| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.885813148788928, | |
| "eval_steps": 500, | |
| "global_step": 180, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02768166089965398, | |
| "grad_norm": 3.3037207391517223, | |
| "learning_rate": 9.999238475781957e-05, | |
| "loss": 1.7593, | |
| "num_input_tokens_seen": 262144, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.05536332179930796, | |
| "grad_norm": 1.9466815384743432, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 1.7063, | |
| "num_input_tokens_seen": 524288, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.08304498269896193, | |
| "grad_norm": 1.2915868553800258, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 1.5574, | |
| "num_input_tokens_seen": 786432, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.11072664359861592, | |
| "grad_norm": 0.6946491855557626, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 1.4176, | |
| "num_input_tokens_seen": 1048576, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.1384083044982699, | |
| "grad_norm": 0.45365345106107186, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 1.3465, | |
| "num_input_tokens_seen": 1310720, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.16608996539792387, | |
| "grad_norm": 0.4850172145419489, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 1.2969, | |
| "num_input_tokens_seen": 1572864, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.19377162629757785, | |
| "grad_norm": 0.3136871486116671, | |
| "learning_rate": 9.962730758206611e-05, | |
| "loss": 1.2299, | |
| "num_input_tokens_seen": 1835008, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.22145328719723184, | |
| "grad_norm": 0.3751301661684622, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 1.1664, | |
| "num_input_tokens_seen": 2097152, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.2491349480968858, | |
| "grad_norm": 0.3023645231047452, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 1.1343, | |
| "num_input_tokens_seen": 2359296, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.2768166089965398, | |
| "grad_norm": 0.28189750356788335, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 1.1138, | |
| "num_input_tokens_seen": 2621440, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3044982698961938, | |
| "grad_norm": 0.3415169757562131, | |
| "learning_rate": 9.908135917238321e-05, | |
| "loss": 1.0732, | |
| "num_input_tokens_seen": 2883584, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.33217993079584773, | |
| "grad_norm": 0.3054483418783897, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 1.0499, | |
| "num_input_tokens_seen": 3145728, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.35986159169550175, | |
| "grad_norm": 0.3164128467566228, | |
| "learning_rate": 9.871850323926177e-05, | |
| "loss": 1.0125, | |
| "num_input_tokens_seen": 3407872, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.3875432525951557, | |
| "grad_norm": 0.26489187095994526, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 0.9943, | |
| "num_input_tokens_seen": 3670016, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.41522491349480967, | |
| "grad_norm": 0.28461733265235484, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 0.9761, | |
| "num_input_tokens_seen": 3932160, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.4429065743944637, | |
| "grad_norm": 0.2855246448932269, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 0.9658, | |
| "num_input_tokens_seen": 4194304, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 0.2516090681914765, | |
| "learning_rate": 9.781523779815179e-05, | |
| "loss": 0.9443, | |
| "num_input_tokens_seen": 4456448, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.4982698961937716, | |
| "grad_norm": 0.2608567765975221, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.9279, | |
| "num_input_tokens_seen": 4718592, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.5259515570934256, | |
| "grad_norm": 0.23894892950600116, | |
| "learning_rate": 9.727592877996585e-05, | |
| "loss": 0.8942, | |
| "num_input_tokens_seen": 4980736, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.5536332179930796, | |
| "grad_norm": 0.24631380564109917, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.888, | |
| "num_input_tokens_seen": 5242880, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.5813148788927336, | |
| "grad_norm": 0.24993538995507514, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 0.8784, | |
| "num_input_tokens_seen": 5505024, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.6089965397923875, | |
| "grad_norm": 0.24239907887203402, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 0.8832, | |
| "num_input_tokens_seen": 5767168, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.6366782006920415, | |
| "grad_norm": 0.27319705128967686, | |
| "learning_rate": 9.602524267262203e-05, | |
| "loss": 0.8673, | |
| "num_input_tokens_seen": 6029312, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.6643598615916955, | |
| "grad_norm": 0.24710997608486757, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.857, | |
| "num_input_tokens_seen": 6291456, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.6920415224913494, | |
| "grad_norm": 0.23862113948586505, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 0.8282, | |
| "num_input_tokens_seen": 6553600, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.7197231833910035, | |
| "grad_norm": 0.22499698855453915, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 0.7924, | |
| "num_input_tokens_seen": 6815744, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.7474048442906575, | |
| "grad_norm": 0.24543105004281635, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 0.8304, | |
| "num_input_tokens_seen": 7077888, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.7750865051903114, | |
| "grad_norm": 0.2373126575604043, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.8041, | |
| "num_input_tokens_seen": 7340032, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.8027681660899654, | |
| "grad_norm": 0.22893825349312907, | |
| "learning_rate": 9.373098535696979e-05, | |
| "loss": 0.8165, | |
| "num_input_tokens_seen": 7602176, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.8304498269896193, | |
| "grad_norm": 0.21905173932955566, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 0.7923, | |
| "num_input_tokens_seen": 7864320, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.8581314878892734, | |
| "grad_norm": 0.2292130616351926, | |
| "learning_rate": 9.285836503510562e-05, | |
| "loss": 0.8111, | |
| "num_input_tokens_seen": 8126464, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.8858131487889274, | |
| "grad_norm": 0.2214107878966629, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 0.7977, | |
| "num_input_tokens_seen": 8388608, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.9134948096885813, | |
| "grad_norm": 0.2198438897545508, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 0.8013, | |
| "num_input_tokens_seen": 8650752, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 0.22444446062256257, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 0.7667, | |
| "num_input_tokens_seen": 8912896, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.9688581314878892, | |
| "grad_norm": 0.23388161799831605, | |
| "learning_rate": 9.09576022144496e-05, | |
| "loss": 0.7651, | |
| "num_input_tokens_seen": 9175040, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.9965397923875432, | |
| "grad_norm": 0.218209848371025, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.7741, | |
| "num_input_tokens_seen": 9437184, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.6085418369298928, | |
| "learning_rate": 8.993177550236464e-05, | |
| "loss": 0.7389, | |
| "num_input_tokens_seen": 9469952, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.027681660899654, | |
| "grad_norm": 0.260963270509872, | |
| "learning_rate": 8.940053768033609e-05, | |
| "loss": 0.6296, | |
| "num_input_tokens_seen": 9732096, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.055363321799308, | |
| "grad_norm": 0.25846705719490826, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 0.6208, | |
| "num_input_tokens_seen": 9994240, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.0830449826989619, | |
| "grad_norm": 0.27925385338634273, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.6086, | |
| "num_input_tokens_seen": 10256384, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.1107266435986158, | |
| "grad_norm": 0.2565275069957051, | |
| "learning_rate": 8.773547901113862e-05, | |
| "loss": 0.5841, | |
| "num_input_tokens_seen": 10518528, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.1384083044982698, | |
| "grad_norm": 0.25188647848207807, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 0.6112, | |
| "num_input_tokens_seen": 10780672, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.1660899653979238, | |
| "grad_norm": 0.22581137918838712, | |
| "learning_rate": 8.656768508095853e-05, | |
| "loss": 0.5926, | |
| "num_input_tokens_seen": 11042816, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.193771626297578, | |
| "grad_norm": 0.26278248584610486, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.5858, | |
| "num_input_tokens_seen": 11304960, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.221453287197232, | |
| "grad_norm": 0.2357437535857933, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.5958, | |
| "num_input_tokens_seen": 11567104, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.2491349480968859, | |
| "grad_norm": 0.23313547992312228, | |
| "learning_rate": 8.473291852294987e-05, | |
| "loss": 0.5908, | |
| "num_input_tokens_seen": 11829248, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.2768166089965398, | |
| "grad_norm": 0.23869112354639369, | |
| "learning_rate": 8.409991800312493e-05, | |
| "loss": 0.5844, | |
| "num_input_tokens_seen": 12091392, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.3044982698961938, | |
| "grad_norm": 0.2393988026622817, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.586, | |
| "num_input_tokens_seen": 12353536, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.3321799307958477, | |
| "grad_norm": 0.23192089469806965, | |
| "learning_rate": 8.280295144952536e-05, | |
| "loss": 0.5676, | |
| "num_input_tokens_seen": 12615680, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.3598615916955017, | |
| "grad_norm": 0.24624021606130117, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 0.5668, | |
| "num_input_tokens_seen": 12877824, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.3875432525951557, | |
| "grad_norm": 0.2360682949671219, | |
| "learning_rate": 8.146601955249188e-05, | |
| "loss": 0.565, | |
| "num_input_tokens_seen": 13139968, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.4152249134948096, | |
| "grad_norm": 0.23708325807568364, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.5746, | |
| "num_input_tokens_seen": 13402112, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.4429065743944638, | |
| "grad_norm": 0.2253072048492113, | |
| "learning_rate": 8.009075115760243e-05, | |
| "loss": 0.5723, | |
| "num_input_tokens_seen": 13664256, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.4705882352941178, | |
| "grad_norm": 0.2294155894631647, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 0.5646, | |
| "num_input_tokens_seen": 13926400, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.4982698961937717, | |
| "grad_norm": 0.23437110809294137, | |
| "learning_rate": 7.86788218175523e-05, | |
| "loss": 0.5577, | |
| "num_input_tokens_seen": 14188544, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.5259515570934257, | |
| "grad_norm": 0.2351650353703212, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 0.5685, | |
| "num_input_tokens_seen": 14450688, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.5536332179930796, | |
| "grad_norm": 0.2339546069470911, | |
| "learning_rate": 7.723195175075136e-05, | |
| "loss": 0.5603, | |
| "num_input_tokens_seen": 14712832, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.5813148788927336, | |
| "grad_norm": 0.22732686546761338, | |
| "learning_rate": 7.649596321166024e-05, | |
| "loss": 0.532, | |
| "num_input_tokens_seen": 14974976, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.6089965397923875, | |
| "grad_norm": 0.2234772514246162, | |
| "learning_rate": 7.575190374550272e-05, | |
| "loss": 0.5428, | |
| "num_input_tokens_seen": 15237120, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.6366782006920415, | |
| "grad_norm": 0.22752739648073927, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.5542, | |
| "num_input_tokens_seen": 15499264, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.6643598615916955, | |
| "grad_norm": 0.23581308611034632, | |
| "learning_rate": 7.424048101231686e-05, | |
| "loss": 0.5356, | |
| "num_input_tokens_seen": 15761408, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.6920415224913494, | |
| "grad_norm": 0.23497586610456628, | |
| "learning_rate": 7.347357813929454e-05, | |
| "loss": 0.5602, | |
| "num_input_tokens_seen": 16023552, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.7197231833910034, | |
| "grad_norm": 0.22461223215897164, | |
| "learning_rate": 7.269952498697734e-05, | |
| "loss": 0.5402, | |
| "num_input_tokens_seen": 16285696, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.7474048442906573, | |
| "grad_norm": 0.22743168302580066, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 0.5454, | |
| "num_input_tokens_seen": 16547840, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.7750865051903113, | |
| "grad_norm": 0.22933972591087348, | |
| "learning_rate": 7.113091308703498e-05, | |
| "loss": 0.5492, | |
| "num_input_tokens_seen": 16809984, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.8027681660899653, | |
| "grad_norm": 0.2311975181754483, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 0.5609, | |
| "num_input_tokens_seen": 17072128, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.8304498269896192, | |
| "grad_norm": 0.22768998479641345, | |
| "learning_rate": 6.953655642446368e-05, | |
| "loss": 0.5447, | |
| "num_input_tokens_seen": 17334272, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.8581314878892734, | |
| "grad_norm": 0.22513110514606996, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 0.5465, | |
| "num_input_tokens_seen": 17596416, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.8858131487889274, | |
| "grad_norm": 0.22925906212313765, | |
| "learning_rate": 6.7918397477265e-05, | |
| "loss": 0.5395, | |
| "num_input_tokens_seen": 17858560, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.9134948096885813, | |
| "grad_norm": 0.24171793988030507, | |
| "learning_rate": 6.710100716628344e-05, | |
| "loss": 0.5508, | |
| "num_input_tokens_seen": 18120704, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.9411764705882353, | |
| "grad_norm": 0.22205245882198657, | |
| "learning_rate": 6.627840772285784e-05, | |
| "loss": 0.5394, | |
| "num_input_tokens_seen": 18382848, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.9688581314878892, | |
| "grad_norm": 0.23312257533351413, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.5201, | |
| "num_input_tokens_seen": 18644992, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.9965397923875432, | |
| "grad_norm": 0.2283413128494737, | |
| "learning_rate": 6.461858523613684e-05, | |
| "loss": 0.5266, | |
| "num_input_tokens_seen": 18907136, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.615769018644783, | |
| "learning_rate": 6.378186779084995e-05, | |
| "loss": 0.4585, | |
| "num_input_tokens_seen": 18939904, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.027681660899654, | |
| "grad_norm": 0.3393218615477358, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 0.3593, | |
| "num_input_tokens_seen": 19202048, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.055363321799308, | |
| "grad_norm": 0.37547631344239085, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 0.3462, | |
| "num_input_tokens_seen": 19464192, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.083044982698962, | |
| "grad_norm": 0.44751782846446986, | |
| "learning_rate": 6.124755271719325e-05, | |
| "loss": 0.3574, | |
| "num_input_tokens_seen": 19726336, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.110726643598616, | |
| "grad_norm": 0.29143900433267883, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 0.3385, | |
| "num_input_tokens_seen": 19988480, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.13840830449827, | |
| "grad_norm": 0.2779214345738146, | |
| "learning_rate": 5.9540449768827246e-05, | |
| "loss": 0.3238, | |
| "num_input_tokens_seen": 20250624, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.1660899653979238, | |
| "grad_norm": 0.2672130552638047, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 0.3402, | |
| "num_input_tokens_seen": 20512768, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.1937716262975777, | |
| "grad_norm": 0.28595957538502975, | |
| "learning_rate": 5.782172325201155e-05, | |
| "loss": 0.3172, | |
| "num_input_tokens_seen": 20774912, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.2214532871972317, | |
| "grad_norm": 0.3072159523724194, | |
| "learning_rate": 5.695865504800327e-05, | |
| "loss": 0.3305, | |
| "num_input_tokens_seen": 21037056, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.2491349480968856, | |
| "grad_norm": 0.2978478923918209, | |
| "learning_rate": 5.6093467170257374e-05, | |
| "loss": 0.3271, | |
| "num_input_tokens_seen": 21299200, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.2768166089965396, | |
| "grad_norm": 0.3051897546416077, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 0.3434, | |
| "num_input_tokens_seen": 21561344, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.304498269896194, | |
| "grad_norm": 0.2787207011014885, | |
| "learning_rate": 5.435778713738292e-05, | |
| "loss": 0.3132, | |
| "num_input_tokens_seen": 21823488, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.3321799307958475, | |
| "grad_norm": 0.26770104916796217, | |
| "learning_rate": 5.348782368720626e-05, | |
| "loss": 0.3132, | |
| "num_input_tokens_seen": 22085632, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.359861591695502, | |
| "grad_norm": 0.27765670859127994, | |
| "learning_rate": 5.26167978121472e-05, | |
| "loss": 0.3166, | |
| "num_input_tokens_seen": 22347776, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.387543252595156, | |
| "grad_norm": 0.2771681582041151, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 0.3226, | |
| "num_input_tokens_seen": 22609920, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.41522491349481, | |
| "grad_norm": 0.27567715251532016, | |
| "learning_rate": 5.0872620321864185e-05, | |
| "loss": 0.3263, | |
| "num_input_tokens_seen": 22872064, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 2.442906574394464, | |
| "grad_norm": 0.270788972141914, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3151, | |
| "num_input_tokens_seen": 23134208, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.4705882352941178, | |
| "grad_norm": 0.2777173897871249, | |
| "learning_rate": 4.912737967813583e-05, | |
| "loss": 0.3354, | |
| "num_input_tokens_seen": 23396352, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.4982698961937717, | |
| "grad_norm": 0.2740028365423069, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 0.312, | |
| "num_input_tokens_seen": 23658496, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 2.5259515570934257, | |
| "grad_norm": 0.2906101508912602, | |
| "learning_rate": 4.738320218785281e-05, | |
| "loss": 0.3192, | |
| "num_input_tokens_seen": 23920640, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.5536332179930796, | |
| "grad_norm": 0.2657438173512922, | |
| "learning_rate": 4.6512176312793736e-05, | |
| "loss": 0.3006, | |
| "num_input_tokens_seen": 24182784, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.5813148788927336, | |
| "grad_norm": 0.27380335660711086, | |
| "learning_rate": 4.564221286261709e-05, | |
| "loss": 0.328, | |
| "num_input_tokens_seen": 24444928, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.6089965397923875, | |
| "grad_norm": 0.2653472809223449, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 0.3131, | |
| "num_input_tokens_seen": 24707072, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.6366782006920415, | |
| "grad_norm": 0.2763551950260926, | |
| "learning_rate": 4.390653282974264e-05, | |
| "loss": 0.3307, | |
| "num_input_tokens_seen": 24969216, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.6643598615916955, | |
| "grad_norm": 0.2684442029630596, | |
| "learning_rate": 4.3041344951996746e-05, | |
| "loss": 0.3079, | |
| "num_input_tokens_seen": 25231360, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 2.6920415224913494, | |
| "grad_norm": 0.2788436256991694, | |
| "learning_rate": 4.2178276747988446e-05, | |
| "loss": 0.3177, | |
| "num_input_tokens_seen": 25493504, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 2.7197231833910034, | |
| "grad_norm": 0.29513825800876553, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 0.3255, | |
| "num_input_tokens_seen": 25755648, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.7474048442906573, | |
| "grad_norm": 0.2735536152349423, | |
| "learning_rate": 4.045955023117276e-05, | |
| "loss": 0.3054, | |
| "num_input_tokens_seen": 26017792, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 2.7750865051903113, | |
| "grad_norm": 0.2621662310034819, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 0.3146, | |
| "num_input_tokens_seen": 26279936, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 2.8027681660899653, | |
| "grad_norm": 0.2681277716926742, | |
| "learning_rate": 3.875244728280676e-05, | |
| "loss": 0.3081, | |
| "num_input_tokens_seen": 26542080, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 2.830449826989619, | |
| "grad_norm": 0.2657552072157264, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 0.3308, | |
| "num_input_tokens_seen": 26804224, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 2.858131487889273, | |
| "grad_norm": 0.2508780313662121, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 0.297, | |
| "num_input_tokens_seen": 27066368, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.8858131487889276, | |
| "grad_norm": 0.2735729639908825, | |
| "learning_rate": 3.6218132209150045e-05, | |
| "loss": 0.3246, | |
| "num_input_tokens_seen": 27328512, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.913494809688581, | |
| "grad_norm": 0.25719044704692595, | |
| "learning_rate": 3.5381414763863166e-05, | |
| "loss": 0.3026, | |
| "num_input_tokens_seen": 27590656, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "grad_norm": 0.27065986320784646, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.3162, | |
| "num_input_tokens_seen": 27852800, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.968858131487889, | |
| "grad_norm": 0.2778157950887113, | |
| "learning_rate": 3.372159227714218e-05, | |
| "loss": 0.3219, | |
| "num_input_tokens_seen": 28114944, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.9965397923875434, | |
| "grad_norm": 0.2837474501991812, | |
| "learning_rate": 3.289899283371657e-05, | |
| "loss": 0.3192, | |
| "num_input_tokens_seen": 28377088, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.6760488934419766, | |
| "learning_rate": 3.2081602522734986e-05, | |
| "loss": 0.2184, | |
| "num_input_tokens_seen": 28409856, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 3.027681660899654, | |
| "grad_norm": 0.2937426397528417, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 0.174, | |
| "num_input_tokens_seen": 28672000, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 3.055363321799308, | |
| "grad_norm": 0.27074426450641464, | |
| "learning_rate": 3.046344357553632e-05, | |
| "loss": 0.161, | |
| "num_input_tokens_seen": 28934144, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 3.083044982698962, | |
| "grad_norm": 0.2944447037701877, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 0.1518, | |
| "num_input_tokens_seen": 29196288, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 3.110726643598616, | |
| "grad_norm": 0.3910554135151892, | |
| "learning_rate": 2.886908691296504e-05, | |
| "loss": 0.1487, | |
| "num_input_tokens_seen": 29458432, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 3.13840830449827, | |
| "grad_norm": 0.3929096409331864, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 0.1516, | |
| "num_input_tokens_seen": 29720576, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 3.1660899653979238, | |
| "grad_norm": 0.3467034550293285, | |
| "learning_rate": 2.7300475013022663e-05, | |
| "loss": 0.141, | |
| "num_input_tokens_seen": 29982720, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 3.1937716262975777, | |
| "grad_norm": 0.32457527007440695, | |
| "learning_rate": 2.6526421860705473e-05, | |
| "loss": 0.1516, | |
| "num_input_tokens_seen": 30244864, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 3.2214532871972317, | |
| "grad_norm": 0.2776671470971107, | |
| "learning_rate": 2.575951898768315e-05, | |
| "loss": 0.15, | |
| "num_input_tokens_seen": 30507008, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 3.2491349480968856, | |
| "grad_norm": 0.26780992442827467, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 0.1482, | |
| "num_input_tokens_seen": 30769152, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.2768166089965396, | |
| "grad_norm": 0.2788844877876364, | |
| "learning_rate": 2.4248096254497288e-05, | |
| "loss": 0.134, | |
| "num_input_tokens_seen": 31031296, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 3.304498269896194, | |
| "grad_norm": 0.29153191468092, | |
| "learning_rate": 2.350403678833976e-05, | |
| "loss": 0.1544, | |
| "num_input_tokens_seen": 31293440, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 3.3321799307958475, | |
| "grad_norm": 0.2804101178824654, | |
| "learning_rate": 2.2768048249248648e-05, | |
| "loss": 0.1341, | |
| "num_input_tokens_seen": 31555584, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 3.359861591695502, | |
| "grad_norm": 0.27608972584537617, | |
| "learning_rate": 2.2040354826462668e-05, | |
| "loss": 0.1386, | |
| "num_input_tokens_seen": 31817728, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 3.387543252595156, | |
| "grad_norm": 0.29062128593141306, | |
| "learning_rate": 2.132117818244771e-05, | |
| "loss": 0.1444, | |
| "num_input_tokens_seen": 32079872, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 3.41522491349481, | |
| "grad_norm": 0.2913229956296559, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 0.1369, | |
| "num_input_tokens_seen": 32342016, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 3.442906574394464, | |
| "grad_norm": 0.2930687487758551, | |
| "learning_rate": 1.9909248842397584e-05, | |
| "loss": 0.1459, | |
| "num_input_tokens_seen": 32604160, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 3.4705882352941178, | |
| "grad_norm": 0.29968241333359413, | |
| "learning_rate": 1.9216926233717085e-05, | |
| "loss": 0.1446, | |
| "num_input_tokens_seen": 32866304, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 3.4982698961937717, | |
| "grad_norm": 0.2939207075270168, | |
| "learning_rate": 1.8533980447508137e-05, | |
| "loss": 0.1461, | |
| "num_input_tokens_seen": 33128448, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 3.5259515570934257, | |
| "grad_norm": 0.28480194753552124, | |
| "learning_rate": 1.7860619515673033e-05, | |
| "loss": 0.1451, | |
| "num_input_tokens_seen": 33390592, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 3.5536332179930796, | |
| "grad_norm": 0.2721877075729539, | |
| "learning_rate": 1.7197048550474643e-05, | |
| "loss": 0.138, | |
| "num_input_tokens_seen": 33652736, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 3.5813148788927336, | |
| "grad_norm": 0.2935009290783486, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 0.1375, | |
| "num_input_tokens_seen": 33914880, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 3.6089965397923875, | |
| "grad_norm": 0.28323510348833286, | |
| "learning_rate": 1.5900081996875083e-05, | |
| "loss": 0.1286, | |
| "num_input_tokens_seen": 34177024, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 3.6366782006920415, | |
| "grad_norm": 0.26924320565669435, | |
| "learning_rate": 1.526708147705013e-05, | |
| "loss": 0.1357, | |
| "num_input_tokens_seen": 34439168, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 3.6643598615916955, | |
| "grad_norm": 0.26984947695360095, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 0.1295, | |
| "num_input_tokens_seen": 34701312, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 3.6920415224913494, | |
| "grad_norm": 0.27843611711833605, | |
| "learning_rate": 1.4033009983067452e-05, | |
| "loss": 0.1298, | |
| "num_input_tokens_seen": 34963456, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 3.7197231833910034, | |
| "grad_norm": 0.2919231187162401, | |
| "learning_rate": 1.3432314919041478e-05, | |
| "loss": 0.144, | |
| "num_input_tokens_seen": 35225600, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 3.7474048442906573, | |
| "grad_norm": 0.28950587722594623, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 0.1388, | |
| "num_input_tokens_seen": 35487744, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 3.7750865051903113, | |
| "grad_norm": 0.28052387866626566, | |
| "learning_rate": 1.22645209888614e-05, | |
| "loss": 0.139, | |
| "num_input_tokens_seen": 35749888, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 3.8027681660899653, | |
| "grad_norm": 0.2676754714326058, | |
| "learning_rate": 1.1697777844051105e-05, | |
| "loss": 0.1317, | |
| "num_input_tokens_seen": 36012032, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.830449826989619, | |
| "grad_norm": 0.26822684247808565, | |
| "learning_rate": 1.1142701927151456e-05, | |
| "loss": 0.1265, | |
| "num_input_tokens_seen": 36274176, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 3.858131487889273, | |
| "grad_norm": 0.2770704242753118, | |
| "learning_rate": 1.0599462319663905e-05, | |
| "loss": 0.138, | |
| "num_input_tokens_seen": 36536320, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 3.8858131487889276, | |
| "grad_norm": 0.2708141817663477, | |
| "learning_rate": 1.006822449763537e-05, | |
| "loss": 0.1348, | |
| "num_input_tokens_seen": 36798464, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 3.913494809688581, | |
| "grad_norm": 0.2863394961107606, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 0.135, | |
| "num_input_tokens_seen": 37060608, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 3.9411764705882355, | |
| "grad_norm": 0.2780251237592949, | |
| "learning_rate": 9.042397785550405e-06, | |
| "loss": 0.1362, | |
| "num_input_tokens_seen": 37322752, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 3.968858131487889, | |
| "grad_norm": 0.28286318190957144, | |
| "learning_rate": 8.548121372247918e-06, | |
| "loss": 0.1424, | |
| "num_input_tokens_seen": 37584896, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 3.9965397923875434, | |
| "grad_norm": 0.2824153687738419, | |
| "learning_rate": 8.066471602728803e-06, | |
| "loss": 0.1374, | |
| "num_input_tokens_seen": 37847040, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.6491532630662947, | |
| "learning_rate": 7.597595192178702e-06, | |
| "loss": 0.1011, | |
| "num_input_tokens_seen": 37879808, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 4.027681660899654, | |
| "grad_norm": 0.20270128714943275, | |
| "learning_rate": 7.1416349648943894e-06, | |
| "loss": 0.0651, | |
| "num_input_tokens_seen": 38141952, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 4.055363321799308, | |
| "grad_norm": 0.20307977493176613, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 0.0748, | |
| "num_input_tokens_seen": 38404096, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 4.083044982698962, | |
| "grad_norm": 0.194487047228382, | |
| "learning_rate": 6.269014643030213e-06, | |
| "loss": 0.066, | |
| "num_input_tokens_seen": 38666240, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 4.110726643598616, | |
| "grad_norm": 0.20122163698091994, | |
| "learning_rate": 5.852620357053651e-06, | |
| "loss": 0.069, | |
| "num_input_tokens_seen": 38928384, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 4.13840830449827, | |
| "grad_norm": 0.22899568675507392, | |
| "learning_rate": 5.449673790581611e-06, | |
| "loss": 0.0674, | |
| "num_input_tokens_seen": 39190528, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 4.166089965397924, | |
| "grad_norm": 0.22855432728321723, | |
| "learning_rate": 5.060297685041659e-06, | |
| "loss": 0.0696, | |
| "num_input_tokens_seen": 39452672, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 4.193771626297578, | |
| "grad_norm": 0.2110278342096293, | |
| "learning_rate": 4.684610648167503e-06, | |
| "loss": 0.0622, | |
| "num_input_tokens_seen": 39714816, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 4.221453287197232, | |
| "grad_norm": 0.22935072238631152, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 0.0654, | |
| "num_input_tokens_seen": 39976960, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 4.249134948096886, | |
| "grad_norm": 0.23955884885861536, | |
| "learning_rate": 3.974757327377981e-06, | |
| "loss": 0.066, | |
| "num_input_tokens_seen": 40239104, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 4.27681660899654, | |
| "grad_norm": 0.2470523646097344, | |
| "learning_rate": 3.6408072716606346e-06, | |
| "loss": 0.0709, | |
| "num_input_tokens_seen": 40501248, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 4.304498269896194, | |
| "grad_norm": 0.25472932869095477, | |
| "learning_rate": 3.3209786751399187e-06, | |
| "loss": 0.0628, | |
| "num_input_tokens_seen": 40763392, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 4.3321799307958475, | |
| "grad_norm": 0.27556890029509057, | |
| "learning_rate": 3.0153689607045845e-06, | |
| "loss": 0.0646, | |
| "num_input_tokens_seen": 41025536, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.359861591695502, | |
| "grad_norm": 0.2380980366901671, | |
| "learning_rate": 2.724071220034158e-06, | |
| "loss": 0.0636, | |
| "num_input_tokens_seen": 41287680, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 4.387543252595155, | |
| "grad_norm": 0.25550042168413734, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 0.0667, | |
| "num_input_tokens_seen": 41549824, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 4.41522491349481, | |
| "grad_norm": 0.24537341237199822, | |
| "learning_rate": 2.1847622018482283e-06, | |
| "loss": 0.0633, | |
| "num_input_tokens_seen": 41811968, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 4.442906574394463, | |
| "grad_norm": 0.25994322940034775, | |
| "learning_rate": 1.9369152030840556e-06, | |
| "loss": 0.0689, | |
| "num_input_tokens_seen": 42074112, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 4.470588235294118, | |
| "grad_norm": 0.2439059244202511, | |
| "learning_rate": 1.70370868554659e-06, | |
| "loss": 0.0657, | |
| "num_input_tokens_seen": 42336256, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 4.498269896193771, | |
| "grad_norm": 0.2506318904052677, | |
| "learning_rate": 1.4852136862001764e-06, | |
| "loss": 0.0618, | |
| "num_input_tokens_seen": 42598400, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 4.525951557093426, | |
| "grad_norm": 0.23653008472997497, | |
| "learning_rate": 1.2814967607382432e-06, | |
| "loss": 0.0629, | |
| "num_input_tokens_seen": 42860544, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 4.553633217993079, | |
| "grad_norm": 0.26312867947287943, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 0.0706, | |
| "num_input_tokens_seen": 43122688, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 4.581314878892734, | |
| "grad_norm": 0.23016888232856603, | |
| "learning_rate": 9.186408276168013e-07, | |
| "loss": 0.0613, | |
| "num_input_tokens_seen": 43384832, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 4.608996539792388, | |
| "grad_norm": 0.24290205151293878, | |
| "learning_rate": 7.596123493895991e-07, | |
| "loss": 0.06, | |
| "num_input_tokens_seen": 43646976, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 4.6366782006920415, | |
| "grad_norm": 0.24284900083654315, | |
| "learning_rate": 6.15582970243117e-07, | |
| "loss": 0.0649, | |
| "num_input_tokens_seen": 43909120, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 4.664359861591695, | |
| "grad_norm": 0.24318407581731105, | |
| "learning_rate": 4.865965629214819e-07, | |
| "loss": 0.0665, | |
| "num_input_tokens_seen": 44171264, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 4.692041522491349, | |
| "grad_norm": 0.24512300511792165, | |
| "learning_rate": 3.7269241793390085e-07, | |
| "loss": 0.0718, | |
| "num_input_tokens_seen": 44433408, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 4.719723183391004, | |
| "grad_norm": 0.24461464645997066, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 0.0686, | |
| "num_input_tokens_seen": 44695552, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 4.747404844290657, | |
| "grad_norm": 0.24433953012382653, | |
| "learning_rate": 1.9026509541272275e-07, | |
| "loss": 0.069, | |
| "num_input_tokens_seen": 44957696, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 4.775086505190312, | |
| "grad_norm": 0.23366843489201422, | |
| "learning_rate": 1.2179748700879012e-07, | |
| "loss": 0.0581, | |
| "num_input_tokens_seen": 45219840, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 4.802768166089965, | |
| "grad_norm": 0.23890311459194627, | |
| "learning_rate": 6.852326227130834e-08, | |
| "loss": 0.0647, | |
| "num_input_tokens_seen": 45481984, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 4.83044982698962, | |
| "grad_norm": 0.23100179334343116, | |
| "learning_rate": 3.04586490452119e-08, | |
| "loss": 0.0635, | |
| "num_input_tokens_seen": 45744128, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 4.858131487889273, | |
| "grad_norm": 0.2425970529110237, | |
| "learning_rate": 7.615242180436522e-09, | |
| "loss": 0.063, | |
| "num_input_tokens_seen": 46006272, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 4.885813148788928, | |
| "grad_norm": 0.2508832655720532, | |
| "learning_rate": 0.0, | |
| "loss": 0.0662, | |
| "num_input_tokens_seen": 46268416, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 4.885813148788928, | |
| "num_input_tokens_seen": 46268416, | |
| "step": 180, | |
| "total_flos": 366014160175104.0, | |
| "train_loss": 0.42912846936119925, | |
| "train_runtime": 30969.1597, | |
| "train_samples_per_second": 1.491, | |
| "train_steps_per_second": 0.006 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 180, | |
| "num_input_tokens_seen": 46268416, | |
| "num_train_epochs": 5, | |
| "save_steps": 1000000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 366014160175104.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |