| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.4934383202099737, | |
| "eval_steps": 500, | |
| "global_step": 285, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005249343832020997, | |
| "grad_norm": 1.1348930782232016, | |
| "learning_rate": 1.5000000000000002e-07, | |
| "loss": 1.1087, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.010498687664041995, | |
| "grad_norm": 1.123696373079589, | |
| "learning_rate": 3.0000000000000004e-07, | |
| "loss": 1.1356, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.015748031496062992, | |
| "grad_norm": 1.0989081863562118, | |
| "learning_rate": 4.5e-07, | |
| "loss": 1.1158, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.02099737532808399, | |
| "grad_norm": 1.0628548113414964, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 1.0986, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.026246719160104987, | |
| "grad_norm": 1.0629069543612368, | |
| "learning_rate": 7.5e-07, | |
| "loss": 1.0727, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.031496062992125984, | |
| "grad_norm": 1.1219311917213644, | |
| "learning_rate": 9e-07, | |
| "loss": 1.1513, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.03674540682414698, | |
| "grad_norm": 1.068318638334139, | |
| "learning_rate": 1.05e-06, | |
| "loss": 1.0978, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.04199475065616798, | |
| "grad_norm": 1.0335025624008565, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 1.0932, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.047244094488188976, | |
| "grad_norm": 0.9514112971268772, | |
| "learning_rate": 1.35e-06, | |
| "loss": 1.1046, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.05249343832020997, | |
| "grad_norm": 0.8944230714776324, | |
| "learning_rate": 1.5e-06, | |
| "loss": 1.0638, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05774278215223097, | |
| "grad_norm": 0.8720343077794245, | |
| "learning_rate": 1.65e-06, | |
| "loss": 1.1132, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.06299212598425197, | |
| "grad_norm": 0.7519518665820406, | |
| "learning_rate": 1.8e-06, | |
| "loss": 1.0788, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.06824146981627296, | |
| "grad_norm": 0.7768466543241798, | |
| "learning_rate": 1.95e-06, | |
| "loss": 1.0795, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.07349081364829396, | |
| "grad_norm": 0.7109922479048013, | |
| "learning_rate": 2.1e-06, | |
| "loss": 1.1012, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.07874015748031496, | |
| "grad_norm": 0.6312078880187205, | |
| "learning_rate": 2.25e-06, | |
| "loss": 1.0851, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.08398950131233596, | |
| "grad_norm": 0.5514473048370377, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 1.1041, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.08923884514435695, | |
| "grad_norm": 0.6271281070432462, | |
| "learning_rate": 2.55e-06, | |
| "loss": 1.0855, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.09448818897637795, | |
| "grad_norm": 0.7059888078645049, | |
| "learning_rate": 2.7e-06, | |
| "loss": 1.0473, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.09973753280839895, | |
| "grad_norm": 0.7226157330393405, | |
| "learning_rate": 2.85e-06, | |
| "loss": 1.0665, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.10498687664041995, | |
| "grad_norm": 0.7244742832208652, | |
| "learning_rate": 3e-06, | |
| "loss": 1.0604, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.11023622047244094, | |
| "grad_norm": 0.7088251146482789, | |
| "learning_rate": 3.1500000000000003e-06, | |
| "loss": 1.0516, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.11548556430446194, | |
| "grad_norm": 0.5987242362229293, | |
| "learning_rate": 3.3e-06, | |
| "loss": 1.084, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.12073490813648294, | |
| "grad_norm": 0.5730637810768702, | |
| "learning_rate": 3.45e-06, | |
| "loss": 1.0621, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.12598425196850394, | |
| "grad_norm": 0.5894968443138215, | |
| "learning_rate": 3.6e-06, | |
| "loss": 1.0797, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.13123359580052493, | |
| "grad_norm": 0.5798124303184627, | |
| "learning_rate": 3.75e-06, | |
| "loss": 1.0035, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.13648293963254593, | |
| "grad_norm": 0.643205751513686, | |
| "learning_rate": 3.9e-06, | |
| "loss": 1.0455, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.14173228346456693, | |
| "grad_norm": 0.5621970774702022, | |
| "learning_rate": 4.05e-06, | |
| "loss": 1.0576, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.14698162729658792, | |
| "grad_norm": 0.5506084571895594, | |
| "learning_rate": 4.2e-06, | |
| "loss": 1.0298, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.15223097112860892, | |
| "grad_norm": 0.48741149421912777, | |
| "learning_rate": 4.35e-06, | |
| "loss": 1.0018, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.15748031496062992, | |
| "grad_norm": 0.46403007703544275, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.9872, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.16272965879265092, | |
| "grad_norm": 0.4754381818573106, | |
| "learning_rate": 4.65e-06, | |
| "loss": 1.0271, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.1679790026246719, | |
| "grad_norm": 0.9362850890979981, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 1.0437, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.1732283464566929, | |
| "grad_norm": 0.47391181595772164, | |
| "learning_rate": 4.95e-06, | |
| "loss": 1.0437, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.1784776902887139, | |
| "grad_norm": 0.5276920454851337, | |
| "learning_rate": 5.1e-06, | |
| "loss": 1.0557, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1837270341207349, | |
| "grad_norm": 0.4616075133913133, | |
| "learning_rate": 5.2500000000000006e-06, | |
| "loss": 1.0465, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.1889763779527559, | |
| "grad_norm": 0.4555174555636226, | |
| "learning_rate": 5.4e-06, | |
| "loss": 1.0588, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.1942257217847769, | |
| "grad_norm": 0.5071864534648831, | |
| "learning_rate": 5.55e-06, | |
| "loss": 1.044, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.1994750656167979, | |
| "grad_norm": 0.4851367263882934, | |
| "learning_rate": 5.7e-06, | |
| "loss": 1.0464, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2047244094488189, | |
| "grad_norm": 0.44188022228811896, | |
| "learning_rate": 5.85e-06, | |
| "loss": 1.0182, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.2099737532808399, | |
| "grad_norm": 0.43420740120454643, | |
| "learning_rate": 6e-06, | |
| "loss": 1.0188, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2152230971128609, | |
| "grad_norm": 0.4291543441241407, | |
| "learning_rate": 5.9998719351101036e-06, | |
| "loss": 1.0245, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2204724409448819, | |
| "grad_norm": 0.43326370236005163, | |
| "learning_rate": 5.999487751374158e-06, | |
| "loss": 1.0238, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.22572178477690288, | |
| "grad_norm": 0.427571644972227, | |
| "learning_rate": 5.998847481592462e-06, | |
| "loss": 1.0311, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.23097112860892388, | |
| "grad_norm": 0.4215063088273006, | |
| "learning_rate": 5.997951180429069e-06, | |
| "loss": 0.9925, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.23622047244094488, | |
| "grad_norm": 0.4206536914503675, | |
| "learning_rate": 5.996798924407118e-06, | |
| "loss": 1.003, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.24146981627296588, | |
| "grad_norm": 0.40910969064965136, | |
| "learning_rate": 5.995390811902302e-06, | |
| "loss": 0.9949, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.24671916010498687, | |
| "grad_norm": 0.4165775049327623, | |
| "learning_rate": 5.993726963134471e-06, | |
| "loss": 0.9734, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.25196850393700787, | |
| "grad_norm": 0.3832235501001726, | |
| "learning_rate": 5.9918075201573645e-06, | |
| "loss": 0.9485, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.2572178477690289, | |
| "grad_norm": 0.37002495168808525, | |
| "learning_rate": 5.9896326468464835e-06, | |
| "loss": 0.9358, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.26246719160104987, | |
| "grad_norm": 0.44836853406053057, | |
| "learning_rate": 5.987202528885104e-06, | |
| "loss": 0.9982, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2677165354330709, | |
| "grad_norm": 0.4080608606117312, | |
| "learning_rate": 5.984517373748417e-06, | |
| "loss": 1.0129, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.27296587926509186, | |
| "grad_norm": 0.4001550595702573, | |
| "learning_rate": 5.981577410685822e-06, | |
| "loss": 0.9788, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.2782152230971129, | |
| "grad_norm": 0.41021488877460305, | |
| "learning_rate": 5.978382890701347e-06, | |
| "loss": 1.0262, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.28346456692913385, | |
| "grad_norm": 0.39997016380492506, | |
| "learning_rate": 5.9749340865322284e-06, | |
| "loss": 1.0275, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.2887139107611549, | |
| "grad_norm": 0.3839823787027912, | |
| "learning_rate": 5.971231292625615e-06, | |
| "loss": 0.9374, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.29396325459317585, | |
| "grad_norm": 0.4125068495663659, | |
| "learning_rate": 5.967274825113438e-06, | |
| "loss": 0.9954, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.2992125984251969, | |
| "grad_norm": 0.3908377197765856, | |
| "learning_rate": 5.963065021785414e-06, | |
| "loss": 0.9671, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.30446194225721784, | |
| "grad_norm": 0.3850488592862481, | |
| "learning_rate": 5.958602242060207e-06, | |
| "loss": 0.9657, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.30971128608923887, | |
| "grad_norm": 0.3877990366088493, | |
| "learning_rate": 5.95388686695475e-06, | |
| "loss": 0.9678, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.31496062992125984, | |
| "grad_norm": 0.40470471194287355, | |
| "learning_rate": 5.948919299051706e-06, | |
| "loss": 1.0149, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.32020997375328086, | |
| "grad_norm": 0.42889495063392963, | |
| "learning_rate": 5.943699962465096e-06, | |
| "loss": 1.033, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.32545931758530183, | |
| "grad_norm": 0.39164358737100274, | |
| "learning_rate": 5.9382293028040985e-06, | |
| "loss": 0.9761, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.33070866141732286, | |
| "grad_norm": 0.3869342590567232, | |
| "learning_rate": 5.9325077871349975e-06, | |
| "loss": 0.9982, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.3359580052493438, | |
| "grad_norm": 0.39264627926569035, | |
| "learning_rate": 5.9265359039413105e-06, | |
| "loss": 0.9667, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.34120734908136485, | |
| "grad_norm": 0.3887717698297268, | |
| "learning_rate": 5.920314163082079e-06, | |
| "loss": 0.9806, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.3464566929133858, | |
| "grad_norm": 0.40896336915084297, | |
| "learning_rate": 5.913843095748342e-06, | |
| "loss": 1.0135, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.35170603674540685, | |
| "grad_norm": 0.3610209560875707, | |
| "learning_rate": 5.907123254417783e-06, | |
| "loss": 0.956, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.3569553805774278, | |
| "grad_norm": 0.38154744815823505, | |
| "learning_rate": 5.9001552128075625e-06, | |
| "loss": 1.0045, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.36220472440944884, | |
| "grad_norm": 0.4094826396119445, | |
| "learning_rate": 5.892939565825335e-06, | |
| "loss": 1.0069, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.3674540682414698, | |
| "grad_norm": 0.39129138622932325, | |
| "learning_rate": 5.885476929518457e-06, | |
| "loss": 0.9525, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.37270341207349084, | |
| "grad_norm": 0.3712890701175899, | |
| "learning_rate": 5.8777679410213956e-06, | |
| "loss": 0.9792, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.3779527559055118, | |
| "grad_norm": 0.4086264062600148, | |
| "learning_rate": 5.869813258501323e-06, | |
| "loss": 0.9926, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.38320209973753283, | |
| "grad_norm": 0.368975878599487, | |
| "learning_rate": 5.861613561101934e-06, | |
| "loss": 0.9643, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.3884514435695538, | |
| "grad_norm": 0.36792811629461203, | |
| "learning_rate": 5.853169548885461e-06, | |
| "loss": 0.9867, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.3937007874015748, | |
| "grad_norm": 0.3566251893981936, | |
| "learning_rate": 5.844481942772898e-06, | |
| "loss": 1.0069, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.3989501312335958, | |
| "grad_norm": 0.4578529359685586, | |
| "learning_rate": 5.835551484482459e-06, | |
| "loss": 1.0173, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.4041994750656168, | |
| "grad_norm": 0.3935925285922137, | |
| "learning_rate": 5.826378936466249e-06, | |
| "loss": 0.9743, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.4094488188976378, | |
| "grad_norm": 0.4109939217838428, | |
| "learning_rate": 5.81696508184517e-06, | |
| "loss": 0.9866, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.4146981627296588, | |
| "grad_norm": 0.3839870332489822, | |
| "learning_rate": 5.807310724342058e-06, | |
| "loss": 0.9516, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.4199475065616798, | |
| "grad_norm": 0.3774576797883406, | |
| "learning_rate": 5.797416688213067e-06, | |
| "loss": 0.9895, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.4251968503937008, | |
| "grad_norm": 0.3817468964498129, | |
| "learning_rate": 5.787283818177297e-06, | |
| "loss": 0.9632, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.4304461942257218, | |
| "grad_norm": 0.60843002346461, | |
| "learning_rate": 5.776912979344669e-06, | |
| "loss": 1.0166, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.4356955380577428, | |
| "grad_norm": 0.3858713700245362, | |
| "learning_rate": 5.766305057142073e-06, | |
| "loss": 0.9976, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.4409448818897638, | |
| "grad_norm": 0.3724153436541016, | |
| "learning_rate": 5.755460957237769e-06, | |
| "loss": 0.9645, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.4461942257217848, | |
| "grad_norm": 0.38201105695018567, | |
| "learning_rate": 5.744381605464064e-06, | |
| "loss": 0.9899, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.45144356955380577, | |
| "grad_norm": 0.38383930861007165, | |
| "learning_rate": 5.7330679477382655e-06, | |
| "loss": 0.9919, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.4566929133858268, | |
| "grad_norm": 0.4078870418259581, | |
| "learning_rate": 5.7215209499819296e-06, | |
| "loss": 0.9797, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.46194225721784776, | |
| "grad_norm": 0.38463767466523974, | |
| "learning_rate": 5.709741598038387e-06, | |
| "loss": 0.9597, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.4671916010498688, | |
| "grad_norm": 0.36309855116472584, | |
| "learning_rate": 5.697730897588577e-06, | |
| "loss": 0.9737, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.47244094488188976, | |
| "grad_norm": 0.4106701446638758, | |
| "learning_rate": 5.685489874065187e-06, | |
| "loss": 0.9683, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.4776902887139108, | |
| "grad_norm": 0.37110409255145443, | |
| "learning_rate": 5.673019572565103e-06, | |
| "loss": 1.0418, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.48293963254593175, | |
| "grad_norm": 0.3558357783330656, | |
| "learning_rate": 5.660321057760186e-06, | |
| "loss": 1.0055, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.4881889763779528, | |
| "grad_norm": 0.40499489938404787, | |
| "learning_rate": 5.6473954138063674e-06, | |
| "loss": 1.0113, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.49343832020997375, | |
| "grad_norm": 0.39428526462199764, | |
| "learning_rate": 5.634243744251094e-06, | |
| "loss": 0.9875, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.49868766404199477, | |
| "grad_norm": 0.3711741011240413, | |
| "learning_rate": 5.620867171939109e-06, | |
| "loss": 0.9749, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.5039370078740157, | |
| "grad_norm": 0.3961340085644134, | |
| "learning_rate": 5.607266838916585e-06, | |
| "loss": 0.982, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.5091863517060368, | |
| "grad_norm": 0.3784646685814138, | |
| "learning_rate": 5.593443906333624e-06, | |
| "loss": 0.9957, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.5144356955380578, | |
| "grad_norm": 0.3750460397069026, | |
| "learning_rate": 5.579399554345118e-06, | |
| "loss": 0.9755, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.5196850393700787, | |
| "grad_norm": 0.3746718538274792, | |
| "learning_rate": 5.565134982009994e-06, | |
| "loss": 0.9736, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.5249343832020997, | |
| "grad_norm": 0.38418890409196027, | |
| "learning_rate": 5.550651407188843e-06, | |
| "loss": 0.9506, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5301837270341208, | |
| "grad_norm": 0.422976375435725, | |
| "learning_rate": 5.535950066439941e-06, | |
| "loss": 1.0141, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.5354330708661418, | |
| "grad_norm": 0.38354451243133536, | |
| "learning_rate": 5.521032214913679e-06, | |
| "loss": 0.9618, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.5406824146981627, | |
| "grad_norm": 0.38257660011773076, | |
| "learning_rate": 5.505899126245397e-06, | |
| "loss": 0.939, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.5459317585301837, | |
| "grad_norm": 0.3768438915225408, | |
| "learning_rate": 5.490552092446652e-06, | |
| "loss": 0.9675, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.5511811023622047, | |
| "grad_norm": 0.3749655286727107, | |
| "learning_rate": 5.474992423794907e-06, | |
| "loss": 0.9592, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.5564304461942258, | |
| "grad_norm": 0.38461916993489687, | |
| "learning_rate": 5.459221448721664e-06, | |
| "loss": 0.9623, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.5616797900262467, | |
| "grad_norm": 0.35648642966931204, | |
| "learning_rate": 5.443240513699045e-06, | |
| "loss": 0.985, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.5669291338582677, | |
| "grad_norm": 0.4051560712719681, | |
| "learning_rate": 5.427050983124842e-06, | |
| "loss": 0.9407, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.5721784776902887, | |
| "grad_norm": 0.3769879713701903, | |
| "learning_rate": 5.410654239206021e-06, | |
| "loss": 0.968, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.5774278215223098, | |
| "grad_norm": 0.3746822083724367, | |
| "learning_rate": 5.394051681840719e-06, | |
| "loss": 0.9497, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.5826771653543307, | |
| "grad_norm": 0.3987231911136733, | |
| "learning_rate": 5.3772447284987216e-06, | |
| "loss": 0.961, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.5879265091863517, | |
| "grad_norm": 0.37848222525971176, | |
| "learning_rate": 5.36023481410045e-06, | |
| "loss": 0.9707, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.5931758530183727, | |
| "grad_norm": 0.3794904855253974, | |
| "learning_rate": 5.343023390894446e-06, | |
| "loss": 0.9714, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.5984251968503937, | |
| "grad_norm": 0.37452267525256994, | |
| "learning_rate": 5.325611928333389e-06, | |
| "loss": 0.9406, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.6036745406824147, | |
| "grad_norm": 0.39474437059829304, | |
| "learning_rate": 5.308001912948637e-06, | |
| "loss": 0.9626, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.6089238845144357, | |
| "grad_norm": 0.4023921986663554, | |
| "learning_rate": 5.290194848223309e-06, | |
| "loss": 0.9889, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.6141732283464567, | |
| "grad_norm": 0.39963771712171875, | |
| "learning_rate": 5.272192254463929e-06, | |
| "loss": 0.9639, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.6194225721784777, | |
| "grad_norm": 0.3893586064595733, | |
| "learning_rate": 5.2539956686706205e-06, | |
| "loss": 0.9469, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.6246719160104987, | |
| "grad_norm": 0.4651495625439333, | |
| "learning_rate": 5.2356066444058875e-06, | |
| "loss": 0.9658, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.6299212598425197, | |
| "grad_norm": 0.39599728107932586, | |
| "learning_rate": 5.217026751661978e-06, | |
| "loss": 1.0137, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.6351706036745407, | |
| "grad_norm": 0.406988761369817, | |
| "learning_rate": 5.198257576726835e-06, | |
| "loss": 0.9306, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.6404199475065617, | |
| "grad_norm": 0.3611939094322339, | |
| "learning_rate": 5.179300722048673e-06, | |
| "loss": 0.9462, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.6456692913385826, | |
| "grad_norm": 0.3809841775392484, | |
| "learning_rate": 5.1601578060991645e-06, | |
| "loss": 0.953, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.6509186351706037, | |
| "grad_norm": 0.46022843064705843, | |
| "learning_rate": 5.1408304632352575e-06, | |
| "loss": 0.9422, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.6561679790026247, | |
| "grad_norm": 0.3979704646560941, | |
| "learning_rate": 5.1213203435596425e-06, | |
| "loss": 0.9751, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.6614173228346457, | |
| "grad_norm": 0.39388496260457084, | |
| "learning_rate": 5.101629112779873e-06, | |
| "loss": 0.9722, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.3899148438115094, | |
| "learning_rate": 5.08175845206615e-06, | |
| "loss": 0.9652, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.6719160104986877, | |
| "grad_norm": 0.37391882787694275, | |
| "learning_rate": 5.061710057907788e-06, | |
| "loss": 0.9621, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.6771653543307087, | |
| "grad_norm": 0.39500875865406576, | |
| "learning_rate": 5.041485641968385e-06, | |
| "loss": 0.9899, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.6824146981627297, | |
| "grad_norm": 0.37540362490802714, | |
| "learning_rate": 5.021086930939672e-06, | |
| "loss": 0.9472, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.6876640419947506, | |
| "grad_norm": 0.3940788728379769, | |
| "learning_rate": 5.000515666394105e-06, | |
| "loss": 0.9479, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.6929133858267716, | |
| "grad_norm": 0.3919125365655477, | |
| "learning_rate": 4.979773604636169e-06, | |
| "loss": 0.9624, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.6981627296587927, | |
| "grad_norm": 0.3804552314744538, | |
| "learning_rate": 4.958862516552433e-06, | |
| "loss": 0.9806, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.7034120734908137, | |
| "grad_norm": 0.3674434286105591, | |
| "learning_rate": 4.937784187460362e-06, | |
| "loss": 0.9511, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.7086614173228346, | |
| "grad_norm": 0.4109777494732396, | |
| "learning_rate": 4.916540416955884e-06, | |
| "loss": 0.9943, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.7139107611548556, | |
| "grad_norm": 0.40231567788837497, | |
| "learning_rate": 4.895133018759753e-06, | |
| "loss": 0.9798, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.7191601049868767, | |
| "grad_norm": 0.3721834479908975, | |
| "learning_rate": 4.873563820562698e-06, | |
| "loss": 0.9504, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.7244094488188977, | |
| "grad_norm": 0.36127526200518306, | |
| "learning_rate": 4.851834663869379e-06, | |
| "loss": 0.9517, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.7296587926509186, | |
| "grad_norm": 0.3513827139135777, | |
| "learning_rate": 4.82994740384117e-06, | |
| "loss": 0.9835, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.7349081364829396, | |
| "grad_norm": 0.36760728272750326, | |
| "learning_rate": 4.80790390913777e-06, | |
| "loss": 0.9503, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.7401574803149606, | |
| "grad_norm": 0.36275280721999276, | |
| "learning_rate": 4.785706061757656e-06, | |
| "loss": 0.9743, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.7454068241469817, | |
| "grad_norm": 0.3733380512329921, | |
| "learning_rate": 4.763355756877419e-06, | |
| "loss": 0.9384, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.7506561679790026, | |
| "grad_norm": 0.3801691027568987, | |
| "learning_rate": 4.740854902689947e-06, | |
| "loss": 0.9296, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.7559055118110236, | |
| "grad_norm": 0.39053906811778566, | |
| "learning_rate": 4.718205420241516e-06, | |
| "loss": 0.9488, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.7611548556430446, | |
| "grad_norm": 0.3923993707534958, | |
| "learning_rate": 4.695409243267776e-06, | |
| "loss": 0.9383, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.7664041994750657, | |
| "grad_norm": 0.364792552828712, | |
| "learning_rate": 4.672468318028657e-06, | |
| "loss": 0.9193, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.7716535433070866, | |
| "grad_norm": 0.35070825551906964, | |
| "learning_rate": 4.649384603142202e-06, | |
| "loss": 0.9164, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.7769028871391076, | |
| "grad_norm": 0.37099778180795795, | |
| "learning_rate": 4.626160069417348e-06, | |
| "loss": 0.9425, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.7821522309711286, | |
| "grad_norm": 0.36954118968922517, | |
| "learning_rate": 4.602796699685665e-06, | |
| "loss": 0.9265, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.7874015748031497, | |
| "grad_norm": 0.4076466706382121, | |
| "learning_rate": 4.579296488632067e-06, | |
| "loss": 1.0133, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.7926509186351706, | |
| "grad_norm": 0.4015334925568992, | |
| "learning_rate": 4.5556614426245165e-06, | |
| "loss": 0.9486, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.7979002624671916, | |
| "grad_norm": 0.39628644809730684, | |
| "learning_rate": 4.5318935795427206e-06, | |
| "loss": 0.9605, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.8031496062992126, | |
| "grad_norm": 0.36792154742540445, | |
| "learning_rate": 4.507994928605862e-06, | |
| "loss": 0.9287, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.8083989501312336, | |
| "grad_norm": 0.3887839296706913, | |
| "learning_rate": 4.483967530199337e-06, | |
| "loss": 0.951, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.8136482939632546, | |
| "grad_norm": 0.36716852968968616, | |
| "learning_rate": 4.459813435700569e-06, | |
| "loss": 0.9702, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.8188976377952756, | |
| "grad_norm": 0.3533521076976156, | |
| "learning_rate": 4.4355347073038595e-06, | |
| "loss": 0.9612, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.8241469816272966, | |
| "grad_norm": 0.3499649930079787, | |
| "learning_rate": 4.411133417844328e-06, | |
| "loss": 0.9599, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.8293963254593176, | |
| "grad_norm": 0.38582146832565867, | |
| "learning_rate": 4.38661165062094e-06, | |
| "loss": 0.9894, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.8346456692913385, | |
| "grad_norm": 0.39040836855795735, | |
| "learning_rate": 4.36197149921864e-06, | |
| "loss": 0.9747, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.8398950131233596, | |
| "grad_norm": 0.3798580758700489, | |
| "learning_rate": 4.3372150673296155e-06, | |
| "loss": 0.9654, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.8451443569553806, | |
| "grad_norm": 0.3764456540061034, | |
| "learning_rate": 4.3123444685736795e-06, | |
| "loss": 0.9823, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.8503937007874016, | |
| "grad_norm": 0.3771195417830333, | |
| "learning_rate": 4.287361826317827e-06, | |
| "loss": 0.9456, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.8556430446194225, | |
| "grad_norm": 0.37650137746409273, | |
| "learning_rate": 4.262269273494946e-06, | |
| "loss": 1.0022, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.8608923884514436, | |
| "grad_norm": 0.38148353077474145, | |
| "learning_rate": 4.237068952421711e-06, | |
| "loss": 0.964, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.8661417322834646, | |
| "grad_norm": 0.3982519128695332, | |
| "learning_rate": 4.2117630146156845e-06, | |
| "loss": 0.9673, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.8713910761154856, | |
| "grad_norm": 0.36000775624632003, | |
| "learning_rate": 4.186353620611627e-06, | |
| "loss": 0.9359, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.8766404199475065, | |
| "grad_norm": 0.36850454735662447, | |
| "learning_rate": 4.160842939777036e-06, | |
| "loss": 0.9422, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.8818897637795275, | |
| "grad_norm": 0.37804115639757085, | |
| "learning_rate": 4.135233150126931e-06, | |
| "loss": 0.9454, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.8871391076115486, | |
| "grad_norm": 0.3689383402086321, | |
| "learning_rate": 4.109526438137908e-06, | |
| "loss": 0.9455, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.8923884514435696, | |
| "grad_norm": 0.46527154775209717, | |
| "learning_rate": 4.08372499856146e-06, | |
| "loss": 0.9386, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.8976377952755905, | |
| "grad_norm": 0.45653306710128705, | |
| "learning_rate": 4.0578310342365975e-06, | |
| "loss": 0.9616, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.9028871391076115, | |
| "grad_norm": 0.3773630567359451, | |
| "learning_rate": 4.031846755901785e-06, | |
| "loss": 0.9285, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.9081364829396326, | |
| "grad_norm": 0.3644595191521506, | |
| "learning_rate": 4.005774382006182e-06, | |
| "loss": 0.9663, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.9133858267716536, | |
| "grad_norm": 0.3539767481135477, | |
| "learning_rate": 3.97961613852025e-06, | |
| "loss": 0.9564, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.9186351706036745, | |
| "grad_norm": 0.3819676152776953, | |
| "learning_rate": 3.953374258745705e-06, | |
| "loss": 0.9607, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.9238845144356955, | |
| "grad_norm": 0.38397675786726637, | |
| "learning_rate": 3.927050983124842e-06, | |
| "loss": 0.9539, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.9291338582677166, | |
| "grad_norm": 0.3979084367711538, | |
| "learning_rate": 3.900648559049258e-06, | |
| "loss": 0.9505, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.9343832020997376, | |
| "grad_norm": 0.3756154385935223, | |
| "learning_rate": 3.874169240667974e-06, | |
| "loss": 0.9519, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.9396325459317585, | |
| "grad_norm": 0.40551973597201274, | |
| "learning_rate": 3.847615288694985e-06, | |
| "loss": 0.9727, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.9448818897637795, | |
| "grad_norm": 0.4149625851710124, | |
| "learning_rate": 3.820988970216249e-06, | |
| "loss": 0.9464, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.9501312335958005, | |
| "grad_norm": 0.35739115830542967, | |
| "learning_rate": 3.7942925584961272e-06, | |
| "loss": 0.9427, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.9553805774278216, | |
| "grad_norm": 0.3759540038847051, | |
| "learning_rate": 3.767528332783307e-06, | |
| "loss": 0.9679, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.9606299212598425, | |
| "grad_norm": 0.3525867658299593, | |
| "learning_rate": 3.740698578116199e-06, | |
| "loss": 0.9183, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.9658792650918635, | |
| "grad_norm": 0.3557123352774738, | |
| "learning_rate": 3.7138055851278564e-06, | |
| "loss": 0.9383, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.9711286089238845, | |
| "grad_norm": 0.3623514252763418, | |
| "learning_rate": 3.6868516498504025e-06, | |
| "loss": 0.9246, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.9763779527559056, | |
| "grad_norm": 0.38495496418054853, | |
| "learning_rate": 3.6598390735190066e-06, | |
| "loss": 0.9612, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.9816272965879265, | |
| "grad_norm": 0.3648599004428126, | |
| "learning_rate": 3.63277016237541e-06, | |
| "loss": 0.9293, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.9868766404199475, | |
| "grad_norm": 0.38871547084803876, | |
| "learning_rate": 3.6056472274710305e-06, | |
| "loss": 0.9973, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.9921259842519685, | |
| "grad_norm": 0.38590844403642666, | |
| "learning_rate": 3.578472584469651e-06, | |
| "loss": 0.9457, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.9973753280839895, | |
| "grad_norm": 0.3872507088649178, | |
| "learning_rate": 3.5512485534497116e-06, | |
| "loss": 0.9462, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.3872507088649178, | |
| "learning_rate": 3.523977458706237e-06, | |
| "loss": 0.9693, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.005249343832021, | |
| "grad_norm": 0.6232728744646114, | |
| "learning_rate": 3.49666162855239e-06, | |
| "loss": 0.887, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.010498687664042, | |
| "grad_norm": 0.4149641950734625, | |
| "learning_rate": 3.469303395120693e-06, | |
| "loss": 0.8826, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.015748031496063, | |
| "grad_norm": 0.37273340109017755, | |
| "learning_rate": 3.441905094163913e-06, | |
| "loss": 0.8893, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.020997375328084, | |
| "grad_norm": 0.4113832689982837, | |
| "learning_rate": 3.414469064855647e-06, | |
| "loss": 0.9205, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.026246719160105, | |
| "grad_norm": 0.49485155842511663, | |
| "learning_rate": 3.3869976495906104e-06, | |
| "loss": 0.9074, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.031496062992126, | |
| "grad_norm": 0.3736781934252868, | |
| "learning_rate": 3.3594931937846498e-06, | |
| "loss": 0.8966, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.036745406824147, | |
| "grad_norm": 0.3758650059773124, | |
| "learning_rate": 3.3319580456745023e-06, | |
| "loss": 0.8759, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.041994750656168, | |
| "grad_norm": 0.4056031624712629, | |
| "learning_rate": 3.3043945561173092e-06, | |
| "loss": 0.8788, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.047244094488189, | |
| "grad_norm": 0.36344982085137467, | |
| "learning_rate": 3.2768050783899063e-06, | |
| "loss": 0.873, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.05249343832021, | |
| "grad_norm": 0.3760103676246, | |
| "learning_rate": 3.249191967987912e-06, | |
| "loss": 0.899, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.057742782152231, | |
| "grad_norm": 0.39433477834527153, | |
| "learning_rate": 3.221557582424622e-06, | |
| "loss": 0.9019, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.0629921259842519, | |
| "grad_norm": 0.3595753440791428, | |
| "learning_rate": 3.1939042810297328e-06, | |
| "loss": 0.8781, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.068241469816273, | |
| "grad_norm": 0.3743448170598354, | |
| "learning_rate": 3.16623442474791e-06, | |
| "loss": 0.8689, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.073490813648294, | |
| "grad_norm": 0.3618551186966609, | |
| "learning_rate": 3.138550375937219e-06, | |
| "loss": 0.9094, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.078740157480315, | |
| "grad_norm": 0.36577516842050983, | |
| "learning_rate": 3.1108544981674356e-06, | |
| "loss": 0.8668, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.083989501312336, | |
| "grad_norm": 0.3985134455319658, | |
| "learning_rate": 3.0831491560182495e-06, | |
| "loss": 0.9016, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.0892388451443569, | |
| "grad_norm": 0.37808489525197075, | |
| "learning_rate": 3.0554367148773897e-06, | |
| "loss": 0.895, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.094488188976378, | |
| "grad_norm": 0.4112784941005797, | |
| "learning_rate": 3.027719540738673e-06, | |
| "loss": 0.859, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.099737532808399, | |
| "grad_norm": 0.3830296759827936, | |
| "learning_rate": 3e-06, | |
| "loss": 0.8569, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.10498687664042, | |
| "grad_norm": 0.3930755503999148, | |
| "learning_rate": 2.972280459261328e-06, | |
| "loss": 0.8774, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.110236220472441, | |
| "grad_norm": 0.36738851637178116, | |
| "learning_rate": 2.944563285122611e-06, | |
| "loss": 0.9086, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.1154855643044619, | |
| "grad_norm": 0.3897160841039193, | |
| "learning_rate": 2.9168508439817515e-06, | |
| "loss": 0.889, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.120734908136483, | |
| "grad_norm": 0.39858146379374537, | |
| "learning_rate": 2.889145501832566e-06, | |
| "loss": 0.8964, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.125984251968504, | |
| "grad_norm": 0.3739395525411432, | |
| "learning_rate": 2.861449624062782e-06, | |
| "loss": 0.8884, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.1312335958005248, | |
| "grad_norm": 0.3755768464864809, | |
| "learning_rate": 2.83376557525209e-06, | |
| "loss": 0.851, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.136482939632546, | |
| "grad_norm": 0.38260315757882735, | |
| "learning_rate": 2.8060957189702674e-06, | |
| "loss": 0.9152, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.141732283464567, | |
| "grad_norm": 0.4205379839527009, | |
| "learning_rate": 2.7784424175753784e-06, | |
| "loss": 0.8683, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.1469816272965878, | |
| "grad_norm": 0.38325260941818995, | |
| "learning_rate": 2.7508080320120888e-06, | |
| "loss": 0.8943, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.152230971128609, | |
| "grad_norm": 0.3763198826603672, | |
| "learning_rate": 2.7231949216100943e-06, | |
| "loss": 0.8676, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.1574803149606299, | |
| "grad_norm": 0.3767162287387105, | |
| "learning_rate": 2.6956054438826918e-06, | |
| "loss": 0.8482, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.162729658792651, | |
| "grad_norm": 0.3486273740901837, | |
| "learning_rate": 2.668041954325498e-06, | |
| "loss": 0.8879, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.167979002624672, | |
| "grad_norm": 0.39084218665366566, | |
| "learning_rate": 2.640506806215351e-06, | |
| "loss": 0.8679, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.1732283464566928, | |
| "grad_norm": 0.3538552501730603, | |
| "learning_rate": 2.613002350409391e-06, | |
| "loss": 0.8871, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.178477690288714, | |
| "grad_norm": 0.36544200913577, | |
| "learning_rate": 2.585530935144354e-06, | |
| "loss": 0.8616, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.1837270341207349, | |
| "grad_norm": 0.3985990462573467, | |
| "learning_rate": 2.558094905836087e-06, | |
| "loss": 0.8917, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.188976377952756, | |
| "grad_norm": 0.42608518999556655, | |
| "learning_rate": 2.5306966048793067e-06, | |
| "loss": 0.8817, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.194225721784777, | |
| "grad_norm": 0.37952769789031354, | |
| "learning_rate": 2.5033383714476097e-06, | |
| "loss": 0.8985, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.1994750656167978, | |
| "grad_norm": 0.40804864076806885, | |
| "learning_rate": 2.4760225412937633e-06, | |
| "loss": 0.9073, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.204724409448819, | |
| "grad_norm": 0.4167713152946991, | |
| "learning_rate": 2.4487514465502885e-06, | |
| "loss": 0.8566, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.20997375328084, | |
| "grad_norm": 0.4022153540631621, | |
| "learning_rate": 2.42152741553035e-06, | |
| "loss": 0.8713, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.2152230971128608, | |
| "grad_norm": 0.4222065137992956, | |
| "learning_rate": 2.39435277252897e-06, | |
| "loss": 0.9035, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.220472440944882, | |
| "grad_norm": 0.3666365807384159, | |
| "learning_rate": 2.3672298376245908e-06, | |
| "loss": 0.8637, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.2257217847769029, | |
| "grad_norm": 0.3976853335036615, | |
| "learning_rate": 2.3401609264809953e-06, | |
| "loss": 0.9398, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.2309711286089238, | |
| "grad_norm": 0.37956934109451046, | |
| "learning_rate": 2.3131483501495985e-06, | |
| "loss": 0.8353, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.236220472440945, | |
| "grad_norm": 0.33722056538083744, | |
| "learning_rate": 2.2861944148721446e-06, | |
| "loss": 0.8786, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.2414698162729658, | |
| "grad_norm": 0.49777382093647954, | |
| "learning_rate": 2.2593014218838e-06, | |
| "loss": 0.8834, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.246719160104987, | |
| "grad_norm": 0.35315516410389436, | |
| "learning_rate": 2.232471667216693e-06, | |
| "loss": 0.8442, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.2519685039370079, | |
| "grad_norm": 0.3816124424363711, | |
| "learning_rate": 2.2057074415038725e-06, | |
| "loss": 0.8573, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.257217847769029, | |
| "grad_norm": 0.36319142999803095, | |
| "learning_rate": 2.1790110297837514e-06, | |
| "loss": 0.8481, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.26246719160105, | |
| "grad_norm": 0.34672889281207053, | |
| "learning_rate": 2.152384711305015e-06, | |
| "loss": 0.8623, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.2677165354330708, | |
| "grad_norm": 0.37448151544392105, | |
| "learning_rate": 2.1258307593320262e-06, | |
| "loss": 0.8751, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.272965879265092, | |
| "grad_norm": 0.37082567424502005, | |
| "learning_rate": 2.099351440950742e-06, | |
| "loss": 0.8914, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.2782152230971129, | |
| "grad_norm": 0.39074992783073415, | |
| "learning_rate": 2.072949016875158e-06, | |
| "loss": 0.9222, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.2834645669291338, | |
| "grad_norm": 0.4150437401629804, | |
| "learning_rate": 2.046625741254295e-06, | |
| "loss": 0.9475, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.288713910761155, | |
| "grad_norm": 0.4504166670407193, | |
| "learning_rate": 2.0203838614797505e-06, | |
| "loss": 0.9026, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.2939632545931758, | |
| "grad_norm": 0.38345958484903814, | |
| "learning_rate": 1.994225617993819e-06, | |
| "loss": 0.9074, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.2992125984251968, | |
| "grad_norm": 0.37086048031752866, | |
| "learning_rate": 1.9681532440982154e-06, | |
| "loss": 0.8755, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.304461942257218, | |
| "grad_norm": 0.3775524407980251, | |
| "learning_rate": 1.942168965763402e-06, | |
| "loss": 0.8986, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.3097112860892388, | |
| "grad_norm": 0.364796377340789, | |
| "learning_rate": 1.916275001438541e-06, | |
| "loss": 0.867, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.3149606299212597, | |
| "grad_norm": 0.3705604843330414, | |
| "learning_rate": 1.8904735618620928e-06, | |
| "loss": 0.8875, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.3202099737532809, | |
| "grad_norm": 0.3847344001283667, | |
| "learning_rate": 1.8647668498730693e-06, | |
| "loss": 0.8678, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.3254593175853018, | |
| "grad_norm": 0.3507183610862785, | |
| "learning_rate": 1.8391570602229647e-06, | |
| "loss": 0.8895, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.330708661417323, | |
| "grad_norm": 0.34464955572346173, | |
| "learning_rate": 1.8136463793883725e-06, | |
| "loss": 0.9112, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.3359580052493438, | |
| "grad_norm": 0.3804540728076062, | |
| "learning_rate": 1.7882369853843155e-06, | |
| "loss": 0.8818, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.341207349081365, | |
| "grad_norm": 0.38671544491057547, | |
| "learning_rate": 1.76293104757829e-06, | |
| "loss": 0.8712, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.3464566929133859, | |
| "grad_norm": 0.35028636565033566, | |
| "learning_rate": 1.7377307265050559e-06, | |
| "loss": 0.8795, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.3517060367454068, | |
| "grad_norm": 0.3596694021401425, | |
| "learning_rate": 1.7126381736821732e-06, | |
| "loss": 0.8791, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.356955380577428, | |
| "grad_norm": 0.3833574983214166, | |
| "learning_rate": 1.6876555314263213e-06, | |
| "loss": 0.9108, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.3622047244094488, | |
| "grad_norm": 0.3701840047085969, | |
| "learning_rate": 1.6627849326703855e-06, | |
| "loss": 0.8695, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.3674540682414698, | |
| "grad_norm": 0.36098816535443995, | |
| "learning_rate": 1.6380285007813598e-06, | |
| "loss": 0.876, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.372703412073491, | |
| "grad_norm": 0.3900890284585014, | |
| "learning_rate": 1.6133883493790609e-06, | |
| "loss": 0.8498, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.3779527559055118, | |
| "grad_norm": 0.34906551126755136, | |
| "learning_rate": 1.5888665821556724e-06, | |
| "loss": 0.8513, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.3832020997375327, | |
| "grad_norm": 0.3753732283477496, | |
| "learning_rate": 1.5644652926961407e-06, | |
| "loss": 0.8714, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.3884514435695539, | |
| "grad_norm": 0.34748864593560347, | |
| "learning_rate": 1.5401865642994315e-06, | |
| "loss": 0.9124, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.3937007874015748, | |
| "grad_norm": 0.36698053817770165, | |
| "learning_rate": 1.5160324698006642e-06, | |
| "loss": 0.8814, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.3989501312335957, | |
| "grad_norm": 0.4000964153653425, | |
| "learning_rate": 1.4920050713941398e-06, | |
| "loss": 0.9082, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.4041994750656168, | |
| "grad_norm": 0.3985391177875817, | |
| "learning_rate": 1.4681064204572798e-06, | |
| "loss": 0.8749, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.4094488188976377, | |
| "grad_norm": 0.3578122677174226, | |
| "learning_rate": 1.4443385573754837e-06, | |
| "loss": 0.8608, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.4146981627296589, | |
| "grad_norm": 0.3576093239254431, | |
| "learning_rate": 1.4207035113679322e-06, | |
| "loss": 0.8798, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.4199475065616798, | |
| "grad_norm": 0.35299639204379674, | |
| "learning_rate": 1.3972033003143348e-06, | |
| "loss": 0.8972, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.425196850393701, | |
| "grad_norm": 0.3937775289907907, | |
| "learning_rate": 1.3738399305826516e-06, | |
| "loss": 0.8736, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.4304461942257218, | |
| "grad_norm": 0.3691998032129419, | |
| "learning_rate": 1.3506153968577983e-06, | |
| "loss": 0.8667, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.4356955380577427, | |
| "grad_norm": 0.35764876894907843, | |
| "learning_rate": 1.3275316819713435e-06, | |
| "loss": 0.882, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.4409448818897639, | |
| "grad_norm": 0.3859579688778526, | |
| "learning_rate": 1.3045907567322243e-06, | |
| "loss": 0.844, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.4461942257217848, | |
| "grad_norm": 0.3736621084680505, | |
| "learning_rate": 1.2817945797584844e-06, | |
| "loss": 0.8525, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.4514435695538057, | |
| "grad_norm": 0.36602372507940695, | |
| "learning_rate": 1.2591450973100532e-06, | |
| "loss": 0.8577, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.4566929133858268, | |
| "grad_norm": 0.37926054124030645, | |
| "learning_rate": 1.236644243122581e-06, | |
| "loss": 0.8837, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.4619422572178478, | |
| "grad_norm": 0.3680022216795608, | |
| "learning_rate": 1.214293938242344e-06, | |
| "loss": 0.8984, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.4671916010498687, | |
| "grad_norm": 0.37824901927870175, | |
| "learning_rate": 1.1920960908622313e-06, | |
| "loss": 0.8745, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.4724409448818898, | |
| "grad_norm": 0.3489273490529577, | |
| "learning_rate": 1.17005259615883e-06, | |
| "loss": 0.8628, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.4776902887139107, | |
| "grad_norm": 0.3735770062938505, | |
| "learning_rate": 1.1481653361306215e-06, | |
| "loss": 0.8619, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.4829396325459316, | |
| "grad_norm": 0.3458041443504503, | |
| "learning_rate": 1.1264361794373032e-06, | |
| "loss": 0.8761, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.4881889763779528, | |
| "grad_norm": 0.35998420937846626, | |
| "learning_rate": 1.104866981240248e-06, | |
| "loss": 0.8844, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.4934383202099737, | |
| "grad_norm": 0.4029178073367971, | |
| "learning_rate": 1.0834595830441168e-06, | |
| "loss": 0.8511, | |
| "step": 285 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 380, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 95, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.000110270806753e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |