| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.8596001859600186, | |
| "eval_steps": 500, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004649000464900047, | |
| "grad_norm": 0.3293333649635315, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 1.2004, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.009298000929800094, | |
| "grad_norm": 0.3225855827331543, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 1.2289, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.01394700139470014, | |
| "grad_norm": 0.3228434920310974, | |
| "learning_rate": 5e-05, | |
| "loss": 1.2513, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.018596001859600187, | |
| "grad_norm": 0.3422330915927887, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 1.2952, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.023245002324500233, | |
| "grad_norm": 0.35845088958740234, | |
| "learning_rate": 8.333333333333334e-05, | |
| "loss": 1.3072, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02789400278940028, | |
| "grad_norm": 0.41468068957328796, | |
| "learning_rate": 0.0001, | |
| "loss": 1.4565, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.032543003254300325, | |
| "grad_norm": 0.45462965965270996, | |
| "learning_rate": 0.00011666666666666668, | |
| "loss": 1.4221, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.037192003719200374, | |
| "grad_norm": 0.5441383719444275, | |
| "learning_rate": 0.00013333333333333334, | |
| "loss": 1.4777, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.04184100418410042, | |
| "grad_norm": 0.7736104726791382, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 1.5793, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.046490004649000466, | |
| "grad_norm": 1.1128149032592773, | |
| "learning_rate": 0.0001666666666666667, | |
| "loss": 1.7799, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05113900511390051, | |
| "grad_norm": 1.295057773590088, | |
| "learning_rate": 0.00018333333333333334, | |
| "loss": 1.6539, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.05578800557880056, | |
| "grad_norm": 2.0307557582855225, | |
| "learning_rate": 0.0002, | |
| "loss": 1.6614, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.06043700604370061, | |
| "grad_norm": 1.489993691444397, | |
| "learning_rate": 0.0001999970498341241, | |
| "loss": 1.3891, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.06508600650860065, | |
| "grad_norm": 0.45855778455734253, | |
| "learning_rate": 0.0001999881995299069, | |
| "loss": 1.1051, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0697350069735007, | |
| "grad_norm": 0.464190274477005, | |
| "learning_rate": 0.0001999734496675677, | |
| "loss": 1.1558, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.07438400743840075, | |
| "grad_norm": 0.47878339886665344, | |
| "learning_rate": 0.00019995280121409636, | |
| "loss": 1.1533, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.07903300790330078, | |
| "grad_norm": 0.4506080448627472, | |
| "learning_rate": 0.00019992625552318972, | |
| "loss": 1.2074, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.08368200836820083, | |
| "grad_norm": 0.4147086441516876, | |
| "learning_rate": 0.00019989381433516316, | |
| "loss": 1.274, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.08833100883310088, | |
| "grad_norm": 0.3981817066669464, | |
| "learning_rate": 0.00019985547977683643, | |
| "loss": 1.2457, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.09298000929800093, | |
| "grad_norm": 0.3968053162097931, | |
| "learning_rate": 0.00019981125436139405, | |
| "loss": 1.1961, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09762900976290098, | |
| "grad_norm": 0.5115089416503906, | |
| "learning_rate": 0.00019976114098822073, | |
| "loss": 1.228, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.10227801022780102, | |
| "grad_norm": 0.5449497103691101, | |
| "learning_rate": 0.00019970514294271124, | |
| "loss": 1.1062, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.10692701069270107, | |
| "grad_norm": 0.5306552052497864, | |
| "learning_rate": 0.00019964326389605496, | |
| "loss": 1.0354, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.11157601115760112, | |
| "grad_norm": 0.5531755089759827, | |
| "learning_rate": 0.00019957550790499526, | |
| "loss": 1.0759, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.11622501162250116, | |
| "grad_norm": 0.9215659499168396, | |
| "learning_rate": 0.00019950187941156355, | |
| "loss": 1.0, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.12087401208740121, | |
| "grad_norm": 0.2715967297554016, | |
| "learning_rate": 0.00019942238324278803, | |
| "loss": 1.0803, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.12552301255230125, | |
| "grad_norm": 0.24064235389232635, | |
| "learning_rate": 0.00019933702461037716, | |
| "loss": 1.0892, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.1301720130172013, | |
| "grad_norm": 0.18895158171653748, | |
| "learning_rate": 0.00019924580911037827, | |
| "loss": 1.1303, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.13482101348210135, | |
| "grad_norm": 0.1712905615568161, | |
| "learning_rate": 0.00019914874272281032, | |
| "loss": 1.1635, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.1394700139470014, | |
| "grad_norm": 0.18927285075187683, | |
| "learning_rate": 0.00019904583181127206, | |
| "loss": 1.1706, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.14411901441190145, | |
| "grad_norm": 0.2120945006608963, | |
| "learning_rate": 0.0001989370831225248, | |
| "loss": 1.1774, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.1487680148768015, | |
| "grad_norm": 0.2811720371246338, | |
| "learning_rate": 0.00019882250378605015, | |
| "loss": 1.1867, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.15341701534170155, | |
| "grad_norm": 0.3283132016658783, | |
| "learning_rate": 0.00019870210131358253, | |
| "loss": 1.2136, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.15806601580660157, | |
| "grad_norm": 0.3853696286678314, | |
| "learning_rate": 0.0001985758835986167, | |
| "loss": 1.1155, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.16271501627150162, | |
| "grad_norm": 0.3992000222206116, | |
| "learning_rate": 0.0001984438589158903, | |
| "loss": 1.071, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.16736401673640167, | |
| "grad_norm": 0.3955259621143341, | |
| "learning_rate": 0.0001983060359208415, | |
| "loss": 0.9018, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.17201301720130172, | |
| "grad_norm": 0.44543391466140747, | |
| "learning_rate": 0.00019816242364904132, | |
| "loss": 0.8921, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.17666201766620176, | |
| "grad_norm": 0.5863587856292725, | |
| "learning_rate": 0.00019801303151560138, | |
| "loss": 1.0336, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.18131101813110181, | |
| "grad_norm": 0.13918030261993408, | |
| "learning_rate": 0.0001978578693145566, | |
| "loss": 1.1035, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.18596001859600186, | |
| "grad_norm": 0.14183400571346283, | |
| "learning_rate": 0.00019769694721822337, | |
| "loss": 1.103, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1906090190609019, | |
| "grad_norm": 0.11069459468126297, | |
| "learning_rate": 0.00019753027577653213, | |
| "loss": 1.11, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.19525801952580196, | |
| "grad_norm": 0.1001875028014183, | |
| "learning_rate": 0.00019735786591633633, | |
| "loss": 1.1537, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.199907019990702, | |
| "grad_norm": 0.137644425034523, | |
| "learning_rate": 0.0001971797289406956, | |
| "loss": 1.1678, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.20455602045560203, | |
| "grad_norm": 0.13967232406139374, | |
| "learning_rate": 0.00019699587652813503, | |
| "loss": 1.147, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.20920502092050208, | |
| "grad_norm": 0.19884178042411804, | |
| "learning_rate": 0.00019680632073187931, | |
| "loss": 1.1424, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.21385402138540213, | |
| "grad_norm": 0.22492916882038116, | |
| "learning_rate": 0.00019661107397906275, | |
| "loss": 1.061, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.21850302185030218, | |
| "grad_norm": 0.2543758153915405, | |
| "learning_rate": 0.00019641014906991437, | |
| "loss": 0.9985, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.22315202231520223, | |
| "grad_norm": 0.3426196873188019, | |
| "learning_rate": 0.00019620355917691884, | |
| "loss": 0.935, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.22780102278010228, | |
| "grad_norm": 0.39695194363594055, | |
| "learning_rate": 0.00019599131784395297, | |
| "loss": 0.8601, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.23245002324500233, | |
| "grad_norm": 0.762519359588623, | |
| "learning_rate": 0.00019577343898539748, | |
| "loss": 0.9861, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.23709902370990238, | |
| "grad_norm": 0.10567212849855423, | |
| "learning_rate": 0.00019554993688522524, | |
| "loss": 1.0594, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.24174802417480243, | |
| "grad_norm": 0.11329612880945206, | |
| "learning_rate": 0.00019532082619606436, | |
| "loss": 1.0659, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.24639702463970248, | |
| "grad_norm": 0.1084655150771141, | |
| "learning_rate": 0.00019508612193823793, | |
| "loss": 1.0896, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.2510460251046025, | |
| "grad_norm": 0.1181301698088646, | |
| "learning_rate": 0.00019484583949877908, | |
| "loss": 1.1282, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.2556950255695026, | |
| "grad_norm": 0.14401216804981232, | |
| "learning_rate": 0.00019459999463042237, | |
| "loss": 1.118, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.2603440260344026, | |
| "grad_norm": 0.1677497774362564, | |
| "learning_rate": 0.00019434860345057096, | |
| "loss": 1.1425, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.2649930264993027, | |
| "grad_norm": 0.20358914136886597, | |
| "learning_rate": 0.00019409168244023987, | |
| "loss": 1.1249, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.2696420269642027, | |
| "grad_norm": 0.26533111929893494, | |
| "learning_rate": 0.00019382924844297582, | |
| "loss": 1.1438, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.2742910274291027, | |
| "grad_norm": 0.33439338207244873, | |
| "learning_rate": 0.0001935613186637526, | |
| "loss": 1.0949, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.2789400278940028, | |
| "grad_norm": 0.4424538016319275, | |
| "learning_rate": 0.0001932879106678434, | |
| "loss": 1.0185, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2835890283589028, | |
| "grad_norm": 0.4415128827095032, | |
| "learning_rate": 0.00019300904237966906, | |
| "loss": 0.9373, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.2882380288238029, | |
| "grad_norm": 0.4046485424041748, | |
| "learning_rate": 0.00019272473208162313, | |
| "loss": 0.8202, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.2928870292887029, | |
| "grad_norm": 0.324368417263031, | |
| "learning_rate": 0.00019243499841287308, | |
| "loss": 0.9427, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.297536029753603, | |
| "grad_norm": 0.07730630785226822, | |
| "learning_rate": 0.00019213986036813863, | |
| "loss": 1.0637, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.302185030218503, | |
| "grad_norm": 0.08834154158830643, | |
| "learning_rate": 0.0001918393372964461, | |
| "loss": 1.0867, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.3068340306834031, | |
| "grad_norm": 0.09792964160442352, | |
| "learning_rate": 0.00019153344889986023, | |
| "loss": 1.0948, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.3114830311483031, | |
| "grad_norm": 0.08694823086261749, | |
| "learning_rate": 0.0001912222152321923, | |
| "loss": 1.1405, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.31613203161320313, | |
| "grad_norm": 0.12065356969833374, | |
| "learning_rate": 0.0001909056566976856, | |
| "loss": 1.1431, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.3207810320781032, | |
| "grad_norm": 0.1258689910173416, | |
| "learning_rate": 0.00019058379404967757, | |
| "loss": 1.1709, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.32543003254300323, | |
| "grad_norm": 0.14431503415107727, | |
| "learning_rate": 0.0001902566483892393, | |
| "loss": 1.0588, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.3300790330079033, | |
| "grad_norm": 0.1783875674009323, | |
| "learning_rate": 0.00018992424116379228, | |
| "loss": 1.0258, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.33472803347280333, | |
| "grad_norm": 0.23002314567565918, | |
| "learning_rate": 0.00018958659416570212, | |
| "loss": 0.9455, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.3393770339377034, | |
| "grad_norm": 0.305070161819458, | |
| "learning_rate": 0.00018924372953084997, | |
| "loss": 0.9194, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.34402603440260343, | |
| "grad_norm": 0.3687322437763214, | |
| "learning_rate": 0.0001888956697371813, | |
| "loss": 0.883, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.3486750348675035, | |
| "grad_norm": 0.6616573333740234, | |
| "learning_rate": 0.00018854243760323223, | |
| "loss": 0.872, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.35332403533240353, | |
| "grad_norm": 0.09560154378414154, | |
| "learning_rate": 0.0001881840562866336, | |
| "loss": 1.0362, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.3579730357973036, | |
| "grad_norm": 0.09787497669458389, | |
| "learning_rate": 0.00018782054928259277, | |
| "loss": 1.0836, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.36262203626220363, | |
| "grad_norm": 0.12214916199445724, | |
| "learning_rate": 0.0001874519404223533, | |
| "loss": 1.0993, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.36727103672710365, | |
| "grad_norm": 0.1271420419216156, | |
| "learning_rate": 0.00018707825387163248, | |
| "loss": 1.1327, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.3719200371920037, | |
| "grad_norm": 0.15113487839698792, | |
| "learning_rate": 0.00018669951412903725, | |
| "loss": 1.1284, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.37656903765690375, | |
| "grad_norm": 0.1705760657787323, | |
| "learning_rate": 0.00018631574602445792, | |
| "loss": 1.1973, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.3812180381218038, | |
| "grad_norm": 0.1751885861158371, | |
| "learning_rate": 0.0001859269747174404, | |
| "loss": 1.1302, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.38586703858670385, | |
| "grad_norm": 0.23309065401554108, | |
| "learning_rate": 0.00018553322569553682, | |
| "loss": 1.1091, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.3905160390516039, | |
| "grad_norm": 0.30925455689430237, | |
| "learning_rate": 0.0001851345247726344, | |
| "loss": 1.0555, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.39516503951650395, | |
| "grad_norm": 0.41445159912109375, | |
| "learning_rate": 0.00018473089808726336, | |
| "loss": 0.9417, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.399814039981404, | |
| "grad_norm": 0.3391374945640564, | |
| "learning_rate": 0.00018432237210088307, | |
| "loss": 0.9262, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.40446304044630405, | |
| "grad_norm": 0.43099090456962585, | |
| "learning_rate": 0.00018390897359614748, | |
| "loss": 0.77, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.40911204091120407, | |
| "grad_norm": 0.3982544243335724, | |
| "learning_rate": 0.00018349072967514896, | |
| "loss": 0.9355, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.41376104137610414, | |
| "grad_norm": 0.06400130689144135, | |
| "learning_rate": 0.00018306766775764196, | |
| "loss": 1.0571, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.41841004184100417, | |
| "grad_norm": 0.0744890421628952, | |
| "learning_rate": 0.00018263981557924483, | |
| "loss": 1.0519, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.42305904230590424, | |
| "grad_norm": 0.08555030077695847, | |
| "learning_rate": 0.00018220720118962205, | |
| "loss": 1.1044, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.42770804277080426, | |
| "grad_norm": 0.09533877670764923, | |
| "learning_rate": 0.00018176985295064487, | |
| "loss": 1.1114, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.43235704323570434, | |
| "grad_norm": 0.11658069491386414, | |
| "learning_rate": 0.00018132779953453226, | |
| "loss": 1.1064, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.43700604370060436, | |
| "grad_norm": 0.13404923677444458, | |
| "learning_rate": 0.00018088106992197091, | |
| "loss": 1.1292, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.44165504416550444, | |
| "grad_norm": 0.16842801868915558, | |
| "learning_rate": 0.00018042969340021546, | |
| "loss": 1.1256, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.44630404463040446, | |
| "grad_norm": 0.24980179965496063, | |
| "learning_rate": 0.00017997369956116845, | |
| "loss": 1.0524, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.4509530450953045, | |
| "grad_norm": 0.27264076471328735, | |
| "learning_rate": 0.00017951311829944014, | |
| "loss": 0.9696, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.45560204556020456, | |
| "grad_norm": 0.33832067251205444, | |
| "learning_rate": 0.00017904797981038874, | |
| "loss": 0.9045, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.4602510460251046, | |
| "grad_norm": 0.40327176451683044, | |
| "learning_rate": 0.00017857831458814098, | |
| "loss": 0.8434, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.46490004649000466, | |
| "grad_norm": 0.7074009776115417, | |
| "learning_rate": 0.00017810415342359257, | |
| "loss": 0.9095, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4695490469549047, | |
| "grad_norm": 0.0739617720246315, | |
| "learning_rate": 0.00017762552740238998, | |
| "loss": 1.0309, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.47419804741980476, | |
| "grad_norm": 0.0909004807472229, | |
| "learning_rate": 0.00017714246790289214, | |
| "loss": 1.0933, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.4788470478847048, | |
| "grad_norm": 0.0961037203669548, | |
| "learning_rate": 0.0001766550065941136, | |
| "loss": 1.0684, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.48349604834960486, | |
| "grad_norm": 0.11135982722043991, | |
| "learning_rate": 0.00017616317543364804, | |
| "loss": 1.118, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.4881450488145049, | |
| "grad_norm": 0.11919623613357544, | |
| "learning_rate": 0.00017566700666557346, | |
| "loss": 1.1175, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.49279404927940496, | |
| "grad_norm": 0.1320277750492096, | |
| "learning_rate": 0.00017516653281833794, | |
| "loss": 1.113, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.497443049744305, | |
| "grad_norm": 0.15232138335704803, | |
| "learning_rate": 0.00017466178670262747, | |
| "loss": 1.1451, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.502092050209205, | |
| "grad_norm": 0.17556369304656982, | |
| "learning_rate": 0.00017415280140921463, | |
| "loss": 1.0489, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.506741050674105, | |
| "grad_norm": 0.2791363596916199, | |
| "learning_rate": 0.00017363961030678927, | |
| "loss": 1.0767, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.5113900511390052, | |
| "grad_norm": 0.39919206500053406, | |
| "learning_rate": 0.00017312224703977094, | |
| "loss": 0.999, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.5160390516039052, | |
| "grad_norm": 0.41332653164863586, | |
| "learning_rate": 0.00017260074552610306, | |
| "loss": 0.8625, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.5206880520688052, | |
| "grad_norm": 0.41465967893600464, | |
| "learning_rate": 0.00017207513995502939, | |
| "loss": 0.8294, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.5253370525337052, | |
| "grad_norm": 0.35807111859321594, | |
| "learning_rate": 0.00017154546478485264, | |
| "loss": 0.9328, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.5299860529986054, | |
| "grad_norm": 0.18978600203990936, | |
| "learning_rate": 0.0001710117547406753, | |
| "loss": 1.0327, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.5346350534635054, | |
| "grad_norm": 0.16808465123176575, | |
| "learning_rate": 0.00017047404481212314, | |
| "loss": 1.0977, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.5392840539284054, | |
| "grad_norm": 0.1754007339477539, | |
| "learning_rate": 0.0001699323702510513, | |
| "loss": 1.063, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.5439330543933054, | |
| "grad_norm": 0.15331235527992249, | |
| "learning_rate": 0.0001693867665692333, | |
| "loss": 1.1018, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.5485820548582054, | |
| "grad_norm": 0.13346055150032043, | |
| "learning_rate": 0.00016883726953603273, | |
| "loss": 1.1215, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.5532310553231056, | |
| "grad_norm": 0.14829131960868835, | |
| "learning_rate": 0.00016828391517605845, | |
| "loss": 1.1484, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.5578800557880056, | |
| "grad_norm": 0.14702975749969482, | |
| "learning_rate": 0.0001677267397668026, | |
| "loss": 1.0535, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.5625290562529056, | |
| "grad_norm": 0.21014554798603058, | |
| "learning_rate": 0.00016716577983626259, | |
| "loss": 1.1138, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.5671780567178056, | |
| "grad_norm": 0.2756856679916382, | |
| "learning_rate": 0.000166601072160546, | |
| "loss": 0.9879, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.5718270571827058, | |
| "grad_norm": 0.39934659004211426, | |
| "learning_rate": 0.0001660326537614599, | |
| "loss": 0.9133, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.5764760576476058, | |
| "grad_norm": 0.37866297364234924, | |
| "learning_rate": 0.0001654605619040835, | |
| "loss": 0.9116, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.5811250581125058, | |
| "grad_norm": 0.479396253824234, | |
| "learning_rate": 0.00016488483409432504, | |
| "loss": 0.8499, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5857740585774058, | |
| "grad_norm": 0.0876205563545227, | |
| "learning_rate": 0.00016430550807646323, | |
| "loss": 1.0378, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.5904230590423059, | |
| "grad_norm": 0.09576641768217087, | |
| "learning_rate": 0.00016372262183067247, | |
| "loss": 1.0617, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.595072059507206, | |
| "grad_norm": 0.08960004895925522, | |
| "learning_rate": 0.00016313621357053306, | |
| "loss": 1.0774, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.599721059972106, | |
| "grad_norm": 0.0907701924443245, | |
| "learning_rate": 0.00016254632174052578, | |
| "loss": 1.0989, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.604370060437006, | |
| "grad_norm": 0.09422672539949417, | |
| "learning_rate": 0.00016195298501351177, | |
| "loss": 1.1659, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.609019060901906, | |
| "grad_norm": 0.10595980286598206, | |
| "learning_rate": 0.00016135624228819683, | |
| "loss": 1.1642, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.6136680613668062, | |
| "grad_norm": 0.12369013577699661, | |
| "learning_rate": 0.00016075613268658157, | |
| "loss": 1.1369, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.6183170618317062, | |
| "grad_norm": 0.16726712882518768, | |
| "learning_rate": 0.00016015269555139642, | |
| "loss": 1.0458, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.6229660622966062, | |
| "grad_norm": 0.21234917640686035, | |
| "learning_rate": 0.00015954597044352234, | |
| "loss": 1.0013, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.6276150627615062, | |
| "grad_norm": 0.28728505969047546, | |
| "learning_rate": 0.00015893599713939728, | |
| "loss": 0.9075, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.6322640632264063, | |
| "grad_norm": 0.4090898334980011, | |
| "learning_rate": 0.00015832281562840856, | |
| "loss": 0.9677, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.6369130636913064, | |
| "grad_norm": 0.3748779296875, | |
| "learning_rate": 0.000157706466110271, | |
| "loss": 0.7873, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.6415620641562064, | |
| "grad_norm": 0.31243249773979187, | |
| "learning_rate": 0.00015708698899239172, | |
| "loss": 0.8241, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.6462110646211064, | |
| "grad_norm": 0.08351978659629822, | |
| "learning_rate": 0.00015646442488722074, | |
| "loss": 1.0431, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.6508600650860065, | |
| "grad_norm": 0.08479636162519455, | |
| "learning_rate": 0.00015583881460958868, | |
| "loss": 1.0725, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.6555090655509066, | |
| "grad_norm": 0.09441729635000229, | |
| "learning_rate": 0.000155210199174031, | |
| "loss": 1.076, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.6601580660158066, | |
| "grad_norm": 0.10794027149677277, | |
| "learning_rate": 0.0001545786197920989, | |
| "loss": 1.1112, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.6648070664807066, | |
| "grad_norm": 0.11890177428722382, | |
| "learning_rate": 0.00015394411786965776, | |
| "loss": 1.1748, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.6694560669456067, | |
| "grad_norm": 0.13239571452140808, | |
| "learning_rate": 0.0001533067350041725, | |
| "loss": 1.1075, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.6741050674105067, | |
| "grad_norm": 0.1535806506872177, | |
| "learning_rate": 0.00015266651298198033, | |
| "loss": 1.1259, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.6787540678754068, | |
| "grad_norm": 0.19703824818134308, | |
| "learning_rate": 0.00015202349377555166, | |
| "loss": 1.0655, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.6834030683403068, | |
| "grad_norm": 0.2627493441104889, | |
| "learning_rate": 0.00015137771954073804, | |
| "loss": 0.9644, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.6880520688052069, | |
| "grad_norm": 0.3154362142086029, | |
| "learning_rate": 0.0001507292326140085, | |
| "loss": 0.8241, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.6927010692701069, | |
| "grad_norm": 0.3660978078842163, | |
| "learning_rate": 0.0001500780755096743, | |
| "loss": 0.9218, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.697350069735007, | |
| "grad_norm": 0.4281309247016907, | |
| "learning_rate": 0.00014942429091710141, | |
| "loss": 0.7471, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.701999070199907, | |
| "grad_norm": 0.058116402477025986, | |
| "learning_rate": 0.00014876792169791193, | |
| "loss": 1.0336, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.7066480706648071, | |
| "grad_norm": 0.06079982966184616, | |
| "learning_rate": 0.00014810901088317414, | |
| "loss": 1.0446, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.7112970711297071, | |
| "grad_norm": 0.06905698031187057, | |
| "learning_rate": 0.00014744760167058137, | |
| "loss": 1.0841, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.7159460715946072, | |
| "grad_norm": 0.07570036500692368, | |
| "learning_rate": 0.00014678373742162007, | |
| "loss": 1.0895, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.7205950720595072, | |
| "grad_norm": 0.08454253524541855, | |
| "learning_rate": 0.00014611746165872698, | |
| "loss": 1.1083, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.7252440725244073, | |
| "grad_norm": 0.10023923218250275, | |
| "learning_rate": 0.00014544881806243583, | |
| "loss": 1.0951, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.7298930729893073, | |
| "grad_norm": 0.12155482918024063, | |
| "learning_rate": 0.00014477785046851385, | |
| "loss": 1.092, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.7345420734542073, | |
| "grad_norm": 0.15077327191829681, | |
| "learning_rate": 0.00014410460286508762, | |
| "loss": 1.0582, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.7391910739191074, | |
| "grad_norm": 0.19081415235996246, | |
| "learning_rate": 0.00014342911938975948, | |
| "loss": 1.0055, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.7438400743840075, | |
| "grad_norm": 0.24974983930587769, | |
| "learning_rate": 0.0001427514443267139, | |
| "loss": 0.9043, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.7484890748489075, | |
| "grad_norm": 0.33478638529777527, | |
| "learning_rate": 0.00014207162210381404, | |
| "loss": 0.8285, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.7531380753138075, | |
| "grad_norm": 0.34924110770225525, | |
| "learning_rate": 0.0001413896972896894, | |
| "loss": 0.774, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.7577870757787076, | |
| "grad_norm": 0.2992526888847351, | |
| "learning_rate": 0.00014070571459081366, | |
| "loss": 0.8779, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.7624360762436077, | |
| "grad_norm": 0.07906091213226318, | |
| "learning_rate": 0.0001400197188485739, | |
| "loss": 1.047, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.7670850767085077, | |
| "grad_norm": 0.09800871461629868, | |
| "learning_rate": 0.00013933175503633068, | |
| "loss": 1.0439, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.7717340771734077, | |
| "grad_norm": 0.09857136756181717, | |
| "learning_rate": 0.00013864186825646995, | |
| "loss": 1.0522, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.7763830776383077, | |
| "grad_norm": 0.1110193282365799, | |
| "learning_rate": 0.00013795010373744582, | |
| "loss": 1.1126, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.7810320781032078, | |
| "grad_norm": 0.11846626549959183, | |
| "learning_rate": 0.00013725650683081556, | |
| "loss": 1.0925, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.7856810785681079, | |
| "grad_norm": 0.14214776456356049, | |
| "learning_rate": 0.00013656112300826646, | |
| "loss": 1.1323, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.7903300790330079, | |
| "grad_norm": 0.1606011986732483, | |
| "learning_rate": 0.00013586399785863454, | |
| "loss": 1.0505, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.7949790794979079, | |
| "grad_norm": 0.18738731741905212, | |
| "learning_rate": 0.000135165177084916, | |
| "loss": 1.0334, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.799628079962808, | |
| "grad_norm": 0.23303773999214172, | |
| "learning_rate": 0.0001344647065012709, | |
| "loss": 0.9471, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.8042770804277081, | |
| "grad_norm": 0.27448582649230957, | |
| "learning_rate": 0.00013376263203001938, | |
| "loss": 0.8672, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.8089260808926081, | |
| "grad_norm": 0.29609808325767517, | |
| "learning_rate": 0.0001330589996986315, | |
| "loss": 0.7936, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.8135750813575081, | |
| "grad_norm": 0.43915602564811707, | |
| "learning_rate": 0.00013235385563670934, | |
| "loss": 0.7688, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.8182240818224081, | |
| "grad_norm": 0.06484824419021606, | |
| "learning_rate": 0.00013164724607296285, | |
| "loss": 1.0403, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.8228730822873083, | |
| "grad_norm": 0.07361859828233719, | |
| "learning_rate": 0.00013093921733217916, | |
| "loss": 1.0539, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.8275220827522083, | |
| "grad_norm": 0.08026642352342606, | |
| "learning_rate": 0.00013022981583218565, | |
| "loss": 1.0596, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.8321710832171083, | |
| "grad_norm": 0.08283592760562897, | |
| "learning_rate": 0.0001295190880808067, | |
| "loss": 1.0435, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.8368200836820083, | |
| "grad_norm": 0.11680889129638672, | |
| "learning_rate": 0.00012880708067281477, | |
| "loss": 1.1464, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.8414690841469085, | |
| "grad_norm": 0.10784471035003662, | |
| "learning_rate": 0.00012809384028687553, | |
| "loss": 1.1004, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.8461180846118085, | |
| "grad_norm": 0.1224328875541687, | |
| "learning_rate": 0.00012737941368248792, | |
| "loss": 1.0699, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.8507670850767085, | |
| "grad_norm": 0.15732775628566742, | |
| "learning_rate": 0.0001266638476969183, | |
| "loss": 1.0579, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.8554160855416085, | |
| "grad_norm": 0.18987177312374115, | |
| "learning_rate": 0.00012594718924213008, | |
| "loss": 1.0212, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.8600650860065086, | |
| "grad_norm": 0.250615656375885, | |
| "learning_rate": 0.00012522948530170806, | |
| "loss": 0.9817, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.8647140864714087, | |
| "grad_norm": 0.28331658244132996, | |
| "learning_rate": 0.00012451078292777837, | |
| "loss": 0.8079, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.8693630869363087, | |
| "grad_norm": 0.3439493477344513, | |
| "learning_rate": 0.0001237911292379237, | |
| "loss": 0.7382, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.8740120874012087, | |
| "grad_norm": 0.35732191801071167, | |
| "learning_rate": 0.00012307057141209415, | |
| "loss": 0.9792, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.8786610878661087, | |
| "grad_norm": 0.10152771323919296, | |
| "learning_rate": 0.0001223491566895144, | |
| "loss": 1.0674, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.8833100883310089, | |
| "grad_norm": 0.11470180004835129, | |
| "learning_rate": 0.00012162693236558658, | |
| "loss": 1.0276, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.8879590887959089, | |
| "grad_norm": 0.1217978298664093, | |
| "learning_rate": 0.00012090394578878974, | |
| "loss": 1.0734, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.8926080892608089, | |
| "grad_norm": 0.11547485738992691, | |
| "learning_rate": 0.0001201802443575756, | |
| "loss": 1.0862, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.897257089725709, | |
| "grad_norm": 0.13204635679721832, | |
| "learning_rate": 0.00011945587551726116, | |
| "loss": 1.1112, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.901906090190609, | |
| "grad_norm": 0.12314517050981522, | |
| "learning_rate": 0.00011873088675691835, | |
| "loss": 1.1342, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.9065550906555091, | |
| "grad_norm": 0.14891035854816437, | |
| "learning_rate": 0.00011800532560626048, | |
| "loss": 1.0975, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.9112040911204091, | |
| "grad_norm": 0.17384392023086548, | |
| "learning_rate": 0.0001172792396325264, | |
| "loss": 1.0599, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.9158530915853091, | |
| "grad_norm": 0.22434796392917633, | |
| "learning_rate": 0.00011655267643736194, | |
| "loss": 1.0202, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.9205020920502092, | |
| "grad_norm": 0.26256677508354187, | |
| "learning_rate": 0.00011582568365369924, | |
| "loss": 0.9057, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.9251510925151093, | |
| "grad_norm": 0.30824849009513855, | |
| "learning_rate": 0.00011509830894263387, | |
| "loss": 0.8073, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.9298000929800093, | |
| "grad_norm": 0.4767857789993286, | |
| "learning_rate": 0.00011437059999030035, | |
| "loss": 0.806, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.9344490934449093, | |
| "grad_norm": 0.05803222209215164, | |
| "learning_rate": 0.00011364260450474575, | |
| "loss": 1.0481, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.9390980939098094, | |
| "grad_norm": 0.06460921466350555, | |
| "learning_rate": 0.00011291437021280205, | |
| "loss": 1.0617, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.9437470943747094, | |
| "grad_norm": 0.07504323869943619, | |
| "learning_rate": 0.0001121859448569572, | |
| "loss": 1.0995, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.9483960948396095, | |
| "grad_norm": 0.07197124511003494, | |
| "learning_rate": 0.00011145737619222516, | |
| "loss": 1.0629, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.9530450953045095, | |
| "grad_norm": 0.08467988669872284, | |
| "learning_rate": 0.0001107287119830151, | |
| "loss": 1.0961, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.9576940957694096, | |
| "grad_norm": 0.10283592343330383, | |
| "learning_rate": 0.00011000000000000002, | |
| "loss": 1.1351, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.9623430962343096, | |
| "grad_norm": 0.11667031794786453, | |
| "learning_rate": 0.00010927128801698494, | |
| "loss": 1.0459, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.9669920966992097, | |
| "grad_norm": 0.1426560878753662, | |
| "learning_rate": 0.00010854262380777486, | |
| "loss": 1.0367, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.9716410971641097, | |
| "grad_norm": 0.18659153580665588, | |
| "learning_rate": 0.00010781405514304284, | |
| "loss": 0.9566, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.9762900976290098, | |
| "grad_norm": 0.2517950236797333, | |
| "learning_rate": 0.000107085629787198, | |
| "loss": 0.9468, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.9809390980939098, | |
| "grad_norm": 0.27402400970458984, | |
| "learning_rate": 0.0001063573954952543, | |
| "loss": 0.8093, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.9855880985588099, | |
| "grad_norm": 0.339691698551178, | |
| "learning_rate": 0.0001056294000096997, | |
| "loss": 0.7326, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.9902370990237099, | |
| "grad_norm": 0.36932969093322754, | |
| "learning_rate": 0.00010490169105736613, | |
| "loss": 0.9435, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.99488609948861, | |
| "grad_norm": 0.09353512525558472, | |
| "learning_rate": 0.0001041743163463008, | |
| "loss": 1.1068, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.99953509995351, | |
| "grad_norm": 0.249754399061203, | |
| "learning_rate": 0.00010344732356263808, | |
| "loss": 0.8395, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.00418410041841, | |
| "grad_norm": 0.0985850915312767, | |
| "learning_rate": 0.00010272076036747365, | |
| "loss": 0.9939, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.00883310088331, | |
| "grad_norm": 0.06556614488363266, | |
| "learning_rate": 0.00010199467439373956, | |
| "loss": 1.0809, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.01348210134821, | |
| "grad_norm": 0.07044567912817001, | |
| "learning_rate": 0.00010126911324308168, | |
| "loss": 1.0733, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.0181311018131103, | |
| "grad_norm": 0.08298351615667343, | |
| "learning_rate": 0.00010054412448273886, | |
| "loss": 1.0562, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.0227801022780103, | |
| "grad_norm": 0.09130789339542389, | |
| "learning_rate": 9.981975564242443e-05, | |
| "loss": 1.0767, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.0274291027429103, | |
| "grad_norm": 0.10151571035385132, | |
| "learning_rate": 9.909605421121028e-05, | |
| "loss": 1.1066, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.0320781032078103, | |
| "grad_norm": 0.11376982927322388, | |
| "learning_rate": 9.837306763441345e-05, | |
| "loss": 1.0909, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.0367271036727104, | |
| "grad_norm": 0.14317406713962555, | |
| "learning_rate": 9.765084331048567e-05, | |
| "loss": 1.0539, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.0413761041376104, | |
| "grad_norm": 0.183615580201149, | |
| "learning_rate": 9.692942858790591e-05, | |
| "loss": 0.9804, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.0460251046025104, | |
| "grad_norm": 0.2267308384180069, | |
| "learning_rate": 9.620887076207632e-05, | |
| "loss": 0.9, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.0506741050674104, | |
| "grad_norm": 0.25844672322273254, | |
| "learning_rate": 9.548921707222163e-05, | |
| "loss": 0.7342, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.0553231055323105, | |
| "grad_norm": 0.35192665457725525, | |
| "learning_rate": 9.477051469829196e-05, | |
| "loss": 0.7048, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.0599721059972107, | |
| "grad_norm": 0.28442806005477905, | |
| "learning_rate": 9.405281075786995e-05, | |
| "loss": 0.7852, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.0646211064621107, | |
| "grad_norm": 0.06501670181751251, | |
| "learning_rate": 9.333615230308173e-05, | |
| "loss": 1.0592, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.0692701069270107, | |
| "grad_norm": 0.06888816505670547, | |
| "learning_rate": 9.26205863175121e-05, | |
| "loss": 1.0582, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.0739191073919108, | |
| "grad_norm": 0.07344524562358856, | |
| "learning_rate": 9.190615971312446e-05, | |
| "loss": 1.0434, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.0785681078568108, | |
| "grad_norm": 0.08665701746940613, | |
| "learning_rate": 9.119291932718525e-05, | |
| "loss": 1.0843, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.0832171083217108, | |
| "grad_norm": 0.09859387576580048, | |
| "learning_rate": 9.048091191919332e-05, | |
| "loss": 1.0871, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.0878661087866108, | |
| "grad_norm": 0.11464710533618927, | |
| "learning_rate": 8.97701841678144e-05, | |
| "loss": 1.0692, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.0925151092515109, | |
| "grad_norm": 0.14108416438102722, | |
| "learning_rate": 8.906078266782087e-05, | |
| "loss": 1.0655, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.0971641097164109, | |
| "grad_norm": 0.1800997108221054, | |
| "learning_rate": 8.835275392703721e-05, | |
| "loss": 1.0205, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.1018131101813111, | |
| "grad_norm": 0.2220492660999298, | |
| "learning_rate": 8.764614436329066e-05, | |
| "loss": 0.9219, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.1064621106462111, | |
| "grad_norm": 0.2564367353916168, | |
| "learning_rate": 8.694100030136849e-05, | |
| "loss": 0.7692, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 0.2938039004802704, | |
| "learning_rate": 8.623736796998063e-05, | |
| "loss": 0.7331, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.1157601115760112, | |
| "grad_norm": 0.37189003825187683, | |
| "learning_rate": 8.553529349872916e-05, | |
| "loss": 0.5737, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.1204091120409112, | |
| "grad_norm": 0.21121616661548615, | |
| "learning_rate": 8.4834822915084e-05, | |
| "loss": 1.0279, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.1250581125058112, | |
| "grad_norm": 0.08440238237380981, | |
| "learning_rate": 8.413600214136548e-05, | |
| "loss": 1.039, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.1297071129707112, | |
| "grad_norm": 0.08833085745573044, | |
| "learning_rate": 8.343887699173356e-05, | |
| "loss": 1.065, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.1343561134356113, | |
| "grad_norm": 0.09478267282247543, | |
| "learning_rate": 8.274349316918446e-05, | |
| "loss": 1.0549, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.1390051139005113, | |
| "grad_norm": 0.1044965535402298, | |
| "learning_rate": 8.204989626255422e-05, | |
| "loss": 1.1372, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.1436541143654115, | |
| "grad_norm": 0.13615119457244873, | |
| "learning_rate": 8.135813174353008e-05, | |
| "loss": 1.1241, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.1483031148303116, | |
| "grad_norm": 0.1386057436466217, | |
| "learning_rate": 8.066824496366937e-05, | |
| "loss": 1.1088, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.1529521152952116, | |
| "grad_norm": 0.16772480309009552, | |
| "learning_rate": 7.998028115142617e-05, | |
| "loss": 1.0241, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.1576011157601116, | |
| "grad_norm": 0.2133951038122177, | |
| "learning_rate": 7.929428540918635e-05, | |
| "loss": 0.9239, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.1622501162250116, | |
| "grad_norm": 0.2830871343612671, | |
| "learning_rate": 7.86103027103106e-05, | |
| "loss": 0.8394, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.1668991166899116, | |
| "grad_norm": 0.29441362619400024, | |
| "learning_rate": 7.792837789618595e-05, | |
| "loss": 0.7265, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.1715481171548117, | |
| "grad_norm": 0.33453473448753357, | |
| "learning_rate": 7.724855567328613e-05, | |
| "loss": 0.6728, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.1761971176197117, | |
| "grad_norm": 0.3466179370880127, | |
| "learning_rate": 7.657088061024054e-05, | |
| "loss": 0.7098, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.1808461180846117, | |
| "grad_norm": 0.0637335404753685, | |
| "learning_rate": 7.58953971349124e-05, | |
| "loss": 1.0493, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.185495118549512, | |
| "grad_norm": 0.07033415138721466, | |
| "learning_rate": 7.522214953148618e-05, | |
| "loss": 1.04, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.190144119014412, | |
| "grad_norm": 0.07706312835216522, | |
| "learning_rate": 7.455118193756419e-05, | |
| "loss": 1.0632, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.194793119479312, | |
| "grad_norm": 0.09520713984966278, | |
| "learning_rate": 7.388253834127306e-05, | |
| "loss": 1.1191, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.199442119944212, | |
| "grad_norm": 0.10701565444469452, | |
| "learning_rate": 7.321626257837996e-05, | |
| "loss": 1.0896, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.204091120409112, | |
| "grad_norm": 0.11862656474113464, | |
| "learning_rate": 7.255239832941866e-05, | |
| "loss": 1.0893, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.208740120874012, | |
| "grad_norm": 0.1468852460384369, | |
| "learning_rate": 7.189098911682592e-05, | |
| "loss": 1.0755, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.213389121338912, | |
| "grad_norm": 0.18458393216133118, | |
| "learning_rate": 7.123207830208806e-05, | |
| "loss": 0.951, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.218038121803812, | |
| "grad_norm": 0.22184807062149048, | |
| "learning_rate": 7.05757090828986e-05, | |
| "loss": 0.8779, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.2226871222687121, | |
| "grad_norm": 0.3002167344093323, | |
| "learning_rate": 6.992192449032571e-05, | |
| "loss": 0.8529, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.2273361227336124, | |
| "grad_norm": 0.3212537169456482, | |
| "learning_rate": 6.927076738599152e-05, | |
| "loss": 0.6355, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.2319851231985124, | |
| "grad_norm": 0.3631160259246826, | |
| "learning_rate": 6.862228045926202e-05, | |
| "loss": 0.4947, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.2366341236634124, | |
| "grad_norm": 0.19289681315422058, | |
| "learning_rate": 6.797650622444836e-05, | |
| "loss": 0.9966, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.2412831241283124, | |
| "grad_norm": 0.07654353976249695, | |
| "learning_rate": 6.733348701801969e-05, | |
| "loss": 1.0453, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.2459321245932125, | |
| "grad_norm": 0.08304441720247269, | |
| "learning_rate": 6.669326499582755e-05, | |
| "loss": 1.0742, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.2505811250581125, | |
| "grad_norm": 0.0931682139635086, | |
| "learning_rate": 6.605588213034227e-05, | |
| "loss": 1.0973, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.2552301255230125, | |
| "grad_norm": 0.10273009538650513, | |
| "learning_rate": 6.542138020790116e-05, | |
| "loss": 1.0871, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.2598791259879125, | |
| "grad_norm": 0.1231117695569992, | |
| "learning_rate": 6.478980082596905e-05, | |
| "loss": 1.1266, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.2645281264528125, | |
| "grad_norm": 0.13809573650360107, | |
| "learning_rate": 6.416118539041135e-05, | |
| "loss": 1.0663, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.2691771269177128, | |
| "grad_norm": 0.17248062789440155, | |
| "learning_rate": 6.353557511277928e-05, | |
| "loss": 1.0427, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.2738261273826128, | |
| "grad_norm": 0.22152015566825867, | |
| "learning_rate": 6.291301100760829e-05, | |
| "loss": 0.9226, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.2784751278475128, | |
| "grad_norm": 0.2788388729095459, | |
| "learning_rate": 6.229353388972901e-05, | |
| "loss": 0.8773, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.2831241283124128, | |
| "grad_norm": 0.30847859382629395, | |
| "learning_rate": 6.167718437159147e-05, | |
| "loss": 0.7012, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.2877731287773129, | |
| "grad_norm": 0.3570963442325592, | |
| "learning_rate": 6.106400286060274e-05, | |
| "loss": 0.6076, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.292422129242213, | |
| "grad_norm": 0.3086980879306793, | |
| "learning_rate": 6.045402955647769e-05, | |
| "loss": 0.7205, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.297071129707113, | |
| "grad_norm": 0.0752180814743042, | |
| "learning_rate": 5.98473044486036e-05, | |
| "loss": 1.0282, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.301720130172013, | |
| "grad_norm": 0.07571510970592499, | |
| "learning_rate": 5.924386731341842e-05, | |
| "loss": 1.0712, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.306369130636913, | |
| "grad_norm": 0.08388427644968033, | |
| "learning_rate": 5.864375771180317e-05, | |
| "loss": 1.0425, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.3110181311018132, | |
| "grad_norm": 0.0906265452504158, | |
| "learning_rate": 5.804701498648828e-05, | |
| "loss": 1.0463, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.3156671315667132, | |
| "grad_norm": 0.10397284477949142, | |
| "learning_rate": 5.7453678259474234e-05, | |
| "loss": 1.0913, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.3203161320316132, | |
| "grad_norm": 0.12175474315881729, | |
| "learning_rate": 5.686378642946699e-05, | |
| "loss": 1.0896, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.3249651324965133, | |
| "grad_norm": 0.14317680895328522, | |
| "learning_rate": 5.627737816932754e-05, | |
| "loss": 1.0503, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.3296141329614133, | |
| "grad_norm": 0.19962027668952942, | |
| "learning_rate": 5.569449192353678e-05, | |
| "loss": 1.0168, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.3342631334263133, | |
| "grad_norm": 0.253571093082428, | |
| "learning_rate": 5.511516590567499e-05, | |
| "loss": 0.8714, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.3389121338912133, | |
| "grad_norm": 0.2986457049846649, | |
| "learning_rate": 5.453943809591654e-05, | |
| "loss": 0.7573, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.3435611343561136, | |
| "grad_norm": 0.37284067273139954, | |
| "learning_rate": 5.396734623854012e-05, | |
| "loss": 0.6708, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.3482101348210134, | |
| "grad_norm": 0.4108807444572449, | |
| "learning_rate": 5.3398927839453996e-05, | |
| "loss": 0.549, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.3528591352859136, | |
| "grad_norm": 0.21761813759803772, | |
| "learning_rate": 5.283422016373745e-05, | |
| "loss": 0.9454, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.3575081357508136, | |
| "grad_norm": 0.07017389684915543, | |
| "learning_rate": 5.227326023319743e-05, | |
| "loss": 1.06, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.3621571362157137, | |
| "grad_norm": 0.07673922181129456, | |
| "learning_rate": 5.17160848239416e-05, | |
| "loss": 1.061, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.3668061366806137, | |
| "grad_norm": 0.08843539655208588, | |
| "learning_rate": 5.1162730463967304e-05, | |
| "loss": 1.0668, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.3714551371455137, | |
| "grad_norm": 0.10018595308065414, | |
| "learning_rate": 5.061323343076672e-05, | |
| "loss": 1.0926, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.3761041376104137, | |
| "grad_norm": 0.11475757509469986, | |
| "learning_rate": 5.006762974894872e-05, | |
| "loss": 1.0878, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.3807531380753137, | |
| "grad_norm": 0.14695028960704803, | |
| "learning_rate": 4.9525955187876885e-05, | |
| "loss": 1.088, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.385402138540214, | |
| "grad_norm": 0.18600180745124817, | |
| "learning_rate": 4.898824525932471e-05, | |
| "loss": 1.0019, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.3900511390051138, | |
| "grad_norm": 0.21412943303585052, | |
| "learning_rate": 4.845453521514738e-05, | |
| "loss": 0.9161, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.394700139470014, | |
| "grad_norm": 0.2769903838634491, | |
| "learning_rate": 4.7924860044970615e-05, | |
| "loss": 0.8767, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.399349139934914, | |
| "grad_norm": 0.37584012746810913, | |
| "learning_rate": 4.739925447389698e-05, | |
| "loss": 0.8159, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.403998140399814, | |
| "grad_norm": 0.35033103823661804, | |
| "learning_rate": 4.687775296022908e-05, | |
| "loss": 0.5912, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.408647140864714, | |
| "grad_norm": 0.3620702624320984, | |
| "learning_rate": 4.6360389693210735e-05, | |
| "loss": 0.7819, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.4132961413296141, | |
| "grad_norm": 0.06753652542829514, | |
| "learning_rate": 4.5847198590785394e-05, | |
| "loss": 1.023, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.4179451417945141, | |
| "grad_norm": 0.07456682622432709, | |
| "learning_rate": 4.5338213297372534e-05, | |
| "loss": 1.0321, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.4225941422594142, | |
| "grad_norm": 0.08220840245485306, | |
| "learning_rate": 4.4833467181662086e-05, | |
| "loss": 1.0518, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.4272431427243144, | |
| "grad_norm": 0.09785965085029602, | |
| "learning_rate": 4.4332993334426576e-05, | |
| "loss": 1.0736, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.4318921431892142, | |
| "grad_norm": 0.10714032500982285, | |
| "learning_rate": 4.383682456635199e-05, | |
| "loss": 1.0766, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.4365411436541144, | |
| "grad_norm": 0.12374605983495712, | |
| "learning_rate": 4.3344993405886425e-05, | |
| "loss": 1.1013, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.4411901441190145, | |
| "grad_norm": 0.15557971596717834, | |
| "learning_rate": 4.285753209710786e-05, | |
| "loss": 1.0673, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.4458391445839145, | |
| "grad_norm": 0.1834760308265686, | |
| "learning_rate": 4.2374472597610044e-05, | |
| "loss": 0.9611, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.4504881450488145, | |
| "grad_norm": 0.24258282780647278, | |
| "learning_rate": 4.1895846576407424e-05, | |
| "loss": 0.9162, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.4551371455137145, | |
| "grad_norm": 0.31023555994033813, | |
| "learning_rate": 4.1421685411859046e-05, | |
| "loss": 0.8321, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.4597861459786146, | |
| "grad_norm": 0.3393552601337433, | |
| "learning_rate": 4.095202018961125e-05, | |
| "loss": 0.6979, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.4644351464435146, | |
| "grad_norm": 0.3873143792152405, | |
| "learning_rate": 4.048688170055989e-05, | |
| "loss": 0.5101, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.4690841469084148, | |
| "grad_norm": 0.20240376889705658, | |
| "learning_rate": 4.002630043883159e-05, | |
| "loss": 1.0254, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.4737331473733146, | |
| "grad_norm": 0.08021257072687149, | |
| "learning_rate": 3.9570306599784544e-05, | |
| "loss": 1.033, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.4783821478382149, | |
| "grad_norm": 0.079825259745121, | |
| "learning_rate": 3.911893007802913e-05, | |
| "loss": 1.0105, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.4830311483031149, | |
| "grad_norm": 0.09591138362884521, | |
| "learning_rate": 3.8672200465467765e-05, | |
| "loss": 1.0867, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.487680148768015, | |
| "grad_norm": 0.10192592442035675, | |
| "learning_rate": 3.8230147049355147e-05, | |
| "loss": 1.1238, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.492329149232915, | |
| "grad_norm": 0.11998689919710159, | |
| "learning_rate": 3.779279881037797e-05, | |
| "loss": 1.099, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.496978149697815, | |
| "grad_norm": 0.14268672466278076, | |
| "learning_rate": 3.7360184420755165e-05, | |
| "loss": 1.0832, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.501627150162715, | |
| "grad_norm": 0.1779240220785141, | |
| "learning_rate": 3.693233224235806e-05, | |
| "loss": 1.0322, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.506276150627615, | |
| "grad_norm": 0.23235994577407837, | |
| "learning_rate": 3.650927032485101e-05, | |
| "loss": 0.987, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.5109251510925152, | |
| "grad_norm": 0.2786919176578522, | |
| "learning_rate": 3.609102640385254e-05, | |
| "loss": 0.7974, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.515574151557415, | |
| "grad_norm": 0.3645856976509094, | |
| "learning_rate": 3.567762789911693e-05, | |
| "loss": 0.7208, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.5202231520223153, | |
| "grad_norm": 0.4197288751602173, | |
| "learning_rate": 3.526910191273665e-05, | |
| "loss": 0.5941, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.524872152487215, | |
| "grad_norm": 0.43531715869903564, | |
| "learning_rate": 3.486547522736562e-05, | |
| "loss": 0.8169, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.5295211529521153, | |
| "grad_norm": 0.07292664051055908, | |
| "learning_rate": 3.44667743044632e-05, | |
| "loss": 1.026, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.5341701534170153, | |
| "grad_norm": 0.07665824145078659, | |
| "learning_rate": 3.407302528255961e-05, | |
| "loss": 1.0711, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.5388191538819154, | |
| "grad_norm": 0.09035459905862808, | |
| "learning_rate": 3.36842539755421e-05, | |
| "loss": 1.058, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.5434681543468154, | |
| "grad_norm": 0.10302021354436874, | |
| "learning_rate": 3.3300485870962776e-05, | |
| "loss": 1.0717, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.5481171548117154, | |
| "grad_norm": 0.11944916844367981, | |
| "learning_rate": 3.292174612836757e-05, | |
| "loss": 1.1738, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.5527661552766157, | |
| "grad_norm": 0.1299082338809967, | |
| "learning_rate": 3.254805957764673e-05, | |
| "loss": 1.1078, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.5574151557415155, | |
| "grad_norm": 0.15316098928451538, | |
| "learning_rate": 3.217945071740724e-05, | |
| "loss": 1.0111, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.5620641562064157, | |
| "grad_norm": 0.19828684628009796, | |
| "learning_rate": 3.1815943713366404e-05, | |
| "loss": 1.0292, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.5667131566713157, | |
| "grad_norm": 0.2431751936674118, | |
| "learning_rate": 3.145756239676779e-05, | |
| "loss": 0.867, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.5713621571362157, | |
| "grad_norm": 0.3115411698818207, | |
| "learning_rate": 3.110433026281872e-05, | |
| "loss": 0.7307, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.5760111576011158, | |
| "grad_norm": 0.37397709488868713, | |
| "learning_rate": 3.075627046915003e-05, | |
| "loss": 0.6894, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.5806601580660158, | |
| "grad_norm": 0.4691689610481262, | |
| "learning_rate": 3.041340583429789e-05, | |
| "loss": 0.6568, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.5853091585309158, | |
| "grad_norm": 0.27498671412467957, | |
| "learning_rate": 3.0075758836207716e-05, | |
| "loss": 1.0342, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.5899581589958158, | |
| "grad_norm": 0.07807187736034393, | |
| "learning_rate": 2.9743351610760716e-05, | |
| "loss": 1.0501, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.594607159460716, | |
| "grad_norm": 0.08518065512180328, | |
| "learning_rate": 2.941620595032246e-05, | |
| "loss": 1.0657, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.5992561599256159, | |
| "grad_norm": 0.09096917510032654, | |
| "learning_rate": 2.9094343302314432e-05, | |
| "loss": 1.0534, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.6039051603905161, | |
| "grad_norm": 0.10106653720140457, | |
| "learning_rate": 2.8777784767807727e-05, | |
| "loss": 1.0949, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.6085541608554161, | |
| "grad_norm": 0.11770515143871307, | |
| "learning_rate": 2.846655110013978e-05, | |
| "loss": 1.1013, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.6132031613203162, | |
| "grad_norm": 0.14044031500816345, | |
| "learning_rate": 2.816066270355391e-05, | |
| "loss": 1.0657, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.6178521617852162, | |
| "grad_norm": 0.1705469787120819, | |
| "learning_rate": 2.78601396318614e-05, | |
| "loss": 1.0108, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.6225011622501162, | |
| "grad_norm": 0.23262286186218262, | |
| "learning_rate": 2.7565001587126922e-05, | |
| "loss": 0.9104, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.6271501627150162, | |
| "grad_norm": 0.3084715008735657, | |
| "learning_rate": 2.7275267918376912e-05, | |
| "loss": 0.8493, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.6317991631799162, | |
| "grad_norm": 0.34869176149368286, | |
| "learning_rate": 2.6990957620330954e-05, | |
| "loss": 0.7103, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.6364481636448165, | |
| "grad_norm": 0.40754231810569763, | |
| "learning_rate": 2.6712089332156633e-05, | |
| "loss": 0.606, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.6410971641097163, | |
| "grad_norm": 0.38754644989967346, | |
| "learning_rate": 2.6438681336247417e-05, | |
| "loss": 0.7465, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.6457461645746165, | |
| "grad_norm": 0.07341364026069641, | |
| "learning_rate": 2.6170751557024197e-05, | |
| "loss": 1.0565, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.6503951650395166, | |
| "grad_norm": 0.08013252913951874, | |
| "learning_rate": 2.5908317559760138e-05, | |
| "loss": 1.0146, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.6550441655044166, | |
| "grad_norm": 0.09281094372272491, | |
| "learning_rate": 2.5651396549429086e-05, | |
| "loss": 1.0903, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.6596931659693166, | |
| "grad_norm": 0.09770162403583527, | |
| "learning_rate": 2.540000536957765e-05, | |
| "loss": 1.1007, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.6643421664342166, | |
| "grad_norm": 0.11755497008562088, | |
| "learning_rate": 2.515416050122092e-05, | |
| "loss": 1.1027, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.6689911668991166, | |
| "grad_norm": 0.1314418464899063, | |
| "learning_rate": 2.4913878061762094e-05, | |
| "loss": 1.0532, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.6736401673640167, | |
| "grad_norm": 0.1723410040140152, | |
| "learning_rate": 2.4679173803935662e-05, | |
| "loss": 1.0447, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.678289167828917, | |
| "grad_norm": 0.22374863922595978, | |
| "learning_rate": 2.4450063114774784e-05, | |
| "loss": 1.0271, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.6829381682938167, | |
| "grad_norm": 0.2649782598018646, | |
| "learning_rate": 2.4226561014602522e-05, | |
| "loss": 0.8288, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.687587168758717, | |
| "grad_norm": 0.3294927477836609, | |
| "learning_rate": 2.400868215604706e-05, | |
| "loss": 0.7618, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.692236169223617, | |
| "grad_norm": 0.390989750623703, | |
| "learning_rate": 2.3796440823081167e-05, | |
| "loss": 0.6498, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.696885169688517, | |
| "grad_norm": 0.47515323758125305, | |
| "learning_rate": 2.358985093008566e-05, | |
| "loss": 0.6206, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.701534170153417, | |
| "grad_norm": 0.18342038989067078, | |
| "learning_rate": 2.3388926020937286e-05, | |
| "loss": 0.9768, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.706183170618317, | |
| "grad_norm": 0.08224090933799744, | |
| "learning_rate": 2.3193679268120718e-05, | |
| "loss": 1.0393, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.710832171083217, | |
| "grad_norm": 0.08616173267364502, | |
| "learning_rate": 2.3004123471865e-05, | |
| "loss": 1.0021, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.715481171548117, | |
| "grad_norm": 0.08873171359300613, | |
| "learning_rate": 2.2820271059304412e-05, | |
| "loss": 1.066, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.7201301720130173, | |
| "grad_norm": 0.10707499831914902, | |
| "learning_rate": 2.2642134083663678e-05, | |
| "loss": 1.1172, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.7247791724779171, | |
| "grad_norm": 0.1225440576672554, | |
| "learning_rate": 2.2469724223467866e-05, | |
| "loss": 1.0865, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.7294281729428174, | |
| "grad_norm": 0.1396765112876892, | |
| "learning_rate": 2.2303052781776664e-05, | |
| "loss": 1.0273, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.7340771734077174, | |
| "grad_norm": 0.17938072979450226, | |
| "learning_rate": 2.2142130685443382e-05, | |
| "loss": 1.1046, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.7387261738726174, | |
| "grad_norm": 0.2324485331773758, | |
| "learning_rate": 2.198696848439865e-05, | |
| "loss": 0.9628, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.7433751743375174, | |
| "grad_norm": 0.2933482527732849, | |
| "learning_rate": 2.1837576350958686e-05, | |
| "loss": 0.8228, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.7480241748024175, | |
| "grad_norm": 0.3423949182033539, | |
| "learning_rate": 2.169396407915849e-05, | |
| "loss": 0.7043, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.7526731752673175, | |
| "grad_norm": 0.38756224513053894, | |
| "learning_rate": 2.155614108410968e-05, | |
| "loss": 0.6579, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.7573221757322175, | |
| "grad_norm": 0.4097736179828644, | |
| "learning_rate": 2.142411640138332e-05, | |
| "loss": 0.7569, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.7619711761971177, | |
| "grad_norm": 0.07299873232841492, | |
| "learning_rate": 2.129789868641749e-05, | |
| "loss": 1.0571, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.7666201766620175, | |
| "grad_norm": 0.07895659655332565, | |
| "learning_rate": 2.1177496213949837e-05, | |
| "loss": 1.0487, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.7712691771269178, | |
| "grad_norm": 0.09031179547309875, | |
| "learning_rate": 2.1062916877475198e-05, | |
| "loss": 1.0911, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.7759181775918178, | |
| "grad_norm": 0.09883815795183182, | |
| "learning_rate": 2.0954168188727962e-05, | |
| "loss": 1.0619, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.7805671780567178, | |
| "grad_norm": 0.11488756537437439, | |
| "learning_rate": 2.0851257277189703e-05, | |
| "loss": 1.0788, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.7852161785216178, | |
| "grad_norm": 0.14409998059272766, | |
| "learning_rate": 2.0754190889621745e-05, | |
| "loss": 1.0902, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.7898651789865179, | |
| "grad_norm": 0.1650954782962799, | |
| "learning_rate": 2.0662975389622843e-05, | |
| "loss": 0.999, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.794514179451418, | |
| "grad_norm": 0.22679570317268372, | |
| "learning_rate": 2.0577616757212016e-05, | |
| "loss": 0.9647, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.799163179916318, | |
| "grad_norm": 0.284229040145874, | |
| "learning_rate": 2.0498120588436466e-05, | |
| "loss": 0.85, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.8038121803812182, | |
| "grad_norm": 0.32825493812561035, | |
| "learning_rate": 2.0424492095004746e-05, | |
| "loss": 0.7025, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.808461180846118, | |
| "grad_norm": 0.37925466895103455, | |
| "learning_rate": 2.0356736103945047e-05, | |
| "loss": 0.6595, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.8131101813110182, | |
| "grad_norm": 0.45006221532821655, | |
| "learning_rate": 2.029485705728876e-05, | |
| "loss": 0.4914, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.8177591817759182, | |
| "grad_norm": 0.2308957725763321, | |
| "learning_rate": 2.023885901177926e-05, | |
| "loss": 1.0114, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.8224081822408182, | |
| "grad_norm": 0.07487577944993973, | |
| "learning_rate": 2.0188745638605954e-05, | |
| "loss": 1.0175, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.8270571827057183, | |
| "grad_norm": 0.07909037917852402, | |
| "learning_rate": 2.014452022316358e-05, | |
| "loss": 1.0365, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.8317061831706183, | |
| "grad_norm": 0.09483642131090164, | |
| "learning_rate": 2.010618566483684e-05, | |
| "loss": 1.0723, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.8363551836355185, | |
| "grad_norm": 0.1127537190914154, | |
| "learning_rate": 2.00737444768103e-05, | |
| "loss": 1.0905, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.8410041841004183, | |
| "grad_norm": 0.12364522367715836, | |
| "learning_rate": 2.0047198785903658e-05, | |
| "loss": 1.1045, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.8456531845653186, | |
| "grad_norm": 0.14588101208209991, | |
| "learning_rate": 2.002655033243228e-05, | |
| "loss": 1.0753, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.8503021850302184, | |
| "grad_norm": 0.196581169962883, | |
| "learning_rate": 2.0011800470093105e-05, | |
| "loss": 1.0165, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.8549511854951186, | |
| "grad_norm": 0.24468840658664703, | |
| "learning_rate": 2.0002950165875934e-05, | |
| "loss": 0.9067, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.8596001859600186, | |
| "grad_norm": 0.32478219270706177, | |
| "learning_rate": 2e-05, | |
| "loss": 0.8752, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.3538115420584673e+18, | |
| "train_batch_size": 6, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |