| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 2000, | |
| "global_step": 80, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "crossentropy": 2.487246513366699, | |
| "epoch": 0.125, | |
| "grad_norm": 0.03208793327212334, | |
| "learning_rate": 0.001, | |
| "loss": 2.4872, | |
| "step": 1 | |
| }, | |
| { | |
| "crossentropy": 2.379896402359009, | |
| "epoch": 0.25, | |
| "grad_norm": 0.03239322826266289, | |
| "learning_rate": 0.002, | |
| "loss": 2.3799, | |
| "step": 2 | |
| }, | |
| { | |
| "crossentropy": 2.5010085105895996, | |
| "epoch": 0.375, | |
| "grad_norm": 0.03320762887597084, | |
| "learning_rate": 0.003, | |
| "loss": 2.501, | |
| "step": 3 | |
| }, | |
| { | |
| "crossentropy": 2.688127040863037, | |
| "epoch": 0.5, | |
| "grad_norm": 0.03224330395460129, | |
| "learning_rate": 0.004, | |
| "loss": 2.6881, | |
| "step": 4 | |
| }, | |
| { | |
| "crossentropy": 2.4009199142456055, | |
| "epoch": 0.625, | |
| "grad_norm": 0.029966533184051514, | |
| "learning_rate": 0.005, | |
| "loss": 2.4009, | |
| "step": 5 | |
| }, | |
| { | |
| "crossentropy": 2.474385976791382, | |
| "epoch": 0.75, | |
| "grad_norm": 0.030283687636256218, | |
| "learning_rate": 0.006, | |
| "loss": 2.4744, | |
| "step": 6 | |
| }, | |
| { | |
| "crossentropy": 2.5291025638580322, | |
| "epoch": 0.875, | |
| "grad_norm": 0.03127186745405197, | |
| "learning_rate": 0.006999999999999999, | |
| "loss": 2.5291, | |
| "step": 7 | |
| }, | |
| { | |
| "crossentropy": 2.3309171199798584, | |
| "epoch": 1.0, | |
| "grad_norm": 0.030661335214972496, | |
| "learning_rate": 0.008, | |
| "loss": 2.3309, | |
| "step": 8 | |
| }, | |
| { | |
| "crossentropy": 2.4654123783111572, | |
| "epoch": 1.125, | |
| "grad_norm": 0.029608242213726044, | |
| "learning_rate": 0.009000000000000001, | |
| "loss": 2.4654, | |
| "step": 9 | |
| }, | |
| { | |
| "crossentropy": 2.468688488006592, | |
| "epoch": 1.25, | |
| "grad_norm": 0.03335599973797798, | |
| "learning_rate": 0.01, | |
| "loss": 2.4687, | |
| "step": 10 | |
| }, | |
| { | |
| "crossentropy": 2.3773512840270996, | |
| "epoch": 1.375, | |
| "grad_norm": 0.031801655888557434, | |
| "learning_rate": 0.009994965332706574, | |
| "loss": 2.3774, | |
| "step": 11 | |
| }, | |
| { | |
| "crossentropy": 2.5056920051574707, | |
| "epoch": 1.5, | |
| "grad_norm": 0.03860826417803764, | |
| "learning_rate": 0.009979871469976196, | |
| "loss": 2.5057, | |
| "step": 12 | |
| }, | |
| { | |
| "crossentropy": 2.4569733142852783, | |
| "epoch": 1.625, | |
| "grad_norm": 0.03568004071712494, | |
| "learning_rate": 0.009954748808839673, | |
| "loss": 2.457, | |
| "step": 13 | |
| }, | |
| { | |
| "crossentropy": 2.4127414226531982, | |
| "epoch": 1.75, | |
| "grad_norm": 0.03557576611638069, | |
| "learning_rate": 0.009919647942993149, | |
| "loss": 2.4127, | |
| "step": 14 | |
| }, | |
| { | |
| "crossentropy": 2.4089269638061523, | |
| "epoch": 1.875, | |
| "grad_norm": 0.0321757011115551, | |
| "grad_norm_var": 5.793923908080815e-06, | |
| "learning_rate": 0.009874639560909117, | |
| "loss": 2.4089, | |
| "step": 15 | |
| }, | |
| { | |
| "crossentropy": 2.385178804397583, | |
| "epoch": 2.0, | |
| "grad_norm": 0.033555347472429276, | |
| "grad_norm_var": 5.852987523136916e-06, | |
| "learning_rate": 0.009819814303479266, | |
| "loss": 2.3852, | |
| "step": 16 | |
| }, | |
| { | |
| "crossentropy": 2.4342517852783203, | |
| "epoch": 2.125, | |
| "grad_norm": 0.034557584673166275, | |
| "grad_norm_var": 6.127065953599033e-06, | |
| "learning_rate": 0.009755282581475769, | |
| "loss": 2.4343, | |
| "step": 17 | |
| }, | |
| { | |
| "crossentropy": 2.4237754344940186, | |
| "epoch": 2.25, | |
| "grad_norm": 0.03371824324131012, | |
| "grad_norm_var": 6.168768579761655e-06, | |
| "learning_rate": 0.009681174353198686, | |
| "loss": 2.4238, | |
| "step": 18 | |
| }, | |
| { | |
| "crossentropy": 2.4231510162353516, | |
| "epoch": 2.375, | |
| "grad_norm": 0.02907504141330719, | |
| "grad_norm_var": 6.557002911025779e-06, | |
| "learning_rate": 0.009597638862757255, | |
| "loss": 2.4232, | |
| "step": 19 | |
| }, | |
| { | |
| "crossentropy": 2.3850157260894775, | |
| "epoch": 2.5, | |
| "grad_norm": 0.03278205543756485, | |
| "grad_norm_var": 6.122522196004565e-06, | |
| "learning_rate": 0.009504844339512096, | |
| "loss": 2.385, | |
| "step": 20 | |
| }, | |
| { | |
| "crossentropy": 2.3287932872772217, | |
| "epoch": 2.625, | |
| "grad_norm": 0.03301858901977539, | |
| "grad_norm_var": 5.930476905840337e-06, | |
| "learning_rate": 0.00940297765928369, | |
| "loss": 2.3288, | |
| "step": 21 | |
| }, | |
| { | |
| "crossentropy": 2.3058865070343018, | |
| "epoch": 2.75, | |
| "grad_norm": 0.03293571248650551, | |
| "grad_norm_var": 5.5371734660507875e-06, | |
| "learning_rate": 0.009292243968009331, | |
| "loss": 2.3059, | |
| "step": 22 | |
| }, | |
| { | |
| "crossentropy": 2.2683637142181396, | |
| "epoch": 2.875, | |
| "grad_norm": 0.03541896864771843, | |
| "grad_norm_var": 5.362674881745298e-06, | |
| "learning_rate": 0.009172866268606514, | |
| "loss": 2.2684, | |
| "step": 23 | |
| }, | |
| { | |
| "crossentropy": 2.411990165710449, | |
| "epoch": 3.0, | |
| "grad_norm": 0.03502194955945015, | |
| "grad_norm_var": 4.411311488799798e-06, | |
| "learning_rate": 0.009045084971874737, | |
| "loss": 2.412, | |
| "step": 24 | |
| }, | |
| { | |
| "crossentropy": 2.2515363693237305, | |
| "epoch": 3.125, | |
| "grad_norm": 0.03379856050014496, | |
| "grad_norm_var": 4.1744788750110175e-06, | |
| "learning_rate": 0.008909157412340149, | |
| "loss": 2.2515, | |
| "step": 25 | |
| }, | |
| { | |
| "crossentropy": 2.3486626148223877, | |
| "epoch": 3.25, | |
| "grad_norm": 0.034475792199373245, | |
| "grad_norm_var": 2.7198637048790144e-06, | |
| "learning_rate": 0.008765357330018056, | |
| "loss": 2.3487, | |
| "step": 26 | |
| }, | |
| { | |
| "crossentropy": 2.379669666290283, | |
| "epoch": 3.375, | |
| "grad_norm": 0.03319563344120979, | |
| "grad_norm_var": 2.4738877727155267e-06, | |
| "learning_rate": 0.008613974319136958, | |
| "loss": 2.3797, | |
| "step": 27 | |
| }, | |
| { | |
| "crossentropy": 2.3640384674072266, | |
| "epoch": 3.5, | |
| "grad_norm": 0.03426358476281166, | |
| "grad_norm_var": 2.2389126654258237e-06, | |
| "learning_rate": 0.008455313244934324, | |
| "loss": 2.364, | |
| "step": 28 | |
| }, | |
| { | |
| "crossentropy": 2.324143409729004, | |
| "epoch": 3.625, | |
| "grad_norm": 0.0358288437128067, | |
| "grad_norm_var": 2.410602035355508e-06, | |
| "learning_rate": 0.008289693629698563, | |
| "loss": 2.3241, | |
| "step": 29 | |
| }, | |
| { | |
| "crossentropy": 2.159423828125, | |
| "epoch": 3.75, | |
| "grad_norm": 0.031966786831617355, | |
| "grad_norm_var": 2.612506091834166e-06, | |
| "learning_rate": 0.008117449009293669, | |
| "loss": 2.1594, | |
| "step": 30 | |
| }, | |
| { | |
| "crossentropy": 2.2669661045074463, | |
| "epoch": 3.875, | |
| "grad_norm": 0.0339609794318676, | |
| "grad_norm_var": 2.61687730451927e-06, | |
| "learning_rate": 0.007938926261462366, | |
| "loss": 2.267, | |
| "step": 31 | |
| }, | |
| { | |
| "crossentropy": 2.482945680618286, | |
| "epoch": 4.0, | |
| "grad_norm": 0.03357086703181267, | |
| "grad_norm_var": 2.5635888162410063e-06, | |
| "learning_rate": 0.007754484907260513, | |
| "loss": 2.4829, | |
| "step": 32 | |
| }, | |
| { | |
| "crossentropy": 2.1757168769836426, | |
| "epoch": 4.125, | |
| "grad_norm": 0.032582979649305344, | |
| "grad_norm_var": 1.2068945133852464e-06, | |
| "learning_rate": 0.007564496387029531, | |
| "loss": 2.1757, | |
| "step": 33 | |
| }, | |
| { | |
| "crossentropy": 2.4300379753112793, | |
| "epoch": 4.25, | |
| "grad_norm": 0.03353552147746086, | |
| "grad_norm_var": 1.1362555840309259e-06, | |
| "learning_rate": 0.007369343312364994, | |
| "loss": 2.43, | |
| "step": 34 | |
| }, | |
| { | |
| "crossentropy": 2.1233766078948975, | |
| "epoch": 4.375, | |
| "grad_norm": 0.034830041229724884, | |
| "grad_norm_var": 1.1319644129707703e-06, | |
| "learning_rate": 0.007169418695587791, | |
| "loss": 2.1234, | |
| "step": 35 | |
| }, | |
| { | |
| "crossentropy": 2.2619526386260986, | |
| "epoch": 4.5, | |
| "grad_norm": 0.033126723021268845, | |
| "grad_norm_var": 1.1071727437842251e-06, | |
| "learning_rate": 0.006965125158269619, | |
| "loss": 2.262, | |
| "step": 36 | |
| }, | |
| { | |
| "crossentropy": 2.3065433502197266, | |
| "epoch": 4.625, | |
| "grad_norm": 0.0348593033850193, | |
| "grad_norm_var": 1.0216560744425243e-06, | |
| "learning_rate": 0.0067568741204067145, | |
| "loss": 2.3065, | |
| "step": 37 | |
| }, | |
| { | |
| "crossentropy": 2.2902674674987793, | |
| "epoch": 4.75, | |
| "grad_norm": 0.0360955074429512, | |
| "grad_norm_var": 1.2434575549111667e-06, | |
| "learning_rate": 0.006545084971874737, | |
| "loss": 2.2903, | |
| "step": 38 | |
| }, | |
| { | |
| "crossentropy": 2.2587778568267822, | |
| "epoch": 4.875, | |
| "grad_norm": 0.03467360511422157, | |
| "grad_norm_var": 1.2055615432713293e-06, | |
| "learning_rate": 0.006330184227833375, | |
| "loss": 2.2588, | |
| "step": 39 | |
| }, | |
| { | |
| "crossentropy": 2.3978285789489746, | |
| "epoch": 5.0, | |
| "grad_norm": 0.035668086260557175, | |
| "grad_norm_var": 1.3685657271088221e-06, | |
| "learning_rate": 0.006112604669781572, | |
| "loss": 2.3978, | |
| "step": 40 | |
| }, | |
| { | |
| "crossentropy": 2.3564870357513428, | |
| "epoch": 5.125, | |
| "grad_norm": 0.0373300276696682, | |
| "grad_norm_var": 2.0190065310690096e-06, | |
| "learning_rate": 0.005892784473993183, | |
| "loss": 2.3565, | |
| "step": 41 | |
| }, | |
| { | |
| "crossentropy": 2.202275514602661, | |
| "epoch": 5.25, | |
| "grad_norm": 0.03768543526530266, | |
| "grad_norm_var": 2.6563097811015932e-06, | |
| "learning_rate": 0.0056711663290882775, | |
| "loss": 2.2023, | |
| "step": 42 | |
| }, | |
| { | |
| "crossentropy": 2.192146062850952, | |
| "epoch": 5.375, | |
| "grad_norm": 0.03134535253047943, | |
| "grad_norm_var": 3.228640330968311e-06, | |
| "learning_rate": 0.005448196544517168, | |
| "loss": 2.1921, | |
| "step": 43 | |
| }, | |
| { | |
| "crossentropy": 2.2519092559814453, | |
| "epoch": 5.5, | |
| "grad_norm": 0.032612044364213943, | |
| "grad_norm_var": 3.044945465432008e-06, | |
| "learning_rate": 0.005224324151752576, | |
| "loss": 2.2519, | |
| "step": 44 | |
| }, | |
| { | |
| "crossentropy": 2.3610141277313232, | |
| "epoch": 5.625, | |
| "grad_norm": 0.03532750904560089, | |
| "grad_norm_var": 3.073519780112897e-06, | |
| "learning_rate": 0.005, | |
| "loss": 2.361, | |
| "step": 45 | |
| }, | |
| { | |
| "crossentropy": 2.2574493885040283, | |
| "epoch": 5.75, | |
| "grad_norm": 0.03538508713245392, | |
| "grad_norm_var": 3.047191916038562e-06, | |
| "learning_rate": 0.004775675848247427, | |
| "loss": 2.2574, | |
| "step": 46 | |
| }, | |
| { | |
| "crossentropy": 2.170254707336426, | |
| "epoch": 5.875, | |
| "grad_norm": 0.03475534915924072, | |
| "grad_norm_var": 2.9654755954788835e-06, | |
| "learning_rate": 0.004551803455482833, | |
| "loss": 2.1703, | |
| "step": 47 | |
| }, | |
| { | |
| "crossentropy": 2.2004218101501465, | |
| "epoch": 6.0, | |
| "grad_norm": 0.0375908762216568, | |
| "grad_norm_var": 3.107626395358089e-06, | |
| "learning_rate": 0.004328833670911724, | |
| "loss": 2.2004, | |
| "step": 48 | |
| }, | |
| { | |
| "crossentropy": 2.162071943283081, | |
| "epoch": 6.125, | |
| "grad_norm": 0.031826432794332504, | |
| "grad_norm_var": 4.072774386597716e-06, | |
| "learning_rate": 0.004107215526006817, | |
| "loss": 2.1621, | |
| "step": 49 | |
| }, | |
| { | |
| "crossentropy": 2.144589424133301, | |
| "epoch": 6.25, | |
| "grad_norm": 0.03268062323331833, | |
| "grad_norm_var": 4.202360731680779e-06, | |
| "learning_rate": 0.003887395330218428, | |
| "loss": 2.1446, | |
| "step": 50 | |
| }, | |
| { | |
| "crossentropy": 2.202075481414795, | |
| "epoch": 6.375, | |
| "grad_norm": 0.03384561836719513, | |
| "grad_norm_var": 4.294842472950198e-06, | |
| "learning_rate": 0.003669815772166625, | |
| "loss": 2.2021, | |
| "step": 51 | |
| }, | |
| { | |
| "crossentropy": 2.3169946670532227, | |
| "epoch": 6.5, | |
| "grad_norm": 0.03661832585930824, | |
| "grad_norm_var": 4.387942230220839e-06, | |
| "learning_rate": 0.003454915028125263, | |
| "loss": 2.317, | |
| "step": 52 | |
| }, | |
| { | |
| "crossentropy": 2.191814422607422, | |
| "epoch": 6.625, | |
| "grad_norm": 0.03522748872637749, | |
| "grad_norm_var": 4.380226970383122e-06, | |
| "learning_rate": 0.003243125879593286, | |
| "loss": 2.1918, | |
| "step": 53 | |
| }, | |
| { | |
| "crossentropy": 2.196298360824585, | |
| "epoch": 6.75, | |
| "grad_norm": 0.03524046018719673, | |
| "grad_norm_var": 4.3576876356847075e-06, | |
| "learning_rate": 0.0030348748417303823, | |
| "loss": 2.1963, | |
| "step": 54 | |
| }, | |
| { | |
| "crossentropy": 2.2684872150421143, | |
| "epoch": 6.875, | |
| "grad_norm": 0.03640512377023697, | |
| "grad_norm_var": 4.452811959714381e-06, | |
| "learning_rate": 0.00283058130441221, | |
| "loss": 2.2685, | |
| "step": 55 | |
| }, | |
| { | |
| "crossentropy": 2.3180689811706543, | |
| "epoch": 7.0, | |
| "grad_norm": 0.03479791432619095, | |
| "grad_norm_var": 4.097831569832335e-06, | |
| "learning_rate": 0.002630656687635007, | |
| "loss": 2.3181, | |
| "step": 56 | |
| }, | |
| { | |
| "crossentropy": 2.201280117034912, | |
| "epoch": 7.125, | |
| "grad_norm": 0.03417252376675606, | |
| "grad_norm_var": 2.7744502044233416e-06, | |
| "learning_rate": 0.00243550361297047, | |
| "loss": 2.2013, | |
| "step": 57 | |
| }, | |
| { | |
| "crossentropy": 2.3339715003967285, | |
| "epoch": 7.25, | |
| "grad_norm": 0.033313509076833725, | |
| "grad_norm_var": 2.5884423837709467e-06, | |
| "learning_rate": 0.002245515092739488, | |
| "loss": 2.334, | |
| "step": 58 | |
| }, | |
| { | |
| "crossentropy": 2.1603586673736572, | |
| "epoch": 7.375, | |
| "grad_norm": 0.03365428000688553, | |
| "grad_norm_var": 2.684439790351147e-06, | |
| "learning_rate": 0.0020610737385376348, | |
| "loss": 2.1604, | |
| "step": 59 | |
| }, | |
| { | |
| "crossentropy": 2.2999627590179443, | |
| "epoch": 7.5, | |
| "grad_norm": 0.034641556441783905, | |
| "grad_norm_var": 2.6678187146791006e-06, | |
| "learning_rate": 0.0018825509907063327, | |
| "loss": 2.3, | |
| "step": 60 | |
| }, | |
| { | |
| "crossentropy": 2.1747264862060547, | |
| "epoch": 7.625, | |
| "grad_norm": 0.03396380692720413, | |
| "grad_norm_var": 2.7140570016346197e-06, | |
| "learning_rate": 0.001710306370301437, | |
| "loss": 2.1747, | |
| "step": 61 | |
| }, | |
| { | |
| "crossentropy": 2.120774269104004, | |
| "epoch": 7.75, | |
| "grad_norm": 0.03624221310019493, | |
| "grad_norm_var": 2.3210148057607623e-06, | |
| "learning_rate": 0.0015446867550656768, | |
| "loss": 2.1208, | |
| "step": 62 | |
| }, | |
| { | |
| "crossentropy": 2.2385036945343018, | |
| "epoch": 7.875, | |
| "grad_norm": 0.034913428127765656, | |
| "grad_norm_var": 1.732991609125894e-06, | |
| "learning_rate": 0.0013860256808630427, | |
| "loss": 2.2385, | |
| "step": 63 | |
| }, | |
| { | |
| "crossentropy": 2.1571521759033203, | |
| "epoch": 8.0, | |
| "grad_norm": 0.03277244418859482, | |
| "grad_norm_var": 1.4490052253732697e-06, | |
| "learning_rate": 0.0012346426699819458, | |
| "loss": 2.1572, | |
| "step": 64 | |
| }, | |
| { | |
| "crossentropy": 2.2847859859466553, | |
| "epoch": 8.125, | |
| "grad_norm": 0.033221788704395294, | |
| "grad_norm_var": 1.512194472239048e-06, | |
| "learning_rate": 0.001090842587659851, | |
| "loss": 2.2848, | |
| "step": 65 | |
| }, | |
| { | |
| "crossentropy": 2.253898859024048, | |
| "epoch": 8.25, | |
| "grad_norm": 0.03283696994185448, | |
| "grad_norm_var": 1.361639072858939e-06, | |
| "learning_rate": 0.0009549150281252633, | |
| "loss": 2.2539, | |
| "step": 66 | |
| }, | |
| { | |
| "crossentropy": 2.1402039527893066, | |
| "epoch": 8.375, | |
| "grad_norm": 0.03309963271021843, | |
| "grad_norm_var": 1.3845661239702636e-06, | |
| "learning_rate": 0.0008271337313934868, | |
| "loss": 2.1402, | |
| "step": 67 | |
| }, | |
| { | |
| "crossentropy": 2.1381993293762207, | |
| "epoch": 8.5, | |
| "grad_norm": 0.03431132435798645, | |
| "grad_norm_var": 1.3068839073976725e-06, | |
| "learning_rate": 0.0007077560319906695, | |
| "loss": 2.1382, | |
| "step": 68 | |
| }, | |
| { | |
| "crossentropy": 2.1610183715820312, | |
| "epoch": 8.625, | |
| "grad_norm": 0.03382508084177971, | |
| "grad_norm_var": 9.367598844583584e-07, | |
| "learning_rate": 0.00059702234071631, | |
| "loss": 2.161, | |
| "step": 69 | |
| }, | |
| { | |
| "crossentropy": 2.2194838523864746, | |
| "epoch": 8.75, | |
| "grad_norm": 0.03652471676468849, | |
| "grad_norm_var": 1.3163803696068672e-06, | |
| "learning_rate": 0.0004951556604879049, | |
| "loss": 2.2195, | |
| "step": 70 | |
| }, | |
| { | |
| "crossentropy": 2.179126262664795, | |
| "epoch": 8.875, | |
| "grad_norm": 0.03386425971984863, | |
| "grad_norm_var": 1.2798076699575718e-06, | |
| "learning_rate": 0.0004023611372427471, | |
| "loss": 2.1791, | |
| "step": 71 | |
| }, | |
| { | |
| "crossentropy": 2.2535359859466553, | |
| "epoch": 9.0, | |
| "grad_norm": 0.032655857503414154, | |
| "grad_norm_var": 1.3903296123838708e-06, | |
| "learning_rate": 0.00031882564680131396, | |
| "loss": 2.2535, | |
| "step": 72 | |
| }, | |
| { | |
| "crossentropy": 2.083571434020996, | |
| "epoch": 9.125, | |
| "grad_norm": 0.03222977742552757, | |
| "grad_norm_var": 1.6381791075122573e-06, | |
| "learning_rate": 0.00024471741852423234, | |
| "loss": 2.0836, | |
| "step": 73 | |
| }, | |
| { | |
| "crossentropy": 2.1447954177856445, | |
| "epoch": 9.25, | |
| "grad_norm": 0.03370295464992523, | |
| "grad_norm_var": 1.585818962655111e-06, | |
| "learning_rate": 0.0001801856965207338, | |
| "loss": 2.1448, | |
| "step": 74 | |
| }, | |
| { | |
| "crossentropy": 2.3297011852264404, | |
| "epoch": 9.375, | |
| "grad_norm": 0.03342713788151741, | |
| "grad_norm_var": 1.5866984901570397e-06, | |
| "learning_rate": 0.0001253604390908819, | |
| "loss": 2.3297, | |
| "step": 75 | |
| }, | |
| { | |
| "crossentropy": 2.184936761856079, | |
| "epoch": 9.5, | |
| "grad_norm": 0.03379293903708458, | |
| "grad_norm_var": 1.128480817529492e-06, | |
| "learning_rate": 8.035205700685166e-05, | |
| "loss": 2.1849, | |
| "step": 76 | |
| }, | |
| { | |
| "crossentropy": 2.244959592819214, | |
| "epoch": 9.625, | |
| "grad_norm": 0.03535815700888634, | |
| "grad_norm_var": 1.2224064569112125e-06, | |
| "learning_rate": 4.52511911603265e-05, | |
| "loss": 2.245, | |
| "step": 77 | |
| }, | |
| { | |
| "crossentropy": 2.145529270172119, | |
| "epoch": 9.75, | |
| "grad_norm": 0.03338780626654625, | |
| "grad_norm_var": 1.180987359381859e-06, | |
| "learning_rate": 2.012853002380466e-05, | |
| "loss": 2.1455, | |
| "step": 78 | |
| }, | |
| { | |
| "crossentropy": 2.282687187194824, | |
| "epoch": 9.875, | |
| "grad_norm": 0.03313542157411575, | |
| "grad_norm_var": 1.1489689212500093e-06, | |
| "learning_rate": 5.034667293427053e-06, | |
| "loss": 2.2827, | |
| "step": 79 | |
| }, | |
| { | |
| "crossentropy": 2.1934750080108643, | |
| "epoch": 10.0, | |
| "grad_norm": 0.03277941048145294, | |
| "grad_norm_var": 1.18509241838338e-06, | |
| "learning_rate": 0.0, | |
| "loss": 2.1935, | |
| "step": 80 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 80, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 2000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.384645494177792e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |