498008728,
"learning_rate": 9.068759265665384e-06,
"loss": 0.0354,
"step": 1630
},
{
"epoch": 16.4,
"grad_norm": 0.7133916020393372,
"learning_rate": 8.599558442598998e-06,
"loss": 0.036,
"step": 1640
},
{
"epoch": 16.5,
"grad_norm": 0.765894889831543,
"learning_rate": 8.141676086873572e-06,
"loss": 0.0371,
"step": 1650
},
{
"epoch": 16.6,
"grad_norm": 0.49927645921707153,
"learning_rate": 7.695237378953223e-06,
"loss": 0.0398,
"step": 1660
},
{
"epoch": 16.7,
"grad_norm": 1.3441693782806396,
"learning_rate": 7.260364370723044e-06,
"loss": 0.0393,
"step": 1670
},
{
"epoch": 16.8,
"grad_norm": 0.6928268671035767,
"learning_rate": 6.837175952121306e-06,
"loss": 0.0333,
"step": 1680
},
{
"epoch": 16.9,
"grad_norm": 0.4891086220741272,
"learning_rate": 6.425787818636131e-06,
"loss": 0.0357,
"step": 1690
},
{
"epoch": 17.0,
"grad_norm": 1.0727057456970215,
"learning_rate": 6.026312439675552e-06,
"loss": 0.0403,
"step": 1700
},
{
"epoch": 17.1,
"grad_norm": 0.5739259123802185,
"learning_rate": 5.6388590278194096e-06,
"loss": 0.0366,
"step": 1710
},
{
"epoch": 17.2,
"grad_norm": 0.5994676351547241,
"learning_rate": 5.263533508961827e-06,
"loss": 0.0365,
"step": 1720
},
{
"epoch": 17.3,
"grad_norm": 0.7043574452400208,
"learning_rate": 4.900438493352055e-06,
"loss": 0.0404,
"step": 1730
},
{
"epoch": 17.4,
"grad_norm": 0.4768543839454651,
"learning_rate": 4.549673247541875e-06,
"loss": 0.0419,
"step": 1740
},
{
"epoch": 17.5,
"grad_norm": 0.800778329372406,
"learning_rate": 4.2113336672471245e-06,
"loss": 0.0379,
"step": 1750
},
{
"epoch": 17.6,
"grad_norm": 0.9242330193519592,
"learning_rate": 3.885512251130763e-06,
"loss": 0.0337,
"step": 1760
},
{
"epoch": 17.7,
"grad_norm": 0.37277647852897644,
"learning_rate": 3.5722980755146517e-06,
"loss": 0.0381,
"step": 1770
},
{
"epoch": 17.8,
"grad_norm": 0.8925457000732422,
"learning_rate": 3.271776770026963e-06,
"loss": 0.0383,
"step": 1780
},
{
"epoch": 17.9,
"grad_norm": 0.8878549337387085,
"learning_rate": 2.9840304941919415e-06,
"loss": 0.0405,
"step": 1790
},
{
"epoch": 18.0,
"grad_norm": 0.779109537601471,
"learning_rate": 2.7091379149682685e-06,
"loss": 0.0341,
"step": 1800
},
{
"epoch": 18.1,
"grad_norm": 0.4987022280693054,
"learning_rate": 2.4471741852423237e-06,
"loss": 0.0435,
"step": 1810
},
{
"epoch": 18.2,
"grad_norm": 0.8446706533432007,
"learning_rate": 2.1982109232821178e-06,
"loss": 0.0401,
"step": 1820
},
{
"epoch": 18.3,
"grad_norm": 0.6136398911476135,
"learning_rate": 1.962316193157593e-06,
"loss": 0.0405,
"step": 1830
},
{
"epoch": 18.4,
"grad_norm": 0.5189019441604614,
"learning_rate": 1.7395544861325718e-06,
"loss": 0.035,
"step": 1840
},
{
"epoch": 18.5,
"grad_norm": 1.2081691026687622,
"learning_rate": 1.5299867030334814e-06,
"loss": 0.0301,
"step": 1850
},
{
"epoch": 18.6,
"grad_norm": 0.5128328204154968,
"learning_rate": 1.333670137599713e-06,
"loss": 0.0342,
"step": 1860
},
{
"epoch": 18.7,
"grad_norm": 0.6271176338195801,
"learning_rate": 1.1506584608200367e-06,
"loss": 0.0374,
"step": 1870
},
{
"epoch": 18.8,
"grad_norm": 0.7536263465881348,
"learning_rate": 9.810017062595322e-07,
"loss": 0.0432,
"step": 1880
},
{
"epoch": 18.9,
"grad_norm": 0.70732581615448,
"learning_rate": 8.247462563808817e-07,
"loss": 0.0389,
"step": 1890
},
{
"epoch": 19.0,
"grad_norm": 3.046426296234131,
"learning_rate": 6.819348298638839e-07,
"loss": 0.0339,
"step": 1900
},
{
"epoch": 19.1,
"grad_norm": 0.5528438091278076,
"learning_rate": 5.526064699265753e-07,
"loss": 0.0356,
"step": 1910
},
{
"epoch": 19.2,
"grad_norm": 0.4317626357078552,
"learning_rate": 4.367965336512403e-07,
"loss": 0.0415,
"step": 1920
},
{
"epoch": 19.3,
"grad_norm": 0.7551414966583252,
"learning_rate": 3.3453668231809286e-07,
"loss": 0.0375,
"step": 1930
},
{
"epoch": 19.4,
"grad_norm": 0.737501859664917,
"learning_rate": 2.458548727494292e-07,
"loss": 0.0391,
"step": 1940
},
{
"epoch": 19.5,
"grad_norm": 0.3441050052642822,
"learning_rate": 1.7077534966650766e-07,
"loss": 0.0356,
"step": 1950
},
{
"epoch": 19.6,
"grad_norm": 0.7213258147239685,
"learning_rate": 1.0931863906127327e-07,
"loss": 0.0377,
"step": 1960
},
{
"epoch": 19.7,
"grad_norm": 0.6739200353622437,
"learning_rate": 6.150154258476315e-08,
"loss": 0.0346,
"step": 1970
},
{
"epoch": 19.8,
"grad_norm": 0.4855494499206543,
"learning_rate": 2.7337132953697554e-08,
"loss": 0.0315,
"step": 1980
},
{
"epoch": 19.9,
"grad_norm": 0.8237863779067993,
"learning_rate": 6.834750376549792e-09,
"loss": 0.0403,
"step": 1990
},
{
"epoch": 20.0,
"grad_norm": 0.4662252366542816,
"learning_rate": 0.0,
"loss": 0.0408,
"step": 2000
},
{
"epoch": 20.0,
"step": 2000,
"total_flos": 1.357860048774336e+17,
"train_loss": 0.09518647351861,
"train_runtime": 632.3751,
"train_samples_per_second": 100.257,
"train_steps_per_second": 3.163
}
],
"logging_steps": 10,
"max_steps": 2000,
"num_input_tokens_seen": 0,
"num_train_epochs": 20,
"save_steps": 20000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.357860048774336e+17,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}