| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0006826715212198, | |
| "eval_steps": 275, | |
| "global_step": 1099, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0009102286949596086, | |
| "grad_norm": 0.3924100995063782, | |
| "learning_rate": 2e-05, | |
| "loss": 2.8811, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0009102286949596086, | |
| "eval_loss": 2.6547484397888184, | |
| "eval_runtime": 202.3055, | |
| "eval_samples_per_second": 9.15, | |
| "eval_steps_per_second": 4.577, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0018204573899192173, | |
| "grad_norm": 0.4433031976222992, | |
| "learning_rate": 4e-05, | |
| "loss": 2.8541, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0027306860848788257, | |
| "grad_norm": 0.4557134211063385, | |
| "learning_rate": 6e-05, | |
| "loss": 2.6364, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0036409147798384346, | |
| "grad_norm": 0.3975633978843689, | |
| "learning_rate": 8e-05, | |
| "loss": 2.6666, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.004551143474798043, | |
| "grad_norm": 0.36896416544914246, | |
| "learning_rate": 0.0001, | |
| "loss": 2.745, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0054613721697576514, | |
| "grad_norm": 0.30803996324539185, | |
| "learning_rate": 0.00012, | |
| "loss": 2.7417, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.00637160086471726, | |
| "grad_norm": 0.40208426117897034, | |
| "learning_rate": 0.00014, | |
| "loss": 2.7006, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.007281829559676869, | |
| "grad_norm": 0.5293606519699097, | |
| "learning_rate": 0.00016, | |
| "loss": 2.6852, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.008192058254636477, | |
| "grad_norm": 0.4535931646823883, | |
| "learning_rate": 0.00018, | |
| "loss": 2.6472, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.009102286949596085, | |
| "grad_norm": 0.32121461629867554, | |
| "learning_rate": 0.0002, | |
| "loss": 2.6648, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.010012515644555695, | |
| "grad_norm": 0.3374486565589905, | |
| "learning_rate": 0.00019999958388469571, | |
| "loss": 2.5696, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.010922744339515303, | |
| "grad_norm": 0.3663835823535919, | |
| "learning_rate": 0.00019999833554224577, | |
| "loss": 2.4938, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.011832973034474913, | |
| "grad_norm": 0.32625913619995117, | |
| "learning_rate": 0.00019999625498303932, | |
| "loss": 2.5661, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.01274320172943452, | |
| "grad_norm": 0.382977157831192, | |
| "learning_rate": 0.00019999334222439147, | |
| "loss": 2.5544, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.013653430424394129, | |
| "grad_norm": 0.3954707384109497, | |
| "learning_rate": 0.00019998959729054295, | |
| "loss": 2.5744, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.014563659119353738, | |
| "grad_norm": 0.30874142050743103, | |
| "learning_rate": 0.0001999850202126604, | |
| "loss": 2.5242, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.015473887814313346, | |
| "grad_norm": 0.30474328994750977, | |
| "learning_rate": 0.00019997961102883552, | |
| "loss": 2.4201, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.016384116509272954, | |
| "grad_norm": 0.32679283618927, | |
| "learning_rate": 0.00019997336978408531, | |
| "loss": 2.5165, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.017294345204232564, | |
| "grad_norm": 0.3267504572868347, | |
| "learning_rate": 0.00019996629653035126, | |
| "loss": 2.2016, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.01820457389919217, | |
| "grad_norm": 0.33519116044044495, | |
| "learning_rate": 0.00019995839132649917, | |
| "loss": 2.4004, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01911480259415178, | |
| "grad_norm": 0.36370110511779785, | |
| "learning_rate": 0.00019994965423831854, | |
| "loss": 2.4736, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.02002503128911139, | |
| "grad_norm": 0.36940237879753113, | |
| "learning_rate": 0.0001999400853385221, | |
| "loss": 2.4726, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.020935259984071, | |
| "grad_norm": 0.3428037762641907, | |
| "learning_rate": 0.0001999296847067452, | |
| "loss": 2.4291, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.021845488679030606, | |
| "grad_norm": 0.3404378294944763, | |
| "learning_rate": 0.00019991845242954505, | |
| "loss": 2.0424, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.022755717373990215, | |
| "grad_norm": 0.3714047074317932, | |
| "learning_rate": 0.00019990638860040006, | |
| "loss": 2.2809, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.023665946068949825, | |
| "grad_norm": 0.37741491198539734, | |
| "learning_rate": 0.00019989349331970923, | |
| "loss": 2.2841, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.02457617476390943, | |
| "grad_norm": 0.38504019379615784, | |
| "learning_rate": 0.00019987976669479088, | |
| "loss": 2.2056, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.02548640345886904, | |
| "grad_norm": 0.4182392954826355, | |
| "learning_rate": 0.00019986520883988232, | |
| "loss": 2.3747, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.02639663215382865, | |
| "grad_norm": 0.4266968369483948, | |
| "learning_rate": 0.0001998498198761384, | |
| "loss": 2.3608, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.027306860848788257, | |
| "grad_norm": 0.3748364746570587, | |
| "learning_rate": 0.00019983359993163078, | |
| "loss": 2.0789, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.028217089543747867, | |
| "grad_norm": 0.4157109558582306, | |
| "learning_rate": 0.00019981654914134686, | |
| "loss": 2.3174, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.029127318238707477, | |
| "grad_norm": 0.4328743517398834, | |
| "learning_rate": 0.00019979866764718843, | |
| "loss": 2.3304, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.030037546933667083, | |
| "grad_norm": 0.4545007050037384, | |
| "learning_rate": 0.0001997799555979709, | |
| "loss": 2.3578, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.030947775628626693, | |
| "grad_norm": 0.4287387430667877, | |
| "learning_rate": 0.00019976041314942155, | |
| "loss": 2.1072, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0318580043235863, | |
| "grad_norm": 0.4378172755241394, | |
| "learning_rate": 0.0001997400404641787, | |
| "loss": 2.2507, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.03276823301854591, | |
| "grad_norm": 0.42525431513786316, | |
| "learning_rate": 0.00019971883771179003, | |
| "loss": 2.1549, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.03367846171350552, | |
| "grad_norm": 0.4390096068382263, | |
| "learning_rate": 0.00019969680506871137, | |
| "loss": 2.3869, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.03458869040846513, | |
| "grad_norm": 0.4858846068382263, | |
| "learning_rate": 0.00019967394271830504, | |
| "loss": 2.5711, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.03549891910342474, | |
| "grad_norm": 0.46703851222991943, | |
| "learning_rate": 0.00019965025085083858, | |
| "loss": 2.3567, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.03640914779838434, | |
| "grad_norm": 0.4408855438232422, | |
| "learning_rate": 0.000199625729663483, | |
| "loss": 2.1646, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03731937649334395, | |
| "grad_norm": 0.49953216314315796, | |
| "learning_rate": 0.00019960037936031104, | |
| "loss": 2.3658, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.03822960518830356, | |
| "grad_norm": 0.48640862107276917, | |
| "learning_rate": 0.00019957420015229572, | |
| "loss": 2.4125, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.03913983388326317, | |
| "grad_norm": 0.5187159776687622, | |
| "learning_rate": 0.00019954719225730847, | |
| "loss": 2.3655, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.04005006257822278, | |
| "grad_norm": 0.48923739790916443, | |
| "learning_rate": 0.00019951935590011718, | |
| "loss": 2.257, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.04096029127318239, | |
| "grad_norm": 0.4975845515727997, | |
| "learning_rate": 0.0001994906913123846, | |
| "loss": 2.185, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.041870519968142, | |
| "grad_norm": 0.5635886788368225, | |
| "learning_rate": 0.00019946119873266613, | |
| "loss": 2.6952, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.0427807486631016, | |
| "grad_norm": 0.5890567898750305, | |
| "learning_rate": 0.00019943087840640814, | |
| "loss": 2.5154, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.04369097735806121, | |
| "grad_norm": 0.6770642995834351, | |
| "learning_rate": 0.0001993997305859456, | |
| "loss": 2.7011, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.04460120605302082, | |
| "grad_norm": 0.7621054649353027, | |
| "learning_rate": 0.0001993677555305002, | |
| "loss": 2.5097, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.04551143474798043, | |
| "grad_norm": 1.2259888648986816, | |
| "learning_rate": 0.00019933495350617813, | |
| "loss": 2.516, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04642166344294004, | |
| "grad_norm": 1.261356234550476, | |
| "learning_rate": 0.00019930132478596796, | |
| "loss": 2.7173, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.04733189213789965, | |
| "grad_norm": 0.5333342552185059, | |
| "learning_rate": 0.00019926686964973813, | |
| "loss": 2.7659, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.04824212083285925, | |
| "grad_norm": 0.43384629487991333, | |
| "learning_rate": 0.00019923158838423482, | |
| "loss": 2.4593, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.04915234952781886, | |
| "grad_norm": 0.506197452545166, | |
| "learning_rate": 0.00019919548128307954, | |
| "loss": 2.6135, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.05006257822277847, | |
| "grad_norm": 0.4372877776622772, | |
| "learning_rate": 0.00019915854864676664, | |
| "loss": 2.4952, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.05097280691773808, | |
| "grad_norm": 0.4044889509677887, | |
| "learning_rate": 0.00019912079078266085, | |
| "loss": 2.6152, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.05188303561269769, | |
| "grad_norm": 0.4284885823726654, | |
| "learning_rate": 0.0001990822080049946, | |
| "loss": 2.508, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.0527932643076573, | |
| "grad_norm": 0.4030478298664093, | |
| "learning_rate": 0.0001990428006348656, | |
| "loss": 2.4545, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.053703493002616905, | |
| "grad_norm": 0.37658700346946716, | |
| "learning_rate": 0.00019900256900023413, | |
| "loss": 2.3762, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.054613721697576514, | |
| "grad_norm": 0.3478519022464752, | |
| "learning_rate": 0.00019896151343592008, | |
| "loss": 2.2991, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.055523950392536124, | |
| "grad_norm": 0.38357114791870117, | |
| "learning_rate": 0.00019891963428360043, | |
| "loss": 2.4336, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.056434179087495734, | |
| "grad_norm": 0.37183457612991333, | |
| "learning_rate": 0.00019887693189180633, | |
| "loss": 2.3966, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.057344407782455344, | |
| "grad_norm": 0.3586178123950958, | |
| "learning_rate": 0.00019883340661592015, | |
| "loss": 2.3195, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.05825463647741495, | |
| "grad_norm": 0.37112224102020264, | |
| "learning_rate": 0.00019878905881817252, | |
| "loss": 2.3924, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.059164865172374556, | |
| "grad_norm": 0.37623754143714905, | |
| "learning_rate": 0.00019874388886763944, | |
| "loss": 2.562, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.060075093867334166, | |
| "grad_norm": 0.37167176604270935, | |
| "learning_rate": 0.00019869789714023906, | |
| "loss": 2.4175, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.060985322562293776, | |
| "grad_norm": 0.35394376516342163, | |
| "learning_rate": 0.00019865108401872857, | |
| "loss": 2.2845, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.061895551257253385, | |
| "grad_norm": 0.3475804030895233, | |
| "learning_rate": 0.00019860344989270113, | |
| "loss": 2.0894, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.062805779952213, | |
| "grad_norm": 0.387608140707016, | |
| "learning_rate": 0.0001985549951585825, | |
| "loss": 2.2795, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.0637160086471726, | |
| "grad_norm": 0.38371261954307556, | |
| "learning_rate": 0.00019850572021962788, | |
| "loss": 2.1714, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06462623734213221, | |
| "grad_norm": 0.397544264793396, | |
| "learning_rate": 0.00019845562548591826, | |
| "loss": 2.167, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.06553646603709182, | |
| "grad_norm": 0.394674688577652, | |
| "learning_rate": 0.00019840471137435746, | |
| "loss": 2.266, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.06644669473205143, | |
| "grad_norm": 0.3893704116344452, | |
| "learning_rate": 0.00019835297830866826, | |
| "loss": 2.2651, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.06735692342701104, | |
| "grad_norm": 0.430054247379303, | |
| "learning_rate": 0.00019830042671938904, | |
| "loss": 2.4341, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.06826715212197064, | |
| "grad_norm": 0.4117582440376282, | |
| "learning_rate": 0.00019824705704387028, | |
| "loss": 2.2562, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.06917738081693026, | |
| "grad_norm": 0.40804898738861084, | |
| "learning_rate": 0.00019819286972627066, | |
| "loss": 2.1574, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.07008760951188986, | |
| "grad_norm": 0.3954644501209259, | |
| "learning_rate": 0.00019813786521755372, | |
| "loss": 2.0693, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.07099783820684948, | |
| "grad_norm": 0.4126876890659332, | |
| "learning_rate": 0.00019808204397548377, | |
| "loss": 2.2299, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.07190806690180908, | |
| "grad_norm": 0.42160266637802124, | |
| "learning_rate": 0.0001980254064646223, | |
| "loss": 2.1759, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.07281829559676868, | |
| "grad_norm": 0.4152994751930237, | |
| "learning_rate": 0.00019796795315632395, | |
| "loss": 2.1931, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0737285242917283, | |
| "grad_norm": 0.4207664728164673, | |
| "learning_rate": 0.0001979096845287328, | |
| "loss": 2.0727, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.0746387529866879, | |
| "grad_norm": 0.41932612657546997, | |
| "learning_rate": 0.00019785060106677818, | |
| "loss": 2.3067, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.07554898168164752, | |
| "grad_norm": 0.44273534417152405, | |
| "learning_rate": 0.00019779070326217074, | |
| "loss": 2.2838, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.07645921037660712, | |
| "grad_norm": 0.4609782099723816, | |
| "learning_rate": 0.00019772999161339833, | |
| "loss": 2.2605, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.07736943907156674, | |
| "grad_norm": 0.4551486074924469, | |
| "learning_rate": 0.00019766846662572191, | |
| "loss": 2.3159, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.07827966776652634, | |
| "grad_norm": 0.4343845546245575, | |
| "learning_rate": 0.00019760612881117125, | |
| "loss": 2.1787, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.07918989646148594, | |
| "grad_norm": 0.48003000020980835, | |
| "learning_rate": 0.00019754297868854073, | |
| "loss": 2.3255, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.08010012515644556, | |
| "grad_norm": 0.4384368062019348, | |
| "learning_rate": 0.00019747901678338496, | |
| "loss": 2.2881, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.08101035385140516, | |
| "grad_norm": 0.5119354724884033, | |
| "learning_rate": 0.00019741424362801452, | |
| "loss": 2.3529, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.08192058254636478, | |
| "grad_norm": 0.4595087766647339, | |
| "learning_rate": 0.00019734865976149145, | |
| "loss": 2.2495, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.08283081124132438, | |
| "grad_norm": 0.4848781228065491, | |
| "learning_rate": 0.00019728226572962473, | |
| "loss": 2.446, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.083741039936284, | |
| "grad_norm": 0.4848371148109436, | |
| "learning_rate": 0.00019721506208496585, | |
| "loss": 2.2737, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.0846512686312436, | |
| "grad_norm": 0.4835449755191803, | |
| "learning_rate": 0.00019714704938680408, | |
| "loss": 2.224, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.0855614973262032, | |
| "grad_norm": 0.48262766003608704, | |
| "learning_rate": 0.00019707822820116193, | |
| "loss": 2.2449, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.08647172602116282, | |
| "grad_norm": 0.49700936675071716, | |
| "learning_rate": 0.00019700859910079036, | |
| "loss": 2.1856, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.08738195471612242, | |
| "grad_norm": 0.5965728163719177, | |
| "learning_rate": 0.00019693816266516407, | |
| "loss": 2.6448, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.08829218341108204, | |
| "grad_norm": 0.6164146661758423, | |
| "learning_rate": 0.00019686691948047664, | |
| "loss": 2.4917, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.08920241210604164, | |
| "grad_norm": 0.6717427372932434, | |
| "learning_rate": 0.00019679487013963564, | |
| "loss": 2.7093, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.09011264080100125, | |
| "grad_norm": 0.808948278427124, | |
| "learning_rate": 0.00019672201524225776, | |
| "loss": 2.766, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.09102286949596086, | |
| "grad_norm": 1.4683064222335815, | |
| "learning_rate": 0.0001966483553946637, | |
| "loss": 2.7288, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09193309819092046, | |
| "grad_norm": 1.0303999185562134, | |
| "learning_rate": 0.00019657389120987333, | |
| "loss": 2.8246, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.09284332688588008, | |
| "grad_norm": 0.49279651045799255, | |
| "learning_rate": 0.00019649862330760036, | |
| "loss": 2.722, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.09375355558083968, | |
| "grad_norm": 0.43273016810417175, | |
| "learning_rate": 0.00019642255231424729, | |
| "loss": 2.5446, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.0946637842757993, | |
| "grad_norm": 0.4434455335140228, | |
| "learning_rate": 0.00019634567886290025, | |
| "loss": 2.4578, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.0955740129707589, | |
| "grad_norm": 0.4245634973049164, | |
| "learning_rate": 0.00019626800359332362, | |
| "loss": 2.5472, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.0964842416657185, | |
| "grad_norm": 0.42879197001457214, | |
| "learning_rate": 0.00019618952715195475, | |
| "loss": 2.5313, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.09739447036067812, | |
| "grad_norm": 0.4644932150840759, | |
| "learning_rate": 0.0001961102501918986, | |
| "loss": 2.3859, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.09830469905563773, | |
| "grad_norm": 0.40692242980003357, | |
| "learning_rate": 0.00019603017337292236, | |
| "loss": 2.433, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.09921492775059734, | |
| "grad_norm": 0.36667487025260925, | |
| "learning_rate": 0.00019594929736144976, | |
| "loss": 2.3352, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.10012515644555695, | |
| "grad_norm": 0.41123342514038086, | |
| "learning_rate": 0.00019586762283055573, | |
| "loss": 2.3494, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.10103538514051655, | |
| "grad_norm": 0.40255314111709595, | |
| "learning_rate": 0.00019578515045996073, | |
| "loss": 2.4524, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.10194561383547616, | |
| "grad_norm": 0.38136208057403564, | |
| "learning_rate": 0.0001957018809360251, | |
| "loss": 2.3937, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.10285584253043577, | |
| "grad_norm": 0.4538716673851013, | |
| "learning_rate": 0.00019561781495174328, | |
| "loss": 2.3144, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.10376607122539538, | |
| "grad_norm": 0.3876875042915344, | |
| "learning_rate": 0.00019553295320673807, | |
| "loss": 2.2868, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.10467629992035499, | |
| "grad_norm": 0.40529122948646545, | |
| "learning_rate": 0.00019544729640725498, | |
| "loss": 2.4219, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.1055865286153146, | |
| "grad_norm": 0.36852964758872986, | |
| "learning_rate": 0.0001953608452661561, | |
| "loss": 2.3485, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.1064967573102742, | |
| "grad_norm": 0.3608396053314209, | |
| "learning_rate": 0.0001952736005029142, | |
| "loss": 2.2813, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.10740698600523381, | |
| "grad_norm": 0.3647065758705139, | |
| "learning_rate": 0.00019518556284360696, | |
| "loss": 2.2284, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.10831721470019343, | |
| "grad_norm": 0.3692014217376709, | |
| "learning_rate": 0.00019509673302091075, | |
| "loss": 2.3005, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.10922744339515303, | |
| "grad_norm": 0.3861401379108429, | |
| "learning_rate": 0.00019500711177409454, | |
| "loss": 2.3023, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.11013767209011265, | |
| "grad_norm": 0.38329821825027466, | |
| "learning_rate": 0.00019491669984901379, | |
| "loss": 2.0692, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.11104790078507225, | |
| "grad_norm": 0.3725563585758209, | |
| "learning_rate": 0.00019482549799810413, | |
| "loss": 2.2495, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.11195812948003186, | |
| "grad_norm": 0.37789279222488403, | |
| "learning_rate": 0.00019473350698037535, | |
| "loss": 2.1242, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.11286835817499147, | |
| "grad_norm": 0.4199522137641907, | |
| "learning_rate": 0.00019464072756140486, | |
| "loss": 2.2501, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.11377858686995107, | |
| "grad_norm": 0.4012223184108734, | |
| "learning_rate": 0.00019454716051333135, | |
| "loss": 2.2263, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.11468881556491069, | |
| "grad_norm": 0.4205365478992462, | |
| "learning_rate": 0.00019445280661484847, | |
| "loss": 2.1445, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.11559904425987029, | |
| "grad_norm": 0.41000694036483765, | |
| "learning_rate": 0.0001943576666511982, | |
| "loss": 2.198, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.1165092729548299, | |
| "grad_norm": 0.43458715081214905, | |
| "learning_rate": 0.00019426174141416448, | |
| "loss": 2.1884, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.11741950164978951, | |
| "grad_norm": 0.4233211874961853, | |
| "learning_rate": 0.00019416503170206645, | |
| "loss": 2.311, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.11832973034474911, | |
| "grad_norm": 0.411350816488266, | |
| "learning_rate": 0.00019406753831975203, | |
| "loss": 2.1085, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.11923995903970873, | |
| "grad_norm": 0.42890501022338867, | |
| "learning_rate": 0.00019396926207859084, | |
| "loss": 2.0548, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.12015018773466833, | |
| "grad_norm": 0.45003724098205566, | |
| "learning_rate": 0.00019387020379646797, | |
| "loss": 2.2063, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.12106041642962795, | |
| "grad_norm": 0.4295034408569336, | |
| "learning_rate": 0.00019377036429777672, | |
| "loss": 2.1306, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.12197064512458755, | |
| "grad_norm": 0.4310653805732727, | |
| "learning_rate": 0.0001936697444134119, | |
| "loss": 2.1307, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.12288087381954717, | |
| "grad_norm": 0.4377802014350891, | |
| "learning_rate": 0.0001935683449807631, | |
| "loss": 2.2982, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.12379110251450677, | |
| "grad_norm": 0.45586827397346497, | |
| "learning_rate": 0.0001934661668437073, | |
| "loss": 2.2334, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.12470133120946637, | |
| "grad_norm": 0.4612938463687897, | |
| "learning_rate": 0.00019336321085260236, | |
| "loss": 2.2488, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.125611559904426, | |
| "grad_norm": 0.48157835006713867, | |
| "learning_rate": 0.00019325947786427952, | |
| "loss": 2.2068, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.1265217885993856, | |
| "grad_norm": 0.46901848912239075, | |
| "learning_rate": 0.0001931549687420364, | |
| "loss": 2.2526, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.1274320172943452, | |
| "grad_norm": 0.4637078046798706, | |
| "learning_rate": 0.00019304968435562993, | |
| "loss": 2.1521, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.12834224598930483, | |
| "grad_norm": 0.5239959955215454, | |
| "learning_rate": 0.00019294362558126905, | |
| "loss": 2.3988, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.12925247468426443, | |
| "grad_norm": 0.462212473154068, | |
| "learning_rate": 0.00019283679330160726, | |
| "loss": 2.3085, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.13016270337922403, | |
| "grad_norm": 0.49486446380615234, | |
| "learning_rate": 0.00019272918840573558, | |
| "loss": 2.2851, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.13107293207418363, | |
| "grad_norm": 0.5234097838401794, | |
| "learning_rate": 0.00019262081178917482, | |
| "loss": 2.3942, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.13198316076914324, | |
| "grad_norm": 0.525196373462677, | |
| "learning_rate": 0.0001925116643538684, | |
| "loss": 2.3057, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.13289338946410287, | |
| "grad_norm": 0.5571113228797913, | |
| "learning_rate": 0.00019240174700817464, | |
| "loss": 2.2406, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.13380361815906247, | |
| "grad_norm": 0.6122002601623535, | |
| "learning_rate": 0.00019229106066685937, | |
| "loss": 2.42, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.13471384685402207, | |
| "grad_norm": 0.6147793531417847, | |
| "learning_rate": 0.0001921796062510882, | |
| "loss": 2.5769, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.13562407554898168, | |
| "grad_norm": 0.7194758653640747, | |
| "learning_rate": 0.0001920673846884189, | |
| "loss": 2.5508, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.13653430424394128, | |
| "grad_norm": 1.1327760219573975, | |
| "learning_rate": 0.00019195439691279363, | |
| "loss": 2.776, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1374445329389009, | |
| "grad_norm": 0.5275815725326538, | |
| "learning_rate": 0.00019184064386453128, | |
| "loss": 2.6188, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.1383547616338605, | |
| "grad_norm": 0.5287826657295227, | |
| "learning_rate": 0.00019172612649031952, | |
| "loss": 2.6061, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.13926499032882012, | |
| "grad_norm": 0.4138477146625519, | |
| "learning_rate": 0.00019161084574320696, | |
| "loss": 2.355, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.14017521902377972, | |
| "grad_norm": 0.4219213128089905, | |
| "learning_rate": 0.00019149480258259533, | |
| "loss": 2.5719, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.14108544771873932, | |
| "grad_norm": 0.39532896876335144, | |
| "learning_rate": 0.00019137799797423126, | |
| "loss": 2.4734, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.14199567641369895, | |
| "grad_norm": 0.4149189889431, | |
| "learning_rate": 0.00019126043289019852, | |
| "loss": 2.5212, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.14290590510865855, | |
| "grad_norm": 0.45758771896362305, | |
| "learning_rate": 0.00019114210830890969, | |
| "loss": 2.5216, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.14381613380361816, | |
| "grad_norm": 0.4397677779197693, | |
| "learning_rate": 0.00019102302521509815, | |
| "loss": 2.6222, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.14472636249857776, | |
| "grad_norm": 0.4018581211566925, | |
| "learning_rate": 0.00019090318459980986, | |
| "loss": 2.546, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.14563659119353736, | |
| "grad_norm": 0.3948480486869812, | |
| "learning_rate": 0.00019078258746039507, | |
| "loss": 2.5202, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.146546819888497, | |
| "grad_norm": 0.37651899456977844, | |
| "learning_rate": 0.00019066123480050015, | |
| "loss": 2.3166, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.1474570485834566, | |
| "grad_norm": 0.3794865310192108, | |
| "learning_rate": 0.00019053912763005907, | |
| "loss": 2.1592, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.1483672772784162, | |
| "grad_norm": 0.39368098974227905, | |
| "learning_rate": 0.00019041626696528503, | |
| "loss": 2.3273, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.1492775059733758, | |
| "grad_norm": 0.39652615785598755, | |
| "learning_rate": 0.00019029265382866214, | |
| "loss": 2.4105, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.15018773466833543, | |
| "grad_norm": 0.4029604196548462, | |
| "learning_rate": 0.0001901682892489367, | |
| "loss": 2.3814, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.15109796336329503, | |
| "grad_norm": 0.40106695890426636, | |
| "learning_rate": 0.0001900431742611089, | |
| "loss": 2.3963, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.15200819205825464, | |
| "grad_norm": 0.3830929696559906, | |
| "learning_rate": 0.00018991730990642388, | |
| "loss": 2.2472, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.15291842075321424, | |
| "grad_norm": 0.36839723587036133, | |
| "learning_rate": 0.00018979069723236333, | |
| "loss": 2.2463, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.15382864944817384, | |
| "grad_norm": 0.36918777227401733, | |
| "learning_rate": 0.00018966333729263674, | |
| "loss": 2.2286, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.15473887814313347, | |
| "grad_norm": 0.37009552121162415, | |
| "learning_rate": 0.00018953523114717245, | |
| "loss": 2.2823, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.15564910683809308, | |
| "grad_norm": 0.38302165269851685, | |
| "learning_rate": 0.00018940637986210906, | |
| "loss": 2.1199, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.15655933553305268, | |
| "grad_norm": 0.40206897258758545, | |
| "learning_rate": 0.0001892767845097864, | |
| "loss": 2.3254, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.15746956422801228, | |
| "grad_norm": 0.3847138285636902, | |
| "learning_rate": 0.00018914644616873657, | |
| "loss": 2.0739, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.15837979292297188, | |
| "grad_norm": 0.37638360261917114, | |
| "learning_rate": 0.0001890153659236753, | |
| "loss": 2.1024, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.15929002161793152, | |
| "grad_norm": 0.38801488280296326, | |
| "learning_rate": 0.00018888354486549237, | |
| "loss": 2.1634, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.16020025031289112, | |
| "grad_norm": 0.4164554178714752, | |
| "learning_rate": 0.00018875098409124302, | |
| "loss": 2.2987, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.16111047900785072, | |
| "grad_norm": 0.38982319831848145, | |
| "learning_rate": 0.0001886176847041386, | |
| "loss": 2.1834, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.16202070770281032, | |
| "grad_norm": 0.42200496792793274, | |
| "learning_rate": 0.00018848364781353744, | |
| "loss": 2.2793, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.16293093639776993, | |
| "grad_norm": 0.4152745306491852, | |
| "learning_rate": 0.0001883488745349355, | |
| "loss": 2.1988, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.16384116509272956, | |
| "grad_norm": 0.41070255637168884, | |
| "learning_rate": 0.0001882133659899573, | |
| "loss": 1.8985, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.16475139378768916, | |
| "grad_norm": 0.4259546399116516, | |
| "learning_rate": 0.00018807712330634642, | |
| "loss": 2.4285, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.16566162248264876, | |
| "grad_norm": 0.44458258152008057, | |
| "learning_rate": 0.0001879401476179562, | |
| "loss": 2.2561, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.16657185117760837, | |
| "grad_norm": 0.4640187621116638, | |
| "learning_rate": 0.0001878024400647402, | |
| "loss": 2.4091, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.167482079872568, | |
| "grad_norm": 0.40832704305648804, | |
| "learning_rate": 0.00018766400179274286, | |
| "loss": 1.9156, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.1683923085675276, | |
| "grad_norm": 0.4358328878879547, | |
| "learning_rate": 0.00018752483395408987, | |
| "loss": 2.0592, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.1693025372624872, | |
| "grad_norm": 0.45089927315711975, | |
| "learning_rate": 0.00018738493770697852, | |
| "loss": 2.2189, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.1702127659574468, | |
| "grad_norm": 0.45900994539260864, | |
| "learning_rate": 0.00018724431421566823, | |
| "loss": 2.3023, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.1711229946524064, | |
| "grad_norm": 0.4608750641345978, | |
| "learning_rate": 0.00018710296465047075, | |
| "loss": 2.2467, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.17203322334736604, | |
| "grad_norm": 0.44717609882354736, | |
| "learning_rate": 0.0001869608901877404, | |
| "loss": 2.0313, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.17294345204232564, | |
| "grad_norm": 0.40987592935562134, | |
| "learning_rate": 0.0001868180920098644, | |
| "loss": 2.0965, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.17385368073728524, | |
| "grad_norm": 0.502581775188446, | |
| "learning_rate": 0.00018667457130525284, | |
| "loss": 2.1991, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.17476390943224485, | |
| "grad_norm": 0.4770600199699402, | |
| "learning_rate": 0.00018653032926832896, | |
| "loss": 2.2258, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.17567413812720445, | |
| "grad_norm": 0.48825547099113464, | |
| "learning_rate": 0.00018638536709951917, | |
| "loss": 2.2884, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.17658436682216408, | |
| "grad_norm": 0.5004613995552063, | |
| "learning_rate": 0.000186239686005243, | |
| "loss": 2.2927, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.17749459551712368, | |
| "grad_norm": 0.5268414616584778, | |
| "learning_rate": 0.0001860932871979031, | |
| "loss": 2.356, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.17840482421208329, | |
| "grad_norm": 0.5261876583099365, | |
| "learning_rate": 0.00018594617189587512, | |
| "loss": 2.2864, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.1793150529070429, | |
| "grad_norm": 0.6071822643280029, | |
| "learning_rate": 0.00018579834132349772, | |
| "loss": 2.684, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.1802252816020025, | |
| "grad_norm": 0.659428060054779, | |
| "learning_rate": 0.0001856497967110621, | |
| "loss": 2.5811, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.18113551029696212, | |
| "grad_norm": 0.828525960445404, | |
| "learning_rate": 0.00018550053929480202, | |
| "loss": 2.7052, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.18204573899192172, | |
| "grad_norm": 1.9569804668426514, | |
| "learning_rate": 0.00018535057031688335, | |
| "loss": 2.9428, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.18295596768688133, | |
| "grad_norm": 0.6459002494812012, | |
| "learning_rate": 0.0001851998910253939, | |
| "loss": 2.6936, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.18386619638184093, | |
| "grad_norm": 0.6431388854980469, | |
| "learning_rate": 0.0001850485026743328, | |
| "loss": 2.5446, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.18477642507680053, | |
| "grad_norm": 0.5401524305343628, | |
| "learning_rate": 0.00018489640652360022, | |
| "loss": 2.4508, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.18568665377176016, | |
| "grad_norm": 0.4280170500278473, | |
| "learning_rate": 0.00018474360383898694, | |
| "loss": 2.5167, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.18659688246671977, | |
| "grad_norm": 0.4014836847782135, | |
| "learning_rate": 0.00018459009589216364, | |
| "loss": 2.4456, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.18750711116167937, | |
| "grad_norm": 0.41716477274894714, | |
| "learning_rate": 0.0001844358839606705, | |
| "loss": 2.4357, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.18841733985663897, | |
| "grad_norm": 0.40400266647338867, | |
| "learning_rate": 0.00018428096932790632, | |
| "loss": 2.5122, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.1893275685515986, | |
| "grad_norm": 0.4276406764984131, | |
| "learning_rate": 0.00018412535328311814, | |
| "loss": 2.3581, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.1902377972465582, | |
| "grad_norm": 0.42548567056655884, | |
| "learning_rate": 0.0001839690371213903, | |
| "loss": 2.4639, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.1911480259415178, | |
| "grad_norm": 0.3912813365459442, | |
| "learning_rate": 0.0001838120221436338, | |
| "loss": 2.4395, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.1920582546364774, | |
| "grad_norm": 0.40997177362442017, | |
| "learning_rate": 0.00018365430965657526, | |
| "loss": 2.4833, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.192968483331437, | |
| "grad_norm": 0.3765038847923279, | |
| "learning_rate": 0.00018349590097274632, | |
| "loss": 2.2293, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.19387871202639664, | |
| "grad_norm": 0.3960655629634857, | |
| "learning_rate": 0.00018333679741047254, | |
| "loss": 2.2261, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.19478894072135625, | |
| "grad_norm": 0.3758634030818939, | |
| "learning_rate": 0.00018317700029386245, | |
| "loss": 2.0951, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.19569916941631585, | |
| "grad_norm": 0.4159095883369446, | |
| "learning_rate": 0.00018301651095279655, | |
| "loss": 2.3568, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.19660939811127545, | |
| "grad_norm": 0.4104447662830353, | |
| "learning_rate": 0.0001828553307229163, | |
| "loss": 2.2262, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.19751962680623505, | |
| "grad_norm": 0.3676728308200836, | |
| "learning_rate": 0.0001826934609456129, | |
| "loss": 2.2557, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.19842985550119469, | |
| "grad_norm": 0.39037179946899414, | |
| "learning_rate": 0.00018253090296801614, | |
| "loss": 2.2632, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.1993400841961543, | |
| "grad_norm": 0.38477587699890137, | |
| "learning_rate": 0.0001823676581429833, | |
| "loss": 2.1344, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.2002503128911139, | |
| "grad_norm": 0.38563740253448486, | |
| "learning_rate": 0.00018220372782908777, | |
| "loss": 2.2298, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.2011605415860735, | |
| "grad_norm": 0.38280826807022095, | |
| "learning_rate": 0.00018203911339060783, | |
| "loss": 2.0098, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.2020707702810331, | |
| "grad_norm": 0.4058510661125183, | |
| "learning_rate": 0.00018187381619751516, | |
| "loss": 2.2295, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.20298099897599273, | |
| "grad_norm": 0.42094549536705017, | |
| "learning_rate": 0.00018170783762546365, | |
| "loss": 2.1706, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.20389122767095233, | |
| "grad_norm": 0.42036664485931396, | |
| "learning_rate": 0.00018154117905577776, | |
| "loss": 2.1802, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.20480145636591193, | |
| "grad_norm": 0.39867401123046875, | |
| "learning_rate": 0.00018137384187544116, | |
| "loss": 2.165, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.20571168506087154, | |
| "grad_norm": 0.3934602737426758, | |
| "learning_rate": 0.00018120582747708502, | |
| "loss": 1.9683, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.20662191375583117, | |
| "grad_norm": 0.3914499282836914, | |
| "learning_rate": 0.0001810371372589766, | |
| "loss": 1.9805, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.20753214245079077, | |
| "grad_norm": 0.4222523868083954, | |
| "learning_rate": 0.0001808677726250076, | |
| "loss": 2.185, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.20844237114575037, | |
| "grad_norm": 0.3909006118774414, | |
| "learning_rate": 0.00018069773498468223, | |
| "loss": 2.0038, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.20935259984070997, | |
| "grad_norm": 0.42738041281700134, | |
| "learning_rate": 0.00018052702575310588, | |
| "loss": 2.2352, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.21026282853566958, | |
| "grad_norm": 0.4195539653301239, | |
| "learning_rate": 0.00018035564635097298, | |
| "loss": 2.0254, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.2111730572306292, | |
| "grad_norm": 0.4464121460914612, | |
| "learning_rate": 0.00018018359820455536, | |
| "loss": 2.2766, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.2120832859255888, | |
| "grad_norm": 0.42468565702438354, | |
| "learning_rate": 0.00018001088274569038, | |
| "loss": 2.1718, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.2129935146205484, | |
| "grad_norm": 0.41730690002441406, | |
| "learning_rate": 0.00017983750141176895, | |
| "loss": 2.0959, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.21390374331550802, | |
| "grad_norm": 0.4425525665283203, | |
| "learning_rate": 0.0001796634556457236, | |
| "loss": 2.1538, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.21481397201046762, | |
| "grad_norm": 0.4559743106365204, | |
| "learning_rate": 0.0001794887468960165, | |
| "loss": 2.1299, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.21572420070542725, | |
| "grad_norm": 0.43843337893486023, | |
| "learning_rate": 0.00017931337661662727, | |
| "loss": 2.1172, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.21663442940038685, | |
| "grad_norm": 0.46661749482154846, | |
| "learning_rate": 0.0001791373462670411, | |
| "loss": 2.1352, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.21754465809534645, | |
| "grad_norm": 0.4997507333755493, | |
| "learning_rate": 0.00017896065731223644, | |
| "loss": 2.1185, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.21845488679030606, | |
| "grad_norm": 0.480398029088974, | |
| "learning_rate": 0.00017878331122267284, | |
| "loss": 2.3208, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.21936511548526566, | |
| "grad_norm": 0.4819256365299225, | |
| "learning_rate": 0.00017860530947427875, | |
| "loss": 2.3114, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.2202753441802253, | |
| "grad_norm": 0.5086483359336853, | |
| "learning_rate": 0.00017842665354843922, | |
| "loss": 2.1578, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.2211855728751849, | |
| "grad_norm": 0.5267627239227295, | |
| "learning_rate": 0.0001782473449319835, | |
| "loss": 2.3594, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.2220958015701445, | |
| "grad_norm": 0.5265729427337646, | |
| "learning_rate": 0.0001780673851171728, | |
| "loss": 2.2083, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.2230060302651041, | |
| "grad_norm": 0.5762265920639038, | |
| "learning_rate": 0.00017788677560168784, | |
| "loss": 2.4887, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.22391625896006373, | |
| "grad_norm": 0.5499774217605591, | |
| "learning_rate": 0.0001777055178886162, | |
| "loss": 2.4169, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.22482648765502333, | |
| "grad_norm": 0.613876223564148, | |
| "learning_rate": 0.0001775236134864401, | |
| "loss": 2.5177, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.22573671634998294, | |
| "grad_norm": 0.6267974376678467, | |
| "learning_rate": 0.00017734106390902366, | |
| "loss": 2.5212, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.22664694504494254, | |
| "grad_norm": 0.7940219640731812, | |
| "learning_rate": 0.0001771578706756003, | |
| "loss": 2.5769, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.22755717373990214, | |
| "grad_norm": 1.3050237894058228, | |
| "learning_rate": 0.0001769740353107602, | |
| "loss": 2.3235, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.22846740243486177, | |
| "grad_norm": 0.5289978981018066, | |
| "learning_rate": 0.00017678955934443758, | |
| "loss": 2.5637, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.22937763112982137, | |
| "grad_norm": 0.5574918389320374, | |
| "learning_rate": 0.0001766044443118978, | |
| "loss": 2.5542, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.23028785982478098, | |
| "grad_norm": 0.47572481632232666, | |
| "learning_rate": 0.00017641869175372493, | |
| "loss": 2.4325, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.23119808851974058, | |
| "grad_norm": 0.41480204463005066, | |
| "learning_rate": 0.00017623230321580854, | |
| "loss": 2.5176, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.23210831721470018, | |
| "grad_norm": 0.4072042405605316, | |
| "learning_rate": 0.00017604528024933115, | |
| "loss": 2.4798, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.2330185459096598, | |
| "grad_norm": 0.3996100127696991, | |
| "learning_rate": 0.00017585762441075503, | |
| "loss": 2.35, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.23392877460461942, | |
| "grad_norm": 0.4213111698627472, | |
| "learning_rate": 0.00017566933726180964, | |
| "loss": 2.526, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.23483900329957902, | |
| "grad_norm": 0.42595675587654114, | |
| "learning_rate": 0.0001754804203694782, | |
| "loss": 2.4724, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.23574923199453862, | |
| "grad_norm": 0.407141774892807, | |
| "learning_rate": 0.0001752908753059849, | |
| "loss": 2.3349, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.23665946068949822, | |
| "grad_norm": 0.4092378616333008, | |
| "learning_rate": 0.00017510070364878177, | |
| "loss": 2.4487, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.23756968938445785, | |
| "grad_norm": 0.38301384449005127, | |
| "learning_rate": 0.00017490990698053563, | |
| "loss": 2.3245, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.23847991807941746, | |
| "grad_norm": 0.3946523666381836, | |
| "learning_rate": 0.00017471848688911464, | |
| "loss": 2.3614, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.23939014677437706, | |
| "grad_norm": 0.41228440403938293, | |
| "learning_rate": 0.0001745264449675755, | |
| "loss": 2.3525, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.24030037546933666, | |
| "grad_norm": 0.39352720975875854, | |
| "learning_rate": 0.00017433378281414975, | |
| "loss": 2.3851, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.24121060416429627, | |
| "grad_norm": 0.39706528186798096, | |
| "learning_rate": 0.0001741405020322309, | |
| "loss": 2.417, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.2421208328592559, | |
| "grad_norm": 0.3801213204860687, | |
| "learning_rate": 0.00017394660423036075, | |
| "loss": 2.1036, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.2430310615542155, | |
| "grad_norm": 0.37401556968688965, | |
| "learning_rate": 0.00017375209102221613, | |
| "loss": 2.1668, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.2439412902491751, | |
| "grad_norm": 0.3884643614292145, | |
| "learning_rate": 0.00017355696402659548, | |
| "loss": 2.3057, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.2448515189441347, | |
| "grad_norm": 0.38754650950431824, | |
| "learning_rate": 0.00017336122486740548, | |
| "loss": 2.1363, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.24576174763909434, | |
| "grad_norm": 0.3861556053161621, | |
| "learning_rate": 0.00017316487517364721, | |
| "loss": 2.2378, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.24667197633405394, | |
| "grad_norm": 0.3905907869338989, | |
| "learning_rate": 0.000172967916579403, | |
| "loss": 2.2011, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.24758220502901354, | |
| "grad_norm": 0.40004828572273254, | |
| "learning_rate": 0.00017277035072382253, | |
| "loss": 2.1316, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.24849243372397314, | |
| "grad_norm": 0.39059486985206604, | |
| "learning_rate": 0.00017257217925110933, | |
| "loss": 2.1582, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.24940266241893275, | |
| "grad_norm": 0.3994196355342865, | |
| "learning_rate": 0.00017237340381050703, | |
| "loss": 2.0078, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.2503128911138924, | |
| "grad_norm": 0.41038015484809875, | |
| "learning_rate": 0.00017217402605628572, | |
| "loss": 2.0799, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.2503128911138924, | |
| "eval_loss": 2.2957444190979004, | |
| "eval_runtime": 203.8901, | |
| "eval_samples_per_second": 9.078, | |
| "eval_steps_per_second": 4.542, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.251223119808852, | |
| "grad_norm": 0.39477062225341797, | |
| "learning_rate": 0.00017197404764772805, | |
| "loss": 2.1965, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.2521333485038116, | |
| "grad_norm": 0.4311797320842743, | |
| "learning_rate": 0.00017177347024911562, | |
| "loss": 2.2732, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.2530435771987712, | |
| "grad_norm": 0.3932170867919922, | |
| "learning_rate": 0.00017157229552971487, | |
| "loss": 2.1883, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.2539538058937308, | |
| "grad_norm": 0.40245768427848816, | |
| "learning_rate": 0.00017137052516376345, | |
| "loss": 2.1204, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.2548640345886904, | |
| "grad_norm": 0.44616052508354187, | |
| "learning_rate": 0.00017116816083045602, | |
| "loss": 2.359, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.25577426328365, | |
| "grad_norm": 0.4077576994895935, | |
| "learning_rate": 0.0001709652042139306, | |
| "loss": 2.0829, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.25668449197860965, | |
| "grad_norm": 0.40657973289489746, | |
| "learning_rate": 0.0001707616570032542, | |
| "loss": 2.1668, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.25759472067356926, | |
| "grad_norm": 0.4176836907863617, | |
| "learning_rate": 0.00017055752089240907, | |
| "loss": 2.1375, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.25850494936852886, | |
| "grad_norm": 0.4195888936519623, | |
| "learning_rate": 0.00017035279758027832, | |
| "loss": 2.0588, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.25941517806348846, | |
| "grad_norm": 0.42504483461380005, | |
| "learning_rate": 0.00017014748877063214, | |
| "loss": 2.0153, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.26032540675844806, | |
| "grad_norm": 0.45724233984947205, | |
| "learning_rate": 0.00016994159617211317, | |
| "loss": 2.1907, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.26123563545340767, | |
| "grad_norm": 0.4327910542488098, | |
| "learning_rate": 0.00016973512149822274, | |
| "loss": 2.0613, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.26214586414836727, | |
| "grad_norm": 0.446054607629776, | |
| "learning_rate": 0.0001695280664673062, | |
| "loss": 2.1539, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.26305609284332687, | |
| "grad_norm": 0.4568957984447479, | |
| "learning_rate": 0.0001693204328025389, | |
| "loss": 2.2892, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.2639663215382865, | |
| "grad_norm": 0.4926307201385498, | |
| "learning_rate": 0.00016911222223191182, | |
| "loss": 2.253, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.2648765502332461, | |
| "grad_norm": 0.49324727058410645, | |
| "learning_rate": 0.00016890343648821697, | |
| "loss": 2.2756, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.26578677892820574, | |
| "grad_norm": 0.478746235370636, | |
| "learning_rate": 0.0001686940773090333, | |
| "loss": 2.2775, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.26669700762316534, | |
| "grad_norm": 0.508730411529541, | |
| "learning_rate": 0.00016848414643671195, | |
| "loss": 2.3255, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.26760723631812494, | |
| "grad_norm": 0.5452204942703247, | |
| "learning_rate": 0.00016827364561836187, | |
| "loss": 2.4101, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.26851746501308454, | |
| "grad_norm": 0.5368967652320862, | |
| "learning_rate": 0.00016806257660583534, | |
| "loss": 2.3823, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.26942769370804415, | |
| "grad_norm": 0.565154492855072, | |
| "learning_rate": 0.00016785094115571322, | |
| "loss": 2.376, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.27033792240300375, | |
| "grad_norm": 0.623928964138031, | |
| "learning_rate": 0.0001676387410292906, | |
| "loss": 2.4355, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.27124815109796335, | |
| "grad_norm": 0.7393582463264465, | |
| "learning_rate": 0.00016742597799256182, | |
| "loss": 2.699, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.27215837979292296, | |
| "grad_norm": 0.845248818397522, | |
| "learning_rate": 0.000167212653816206, | |
| "loss": 2.6997, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.27306860848788256, | |
| "grad_norm": 1.6121950149536133, | |
| "learning_rate": 0.00016699877027557226, | |
| "loss": 2.7515, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2739788371828422, | |
| "grad_norm": 0.4747787117958069, | |
| "learning_rate": 0.00016678432915066488, | |
| "loss": 2.5881, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.2748890658778018, | |
| "grad_norm": 0.46245065331459045, | |
| "learning_rate": 0.00016656933222612854, | |
| "loss": 2.4625, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.2757992945727614, | |
| "grad_norm": 0.5154780745506287, | |
| "learning_rate": 0.00016635378129123342, | |
| "loss": 2.4195, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.276709523267721, | |
| "grad_norm": 0.46146687865257263, | |
| "learning_rate": 0.00016613767813986044, | |
| "loss": 2.4896, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.2776197519626806, | |
| "grad_norm": 0.42735326290130615, | |
| "learning_rate": 0.0001659210245704861, | |
| "loss": 2.4185, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.27852998065764023, | |
| "grad_norm": 0.40846991539001465, | |
| "learning_rate": 0.00016570382238616777, | |
| "loss": 2.4187, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.27944020935259983, | |
| "grad_norm": 0.3983679711818695, | |
| "learning_rate": 0.00016548607339452853, | |
| "loss": 2.3726, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.28035043804755944, | |
| "grad_norm": 0.4192310869693756, | |
| "learning_rate": 0.00016526777940774204, | |
| "loss": 2.3308, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.28126066674251904, | |
| "grad_norm": 0.4130899906158447, | |
| "learning_rate": 0.00016504894224251778, | |
| "loss": 2.3803, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.28217089543747864, | |
| "grad_norm": 0.3945166766643524, | |
| "learning_rate": 0.0001648295637200856, | |
| "loss": 2.2948, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.2830811241324383, | |
| "grad_norm": 0.4377036392688751, | |
| "learning_rate": 0.0001646096456661807, | |
| "loss": 2.376, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.2839913528273979, | |
| "grad_norm": 0.4076646864414215, | |
| "learning_rate": 0.00016438918991102842, | |
| "loss": 2.202, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.2849015815223575, | |
| "grad_norm": 0.410691499710083, | |
| "learning_rate": 0.000164168198289329, | |
| "loss": 2.3081, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.2858118102173171, | |
| "grad_norm": 0.40779241919517517, | |
| "learning_rate": 0.00016394667264024246, | |
| "loss": 2.3292, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.2867220389122767, | |
| "grad_norm": 0.3915857970714569, | |
| "learning_rate": 0.00016372461480737297, | |
| "loss": 2.313, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.2876322676072363, | |
| "grad_norm": 0.40646892786026, | |
| "learning_rate": 0.00016350202663875386, | |
| "loss": 1.9998, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.2885424963021959, | |
| "grad_norm": 0.3695480525493622, | |
| "learning_rate": 0.00016327890998683192, | |
| "loss": 2.0481, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.2894527249971555, | |
| "grad_norm": 0.37870198488235474, | |
| "learning_rate": 0.00016305526670845226, | |
| "loss": 2.1776, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.2903629536921151, | |
| "grad_norm": 0.3980102837085724, | |
| "learning_rate": 0.0001628310986648427, | |
| "loss": 2.21, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.2912731823870747, | |
| "grad_norm": 0.42144468426704407, | |
| "learning_rate": 0.0001626064077215983, | |
| "loss": 2.3021, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.2921834110820344, | |
| "grad_norm": 0.39952829480171204, | |
| "learning_rate": 0.00016238119574866588, | |
| "loss": 2.1492, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.293093639776994, | |
| "grad_norm": 0.42757153511047363, | |
| "learning_rate": 0.0001621554646203284, | |
| "loss": 1.8572, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.2940038684719536, | |
| "grad_norm": 0.418364018201828, | |
| "learning_rate": 0.00016192921621518944, | |
| "loss": 2.1835, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.2949140971669132, | |
| "grad_norm": 0.41788679361343384, | |
| "learning_rate": 0.0001617024524161574, | |
| "loss": 2.2657, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.2958243258618728, | |
| "grad_norm": 0.44890159368515015, | |
| "learning_rate": 0.0001614751751104301, | |
| "loss": 2.2471, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.2967345545568324, | |
| "grad_norm": 0.4491661787033081, | |
| "learning_rate": 0.0001612473861894788, | |
| "loss": 1.971, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.297644783251792, | |
| "grad_norm": 0.3881804943084717, | |
| "learning_rate": 0.00016101908754903268, | |
| "loss": 2.0647, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.2985550119467516, | |
| "grad_norm": 0.42832571268081665, | |
| "learning_rate": 0.00016079028108906282, | |
| "loss": 2.1406, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.2994652406417112, | |
| "grad_norm": 0.4422348439693451, | |
| "learning_rate": 0.00016056096871376667, | |
| "loss": 2.0549, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.30037546933667086, | |
| "grad_norm": 0.4560178220272064, | |
| "learning_rate": 0.00016033115233155202, | |
| "loss": 2.2055, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.30128569803163047, | |
| "grad_norm": 0.4027805030345917, | |
| "learning_rate": 0.0001601008338550211, | |
| "loss": 2.0151, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.30219592672659007, | |
| "grad_norm": 0.4404117465019226, | |
| "learning_rate": 0.00015987001520095478, | |
| "loss": 2.1838, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.3031061554215497, | |
| "grad_norm": 0.47021403908729553, | |
| "learning_rate": 0.00015963869829029658, | |
| "loss": 2.1402, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.3040163841165093, | |
| "grad_norm": 0.44921791553497314, | |
| "learning_rate": 0.00015940688504813662, | |
| "loss": 2.2947, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.3049266128114689, | |
| "grad_norm": 0.4406517744064331, | |
| "learning_rate": 0.00015917457740369565, | |
| "loss": 2.142, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.3058368415064285, | |
| "grad_norm": 0.43675345182418823, | |
| "learning_rate": 0.000158941777290309, | |
| "loss": 2.0953, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.3067470702013881, | |
| "grad_norm": 0.4730078876018524, | |
| "learning_rate": 0.00015870848664541044, | |
| "loss": 2.2471, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.3076572988963477, | |
| "grad_norm": 0.43952468037605286, | |
| "learning_rate": 0.00015847470741051618, | |
| "loss": 2.1458, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.3085675275913073, | |
| "grad_norm": 0.4473972022533417, | |
| "learning_rate": 0.00015824044153120852, | |
| "loss": 2.1065, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.30947775628626695, | |
| "grad_norm": 0.49501466751098633, | |
| "learning_rate": 0.00015800569095711982, | |
| "loss": 2.1574, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.31038798498122655, | |
| "grad_norm": 0.4756048023700714, | |
| "learning_rate": 0.00015777045764191625, | |
| "loss": 2.019, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.31129821367618615, | |
| "grad_norm": 0.491579532623291, | |
| "learning_rate": 0.00015753474354328142, | |
| "loss": 2.2317, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.31220844237114576, | |
| "grad_norm": 0.5067722797393799, | |
| "learning_rate": 0.00015729855062290022, | |
| "loss": 2.4042, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.31311867106610536, | |
| "grad_norm": 0.5405014157295227, | |
| "learning_rate": 0.00015706188084644242, | |
| "loss": 2.243, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.31402889976106496, | |
| "grad_norm": 0.536719799041748, | |
| "learning_rate": 0.00015682473618354635, | |
| "loss": 2.2665, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.31493912845602456, | |
| "grad_norm": 0.5940792560577393, | |
| "learning_rate": 0.0001565871186078025, | |
| "loss": 2.4295, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.31584935715098417, | |
| "grad_norm": 0.6351253986358643, | |
| "learning_rate": 0.00015634903009673705, | |
| "loss": 2.5128, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.31675958584594377, | |
| "grad_norm": 0.7466199398040771, | |
| "learning_rate": 0.00015611047263179548, | |
| "loss": 2.56, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.31766981454090343, | |
| "grad_norm": 0.84588623046875, | |
| "learning_rate": 0.000155871448198326, | |
| "loss": 2.6547, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.31858004323586303, | |
| "grad_norm": 1.7304902076721191, | |
| "learning_rate": 0.0001556319587855631, | |
| "loss": 2.7543, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.31949027193082263, | |
| "grad_norm": 0.5234479308128357, | |
| "learning_rate": 0.00015539200638661104, | |
| "loss": 2.6642, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.32040050062578224, | |
| "grad_norm": 0.5147595405578613, | |
| "learning_rate": 0.00015515159299842707, | |
| "loss": 2.4943, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.32131072932074184, | |
| "grad_norm": 0.4990614950656891, | |
| "learning_rate": 0.00015491072062180503, | |
| "loss": 2.4958, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.32222095801570144, | |
| "grad_norm": 0.45430535078048706, | |
| "learning_rate": 0.00015466939126135856, | |
| "loss": 2.4471, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.32313118671066104, | |
| "grad_norm": 0.44151967763900757, | |
| "learning_rate": 0.00015442760692550443, | |
| "loss": 2.2696, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.32404141540562065, | |
| "grad_norm": 0.4359540641307831, | |
| "learning_rate": 0.00015418536962644592, | |
| "loss": 2.4827, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.32495164410058025, | |
| "grad_norm": 0.4277969002723694, | |
| "learning_rate": 0.00015394268138015598, | |
| "loss": 2.4197, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.32586187279553985, | |
| "grad_norm": 0.42158129811286926, | |
| "learning_rate": 0.00015369954420636048, | |
| "loss": 2.3935, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.3267721014904995, | |
| "grad_norm": 0.4236292243003845, | |
| "learning_rate": 0.00015345596012852138, | |
| "loss": 2.3498, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.3276823301854591, | |
| "grad_norm": 0.39974719285964966, | |
| "learning_rate": 0.00015321193117381996, | |
| "loss": 2.2955, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3285925588804187, | |
| "grad_norm": 0.39424553513526917, | |
| "learning_rate": 0.00015296745937313987, | |
| "loss": 2.2767, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.3295027875753783, | |
| "grad_norm": 0.4082043170928955, | |
| "learning_rate": 0.00015272254676105025, | |
| "loss": 2.2224, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.3304130162703379, | |
| "grad_norm": 0.3964308798313141, | |
| "learning_rate": 0.00015247719537578883, | |
| "loss": 2.2479, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.3313232449652975, | |
| "grad_norm": 0.40646740794181824, | |
| "learning_rate": 0.00015223140725924495, | |
| "loss": 2.3327, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.33223347366025713, | |
| "grad_norm": 0.4101439118385315, | |
| "learning_rate": 0.00015198518445694255, | |
| "loss": 2.4099, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.33314370235521673, | |
| "grad_norm": 0.400878369808197, | |
| "learning_rate": 0.0001517385290180231, | |
| "loss": 2.272, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.33405393105017633, | |
| "grad_norm": 0.37038350105285645, | |
| "learning_rate": 0.00015149144299522873, | |
| "loss": 2.1738, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.334964159745136, | |
| "grad_norm": 0.37709566950798035, | |
| "learning_rate": 0.0001512439284448849, | |
| "loss": 2.1423, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.3358743884400956, | |
| "grad_norm": 0.40142226219177246, | |
| "learning_rate": 0.0001509959874268835, | |
| "loss": 2.2503, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.3367846171350552, | |
| "grad_norm": 0.38601839542388916, | |
| "learning_rate": 0.00015074762200466556, | |
| "loss": 2.15, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3376948458300148, | |
| "grad_norm": 0.39862939715385437, | |
| "learning_rate": 0.00015049883424520414, | |
| "loss": 2.2177, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.3386050745249744, | |
| "grad_norm": 0.40157076716423035, | |
| "learning_rate": 0.00015024962621898715, | |
| "loss": 2.2062, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.339515303219934, | |
| "grad_norm": 0.38866209983825684, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 2.1295, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.3404255319148936, | |
| "grad_norm": 0.40278226137161255, | |
| "learning_rate": 0.00014974995766570855, | |
| "loss": 2.1402, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.3413357606098532, | |
| "grad_norm": 0.41342514753341675, | |
| "learning_rate": 0.00014949950129704162, | |
| "loss": 2.1778, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.3422459893048128, | |
| "grad_norm": 0.4095761775970459, | |
| "learning_rate": 0.00014924863297837378, | |
| "loss": 2.0599, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.3431562179997724, | |
| "grad_norm": 0.4194605350494385, | |
| "learning_rate": 0.00014899735479750794, | |
| "loss": 2.2467, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.3440664466947321, | |
| "grad_norm": 0.41888725757598877, | |
| "learning_rate": 0.00014874566884565807, | |
| "loss": 2.0157, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.3449766753896917, | |
| "grad_norm": 0.41246896982192993, | |
| "learning_rate": 0.00014849357721743168, | |
| "loss": 2.1508, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.3458869040846513, | |
| "grad_norm": 0.4054362177848816, | |
| "learning_rate": 0.00014824108201081247, | |
| "loss": 2.0896, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.3467971327796109, | |
| "grad_norm": 0.4158082902431488, | |
| "learning_rate": 0.00014798818532714279, | |
| "loss": 2.0494, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.3477073614745705, | |
| "grad_norm": 0.43001338839530945, | |
| "learning_rate": 0.00014773488927110633, | |
| "loss": 2.1462, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.3486175901695301, | |
| "grad_norm": 0.4192085862159729, | |
| "learning_rate": 0.00014748119595071034, | |
| "loss": 1.9395, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.3495278188644897, | |
| "grad_norm": 0.45981523394584656, | |
| "learning_rate": 0.0001472271074772683, | |
| "loss": 2.2445, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.3504380475594493, | |
| "grad_norm": 0.4254681169986725, | |
| "learning_rate": 0.00014697262596538227, | |
| "loss": 2.2156, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.3513482762544089, | |
| "grad_norm": 0.4846554100513458, | |
| "learning_rate": 0.00014671775353292525, | |
| "loss": 2.1879, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.35225850494936856, | |
| "grad_norm": 0.4701845645904541, | |
| "learning_rate": 0.00014646249230102366, | |
| "loss": 2.2736, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.35316873364432816, | |
| "grad_norm": 0.45750224590301514, | |
| "learning_rate": 0.00014620684439403962, | |
| "loss": 2.2816, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.35407896233928776, | |
| "grad_norm": 0.4719528555870056, | |
| "learning_rate": 0.00014595081193955324, | |
| "loss": 2.1235, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.35498919103424736, | |
| "grad_norm": 0.49415460228919983, | |
| "learning_rate": 0.000145694397068345, | |
| "loss": 2.1539, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.35589941972920697, | |
| "grad_norm": 0.5116055011749268, | |
| "learning_rate": 0.0001454376019143779, | |
| "loss": 2.1482, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.35680964842416657, | |
| "grad_norm": 0.4707508087158203, | |
| "learning_rate": 0.00014518042861477986, | |
| "loss": 2.1791, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.3577198771191262, | |
| "grad_norm": 0.47449609637260437, | |
| "learning_rate": 0.00014492287930982576, | |
| "loss": 2.1772, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.3586301058140858, | |
| "grad_norm": 0.5449949502944946, | |
| "learning_rate": 0.00014466495614291977, | |
| "loss": 2.454, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.3595403345090454, | |
| "grad_norm": 0.5154662728309631, | |
| "learning_rate": 0.00014440666126057744, | |
| "loss": 2.372, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.360450563204005, | |
| "grad_norm": 0.5283260941505432, | |
| "learning_rate": 0.0001441479968124078, | |
| "loss": 2.3964, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.36136079189896464, | |
| "grad_norm": 0.5910937190055847, | |
| "learning_rate": 0.0001438889649510956, | |
| "loss": 2.5046, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.36227102059392424, | |
| "grad_norm": 0.6620859503746033, | |
| "learning_rate": 0.00014362956783238324, | |
| "loss": 2.642, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.36318124928888385, | |
| "grad_norm": 0.7961871027946472, | |
| "learning_rate": 0.00014336980761505297, | |
| "loss": 2.6575, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.36409147798384345, | |
| "grad_norm": 1.4348353147506714, | |
| "learning_rate": 0.00014310968646090883, | |
| "loss": 2.7044, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.36500170667880305, | |
| "grad_norm": 0.5091971755027771, | |
| "learning_rate": 0.00014284920653475866, | |
| "loss": 2.6281, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.36591193537376265, | |
| "grad_norm": 0.4547927975654602, | |
| "learning_rate": 0.00014258837000439618, | |
| "loss": 2.3868, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.36682216406872226, | |
| "grad_norm": 0.4767512083053589, | |
| "learning_rate": 0.0001423271790405828, | |
| "loss": 2.4612, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.36773239276368186, | |
| "grad_norm": 0.4405704736709595, | |
| "learning_rate": 0.00014206563581702964, | |
| "loss": 2.3659, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.36864262145864146, | |
| "grad_norm": 0.46354222297668457, | |
| "learning_rate": 0.0001418037425103795, | |
| "loss": 2.5197, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.36955285015360106, | |
| "grad_norm": 0.4474616050720215, | |
| "learning_rate": 0.00014154150130018866, | |
| "loss": 2.5156, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.3704630788485607, | |
| "grad_norm": 0.40332525968551636, | |
| "learning_rate": 0.00014127891436890868, | |
| "loss": 2.3852, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.3713733075435203, | |
| "grad_norm": 0.3969736695289612, | |
| "learning_rate": 0.0001410159839018684, | |
| "loss": 2.316, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.37228353623847993, | |
| "grad_norm": 0.40719449520111084, | |
| "learning_rate": 0.0001407527120872557, | |
| "loss": 2.3618, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.37319376493343953, | |
| "grad_norm": 0.4055502116680145, | |
| "learning_rate": 0.00014048910111609915, | |
| "loss": 2.2901, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.37410399362839913, | |
| "grad_norm": 0.3881557285785675, | |
| "learning_rate": 0.0001402251531822499, | |
| "loss": 2.3096, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.37501422232335874, | |
| "grad_norm": 0.38520845770835876, | |
| "learning_rate": 0.00013996087048236358, | |
| "loss": 2.096, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.37592445101831834, | |
| "grad_norm": 0.42187705636024475, | |
| "learning_rate": 0.00013969625521588158, | |
| "loss": 2.3212, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.37683467971327794, | |
| "grad_norm": 0.4028996527194977, | |
| "learning_rate": 0.00013943130958501317, | |
| "loss": 2.2625, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.37774490840823755, | |
| "grad_norm": 0.4009808599948883, | |
| "learning_rate": 0.00013916603579471705, | |
| "loss": 2.3579, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.3786551371031972, | |
| "grad_norm": 0.3909650146961212, | |
| "learning_rate": 0.00013890043605268283, | |
| "loss": 2.219, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.3795653657981568, | |
| "grad_norm": 0.38394787907600403, | |
| "learning_rate": 0.00013863451256931287, | |
| "loss": 2.0305, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.3804755944931164, | |
| "grad_norm": 0.38588935136795044, | |
| "learning_rate": 0.00013836826755770384, | |
| "loss": 2.1602, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.381385823188076, | |
| "grad_norm": 0.40827664732933044, | |
| "learning_rate": 0.00013810170323362816, | |
| "loss": 2.2656, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.3822960518830356, | |
| "grad_norm": 0.38694658875465393, | |
| "learning_rate": 0.0001378348218155158, | |
| "loss": 2.0107, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.3832062805779952, | |
| "grad_norm": 0.4133118987083435, | |
| "learning_rate": 0.00013756762552443553, | |
| "loss": 2.2531, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.3841165092729548, | |
| "grad_norm": 0.45007333159446716, | |
| "learning_rate": 0.00013730011658407676, | |
| "loss": 2.1968, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.3850267379679144, | |
| "grad_norm": 0.4120388329029083, | |
| "learning_rate": 0.00013703229722073065, | |
| "loss": 2.149, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.385936966662874, | |
| "grad_norm": 0.3964555561542511, | |
| "learning_rate": 0.000136764169663272, | |
| "loss": 1.9833, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.38684719535783363, | |
| "grad_norm": 0.4019937813282013, | |
| "learning_rate": 0.00013649573614314044, | |
| "loss": 2.2682, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.3877574240527933, | |
| "grad_norm": 0.4096015691757202, | |
| "learning_rate": 0.00013622699889432184, | |
| "loss": 2.2043, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.3886676527477529, | |
| "grad_norm": 0.4118809998035431, | |
| "learning_rate": 0.00013595796015332984, | |
| "loss": 2.0993, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.3895778814427125, | |
| "grad_norm": 0.40942683815956116, | |
| "learning_rate": 0.00013568862215918717, | |
| "loss": 2.1092, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.3904881101376721, | |
| "grad_norm": 0.44069868326187134, | |
| "learning_rate": 0.00013541898715340716, | |
| "loss": 2.1773, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.3913983388326317, | |
| "grad_norm": 0.4173736870288849, | |
| "learning_rate": 0.00013514905737997473, | |
| "loss": 2.3106, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.3923085675275913, | |
| "grad_norm": 0.4134718179702759, | |
| "learning_rate": 0.00013487883508532815, | |
| "loss": 2.0739, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.3932187962225509, | |
| "grad_norm": 0.4332071840763092, | |
| "learning_rate": 0.00013460832251834011, | |
| "loss": 2.1973, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.3941290249175105, | |
| "grad_norm": 0.44685009121894836, | |
| "learning_rate": 0.00013433752193029886, | |
| "loss": 2.1495, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.3950392536124701, | |
| "grad_norm": 0.443863183259964, | |
| "learning_rate": 0.0001340664355748899, | |
| "loss": 2.098, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.39594948230742977, | |
| "grad_norm": 0.4420264661312103, | |
| "learning_rate": 0.0001337950657081768, | |
| "loss": 2.11, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.39685971100238937, | |
| "grad_norm": 0.4621724784374237, | |
| "learning_rate": 0.00013352341458858265, | |
| "loss": 2.2452, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.397769939697349, | |
| "grad_norm": 0.4760790467262268, | |
| "learning_rate": 0.00013325148447687125, | |
| "loss": 2.2252, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.3986801683923086, | |
| "grad_norm": 0.44758686423301697, | |
| "learning_rate": 0.0001329792776361282, | |
| "loss": 2.0233, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.3995903970872682, | |
| "grad_norm": 0.4756091237068176, | |
| "learning_rate": 0.00013270679633174218, | |
| "loss": 2.0235, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.4005006257822278, | |
| "grad_norm": 0.46966737508773804, | |
| "learning_rate": 0.00013243404283138597, | |
| "loss": 2.1165, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.4014108544771874, | |
| "grad_norm": 0.49555307626724243, | |
| "learning_rate": 0.00013216101940499768, | |
| "loss": 2.0503, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.402321083172147, | |
| "grad_norm": 0.5097188353538513, | |
| "learning_rate": 0.00013188772832476188, | |
| "loss": 2.164, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.4032313118671066, | |
| "grad_norm": 0.5129069685935974, | |
| "learning_rate": 0.00013161417186509052, | |
| "loss": 2.2285, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.4041415405620662, | |
| "grad_norm": 0.5010905265808105, | |
| "learning_rate": 0.00013134035230260427, | |
| "loss": 2.0997, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.40505176925702585, | |
| "grad_norm": 0.5693898797035217, | |
| "learning_rate": 0.00013106627191611332, | |
| "loss": 2.2556, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.40596199795198545, | |
| "grad_norm": 0.5993850231170654, | |
| "learning_rate": 0.0001307919329865985, | |
| "loss": 2.4551, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.40687222664694506, | |
| "grad_norm": 0.6662326455116272, | |
| "learning_rate": 0.00013051733779719234, | |
| "loss": 2.5501, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.40778245534190466, | |
| "grad_norm": 0.7213377356529236, | |
| "learning_rate": 0.00013024248863316012, | |
| "loss": 2.5791, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.40869268403686426, | |
| "grad_norm": 0.8552189469337463, | |
| "learning_rate": 0.00012996738778188067, | |
| "loss": 2.5783, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.40960291273182386, | |
| "grad_norm": 1.2669048309326172, | |
| "learning_rate": 0.0001296920375328275, | |
| "loss": 2.2038, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.41051314142678347, | |
| "grad_norm": 0.4974716901779175, | |
| "learning_rate": 0.00012941644017754964, | |
| "loss": 2.4347, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.41142337012174307, | |
| "grad_norm": 0.44350573420524597, | |
| "learning_rate": 0.00012914059800965268, | |
| "loss": 2.5506, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.4123335988167027, | |
| "grad_norm": 0.43674564361572266, | |
| "learning_rate": 0.0001288645133247795, | |
| "loss": 2.4322, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.41324382751166233, | |
| "grad_norm": 0.46389102935791016, | |
| "learning_rate": 0.00012858818842059145, | |
| "loss": 2.4429, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.41415405620662193, | |
| "grad_norm": 0.4361862540245056, | |
| "learning_rate": 0.00012831162559674887, | |
| "loss": 2.4228, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.41506428490158154, | |
| "grad_norm": 0.4095577001571655, | |
| "learning_rate": 0.0001280348271548923, | |
| "loss": 2.3177, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.41597451359654114, | |
| "grad_norm": 0.41431957483291626, | |
| "learning_rate": 0.00012775779539862304, | |
| "loss": 2.592, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.41688474229150074, | |
| "grad_norm": 0.4042568802833557, | |
| "learning_rate": 0.0001274805326334842, | |
| "loss": 2.3442, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.41779497098646035, | |
| "grad_norm": 0.4499073922634125, | |
| "learning_rate": 0.00012720304116694138, | |
| "loss": 2.4013, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.41870519968141995, | |
| "grad_norm": 0.41216856241226196, | |
| "learning_rate": 0.00012692532330836346, | |
| "loss": 2.3945, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.41961542837637955, | |
| "grad_norm": 0.423570841550827, | |
| "learning_rate": 0.00012664738136900348, | |
| "loss": 2.3147, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.42052565707133915, | |
| "grad_norm": 0.40256237983703613, | |
| "learning_rate": 0.00012636921766197943, | |
| "loss": 2.3273, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.42143588576629876, | |
| "grad_norm": 0.39139389991760254, | |
| "learning_rate": 0.0001260908345022547, | |
| "loss": 2.1793, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.4223461144612584, | |
| "grad_norm": 0.4252980351448059, | |
| "learning_rate": 0.00012581223420661913, | |
| "loss": 2.4073, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.423256343156218, | |
| "grad_norm": 0.40091437101364136, | |
| "learning_rate": 0.00012553341909366978, | |
| "loss": 2.0757, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.4241665718511776, | |
| "grad_norm": 0.3780224323272705, | |
| "learning_rate": 0.00012525439148379128, | |
| "loss": 2.1467, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.4250768005461372, | |
| "grad_norm": 0.3823038637638092, | |
| "learning_rate": 0.00012497515369913685, | |
| "loss": 2.0455, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.4259870292410968, | |
| "grad_norm": 0.3887813091278076, | |
| "learning_rate": 0.00012469570806360875, | |
| "loss": 2.1608, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.42689725793605643, | |
| "grad_norm": 0.3840020000934601, | |
| "learning_rate": 0.00012441605690283915, | |
| "loss": 2.0606, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.42780748663101603, | |
| "grad_norm": 0.402353972196579, | |
| "learning_rate": 0.00012413620254417057, | |
| "loss": 2.1491, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.42871771532597563, | |
| "grad_norm": 0.3938208818435669, | |
| "learning_rate": 0.00012385614731663666, | |
| "loss": 2.1968, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.42962794402093524, | |
| "grad_norm": 0.42270010709762573, | |
| "learning_rate": 0.00012357589355094275, | |
| "loss": 2.0428, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.4305381727158949, | |
| "grad_norm": 0.44319435954093933, | |
| "learning_rate": 0.0001232954435794464, | |
| "loss": 2.3332, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.4314484014108545, | |
| "grad_norm": 0.39667901396751404, | |
| "learning_rate": 0.00012301479973613822, | |
| "loss": 2.1076, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.4323586301058141, | |
| "grad_norm": 0.4128730595111847, | |
| "learning_rate": 0.00012273396435662212, | |
| "loss": 2.0687, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.4332688588007737, | |
| "grad_norm": 0.4556971788406372, | |
| "learning_rate": 0.00012245293977809605, | |
| "loss": 2.1699, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.4341790874957333, | |
| "grad_norm": 0.43817293643951416, | |
| "learning_rate": 0.0001221717283393326, | |
| "loss": 2.2619, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.4350893161906929, | |
| "grad_norm": 0.4148106873035431, | |
| "learning_rate": 0.0001218903323806595, | |
| "loss": 2.0805, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.4359995448856525, | |
| "grad_norm": 0.428019255399704, | |
| "learning_rate": 0.00012160875424393996, | |
| "loss": 2.1673, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.4369097735806121, | |
| "grad_norm": 0.44368696212768555, | |
| "learning_rate": 0.00012132699627255347, | |
| "loss": 2.1894, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.4378200022755717, | |
| "grad_norm": 0.44192251563072205, | |
| "learning_rate": 0.00012104506081137608, | |
| "loss": 2.131, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.4387302309705313, | |
| "grad_norm": 0.45728325843811035, | |
| "learning_rate": 0.00012076295020676103, | |
| "loss": 2.1606, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.439640459665491, | |
| "grad_norm": 0.45007196068763733, | |
| "learning_rate": 0.00012048066680651908, | |
| "loss": 2.1161, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.4405506883604506, | |
| "grad_norm": 0.46728429198265076, | |
| "learning_rate": 0.00012019821295989912, | |
| "loss": 2.2119, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.4414609170554102, | |
| "grad_norm": 0.44168952107429504, | |
| "learning_rate": 0.00011991559101756852, | |
| "loss": 2.1377, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.4423711457503698, | |
| "grad_norm": 0.43559008836746216, | |
| "learning_rate": 0.00011963280333159358, | |
| "loss": 2.054, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.4432813744453294, | |
| "grad_norm": 0.45376813411712646, | |
| "learning_rate": 0.00011934985225541998, | |
| "loss": 2.1481, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.444191603140289, | |
| "grad_norm": 0.49441829323768616, | |
| "learning_rate": 0.00011906674014385318, | |
| "loss": 2.0623, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.4451018318352486, | |
| "grad_norm": 0.4792848229408264, | |
| "learning_rate": 0.00011878346935303883, | |
| "loss": 2.2903, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.4460120605302082, | |
| "grad_norm": 0.4943479597568512, | |
| "learning_rate": 0.00011850004224044315, | |
| "loss": 2.3084, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.4469222892251678, | |
| "grad_norm": 0.5104116797447205, | |
| "learning_rate": 0.00011821646116483335, | |
| "loss": 2.2815, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.44783251792012746, | |
| "grad_norm": 0.4758678674697876, | |
| "learning_rate": 0.00011793272848625797, | |
| "loss": 2.0609, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.44874274661508706, | |
| "grad_norm": 0.5053780674934387, | |
| "learning_rate": 0.0001176488465660271, | |
| "loss": 2.1577, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.44965297531004667, | |
| "grad_norm": 0.5221775770187378, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 2.2924, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.45056320400500627, | |
| "grad_norm": 0.5941824913024902, | |
| "learning_rate": 0.00011708064445203042, | |
| "loss": 2.3528, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.45147343269996587, | |
| "grad_norm": 0.5934675931930542, | |
| "learning_rate": 0.00011679632898701649, | |
| "loss": 2.4301, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.4523836613949255, | |
| "grad_norm": 0.6429802179336548, | |
| "learning_rate": 0.0001165118737378116, | |
| "loss": 2.605, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.4532938900898851, | |
| "grad_norm": 0.7099617719650269, | |
| "learning_rate": 0.00011622728107173946, | |
| "loss": 2.4261, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.4542041187848447, | |
| "grad_norm": 0.8483153581619263, | |
| "learning_rate": 0.00011594255335726724, | |
| "loss": 2.5175, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.4551143474798043, | |
| "grad_norm": 1.6742238998413086, | |
| "learning_rate": 0.00011565769296398618, | |
| "loss": 2.6628, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4560245761747639, | |
| "grad_norm": 0.4585944712162018, | |
| "learning_rate": 0.00011537270226259169, | |
| "loss": 2.5794, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.45693480486972354, | |
| "grad_norm": 0.44302108883857727, | |
| "learning_rate": 0.00011508758362486358, | |
| "loss": 2.3893, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.45784503356468315, | |
| "grad_norm": 0.44478365778923035, | |
| "learning_rate": 0.00011480233942364645, | |
| "loss": 2.3204, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.45875526225964275, | |
| "grad_norm": 0.432339608669281, | |
| "learning_rate": 0.00011451697203282982, | |
| "loss": 2.3721, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.45966549095460235, | |
| "grad_norm": 0.4154585301876068, | |
| "learning_rate": 0.00011423148382732853, | |
| "loss": 2.4017, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.46057571964956195, | |
| "grad_norm": 0.4535946846008301, | |
| "learning_rate": 0.00011394587718306275, | |
| "loss": 2.5771, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.46148594834452156, | |
| "grad_norm": 0.4111529588699341, | |
| "learning_rate": 0.00011366015447693837, | |
| "loss": 2.2596, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.46239617703948116, | |
| "grad_norm": 0.431909441947937, | |
| "learning_rate": 0.0001133743180868273, | |
| "loss": 2.3504, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.46330640573444076, | |
| "grad_norm": 0.4069095849990845, | |
| "learning_rate": 0.00011308837039154739, | |
| "loss": 2.2642, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.46421663442940037, | |
| "grad_norm": 0.4086742401123047, | |
| "learning_rate": 0.0001128023137708429, | |
| "loss": 2.2712, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.46512686312435997, | |
| "grad_norm": 0.4243764579296112, | |
| "learning_rate": 0.0001125161506053646, | |
| "loss": 2.3859, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.4660370918193196, | |
| "grad_norm": 0.39640411734580994, | |
| "learning_rate": 0.00011222988327664997, | |
| "loss": 2.1983, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.46694732051427923, | |
| "grad_norm": 0.4024841785430908, | |
| "learning_rate": 0.00011194351416710324, | |
| "loss": 2.2473, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.46785754920923883, | |
| "grad_norm": 0.3982551693916321, | |
| "learning_rate": 0.00011165704565997593, | |
| "loss": 2.1234, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.46876777790419843, | |
| "grad_norm": 0.4064629375934601, | |
| "learning_rate": 0.00011137048013934656, | |
| "loss": 2.1587, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.46967800659915804, | |
| "grad_norm": 0.3930409550666809, | |
| "learning_rate": 0.00011108381999010111, | |
| "loss": 2.2324, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 0.38076552748680115, | |
| "learning_rate": 0.00011079706759791311, | |
| "loss": 2.1943, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.47149846398907724, | |
| "grad_norm": 0.38611575961112976, | |
| "learning_rate": 0.00011051022534922371, | |
| "loss": 2.1572, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.47240869268403685, | |
| "grad_norm": 0.3925657868385315, | |
| "learning_rate": 0.00011022329563122191, | |
| "loss": 2.232, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.47331892137899645, | |
| "grad_norm": 0.401633620262146, | |
| "learning_rate": 0.00010993628083182467, | |
| "loss": 2.1639, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.4742291500739561, | |
| "grad_norm": 0.3987519443035126, | |
| "learning_rate": 0.000109649183339657, | |
| "loss": 2.1657, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.4751393787689157, | |
| "grad_norm": 0.3978632986545563, | |
| "learning_rate": 0.00010936200554403209, | |
| "loss": 2.1929, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.4760496074638753, | |
| "grad_norm": 0.40962356328964233, | |
| "learning_rate": 0.00010907474983493144, | |
| "loss": 2.167, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.4769598361588349, | |
| "grad_norm": 0.41112762689590454, | |
| "learning_rate": 0.00010878741860298503, | |
| "loss": 2.178, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.4778700648537945, | |
| "grad_norm": 0.4231667220592499, | |
| "learning_rate": 0.00010850001423945126, | |
| "loss": 1.995, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.4787802935487541, | |
| "grad_norm": 0.42191213369369507, | |
| "learning_rate": 0.00010821253913619726, | |
| "loss": 2.1639, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.4796905222437137, | |
| "grad_norm": 0.41400229930877686, | |
| "learning_rate": 0.00010792499568567884, | |
| "loss": 2.129, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.4806007509386733, | |
| "grad_norm": 0.4085232615470886, | |
| "learning_rate": 0.00010763738628092062, | |
| "loss": 2.0848, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.48151097963363293, | |
| "grad_norm": 0.4125650227069855, | |
| "learning_rate": 0.00010734971331549603, | |
| "loss": 1.9966, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.48242120832859253, | |
| "grad_norm": 0.42445600032806396, | |
| "learning_rate": 0.00010706197918350758, | |
| "loss": 1.9826, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.4833314370235522, | |
| "grad_norm": 0.45885640382766724, | |
| "learning_rate": 0.0001067741862795668, | |
| "loss": 2.1072, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.4842416657185118, | |
| "grad_norm": 0.45989179611206055, | |
| "learning_rate": 0.0001064863369987743, | |
| "loss": 2.4108, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.4851518944134714, | |
| "grad_norm": 0.4582422375679016, | |
| "learning_rate": 0.00010619843373669993, | |
| "loss": 2.1733, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.486062123108431, | |
| "grad_norm": 0.4393966495990753, | |
| "learning_rate": 0.00010591047888936274, | |
| "loss": 2.1335, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.4869723518033906, | |
| "grad_norm": 0.4389214813709259, | |
| "learning_rate": 0.00010562247485321115, | |
| "loss": 2.0692, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.4878825804983502, | |
| "grad_norm": 0.4686262905597687, | |
| "learning_rate": 0.00010533442402510284, | |
| "loss": 2.226, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.4887928091933098, | |
| "grad_norm": 0.47556954622268677, | |
| "learning_rate": 0.00010504632880228498, | |
| "loss": 2.2484, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.4897030378882694, | |
| "grad_norm": 0.46617594361305237, | |
| "learning_rate": 0.00010475819158237425, | |
| "loss": 2.2602, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.490613266583229, | |
| "grad_norm": 0.47310179471969604, | |
| "learning_rate": 0.00010447001476333673, | |
| "loss": 2.089, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.49152349527818867, | |
| "grad_norm": 0.507408618927002, | |
| "learning_rate": 0.00010418180074346815, | |
| "loss": 2.2762, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.4924337239731483, | |
| "grad_norm": 0.501750111579895, | |
| "learning_rate": 0.00010389355192137377, | |
| "loss": 2.1129, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.4933439526681079, | |
| "grad_norm": 0.5230852365493774, | |
| "learning_rate": 0.00010360527069594859, | |
| "loss": 2.4063, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.4942541813630675, | |
| "grad_norm": 0.499337762594223, | |
| "learning_rate": 0.00010331695946635708, | |
| "loss": 2.1379, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.4951644100580271, | |
| "grad_norm": 0.5291334390640259, | |
| "learning_rate": 0.00010302862063201367, | |
| "loss": 2.2284, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.4960746387529867, | |
| "grad_norm": 0.5520262718200684, | |
| "learning_rate": 0.00010274025659256232, | |
| "loss": 2.2331, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.4969848674479463, | |
| "grad_norm": 0.5880653262138367, | |
| "learning_rate": 0.00010245186974785685, | |
| "loss": 2.369, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.4978950961429059, | |
| "grad_norm": 0.5887028574943542, | |
| "learning_rate": 0.00010216346249794087, | |
| "loss": 2.3323, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.4988053248378655, | |
| "grad_norm": 0.664823055267334, | |
| "learning_rate": 0.00010187503724302776, | |
| "loss": 2.4451, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.4997155535328251, | |
| "grad_norm": 0.8141515254974365, | |
| "learning_rate": 0.00010158659638348081, | |
| "loss": 2.4308, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.5006257822277848, | |
| "grad_norm": 1.4680793285369873, | |
| "learning_rate": 0.0001012981423197931, | |
| "loss": 2.6184, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5006257822277848, | |
| "eval_loss": 2.2486555576324463, | |
| "eval_runtime": 203.9397, | |
| "eval_samples_per_second": 9.076, | |
| "eval_steps_per_second": 4.541, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5015360109227444, | |
| "grad_norm": 0.44648221135139465, | |
| "learning_rate": 0.00010100967745256766, | |
| "loss": 2.4608, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.502446239617704, | |
| "grad_norm": 0.4537757635116577, | |
| "learning_rate": 0.00010072120418249745, | |
| "loss": 2.3225, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.5033564683126636, | |
| "grad_norm": 0.45517390966415405, | |
| "learning_rate": 0.00010043272491034523, | |
| "loss": 2.4937, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.5042666970076232, | |
| "grad_norm": 0.4309350848197937, | |
| "learning_rate": 0.00010014424203692388, | |
| "loss": 2.3752, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.5051769257025828, | |
| "grad_norm": 0.4162474274635315, | |
| "learning_rate": 9.985575796307615e-05, | |
| "loss": 2.356, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.5060871543975424, | |
| "grad_norm": 0.454484224319458, | |
| "learning_rate": 9.956727508965481e-05, | |
| "loss": 2.3114, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.506997383092502, | |
| "grad_norm": 0.4174659550189972, | |
| "learning_rate": 9.927879581750259e-05, | |
| "loss": 2.2907, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.5079076117874616, | |
| "grad_norm": 0.42208272218704224, | |
| "learning_rate": 9.899032254743235e-05, | |
| "loss": 2.3062, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.5088178404824212, | |
| "grad_norm": 0.41622260212898254, | |
| "learning_rate": 9.870185768020693e-05, | |
| "loss": 2.3293, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.5097280691773808, | |
| "grad_norm": 0.45536091923713684, | |
| "learning_rate": 9.84134036165192e-05, | |
| "loss": 2.2693, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.5106382978723404, | |
| "grad_norm": 0.3866654634475708, | |
| "learning_rate": 9.812496275697226e-05, | |
| "loss": 2.1254, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.5115485265673, | |
| "grad_norm": 0.41163724660873413, | |
| "learning_rate": 9.783653750205915e-05, | |
| "loss": 2.2173, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.5124587552622596, | |
| "grad_norm": 0.4118368327617645, | |
| "learning_rate": 9.754813025214317e-05, | |
| "loss": 2.2477, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.5133689839572193, | |
| "grad_norm": 0.4303027093410492, | |
| "learning_rate": 9.725974340743769e-05, | |
| "loss": 2.3855, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.5142792126521789, | |
| "grad_norm": 0.41484424471855164, | |
| "learning_rate": 9.697137936798634e-05, | |
| "loss": 2.289, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.5151894413471385, | |
| "grad_norm": 0.397540807723999, | |
| "learning_rate": 9.668304053364294e-05, | |
| "loss": 2.1869, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.5160996700420981, | |
| "grad_norm": 0.38194504380226135, | |
| "learning_rate": 9.639472930405143e-05, | |
| "loss": 2.1475, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.5170098987370577, | |
| "grad_norm": 0.42346352338790894, | |
| "learning_rate": 9.610644807862625e-05, | |
| "loss": 2.1869, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.5179201274320173, | |
| "grad_norm": 0.4076716899871826, | |
| "learning_rate": 9.581819925653188e-05, | |
| "loss": 2.1979, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.5188303561269769, | |
| "grad_norm": 0.3888402581214905, | |
| "learning_rate": 9.552998523666326e-05, | |
| "loss": 2.131, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.5197405848219365, | |
| "grad_norm": 0.3931845426559448, | |
| "learning_rate": 9.524180841762577e-05, | |
| "loss": 2.07, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.5206508135168961, | |
| "grad_norm": 0.38014844059944153, | |
| "learning_rate": 9.495367119771503e-05, | |
| "loss": 1.8905, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.5215610422118557, | |
| "grad_norm": 0.4095986783504486, | |
| "learning_rate": 9.46655759748972e-05, | |
| "loss": 2.1217, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.5224712709068153, | |
| "grad_norm": 0.3965059518814087, | |
| "learning_rate": 9.437752514678887e-05, | |
| "loss": 2.0446, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.5233814996017749, | |
| "grad_norm": 0.42244595289230347, | |
| "learning_rate": 9.408952111063727e-05, | |
| "loss": 2.1184, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.5242917282967345, | |
| "grad_norm": 0.4098374843597412, | |
| "learning_rate": 9.380156626330009e-05, | |
| "loss": 2.0365, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.5252019569916941, | |
| "grad_norm": 0.424630731344223, | |
| "learning_rate": 9.35136630012257e-05, | |
| "loss": 2.1321, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.5261121856866537, | |
| "grad_norm": 0.431257039308548, | |
| "learning_rate": 9.322581372043321e-05, | |
| "loss": 2.1883, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.5270224143816133, | |
| "grad_norm": 0.40972310304641724, | |
| "learning_rate": 9.293802081649243e-05, | |
| "loss": 2.0498, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.527932643076573, | |
| "grad_norm": 0.4238174855709076, | |
| "learning_rate": 9.265028668450402e-05, | |
| "loss": 2.0831, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.5288428717715326, | |
| "grad_norm": 0.43082720041275024, | |
| "learning_rate": 9.23626137190794e-05, | |
| "loss": 2.1704, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.5297531004664922, | |
| "grad_norm": 0.4657386541366577, | |
| "learning_rate": 9.207500431432115e-05, | |
| "loss": 2.1338, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.5306633291614519, | |
| "grad_norm": 0.45546218752861023, | |
| "learning_rate": 9.178746086380275e-05, | |
| "loss": 2.1469, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.5315735578564115, | |
| "grad_norm": 0.4571657180786133, | |
| "learning_rate": 9.149998576054874e-05, | |
| "loss": 2.2036, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.5324837865513711, | |
| "grad_norm": 0.4736308753490448, | |
| "learning_rate": 9.121258139701502e-05, | |
| "loss": 2.2127, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.5333940152463307, | |
| "grad_norm": 0.46164044737815857, | |
| "learning_rate": 9.092525016506858e-05, | |
| "loss": 2.1177, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.5343042439412903, | |
| "grad_norm": 0.48902738094329834, | |
| "learning_rate": 9.063799445596795e-05, | |
| "loss": 2.2188, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.5352144726362499, | |
| "grad_norm": 0.4848826825618744, | |
| "learning_rate": 9.035081666034304e-05, | |
| "loss": 2.2372, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.5361247013312095, | |
| "grad_norm": 0.49447596073150635, | |
| "learning_rate": 9.006371916817534e-05, | |
| "loss": 2.2786, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.5370349300261691, | |
| "grad_norm": 0.45803365111351013, | |
| "learning_rate": 8.977670436877811e-05, | |
| "loss": 2.0856, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.5379451587211287, | |
| "grad_norm": 0.5051137208938599, | |
| "learning_rate": 8.948977465077632e-05, | |
| "loss": 2.2219, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.5388553874160883, | |
| "grad_norm": 0.5000495910644531, | |
| "learning_rate": 8.920293240208694e-05, | |
| "loss": 2.2151, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.5397656161110479, | |
| "grad_norm": 0.5208660960197449, | |
| "learning_rate": 8.891618000989891e-05, | |
| "loss": 2.3381, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.5406758448060075, | |
| "grad_norm": 0.5478562116622925, | |
| "learning_rate": 8.862951986065345e-05, | |
| "loss": 2.1592, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.5415860735009671, | |
| "grad_norm": 0.5808135271072388, | |
| "learning_rate": 8.83429543400241e-05, | |
| "loss": 2.3678, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.5424963021959267, | |
| "grad_norm": 0.5837730765342712, | |
| "learning_rate": 8.805648583289674e-05, | |
| "loss": 2.3397, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.5434065308908863, | |
| "grad_norm": 0.6297701597213745, | |
| "learning_rate": 8.777011672335008e-05, | |
| "loss": 2.4747, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.5443167595858459, | |
| "grad_norm": 0.735091507434845, | |
| "learning_rate": 8.748384939463543e-05, | |
| "loss": 2.5623, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.5452269882808055, | |
| "grad_norm": 0.8794751167297363, | |
| "learning_rate": 8.719768622915714e-05, | |
| "loss": 2.557, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.5461372169757651, | |
| "grad_norm": 1.7257400751113892, | |
| "learning_rate": 8.691162960845264e-05, | |
| "loss": 2.7309, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5470474456707247, | |
| "grad_norm": 0.42087942361831665, | |
| "learning_rate": 8.662568191317273e-05, | |
| "loss": 2.3698, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.5479576743656844, | |
| "grad_norm": 0.45860013365745544, | |
| "learning_rate": 8.633984552306164e-05, | |
| "loss": 2.4207, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.548867903060644, | |
| "grad_norm": 0.4520268738269806, | |
| "learning_rate": 8.605412281693727e-05, | |
| "loss": 2.5066, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.5497781317556036, | |
| "grad_norm": 0.42296549677848816, | |
| "learning_rate": 8.57685161726715e-05, | |
| "loss": 2.3086, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.5506883604505632, | |
| "grad_norm": 0.4488202631473541, | |
| "learning_rate": 8.548302796717019e-05, | |
| "loss": 2.3949, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.5515985891455228, | |
| "grad_norm": 0.43144237995147705, | |
| "learning_rate": 8.519766057635355e-05, | |
| "loss": 2.3862, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.5525088178404824, | |
| "grad_norm": 0.43359822034835815, | |
| "learning_rate": 8.491241637513644e-05, | |
| "loss": 2.2576, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.553419046535442, | |
| "grad_norm": 0.41321370005607605, | |
| "learning_rate": 8.462729773740832e-05, | |
| "loss": 2.294, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.5543292752304017, | |
| "grad_norm": 0.4104655981063843, | |
| "learning_rate": 8.434230703601384e-05, | |
| "loss": 2.2005, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.5552395039253613, | |
| "grad_norm": 0.4300507605075836, | |
| "learning_rate": 8.405744664273278e-05, | |
| "loss": 2.4242, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5561497326203209, | |
| "grad_norm": 0.42208558320999146, | |
| "learning_rate": 8.37727189282606e-05, | |
| "loss": 2.2788, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.5570599613152805, | |
| "grad_norm": 0.4338865280151367, | |
| "learning_rate": 8.34881262621884e-05, | |
| "loss": 2.4823, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.5579701900102401, | |
| "grad_norm": 0.399996280670166, | |
| "learning_rate": 8.320367101298351e-05, | |
| "loss": 2.1704, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.5588804187051997, | |
| "grad_norm": 0.4285057783126831, | |
| "learning_rate": 8.291935554796962e-05, | |
| "loss": 2.3405, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.5597906474001593, | |
| "grad_norm": 0.41062313318252563, | |
| "learning_rate": 8.263518223330697e-05, | |
| "loss": 2.2424, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.5607008760951189, | |
| "grad_norm": 0.40377363562583923, | |
| "learning_rate": 8.235115343397295e-05, | |
| "loss": 2.2611, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.5616111047900785, | |
| "grad_norm": 0.38683268427848816, | |
| "learning_rate": 8.206727151374207e-05, | |
| "loss": 2.0895, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.5625213334850381, | |
| "grad_norm": 0.3813983201980591, | |
| "learning_rate": 8.178353883516664e-05, | |
| "loss": 2.0735, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.5634315621799977, | |
| "grad_norm": 0.40514618158340454, | |
| "learning_rate": 8.149995775955686e-05, | |
| "loss": 2.2261, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.5643417908749573, | |
| "grad_norm": 0.3858005106449127, | |
| "learning_rate": 8.121653064696118e-05, | |
| "loss": 2.0794, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.565252019569917, | |
| "grad_norm": 0.4032715857028961, | |
| "learning_rate": 8.093325985614685e-05, | |
| "loss": 2.2016, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.5661622482648766, | |
| "grad_norm": 0.40262308716773987, | |
| "learning_rate": 8.065014774458003e-05, | |
| "loss": 2.1193, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.5670724769598362, | |
| "grad_norm": 0.4016035497188568, | |
| "learning_rate": 8.036719666840647e-05, | |
| "loss": 2.0265, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.5679827056547958, | |
| "grad_norm": 0.42149612307548523, | |
| "learning_rate": 8.008440898243149e-05, | |
| "loss": 2.1186, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.5688929343497554, | |
| "grad_norm": 0.40055879950523376, | |
| "learning_rate": 7.980178704010089e-05, | |
| "loss": 2.0066, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.569803163044715, | |
| "grad_norm": 0.40722012519836426, | |
| "learning_rate": 7.951933319348095e-05, | |
| "loss": 2.0262, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.5707133917396746, | |
| "grad_norm": 0.4256587624549866, | |
| "learning_rate": 7.923704979323899e-05, | |
| "loss": 2.0765, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.5716236204346342, | |
| "grad_norm": 0.452120840549469, | |
| "learning_rate": 7.895493918862396e-05, | |
| "loss": 2.2261, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.5725338491295938, | |
| "grad_norm": 0.45264679193496704, | |
| "learning_rate": 7.867300372744657e-05, | |
| "loss": 2.1382, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.5734440778245534, | |
| "grad_norm": 0.4367516040802002, | |
| "learning_rate": 7.839124575606004e-05, | |
| "loss": 2.1152, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.574354306519513, | |
| "grad_norm": 0.41711944341659546, | |
| "learning_rate": 7.810966761934053e-05, | |
| "loss": 2.0499, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.5752645352144726, | |
| "grad_norm": 0.4274366497993469, | |
| "learning_rate": 7.782827166066739e-05, | |
| "loss": 2.0843, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.5761747639094322, | |
| "grad_norm": 0.4412054717540741, | |
| "learning_rate": 7.754706022190398e-05, | |
| "loss": 2.1291, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.5770849926043918, | |
| "grad_norm": 0.43048036098480225, | |
| "learning_rate": 7.726603564337791e-05, | |
| "loss": 2.0492, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.5779952212993514, | |
| "grad_norm": 0.4570690393447876, | |
| "learning_rate": 7.69852002638618e-05, | |
| "loss": 2.2868, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.578905449994311, | |
| "grad_norm": 0.4373176097869873, | |
| "learning_rate": 7.670455642055361e-05, | |
| "loss": 2.106, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.5798156786892706, | |
| "grad_norm": 0.46953141689300537, | |
| "learning_rate": 7.642410644905726e-05, | |
| "loss": 2.2209, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.5807259073842302, | |
| "grad_norm": 0.48714450001716614, | |
| "learning_rate": 7.614385268336336e-05, | |
| "loss": 2.2923, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.5816361360791898, | |
| "grad_norm": 0.45921769738197327, | |
| "learning_rate": 7.586379745582944e-05, | |
| "loss": 2.1636, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.5825463647741494, | |
| "grad_norm": 0.4652685821056366, | |
| "learning_rate": 7.558394309716088e-05, | |
| "loss": 2.205, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.5834565934691092, | |
| "grad_norm": 0.48991554975509644, | |
| "learning_rate": 7.530429193639128e-05, | |
| "loss": 2.1805, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.5843668221640688, | |
| "grad_norm": 0.515925407409668, | |
| "learning_rate": 7.502484630086318e-05, | |
| "loss": 2.2075, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.5852770508590284, | |
| "grad_norm": 0.497593492269516, | |
| "learning_rate": 7.474560851620873e-05, | |
| "loss": 2.0497, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.586187279553988, | |
| "grad_norm": 0.5590106248855591, | |
| "learning_rate": 7.446658090633026e-05, | |
| "loss": 2.273, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.5870975082489476, | |
| "grad_norm": 0.570801854133606, | |
| "learning_rate": 7.41877657933809e-05, | |
| "loss": 2.4023, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.5880077369439072, | |
| "grad_norm": 0.5791385173797607, | |
| "learning_rate": 7.390916549774536e-05, | |
| "loss": 2.2391, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.5889179656388668, | |
| "grad_norm": 0.677937924861908, | |
| "learning_rate": 7.363078233802063e-05, | |
| "loss": 2.6502, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.5898281943338264, | |
| "grad_norm": 0.6995664238929749, | |
| "learning_rate": 7.335261863099651e-05, | |
| "loss": 2.4716, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.590738423028786, | |
| "grad_norm": 0.8106637597084045, | |
| "learning_rate": 7.307467669163655e-05, | |
| "loss": 2.3574, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.5916486517237456, | |
| "grad_norm": 1.577392816543579, | |
| "learning_rate": 7.279695883305866e-05, | |
| "loss": 2.1657, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.5925588804187052, | |
| "grad_norm": 0.43565502762794495, | |
| "learning_rate": 7.251946736651582e-05, | |
| "loss": 2.4673, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.5934691091136648, | |
| "grad_norm": 0.42318448424339294, | |
| "learning_rate": 7.224220460137701e-05, | |
| "loss": 2.409, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.5943793378086244, | |
| "grad_norm": 0.41891592741012573, | |
| "learning_rate": 7.196517284510773e-05, | |
| "loss": 2.3842, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.595289566503584, | |
| "grad_norm": 0.4418095350265503, | |
| "learning_rate": 7.168837440325114e-05, | |
| "loss": 2.3998, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.5961997951985436, | |
| "grad_norm": 0.43118715286254883, | |
| "learning_rate": 7.141181157940859e-05, | |
| "loss": 2.3845, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.5971100238935032, | |
| "grad_norm": 0.41435372829437256, | |
| "learning_rate": 7.11354866752205e-05, | |
| "loss": 2.3077, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.5980202525884628, | |
| "grad_norm": 0.4133334457874298, | |
| "learning_rate": 7.085940199034735e-05, | |
| "loss": 2.385, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.5989304812834224, | |
| "grad_norm": 0.4240586757659912, | |
| "learning_rate": 7.058355982245037e-05, | |
| "loss": 2.3835, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.599840709978382, | |
| "grad_norm": 0.4071354568004608, | |
| "learning_rate": 7.030796246717255e-05, | |
| "loss": 2.0752, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.6007509386733417, | |
| "grad_norm": 0.44435110688209534, | |
| "learning_rate": 7.003261221811934e-05, | |
| "loss": 2.4811, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.6016611673683013, | |
| "grad_norm": 0.4175739884376526, | |
| "learning_rate": 6.97575113668399e-05, | |
| "loss": 2.3717, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.6025713960632609, | |
| "grad_norm": 0.41801995038986206, | |
| "learning_rate": 6.948266220280771e-05, | |
| "loss": 2.3406, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.6034816247582205, | |
| "grad_norm": 0.42795541882514954, | |
| "learning_rate": 6.920806701340155e-05, | |
| "loss": 2.1457, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.6043918534531801, | |
| "grad_norm": 0.39928749203681946, | |
| "learning_rate": 6.893372808388675e-05, | |
| "loss": 2.2435, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.6053020821481397, | |
| "grad_norm": 0.39942193031311035, | |
| "learning_rate": 6.865964769739575e-05, | |
| "loss": 2.1791, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.6062123108430993, | |
| "grad_norm": 0.3968772888183594, | |
| "learning_rate": 6.838582813490947e-05, | |
| "loss": 2.106, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.607122539538059, | |
| "grad_norm": 0.41822007298469543, | |
| "learning_rate": 6.811227167523815e-05, | |
| "loss": 2.2538, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.6080327682330186, | |
| "grad_norm": 0.39328843355178833, | |
| "learning_rate": 6.783898059500233e-05, | |
| "loss": 2.1372, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.6089429969279782, | |
| "grad_norm": 0.39368972182273865, | |
| "learning_rate": 6.756595716861407e-05, | |
| "loss": 2.1002, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.6098532256229378, | |
| "grad_norm": 0.41033080220222473, | |
| "learning_rate": 6.729320366825784e-05, | |
| "loss": 2.0963, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.6107634543178974, | |
| "grad_norm": 0.40454182028770447, | |
| "learning_rate": 6.702072236387182e-05, | |
| "loss": 2.0902, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.611673683012857, | |
| "grad_norm": 0.40117865800857544, | |
| "learning_rate": 6.674851552312878e-05, | |
| "loss": 2.086, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.6125839117078166, | |
| "grad_norm": 0.39771145582199097, | |
| "learning_rate": 6.647658541141735e-05, | |
| "loss": 1.977, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.6134941404027762, | |
| "grad_norm": 0.4230089485645294, | |
| "learning_rate": 6.620493429182323e-05, | |
| "loss": 2.1687, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.6144043690977358, | |
| "grad_norm": 0.40441179275512695, | |
| "learning_rate": 6.593356442511015e-05, | |
| "loss": 2.1608, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.6153145977926954, | |
| "grad_norm": 0.40813207626342773, | |
| "learning_rate": 6.566247806970119e-05, | |
| "loss": 2.0103, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.616224826487655, | |
| "grad_norm": 0.4276919364929199, | |
| "learning_rate": 6.539167748165994e-05, | |
| "loss": 2.0024, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.6171350551826146, | |
| "grad_norm": 0.4275762140750885, | |
| "learning_rate": 6.512116491467185e-05, | |
| "loss": 2.1589, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.6180452838775743, | |
| "grad_norm": 0.4258089065551758, | |
| "learning_rate": 6.485094262002529e-05, | |
| "loss": 1.9598, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.6189555125725339, | |
| "grad_norm": 0.436363160610199, | |
| "learning_rate": 6.458101284659286e-05, | |
| "loss": 2.2201, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.6198657412674935, | |
| "grad_norm": 0.41596871614456177, | |
| "learning_rate": 6.431137784081282e-05, | |
| "loss": 2.0394, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.6207759699624531, | |
| "grad_norm": 0.46983107924461365, | |
| "learning_rate": 6.404203984667019e-05, | |
| "loss": 2.0277, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.6216861986574127, | |
| "grad_norm": 0.4560098350048065, | |
| "learning_rate": 6.377300110567821e-05, | |
| "loss": 2.2369, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.6225964273523723, | |
| "grad_norm": 0.45281484723091125, | |
| "learning_rate": 6.350426385685957e-05, | |
| "loss": 2.2576, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.6235066560473319, | |
| "grad_norm": 0.45989713072776794, | |
| "learning_rate": 6.323583033672799e-05, | |
| "loss": 2.1294, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.6244168847422915, | |
| "grad_norm": 0.46797510981559753, | |
| "learning_rate": 6.296770277926937e-05, | |
| "loss": 2.0688, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.6253271134372511, | |
| "grad_norm": 0.5156514644622803, | |
| "learning_rate": 6.269988341592328e-05, | |
| "loss": 2.1114, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.6262373421322107, | |
| "grad_norm": 0.48389503359794617, | |
| "learning_rate": 6.243237447556449e-05, | |
| "loss": 2.0931, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.6271475708271703, | |
| "grad_norm": 0.4714515209197998, | |
| "learning_rate": 6.216517818448423e-05, | |
| "loss": 2.0982, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.6280577995221299, | |
| "grad_norm": 0.5029696226119995, | |
| "learning_rate": 6.189829676637182e-05, | |
| "loss": 2.2145, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.6289680282170895, | |
| "grad_norm": 0.49942919611930847, | |
| "learning_rate": 6.163173244229619e-05, | |
| "loss": 2.1344, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.6298782569120491, | |
| "grad_norm": 0.5015403032302856, | |
| "learning_rate": 6.136548743068713e-05, | |
| "loss": 2.0721, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.6307884856070087, | |
| "grad_norm": 0.5125733613967896, | |
| "learning_rate": 6.109956394731722e-05, | |
| "loss": 2.0609, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.6316987143019683, | |
| "grad_norm": 0.5657601952552795, | |
| "learning_rate": 6.083396420528298e-05, | |
| "loss": 2.4455, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.6326089429969279, | |
| "grad_norm": 0.5578758120536804, | |
| "learning_rate": 6.056869041498687e-05, | |
| "loss": 2.205, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.6335191716918875, | |
| "grad_norm": 0.6066897511482239, | |
| "learning_rate": 6.030374478411847e-05, | |
| "loss": 2.3107, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.6344294003868471, | |
| "grad_norm": 0.6329779028892517, | |
| "learning_rate": 6.0039129517636435e-05, | |
| "loss": 2.3413, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.6353396290818069, | |
| "grad_norm": 0.7217928767204285, | |
| "learning_rate": 5.9774846817750105e-05, | |
| "loss": 2.4855, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.6362498577767665, | |
| "grad_norm": 0.9335882067680359, | |
| "learning_rate": 5.951089888390087e-05, | |
| "loss": 2.7768, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.6371600864717261, | |
| "grad_norm": 1.5800278186798096, | |
| "learning_rate": 5.924728791274432e-05, | |
| "loss": 2.5227, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.6380703151666857, | |
| "grad_norm": 0.4945278763771057, | |
| "learning_rate": 5.89840160981316e-05, | |
| "loss": 2.4396, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.6389805438616453, | |
| "grad_norm": 0.4547106921672821, | |
| "learning_rate": 5.872108563109131e-05, | |
| "loss": 2.3647, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.6398907725566049, | |
| "grad_norm": 0.46132609248161316, | |
| "learning_rate": 5.845849869981137e-05, | |
| "loss": 2.5182, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.6408010012515645, | |
| "grad_norm": 0.41916942596435547, | |
| "learning_rate": 5.819625748962049e-05, | |
| "loss": 2.3084, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.6417112299465241, | |
| "grad_norm": 0.4376005530357361, | |
| "learning_rate": 5.79343641829704e-05, | |
| "loss": 2.3951, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.6426214586414837, | |
| "grad_norm": 0.4435478448867798, | |
| "learning_rate": 5.7672820959417254e-05, | |
| "loss": 2.3179, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.6435316873364433, | |
| "grad_norm": 0.43887507915496826, | |
| "learning_rate": 5.741162999560386e-05, | |
| "loss": 2.2442, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.6444419160314029, | |
| "grad_norm": 0.45083487033843994, | |
| "learning_rate": 5.7150793465241346e-05, | |
| "loss": 2.3397, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.6453521447263625, | |
| "grad_norm": 0.4538663625717163, | |
| "learning_rate": 5.68903135390912e-05, | |
| "loss": 2.1915, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.6462623734213221, | |
| "grad_norm": 0.41724029183387756, | |
| "learning_rate": 5.663019238494704e-05, | |
| "loss": 2.314, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.6471726021162817, | |
| "grad_norm": 0.4224849343299866, | |
| "learning_rate": 5.637043216761678e-05, | |
| "loss": 2.1712, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.6480828308112413, | |
| "grad_norm": 0.4206700921058655, | |
| "learning_rate": 5.611103504890444e-05, | |
| "loss": 2.2096, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.6489930595062009, | |
| "grad_norm": 0.41635435819625854, | |
| "learning_rate": 5.5852003187592226e-05, | |
| "loss": 2.3813, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.6499032882011605, | |
| "grad_norm": 0.4309711456298828, | |
| "learning_rate": 5.559333873942259e-05, | |
| "loss": 2.3199, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.6508135168961201, | |
| "grad_norm": 0.4129570424556732, | |
| "learning_rate": 5.533504385708024e-05, | |
| "loss": 2.2384, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.6517237455910797, | |
| "grad_norm": 0.40963393449783325, | |
| "learning_rate": 5.5077120690174246e-05, | |
| "loss": 2.1162, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.6526339742860394, | |
| "grad_norm": 0.4129562973976135, | |
| "learning_rate": 5.481957138522018e-05, | |
| "loss": 2.2232, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.653544202980999, | |
| "grad_norm": 0.43267586827278137, | |
| "learning_rate": 5.456239808562209e-05, | |
| "loss": 2.208, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.6544544316759586, | |
| "grad_norm": 0.4190649688243866, | |
| "learning_rate": 5.4305602931655045e-05, | |
| "loss": 2.0585, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.6553646603709182, | |
| "grad_norm": 0.41037124395370483, | |
| "learning_rate": 5.404918806044679e-05, | |
| "loss": 2.1371, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.6562748890658778, | |
| "grad_norm": 0.43521812558174133, | |
| "learning_rate": 5.379315560596038e-05, | |
| "loss": 2.2227, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.6571851177608374, | |
| "grad_norm": 0.4168831408023834, | |
| "learning_rate": 5.3537507698976365e-05, | |
| "loss": 1.962, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.658095346455797, | |
| "grad_norm": 0.4054042100906372, | |
| "learning_rate": 5.328224646707479e-05, | |
| "loss": 2.0154, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.6590055751507566, | |
| "grad_norm": 0.4229868948459625, | |
| "learning_rate": 5.3027374034617785e-05, | |
| "loss": 2.1075, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.6599158038457162, | |
| "grad_norm": 0.4285801351070404, | |
| "learning_rate": 5.277289252273174e-05, | |
| "loss": 2.0953, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.6608260325406758, | |
| "grad_norm": 0.43290793895721436, | |
| "learning_rate": 5.251880404928971e-05, | |
| "loss": 2.3198, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.6617362612356354, | |
| "grad_norm": 0.4112420082092285, | |
| "learning_rate": 5.226511072889371e-05, | |
| "loss": 2.1214, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 0.662646489930595, | |
| "grad_norm": 0.4102923572063446, | |
| "learning_rate": 5.201181467285723e-05, | |
| "loss": 1.8327, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.6635567186255547, | |
| "grad_norm": 0.45263952016830444, | |
| "learning_rate": 5.175891798918757e-05, | |
| "loss": 2.1415, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.6644669473205143, | |
| "grad_norm": 0.4748065769672394, | |
| "learning_rate": 5.1506422782568345e-05, | |
| "loss": 2.0534, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.6653771760154739, | |
| "grad_norm": 0.42939677834510803, | |
| "learning_rate": 5.125433115434197e-05, | |
| "loss": 1.8938, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 0.6662874047104335, | |
| "grad_norm": 0.4550830125808716, | |
| "learning_rate": 5.100264520249205e-05, | |
| "loss": 2.1615, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.6671976334053931, | |
| "grad_norm": 0.47507891058921814, | |
| "learning_rate": 5.0751367021626215e-05, | |
| "loss": 2.1034, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 0.6681078621003527, | |
| "grad_norm": 0.44998088479042053, | |
| "learning_rate": 5.050049870295841e-05, | |
| "loss": 1.9552, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.6690180907953123, | |
| "grad_norm": 0.4823133647441864, | |
| "learning_rate": 5.025004233429145e-05, | |
| "loss": 2.0965, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.669928319490272, | |
| "grad_norm": 0.4999094605445862, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 2.2037, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.6708385481852316, | |
| "grad_norm": 0.4802190661430359, | |
| "learning_rate": 4.9750373781012885e-05, | |
| "loss": 2.0832, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.6717487768801912, | |
| "grad_norm": 0.4969741702079773, | |
| "learning_rate": 4.950116575479586e-05, | |
| "loss": 2.0196, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.6726590055751508, | |
| "grad_norm": 0.47610992193222046, | |
| "learning_rate": 4.9252377995334444e-05, | |
| "loss": 1.9983, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 0.6735692342701104, | |
| "grad_norm": 0.5233409404754639, | |
| "learning_rate": 4.90040125731165e-05, | |
| "loss": 2.2088, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.67447946296507, | |
| "grad_norm": 0.5126591920852661, | |
| "learning_rate": 4.87560715551151e-05, | |
| "loss": 2.0659, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.6753896916600296, | |
| "grad_norm": 0.5209243893623352, | |
| "learning_rate": 4.85085570047713e-05, | |
| "loss": 2.2167, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.6762999203549892, | |
| "grad_norm": 0.5382196307182312, | |
| "learning_rate": 4.826147098197691e-05, | |
| "loss": 2.234, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 0.6772101490499488, | |
| "grad_norm": 0.5741109848022461, | |
| "learning_rate": 4.8014815543057475e-05, | |
| "loss": 2.2118, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.6781203777449084, | |
| "grad_norm": 0.577425479888916, | |
| "learning_rate": 4.776859274075506e-05, | |
| "loss": 2.2446, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.679030606439868, | |
| "grad_norm": 0.614007294178009, | |
| "learning_rate": 4.752280462421117e-05, | |
| "loss": 2.3033, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.6799408351348276, | |
| "grad_norm": 0.654388427734375, | |
| "learning_rate": 4.727745323894976e-05, | |
| "loss": 2.33, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.6808510638297872, | |
| "grad_norm": 0.7294265627861023, | |
| "learning_rate": 4.703254062686017e-05, | |
| "loss": 2.5732, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.6817612925247468, | |
| "grad_norm": 0.8145880699157715, | |
| "learning_rate": 4.678806882618003e-05, | |
| "loss": 2.4708, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 0.6826715212197064, | |
| "grad_norm": 1.3215488195419312, | |
| "learning_rate": 4.654403987147865e-05, | |
| "loss": 2.566, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.683581749914666, | |
| "grad_norm": 0.4625120759010315, | |
| "learning_rate": 4.630045579363957e-05, | |
| "loss": 2.4207, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 0.6844919786096256, | |
| "grad_norm": 0.4124356508255005, | |
| "learning_rate": 4.605731861984401e-05, | |
| "loss": 2.2606, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.6854022073045852, | |
| "grad_norm": 0.42320266366004944, | |
| "learning_rate": 4.5814630373554115e-05, | |
| "loss": 2.3107, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.6863124359995448, | |
| "grad_norm": 0.4337674379348755, | |
| "learning_rate": 4.557239307449561e-05, | |
| "loss": 2.5028, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.6872226646945045, | |
| "grad_norm": 0.44273191690444946, | |
| "learning_rate": 4.5330608738641486e-05, | |
| "loss": 2.42, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.6881328933894642, | |
| "grad_norm": 0.40541112422943115, | |
| "learning_rate": 4.508927937819499e-05, | |
| "loss": 2.1912, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.6890431220844238, | |
| "grad_norm": 0.4108361601829529, | |
| "learning_rate": 4.484840700157295e-05, | |
| "loss": 2.2871, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 0.6899533507793834, | |
| "grad_norm": 0.4105437994003296, | |
| "learning_rate": 4.4607993613388976e-05, | |
| "loss": 2.244, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.690863579474343, | |
| "grad_norm": 0.42231062054634094, | |
| "learning_rate": 4.436804121443689e-05, | |
| "loss": 2.3462, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.6917738081693026, | |
| "grad_norm": 0.42183414101600647, | |
| "learning_rate": 4.412855180167406e-05, | |
| "loss": 2.3269, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.6926840368642622, | |
| "grad_norm": 0.4242246448993683, | |
| "learning_rate": 4.388952736820453e-05, | |
| "loss": 2.2572, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 0.6935942655592218, | |
| "grad_norm": 0.41937458515167236, | |
| "learning_rate": 4.365096990326297e-05, | |
| "loss": 2.1373, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.6945044942541814, | |
| "grad_norm": 0.4179951250553131, | |
| "learning_rate": 4.3412881392197526e-05, | |
| "loss": 2.2584, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 0.695414722949141, | |
| "grad_norm": 0.4121822416782379, | |
| "learning_rate": 4.317526381645363e-05, | |
| "loss": 2.2389, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.6963249516441006, | |
| "grad_norm": 0.400831401348114, | |
| "learning_rate": 4.293811915355761e-05, | |
| "loss": 2.1734, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.6972351803390602, | |
| "grad_norm": 0.426580011844635, | |
| "learning_rate": 4.270144937709981e-05, | |
| "loss": 2.1537, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.6981454090340198, | |
| "grad_norm": 0.4221407175064087, | |
| "learning_rate": 4.2465256456718615e-05, | |
| "loss": 2.1184, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 0.6990556377289794, | |
| "grad_norm": 0.3915541172027588, | |
| "learning_rate": 4.222954235808378e-05, | |
| "loss": 2.0476, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.699965866423939, | |
| "grad_norm": 0.39140835404396057, | |
| "learning_rate": 4.19943090428802e-05, | |
| "loss": 1.9244, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 0.7008760951188986, | |
| "grad_norm": 0.4446852505207062, | |
| "learning_rate": 4.175955846879151e-05, | |
| "loss": 2.1622, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.7017863238138582, | |
| "grad_norm": 0.4065980911254883, | |
| "learning_rate": 4.1525292589483843e-05, | |
| "loss": 1.9522, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.7026965525088178, | |
| "grad_norm": 0.41551730036735535, | |
| "learning_rate": 4.129151335458957e-05, | |
| "loss": 1.9784, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.7036067812037774, | |
| "grad_norm": 0.4159603416919708, | |
| "learning_rate": 4.105822270969102e-05, | |
| "loss": 2.0386, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 0.7045170098987371, | |
| "grad_norm": 0.4357737600803375, | |
| "learning_rate": 4.0825422596304396e-05, | |
| "loss": 2.1812, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.7054272385936967, | |
| "grad_norm": 0.43295785784721375, | |
| "learning_rate": 4.059311495186338e-05, | |
| "loss": 2.1118, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.7063374672886563, | |
| "grad_norm": 0.4281920790672302, | |
| "learning_rate": 4.036130170970341e-05, | |
| "loss": 2.1594, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.7072476959836159, | |
| "grad_norm": 0.45081713795661926, | |
| "learning_rate": 4.012998479904525e-05, | |
| "loss": 2.2, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.7081579246785755, | |
| "grad_norm": 0.4478646218776703, | |
| "learning_rate": 3.9899166144978904e-05, | |
| "loss": 2.1344, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.7090681533735351, | |
| "grad_norm": 0.44155117869377136, | |
| "learning_rate": 3.966884766844803e-05, | |
| "loss": 2.1462, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 0.7099783820684947, | |
| "grad_norm": 0.4430200755596161, | |
| "learning_rate": 3.943903128623335e-05, | |
| "loss": 2.0047, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.7108886107634543, | |
| "grad_norm": 0.43685182929039, | |
| "learning_rate": 3.920971891093718e-05, | |
| "loss": 1.9843, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.7117988394584139, | |
| "grad_norm": 0.44161438941955566, | |
| "learning_rate": 3.8980912450967366e-05, | |
| "loss": 2.02, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.7127090681533735, | |
| "grad_norm": 0.45051446557044983, | |
| "learning_rate": 3.875261381052121e-05, | |
| "loss": 2.0351, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.7136192968483331, | |
| "grad_norm": 0.47110506892204285, | |
| "learning_rate": 3.852482488956992e-05, | |
| "loss": 2.0375, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.7145295255432927, | |
| "grad_norm": 0.49640730023384094, | |
| "learning_rate": 3.829754758384262e-05, | |
| "loss": 2.301, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.7154397542382523, | |
| "grad_norm": 0.5000788569450378, | |
| "learning_rate": 3.807078378481059e-05, | |
| "loss": 2.3439, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.716349982933212, | |
| "grad_norm": 0.4984182119369507, | |
| "learning_rate": 3.784453537967161e-05, | |
| "loss": 2.1625, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 0.7172602116281716, | |
| "grad_norm": 0.4947352409362793, | |
| "learning_rate": 3.761880425133413e-05, | |
| "loss": 2.1349, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.7181704403231312, | |
| "grad_norm": 0.48531001806259155, | |
| "learning_rate": 3.7393592278401704e-05, | |
| "loss": 2.0905, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.7190806690180908, | |
| "grad_norm": 0.5409421324729919, | |
| "learning_rate": 3.7168901335157315e-05, | |
| "loss": 2.4218, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.7199908977130504, | |
| "grad_norm": 0.5206452012062073, | |
| "learning_rate": 3.694473329154778e-05, | |
| "loss": 1.98, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 0.72090112640801, | |
| "grad_norm": 0.5477956533432007, | |
| "learning_rate": 3.672109001316809e-05, | |
| "loss": 2.4692, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.7218113551029697, | |
| "grad_norm": 0.5402776002883911, | |
| "learning_rate": 3.649797336124615e-05, | |
| "loss": 2.0121, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 0.7227215837979293, | |
| "grad_norm": 0.529184877872467, | |
| "learning_rate": 3.6275385192627056e-05, | |
| "loss": 2.1043, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.7236318124928889, | |
| "grad_norm": 0.5657643675804138, | |
| "learning_rate": 3.6053327359757535e-05, | |
| "loss": 2.1002, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.7245420411878485, | |
| "grad_norm": 0.6207425594329834, | |
| "learning_rate": 3.583180171067101e-05, | |
| "loss": 2.3302, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.7254522698828081, | |
| "grad_norm": 0.6927235722541809, | |
| "learning_rate": 3.5610810088971625e-05, | |
| "loss": 2.446, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 0.7263624985777677, | |
| "grad_norm": 0.7733549475669861, | |
| "learning_rate": 3.5390354333819344e-05, | |
| "loss": 2.6638, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.9281876087188721, | |
| "learning_rate": 3.517043627991441e-05, | |
| "loss": 2.5248, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 0.7281829559676869, | |
| "grad_norm": 1.9464212656021118, | |
| "learning_rate": 3.4951057757482205e-05, | |
| "loss": 2.603, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.7290931846626465, | |
| "grad_norm": 0.42626360058784485, | |
| "learning_rate": 3.4732220592257946e-05, | |
| "loss": 2.5094, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.7300034133576061, | |
| "grad_norm": 0.42765355110168457, | |
| "learning_rate": 3.45139266054715e-05, | |
| "loss": 2.172, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.7309136420525657, | |
| "grad_norm": 0.42991819977760315, | |
| "learning_rate": 3.429617761383222e-05, | |
| "loss": 2.2523, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 0.7318238707475253, | |
| "grad_norm": 0.4252484142780304, | |
| "learning_rate": 3.40789754295139e-05, | |
| "loss": 2.4049, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.7327340994424849, | |
| "grad_norm": 0.4209078252315521, | |
| "learning_rate": 3.3862321860139576e-05, | |
| "loss": 2.4249, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.7336443281374445, | |
| "grad_norm": 0.4354207515716553, | |
| "learning_rate": 3.364621870876659e-05, | |
| "loss": 2.4072, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.7345545568324041, | |
| "grad_norm": 0.4264775514602661, | |
| "learning_rate": 3.343066777387148e-05, | |
| "loss": 2.3709, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.7354647855273637, | |
| "grad_norm": 0.41826266050338745, | |
| "learning_rate": 3.3215670849335155e-05, | |
| "loss": 2.2593, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.7363750142223233, | |
| "grad_norm": 0.4268350899219513, | |
| "learning_rate": 3.300122972442773e-05, | |
| "loss": 2.3383, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 0.7372852429172829, | |
| "grad_norm": 0.40746137499809265, | |
| "learning_rate": 3.278734618379402e-05, | |
| "loss": 2.1898, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.7381954716122425, | |
| "grad_norm": 0.4189973771572113, | |
| "learning_rate": 3.257402200743821e-05, | |
| "loss": 2.3329, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 0.7391057003072021, | |
| "grad_norm": 0.44023388624191284, | |
| "learning_rate": 3.2361258970709397e-05, | |
| "loss": 2.3947, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.7400159290021618, | |
| "grad_norm": 0.41705650091171265, | |
| "learning_rate": 3.21490588442868e-05, | |
| "loss": 2.261, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.7409261576971214, | |
| "grad_norm": 0.431820273399353, | |
| "learning_rate": 3.19374233941647e-05, | |
| "loss": 2.4248, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.741836386392081, | |
| "grad_norm": 0.4188806414604187, | |
| "learning_rate": 3.172635438163816e-05, | |
| "loss": 2.2794, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.7427466150870407, | |
| "grad_norm": 0.41158682107925415, | |
| "learning_rate": 3.1515853563288076e-05, | |
| "loss": 2.1274, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.7436568437820003, | |
| "grad_norm": 0.4038848876953125, | |
| "learning_rate": 3.130592269096671e-05, | |
| "loss": 2.0359, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 0.7445670724769599, | |
| "grad_norm": 0.4189314842224121, | |
| "learning_rate": 3.1096563511783014e-05, | |
| "loss": 2.1405, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.7454773011719195, | |
| "grad_norm": 0.39889493584632874, | |
| "learning_rate": 3.08877777680882e-05, | |
| "loss": 2.0187, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.7463875298668791, | |
| "grad_norm": 0.38899531960487366, | |
| "learning_rate": 3.0679567197461134e-05, | |
| "loss": 2.0957, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.7472977585618387, | |
| "grad_norm": 0.41540107131004333, | |
| "learning_rate": 3.047193353269382e-05, | |
| "loss": 2.1757, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 0.7482079872567983, | |
| "grad_norm": 0.41734689474105835, | |
| "learning_rate": 3.0264878501777306e-05, | |
| "loss": 2.0902, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.7491182159517579, | |
| "grad_norm": 0.44342851638793945, | |
| "learning_rate": 3.005840382788685e-05, | |
| "loss": 2.087, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 0.7500284446467175, | |
| "grad_norm": 0.430385559797287, | |
| "learning_rate": 2.9852511229367865e-05, | |
| "loss": 2.1539, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.7509386733416771, | |
| "grad_norm": 0.4446250796318054, | |
| "learning_rate": 2.9647202419721687e-05, | |
| "loss": 2.2294, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.7509386733416771, | |
| "eval_loss": 2.2184133529663086, | |
| "eval_runtime": 204.0107, | |
| "eval_samples_per_second": 9.073, | |
| "eval_steps_per_second": 4.539, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.7518489020366367, | |
| "grad_norm": 0.4406190812587738, | |
| "learning_rate": 2.944247910759097e-05, | |
| "loss": 2.163, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.7527591307315963, | |
| "grad_norm": 0.43820223212242126, | |
| "learning_rate": 2.9238342996745817e-05, | |
| "loss": 2.1481, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 0.7536693594265559, | |
| "grad_norm": 0.4559546709060669, | |
| "learning_rate": 2.9034795786069436e-05, | |
| "loss": 2.1513, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.7545795881215155, | |
| "grad_norm": 0.41954734921455383, | |
| "learning_rate": 2.8831839169543996e-05, | |
| "loss": 1.9627, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 0.7554898168164751, | |
| "grad_norm": 0.4652347266674042, | |
| "learning_rate": 2.862947483623659e-05, | |
| "loss": 2.1275, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.7564000455114347, | |
| "grad_norm": 0.454685777425766, | |
| "learning_rate": 2.8427704470285144e-05, | |
| "loss": 2.1943, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.7573102742063944, | |
| "grad_norm": 0.45541757345199585, | |
| "learning_rate": 2.8226529750884402e-05, | |
| "loss": 2.0786, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.758220502901354, | |
| "grad_norm": 0.46581897139549255, | |
| "learning_rate": 2.8025952352271958e-05, | |
| "loss": 2.1656, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 0.7591307315963136, | |
| "grad_norm": 0.47137272357940674, | |
| "learning_rate": 2.7825973943714335e-05, | |
| "loss": 2.1557, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.7600409602912732, | |
| "grad_norm": 0.462990403175354, | |
| "learning_rate": 2.7626596189492983e-05, | |
| "loss": 2.0477, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.7609511889862328, | |
| "grad_norm": 0.4760456085205078, | |
| "learning_rate": 2.7427820748890685e-05, | |
| "loss": 2.1504, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 0.7618614176811924, | |
| "grad_norm": 0.46250203251838684, | |
| "learning_rate": 2.7229649276177503e-05, | |
| "loss": 2.1027, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.762771646376152, | |
| "grad_norm": 0.48775410652160645, | |
| "learning_rate": 2.7032083420597e-05, | |
| "loss": 2.2122, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 0.7636818750711116, | |
| "grad_norm": 0.4987994432449341, | |
| "learning_rate": 2.683512482635281e-05, | |
| "loss": 2.1812, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 0.7645921037660712, | |
| "grad_norm": 0.5075133442878723, | |
| "learning_rate": 2.6638775132594553e-05, | |
| "loss": 2.2842, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.7655023324610308, | |
| "grad_norm": 0.49660152196884155, | |
| "learning_rate": 2.6443035973404496e-05, | |
| "loss": 2.1107, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 0.7664125611559904, | |
| "grad_norm": 0.5171281695365906, | |
| "learning_rate": 2.624790897778391e-05, | |
| "loss": 2.1871, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 0.76732278985095, | |
| "grad_norm": 0.5493224859237671, | |
| "learning_rate": 2.605339576963929e-05, | |
| "loss": 2.3867, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 0.7682330185459096, | |
| "grad_norm": 0.534160315990448, | |
| "learning_rate": 2.585949796776912e-05, | |
| "loss": 2.255, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 0.7691432472408692, | |
| "grad_norm": 0.5723463892936707, | |
| "learning_rate": 2.5666217185850262e-05, | |
| "loss": 2.321, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.7700534759358288, | |
| "grad_norm": 0.5741109848022461, | |
| "learning_rate": 2.5473555032424533e-05, | |
| "loss": 2.1497, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 0.7709637046307884, | |
| "grad_norm": 0.6545650959014893, | |
| "learning_rate": 2.528151311088537e-05, | |
| "loss": 2.3931, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 0.771873933325748, | |
| "grad_norm": 0.756486713886261, | |
| "learning_rate": 2.50900930194644e-05, | |
| "loss": 2.7378, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 0.7727841620207077, | |
| "grad_norm": 0.8643143773078918, | |
| "learning_rate": 2.4899296351218227e-05, | |
| "loss": 2.3271, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 0.7736943907156673, | |
| "grad_norm": 1.6659263372421265, | |
| "learning_rate": 2.4709124694015116e-05, | |
| "loss": 2.4264, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.774604619410627, | |
| "grad_norm": 0.4279872477054596, | |
| "learning_rate": 2.451957963052185e-05, | |
| "loss": 2.43, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 0.7755148481055866, | |
| "grad_norm": 0.41057080030441284, | |
| "learning_rate": 2.433066273819037e-05, | |
| "loss": 2.3079, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 0.7764250768005462, | |
| "grad_norm": 0.4251214563846588, | |
| "learning_rate": 2.4142375589244957e-05, | |
| "loss": 2.3776, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 0.7773353054955058, | |
| "grad_norm": 0.41808971762657166, | |
| "learning_rate": 2.3954719750668907e-05, | |
| "loss": 2.1376, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 0.7782455341904654, | |
| "grad_norm": 0.408611923456192, | |
| "learning_rate": 2.3767696784191463e-05, | |
| "loss": 2.2524, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.779155762885425, | |
| "grad_norm": 0.42068326473236084, | |
| "learning_rate": 2.3581308246275103e-05, | |
| "loss": 2.3363, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 0.7800659915803846, | |
| "grad_norm": 0.4118780493736267, | |
| "learning_rate": 2.339555568810221e-05, | |
| "loss": 2.1718, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 0.7809762202753442, | |
| "grad_norm": 0.42966723442077637, | |
| "learning_rate": 2.321044065556246e-05, | |
| "loss": 2.4884, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.7818864489703038, | |
| "grad_norm": 0.42242035269737244, | |
| "learning_rate": 2.302596468923981e-05, | |
| "loss": 2.3057, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 0.7827966776652634, | |
| "grad_norm": 0.41452938318252563, | |
| "learning_rate": 2.284212932439972e-05, | |
| "loss": 2.3253, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.783706906360223, | |
| "grad_norm": 0.39578771591186523, | |
| "learning_rate": 2.265893609097637e-05, | |
| "loss": 2.0894, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 0.7846171350551826, | |
| "grad_norm": 0.42091935873031616, | |
| "learning_rate": 2.247638651355991e-05, | |
| "loss": 2.3383, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 0.7855273637501422, | |
| "grad_norm": 0.4203489124774933, | |
| "learning_rate": 2.229448211138382e-05, | |
| "loss": 2.3549, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 0.7864375924451018, | |
| "grad_norm": 0.4158013164997101, | |
| "learning_rate": 2.211322439831218e-05, | |
| "loss": 2.1964, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.7873478211400614, | |
| "grad_norm": 0.4121032655239105, | |
| "learning_rate": 2.1932614882827197e-05, | |
| "loss": 2.2676, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.788258049835021, | |
| "grad_norm": 0.4054934084415436, | |
| "learning_rate": 2.1752655068016515e-05, | |
| "loss": 2.119, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 0.7891682785299806, | |
| "grad_norm": 0.4000680148601532, | |
| "learning_rate": 2.1573346451560794e-05, | |
| "loss": 2.0798, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 0.7900785072249402, | |
| "grad_norm": 0.4024758040904999, | |
| "learning_rate": 2.139469052572127e-05, | |
| "loss": 1.9786, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 0.7909887359198998, | |
| "grad_norm": 0.422947496175766, | |
| "learning_rate": 2.1216688777327154e-05, | |
| "loss": 2.077, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 0.7918989646148595, | |
| "grad_norm": 0.4274744689464569, | |
| "learning_rate": 2.1039342687763586e-05, | |
| "loss": 2.2014, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.7928091933098191, | |
| "grad_norm": 0.4078015387058258, | |
| "learning_rate": 2.0862653732958915e-05, | |
| "loss": 1.9707, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 0.7937194220047787, | |
| "grad_norm": 0.4366302788257599, | |
| "learning_rate": 2.0686623383372715e-05, | |
| "loss": 2.1645, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 0.7946296506997383, | |
| "grad_norm": 0.4288860559463501, | |
| "learning_rate": 2.051125310398353e-05, | |
| "loss": 2.0476, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 0.795539879394698, | |
| "grad_norm": 0.4501003623008728, | |
| "learning_rate": 2.03365443542764e-05, | |
| "loss": 2.1982, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 0.7964501080896575, | |
| "grad_norm": 0.4465639293193817, | |
| "learning_rate": 2.016249858823106e-05, | |
| "loss": 2.0277, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.7973603367846172, | |
| "grad_norm": 0.45942404866218567, | |
| "learning_rate": 1.998911725430963e-05, | |
| "loss": 2.1634, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 0.7982705654795768, | |
| "grad_norm": 0.4460104703903198, | |
| "learning_rate": 1.981640179544466e-05, | |
| "loss": 2.3188, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 0.7991807941745364, | |
| "grad_norm": 0.4495980143547058, | |
| "learning_rate": 1.964435364902705e-05, | |
| "loss": 2.1365, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 0.800091022869496, | |
| "grad_norm": 0.4429667294025421, | |
| "learning_rate": 1.947297424689414e-05, | |
| "loss": 2.1187, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 0.8010012515644556, | |
| "grad_norm": 0.4700337052345276, | |
| "learning_rate": 1.93022650153178e-05, | |
| "loss": 2.0205, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.8019114802594152, | |
| "grad_norm": 0.45045018196105957, | |
| "learning_rate": 1.913222737499243e-05, | |
| "loss": 2.0114, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 0.8028217089543748, | |
| "grad_norm": 0.441072016954422, | |
| "learning_rate": 1.8962862741023423e-05, | |
| "loss": 1.9489, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 0.8037319376493344, | |
| "grad_norm": 0.4684150516986847, | |
| "learning_rate": 1.879417252291502e-05, | |
| "loss": 2.1999, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 0.804642166344294, | |
| "grad_norm": 0.4659992754459381, | |
| "learning_rate": 1.8626158124558858e-05, | |
| "loss": 2.1043, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 0.8055523950392536, | |
| "grad_norm": 0.49253416061401367, | |
| "learning_rate": 1.8458820944222255e-05, | |
| "loss": 2.2101, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.8064626237342132, | |
| "grad_norm": 0.4710872769355774, | |
| "learning_rate": 1.829216237453637e-05, | |
| "loss": 2.1022, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 0.8073728524291728, | |
| "grad_norm": 0.4951106309890747, | |
| "learning_rate": 1.8126183802484865e-05, | |
| "loss": 2.2409, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 0.8082830811241324, | |
| "grad_norm": 0.48583388328552246, | |
| "learning_rate": 1.7960886609392214e-05, | |
| "loss": 2.0319, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 0.8091933098190921, | |
| "grad_norm": 0.4637562930583954, | |
| "learning_rate": 1.7796272170912253e-05, | |
| "loss": 1.8175, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 0.8101035385140517, | |
| "grad_norm": 0.5015937089920044, | |
| "learning_rate": 1.763234185701673e-05, | |
| "loss": 2.2035, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.8110137672090113, | |
| "grad_norm": 0.5073252320289612, | |
| "learning_rate": 1.7469097031983893e-05, | |
| "loss": 2.0869, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 0.8119239959039709, | |
| "grad_norm": 0.5104179978370667, | |
| "learning_rate": 1.730653905438714e-05, | |
| "loss": 2.0663, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 0.8128342245989305, | |
| "grad_norm": 0.5239804983139038, | |
| "learning_rate": 1.7144669277083712e-05, | |
| "loss": 2.1529, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 0.8137444532938901, | |
| "grad_norm": 0.5428870320320129, | |
| "learning_rate": 1.6983489047203483e-05, | |
| "loss": 2.0934, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 0.8146546819888497, | |
| "grad_norm": 0.5535885691642761, | |
| "learning_rate": 1.6822999706137567e-05, | |
| "loss": 2.0925, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.8155649106838093, | |
| "grad_norm": 0.6293628811836243, | |
| "learning_rate": 1.6663202589527473e-05, | |
| "loss": 2.3585, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 0.8164751393787689, | |
| "grad_norm": 0.7003004550933838, | |
| "learning_rate": 1.6504099027253706e-05, | |
| "loss": 2.4135, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 0.8173853680737285, | |
| "grad_norm": 0.7463914155960083, | |
| "learning_rate": 1.634569034342476e-05, | |
| "loss": 2.578, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 0.8182955967686881, | |
| "grad_norm": 0.8569872975349426, | |
| "learning_rate": 1.6187977856366253e-05, | |
| "loss": 2.5576, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 0.8192058254636477, | |
| "grad_norm": 1.334454894065857, | |
| "learning_rate": 1.6030962878609725e-05, | |
| "loss": 2.4123, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.8201160541586073, | |
| "grad_norm": 0.4524904489517212, | |
| "learning_rate": 1.587464671688187e-05, | |
| "loss": 2.4774, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 0.8210262828535669, | |
| "grad_norm": 0.42647942900657654, | |
| "learning_rate": 1.5719030672093717e-05, | |
| "loss": 2.3709, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 0.8219365115485265, | |
| "grad_norm": 0.41805699467658997, | |
| "learning_rate": 1.5564116039329545e-05, | |
| "loss": 2.2022, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 0.8228467402434861, | |
| "grad_norm": 0.4297042489051819, | |
| "learning_rate": 1.5409904107836358e-05, | |
| "loss": 2.2815, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 0.8237569689384457, | |
| "grad_norm": 0.4077490568161011, | |
| "learning_rate": 1.5256396161013075e-05, | |
| "loss": 2.332, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.8246671976334053, | |
| "grad_norm": 0.42278701066970825, | |
| "learning_rate": 1.5103593476399791e-05, | |
| "loss": 2.3218, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 0.825577426328365, | |
| "grad_norm": 0.4234485626220703, | |
| "learning_rate": 1.495149732566723e-05, | |
| "loss": 2.2363, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 0.8264876550233247, | |
| "grad_norm": 0.410879909992218, | |
| "learning_rate": 1.4800108974606119e-05, | |
| "loss": 2.2881, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 0.8273978837182843, | |
| "grad_norm": 0.4263591170310974, | |
| "learning_rate": 1.4649429683116644e-05, | |
| "loss": 2.1476, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 0.8283081124132439, | |
| "grad_norm": 0.4317147433757782, | |
| "learning_rate": 1.4499460705197998e-05, | |
| "loss": 2.2372, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.8292183411082035, | |
| "grad_norm": 0.423459529876709, | |
| "learning_rate": 1.4350203288937936e-05, | |
| "loss": 2.3593, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 0.8301285698031631, | |
| "grad_norm": 0.40977466106414795, | |
| "learning_rate": 1.4201658676502294e-05, | |
| "loss": 2.1842, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 0.8310387984981227, | |
| "grad_norm": 0.41937538981437683, | |
| "learning_rate": 1.4053828104124867e-05, | |
| "loss": 2.251, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 0.8319490271930823, | |
| "grad_norm": 0.407067209482193, | |
| "learning_rate": 1.3906712802096933e-05, | |
| "loss": 2.027, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 0.8328592558880419, | |
| "grad_norm": 0.42164286971092224, | |
| "learning_rate": 1.3760313994757001e-05, | |
| "loss": 2.2382, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.8337694845830015, | |
| "grad_norm": 0.41477370262145996, | |
| "learning_rate": 1.361463290048085e-05, | |
| "loss": 2.0209, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 0.8346797132779611, | |
| "grad_norm": 0.39464712142944336, | |
| "learning_rate": 1.3469670731671046e-05, | |
| "loss": 2.063, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 0.8355899419729207, | |
| "grad_norm": 0.4041547179222107, | |
| "learning_rate": 1.3325428694747177e-05, | |
| "loss": 2.0052, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 0.8365001706678803, | |
| "grad_norm": 0.40053966641426086, | |
| "learning_rate": 1.3181907990135622e-05, | |
| "loss": 2.0691, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 0.8374103993628399, | |
| "grad_norm": 0.4019843637943268, | |
| "learning_rate": 1.3039109812259598e-05, | |
| "loss": 2.0359, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.8383206280577995, | |
| "grad_norm": 0.42373621463775635, | |
| "learning_rate": 1.2897035349529263e-05, | |
| "loss": 2.0603, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 0.8392308567527591, | |
| "grad_norm": 0.42414146661758423, | |
| "learning_rate": 1.2755685784331783e-05, | |
| "loss": 2.0417, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 0.8401410854477187, | |
| "grad_norm": 0.42916223406791687, | |
| "learning_rate": 1.2615062293021507e-05, | |
| "loss": 2.0517, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 0.8410513141426783, | |
| "grad_norm": 0.44549453258514404, | |
| "learning_rate": 1.2475166045910159e-05, | |
| "loss": 2.2541, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 0.8419615428376379, | |
| "grad_norm": 0.4355100393295288, | |
| "learning_rate": 1.2335998207257137e-05, | |
| "loss": 2.1341, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.8428717715325975, | |
| "grad_norm": 0.447841078042984, | |
| "learning_rate": 1.2197559935259795e-05, | |
| "loss": 2.2051, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 0.8437820002275572, | |
| "grad_norm": 0.4368191659450531, | |
| "learning_rate": 1.20598523820438e-05, | |
| "loss": 1.8763, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 0.8446922289225168, | |
| "grad_norm": 0.4553552269935608, | |
| "learning_rate": 1.1922876693653585e-05, | |
| "loss": 2.0406, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 0.8456024576174764, | |
| "grad_norm": 0.4466502070426941, | |
| "learning_rate": 1.1786634010042719e-05, | |
| "loss": 1.9599, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 0.846512686312436, | |
| "grad_norm": 0.43716827034950256, | |
| "learning_rate": 1.1651125465064516e-05, | |
| "loss": 2.0089, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.8474229150073956, | |
| "grad_norm": 0.4619704782962799, | |
| "learning_rate": 1.1516352186462586e-05, | |
| "loss": 2.0697, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.8483331437023552, | |
| "grad_norm": 0.44224247336387634, | |
| "learning_rate": 1.13823152958614e-05, | |
| "loss": 1.8982, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 0.8492433723973148, | |
| "grad_norm": 0.45818379521369934, | |
| "learning_rate": 1.1249015908756998e-05, | |
| "loss": 1.9608, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 0.8501536010922744, | |
| "grad_norm": 0.4756350517272949, | |
| "learning_rate": 1.1116455134507664e-05, | |
| "loss": 2.0776, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 0.851063829787234, | |
| "grad_norm": 0.46378758549690247, | |
| "learning_rate": 1.098463407632474e-05, | |
| "loss": 2.0733, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.8519740584821937, | |
| "grad_norm": 0.4653812050819397, | |
| "learning_rate": 1.0853553831263418e-05, | |
| "loss": 2.0823, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 0.8528842871771533, | |
| "grad_norm": 0.4805110991001129, | |
| "learning_rate": 1.0723215490213634e-05, | |
| "loss": 2.1113, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 0.8537945158721129, | |
| "grad_norm": 0.49554312229156494, | |
| "learning_rate": 1.0593620137890948e-05, | |
| "loss": 2.2246, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 0.8547047445670725, | |
| "grad_norm": 0.5131683349609375, | |
| "learning_rate": 1.0464768852827545e-05, | |
| "loss": 2.1736, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 0.8556149732620321, | |
| "grad_norm": 0.5109636187553406, | |
| "learning_rate": 1.0336662707363287e-05, | |
| "loss": 2.1027, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.8565252019569917, | |
| "grad_norm": 0.5181056261062622, | |
| "learning_rate": 1.0209302767636664e-05, | |
| "loss": 2.2091, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 0.8574354306519513, | |
| "grad_norm": 0.542943000793457, | |
| "learning_rate": 1.0082690093576163e-05, | |
| "loss": 2.2451, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 0.8583456593469109, | |
| "grad_norm": 0.5382577180862427, | |
| "learning_rate": 9.95682573889114e-06, | |
| "loss": 2.239, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 0.8592558880418705, | |
| "grad_norm": 0.5316376686096191, | |
| "learning_rate": 9.831710751063283e-06, | |
| "loss": 1.978, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 0.8601661167368301, | |
| "grad_norm": 0.5699295997619629, | |
| "learning_rate": 9.707346171337894e-06, | |
| "loss": 2.293, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.8610763454317898, | |
| "grad_norm": 0.6195168495178223, | |
| "learning_rate": 9.583733034714981e-06, | |
| "loss": 2.3735, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 0.8619865741267494, | |
| "grad_norm": 0.6604196429252625, | |
| "learning_rate": 9.460872369940955e-06, | |
| "loss": 2.4664, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 0.862896802821709, | |
| "grad_norm": 0.7245504856109619, | |
| "learning_rate": 9.338765199499854e-06, | |
| "loss": 2.4597, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 0.8638070315166686, | |
| "grad_norm": 0.8482882976531982, | |
| "learning_rate": 9.217412539604942e-06, | |
| "loss": 2.6484, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 0.8647172602116282, | |
| "grad_norm": 1.452289342880249, | |
| "learning_rate": 9.096815400190172e-06, | |
| "loss": 2.4219, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.8656274889065878, | |
| "grad_norm": 0.43866923451423645, | |
| "learning_rate": 8.97697478490188e-06, | |
| "loss": 2.5448, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 0.8665377176015474, | |
| "grad_norm": 0.4115692377090454, | |
| "learning_rate": 8.857891691090337e-06, | |
| "loss": 2.3454, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 0.867447946296507, | |
| "grad_norm": 0.4357510507106781, | |
| "learning_rate": 8.739567109801494e-06, | |
| "loss": 2.3642, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 0.8683581749914666, | |
| "grad_norm": 0.4002648890018463, | |
| "learning_rate": 8.62200202576875e-06, | |
| "loss": 2.2382, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 0.8692684036864262, | |
| "grad_norm": 0.41515567898750305, | |
| "learning_rate": 8.505197417404687e-06, | |
| "loss": 2.1776, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.8701786323813858, | |
| "grad_norm": 0.43014460802078247, | |
| "learning_rate": 8.38915425679304e-06, | |
| "loss": 2.4619, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 0.8710888610763454, | |
| "grad_norm": 0.42643028497695923, | |
| "learning_rate": 8.273873509680519e-06, | |
| "loss": 2.43, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 0.871999089771305, | |
| "grad_norm": 0.43487077951431274, | |
| "learning_rate": 8.15935613546872e-06, | |
| "loss": 2.3027, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 0.8729093184662646, | |
| "grad_norm": 0.42600685358047485, | |
| "learning_rate": 8.045603087206388e-06, | |
| "loss": 2.2503, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 0.8738195471612242, | |
| "grad_norm": 0.41962459683418274, | |
| "learning_rate": 7.932615311581126e-06, | |
| "loss": 2.2857, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.8747297758561838, | |
| "grad_norm": 0.4238678812980652, | |
| "learning_rate": 7.820393748911791e-06, | |
| "loss": 2.2728, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 0.8756400045511434, | |
| "grad_norm": 0.40477514266967773, | |
| "learning_rate": 7.708939333140642e-06, | |
| "loss": 2.2017, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 0.876550233246103, | |
| "grad_norm": 0.4259748160839081, | |
| "learning_rate": 7.598252991825372e-06, | |
| "loss": 2.1998, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 0.8774604619410626, | |
| "grad_norm": 0.4023091793060303, | |
| "learning_rate": 7.488335646131628e-06, | |
| "loss": 2.1225, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 0.8783706906360224, | |
| "grad_norm": 0.4074653387069702, | |
| "learning_rate": 7.3791882108251945e-06, | |
| "loss": 2.0988, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.879280919330982, | |
| "grad_norm": 0.40571245551109314, | |
| "learning_rate": 7.270811594264437e-06, | |
| "loss": 2.1729, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 0.8801911480259416, | |
| "grad_norm": 0.3894023895263672, | |
| "learning_rate": 7.163206698392744e-06, | |
| "loss": 2.0464, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 0.8811013767209012, | |
| "grad_norm": 0.40034767985343933, | |
| "learning_rate": 7.056374418730971e-06, | |
| "loss": 2.1144, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 0.8820116054158608, | |
| "grad_norm": 0.3878059387207031, | |
| "learning_rate": 6.950315644370075e-06, | |
| "loss": 1.8818, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 0.8829218341108204, | |
| "grad_norm": 0.38767677545547485, | |
| "learning_rate": 6.845031257963619e-06, | |
| "loss": 2.0166, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.88383206280578, | |
| "grad_norm": 0.40807151794433594, | |
| "learning_rate": 6.740522135720517e-06, | |
| "loss": 2.2045, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 0.8847422915007396, | |
| "grad_norm": 0.41804906725883484, | |
| "learning_rate": 6.636789147397637e-06, | |
| "loss": 2.1742, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 0.8856525201956992, | |
| "grad_norm": 0.420526921749115, | |
| "learning_rate": 6.533833156292679e-06, | |
| "loss": 1.9607, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 0.8865627488906588, | |
| "grad_norm": 0.42398571968078613, | |
| "learning_rate": 6.431655019236948e-06, | |
| "loss": 2.1077, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 0.8874729775856184, | |
| "grad_norm": 0.43108931183815, | |
| "learning_rate": 6.3302555865880965e-06, | |
| "loss": 2.0987, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.888383206280578, | |
| "grad_norm": 0.4478430151939392, | |
| "learning_rate": 6.229635702223324e-06, | |
| "loss": 2.2323, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 0.8892934349755376, | |
| "grad_norm": 0.4497049152851105, | |
| "learning_rate": 6.129796203532057e-06, | |
| "loss": 2.2498, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 0.8902036636704972, | |
| "grad_norm": 0.45716363191604614, | |
| "learning_rate": 6.030737921409169e-06, | |
| "loss": 2.1167, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 0.8911138923654568, | |
| "grad_norm": 0.4461272060871124, | |
| "learning_rate": 5.932461680248014e-06, | |
| "loss": 2.0598, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 0.8920241210604164, | |
| "grad_norm": 0.41892901062965393, | |
| "learning_rate": 5.834968297933541e-06, | |
| "loss": 2.0647, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.892934349755376, | |
| "grad_norm": 0.43655654788017273, | |
| "learning_rate": 5.738258585835532e-06, | |
| "loss": 1.9849, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 0.8938445784503356, | |
| "grad_norm": 0.45647647976875305, | |
| "learning_rate": 5.6423333488018095e-06, | |
| "loss": 2.1166, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 0.8947548071452952, | |
| "grad_norm": 0.46431490778923035, | |
| "learning_rate": 5.547193385151561e-06, | |
| "loss": 2.0456, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 0.8956650358402549, | |
| "grad_norm": 0.45279815793037415, | |
| "learning_rate": 5.45283948666866e-06, | |
| "loss": 2.0964, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 0.8965752645352145, | |
| "grad_norm": 0.47188904881477356, | |
| "learning_rate": 5.359272438595153e-06, | |
| "loss": 2.1394, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.8974854932301741, | |
| "grad_norm": 0.47204941511154175, | |
| "learning_rate": 5.266493019624663e-06, | |
| "loss": 2.0513, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 0.8983957219251337, | |
| "grad_norm": 0.46656250953674316, | |
| "learning_rate": 5.1745020018958866e-06, | |
| "loss": 1.9721, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 0.8993059506200933, | |
| "grad_norm": 0.47728633880615234, | |
| "learning_rate": 5.083300150986259e-06, | |
| "loss": 2.0223, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 0.9002161793150529, | |
| "grad_norm": 0.5116147398948669, | |
| "learning_rate": 4.992888225905468e-06, | |
| "loss": 2.108, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 0.9011264080100125, | |
| "grad_norm": 0.5168989300727844, | |
| "learning_rate": 4.903266979089249e-06, | |
| "loss": 2.1711, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.9020366367049721, | |
| "grad_norm": 0.5194457769393921, | |
| "learning_rate": 4.8144371563930476e-06, | |
| "loss": 2.2852, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 0.9029468653999317, | |
| "grad_norm": 0.5500402450561523, | |
| "learning_rate": 4.726399497085832e-06, | |
| "loss": 2.2758, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 0.9038570940948913, | |
| "grad_norm": 0.5453543066978455, | |
| "learning_rate": 4.6391547338439536e-06, | |
| "loss": 2.1761, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 0.904767322789851, | |
| "grad_norm": 0.5767526030540466, | |
| "learning_rate": 4.552703592745033e-06, | |
| "loss": 2.2521, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 0.9056775514848106, | |
| "grad_norm": 0.5596179962158203, | |
| "learning_rate": 4.467046793261931e-06, | |
| "loss": 2.2413, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.9065877801797702, | |
| "grad_norm": 0.610089123249054, | |
| "learning_rate": 4.3821850482567595e-06, | |
| "loss": 2.448, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 0.9074980088747298, | |
| "grad_norm": 0.6238061189651489, | |
| "learning_rate": 4.298119063974914e-06, | |
| "loss": 2.191, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 0.9084082375696894, | |
| "grad_norm": 0.7346171736717224, | |
| "learning_rate": 4.214849540039267e-06, | |
| "loss": 2.5612, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 0.909318466264649, | |
| "grad_norm": 0.8779885768890381, | |
| "learning_rate": 4.132377169444279e-06, | |
| "loss": 2.5268, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 0.9102286949596086, | |
| "grad_norm": 1.4997063875198364, | |
| "learning_rate": 4.050702638550275e-06, | |
| "loss": 2.3846, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.9111389236545682, | |
| "grad_norm": 0.4338892996311188, | |
| "learning_rate": 3.969826627077655e-06, | |
| "loss": 2.5665, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 0.9120491523495278, | |
| "grad_norm": 0.4329196810722351, | |
| "learning_rate": 3.889749808101395e-06, | |
| "loss": 2.3979, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 0.9129593810444874, | |
| "grad_norm": 0.43765294551849365, | |
| "learning_rate": 3.810472848045266e-06, | |
| "loss": 2.6044, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 0.9138696097394471, | |
| "grad_norm": 0.4189110994338989, | |
| "learning_rate": 3.7319964066763858e-06, | |
| "loss": 2.3877, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 0.9147798384344067, | |
| "grad_norm": 0.4291215240955353, | |
| "learning_rate": 3.6543211370997587e-06, | |
| "loss": 2.3827, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.9156900671293663, | |
| "grad_norm": 0.4351893365383148, | |
| "learning_rate": 3.5774476857527107e-06, | |
| "loss": 2.3855, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 0.9166002958243259, | |
| "grad_norm": 0.421740859746933, | |
| "learning_rate": 3.5013766923996604e-06, | |
| "loss": 2.2888, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 0.9175105245192855, | |
| "grad_norm": 0.41383060812950134, | |
| "learning_rate": 3.426108790126681e-06, | |
| "loss": 2.3313, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 0.9184207532142451, | |
| "grad_norm": 0.4178398549556732, | |
| "learning_rate": 3.3516446053363015e-06, | |
| "loss": 2.21, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 0.9193309819092047, | |
| "grad_norm": 0.4307008385658264, | |
| "learning_rate": 3.2779847577422697e-06, | |
| "loss": 2.2395, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.9202412106041643, | |
| "grad_norm": 0.4171479046344757, | |
| "learning_rate": 3.2051298603643753e-06, | |
| "loss": 2.0993, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 0.9211514392991239, | |
| "grad_norm": 0.4234906733036041, | |
| "learning_rate": 3.133080519523368e-06, | |
| "loss": 2.3472, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 0.9220616679940835, | |
| "grad_norm": 0.4258922040462494, | |
| "learning_rate": 3.0618373348359264e-06, | |
| "loss": 2.3281, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 0.9229718966890431, | |
| "grad_norm": 0.41497015953063965, | |
| "learning_rate": 2.991400899209651e-06, | |
| "loss": 2.11, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 0.9238821253840027, | |
| "grad_norm": 0.4151165187358856, | |
| "learning_rate": 2.921771798838069e-06, | |
| "loss": 2.0979, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.9247923540789623, | |
| "grad_norm": 0.4119180142879486, | |
| "learning_rate": 2.852950613195915e-06, | |
| "loss": 2.1067, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 0.9257025827739219, | |
| "grad_norm": 0.3973481357097626, | |
| "learning_rate": 2.784937915034169e-06, | |
| "loss": 2.1104, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 0.9266128114688815, | |
| "grad_norm": 0.41312044858932495, | |
| "learning_rate": 2.717734270375272e-06, | |
| "loss": 2.217, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 0.9275230401638411, | |
| "grad_norm": 0.3981812298297882, | |
| "learning_rate": 2.6513402385085704e-06, | |
| "loss": 2.035, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 0.9284332688588007, | |
| "grad_norm": 0.39964228868484497, | |
| "learning_rate": 2.585756371985493e-06, | |
| "loss": 1.9749, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.9293434975537603, | |
| "grad_norm": 0.4008193910121918, | |
| "learning_rate": 2.520983216615047e-06, | |
| "loss": 2.0877, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 0.9302537262487199, | |
| "grad_norm": 0.4075929820537567, | |
| "learning_rate": 2.4570213114592954e-06, | |
| "loss": 2.0702, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 0.9311639549436797, | |
| "grad_norm": 0.4271462857723236, | |
| "learning_rate": 2.393871188828767e-06, | |
| "loss": 2.1174, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 0.9320741836386393, | |
| "grad_norm": 0.410427063703537, | |
| "learning_rate": 2.3315333742780942e-06, | |
| "loss": 2.0849, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 0.9329844123335989, | |
| "grad_norm": 0.41201114654541016, | |
| "learning_rate": 2.270008386601685e-06, | |
| "loss": 2.0695, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.9338946410285585, | |
| "grad_norm": 0.41850075125694275, | |
| "learning_rate": 2.2092967378292915e-06, | |
| "loss": 2.0697, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 0.9348048697235181, | |
| "grad_norm": 0.429149866104126, | |
| "learning_rate": 2.1493989332218468e-06, | |
| "loss": 2.1187, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 0.9357150984184777, | |
| "grad_norm": 0.4181869328022003, | |
| "learning_rate": 2.0903154712672237e-06, | |
| "loss": 1.8432, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 0.9366253271134373, | |
| "grad_norm": 0.43963634967803955, | |
| "learning_rate": 2.032046843676061e-06, | |
| "loss": 2.2287, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 0.9375355558083969, | |
| "grad_norm": 0.4306661784648895, | |
| "learning_rate": 1.974593535377722e-06, | |
| "loss": 1.9793, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.9384457845033565, | |
| "grad_norm": 0.4563003480434418, | |
| "learning_rate": 1.917956024516243e-06, | |
| "loss": 1.932, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 0.9393560131983161, | |
| "grad_norm": 0.4930243492126465, | |
| "learning_rate": 1.8621347824462787e-06, | |
| "loss": 2.2763, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 0.9402662418932757, | |
| "grad_norm": 0.4600037932395935, | |
| "learning_rate": 1.8071302737293295e-06, | |
| "loss": 2.1954, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 0.47071364521980286, | |
| "learning_rate": 1.752942956129744e-06, | |
| "loss": 2.1367, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 0.9420866992831949, | |
| "grad_norm": 0.4597410559654236, | |
| "learning_rate": 1.6995732806109554e-06, | |
| "loss": 2.0846, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.9429969279781545, | |
| "grad_norm": 0.4716775715351105, | |
| "learning_rate": 1.6470216913317626e-06, | |
| "loss": 2.08, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 0.9439071566731141, | |
| "grad_norm": 0.4849373996257782, | |
| "learning_rate": 1.5952886256425547e-06, | |
| "loss": 2.1608, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 0.9448173853680737, | |
| "grad_norm": 0.5041067004203796, | |
| "learning_rate": 1.5443745140817366e-06, | |
| "loss": 2.2022, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 0.9457276140630333, | |
| "grad_norm": 0.503990888595581, | |
| "learning_rate": 1.4942797803721543e-06, | |
| "loss": 1.9996, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 0.9466378427579929, | |
| "grad_norm": 0.5079309940338135, | |
| "learning_rate": 1.4450048414174854e-06, | |
| "loss": 2.1168, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.9475480714529525, | |
| "grad_norm": 0.5187703371047974, | |
| "learning_rate": 1.3965501072988663e-06, | |
| "loss": 2.0721, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 0.9484583001479122, | |
| "grad_norm": 0.5252265930175781, | |
| "learning_rate": 1.348915981271437e-06, | |
| "loss": 2.1585, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 0.9493685288428718, | |
| "grad_norm": 0.5392215847969055, | |
| "learning_rate": 1.3021028597609675e-06, | |
| "loss": 2.2417, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 0.9502787575378314, | |
| "grad_norm": 0.5682265162467957, | |
| "learning_rate": 1.2561111323605712e-06, | |
| "loss": 2.1904, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 0.951188986232791, | |
| "grad_norm": 0.5829483270645142, | |
| "learning_rate": 1.2109411818274852e-06, | |
| "loss": 2.2045, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.9520992149277506, | |
| "grad_norm": 0.6208778023719788, | |
| "learning_rate": 1.1665933840798838e-06, | |
| "loss": 2.1755, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 0.9530094436227102, | |
| "grad_norm": 0.6663662195205688, | |
| "learning_rate": 1.1230681081936923e-06, | |
| "loss": 2.4264, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 0.9539196723176698, | |
| "grad_norm": 0.7317161560058594, | |
| "learning_rate": 1.0803657163995895e-06, | |
| "loss": 2.4505, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 0.9548299010126294, | |
| "grad_norm": 0.8815041184425354, | |
| "learning_rate": 1.0384865640799435e-06, | |
| "loss": 2.4768, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 0.955740129707589, | |
| "grad_norm": 1.4678566455841064, | |
| "learning_rate": 9.974309997658915e-07, | |
| "loss": 2.6858, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.9566503584025486, | |
| "grad_norm": 0.4141276180744171, | |
| "learning_rate": 9.57199365134387e-07, | |
| "loss": 2.4886, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 0.9575605870975082, | |
| "grad_norm": 0.43364688754081726, | |
| "learning_rate": 9.177919950054237e-07, | |
| "loss": 2.4077, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 0.9584708157924678, | |
| "grad_norm": 0.42708560824394226, | |
| "learning_rate": 8.792092173391831e-07, | |
| "loss": 2.3588, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 0.9593810444874274, | |
| "grad_norm": 0.4065101146697998, | |
| "learning_rate": 8.41451353233369e-07, | |
| "loss": 2.3254, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 0.960291273182387, | |
| "grad_norm": 0.42761117219924927, | |
| "learning_rate": 8.04518716920466e-07, | |
| "loss": 2.4209, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.9612015018773467, | |
| "grad_norm": 0.4246588349342346, | |
| "learning_rate": 7.684116157651966e-07, | |
| "loss": 2.3689, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 0.9621117305723063, | |
| "grad_norm": 0.4166822135448456, | |
| "learning_rate": 7.331303502618903e-07, | |
| "loss": 2.3942, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 0.9630219592672659, | |
| "grad_norm": 0.4337478578090668, | |
| "learning_rate": 6.986752140320518e-07, | |
| "loss": 2.2818, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 0.9639321879622255, | |
| "grad_norm": 0.4047817885875702, | |
| "learning_rate": 6.650464938218637e-07, | |
| "loss": 2.2848, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 0.9648424166571851, | |
| "grad_norm": 0.4132625162601471, | |
| "learning_rate": 6.322444694998319e-07, | |
| "loss": 2.3324, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.9657526453521448, | |
| "grad_norm": 0.4332369863986969, | |
| "learning_rate": 6.002694140544329e-07, | |
| "loss": 2.2694, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 0.9666628740471044, | |
| "grad_norm": 0.4239719808101654, | |
| "learning_rate": 5.691215935918815e-07, | |
| "loss": 2.1016, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 0.967573102742064, | |
| "grad_norm": 0.40531545877456665, | |
| "learning_rate": 5.388012673338661e-07, | |
| "loss": 2.1497, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 0.9684833314370236, | |
| "grad_norm": 0.4235825538635254, | |
| "learning_rate": 5.093086876154174e-07, | |
| "loss": 2.2406, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 0.9693935601319832, | |
| "grad_norm": 0.4083643853664398, | |
| "learning_rate": 4.80644099882821e-07, | |
| "loss": 2.188, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.9703037888269428, | |
| "grad_norm": 0.41800832748413086, | |
| "learning_rate": 4.5280774269154115e-07, | |
| "loss": 2.1485, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 0.9712140175219024, | |
| "grad_norm": 0.4159460663795471, | |
| "learning_rate": 4.2579984770426686e-07, | |
| "loss": 2.0976, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 0.972124246216862, | |
| "grad_norm": 0.39471137523651123, | |
| "learning_rate": 3.99620639688969e-07, | |
| "loss": 1.9677, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 0.9730344749118216, | |
| "grad_norm": 0.414125919342041, | |
| "learning_rate": 3.742703365170241e-07, | |
| "loss": 2.0815, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 0.9739447036067812, | |
| "grad_norm": 0.4122181236743927, | |
| "learning_rate": 3.497491491614158e-07, | |
| "loss": 1.935, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.9748549323017408, | |
| "grad_norm": 0.4170084595680237, | |
| "learning_rate": 3.260572816949692e-07, | |
| "loss": 2.1813, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 0.9757651609967004, | |
| "grad_norm": 0.41518816351890564, | |
| "learning_rate": 3.0319493128866396e-07, | |
| "loss": 2.0718, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 0.97667538969166, | |
| "grad_norm": 0.4334098696708679, | |
| "learning_rate": 2.8116228820997957e-07, | |
| "loss": 2.263, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 0.9775856183866196, | |
| "grad_norm": 0.4226486086845398, | |
| "learning_rate": 2.5995953582130804e-07, | |
| "loss": 2.1012, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 0.9784958470815792, | |
| "grad_norm": 0.43473172187805176, | |
| "learning_rate": 2.395868505784438e-07, | |
| "loss": 2.1051, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.9794060757765388, | |
| "grad_norm": 0.43599316477775574, | |
| "learning_rate": 2.2004440202911814e-07, | |
| "loss": 2.1165, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 0.9803163044714984, | |
| "grad_norm": 0.4303858280181885, | |
| "learning_rate": 2.0133235281156736e-07, | |
| "loss": 2.0183, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 0.981226533166458, | |
| "grad_norm": 0.4305777847766876, | |
| "learning_rate": 1.83450858653178e-07, | |
| "loss": 2.0601, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 0.9821367618614176, | |
| "grad_norm": 0.43362459540367126, | |
| "learning_rate": 1.664000683692324e-07, | |
| "loss": 1.9406, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 0.9830469905563773, | |
| "grad_norm": 0.4709031283855438, | |
| "learning_rate": 1.5018012386162072e-07, | |
| "loss": 2.3195, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.983957219251337, | |
| "grad_norm": 0.43655630946159363, | |
| "learning_rate": 1.3479116011769767e-07, | |
| "loss": 2.0253, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 0.9848674479462965, | |
| "grad_norm": 0.42346787452697754, | |
| "learning_rate": 1.2023330520911646e-07, | |
| "loss": 1.8358, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 0.9857776766412562, | |
| "grad_norm": 0.455380916595459, | |
| "learning_rate": 1.0650668029079658e-07, | |
| "loss": 2.096, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 0.9866879053362158, | |
| "grad_norm": 0.48610299825668335, | |
| "learning_rate": 9.361139959993549e-08, | |
| "loss": 2.1442, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 0.9875981340311754, | |
| "grad_norm": 0.47684475779533386, | |
| "learning_rate": 8.154757045497619e-08, | |
| "loss": 2.108, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.988508362726135, | |
| "grad_norm": 0.45393744111061096, | |
| "learning_rate": 7.0315293254819e-08, | |
| "loss": 1.9724, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 0.9894185914210946, | |
| "grad_norm": 0.4876580238342285, | |
| "learning_rate": 5.991466147791113e-08, | |
| "loss": 2.0793, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 0.9903288201160542, | |
| "grad_norm": 0.5055583119392395, | |
| "learning_rate": 5.0345761681491746e-08, | |
| "loss": 2.1826, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 0.9912390488110138, | |
| "grad_norm": 0.5002666115760803, | |
| "learning_rate": 4.1608673500859175e-08, | |
| "loss": 2.1531, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 0.9921492775059734, | |
| "grad_norm": 0.503621518611908, | |
| "learning_rate": 3.370346964876036e-08, | |
| "loss": 2.035, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.993059506200933, | |
| "grad_norm": 0.5054223537445068, | |
| "learning_rate": 2.6630215914702495e-08, | |
| "loss": 2.0657, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 0.9939697348958926, | |
| "grad_norm": 0.5747066140174866, | |
| "learning_rate": 2.038897116447558e-08, | |
| "loss": 2.3719, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 0.9948799635908522, | |
| "grad_norm": 0.5756259560585022, | |
| "learning_rate": 1.4979787339619578e-08, | |
| "loss": 2.1818, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 0.9957901922858118, | |
| "grad_norm": 0.6113715767860413, | |
| "learning_rate": 1.0402709457035808e-08, | |
| "loss": 2.3374, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 0.9967004209807714, | |
| "grad_norm": 0.6483973264694214, | |
| "learning_rate": 6.657775608553962e-09, | |
| "loss": 2.307, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.997610649675731, | |
| "grad_norm": 0.7209523320198059, | |
| "learning_rate": 3.745016960665648e-09, | |
| "loss": 2.5389, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 0.9985208783706906, | |
| "grad_norm": 0.7734662294387817, | |
| "learning_rate": 1.6644577542357375e-09, | |
| "loss": 2.4849, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 0.9994311070656502, | |
| "grad_norm": 1.1054223775863647, | |
| "learning_rate": 4.1611530431362453e-10, | |
| "loss": 2.4941, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 1.0006826715212198, | |
| "grad_norm": 3.6098527908325195, | |
| "learning_rate": 0.0, | |
| "loss": 4.772, | |
| "step": 1099 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1099, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 275, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.0148357964895355e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |