| { | |
| "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": 0.419871062040329, | |
| "learning_rate": 2e-05, | |
| "loss": 2.8811, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0009102286949596086, | |
| "eval_loss": 2.6547484397888184, | |
| "eval_runtime": 203.8787, | |
| "eval_samples_per_second": 9.079, | |
| "eval_steps_per_second": 4.542, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0018204573899192173, | |
| "grad_norm": 0.47298941016197205, | |
| "learning_rate": 4e-05, | |
| "loss": 2.8541, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0027306860848788257, | |
| "grad_norm": 0.48272326588630676, | |
| "learning_rate": 6e-05, | |
| "loss": 2.6363, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0036409147798384346, | |
| "grad_norm": 0.4156210422515869, | |
| "learning_rate": 8e-05, | |
| "loss": 2.666, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.004551143474798043, | |
| "grad_norm": 0.375728964805603, | |
| "learning_rate": 0.0001, | |
| "loss": 2.7432, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0054613721697576514, | |
| "grad_norm": 0.3113997280597687, | |
| "learning_rate": 0.00012, | |
| "loss": 2.7398, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.00637160086471726, | |
| "grad_norm": 0.4520895183086395, | |
| "learning_rate": 0.00014, | |
| "loss": 2.7017, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.007281829559676869, | |
| "grad_norm": 0.5530128479003906, | |
| "learning_rate": 0.00016, | |
| "loss": 2.6855, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.008192058254636477, | |
| "grad_norm": 0.43422171473503113, | |
| "learning_rate": 0.00018, | |
| "loss": 2.6454, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.009102286949596085, | |
| "grad_norm": 0.3173878490924835, | |
| "learning_rate": 0.0002, | |
| "loss": 2.6638, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.010012515644555695, | |
| "grad_norm": 0.3479019105434418, | |
| "learning_rate": 0.00019999958388469571, | |
| "loss": 2.5689, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.010922744339515303, | |
| "grad_norm": 0.36122018098831177, | |
| "learning_rate": 0.00019999833554224577, | |
| "loss": 2.4933, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.011832973034474913, | |
| "grad_norm": 0.3206821382045746, | |
| "learning_rate": 0.00019999625498303932, | |
| "loss": 2.5661, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.01274320172943452, | |
| "grad_norm": 0.38150081038475037, | |
| "learning_rate": 0.00019999334222439147, | |
| "loss": 2.5545, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.013653430424394129, | |
| "grad_norm": 0.3939971625804901, | |
| "learning_rate": 0.00019998959729054295, | |
| "loss": 2.5737, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.014563659119353738, | |
| "grad_norm": 0.3015269339084625, | |
| "learning_rate": 0.0001999850202126604, | |
| "loss": 2.524, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.015473887814313346, | |
| "grad_norm": 0.3177410662174225, | |
| "learning_rate": 0.00019997961102883552, | |
| "loss": 2.4203, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.016384116509272954, | |
| "grad_norm": 0.3253481388092041, | |
| "learning_rate": 0.00019997336978408531, | |
| "loss": 2.5163, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.017294345204232564, | |
| "grad_norm": 0.3166387975215912, | |
| "learning_rate": 0.00019996629653035126, | |
| "loss": 2.2021, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.01820457389919217, | |
| "grad_norm": 0.32843250036239624, | |
| "learning_rate": 0.00019995839132649917, | |
| "loss": 2.4009, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01911480259415178, | |
| "grad_norm": 0.36471301317214966, | |
| "learning_rate": 0.00019994965423831854, | |
| "loss": 2.4734, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.02002503128911139, | |
| "grad_norm": 0.36941981315612793, | |
| "learning_rate": 0.0001999400853385221, | |
| "loss": 2.4731, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.020935259984071, | |
| "grad_norm": 0.3467971682548523, | |
| "learning_rate": 0.0001999296847067452, | |
| "loss": 2.4294, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.021845488679030606, | |
| "grad_norm": 0.337288498878479, | |
| "learning_rate": 0.00019991845242954505, | |
| "loss": 2.0426, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.022755717373990215, | |
| "grad_norm": 0.36415189504623413, | |
| "learning_rate": 0.00019990638860040006, | |
| "loss": 2.2803, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.023665946068949825, | |
| "grad_norm": 0.37340012192726135, | |
| "learning_rate": 0.00019989349331970923, | |
| "loss": 2.2841, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.02457617476390943, | |
| "grad_norm": 0.3849464952945709, | |
| "learning_rate": 0.00019987976669479088, | |
| "loss": 2.206, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.02548640345886904, | |
| "grad_norm": 0.41922691464424133, | |
| "learning_rate": 0.00019986520883988232, | |
| "loss": 2.3736, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.02639663215382865, | |
| "grad_norm": 0.41973674297332764, | |
| "learning_rate": 0.0001998498198761384, | |
| "loss": 2.3605, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.027306860848788257, | |
| "grad_norm": 0.3731151223182678, | |
| "learning_rate": 0.00019983359993163078, | |
| "loss": 2.0799, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.028217089543747867, | |
| "grad_norm": 0.41079527139663696, | |
| "learning_rate": 0.00019981654914134686, | |
| "loss": 2.3176, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.029127318238707477, | |
| "grad_norm": 0.43309271335601807, | |
| "learning_rate": 0.00019979866764718843, | |
| "loss": 2.3306, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.030037546933667083, | |
| "grad_norm": 0.45209765434265137, | |
| "learning_rate": 0.0001997799555979709, | |
| "loss": 2.3582, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.030947775628626693, | |
| "grad_norm": 0.4335600435733795, | |
| "learning_rate": 0.00019976041314942155, | |
| "loss": 2.1077, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0318580043235863, | |
| "grad_norm": 0.43455106019973755, | |
| "learning_rate": 0.0001997400404641787, | |
| "loss": 2.2497, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.03276823301854591, | |
| "grad_norm": 0.4201047718524933, | |
| "learning_rate": 0.00019971883771179003, | |
| "loss": 2.1547, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.03367846171350552, | |
| "grad_norm": 0.43784937262535095, | |
| "learning_rate": 0.00019969680506871137, | |
| "loss": 2.3883, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.03458869040846513, | |
| "grad_norm": 0.4830430746078491, | |
| "learning_rate": 0.00019967394271830504, | |
| "loss": 2.5713, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.03549891910342474, | |
| "grad_norm": 0.4729761779308319, | |
| "learning_rate": 0.00019965025085083858, | |
| "loss": 2.3565, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.03640914779838434, | |
| "grad_norm": 0.43940603733062744, | |
| "learning_rate": 0.000199625729663483, | |
| "loss": 2.1644, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03731937649334395, | |
| "grad_norm": 0.5039120316505432, | |
| "learning_rate": 0.00019960037936031104, | |
| "loss": 2.3641, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.03822960518830356, | |
| "grad_norm": 0.487318217754364, | |
| "learning_rate": 0.00019957420015229572, | |
| "loss": 2.4128, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.03913983388326317, | |
| "grad_norm": 0.5216456651687622, | |
| "learning_rate": 0.00019954719225730847, | |
| "loss": 2.3659, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.04005006257822278, | |
| "grad_norm": 0.4877496659755707, | |
| "learning_rate": 0.00019951935590011718, | |
| "loss": 2.2571, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.04096029127318239, | |
| "grad_norm": 0.4974242150783539, | |
| "learning_rate": 0.0001994906913123846, | |
| "loss": 2.1847, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.041870519968142, | |
| "grad_norm": 0.5708010196685791, | |
| "learning_rate": 0.00019946119873266613, | |
| "loss": 2.6954, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.0427807486631016, | |
| "grad_norm": 0.5892224311828613, | |
| "learning_rate": 0.00019943087840640814, | |
| "loss": 2.5149, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.04369097735806121, | |
| "grad_norm": 0.6740038394927979, | |
| "learning_rate": 0.0001993997305859456, | |
| "loss": 2.6992, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.04460120605302082, | |
| "grad_norm": 0.7869267463684082, | |
| "learning_rate": 0.0001993677555305002, | |
| "loss": 2.5103, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.04551143474798043, | |
| "grad_norm": 1.244491457939148, | |
| "learning_rate": 0.00019933495350617813, | |
| "loss": 2.5167, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04642166344294004, | |
| "grad_norm": 1.1311120986938477, | |
| "learning_rate": 0.00019930132478596796, | |
| "loss": 2.7147, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.04733189213789965, | |
| "grad_norm": 0.5104111433029175, | |
| "learning_rate": 0.00019926686964973813, | |
| "loss": 2.7642, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.04824212083285925, | |
| "grad_norm": 0.4251229465007782, | |
| "learning_rate": 0.00019923158838423482, | |
| "loss": 2.4562, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.04915234952781886, | |
| "grad_norm": 0.45901840925216675, | |
| "learning_rate": 0.00019919548128307954, | |
| "loss": 2.6106, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.05006257822277847, | |
| "grad_norm": 0.4177067279815674, | |
| "learning_rate": 0.00019915854864676664, | |
| "loss": 2.4952, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.05097280691773808, | |
| "grad_norm": 0.411510705947876, | |
| "learning_rate": 0.00019912079078266085, | |
| "loss": 2.6141, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.05188303561269769, | |
| "grad_norm": 0.42673829197883606, | |
| "learning_rate": 0.0001990822080049946, | |
| "loss": 2.5089, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.0527932643076573, | |
| "grad_norm": 0.3956531882286072, | |
| "learning_rate": 0.0001990428006348656, | |
| "loss": 2.4539, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.053703493002616905, | |
| "grad_norm": 0.3687964975833893, | |
| "learning_rate": 0.00019900256900023413, | |
| "loss": 2.3752, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.054613721697576514, | |
| "grad_norm": 0.34918078780174255, | |
| "learning_rate": 0.00019896151343592008, | |
| "loss": 2.2978, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.055523950392536124, | |
| "grad_norm": 0.3795469403266907, | |
| "learning_rate": 0.00019891963428360043, | |
| "loss": 2.4324, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.056434179087495734, | |
| "grad_norm": 0.37362977862358093, | |
| "learning_rate": 0.00019887693189180633, | |
| "loss": 2.3966, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.057344407782455344, | |
| "grad_norm": 0.3695915639400482, | |
| "learning_rate": 0.00019883340661592015, | |
| "loss": 2.3188, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.05825463647741495, | |
| "grad_norm": 0.376068651676178, | |
| "learning_rate": 0.00019878905881817252, | |
| "loss": 2.3933, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.059164865172374556, | |
| "grad_norm": 0.3845527470111847, | |
| "learning_rate": 0.00019874388886763944, | |
| "loss": 2.5612, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.060075093867334166, | |
| "grad_norm": 0.3678928315639496, | |
| "learning_rate": 0.00019869789714023906, | |
| "loss": 2.4174, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.060985322562293776, | |
| "grad_norm": 0.3534240126609802, | |
| "learning_rate": 0.00019865108401872857, | |
| "loss": 2.2843, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.061895551257253385, | |
| "grad_norm": 0.3559145927429199, | |
| "learning_rate": 0.00019860344989270113, | |
| "loss": 2.0911, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.062805779952213, | |
| "grad_norm": 0.3940396308898926, | |
| "learning_rate": 0.0001985549951585825, | |
| "loss": 2.2796, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.0637160086471726, | |
| "grad_norm": 0.39386069774627686, | |
| "learning_rate": 0.00019850572021962788, | |
| "loss": 2.1712, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06462623734213221, | |
| "grad_norm": 0.3923673629760742, | |
| "learning_rate": 0.00019845562548591826, | |
| "loss": 2.1665, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.06553646603709182, | |
| "grad_norm": 0.3916044533252716, | |
| "learning_rate": 0.00019840471137435746, | |
| "loss": 2.267, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.06644669473205143, | |
| "grad_norm": 0.4030390679836273, | |
| "learning_rate": 0.00019835297830866826, | |
| "loss": 2.2662, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.06735692342701104, | |
| "grad_norm": 0.4568188786506653, | |
| "learning_rate": 0.00019830042671938904, | |
| "loss": 2.4341, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.06826715212197064, | |
| "grad_norm": 0.4151420593261719, | |
| "learning_rate": 0.00019824705704387028, | |
| "loss": 2.2547, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.06917738081693026, | |
| "grad_norm": 0.4141649007797241, | |
| "learning_rate": 0.00019819286972627066, | |
| "loss": 2.157, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.07008760951188986, | |
| "grad_norm": 0.3980599343776703, | |
| "learning_rate": 0.00019813786521755372, | |
| "loss": 2.0681, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.07099783820684948, | |
| "grad_norm": 0.41139495372772217, | |
| "learning_rate": 0.00019808204397548377, | |
| "loss": 2.2297, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.07190806690180908, | |
| "grad_norm": 0.4236723482608795, | |
| "learning_rate": 0.0001980254064646223, | |
| "loss": 2.1775, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.07281829559676868, | |
| "grad_norm": 0.413194864988327, | |
| "learning_rate": 0.00019796795315632395, | |
| "loss": 2.1926, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0737285242917283, | |
| "grad_norm": 0.4185364246368408, | |
| "learning_rate": 0.0001979096845287328, | |
| "loss": 2.0727, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.0746387529866879, | |
| "grad_norm": 0.41983237862586975, | |
| "learning_rate": 0.00019785060106677818, | |
| "loss": 2.3075, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.07554898168164752, | |
| "grad_norm": 0.44238418340682983, | |
| "learning_rate": 0.00019779070326217074, | |
| "loss": 2.2824, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.07645921037660712, | |
| "grad_norm": 0.45874810218811035, | |
| "learning_rate": 0.00019772999161339833, | |
| "loss": 2.2601, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.07736943907156674, | |
| "grad_norm": 0.4516693651676178, | |
| "learning_rate": 0.00019766846662572191, | |
| "loss": 2.314, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.07827966776652634, | |
| "grad_norm": 0.43325892090797424, | |
| "learning_rate": 0.00019760612881117125, | |
| "loss": 2.1802, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.07918989646148594, | |
| "grad_norm": 0.4844377636909485, | |
| "learning_rate": 0.00019754297868854073, | |
| "loss": 2.3255, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.08010012515644556, | |
| "grad_norm": 0.44133615493774414, | |
| "learning_rate": 0.00019747901678338496, | |
| "loss": 2.2873, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.08101035385140516, | |
| "grad_norm": 0.5170537829399109, | |
| "learning_rate": 0.00019741424362801452, | |
| "loss": 2.3539, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.08192058254636478, | |
| "grad_norm": 0.46766072511672974, | |
| "learning_rate": 0.00019734865976149145, | |
| "loss": 2.2522, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.08283081124132438, | |
| "grad_norm": 0.4890718162059784, | |
| "learning_rate": 0.00019728226572962473, | |
| "loss": 2.4453, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.083741039936284, | |
| "grad_norm": 0.481082946062088, | |
| "learning_rate": 0.00019721506208496585, | |
| "loss": 2.2735, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.0846512686312436, | |
| "grad_norm": 0.4898594319820404, | |
| "learning_rate": 0.00019714704938680408, | |
| "loss": 2.223, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.0855614973262032, | |
| "grad_norm": 0.48203563690185547, | |
| "learning_rate": 0.00019707822820116193, | |
| "loss": 2.2461, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.08647172602116282, | |
| "grad_norm": 0.5068606734275818, | |
| "learning_rate": 0.00019700859910079036, | |
| "loss": 2.1878, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.08738195471612242, | |
| "grad_norm": 0.5902338027954102, | |
| "learning_rate": 0.00019693816266516407, | |
| "loss": 2.6452, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.08829218341108204, | |
| "grad_norm": 0.6118764281272888, | |
| "learning_rate": 0.00019686691948047664, | |
| "loss": 2.4923, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.08920241210604164, | |
| "grad_norm": 0.6774008274078369, | |
| "learning_rate": 0.00019679487013963564, | |
| "loss": 2.7092, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.09011264080100125, | |
| "grad_norm": 0.8197199106216431, | |
| "learning_rate": 0.00019672201524225776, | |
| "loss": 2.7662, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.09102286949596086, | |
| "grad_norm": 1.4529683589935303, | |
| "learning_rate": 0.0001966483553946637, | |
| "loss": 2.7347, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09193309819092046, | |
| "grad_norm": 1.3071473836898804, | |
| "learning_rate": 0.00019657389120987333, | |
| "loss": 2.8331, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.09284332688588008, | |
| "grad_norm": 0.5390121340751648, | |
| "learning_rate": 0.00019649862330760036, | |
| "loss": 2.724, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.09375355558083968, | |
| "grad_norm": 0.43149107694625854, | |
| "learning_rate": 0.00019642255231424729, | |
| "loss": 2.5445, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.0946637842757993, | |
| "grad_norm": 0.4529167115688324, | |
| "learning_rate": 0.00019634567886290025, | |
| "loss": 2.4586, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.0955740129707589, | |
| "grad_norm": 0.4370836317539215, | |
| "learning_rate": 0.00019626800359332362, | |
| "loss": 2.5472, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.0964842416657185, | |
| "grad_norm": 0.43084716796875, | |
| "learning_rate": 0.00019618952715195475, | |
| "loss": 2.5307, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.09739447036067812, | |
| "grad_norm": 0.4851230978965759, | |
| "learning_rate": 0.0001961102501918986, | |
| "loss": 2.3863, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.09830469905563773, | |
| "grad_norm": 0.39606913924217224, | |
| "learning_rate": 0.00019603017337292236, | |
| "loss": 2.4334, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.09921492775059734, | |
| "grad_norm": 0.3680226504802704, | |
| "learning_rate": 0.00019594929736144976, | |
| "loss": 2.3358, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.10012515644555695, | |
| "grad_norm": 0.43170568346977234, | |
| "learning_rate": 0.00019586762283055573, | |
| "loss": 2.3502, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.10103538514051655, | |
| "grad_norm": 0.4067252576351166, | |
| "learning_rate": 0.00019578515045996073, | |
| "loss": 2.4522, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.10194561383547616, | |
| "grad_norm": 0.3645339906215668, | |
| "learning_rate": 0.0001957018809360251, | |
| "loss": 2.3924, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.10285584253043577, | |
| "grad_norm": 0.4669637084007263, | |
| "learning_rate": 0.00019561781495174328, | |
| "loss": 2.3151, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.10376607122539538, | |
| "grad_norm": 0.39779335260391235, | |
| "learning_rate": 0.00019553295320673807, | |
| "loss": 2.286, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.10467629992035499, | |
| "grad_norm": 0.41239792108535767, | |
| "learning_rate": 0.00019544729640725498, | |
| "loss": 2.4233, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.1055865286153146, | |
| "grad_norm": 0.3813844621181488, | |
| "learning_rate": 0.0001953608452661561, | |
| "loss": 2.3497, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.1064967573102742, | |
| "grad_norm": 0.3577946424484253, | |
| "learning_rate": 0.0001952736005029142, | |
| "loss": 2.2817, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.10740698600523381, | |
| "grad_norm": 0.3643464744091034, | |
| "learning_rate": 0.00019518556284360696, | |
| "loss": 2.2296, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.10831721470019343, | |
| "grad_norm": 0.38557112216949463, | |
| "learning_rate": 0.00019509673302091075, | |
| "loss": 2.3039, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.10922744339515303, | |
| "grad_norm": 0.39795589447021484, | |
| "learning_rate": 0.00019500711177409454, | |
| "loss": 2.3028, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.11013767209011265, | |
| "grad_norm": 0.3612115681171417, | |
| "learning_rate": 0.00019491669984901379, | |
| "loss": 2.0678, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.11104790078507225, | |
| "grad_norm": 0.36631131172180176, | |
| "learning_rate": 0.00019482549799810413, | |
| "loss": 2.2505, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.11195812948003186, | |
| "grad_norm": 0.383908748626709, | |
| "learning_rate": 0.00019473350698037535, | |
| "loss": 2.1246, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.11286835817499147, | |
| "grad_norm": 0.4124641716480255, | |
| "learning_rate": 0.00019464072756140486, | |
| "loss": 2.2511, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.11377858686995107, | |
| "grad_norm": 0.3917684555053711, | |
| "learning_rate": 0.00019454716051333135, | |
| "loss": 2.2266, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.11468881556491069, | |
| "grad_norm": 0.40930524468421936, | |
| "learning_rate": 0.00019445280661484847, | |
| "loss": 2.1455, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.11559904425987029, | |
| "grad_norm": 0.4211941063404083, | |
| "learning_rate": 0.0001943576666511982, | |
| "loss": 2.1991, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.1165092729548299, | |
| "grad_norm": 0.4424692392349243, | |
| "learning_rate": 0.00019426174141416448, | |
| "loss": 2.1868, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.11741950164978951, | |
| "grad_norm": 0.41721439361572266, | |
| "learning_rate": 0.00019416503170206645, | |
| "loss": 2.3098, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.11832973034474911, | |
| "grad_norm": 0.40203043818473816, | |
| "learning_rate": 0.00019406753831975203, | |
| "loss": 2.1059, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.11923995903970873, | |
| "grad_norm": 0.4373216927051544, | |
| "learning_rate": 0.00019396926207859084, | |
| "loss": 2.0548, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.12015018773466833, | |
| "grad_norm": 0.46566876769065857, | |
| "learning_rate": 0.00019387020379646797, | |
| "loss": 2.2075, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.12106041642962795, | |
| "grad_norm": 0.4372277557849884, | |
| "learning_rate": 0.00019377036429777672, | |
| "loss": 2.1329, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.12197064512458755, | |
| "grad_norm": 0.42452472448349, | |
| "learning_rate": 0.0001936697444134119, | |
| "loss": 2.1308, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.12288087381954717, | |
| "grad_norm": 0.4379199743270874, | |
| "learning_rate": 0.0001935683449807631, | |
| "loss": 2.2978, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.12379110251450677, | |
| "grad_norm": 0.46032705903053284, | |
| "learning_rate": 0.0001934661668437073, | |
| "loss": 2.2349, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.12470133120946637, | |
| "grad_norm": 0.4644859731197357, | |
| "learning_rate": 0.00019336321085260236, | |
| "loss": 2.2485, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.125611559904426, | |
| "grad_norm": 0.4860181510448456, | |
| "learning_rate": 0.00019325947786427952, | |
| "loss": 2.2059, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.1265217885993856, | |
| "grad_norm": 0.46559709310531616, | |
| "learning_rate": 0.0001931549687420364, | |
| "loss": 2.2564, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.1274320172943452, | |
| "grad_norm": 0.47857019305229187, | |
| "learning_rate": 0.00019304968435562993, | |
| "loss": 2.1526, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.12834224598930483, | |
| "grad_norm": 0.5288405418395996, | |
| "learning_rate": 0.00019294362558126905, | |
| "loss": 2.399, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.12925247468426443, | |
| "grad_norm": 0.4651063084602356, | |
| "learning_rate": 0.00019283679330160726, | |
| "loss": 2.3092, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.13016270337922403, | |
| "grad_norm": 0.4940643310546875, | |
| "learning_rate": 0.00019272918840573558, | |
| "loss": 2.2838, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.13107293207418363, | |
| "grad_norm": 0.5267408490180969, | |
| "learning_rate": 0.00019262081178917482, | |
| "loss": 2.3948, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.13198316076914324, | |
| "grad_norm": 0.5252528190612793, | |
| "learning_rate": 0.0001925116643538684, | |
| "loss": 2.3055, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.13289338946410287, | |
| "grad_norm": 0.5565192699432373, | |
| "learning_rate": 0.00019240174700817464, | |
| "loss": 2.2392, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.13380361815906247, | |
| "grad_norm": 0.6170883774757385, | |
| "learning_rate": 0.00019229106066685937, | |
| "loss": 2.4212, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.13471384685402207, | |
| "grad_norm": 0.623285710811615, | |
| "learning_rate": 0.0001921796062510882, | |
| "loss": 2.5778, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.13562407554898168, | |
| "grad_norm": 0.7146514654159546, | |
| "learning_rate": 0.0001920673846884189, | |
| "loss": 2.5517, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.13653430424394128, | |
| "grad_norm": 1.1342458724975586, | |
| "learning_rate": 0.00019195439691279363, | |
| "loss": 2.7753, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1374445329389009, | |
| "grad_norm": 0.5309027433395386, | |
| "learning_rate": 0.00019184064386453128, | |
| "loss": 2.6207, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.1383547616338605, | |
| "grad_norm": 0.5145514607429504, | |
| "learning_rate": 0.00019172612649031952, | |
| "loss": 2.605, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.13926499032882012, | |
| "grad_norm": 0.41491782665252686, | |
| "learning_rate": 0.00019161084574320696, | |
| "loss": 2.3562, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.14017521902377972, | |
| "grad_norm": 0.4266800880432129, | |
| "learning_rate": 0.00019149480258259533, | |
| "loss": 2.5725, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.14108544771873932, | |
| "grad_norm": 0.3961375951766968, | |
| "learning_rate": 0.00019137799797423126, | |
| "loss": 2.4722, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.14199567641369895, | |
| "grad_norm": 0.41011351346969604, | |
| "learning_rate": 0.00019126043289019852, | |
| "loss": 2.5218, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.14290590510865855, | |
| "grad_norm": 0.4531441032886505, | |
| "learning_rate": 0.00019114210830890969, | |
| "loss": 2.5233, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.14381613380361816, | |
| "grad_norm": 0.447412371635437, | |
| "learning_rate": 0.00019102302521509815, | |
| "loss": 2.624, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.14472636249857776, | |
| "grad_norm": 0.41504496335983276, | |
| "learning_rate": 0.00019090318459980986, | |
| "loss": 2.5468, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.14563659119353736, | |
| "grad_norm": 0.3992937505245209, | |
| "learning_rate": 0.00019078258746039507, | |
| "loss": 2.5191, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.146546819888497, | |
| "grad_norm": 0.37622761726379395, | |
| "learning_rate": 0.00019066123480050015, | |
| "loss": 2.3164, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.1474570485834566, | |
| "grad_norm": 0.3903200626373291, | |
| "learning_rate": 0.00019053912763005907, | |
| "loss": 2.1591, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.1483672772784162, | |
| "grad_norm": 0.4018140733242035, | |
| "learning_rate": 0.00019041626696528503, | |
| "loss": 2.3273, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.1492775059733758, | |
| "grad_norm": 0.4058266878128052, | |
| "learning_rate": 0.00019029265382866214, | |
| "loss": 2.4121, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.15018773466833543, | |
| "grad_norm": 0.41517823934555054, | |
| "learning_rate": 0.0001901682892489367, | |
| "loss": 2.3822, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.15109796336329503, | |
| "grad_norm": 0.40526044368743896, | |
| "learning_rate": 0.0001900431742611089, | |
| "loss": 2.3957, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.15200819205825464, | |
| "grad_norm": 0.39370712637901306, | |
| "learning_rate": 0.00018991730990642388, | |
| "loss": 2.2458, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.15291842075321424, | |
| "grad_norm": 0.37955179810523987, | |
| "learning_rate": 0.00018979069723236333, | |
| "loss": 2.246, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.15382864944817384, | |
| "grad_norm": 0.3796885907649994, | |
| "learning_rate": 0.00018966333729263674, | |
| "loss": 2.2288, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.15473887814313347, | |
| "grad_norm": 0.37720954418182373, | |
| "learning_rate": 0.00018953523114717245, | |
| "loss": 2.2816, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.15564910683809308, | |
| "grad_norm": 0.3945274353027344, | |
| "learning_rate": 0.00018940637986210906, | |
| "loss": 2.1191, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.15655933553305268, | |
| "grad_norm": 0.42212241888046265, | |
| "learning_rate": 0.0001892767845097864, | |
| "loss": 2.3253, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.15746956422801228, | |
| "grad_norm": 0.3840961456298828, | |
| "learning_rate": 0.00018914644616873657, | |
| "loss": 2.0736, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.15837979292297188, | |
| "grad_norm": 0.3896063268184662, | |
| "learning_rate": 0.0001890153659236753, | |
| "loss": 2.1033, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.15929002161793152, | |
| "grad_norm": 0.3891961872577667, | |
| "learning_rate": 0.00018888354486549237, | |
| "loss": 2.1627, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.16020025031289112, | |
| "grad_norm": 0.4236443042755127, | |
| "learning_rate": 0.00018875098409124302, | |
| "loss": 2.2984, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.16111047900785072, | |
| "grad_norm": 0.39355891942977905, | |
| "learning_rate": 0.0001886176847041386, | |
| "loss": 2.1823, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.16202070770281032, | |
| "grad_norm": 0.4226863384246826, | |
| "learning_rate": 0.00018848364781353744, | |
| "loss": 2.2797, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.16293093639776993, | |
| "grad_norm": 0.4181986451148987, | |
| "learning_rate": 0.0001883488745349355, | |
| "loss": 2.1999, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.16384116509272956, | |
| "grad_norm": 0.4138600528240204, | |
| "learning_rate": 0.0001882133659899573, | |
| "loss": 1.8991, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.16475139378768916, | |
| "grad_norm": 0.43699535727500916, | |
| "learning_rate": 0.00018807712330634642, | |
| "loss": 2.4304, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.16566162248264876, | |
| "grad_norm": 0.4426632523536682, | |
| "learning_rate": 0.0001879401476179562, | |
| "loss": 2.2588, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.16657185117760837, | |
| "grad_norm": 0.4601946473121643, | |
| "learning_rate": 0.0001878024400647402, | |
| "loss": 2.4086, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.167482079872568, | |
| "grad_norm": 0.41164615750312805, | |
| "learning_rate": 0.00018766400179274286, | |
| "loss": 1.9165, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.1683923085675276, | |
| "grad_norm": 0.42600390315055847, | |
| "learning_rate": 0.00018752483395408987, | |
| "loss": 2.0585, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.1693025372624872, | |
| "grad_norm": 0.4481484889984131, | |
| "learning_rate": 0.00018738493770697852, | |
| "loss": 2.2189, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.1702127659574468, | |
| "grad_norm": 0.47090041637420654, | |
| "learning_rate": 0.00018724431421566823, | |
| "loss": 2.3016, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.1711229946524064, | |
| "grad_norm": 0.4611455798149109, | |
| "learning_rate": 0.00018710296465047075, | |
| "loss": 2.2475, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.17203322334736604, | |
| "grad_norm": 0.4549846053123474, | |
| "learning_rate": 0.0001869608901877404, | |
| "loss": 2.0331, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.17294345204232564, | |
| "grad_norm": 0.4108966886997223, | |
| "learning_rate": 0.0001868180920098644, | |
| "loss": 2.0985, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.17385368073728524, | |
| "grad_norm": 0.505251944065094, | |
| "learning_rate": 0.00018667457130525284, | |
| "loss": 2.1995, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.17476390943224485, | |
| "grad_norm": 0.4776213467121124, | |
| "learning_rate": 0.00018653032926832896, | |
| "loss": 2.2254, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.17567413812720445, | |
| "grad_norm": 0.4876101016998291, | |
| "learning_rate": 0.00018638536709951917, | |
| "loss": 2.2885, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.17658436682216408, | |
| "grad_norm": 0.49354955554008484, | |
| "learning_rate": 0.000186239686005243, | |
| "loss": 2.2939, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.17749459551712368, | |
| "grad_norm": 0.5280092358589172, | |
| "learning_rate": 0.0001860932871979031, | |
| "loss": 2.3555, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.17840482421208329, | |
| "grad_norm": 0.5320273637771606, | |
| "learning_rate": 0.00018594617189587512, | |
| "loss": 2.2854, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.1793150529070429, | |
| "grad_norm": 0.6043667793273926, | |
| "learning_rate": 0.00018579834132349772, | |
| "loss": 2.6826, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.1802252816020025, | |
| "grad_norm": 0.6498908996582031, | |
| "learning_rate": 0.0001856497967110621, | |
| "loss": 2.5815, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.18113551029696212, | |
| "grad_norm": 0.8005975484848022, | |
| "learning_rate": 0.00018550053929480202, | |
| "loss": 2.7034, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.18204573899192172, | |
| "grad_norm": 1.9555100202560425, | |
| "learning_rate": 0.00018535057031688335, | |
| "loss": 2.9439, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.18295596768688133, | |
| "grad_norm": 0.65727299451828, | |
| "learning_rate": 0.0001851998910253939, | |
| "loss": 2.695, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.18386619638184093, | |
| "grad_norm": 0.6414365172386169, | |
| "learning_rate": 0.0001850485026743328, | |
| "loss": 2.5447, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.18477642507680053, | |
| "grad_norm": 0.5610425472259521, | |
| "learning_rate": 0.00018489640652360022, | |
| "loss": 2.4502, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.18568665377176016, | |
| "grad_norm": 0.4109022319316864, | |
| "learning_rate": 0.00018474360383898694, | |
| "loss": 2.5165, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.18659688246671977, | |
| "grad_norm": 0.4291420578956604, | |
| "learning_rate": 0.00018459009589216364, | |
| "loss": 2.4464, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.18750711116167937, | |
| "grad_norm": 0.44230878353118896, | |
| "learning_rate": 0.0001844358839606705, | |
| "loss": 2.4365, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.18841733985663897, | |
| "grad_norm": 0.4131537675857544, | |
| "learning_rate": 0.00018428096932790632, | |
| "loss": 2.5135, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.1893275685515986, | |
| "grad_norm": 0.41846412420272827, | |
| "learning_rate": 0.00018412535328311814, | |
| "loss": 2.3571, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.1902377972465582, | |
| "grad_norm": 0.4208964407444, | |
| "learning_rate": 0.0001839690371213903, | |
| "loss": 2.4642, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.1911480259415178, | |
| "grad_norm": 0.40538257360458374, | |
| "learning_rate": 0.0001838120221436338, | |
| "loss": 2.4407, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.1920582546364774, | |
| "grad_norm": 0.4170033633708954, | |
| "learning_rate": 0.00018365430965657526, | |
| "loss": 2.4844, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.192968483331437, | |
| "grad_norm": 0.3793712258338928, | |
| "learning_rate": 0.00018349590097274632, | |
| "loss": 2.2296, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.19387871202639664, | |
| "grad_norm": 0.38766035437583923, | |
| "learning_rate": 0.00018333679741047254, | |
| "loss": 2.2268, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.19478894072135625, | |
| "grad_norm": 0.3814389109611511, | |
| "learning_rate": 0.00018317700029386245, | |
| "loss": 2.0952, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.19569916941631585, | |
| "grad_norm": 0.4213506579399109, | |
| "learning_rate": 0.00018301651095279655, | |
| "loss": 2.3572, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.19660939811127545, | |
| "grad_norm": 0.41608527302742004, | |
| "learning_rate": 0.0001828553307229163, | |
| "loss": 2.2262, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.19751962680623505, | |
| "grad_norm": 0.3709307312965393, | |
| "learning_rate": 0.0001826934609456129, | |
| "loss": 2.2569, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.19842985550119469, | |
| "grad_norm": 0.400329053401947, | |
| "learning_rate": 0.00018253090296801614, | |
| "loss": 2.2614, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.1993400841961543, | |
| "grad_norm": 0.38953447341918945, | |
| "learning_rate": 0.0001823676581429833, | |
| "loss": 2.1335, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.2002503128911139, | |
| "grad_norm": 0.39479631185531616, | |
| "learning_rate": 0.00018220372782908777, | |
| "loss": 2.2293, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.2011605415860735, | |
| "grad_norm": 0.3911544382572174, | |
| "learning_rate": 0.00018203911339060783, | |
| "loss": 2.0099, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.2020707702810331, | |
| "grad_norm": 0.40512001514434814, | |
| "learning_rate": 0.00018187381619751516, | |
| "loss": 2.2282, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.20298099897599273, | |
| "grad_norm": 0.42339497804641724, | |
| "learning_rate": 0.00018170783762546365, | |
| "loss": 2.1722, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.20389122767095233, | |
| "grad_norm": 0.4218563139438629, | |
| "learning_rate": 0.00018154117905577776, | |
| "loss": 2.1781, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.20480145636591193, | |
| "grad_norm": 0.39831289649009705, | |
| "learning_rate": 0.00018137384187544116, | |
| "loss": 2.1663, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.20571168506087154, | |
| "grad_norm": 0.39963746070861816, | |
| "learning_rate": 0.00018120582747708502, | |
| "loss": 1.97, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.20662191375583117, | |
| "grad_norm": 0.38624411821365356, | |
| "learning_rate": 0.0001810371372589766, | |
| "loss": 1.9793, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.20753214245079077, | |
| "grad_norm": 0.4173285961151123, | |
| "learning_rate": 0.0001808677726250076, | |
| "loss": 2.1832, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.20844237114575037, | |
| "grad_norm": 0.393627405166626, | |
| "learning_rate": 0.00018069773498468223, | |
| "loss": 2.003, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.20935259984070997, | |
| "grad_norm": 0.4344862699508667, | |
| "learning_rate": 0.00018052702575310588, | |
| "loss": 2.235, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.21026282853566958, | |
| "grad_norm": 0.42050907015800476, | |
| "learning_rate": 0.00018035564635097298, | |
| "loss": 2.0272, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.2111730572306292, | |
| "grad_norm": 0.45956531167030334, | |
| "learning_rate": 0.00018018359820455536, | |
| "loss": 2.277, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.2120832859255888, | |
| "grad_norm": 0.4257940649986267, | |
| "learning_rate": 0.00018001088274569038, | |
| "loss": 2.1714, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.2129935146205484, | |
| "grad_norm": 0.4153771996498108, | |
| "learning_rate": 0.00017983750141176895, | |
| "loss": 2.0942, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.21390374331550802, | |
| "grad_norm": 0.44307154417037964, | |
| "learning_rate": 0.0001796634556457236, | |
| "loss": 2.1548, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.21481397201046762, | |
| "grad_norm": 0.45095738768577576, | |
| "learning_rate": 0.0001794887468960165, | |
| "loss": 2.1294, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.21572420070542725, | |
| "grad_norm": 0.4348011612892151, | |
| "learning_rate": 0.00017931337661662727, | |
| "loss": 2.1176, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.21663442940038685, | |
| "grad_norm": 0.46650028228759766, | |
| "learning_rate": 0.0001791373462670411, | |
| "loss": 2.1333, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.21754465809534645, | |
| "grad_norm": 0.49439772963523865, | |
| "learning_rate": 0.00017896065731223644, | |
| "loss": 2.1169, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.21845488679030606, | |
| "grad_norm": 0.4834192395210266, | |
| "learning_rate": 0.00017878331122267284, | |
| "loss": 2.3213, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.21936511548526566, | |
| "grad_norm": 0.49015435576438904, | |
| "learning_rate": 0.00017860530947427875, | |
| "loss": 2.3145, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.2202753441802253, | |
| "grad_norm": 0.4994793236255646, | |
| "learning_rate": 0.00017842665354843922, | |
| "loss": 2.1586, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.2211855728751849, | |
| "grad_norm": 0.5262652635574341, | |
| "learning_rate": 0.0001782473449319835, | |
| "loss": 2.3597, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.2220958015701445, | |
| "grad_norm": 0.5197060704231262, | |
| "learning_rate": 0.0001780673851171728, | |
| "loss": 2.2091, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.2230060302651041, | |
| "grad_norm": 0.5736340284347534, | |
| "learning_rate": 0.00017788677560168784, | |
| "loss": 2.489, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.22391625896006373, | |
| "grad_norm": 0.5602148771286011, | |
| "learning_rate": 0.0001777055178886162, | |
| "loss": 2.4159, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.22482648765502333, | |
| "grad_norm": 0.6122560501098633, | |
| "learning_rate": 0.0001775236134864401, | |
| "loss": 2.5168, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.22573671634998294, | |
| "grad_norm": 0.6337876915931702, | |
| "learning_rate": 0.00017734106390902366, | |
| "loss": 2.522, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.22664694504494254, | |
| "grad_norm": 0.8038336634635925, | |
| "learning_rate": 0.0001771578706756003, | |
| "loss": 2.5793, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.22755717373990214, | |
| "grad_norm": 1.3161811828613281, | |
| "learning_rate": 0.0001769740353107602, | |
| "loss": 2.3232, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.22846740243486177, | |
| "grad_norm": 0.5344598293304443, | |
| "learning_rate": 0.00017678955934443758, | |
| "loss": 2.5626, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.22937763112982137, | |
| "grad_norm": 0.5922054052352905, | |
| "learning_rate": 0.0001766044443118978, | |
| "loss": 2.5528, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.23028785982478098, | |
| "grad_norm": 0.4852460026741028, | |
| "learning_rate": 0.00017641869175372493, | |
| "loss": 2.4348, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.23119808851974058, | |
| "grad_norm": 0.40997761487960815, | |
| "learning_rate": 0.00017623230321580854, | |
| "loss": 2.5167, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.23210831721470018, | |
| "grad_norm": 0.40511611104011536, | |
| "learning_rate": 0.00017604528024933115, | |
| "loss": 2.479, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.2330185459096598, | |
| "grad_norm": 0.3989182114601135, | |
| "learning_rate": 0.00017585762441075503, | |
| "loss": 2.3482, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.23392877460461942, | |
| "grad_norm": 0.4280647039413452, | |
| "learning_rate": 0.00017566933726180964, | |
| "loss": 2.526, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.23483900329957902, | |
| "grad_norm": 0.43045175075531006, | |
| "learning_rate": 0.0001754804203694782, | |
| "loss": 2.4734, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.23574923199453862, | |
| "grad_norm": 0.41747456789016724, | |
| "learning_rate": 0.0001752908753059849, | |
| "loss": 2.3361, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.23665946068949822, | |
| "grad_norm": 0.4000650942325592, | |
| "learning_rate": 0.00017510070364878177, | |
| "loss": 2.4496, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.23756968938445785, | |
| "grad_norm": 0.3866179287433624, | |
| "learning_rate": 0.00017490990698053563, | |
| "loss": 2.3222, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.23847991807941746, | |
| "grad_norm": 0.4083816409111023, | |
| "learning_rate": 0.00017471848688911464, | |
| "loss": 2.3619, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.23939014677437706, | |
| "grad_norm": 0.4181024432182312, | |
| "learning_rate": 0.0001745264449675755, | |
| "loss": 2.3515, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.24030037546933666, | |
| "grad_norm": 0.3845669627189636, | |
| "learning_rate": 0.00017433378281414975, | |
| "loss": 2.3866, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.24121060416429627, | |
| "grad_norm": 0.3917085826396942, | |
| "learning_rate": 0.0001741405020322309, | |
| "loss": 2.4155, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.2421208328592559, | |
| "grad_norm": 0.3883097171783447, | |
| "learning_rate": 0.00017394660423036075, | |
| "loss": 2.1034, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.2430310615542155, | |
| "grad_norm": 0.37689700722694397, | |
| "learning_rate": 0.00017375209102221613, | |
| "loss": 2.1676, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.2439412902491751, | |
| "grad_norm": 0.38591262698173523, | |
| "learning_rate": 0.00017355696402659548, | |
| "loss": 2.3066, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.2448515189441347, | |
| "grad_norm": 0.3865531384944916, | |
| "learning_rate": 0.00017336122486740548, | |
| "loss": 2.1387, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.24576174763909434, | |
| "grad_norm": 0.38778188824653625, | |
| "learning_rate": 0.00017316487517364721, | |
| "loss": 2.2369, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.24667197633405394, | |
| "grad_norm": 0.38550639152526855, | |
| "learning_rate": 0.000172967916579403, | |
| "loss": 2.2013, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.24758220502901354, | |
| "grad_norm": 0.41831621527671814, | |
| "learning_rate": 0.00017277035072382253, | |
| "loss": 2.1315, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.24849243372397314, | |
| "grad_norm": 0.386983722448349, | |
| "learning_rate": 0.00017257217925110933, | |
| "loss": 2.1582, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.24940266241893275, | |
| "grad_norm": 0.39023295044898987, | |
| "learning_rate": 0.00017237340381050703, | |
| "loss": 2.0061, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.2503128911138924, | |
| "grad_norm": 0.4187031686306, | |
| "learning_rate": 0.00017217402605628572, | |
| "loss": 2.0814, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.2503128911138924, | |
| "eval_loss": 2.2949306964874268, | |
| "eval_runtime": 205.2875, | |
| "eval_samples_per_second": 9.017, | |
| "eval_steps_per_second": 4.511, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.251223119808852, | |
| "grad_norm": 0.39927324652671814, | |
| "learning_rate": 0.00017197404764772805, | |
| "loss": 2.1982, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.2521333485038116, | |
| "grad_norm": 0.4287830889225006, | |
| "learning_rate": 0.00017177347024911562, | |
| "loss": 2.2733, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.2530435771987712, | |
| "grad_norm": 0.3960012197494507, | |
| "learning_rate": 0.00017157229552971487, | |
| "loss": 2.1884, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.2539538058937308, | |
| "grad_norm": 0.40106138586997986, | |
| "learning_rate": 0.00017137052516376345, | |
| "loss": 2.1207, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.2548640345886904, | |
| "grad_norm": 0.4410867393016815, | |
| "learning_rate": 0.00017116816083045602, | |
| "loss": 2.3589, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.25577426328365, | |
| "grad_norm": 0.4092939794063568, | |
| "learning_rate": 0.0001709652042139306, | |
| "loss": 2.0842, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.25668449197860965, | |
| "grad_norm": 0.40820494294166565, | |
| "learning_rate": 0.0001707616570032542, | |
| "loss": 2.1658, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.25759472067356926, | |
| "grad_norm": 0.41664186120033264, | |
| "learning_rate": 0.00017055752089240907, | |
| "loss": 2.1389, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.25850494936852886, | |
| "grad_norm": 0.4125240445137024, | |
| "learning_rate": 0.00017035279758027832, | |
| "loss": 2.0615, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.25941517806348846, | |
| "grad_norm": 0.42702898383140564, | |
| "learning_rate": 0.00017014748877063214, | |
| "loss": 2.017, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.26032540675844806, | |
| "grad_norm": 0.44943541288375854, | |
| "learning_rate": 0.00016994159617211317, | |
| "loss": 2.1901, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.26123563545340767, | |
| "grad_norm": 0.4286860227584839, | |
| "learning_rate": 0.00016973512149822274, | |
| "loss": 2.0643, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.26214586414836727, | |
| "grad_norm": 0.44938111305236816, | |
| "learning_rate": 0.0001695280664673062, | |
| "loss": 2.1539, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.26305609284332687, | |
| "grad_norm": 0.4638296067714691, | |
| "learning_rate": 0.0001693204328025389, | |
| "loss": 2.291, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.2639663215382865, | |
| "grad_norm": 0.49295714497566223, | |
| "learning_rate": 0.00016911222223191182, | |
| "loss": 2.2538, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.2648765502332461, | |
| "grad_norm": 0.48185715079307556, | |
| "learning_rate": 0.00016890343648821697, | |
| "loss": 2.2792, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.26578677892820574, | |
| "grad_norm": 0.4750272035598755, | |
| "learning_rate": 0.0001686940773090333, | |
| "loss": 2.2774, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.26669700762316534, | |
| "grad_norm": 0.5073033571243286, | |
| "learning_rate": 0.00016848414643671195, | |
| "loss": 2.3261, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.26760723631812494, | |
| "grad_norm": 0.5343424081802368, | |
| "learning_rate": 0.00016827364561836187, | |
| "loss": 2.4097, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.26851746501308454, | |
| "grad_norm": 0.5311369895935059, | |
| "learning_rate": 0.00016806257660583534, | |
| "loss": 2.3821, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.26942769370804415, | |
| "grad_norm": 0.5551429986953735, | |
| "learning_rate": 0.00016785094115571322, | |
| "loss": 2.3795, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.27033792240300375, | |
| "grad_norm": 0.6279783248901367, | |
| "learning_rate": 0.0001676387410292906, | |
| "loss": 2.435, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.27124815109796335, | |
| "grad_norm": 0.7317250967025757, | |
| "learning_rate": 0.00016742597799256182, | |
| "loss": 2.6991, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.27215837979292296, | |
| "grad_norm": 0.8485302329063416, | |
| "learning_rate": 0.000167212653816206, | |
| "loss": 2.7005, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.27306860848788256, | |
| "grad_norm": 1.5959185361862183, | |
| "learning_rate": 0.00016699877027557226, | |
| "loss": 2.7536, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2739788371828422, | |
| "grad_norm": 0.4755174219608307, | |
| "learning_rate": 0.00016678432915066488, | |
| "loss": 2.5907, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.2748890658778018, | |
| "grad_norm": 0.45389342308044434, | |
| "learning_rate": 0.00016656933222612854, | |
| "loss": 2.4622, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.2757992945727614, | |
| "grad_norm": 0.4949435591697693, | |
| "learning_rate": 0.00016635378129123342, | |
| "loss": 2.4185, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.276709523267721, | |
| "grad_norm": 0.4521631896495819, | |
| "learning_rate": 0.00016613767813986044, | |
| "loss": 2.4918, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.2776197519626806, | |
| "grad_norm": 0.4228963553905487, | |
| "learning_rate": 0.0001659210245704861, | |
| "loss": 2.4194, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.27852998065764023, | |
| "grad_norm": 0.4170341491699219, | |
| "learning_rate": 0.00016570382238616777, | |
| "loss": 2.4185, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.27944020935259983, | |
| "grad_norm": 0.4123315215110779, | |
| "learning_rate": 0.00016548607339452853, | |
| "loss": 2.3737, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.28035043804755944, | |
| "grad_norm": 0.4320162832736969, | |
| "learning_rate": 0.00016526777940774204, | |
| "loss": 2.3317, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.28126066674251904, | |
| "grad_norm": 0.4118390381336212, | |
| "learning_rate": 0.00016504894224251778, | |
| "loss": 2.3786, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.28217089543747864, | |
| "grad_norm": 0.39763331413269043, | |
| "learning_rate": 0.0001648295637200856, | |
| "loss": 2.2968, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.2830811241324383, | |
| "grad_norm": 0.4391527473926544, | |
| "learning_rate": 0.0001646096456661807, | |
| "loss": 2.3764, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.2839913528273979, | |
| "grad_norm": 0.43077877163887024, | |
| "learning_rate": 0.00016438918991102842, | |
| "loss": 2.2013, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.2849015815223575, | |
| "grad_norm": 0.43149155378341675, | |
| "learning_rate": 0.000164168198289329, | |
| "loss": 2.3097, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.2858118102173171, | |
| "grad_norm": 0.40134817361831665, | |
| "learning_rate": 0.00016394667264024246, | |
| "loss": 2.3306, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.2867220389122767, | |
| "grad_norm": 0.4056681990623474, | |
| "learning_rate": 0.00016372461480737297, | |
| "loss": 2.3146, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.2876322676072363, | |
| "grad_norm": 0.41738027334213257, | |
| "learning_rate": 0.00016350202663875386, | |
| "loss": 1.9997, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.2885424963021959, | |
| "grad_norm": 0.38182246685028076, | |
| "learning_rate": 0.00016327890998683192, | |
| "loss": 2.0466, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.2894527249971555, | |
| "grad_norm": 0.39759719371795654, | |
| "learning_rate": 0.00016305526670845226, | |
| "loss": 2.1788, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.2903629536921151, | |
| "grad_norm": 0.3982352614402771, | |
| "learning_rate": 0.0001628310986648427, | |
| "loss": 2.2115, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.2912731823870747, | |
| "grad_norm": 0.41679051518440247, | |
| "learning_rate": 0.0001626064077215983, | |
| "loss": 2.3036, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.2921834110820344, | |
| "grad_norm": 0.40436604619026184, | |
| "learning_rate": 0.00016238119574866588, | |
| "loss": 2.1493, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.293093639776994, | |
| "grad_norm": 0.4502476751804352, | |
| "learning_rate": 0.0001621554646203284, | |
| "loss": 1.8572, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.2940038684719536, | |
| "grad_norm": 0.44303473830223083, | |
| "learning_rate": 0.00016192921621518944, | |
| "loss": 2.1832, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.2949140971669132, | |
| "grad_norm": 0.4064692258834839, | |
| "learning_rate": 0.0001617024524161574, | |
| "loss": 2.2656, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.2958243258618728, | |
| "grad_norm": 0.4479392170906067, | |
| "learning_rate": 0.0001614751751104301, | |
| "loss": 2.2462, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.2967345545568324, | |
| "grad_norm": 0.4629363715648651, | |
| "learning_rate": 0.0001612473861894788, | |
| "loss": 1.9715, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.297644783251792, | |
| "grad_norm": 0.3991665542125702, | |
| "learning_rate": 0.00016101908754903268, | |
| "loss": 2.0642, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.2985550119467516, | |
| "grad_norm": 0.42503711581230164, | |
| "learning_rate": 0.00016079028108906282, | |
| "loss": 2.1403, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.2994652406417112, | |
| "grad_norm": 0.4499455392360687, | |
| "learning_rate": 0.00016056096871376667, | |
| "loss": 2.0534, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.30037546933667086, | |
| "grad_norm": 0.4549277424812317, | |
| "learning_rate": 0.00016033115233155202, | |
| "loss": 2.2083, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.30128569803163047, | |
| "grad_norm": 0.3974262773990631, | |
| "learning_rate": 0.0001601008338550211, | |
| "loss": 2.0156, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.30219592672659007, | |
| "grad_norm": 0.43566057085990906, | |
| "learning_rate": 0.00015987001520095478, | |
| "loss": 2.1801, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.3031061554215497, | |
| "grad_norm": 0.47677701711654663, | |
| "learning_rate": 0.00015963869829029658, | |
| "loss": 2.1415, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.3040163841165093, | |
| "grad_norm": 0.4603672921657562, | |
| "learning_rate": 0.00015940688504813662, | |
| "loss": 2.2967, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.3049266128114689, | |
| "grad_norm": 0.4428515136241913, | |
| "learning_rate": 0.00015917457740369565, | |
| "loss": 2.1447, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.3058368415064285, | |
| "grad_norm": 0.4379275441169739, | |
| "learning_rate": 0.000158941777290309, | |
| "loss": 2.0957, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.3067470702013881, | |
| "grad_norm": 0.4831966459751129, | |
| "learning_rate": 0.00015870848664541044, | |
| "loss": 2.2457, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.3076572988963477, | |
| "grad_norm": 0.45160865783691406, | |
| "learning_rate": 0.00015847470741051618, | |
| "loss": 2.1441, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.3085675275913073, | |
| "grad_norm": 0.44453370571136475, | |
| "learning_rate": 0.00015824044153120852, | |
| "loss": 2.1073, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.30947775628626695, | |
| "grad_norm": 0.49965375661849976, | |
| "learning_rate": 0.00015800569095711982, | |
| "loss": 2.1574, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.31038798498122655, | |
| "grad_norm": 0.48138341307640076, | |
| "learning_rate": 0.00015777045764191625, | |
| "loss": 2.0205, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.31129821367618615, | |
| "grad_norm": 0.5034924745559692, | |
| "learning_rate": 0.00015753474354328142, | |
| "loss": 2.2319, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.31220844237114576, | |
| "grad_norm": 0.5034711956977844, | |
| "learning_rate": 0.00015729855062290022, | |
| "loss": 2.4066, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.31311867106610536, | |
| "grad_norm": 0.5409703254699707, | |
| "learning_rate": 0.00015706188084644242, | |
| "loss": 2.2435, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.31402889976106496, | |
| "grad_norm": 0.544597327709198, | |
| "learning_rate": 0.00015682473618354635, | |
| "loss": 2.2625, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.31493912845602456, | |
| "grad_norm": 0.6114000082015991, | |
| "learning_rate": 0.0001565871186078025, | |
| "loss": 2.4302, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.31584935715098417, | |
| "grad_norm": 0.6364843845367432, | |
| "learning_rate": 0.00015634903009673705, | |
| "loss": 2.5153, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.31675958584594377, | |
| "grad_norm": 0.7510351538658142, | |
| "learning_rate": 0.00015611047263179548, | |
| "loss": 2.5605, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.31766981454090343, | |
| "grad_norm": 0.8501291275024414, | |
| "learning_rate": 0.000155871448198326, | |
| "loss": 2.6519, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.31858004323586303, | |
| "grad_norm": 1.7441632747650146, | |
| "learning_rate": 0.0001556319587855631, | |
| "loss": 2.7517, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.31949027193082263, | |
| "grad_norm": 0.5301811695098877, | |
| "learning_rate": 0.00015539200638661104, | |
| "loss": 2.6647, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.32040050062578224, | |
| "grad_norm": 0.5063616633415222, | |
| "learning_rate": 0.00015515159299842707, | |
| "loss": 2.4961, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.32131072932074184, | |
| "grad_norm": 0.4843781590461731, | |
| "learning_rate": 0.00015491072062180503, | |
| "loss": 2.496, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.32222095801570144, | |
| "grad_norm": 0.4524553716182709, | |
| "learning_rate": 0.00015466939126135856, | |
| "loss": 2.448, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.32313118671066104, | |
| "grad_norm": 0.43678200244903564, | |
| "learning_rate": 0.00015442760692550443, | |
| "loss": 2.2687, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.32404141540562065, | |
| "grad_norm": 0.4301970303058624, | |
| "learning_rate": 0.00015418536962644592, | |
| "loss": 2.4826, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.32495164410058025, | |
| "grad_norm": 0.42540326714515686, | |
| "learning_rate": 0.00015394268138015598, | |
| "loss": 2.4205, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.32586187279553985, | |
| "grad_norm": 0.4173906445503235, | |
| "learning_rate": 0.00015369954420636048, | |
| "loss": 2.394, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.3267721014904995, | |
| "grad_norm": 0.43184736371040344, | |
| "learning_rate": 0.00015345596012852138, | |
| "loss": 2.3504, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.3276823301854591, | |
| "grad_norm": 0.4002053141593933, | |
| "learning_rate": 0.00015321193117381996, | |
| "loss": 2.2951, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3285925588804187, | |
| "grad_norm": 0.39067134261131287, | |
| "learning_rate": 0.00015296745937313987, | |
| "loss": 2.2768, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.3295027875753783, | |
| "grad_norm": 0.40051525831222534, | |
| "learning_rate": 0.00015272254676105025, | |
| "loss": 2.2235, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.3304130162703379, | |
| "grad_norm": 0.3954068422317505, | |
| "learning_rate": 0.00015247719537578883, | |
| "loss": 2.2502, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.3313232449652975, | |
| "grad_norm": 0.4123362600803375, | |
| "learning_rate": 0.00015223140725924495, | |
| "loss": 2.3309, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.33223347366025713, | |
| "grad_norm": 0.4138774871826172, | |
| "learning_rate": 0.00015198518445694255, | |
| "loss": 2.4107, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.33314370235521673, | |
| "grad_norm": 0.3983847498893738, | |
| "learning_rate": 0.0001517385290180231, | |
| "loss": 2.2718, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.33405393105017633, | |
| "grad_norm": 0.36962834000587463, | |
| "learning_rate": 0.00015149144299522873, | |
| "loss": 2.1744, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.334964159745136, | |
| "grad_norm": 0.37924104928970337, | |
| "learning_rate": 0.0001512439284448849, | |
| "loss": 2.1451, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.3358743884400956, | |
| "grad_norm": 0.39990487694740295, | |
| "learning_rate": 0.0001509959874268835, | |
| "loss": 2.2508, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.3367846171350552, | |
| "grad_norm": 0.3862214684486389, | |
| "learning_rate": 0.00015074762200466556, | |
| "loss": 2.1483, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3376948458300148, | |
| "grad_norm": 0.4037676751613617, | |
| "learning_rate": 0.00015049883424520414, | |
| "loss": 2.2179, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.3386050745249744, | |
| "grad_norm": 0.40439948439598083, | |
| "learning_rate": 0.00015024962621898715, | |
| "loss": 2.2054, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.339515303219934, | |
| "grad_norm": 0.3871942460536957, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 2.129, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.3404255319148936, | |
| "grad_norm": 0.4091387093067169, | |
| "learning_rate": 0.00014974995766570855, | |
| "loss": 2.1395, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.3413357606098532, | |
| "grad_norm": 0.4097527265548706, | |
| "learning_rate": 0.00014949950129704162, | |
| "loss": 2.1789, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.3422459893048128, | |
| "grad_norm": 0.4139934480190277, | |
| "learning_rate": 0.00014924863297837378, | |
| "loss": 2.0611, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.3431562179997724, | |
| "grad_norm": 0.4146927297115326, | |
| "learning_rate": 0.00014899735479750794, | |
| "loss": 2.2488, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.3440664466947321, | |
| "grad_norm": 0.4194958209991455, | |
| "learning_rate": 0.00014874566884565807, | |
| "loss": 2.0164, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.3449766753896917, | |
| "grad_norm": 0.41280898451805115, | |
| "learning_rate": 0.00014849357721743168, | |
| "loss": 2.1503, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.3458869040846513, | |
| "grad_norm": 0.4133208692073822, | |
| "learning_rate": 0.00014824108201081247, | |
| "loss": 2.0895, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.3467971327796109, | |
| "grad_norm": 0.41347819566726685, | |
| "learning_rate": 0.00014798818532714279, | |
| "loss": 2.0479, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.3477073614745705, | |
| "grad_norm": 0.43102580308914185, | |
| "learning_rate": 0.00014773488927110633, | |
| "loss": 2.1458, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.3486175901695301, | |
| "grad_norm": 0.41427451372146606, | |
| "learning_rate": 0.00014748119595071034, | |
| "loss": 1.9396, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.3495278188644897, | |
| "grad_norm": 0.46386152505874634, | |
| "learning_rate": 0.0001472271074772683, | |
| "loss": 2.2446, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.3504380475594493, | |
| "grad_norm": 0.4310764670372009, | |
| "learning_rate": 0.00014697262596538227, | |
| "loss": 2.2144, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.3513482762544089, | |
| "grad_norm": 0.4956878423690796, | |
| "learning_rate": 0.00014671775353292525, | |
| "loss": 2.1875, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.35225850494936856, | |
| "grad_norm": 0.4793931543827057, | |
| "learning_rate": 0.00014646249230102366, | |
| "loss": 2.2733, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.35316873364432816, | |
| "grad_norm": 0.46217313408851624, | |
| "learning_rate": 0.00014620684439403962, | |
| "loss": 2.2812, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.35407896233928776, | |
| "grad_norm": 0.4721885323524475, | |
| "learning_rate": 0.00014595081193955324, | |
| "loss": 2.1223, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.35498919103424736, | |
| "grad_norm": 0.49550965428352356, | |
| "learning_rate": 0.000145694397068345, | |
| "loss": 2.156, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.35589941972920697, | |
| "grad_norm": 0.5109139084815979, | |
| "learning_rate": 0.0001454376019143779, | |
| "loss": 2.1494, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.35680964842416657, | |
| "grad_norm": 0.4725574553012848, | |
| "learning_rate": 0.00014518042861477986, | |
| "loss": 2.1793, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.3577198771191262, | |
| "grad_norm": 0.4739914536476135, | |
| "learning_rate": 0.00014492287930982576, | |
| "loss": 2.1763, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.3586301058140858, | |
| "grad_norm": 0.5420114994049072, | |
| "learning_rate": 0.00014466495614291977, | |
| "loss": 2.4521, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.3595403345090454, | |
| "grad_norm": 0.5225427150726318, | |
| "learning_rate": 0.00014440666126057744, | |
| "loss": 2.372, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.360450563204005, | |
| "grad_norm": 0.5337964296340942, | |
| "learning_rate": 0.0001441479968124078, | |
| "loss": 2.397, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.36136079189896464, | |
| "grad_norm": 0.5906230807304382, | |
| "learning_rate": 0.0001438889649510956, | |
| "loss": 2.506, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.36227102059392424, | |
| "grad_norm": 0.6578875780105591, | |
| "learning_rate": 0.00014362956783238324, | |
| "loss": 2.6408, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.36318124928888385, | |
| "grad_norm": 0.7982918620109558, | |
| "learning_rate": 0.00014336980761505297, | |
| "loss": 2.6612, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.36409147798384345, | |
| "grad_norm": 1.4390262365341187, | |
| "learning_rate": 0.00014310968646090883, | |
| "loss": 2.7073, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.36500170667880305, | |
| "grad_norm": 0.5260487198829651, | |
| "learning_rate": 0.00014284920653475866, | |
| "loss": 2.6269, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.36591193537376265, | |
| "grad_norm": 0.4492892026901245, | |
| "learning_rate": 0.00014258837000439618, | |
| "loss": 2.3863, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.36682216406872226, | |
| "grad_norm": 0.4619944095611572, | |
| "learning_rate": 0.0001423271790405828, | |
| "loss": 2.4595, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.36773239276368186, | |
| "grad_norm": 0.4437786638736725, | |
| "learning_rate": 0.00014206563581702964, | |
| "loss": 2.3674, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.36864262145864146, | |
| "grad_norm": 0.4789164364337921, | |
| "learning_rate": 0.0001418037425103795, | |
| "loss": 2.5203, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.36955285015360106, | |
| "grad_norm": 0.44783228635787964, | |
| "learning_rate": 0.00014154150130018866, | |
| "loss": 2.5183, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.3704630788485607, | |
| "grad_norm": 0.40067169070243835, | |
| "learning_rate": 0.00014127891436890868, | |
| "loss": 2.3846, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.3713733075435203, | |
| "grad_norm": 0.3978015184402466, | |
| "learning_rate": 0.0001410159839018684, | |
| "loss": 2.3146, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.37228353623847993, | |
| "grad_norm": 0.4096076190471649, | |
| "learning_rate": 0.0001407527120872557, | |
| "loss": 2.3617, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.37319376493343953, | |
| "grad_norm": 0.4160764217376709, | |
| "learning_rate": 0.00014048910111609915, | |
| "loss": 2.2909, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.37410399362839913, | |
| "grad_norm": 0.3976461887359619, | |
| "learning_rate": 0.0001402251531822499, | |
| "loss": 2.3111, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.37501422232335874, | |
| "grad_norm": 0.3890199065208435, | |
| "learning_rate": 0.00013996087048236358, | |
| "loss": 2.0969, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.37592445101831834, | |
| "grad_norm": 0.4157082140445709, | |
| "learning_rate": 0.00013969625521588158, | |
| "loss": 2.3205, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.37683467971327794, | |
| "grad_norm": 0.4103608727455139, | |
| "learning_rate": 0.00013943130958501317, | |
| "loss": 2.2622, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.37774490840823755, | |
| "grad_norm": 0.40916207432746887, | |
| "learning_rate": 0.00013916603579471705, | |
| "loss": 2.3585, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.3786551371031972, | |
| "grad_norm": 0.39642858505249023, | |
| "learning_rate": 0.00013890043605268283, | |
| "loss": 2.2196, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.3795653657981568, | |
| "grad_norm": 0.3851282596588135, | |
| "learning_rate": 0.00013863451256931287, | |
| "loss": 2.0298, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.3804755944931164, | |
| "grad_norm": 0.38890305161476135, | |
| "learning_rate": 0.00013836826755770384, | |
| "loss": 2.1601, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.381385823188076, | |
| "grad_norm": 0.41382652521133423, | |
| "learning_rate": 0.00013810170323362816, | |
| "loss": 2.2656, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.3822960518830356, | |
| "grad_norm": 0.3820722699165344, | |
| "learning_rate": 0.0001378348218155158, | |
| "loss": 2.0094, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.3832062805779952, | |
| "grad_norm": 0.4150048494338989, | |
| "learning_rate": 0.00013756762552443553, | |
| "loss": 2.2529, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.3841165092729548, | |
| "grad_norm": 0.452776700258255, | |
| "learning_rate": 0.00013730011658407676, | |
| "loss": 2.1972, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.3850267379679144, | |
| "grad_norm": 0.4173040986061096, | |
| "learning_rate": 0.00013703229722073065, | |
| "loss": 2.1502, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.385936966662874, | |
| "grad_norm": 0.4115488529205322, | |
| "learning_rate": 0.000136764169663272, | |
| "loss": 1.9828, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.38684719535783363, | |
| "grad_norm": 0.4060666561126709, | |
| "learning_rate": 0.00013649573614314044, | |
| "loss": 2.267, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.3877574240527933, | |
| "grad_norm": 0.4049409031867981, | |
| "learning_rate": 0.00013622699889432184, | |
| "loss": 2.2044, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.3886676527477529, | |
| "grad_norm": 0.40970832109451294, | |
| "learning_rate": 0.00013595796015332984, | |
| "loss": 2.0984, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.3895778814427125, | |
| "grad_norm": 0.4141111671924591, | |
| "learning_rate": 0.00013568862215918717, | |
| "loss": 2.109, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.3904881101376721, | |
| "grad_norm": 0.43404263257980347, | |
| "learning_rate": 0.00013541898715340716, | |
| "loss": 2.1763, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.3913983388326317, | |
| "grad_norm": 0.41949963569641113, | |
| "learning_rate": 0.00013514905737997473, | |
| "loss": 2.3086, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.3923085675275913, | |
| "grad_norm": 0.41665390133857727, | |
| "learning_rate": 0.00013487883508532815, | |
| "loss": 2.0726, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.3932187962225509, | |
| "grad_norm": 0.4305708110332489, | |
| "learning_rate": 0.00013460832251834011, | |
| "loss": 2.1975, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.3941290249175105, | |
| "grad_norm": 0.44775405526161194, | |
| "learning_rate": 0.00013433752193029886, | |
| "loss": 2.1503, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.3950392536124701, | |
| "grad_norm": 0.44451820850372314, | |
| "learning_rate": 0.0001340664355748899, | |
| "loss": 2.1004, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.39594948230742977, | |
| "grad_norm": 0.44242945313453674, | |
| "learning_rate": 0.0001337950657081768, | |
| "loss": 2.1074, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.39685971100238937, | |
| "grad_norm": 0.4649699926376343, | |
| "learning_rate": 0.00013352341458858265, | |
| "loss": 2.2468, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.397769939697349, | |
| "grad_norm": 0.4718558192253113, | |
| "learning_rate": 0.00013325148447687125, | |
| "loss": 2.225, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.3986801683923086, | |
| "grad_norm": 0.44748789072036743, | |
| "learning_rate": 0.0001329792776361282, | |
| "loss": 2.0243, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.3995903970872682, | |
| "grad_norm": 0.4730619192123413, | |
| "learning_rate": 0.00013270679633174218, | |
| "loss": 2.0262, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.4005006257822278, | |
| "grad_norm": 0.4742071032524109, | |
| "learning_rate": 0.00013243404283138597, | |
| "loss": 2.1171, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.4014108544771874, | |
| "grad_norm": 0.4963454306125641, | |
| "learning_rate": 0.00013216101940499768, | |
| "loss": 2.051, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.402321083172147, | |
| "grad_norm": 0.5127780437469482, | |
| "learning_rate": 0.00013188772832476188, | |
| "loss": 2.1664, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.4032313118671066, | |
| "grad_norm": 0.5129209756851196, | |
| "learning_rate": 0.00013161417186509052, | |
| "loss": 2.2272, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.4041415405620662, | |
| "grad_norm": 0.5068848133087158, | |
| "learning_rate": 0.00013134035230260427, | |
| "loss": 2.1007, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.40505176925702585, | |
| "grad_norm": 0.5721228718757629, | |
| "learning_rate": 0.00013106627191611332, | |
| "loss": 2.255, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.40596199795198545, | |
| "grad_norm": 0.6085918545722961, | |
| "learning_rate": 0.0001307919329865985, | |
| "loss": 2.456, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.40687222664694506, | |
| "grad_norm": 0.6652196645736694, | |
| "learning_rate": 0.00013051733779719234, | |
| "loss": 2.5504, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.40778245534190466, | |
| "grad_norm": 0.7234418392181396, | |
| "learning_rate": 0.00013024248863316012, | |
| "loss": 2.5796, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.40869268403686426, | |
| "grad_norm": 0.8588744401931763, | |
| "learning_rate": 0.00012996738778188067, | |
| "loss": 2.5756, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.40960291273182386, | |
| "grad_norm": 1.2627683877944946, | |
| "learning_rate": 0.0001296920375328275, | |
| "loss": 2.203, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.41051314142678347, | |
| "grad_norm": 0.4838164746761322, | |
| "learning_rate": 0.00012941644017754964, | |
| "loss": 2.434, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.41142337012174307, | |
| "grad_norm": 0.44005534052848816, | |
| "learning_rate": 0.00012914059800965268, | |
| "loss": 2.55, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.4123335988167027, | |
| "grad_norm": 0.4343414604663849, | |
| "learning_rate": 0.0001288645133247795, | |
| "loss": 2.432, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.41324382751166233, | |
| "grad_norm": 0.4588654339313507, | |
| "learning_rate": 0.00012858818842059145, | |
| "loss": 2.4434, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.41415405620662193, | |
| "grad_norm": 0.4294244647026062, | |
| "learning_rate": 0.00012831162559674887, | |
| "loss": 2.4241, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.41506428490158154, | |
| "grad_norm": 0.40034809708595276, | |
| "learning_rate": 0.0001280348271548923, | |
| "loss": 2.3191, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.41597451359654114, | |
| "grad_norm": 0.40817153453826904, | |
| "learning_rate": 0.00012775779539862304, | |
| "loss": 2.589, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.41688474229150074, | |
| "grad_norm": 0.40605810284614563, | |
| "learning_rate": 0.0001274805326334842, | |
| "loss": 2.3445, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.41779497098646035, | |
| "grad_norm": 0.4386533200740814, | |
| "learning_rate": 0.00012720304116694138, | |
| "loss": 2.4002, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.41870519968141995, | |
| "grad_norm": 0.40985172986984253, | |
| "learning_rate": 0.00012692532330836346, | |
| "loss": 2.3964, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.41961542837637955, | |
| "grad_norm": 0.4220562279224396, | |
| "learning_rate": 0.00012664738136900348, | |
| "loss": 2.3145, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.42052565707133915, | |
| "grad_norm": 0.4068267047405243, | |
| "learning_rate": 0.00012636921766197943, | |
| "loss": 2.3274, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.42143588576629876, | |
| "grad_norm": 0.3973187208175659, | |
| "learning_rate": 0.0001260908345022547, | |
| "loss": 2.1801, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.4223461144612584, | |
| "grad_norm": 0.432224303483963, | |
| "learning_rate": 0.00012581223420661913, | |
| "loss": 2.4079, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.423256343156218, | |
| "grad_norm": 0.3939046859741211, | |
| "learning_rate": 0.00012553341909366978, | |
| "loss": 2.0749, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.4241665718511776, | |
| "grad_norm": 0.36949658393859863, | |
| "learning_rate": 0.00012525439148379128, | |
| "loss": 2.1471, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.4250768005461372, | |
| "grad_norm": 0.3828236758708954, | |
| "learning_rate": 0.00012497515369913685, | |
| "loss": 2.0466, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.4259870292410968, | |
| "grad_norm": 0.3874993920326233, | |
| "learning_rate": 0.00012469570806360875, | |
| "loss": 2.1605, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.42689725793605643, | |
| "grad_norm": 0.3854924738407135, | |
| "learning_rate": 0.00012441605690283915, | |
| "loss": 2.0584, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.42780748663101603, | |
| "grad_norm": 0.40301740169525146, | |
| "learning_rate": 0.00012413620254417057, | |
| "loss": 2.1481, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.42871771532597563, | |
| "grad_norm": 0.3891369104385376, | |
| "learning_rate": 0.00012385614731663666, | |
| "loss": 2.1968, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.42962794402093524, | |
| "grad_norm": 0.4305795729160309, | |
| "learning_rate": 0.00012357589355094275, | |
| "loss": 2.0421, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.4305381727158949, | |
| "grad_norm": 0.44661635160446167, | |
| "learning_rate": 0.0001232954435794464, | |
| "loss": 2.3347, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.4314484014108545, | |
| "grad_norm": 0.3984116315841675, | |
| "learning_rate": 0.00012301479973613822, | |
| "loss": 2.1093, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.4323586301058141, | |
| "grad_norm": 0.4153747856616974, | |
| "learning_rate": 0.00012273396435662212, | |
| "loss": 2.0698, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.4332688588007737, | |
| "grad_norm": 0.4589189887046814, | |
| "learning_rate": 0.00012245293977809605, | |
| "loss": 2.1707, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.4341790874957333, | |
| "grad_norm": 0.43936577439308167, | |
| "learning_rate": 0.0001221717283393326, | |
| "loss": 2.2608, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.4350893161906929, | |
| "grad_norm": 0.4170132279396057, | |
| "learning_rate": 0.0001218903323806595, | |
| "loss": 2.0813, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.4359995448856525, | |
| "grad_norm": 0.43124523758888245, | |
| "learning_rate": 0.00012160875424393996, | |
| "loss": 2.1674, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.4369097735806121, | |
| "grad_norm": 0.4394627511501312, | |
| "learning_rate": 0.00012132699627255347, | |
| "loss": 2.1904, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.4378200022755717, | |
| "grad_norm": 0.4404590427875519, | |
| "learning_rate": 0.00012104506081137608, | |
| "loss": 2.1313, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.4387302309705313, | |
| "grad_norm": 0.4580220878124237, | |
| "learning_rate": 0.00012076295020676103, | |
| "loss": 2.16, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.439640459665491, | |
| "grad_norm": 0.4533630311489105, | |
| "learning_rate": 0.00012048066680651908, | |
| "loss": 2.1153, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.4405506883604506, | |
| "grad_norm": 0.47520536184310913, | |
| "learning_rate": 0.00012019821295989912, | |
| "loss": 2.2152, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.4414609170554102, | |
| "grad_norm": 0.44196072220802307, | |
| "learning_rate": 0.00011991559101756852, | |
| "loss": 2.1375, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.4423711457503698, | |
| "grad_norm": 0.43681493401527405, | |
| "learning_rate": 0.00011963280333159358, | |
| "loss": 2.0552, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.4432813744453294, | |
| "grad_norm": 0.4537602961063385, | |
| "learning_rate": 0.00011934985225541998, | |
| "loss": 2.1473, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.444191603140289, | |
| "grad_norm": 0.4935773015022278, | |
| "learning_rate": 0.00011906674014385318, | |
| "loss": 2.0623, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.4451018318352486, | |
| "grad_norm": 0.4802737236022949, | |
| "learning_rate": 0.00011878346935303883, | |
| "loss": 2.2908, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.4460120605302082, | |
| "grad_norm": 0.5020537376403809, | |
| "learning_rate": 0.00011850004224044315, | |
| "loss": 2.3101, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.4469222892251678, | |
| "grad_norm": 0.5106056332588196, | |
| "learning_rate": 0.00011821646116483335, | |
| "loss": 2.2838, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.44783251792012746, | |
| "grad_norm": 0.473910391330719, | |
| "learning_rate": 0.00011793272848625797, | |
| "loss": 2.0599, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.44874274661508706, | |
| "grad_norm": 0.5086584091186523, | |
| "learning_rate": 0.0001176488465660271, | |
| "loss": 2.1578, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.44965297531004667, | |
| "grad_norm": 0.5282394886016846, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 2.2965, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.45056320400500627, | |
| "grad_norm": 0.5987780094146729, | |
| "learning_rate": 0.00011708064445203042, | |
| "loss": 2.3542, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.45147343269996587, | |
| "grad_norm": 0.5943189859390259, | |
| "learning_rate": 0.00011679632898701649, | |
| "loss": 2.4294, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.4523836613949255, | |
| "grad_norm": 0.6443737149238586, | |
| "learning_rate": 0.0001165118737378116, | |
| "loss": 2.605, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.4532938900898851, | |
| "grad_norm": 0.7082577347755432, | |
| "learning_rate": 0.00011622728107173946, | |
| "loss": 2.4254, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.4542041187848447, | |
| "grad_norm": 0.8503845930099487, | |
| "learning_rate": 0.00011594255335726724, | |
| "loss": 2.5187, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.4551143474798043, | |
| "grad_norm": 1.6775977611541748, | |
| "learning_rate": 0.00011565769296398618, | |
| "loss": 2.6669, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4560245761747639, | |
| "grad_norm": 0.45572495460510254, | |
| "learning_rate": 0.00011537270226259169, | |
| "loss": 2.5806, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.45693480486972354, | |
| "grad_norm": 0.45138293504714966, | |
| "learning_rate": 0.00011508758362486358, | |
| "loss": 2.3935, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.45784503356468315, | |
| "grad_norm": 0.4548013210296631, | |
| "learning_rate": 0.00011480233942364645, | |
| "loss": 2.321, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.45875526225964275, | |
| "grad_norm": 0.434442400932312, | |
| "learning_rate": 0.00011451697203282982, | |
| "loss": 2.375, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.45966549095460235, | |
| "grad_norm": 0.4139295816421509, | |
| "learning_rate": 0.00011423148382732853, | |
| "loss": 2.3997, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.46057571964956195, | |
| "grad_norm": 0.46020230650901794, | |
| "learning_rate": 0.00011394587718306275, | |
| "loss": 2.5745, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.46148594834452156, | |
| "grad_norm": 0.4194343090057373, | |
| "learning_rate": 0.00011366015447693837, | |
| "loss": 2.2597, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.46239617703948116, | |
| "grad_norm": 0.43983832001686096, | |
| "learning_rate": 0.0001133743180868273, | |
| "loss": 2.3511, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.46330640573444076, | |
| "grad_norm": 0.41047292947769165, | |
| "learning_rate": 0.00011308837039154739, | |
| "loss": 2.2614, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.46421663442940037, | |
| "grad_norm": 0.4110110104084015, | |
| "learning_rate": 0.0001128023137708429, | |
| "loss": 2.2719, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.46512686312435997, | |
| "grad_norm": 0.41848358511924744, | |
| "learning_rate": 0.0001125161506053646, | |
| "loss": 2.3872, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.4660370918193196, | |
| "grad_norm": 0.39852631092071533, | |
| "learning_rate": 0.00011222988327664997, | |
| "loss": 2.2001, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.46694732051427923, | |
| "grad_norm": 0.4060978293418884, | |
| "learning_rate": 0.00011194351416710324, | |
| "loss": 2.2474, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.46785754920923883, | |
| "grad_norm": 0.4010358452796936, | |
| "learning_rate": 0.00011165704565997593, | |
| "loss": 2.1262, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.46876777790419843, | |
| "grad_norm": 0.4063378572463989, | |
| "learning_rate": 0.00011137048013934656, | |
| "loss": 2.1583, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.46967800659915804, | |
| "grad_norm": 0.40287846326828003, | |
| "learning_rate": 0.00011108381999010111, | |
| "loss": 2.2351, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 0.3861018717288971, | |
| "learning_rate": 0.00011079706759791311, | |
| "loss": 2.195, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.47149846398907724, | |
| "grad_norm": 0.38855546712875366, | |
| "learning_rate": 0.00011051022534922371, | |
| "loss": 2.1575, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.47240869268403685, | |
| "grad_norm": 0.3941628038883209, | |
| "learning_rate": 0.00011022329563122191, | |
| "loss": 2.2324, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.47331892137899645, | |
| "grad_norm": 0.40604814887046814, | |
| "learning_rate": 0.00010993628083182467, | |
| "loss": 2.1641, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.4742291500739561, | |
| "grad_norm": 0.407815158367157, | |
| "learning_rate": 0.000109649183339657, | |
| "loss": 2.1648, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.4751393787689157, | |
| "grad_norm": 0.400680810213089, | |
| "learning_rate": 0.00010936200554403209, | |
| "loss": 2.1939, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.4760496074638753, | |
| "grad_norm": 0.416537344455719, | |
| "learning_rate": 0.00010907474983493144, | |
| "loss": 2.1694, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.4769598361588349, | |
| "grad_norm": 0.4097869396209717, | |
| "learning_rate": 0.00010878741860298503, | |
| "loss": 2.1785, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.4778700648537945, | |
| "grad_norm": 0.4243004024028778, | |
| "learning_rate": 0.00010850001423945126, | |
| "loss": 1.9963, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.4787802935487541, | |
| "grad_norm": 0.41958731412887573, | |
| "learning_rate": 0.00010821253913619726, | |
| "loss": 2.1629, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.4796905222437137, | |
| "grad_norm": 0.4177284240722656, | |
| "learning_rate": 0.00010792499568567884, | |
| "loss": 2.1276, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.4806007509386733, | |
| "grad_norm": 0.41077664494514465, | |
| "learning_rate": 0.00010763738628092062, | |
| "loss": 2.0852, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.48151097963363293, | |
| "grad_norm": 0.4098223149776459, | |
| "learning_rate": 0.00010734971331549603, | |
| "loss": 1.9977, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.48242120832859253, | |
| "grad_norm": 0.42255935072898865, | |
| "learning_rate": 0.00010706197918350758, | |
| "loss": 1.9822, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.4833314370235522, | |
| "grad_norm": 0.45597127079963684, | |
| "learning_rate": 0.0001067741862795668, | |
| "loss": 2.1072, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.4842416657185118, | |
| "grad_norm": 0.4538208544254303, | |
| "learning_rate": 0.0001064863369987743, | |
| "loss": 2.41, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.4851518944134714, | |
| "grad_norm": 0.4586673676967621, | |
| "learning_rate": 0.00010619843373669993, | |
| "loss": 2.1736, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.486062123108431, | |
| "grad_norm": 0.4433608055114746, | |
| "learning_rate": 0.00010591047888936274, | |
| "loss": 2.1324, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.4869723518033906, | |
| "grad_norm": 0.4421234428882599, | |
| "learning_rate": 0.00010562247485321115, | |
| "loss": 2.0689, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.4878825804983502, | |
| "grad_norm": 0.46843069791793823, | |
| "learning_rate": 0.00010533442402510284, | |
| "loss": 2.2252, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.4887928091933098, | |
| "grad_norm": 0.4747142493724823, | |
| "learning_rate": 0.00010504632880228498, | |
| "loss": 2.2503, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.4897030378882694, | |
| "grad_norm": 0.46643224358558655, | |
| "learning_rate": 0.00010475819158237425, | |
| "loss": 2.2628, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.490613266583229, | |
| "grad_norm": 0.47085490822792053, | |
| "learning_rate": 0.00010447001476333673, | |
| "loss": 2.0888, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.49152349527818867, | |
| "grad_norm": 0.5102598071098328, | |
| "learning_rate": 0.00010418180074346815, | |
| "loss": 2.2736, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.4924337239731483, | |
| "grad_norm": 0.49878573417663574, | |
| "learning_rate": 0.00010389355192137377, | |
| "loss": 2.1107, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.4933439526681079, | |
| "grad_norm": 0.5236616134643555, | |
| "learning_rate": 0.00010360527069594859, | |
| "loss": 2.4099, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.4942541813630675, | |
| "grad_norm": 0.49875032901763916, | |
| "learning_rate": 0.00010331695946635708, | |
| "loss": 2.1381, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.4951644100580271, | |
| "grad_norm": 0.5333012938499451, | |
| "learning_rate": 0.00010302862063201367, | |
| "loss": 2.2274, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.4960746387529867, | |
| "grad_norm": 0.5504993200302124, | |
| "learning_rate": 0.00010274025659256232, | |
| "loss": 2.2348, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.4969848674479463, | |
| "grad_norm": 0.5924202799797058, | |
| "learning_rate": 0.00010245186974785685, | |
| "loss": 2.3686, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.4978950961429059, | |
| "grad_norm": 0.6003567576408386, | |
| "learning_rate": 0.00010216346249794087, | |
| "loss": 2.3336, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.4988053248378655, | |
| "grad_norm": 0.6700019836425781, | |
| "learning_rate": 0.00010187503724302776, | |
| "loss": 2.4446, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.4997155535328251, | |
| "grad_norm": 0.8171781897544861, | |
| "learning_rate": 0.00010158659638348081, | |
| "loss": 2.4278, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.5006257822277848, | |
| "grad_norm": 1.4212020635604858, | |
| "learning_rate": 0.0001012981423197931, | |
| "loss": 2.6229, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5006257822277848, | |
| "eval_loss": 2.2479705810546875, | |
| "eval_runtime": 205.3622, | |
| "eval_samples_per_second": 9.013, | |
| "eval_steps_per_second": 4.509, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5015360109227444, | |
| "grad_norm": 0.4380887746810913, | |
| "learning_rate": 0.00010100967745256766, | |
| "loss": 2.4596, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.502446239617704, | |
| "grad_norm": 0.4522015154361725, | |
| "learning_rate": 0.00010072120418249745, | |
| "loss": 2.3217, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.5033564683126636, | |
| "grad_norm": 0.4581010043621063, | |
| "learning_rate": 0.00010043272491034523, | |
| "loss": 2.4948, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.5042666970076232, | |
| "grad_norm": 0.4353744685649872, | |
| "learning_rate": 0.00010014424203692388, | |
| "loss": 2.3769, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.5051769257025828, | |
| "grad_norm": 0.4150161147117615, | |
| "learning_rate": 9.985575796307615e-05, | |
| "loss": 2.3557, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.5060871543975424, | |
| "grad_norm": 0.4532707631587982, | |
| "learning_rate": 9.956727508965481e-05, | |
| "loss": 2.3114, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.506997383092502, | |
| "grad_norm": 0.42450255155563354, | |
| "learning_rate": 9.927879581750259e-05, | |
| "loss": 2.2911, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.5079076117874616, | |
| "grad_norm": 0.42910221219062805, | |
| "learning_rate": 9.899032254743235e-05, | |
| "loss": 2.3062, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.5088178404824212, | |
| "grad_norm": 0.42122408747673035, | |
| "learning_rate": 9.870185768020693e-05, | |
| "loss": 2.3294, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.5097280691773808, | |
| "grad_norm": 0.4562203884124756, | |
| "learning_rate": 9.84134036165192e-05, | |
| "loss": 2.2674, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.5106382978723404, | |
| "grad_norm": 0.3905525207519531, | |
| "learning_rate": 9.812496275697226e-05, | |
| "loss": 2.1259, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.5115485265673, | |
| "grad_norm": 0.41641199588775635, | |
| "learning_rate": 9.783653750205915e-05, | |
| "loss": 2.2191, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.5124587552622596, | |
| "grad_norm": 0.4090450704097748, | |
| "learning_rate": 9.754813025214317e-05, | |
| "loss": 2.2478, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.5133689839572193, | |
| "grad_norm": 0.4293293356895447, | |
| "learning_rate": 9.725974340743769e-05, | |
| "loss": 2.3854, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.5142792126521789, | |
| "grad_norm": 0.4138126075267792, | |
| "learning_rate": 9.697137936798634e-05, | |
| "loss": 2.2903, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.5151894413471385, | |
| "grad_norm": 0.3979492783546448, | |
| "learning_rate": 9.668304053364294e-05, | |
| "loss": 2.1878, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.5160996700420981, | |
| "grad_norm": 0.38530561327934265, | |
| "learning_rate": 9.639472930405143e-05, | |
| "loss": 2.1464, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.5170098987370577, | |
| "grad_norm": 0.4263143837451935, | |
| "learning_rate": 9.610644807862625e-05, | |
| "loss": 2.1856, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.5179201274320173, | |
| "grad_norm": 0.41127926111221313, | |
| "learning_rate": 9.581819925653188e-05, | |
| "loss": 2.198, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.5188303561269769, | |
| "grad_norm": 0.3917299509048462, | |
| "learning_rate": 9.552998523666326e-05, | |
| "loss": 2.1325, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.5197405848219365, | |
| "grad_norm": 0.39180508255958557, | |
| "learning_rate": 9.524180841762577e-05, | |
| "loss": 2.0702, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.5206508135168961, | |
| "grad_norm": 0.3829837739467621, | |
| "learning_rate": 9.495367119771503e-05, | |
| "loss": 1.8913, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.5215610422118557, | |
| "grad_norm": 0.412706196308136, | |
| "learning_rate": 9.46655759748972e-05, | |
| "loss": 2.1221, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.5224712709068153, | |
| "grad_norm": 0.39748141169548035, | |
| "learning_rate": 9.437752514678887e-05, | |
| "loss": 2.0427, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.5233814996017749, | |
| "grad_norm": 0.42854538559913635, | |
| "learning_rate": 9.408952111063727e-05, | |
| "loss": 2.121, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.5242917282967345, | |
| "grad_norm": 0.414654016494751, | |
| "learning_rate": 9.380156626330009e-05, | |
| "loss": 2.038, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.5252019569916941, | |
| "grad_norm": 0.4241427183151245, | |
| "learning_rate": 9.35136630012257e-05, | |
| "loss": 2.1312, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.5261121856866537, | |
| "grad_norm": 0.42928779125213623, | |
| "learning_rate": 9.322581372043321e-05, | |
| "loss": 2.1875, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.5270224143816133, | |
| "grad_norm": 0.4133308231830597, | |
| "learning_rate": 9.293802081649243e-05, | |
| "loss": 2.0477, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.527932643076573, | |
| "grad_norm": 0.427898645401001, | |
| "learning_rate": 9.265028668450402e-05, | |
| "loss": 2.0833, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.5288428717715326, | |
| "grad_norm": 0.4321751892566681, | |
| "learning_rate": 9.23626137190794e-05, | |
| "loss": 2.1698, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.5297531004664922, | |
| "grad_norm": 0.4715782105922699, | |
| "learning_rate": 9.207500431432115e-05, | |
| "loss": 2.1347, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.5306633291614519, | |
| "grad_norm": 0.45599547028541565, | |
| "learning_rate": 9.178746086380275e-05, | |
| "loss": 2.1469, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.5315735578564115, | |
| "grad_norm": 0.45286545157432556, | |
| "learning_rate": 9.149998576054874e-05, | |
| "loss": 2.2013, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.5324837865513711, | |
| "grad_norm": 0.47089457511901855, | |
| "learning_rate": 9.121258139701502e-05, | |
| "loss": 2.2125, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.5333940152463307, | |
| "grad_norm": 0.46750229597091675, | |
| "learning_rate": 9.092525016506858e-05, | |
| "loss": 2.1186, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.5343042439412903, | |
| "grad_norm": 0.4931905269622803, | |
| "learning_rate": 9.063799445596795e-05, | |
| "loss": 2.2185, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.5352144726362499, | |
| "grad_norm": 0.48538026213645935, | |
| "learning_rate": 9.035081666034304e-05, | |
| "loss": 2.2369, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.5361247013312095, | |
| "grad_norm": 0.4944066107273102, | |
| "learning_rate": 9.006371916817534e-05, | |
| "loss": 2.2771, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.5370349300261691, | |
| "grad_norm": 0.4564894139766693, | |
| "learning_rate": 8.977670436877811e-05, | |
| "loss": 2.0879, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.5379451587211287, | |
| "grad_norm": 0.5046347379684448, | |
| "learning_rate": 8.948977465077632e-05, | |
| "loss": 2.2197, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.5388553874160883, | |
| "grad_norm": 0.49683472514152527, | |
| "learning_rate": 8.920293240208694e-05, | |
| "loss": 2.2152, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.5397656161110479, | |
| "grad_norm": 0.5223331451416016, | |
| "learning_rate": 8.891618000989891e-05, | |
| "loss": 2.3358, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.5406758448060075, | |
| "grad_norm": 0.5552563667297363, | |
| "learning_rate": 8.862951986065345e-05, | |
| "loss": 2.1608, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.5415860735009671, | |
| "grad_norm": 0.5853347778320312, | |
| "learning_rate": 8.83429543400241e-05, | |
| "loss": 2.3679, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.5424963021959267, | |
| "grad_norm": 0.5858141183853149, | |
| "learning_rate": 8.805648583289674e-05, | |
| "loss": 2.341, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.5434065308908863, | |
| "grad_norm": 0.6405509114265442, | |
| "learning_rate": 8.777011672335008e-05, | |
| "loss": 2.4773, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.5443167595858459, | |
| "grad_norm": 0.7342801094055176, | |
| "learning_rate": 8.748384939463543e-05, | |
| "loss": 2.557, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.5452269882808055, | |
| "grad_norm": 0.8813995122909546, | |
| "learning_rate": 8.719768622915714e-05, | |
| "loss": 2.5595, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.5461372169757651, | |
| "grad_norm": 1.722114086151123, | |
| "learning_rate": 8.691162960845264e-05, | |
| "loss": 2.7211, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5470474456707247, | |
| "grad_norm": 0.424265056848526, | |
| "learning_rate": 8.662568191317273e-05, | |
| "loss": 2.3728, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.5479576743656844, | |
| "grad_norm": 0.45933809876441956, | |
| "learning_rate": 8.633984552306164e-05, | |
| "loss": 2.4234, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.548867903060644, | |
| "grad_norm": 0.45455530285835266, | |
| "learning_rate": 8.605412281693727e-05, | |
| "loss": 2.5062, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.5497781317556036, | |
| "grad_norm": 0.4334143400192261, | |
| "learning_rate": 8.57685161726715e-05, | |
| "loss": 2.3087, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.5506883604505632, | |
| "grad_norm": 0.4537433385848999, | |
| "learning_rate": 8.548302796717019e-05, | |
| "loss": 2.395, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.5515985891455228, | |
| "grad_norm": 0.43673837184906006, | |
| "learning_rate": 8.519766057635355e-05, | |
| "loss": 2.3855, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.5525088178404824, | |
| "grad_norm": 0.43078145384788513, | |
| "learning_rate": 8.491241637513644e-05, | |
| "loss": 2.2559, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.553419046535442, | |
| "grad_norm": 0.4094640612602234, | |
| "learning_rate": 8.462729773740832e-05, | |
| "loss": 2.295, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.5543292752304017, | |
| "grad_norm": 0.4126126170158386, | |
| "learning_rate": 8.434230703601384e-05, | |
| "loss": 2.2019, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.5552395039253613, | |
| "grad_norm": 0.4372231066226959, | |
| "learning_rate": 8.405744664273278e-05, | |
| "loss": 2.4243, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5561497326203209, | |
| "grad_norm": 0.42160138487815857, | |
| "learning_rate": 8.37727189282606e-05, | |
| "loss": 2.2805, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.5570599613152805, | |
| "grad_norm": 0.4336857795715332, | |
| "learning_rate": 8.34881262621884e-05, | |
| "loss": 2.4811, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.5579701900102401, | |
| "grad_norm": 0.40520837903022766, | |
| "learning_rate": 8.320367101298351e-05, | |
| "loss": 2.1731, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.5588804187051997, | |
| "grad_norm": 0.42664197087287903, | |
| "learning_rate": 8.291935554796962e-05, | |
| "loss": 2.3403, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.5597906474001593, | |
| "grad_norm": 0.4109039902687073, | |
| "learning_rate": 8.263518223330697e-05, | |
| "loss": 2.2425, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.5607008760951189, | |
| "grad_norm": 0.4032575786113739, | |
| "learning_rate": 8.235115343397295e-05, | |
| "loss": 2.2593, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.5616111047900785, | |
| "grad_norm": 0.3929396867752075, | |
| "learning_rate": 8.206727151374207e-05, | |
| "loss": 2.0896, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.5625213334850381, | |
| "grad_norm": 0.38567835092544556, | |
| "learning_rate": 8.178353883516664e-05, | |
| "loss": 2.0715, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.5634315621799977, | |
| "grad_norm": 0.405369371175766, | |
| "learning_rate": 8.149995775955686e-05, | |
| "loss": 2.2249, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.5643417908749573, | |
| "grad_norm": 0.3889697790145874, | |
| "learning_rate": 8.121653064696118e-05, | |
| "loss": 2.0797, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.565252019569917, | |
| "grad_norm": 0.4065384864807129, | |
| "learning_rate": 8.093325985614685e-05, | |
| "loss": 2.2012, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.5661622482648766, | |
| "grad_norm": 0.4066416323184967, | |
| "learning_rate": 8.065014774458003e-05, | |
| "loss": 2.1183, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.5670724769598362, | |
| "grad_norm": 0.40575870871543884, | |
| "learning_rate": 8.036719666840647e-05, | |
| "loss": 2.0258, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.5679827056547958, | |
| "grad_norm": 0.42911243438720703, | |
| "learning_rate": 8.008440898243149e-05, | |
| "loss": 2.1186, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.5688929343497554, | |
| "grad_norm": 0.4009549021720886, | |
| "learning_rate": 7.980178704010089e-05, | |
| "loss": 2.0049, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.569803163044715, | |
| "grad_norm": 0.41156989336013794, | |
| "learning_rate": 7.951933319348095e-05, | |
| "loss": 2.0272, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.5707133917396746, | |
| "grad_norm": 0.4248954653739929, | |
| "learning_rate": 7.923704979323899e-05, | |
| "loss": 2.077, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.5716236204346342, | |
| "grad_norm": 0.45484524965286255, | |
| "learning_rate": 7.895493918862396e-05, | |
| "loss": 2.2255, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.5725338491295938, | |
| "grad_norm": 0.4571921229362488, | |
| "learning_rate": 7.867300372744657e-05, | |
| "loss": 2.1373, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.5734440778245534, | |
| "grad_norm": 0.44238901138305664, | |
| "learning_rate": 7.839124575606004e-05, | |
| "loss": 2.1147, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.574354306519513, | |
| "grad_norm": 0.4206310510635376, | |
| "learning_rate": 7.810966761934053e-05, | |
| "loss": 2.0508, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.5752645352144726, | |
| "grad_norm": 0.43381330370903015, | |
| "learning_rate": 7.782827166066739e-05, | |
| "loss": 2.0847, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.5761747639094322, | |
| "grad_norm": 0.4460139572620392, | |
| "learning_rate": 7.754706022190398e-05, | |
| "loss": 2.1288, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.5770849926043918, | |
| "grad_norm": 0.4371720850467682, | |
| "learning_rate": 7.726603564337791e-05, | |
| "loss": 2.0476, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.5779952212993514, | |
| "grad_norm": 0.4623599052429199, | |
| "learning_rate": 7.69852002638618e-05, | |
| "loss": 2.2858, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.578905449994311, | |
| "grad_norm": 0.4422992765903473, | |
| "learning_rate": 7.670455642055361e-05, | |
| "loss": 2.1072, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.5798156786892706, | |
| "grad_norm": 0.4804936647415161, | |
| "learning_rate": 7.642410644905726e-05, | |
| "loss": 2.2218, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.5807259073842302, | |
| "grad_norm": 0.48249900341033936, | |
| "learning_rate": 7.614385268336336e-05, | |
| "loss": 2.2916, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.5816361360791898, | |
| "grad_norm": 0.46635982394218445, | |
| "learning_rate": 7.586379745582944e-05, | |
| "loss": 2.1636, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.5825463647741494, | |
| "grad_norm": 0.4670505225658417, | |
| "learning_rate": 7.558394309716088e-05, | |
| "loss": 2.2052, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.5834565934691092, | |
| "grad_norm": 0.49475541710853577, | |
| "learning_rate": 7.530429193639128e-05, | |
| "loss": 2.18, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.5843668221640688, | |
| "grad_norm": 0.5231596231460571, | |
| "learning_rate": 7.502484630086318e-05, | |
| "loss": 2.2095, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.5852770508590284, | |
| "grad_norm": 0.5045900344848633, | |
| "learning_rate": 7.474560851620873e-05, | |
| "loss": 2.053, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.586187279553988, | |
| "grad_norm": 0.5511046051979065, | |
| "learning_rate": 7.446658090633026e-05, | |
| "loss": 2.2706, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.5870975082489476, | |
| "grad_norm": 0.5700446963310242, | |
| "learning_rate": 7.41877657933809e-05, | |
| "loss": 2.3999, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.5880077369439072, | |
| "grad_norm": 0.5792605876922607, | |
| "learning_rate": 7.390916549774536e-05, | |
| "loss": 2.2391, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.5889179656388668, | |
| "grad_norm": 0.6770455241203308, | |
| "learning_rate": 7.363078233802063e-05, | |
| "loss": 2.6564, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.5898281943338264, | |
| "grad_norm": 0.7092955708503723, | |
| "learning_rate": 7.335261863099651e-05, | |
| "loss": 2.4722, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.590738423028786, | |
| "grad_norm": 0.8125056028366089, | |
| "learning_rate": 7.307467669163655e-05, | |
| "loss": 2.3581, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.5916486517237456, | |
| "grad_norm": 1.5941253900527954, | |
| "learning_rate": 7.279695883305866e-05, | |
| "loss": 2.16, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.5925588804187052, | |
| "grad_norm": 0.44398507475852966, | |
| "learning_rate": 7.251946736651582e-05, | |
| "loss": 2.4689, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.5934691091136648, | |
| "grad_norm": 0.4264509975910187, | |
| "learning_rate": 7.224220460137701e-05, | |
| "loss": 2.4081, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.5943793378086244, | |
| "grad_norm": 0.4222484529018402, | |
| "learning_rate": 7.196517284510773e-05, | |
| "loss": 2.3827, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.595289566503584, | |
| "grad_norm": 0.4516051411628723, | |
| "learning_rate": 7.168837440325114e-05, | |
| "loss": 2.399, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.5961997951985436, | |
| "grad_norm": 0.4370306730270386, | |
| "learning_rate": 7.141181157940859e-05, | |
| "loss": 2.3837, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.5971100238935032, | |
| "grad_norm": 0.4236253798007965, | |
| "learning_rate": 7.11354866752205e-05, | |
| "loss": 2.3066, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.5980202525884628, | |
| "grad_norm": 0.41718846559524536, | |
| "learning_rate": 7.085940199034735e-05, | |
| "loss": 2.3841, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.5989304812834224, | |
| "grad_norm": 0.4299750030040741, | |
| "learning_rate": 7.058355982245037e-05, | |
| "loss": 2.3842, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.599840709978382, | |
| "grad_norm": 0.4180915057659149, | |
| "learning_rate": 7.030796246717255e-05, | |
| "loss": 2.0758, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.6007509386733417, | |
| "grad_norm": 0.45195114612579346, | |
| "learning_rate": 7.003261221811934e-05, | |
| "loss": 2.4826, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.6016611673683013, | |
| "grad_norm": 0.4253404140472412, | |
| "learning_rate": 6.97575113668399e-05, | |
| "loss": 2.3705, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.6025713960632609, | |
| "grad_norm": 0.4198931157588959, | |
| "learning_rate": 6.948266220280771e-05, | |
| "loss": 2.3396, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.6034816247582205, | |
| "grad_norm": 0.43457460403442383, | |
| "learning_rate": 6.920806701340155e-05, | |
| "loss": 2.1447, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.6043918534531801, | |
| "grad_norm": 0.40161874890327454, | |
| "learning_rate": 6.893372808388675e-05, | |
| "loss": 2.2443, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.6053020821481397, | |
| "grad_norm": 0.4039609432220459, | |
| "learning_rate": 6.865964769739575e-05, | |
| "loss": 2.1815, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.6062123108430993, | |
| "grad_norm": 0.4061351716518402, | |
| "learning_rate": 6.838582813490947e-05, | |
| "loss": 2.1073, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.607122539538059, | |
| "grad_norm": 0.4206211268901825, | |
| "learning_rate": 6.811227167523815e-05, | |
| "loss": 2.2549, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.6080327682330186, | |
| "grad_norm": 0.3936857283115387, | |
| "learning_rate": 6.783898059500233e-05, | |
| "loss": 2.1373, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.6089429969279782, | |
| "grad_norm": 0.3954029083251953, | |
| "learning_rate": 6.756595716861407e-05, | |
| "loss": 2.1001, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.6098532256229378, | |
| "grad_norm": 0.407713919878006, | |
| "learning_rate": 6.729320366825784e-05, | |
| "loss": 2.0967, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.6107634543178974, | |
| "grad_norm": 0.41096213459968567, | |
| "learning_rate": 6.702072236387182e-05, | |
| "loss": 2.0899, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.611673683012857, | |
| "grad_norm": 0.40465790033340454, | |
| "learning_rate": 6.674851552312878e-05, | |
| "loss": 2.089, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.6125839117078166, | |
| "grad_norm": 0.3991434574127197, | |
| "learning_rate": 6.647658541141735e-05, | |
| "loss": 1.9788, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.6134941404027762, | |
| "grad_norm": 0.42327383160591125, | |
| "learning_rate": 6.620493429182323e-05, | |
| "loss": 2.1672, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.6144043690977358, | |
| "grad_norm": 0.4061299264431, | |
| "learning_rate": 6.593356442511015e-05, | |
| "loss": 2.1617, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.6153145977926954, | |
| "grad_norm": 0.4082658588886261, | |
| "learning_rate": 6.566247806970119e-05, | |
| "loss": 2.0112, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.616224826487655, | |
| "grad_norm": 0.43016669154167175, | |
| "learning_rate": 6.539167748165994e-05, | |
| "loss": 2.0, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.6171350551826146, | |
| "grad_norm": 0.43142226338386536, | |
| "learning_rate": 6.512116491467185e-05, | |
| "loss": 2.1585, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.6180452838775743, | |
| "grad_norm": 0.4271491467952728, | |
| "learning_rate": 6.485094262002529e-05, | |
| "loss": 1.9628, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.6189555125725339, | |
| "grad_norm": 0.44002971053123474, | |
| "learning_rate": 6.458101284659286e-05, | |
| "loss": 2.2214, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.6198657412674935, | |
| "grad_norm": 0.4215126931667328, | |
| "learning_rate": 6.431137784081282e-05, | |
| "loss": 2.0377, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.6207759699624531, | |
| "grad_norm": 0.46792343258857727, | |
| "learning_rate": 6.404203984667019e-05, | |
| "loss": 2.029, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.6216861986574127, | |
| "grad_norm": 0.45737308263778687, | |
| "learning_rate": 6.377300110567821e-05, | |
| "loss": 2.2375, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.6225964273523723, | |
| "grad_norm": 0.4526033401489258, | |
| "learning_rate": 6.350426385685957e-05, | |
| "loss": 2.2562, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.6235066560473319, | |
| "grad_norm": 0.45917776226997375, | |
| "learning_rate": 6.323583033672799e-05, | |
| "loss": 2.1321, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.6244168847422915, | |
| "grad_norm": 0.4713301658630371, | |
| "learning_rate": 6.296770277926937e-05, | |
| "loss": 2.07, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.6253271134372511, | |
| "grad_norm": 0.5036799907684326, | |
| "learning_rate": 6.269988341592328e-05, | |
| "loss": 2.1103, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.6262373421322107, | |
| "grad_norm": 0.4843004643917084, | |
| "learning_rate": 6.243237447556449e-05, | |
| "loss": 2.0936, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.6271475708271703, | |
| "grad_norm": 0.4738497734069824, | |
| "learning_rate": 6.216517818448423e-05, | |
| "loss": 2.1004, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.6280577995221299, | |
| "grad_norm": 0.5081862211227417, | |
| "learning_rate": 6.189829676637182e-05, | |
| "loss": 2.2177, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.6289680282170895, | |
| "grad_norm": 0.5060831904411316, | |
| "learning_rate": 6.163173244229619e-05, | |
| "loss": 2.1342, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.6298782569120491, | |
| "grad_norm": 0.5047332644462585, | |
| "learning_rate": 6.136548743068713e-05, | |
| "loss": 2.0727, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.6307884856070087, | |
| "grad_norm": 0.5174648761749268, | |
| "learning_rate": 6.109956394731722e-05, | |
| "loss": 2.0623, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.6316987143019683, | |
| "grad_norm": 0.5705139636993408, | |
| "learning_rate": 6.083396420528298e-05, | |
| "loss": 2.4454, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.6326089429969279, | |
| "grad_norm": 0.5653088092803955, | |
| "learning_rate": 6.056869041498687e-05, | |
| "loss": 2.2071, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.6335191716918875, | |
| "grad_norm": 0.6105946898460388, | |
| "learning_rate": 6.030374478411847e-05, | |
| "loss": 2.3081, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.6344294003868471, | |
| "grad_norm": 0.6362658143043518, | |
| "learning_rate": 6.0039129517636435e-05, | |
| "loss": 2.3426, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.6353396290818069, | |
| "grad_norm": 0.7242766618728638, | |
| "learning_rate": 5.9774846817750105e-05, | |
| "loss": 2.4877, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.6362498577767665, | |
| "grad_norm": 0.9446219205856323, | |
| "learning_rate": 5.951089888390087e-05, | |
| "loss": 2.7741, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.6371600864717261, | |
| "grad_norm": 1.5826219320297241, | |
| "learning_rate": 5.924728791274432e-05, | |
| "loss": 2.533, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.6380703151666857, | |
| "grad_norm": 0.4994019567966461, | |
| "learning_rate": 5.89840160981316e-05, | |
| "loss": 2.4388, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.6389805438616453, | |
| "grad_norm": 0.45213326811790466, | |
| "learning_rate": 5.872108563109131e-05, | |
| "loss": 2.3644, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.6398907725566049, | |
| "grad_norm": 0.45655617117881775, | |
| "learning_rate": 5.845849869981137e-05, | |
| "loss": 2.5202, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.6408010012515645, | |
| "grad_norm": 0.41640329360961914, | |
| "learning_rate": 5.819625748962049e-05, | |
| "loss": 2.3097, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.6417112299465241, | |
| "grad_norm": 0.43625307083129883, | |
| "learning_rate": 5.79343641829704e-05, | |
| "loss": 2.3931, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.6426214586414837, | |
| "grad_norm": 0.44178506731987, | |
| "learning_rate": 5.7672820959417254e-05, | |
| "loss": 2.3195, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.6435316873364433, | |
| "grad_norm": 0.4416089951992035, | |
| "learning_rate": 5.741162999560386e-05, | |
| "loss": 2.2446, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.6444419160314029, | |
| "grad_norm": 0.4419204890727997, | |
| "learning_rate": 5.7150793465241346e-05, | |
| "loss": 2.34, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.6453521447263625, | |
| "grad_norm": 0.45422717928886414, | |
| "learning_rate": 5.68903135390912e-05, | |
| "loss": 2.1915, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.6462623734213221, | |
| "grad_norm": 0.41635921597480774, | |
| "learning_rate": 5.663019238494704e-05, | |
| "loss": 2.3147, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.6471726021162817, | |
| "grad_norm": 0.4240402579307556, | |
| "learning_rate": 5.637043216761678e-05, | |
| "loss": 2.1693, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.6480828308112413, | |
| "grad_norm": 0.41989627480506897, | |
| "learning_rate": 5.611103504890444e-05, | |
| "loss": 2.2087, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.6489930595062009, | |
| "grad_norm": 0.4187220335006714, | |
| "learning_rate": 5.5852003187592226e-05, | |
| "loss": 2.3818, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.6499032882011605, | |
| "grad_norm": 0.43209201097488403, | |
| "learning_rate": 5.559333873942259e-05, | |
| "loss": 2.3176, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.6508135168961201, | |
| "grad_norm": 0.416291207075119, | |
| "learning_rate": 5.533504385708024e-05, | |
| "loss": 2.2397, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.6517237455910797, | |
| "grad_norm": 0.411857932806015, | |
| "learning_rate": 5.5077120690174246e-05, | |
| "loss": 2.1142, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.6526339742860394, | |
| "grad_norm": 0.4142412543296814, | |
| "learning_rate": 5.481957138522018e-05, | |
| "loss": 2.2226, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.653544202980999, | |
| "grad_norm": 0.4322018325328827, | |
| "learning_rate": 5.456239808562209e-05, | |
| "loss": 2.2078, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.6544544316759586, | |
| "grad_norm": 0.42210081219673157, | |
| "learning_rate": 5.4305602931655045e-05, | |
| "loss": 2.0579, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.6553646603709182, | |
| "grad_norm": 0.4076862335205078, | |
| "learning_rate": 5.404918806044679e-05, | |
| "loss": 2.1348, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.6562748890658778, | |
| "grad_norm": 0.43610820174217224, | |
| "learning_rate": 5.379315560596038e-05, | |
| "loss": 2.2212, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.6571851177608374, | |
| "grad_norm": 0.41793620586395264, | |
| "learning_rate": 5.3537507698976365e-05, | |
| "loss": 1.9606, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.658095346455797, | |
| "grad_norm": 0.40680915117263794, | |
| "learning_rate": 5.328224646707479e-05, | |
| "loss": 2.0167, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.6590055751507566, | |
| "grad_norm": 0.424699991941452, | |
| "learning_rate": 5.3027374034617785e-05, | |
| "loss": 2.1065, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.6599158038457162, | |
| "grad_norm": 0.43258509039878845, | |
| "learning_rate": 5.277289252273174e-05, | |
| "loss": 2.0974, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.6608260325406758, | |
| "grad_norm": 0.4318636655807495, | |
| "learning_rate": 5.251880404928971e-05, | |
| "loss": 2.3214, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.6617362612356354, | |
| "grad_norm": 0.4147786498069763, | |
| "learning_rate": 5.226511072889371e-05, | |
| "loss": 2.1223, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 0.662646489930595, | |
| "grad_norm": 0.4131387770175934, | |
| "learning_rate": 5.201181467285723e-05, | |
| "loss": 1.8335, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.6635567186255547, | |
| "grad_norm": 0.456827849149704, | |
| "learning_rate": 5.175891798918757e-05, | |
| "loss": 2.1428, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.6644669473205143, | |
| "grad_norm": 0.47478604316711426, | |
| "learning_rate": 5.1506422782568345e-05, | |
| "loss": 2.0526, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.6653771760154739, | |
| "grad_norm": 0.4357281029224396, | |
| "learning_rate": 5.125433115434197e-05, | |
| "loss": 1.8949, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 0.6662874047104335, | |
| "grad_norm": 0.45749080181121826, | |
| "learning_rate": 5.100264520249205e-05, | |
| "loss": 2.1637, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.6671976334053931, | |
| "grad_norm": 0.47157686948776245, | |
| "learning_rate": 5.0751367021626215e-05, | |
| "loss": 2.1036, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 0.6681078621003527, | |
| "grad_norm": 0.4490397274494171, | |
| "learning_rate": 5.050049870295841e-05, | |
| "loss": 1.9553, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.6690180907953123, | |
| "grad_norm": 0.4798765480518341, | |
| "learning_rate": 5.025004233429145e-05, | |
| "loss": 2.0954, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.669928319490272, | |
| "grad_norm": 0.5034172534942627, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 2.2027, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.6708385481852316, | |
| "grad_norm": 0.481141060590744, | |
| "learning_rate": 4.9750373781012885e-05, | |
| "loss": 2.0822, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.6717487768801912, | |
| "grad_norm": 0.49731481075286865, | |
| "learning_rate": 4.950116575479586e-05, | |
| "loss": 2.0196, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.6726590055751508, | |
| "grad_norm": 0.48321208357810974, | |
| "learning_rate": 4.9252377995334444e-05, | |
| "loss": 1.9995, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 0.6735692342701104, | |
| "grad_norm": 0.5173296332359314, | |
| "learning_rate": 4.90040125731165e-05, | |
| "loss": 2.2078, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.67447946296507, | |
| "grad_norm": 0.5107961297035217, | |
| "learning_rate": 4.87560715551151e-05, | |
| "loss": 2.0684, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.6753896916600296, | |
| "grad_norm": 0.522492527961731, | |
| "learning_rate": 4.85085570047713e-05, | |
| "loss": 2.2168, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.6762999203549892, | |
| "grad_norm": 0.5417767763137817, | |
| "learning_rate": 4.826147098197691e-05, | |
| "loss": 2.2379, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 0.6772101490499488, | |
| "grad_norm": 0.5735164284706116, | |
| "learning_rate": 4.8014815543057475e-05, | |
| "loss": 2.2132, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.6781203777449084, | |
| "grad_norm": 0.5819071531295776, | |
| "learning_rate": 4.776859274075506e-05, | |
| "loss": 2.2469, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.679030606439868, | |
| "grad_norm": 0.6113678216934204, | |
| "learning_rate": 4.752280462421117e-05, | |
| "loss": 2.3064, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.6799408351348276, | |
| "grad_norm": 0.6506679654121399, | |
| "learning_rate": 4.727745323894976e-05, | |
| "loss": 2.3311, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.6808510638297872, | |
| "grad_norm": 0.7372251152992249, | |
| "learning_rate": 4.703254062686017e-05, | |
| "loss": 2.575, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.6817612925247468, | |
| "grad_norm": 0.8235337734222412, | |
| "learning_rate": 4.678806882618003e-05, | |
| "loss": 2.4711, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 0.6826715212197064, | |
| "grad_norm": 1.295682430267334, | |
| "learning_rate": 4.654403987147865e-05, | |
| "loss": 2.5713, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.683581749914666, | |
| "grad_norm": 0.46585023403167725, | |
| "learning_rate": 4.630045579363957e-05, | |
| "loss": 2.4203, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 0.6844919786096256, | |
| "grad_norm": 0.4163340628147125, | |
| "learning_rate": 4.605731861984401e-05, | |
| "loss": 2.261, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.6854022073045852, | |
| "grad_norm": 0.4268229007720947, | |
| "learning_rate": 4.5814630373554115e-05, | |
| "loss": 2.3078, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.6863124359995448, | |
| "grad_norm": 0.43664565682411194, | |
| "learning_rate": 4.557239307449561e-05, | |
| "loss": 2.5044, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.6872226646945045, | |
| "grad_norm": 0.44144201278686523, | |
| "learning_rate": 4.5330608738641486e-05, | |
| "loss": 2.4192, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.6881328933894642, | |
| "grad_norm": 0.40867024660110474, | |
| "learning_rate": 4.508927937819499e-05, | |
| "loss": 2.1908, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.6890431220844238, | |
| "grad_norm": 0.4147084057331085, | |
| "learning_rate": 4.484840700157295e-05, | |
| "loss": 2.2864, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 0.6899533507793834, | |
| "grad_norm": 0.413703054189682, | |
| "learning_rate": 4.4607993613388976e-05, | |
| "loss": 2.2436, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.690863579474343, | |
| "grad_norm": 0.42309460043907166, | |
| "learning_rate": 4.436804121443689e-05, | |
| "loss": 2.3444, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.6917738081693026, | |
| "grad_norm": 0.4229698181152344, | |
| "learning_rate": 4.412855180167406e-05, | |
| "loss": 2.3264, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.6926840368642622, | |
| "grad_norm": 0.4247763156890869, | |
| "learning_rate": 4.388952736820453e-05, | |
| "loss": 2.257, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 0.6935942655592218, | |
| "grad_norm": 0.42383337020874023, | |
| "learning_rate": 4.365096990326297e-05, | |
| "loss": 2.1349, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.6945044942541814, | |
| "grad_norm": 0.4181368350982666, | |
| "learning_rate": 4.3412881392197526e-05, | |
| "loss": 2.2587, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 0.695414722949141, | |
| "grad_norm": 0.4086921811103821, | |
| "learning_rate": 4.317526381645363e-05, | |
| "loss": 2.2378, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.6963249516441006, | |
| "grad_norm": 0.4001654088497162, | |
| "learning_rate": 4.293811915355761e-05, | |
| "loss": 2.1708, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.6972351803390602, | |
| "grad_norm": 0.42960214614868164, | |
| "learning_rate": 4.270144937709981e-05, | |
| "loss": 2.1556, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.6981454090340198, | |
| "grad_norm": 0.42000848054885864, | |
| "learning_rate": 4.2465256456718615e-05, | |
| "loss": 2.1182, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 0.6990556377289794, | |
| "grad_norm": 0.394368052482605, | |
| "learning_rate": 4.222954235808378e-05, | |
| "loss": 2.0486, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.699965866423939, | |
| "grad_norm": 0.3934192359447479, | |
| "learning_rate": 4.19943090428802e-05, | |
| "loss": 1.9222, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 0.7008760951188986, | |
| "grad_norm": 0.44399651885032654, | |
| "learning_rate": 4.175955846879151e-05, | |
| "loss": 2.1621, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.7017863238138582, | |
| "grad_norm": 0.4081316888332367, | |
| "learning_rate": 4.1525292589483843e-05, | |
| "loss": 1.9534, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.7026965525088178, | |
| "grad_norm": 0.42008188366889954, | |
| "learning_rate": 4.129151335458957e-05, | |
| "loss": 1.9773, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.7036067812037774, | |
| "grad_norm": 0.4209793508052826, | |
| "learning_rate": 4.105822270969102e-05, | |
| "loss": 2.0403, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 0.7045170098987371, | |
| "grad_norm": 0.43969592452049255, | |
| "learning_rate": 4.0825422596304396e-05, | |
| "loss": 2.1796, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.7054272385936967, | |
| "grad_norm": 0.4333605468273163, | |
| "learning_rate": 4.059311495186338e-05, | |
| "loss": 2.1115, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.7063374672886563, | |
| "grad_norm": 0.42669251561164856, | |
| "learning_rate": 4.036130170970341e-05, | |
| "loss": 2.1563, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.7072476959836159, | |
| "grad_norm": 0.45064857602119446, | |
| "learning_rate": 4.012998479904525e-05, | |
| "loss": 2.2003, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.7081579246785755, | |
| "grad_norm": 0.4487575888633728, | |
| "learning_rate": 3.9899166144978904e-05, | |
| "loss": 2.1332, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.7090681533735351, | |
| "grad_norm": 0.4423072636127472, | |
| "learning_rate": 3.966884766844803e-05, | |
| "loss": 2.1449, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 0.7099783820684947, | |
| "grad_norm": 0.4436761736869812, | |
| "learning_rate": 3.943903128623335e-05, | |
| "loss": 2.0025, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.7108886107634543, | |
| "grad_norm": 0.4364500045776367, | |
| "learning_rate": 3.920971891093718e-05, | |
| "loss": 1.9834, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.7117988394584139, | |
| "grad_norm": 0.44009071588516235, | |
| "learning_rate": 3.8980912450967366e-05, | |
| "loss": 2.0204, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.7127090681533735, | |
| "grad_norm": 0.4498206079006195, | |
| "learning_rate": 3.875261381052121e-05, | |
| "loss": 2.0348, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.7136192968483331, | |
| "grad_norm": 0.48029908537864685, | |
| "learning_rate": 3.852482488956992e-05, | |
| "loss": 2.0383, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.7145295255432927, | |
| "grad_norm": 0.4986077845096588, | |
| "learning_rate": 3.829754758384262e-05, | |
| "loss": 2.3006, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.7154397542382523, | |
| "grad_norm": 0.5001522302627563, | |
| "learning_rate": 3.807078378481059e-05, | |
| "loss": 2.3416, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.716349982933212, | |
| "grad_norm": 0.5049505829811096, | |
| "learning_rate": 3.784453537967161e-05, | |
| "loss": 2.1652, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 0.7172602116281716, | |
| "grad_norm": 0.5048404932022095, | |
| "learning_rate": 3.761880425133413e-05, | |
| "loss": 2.1345, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.7181704403231312, | |
| "grad_norm": 0.4869529604911804, | |
| "learning_rate": 3.7393592278401704e-05, | |
| "loss": 2.0906, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.7190806690180908, | |
| "grad_norm": 0.5454188585281372, | |
| "learning_rate": 3.7168901335157315e-05, | |
| "loss": 2.4214, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.7199908977130504, | |
| "grad_norm": 0.5238876938819885, | |
| "learning_rate": 3.694473329154778e-05, | |
| "loss": 1.9798, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 0.72090112640801, | |
| "grad_norm": 0.5545910596847534, | |
| "learning_rate": 3.672109001316809e-05, | |
| "loss": 2.4726, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.7218113551029697, | |
| "grad_norm": 0.542072594165802, | |
| "learning_rate": 3.649797336124615e-05, | |
| "loss": 2.016, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 0.7227215837979293, | |
| "grad_norm": 0.5355279445648193, | |
| "learning_rate": 3.6275385192627056e-05, | |
| "loss": 2.1041, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.7236318124928889, | |
| "grad_norm": 0.5673330426216125, | |
| "learning_rate": 3.6053327359757535e-05, | |
| "loss": 2.1006, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.7245420411878485, | |
| "grad_norm": 0.6170483231544495, | |
| "learning_rate": 3.583180171067101e-05, | |
| "loss": 2.3275, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.7254522698828081, | |
| "grad_norm": 0.6877503991127014, | |
| "learning_rate": 3.5610810088971625e-05, | |
| "loss": 2.4504, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 0.7263624985777677, | |
| "grad_norm": 0.7676892280578613, | |
| "learning_rate": 3.5390354333819344e-05, | |
| "loss": 2.6627, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.9272985458374023, | |
| "learning_rate": 3.517043627991441e-05, | |
| "loss": 2.5253, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 0.7281829559676869, | |
| "grad_norm": 1.9746160507202148, | |
| "learning_rate": 3.4951057757482205e-05, | |
| "loss": 2.5993, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.7290931846626465, | |
| "grad_norm": 0.4283389449119568, | |
| "learning_rate": 3.4732220592257946e-05, | |
| "loss": 2.5104, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.7300034133576061, | |
| "grad_norm": 0.43243661522865295, | |
| "learning_rate": 3.45139266054715e-05, | |
| "loss": 2.1707, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.7309136420525657, | |
| "grad_norm": 0.4262010455131531, | |
| "learning_rate": 3.429617761383222e-05, | |
| "loss": 2.2513, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 0.7318238707475253, | |
| "grad_norm": 0.4308786392211914, | |
| "learning_rate": 3.40789754295139e-05, | |
| "loss": 2.4042, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.7327340994424849, | |
| "grad_norm": 0.42787450551986694, | |
| "learning_rate": 3.3862321860139576e-05, | |
| "loss": 2.4259, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.7336443281374445, | |
| "grad_norm": 0.4381856620311737, | |
| "learning_rate": 3.364621870876659e-05, | |
| "loss": 2.4046, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.7345545568324041, | |
| "grad_norm": 0.4300837814807892, | |
| "learning_rate": 3.343066777387148e-05, | |
| "loss": 2.3713, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.7354647855273637, | |
| "grad_norm": 0.4228179156780243, | |
| "learning_rate": 3.3215670849335155e-05, | |
| "loss": 2.2606, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.7363750142223233, | |
| "grad_norm": 0.4267946779727936, | |
| "learning_rate": 3.300122972442773e-05, | |
| "loss": 2.3377, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 0.7372852429172829, | |
| "grad_norm": 0.41019585728645325, | |
| "learning_rate": 3.278734618379402e-05, | |
| "loss": 2.1903, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.7381954716122425, | |
| "grad_norm": 0.42949211597442627, | |
| "learning_rate": 3.257402200743821e-05, | |
| "loss": 2.3309, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 0.7391057003072021, | |
| "grad_norm": 0.4410031735897064, | |
| "learning_rate": 3.2361258970709397e-05, | |
| "loss": 2.3924, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.7400159290021618, | |
| "grad_norm": 0.4223659336566925, | |
| "learning_rate": 3.21490588442868e-05, | |
| "loss": 2.2614, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.7409261576971214, | |
| "grad_norm": 0.43348926305770874, | |
| "learning_rate": 3.19374233941647e-05, | |
| "loss": 2.4255, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.741836386392081, | |
| "grad_norm": 0.42184650897979736, | |
| "learning_rate": 3.172635438163816e-05, | |
| "loss": 2.2794, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.7427466150870407, | |
| "grad_norm": 0.4127393066883087, | |
| "learning_rate": 3.1515853563288076e-05, | |
| "loss": 2.1242, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.7436568437820003, | |
| "grad_norm": 0.40478646755218506, | |
| "learning_rate": 3.130592269096671e-05, | |
| "loss": 2.035, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 0.7445670724769599, | |
| "grad_norm": 0.41883495450019836, | |
| "learning_rate": 3.1096563511783014e-05, | |
| "loss": 2.1427, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.7454773011719195, | |
| "grad_norm": 0.39868757128715515, | |
| "learning_rate": 3.08877777680882e-05, | |
| "loss": 2.0169, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.7463875298668791, | |
| "grad_norm": 0.392415851354599, | |
| "learning_rate": 3.0679567197461134e-05, | |
| "loss": 2.0969, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.7472977585618387, | |
| "grad_norm": 0.4181436598300934, | |
| "learning_rate": 3.047193353269382e-05, | |
| "loss": 2.1766, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 0.7482079872567983, | |
| "grad_norm": 0.41824692487716675, | |
| "learning_rate": 3.0264878501777306e-05, | |
| "loss": 2.0897, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.7491182159517579, | |
| "grad_norm": 0.442640095949173, | |
| "learning_rate": 3.005840382788685e-05, | |
| "loss": 2.0851, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 0.7500284446467175, | |
| "grad_norm": 0.43662169575691223, | |
| "learning_rate": 2.9852511229367865e-05, | |
| "loss": 2.1546, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.7509386733416771, | |
| "grad_norm": 0.44712746143341064, | |
| "learning_rate": 2.9647202419721687e-05, | |
| "loss": 2.2304, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.7509386733416771, | |
| "eval_loss": 2.2184386253356934, | |
| "eval_runtime": 205.4094, | |
| "eval_samples_per_second": 9.011, | |
| "eval_steps_per_second": 4.508, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.7518489020366367, | |
| "grad_norm": 0.4428364336490631, | |
| "learning_rate": 2.944247910759097e-05, | |
| "loss": 2.1631, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.7527591307315963, | |
| "grad_norm": 0.43727412819862366, | |
| "learning_rate": 2.9238342996745817e-05, | |
| "loss": 2.1495, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 0.7536693594265559, | |
| "grad_norm": 0.45698437094688416, | |
| "learning_rate": 2.9034795786069436e-05, | |
| "loss": 2.1497, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.7545795881215155, | |
| "grad_norm": 0.423408567905426, | |
| "learning_rate": 2.8831839169543996e-05, | |
| "loss": 1.9607, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 0.7554898168164751, | |
| "grad_norm": 0.4666843116283417, | |
| "learning_rate": 2.862947483623659e-05, | |
| "loss": 2.1271, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.7564000455114347, | |
| "grad_norm": 0.46026623249053955, | |
| "learning_rate": 2.8427704470285144e-05, | |
| "loss": 2.1943, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.7573102742063944, | |
| "grad_norm": 0.459824800491333, | |
| "learning_rate": 2.8226529750884402e-05, | |
| "loss": 2.0793, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.758220502901354, | |
| "grad_norm": 0.4680033326148987, | |
| "learning_rate": 2.8025952352271958e-05, | |
| "loss": 2.1652, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 0.7591307315963136, | |
| "grad_norm": 0.47054019570350647, | |
| "learning_rate": 2.7825973943714335e-05, | |
| "loss": 2.1526, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.7600409602912732, | |
| "grad_norm": 0.465668648481369, | |
| "learning_rate": 2.7626596189492983e-05, | |
| "loss": 2.0476, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.7609511889862328, | |
| "grad_norm": 0.47716715931892395, | |
| "learning_rate": 2.7427820748890685e-05, | |
| "loss": 2.1511, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 0.7618614176811924, | |
| "grad_norm": 0.468532532453537, | |
| "learning_rate": 2.7229649276177503e-05, | |
| "loss": 2.1065, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.762771646376152, | |
| "grad_norm": 0.4883919358253479, | |
| "learning_rate": 2.7032083420597e-05, | |
| "loss": 2.2111, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 0.7636818750711116, | |
| "grad_norm": 0.5037781000137329, | |
| "learning_rate": 2.683512482635281e-05, | |
| "loss": 2.1824, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 0.7645921037660712, | |
| "grad_norm": 0.5097115635871887, | |
| "learning_rate": 2.6638775132594553e-05, | |
| "loss": 2.2818, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.7655023324610308, | |
| "grad_norm": 0.5003491640090942, | |
| "learning_rate": 2.6443035973404496e-05, | |
| "loss": 2.112, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 0.7664125611559904, | |
| "grad_norm": 0.5198303461074829, | |
| "learning_rate": 2.624790897778391e-05, | |
| "loss": 2.1864, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 0.76732278985095, | |
| "grad_norm": 0.5522372722625732, | |
| "learning_rate": 2.605339576963929e-05, | |
| "loss": 2.3857, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 0.7682330185459096, | |
| "grad_norm": 0.5393935441970825, | |
| "learning_rate": 2.585949796776912e-05, | |
| "loss": 2.2549, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 0.7691432472408692, | |
| "grad_norm": 0.573794424533844, | |
| "learning_rate": 2.5666217185850262e-05, | |
| "loss": 2.3236, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.7700534759358288, | |
| "grad_norm": 0.5689824819564819, | |
| "learning_rate": 2.5473555032424533e-05, | |
| "loss": 2.1463, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 0.7709637046307884, | |
| "grad_norm": 0.6545232534408569, | |
| "learning_rate": 2.528151311088537e-05, | |
| "loss": 2.3964, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 0.771873933325748, | |
| "grad_norm": 0.7613667845726013, | |
| "learning_rate": 2.50900930194644e-05, | |
| "loss": 2.7365, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 0.7727841620207077, | |
| "grad_norm": 0.8592699766159058, | |
| "learning_rate": 2.4899296351218227e-05, | |
| "loss": 2.3281, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 0.7736943907156673, | |
| "grad_norm": 1.6767549514770508, | |
| "learning_rate": 2.4709124694015116e-05, | |
| "loss": 2.4329, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.774604619410627, | |
| "grad_norm": 0.42721498012542725, | |
| "learning_rate": 2.451957963052185e-05, | |
| "loss": 2.4287, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 0.7755148481055866, | |
| "grad_norm": 0.4105132818222046, | |
| "learning_rate": 2.433066273819037e-05, | |
| "loss": 2.3069, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 0.7764250768005462, | |
| "grad_norm": 0.42730897665023804, | |
| "learning_rate": 2.4142375589244957e-05, | |
| "loss": 2.3786, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 0.7773353054955058, | |
| "grad_norm": 0.41130363941192627, | |
| "learning_rate": 2.3954719750668907e-05, | |
| "loss": 2.1378, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 0.7782455341904654, | |
| "grad_norm": 0.41424882411956787, | |
| "learning_rate": 2.3767696784191463e-05, | |
| "loss": 2.2526, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.779155762885425, | |
| "grad_norm": 0.42201122641563416, | |
| "learning_rate": 2.3581308246275103e-05, | |
| "loss": 2.3379, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 0.7800659915803846, | |
| "grad_norm": 0.4150107204914093, | |
| "learning_rate": 2.339555568810221e-05, | |
| "loss": 2.1722, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 0.7809762202753442, | |
| "grad_norm": 0.432271271944046, | |
| "learning_rate": 2.321044065556246e-05, | |
| "loss": 2.4875, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.7818864489703038, | |
| "grad_norm": 0.4255993068218231, | |
| "learning_rate": 2.302596468923981e-05, | |
| "loss": 2.3043, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 0.7827966776652634, | |
| "grad_norm": 0.4193149507045746, | |
| "learning_rate": 2.284212932439972e-05, | |
| "loss": 2.3238, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.783706906360223, | |
| "grad_norm": 0.3979727625846863, | |
| "learning_rate": 2.265893609097637e-05, | |
| "loss": 2.0908, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 0.7846171350551826, | |
| "grad_norm": 0.42481502890586853, | |
| "learning_rate": 2.247638651355991e-05, | |
| "loss": 2.3404, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 0.7855273637501422, | |
| "grad_norm": 0.41932496428489685, | |
| "learning_rate": 2.229448211138382e-05, | |
| "loss": 2.3529, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 0.7864375924451018, | |
| "grad_norm": 0.4188045263290405, | |
| "learning_rate": 2.211322439831218e-05, | |
| "loss": 2.1973, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.7873478211400614, | |
| "grad_norm": 0.41575223207473755, | |
| "learning_rate": 2.1932614882827197e-05, | |
| "loss": 2.2664, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.788258049835021, | |
| "grad_norm": 0.41103559732437134, | |
| "learning_rate": 2.1752655068016515e-05, | |
| "loss": 2.1176, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 0.7891682785299806, | |
| "grad_norm": 0.3994426727294922, | |
| "learning_rate": 2.1573346451560794e-05, | |
| "loss": 2.0824, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 0.7900785072249402, | |
| "grad_norm": 0.40480148792266846, | |
| "learning_rate": 2.139469052572127e-05, | |
| "loss": 1.9797, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 0.7909887359198998, | |
| "grad_norm": 0.4224672317504883, | |
| "learning_rate": 2.1216688777327154e-05, | |
| "loss": 2.0783, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 0.7918989646148595, | |
| "grad_norm": 0.4260886013507843, | |
| "learning_rate": 2.1039342687763586e-05, | |
| "loss": 2.203, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.7928091933098191, | |
| "grad_norm": 0.41183462738990784, | |
| "learning_rate": 2.0862653732958915e-05, | |
| "loss": 1.9724, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 0.7937194220047787, | |
| "grad_norm": 0.43447592854499817, | |
| "learning_rate": 2.0686623383372715e-05, | |
| "loss": 2.1632, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 0.7946296506997383, | |
| "grad_norm": 0.4297522008419037, | |
| "learning_rate": 2.051125310398353e-05, | |
| "loss": 2.0486, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 0.795539879394698, | |
| "grad_norm": 0.45072224736213684, | |
| "learning_rate": 2.03365443542764e-05, | |
| "loss": 2.1973, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 0.7964501080896575, | |
| "grad_norm": 0.4462050199508667, | |
| "learning_rate": 2.016249858823106e-05, | |
| "loss": 2.0274, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.7973603367846172, | |
| "grad_norm": 0.4606810212135315, | |
| "learning_rate": 1.998911725430963e-05, | |
| "loss": 2.1616, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 0.7982705654795768, | |
| "grad_norm": 0.44487303495407104, | |
| "learning_rate": 1.981640179544466e-05, | |
| "loss": 2.323, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 0.7991807941745364, | |
| "grad_norm": 0.45202627778053284, | |
| "learning_rate": 1.964435364902705e-05, | |
| "loss": 2.1361, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 0.800091022869496, | |
| "grad_norm": 0.44588690996170044, | |
| "learning_rate": 1.947297424689414e-05, | |
| "loss": 2.1173, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 0.8010012515644556, | |
| "grad_norm": 0.46819573640823364, | |
| "learning_rate": 1.93022650153178e-05, | |
| "loss": 2.0187, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.8019114802594152, | |
| "grad_norm": 0.44944408535957336, | |
| "learning_rate": 1.913222737499243e-05, | |
| "loss": 2.0103, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 0.8028217089543748, | |
| "grad_norm": 0.44194296002388, | |
| "learning_rate": 1.8962862741023423e-05, | |
| "loss": 1.9489, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 0.8037319376493344, | |
| "grad_norm": 0.4707835614681244, | |
| "learning_rate": 1.879417252291502e-05, | |
| "loss": 2.1982, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 0.804642166344294, | |
| "grad_norm": 0.4707585573196411, | |
| "learning_rate": 1.8626158124558858e-05, | |
| "loss": 2.1049, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 0.8055523950392536, | |
| "grad_norm": 0.4964425265789032, | |
| "learning_rate": 1.8458820944222255e-05, | |
| "loss": 2.2127, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.8064626237342132, | |
| "grad_norm": 0.4742617607116699, | |
| "learning_rate": 1.829216237453637e-05, | |
| "loss": 2.1019, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 0.8073728524291728, | |
| "grad_norm": 0.49655184149742126, | |
| "learning_rate": 1.8126183802484865e-05, | |
| "loss": 2.2403, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 0.8082830811241324, | |
| "grad_norm": 0.4954749643802643, | |
| "learning_rate": 1.7960886609392214e-05, | |
| "loss": 2.0321, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 0.8091933098190921, | |
| "grad_norm": 0.4694468379020691, | |
| "learning_rate": 1.7796272170912253e-05, | |
| "loss": 1.817, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 0.8101035385140517, | |
| "grad_norm": 0.5026715397834778, | |
| "learning_rate": 1.763234185701673e-05, | |
| "loss": 2.2038, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.8110137672090113, | |
| "grad_norm": 0.5050073862075806, | |
| "learning_rate": 1.7469097031983893e-05, | |
| "loss": 2.0861, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 0.8119239959039709, | |
| "grad_norm": 0.5078185796737671, | |
| "learning_rate": 1.730653905438714e-05, | |
| "loss": 2.0672, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 0.8128342245989305, | |
| "grad_norm": 0.525215744972229, | |
| "learning_rate": 1.7144669277083712e-05, | |
| "loss": 2.1502, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 0.8137444532938901, | |
| "grad_norm": 0.5429519414901733, | |
| "learning_rate": 1.6983489047203483e-05, | |
| "loss": 2.0935, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 0.8146546819888497, | |
| "grad_norm": 0.5544317960739136, | |
| "learning_rate": 1.6822999706137567e-05, | |
| "loss": 2.0943, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.8155649106838093, | |
| "grad_norm": 0.6273201107978821, | |
| "learning_rate": 1.6663202589527473e-05, | |
| "loss": 2.3608, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 0.8164751393787689, | |
| "grad_norm": 0.7101454734802246, | |
| "learning_rate": 1.6504099027253706e-05, | |
| "loss": 2.4168, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 0.8173853680737285, | |
| "grad_norm": 0.7550842761993408, | |
| "learning_rate": 1.634569034342476e-05, | |
| "loss": 2.5798, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 0.8182955967686881, | |
| "grad_norm": 0.8533863425254822, | |
| "learning_rate": 1.6187977856366253e-05, | |
| "loss": 2.5575, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 0.8192058254636477, | |
| "grad_norm": 1.34774911403656, | |
| "learning_rate": 1.6030962878609725e-05, | |
| "loss": 2.4134, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.8201160541586073, | |
| "grad_norm": 0.4584032893180847, | |
| "learning_rate": 1.587464671688187e-05, | |
| "loss": 2.4781, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 0.8210262828535669, | |
| "grad_norm": 0.43342748284339905, | |
| "learning_rate": 1.5719030672093717e-05, | |
| "loss": 2.3685, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 0.8219365115485265, | |
| "grad_norm": 0.4225307106971741, | |
| "learning_rate": 1.5564116039329545e-05, | |
| "loss": 2.2022, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 0.8228467402434861, | |
| "grad_norm": 0.43026039004325867, | |
| "learning_rate": 1.5409904107836358e-05, | |
| "loss": 2.2817, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 0.8237569689384457, | |
| "grad_norm": 0.4114493131637573, | |
| "learning_rate": 1.5256396161013075e-05, | |
| "loss": 2.3298, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.8246671976334053, | |
| "grad_norm": 0.42313718795776367, | |
| "learning_rate": 1.5103593476399791e-05, | |
| "loss": 2.3211, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 0.825577426328365, | |
| "grad_norm": 0.4246841371059418, | |
| "learning_rate": 1.495149732566723e-05, | |
| "loss": 2.2385, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 0.8264876550233247, | |
| "grad_norm": 0.4131985008716583, | |
| "learning_rate": 1.4800108974606119e-05, | |
| "loss": 2.2873, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 0.8273978837182843, | |
| "grad_norm": 0.42265599966049194, | |
| "learning_rate": 1.4649429683116644e-05, | |
| "loss": 2.1486, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 0.8283081124132439, | |
| "grad_norm": 0.4338424801826477, | |
| "learning_rate": 1.4499460705197998e-05, | |
| "loss": 2.2365, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.8292183411082035, | |
| "grad_norm": 0.4278540015220642, | |
| "learning_rate": 1.4350203288937936e-05, | |
| "loss": 2.36, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 0.8301285698031631, | |
| "grad_norm": 0.41379448771476746, | |
| "learning_rate": 1.4201658676502294e-05, | |
| "loss": 2.184, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 0.8310387984981227, | |
| "grad_norm": 0.42351198196411133, | |
| "learning_rate": 1.4053828104124867e-05, | |
| "loss": 2.2505, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 0.8319490271930823, | |
| "grad_norm": 0.40783679485321045, | |
| "learning_rate": 1.3906712802096933e-05, | |
| "loss": 2.0255, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 0.8328592558880419, | |
| "grad_norm": 0.4174416661262512, | |
| "learning_rate": 1.3760313994757001e-05, | |
| "loss": 2.2376, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.8337694845830015, | |
| "grad_norm": 0.41884645819664, | |
| "learning_rate": 1.361463290048085e-05, | |
| "loss": 2.0206, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 0.8346797132779611, | |
| "grad_norm": 0.399498850107193, | |
| "learning_rate": 1.3469670731671046e-05, | |
| "loss": 2.063, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 0.8355899419729207, | |
| "grad_norm": 0.40431055426597595, | |
| "learning_rate": 1.3325428694747177e-05, | |
| "loss": 2.0053, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 0.8365001706678803, | |
| "grad_norm": 0.40479356050491333, | |
| "learning_rate": 1.3181907990135622e-05, | |
| "loss": 2.0693, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 0.8374103993628399, | |
| "grad_norm": 0.4056653678417206, | |
| "learning_rate": 1.3039109812259598e-05, | |
| "loss": 2.0361, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.8383206280577995, | |
| "grad_norm": 0.4257088005542755, | |
| "learning_rate": 1.2897035349529263e-05, | |
| "loss": 2.0589, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 0.8392308567527591, | |
| "grad_norm": 0.43024080991744995, | |
| "learning_rate": 1.2755685784331783e-05, | |
| "loss": 2.0419, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 0.8401410854477187, | |
| "grad_norm": 0.42889195680618286, | |
| "learning_rate": 1.2615062293021507e-05, | |
| "loss": 2.0515, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 0.8410513141426783, | |
| "grad_norm": 0.4491952061653137, | |
| "learning_rate": 1.2475166045910159e-05, | |
| "loss": 2.2535, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 0.8419615428376379, | |
| "grad_norm": 0.43797358870506287, | |
| "learning_rate": 1.2335998207257137e-05, | |
| "loss": 2.1338, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.8428717715325975, | |
| "grad_norm": 0.4491622745990753, | |
| "learning_rate": 1.2197559935259795e-05, | |
| "loss": 2.2059, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 0.8437820002275572, | |
| "grad_norm": 0.43628188967704773, | |
| "learning_rate": 1.20598523820438e-05, | |
| "loss": 1.8784, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 0.8446922289225168, | |
| "grad_norm": 0.45739004015922546, | |
| "learning_rate": 1.1922876693653585e-05, | |
| "loss": 2.0433, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 0.8456024576174764, | |
| "grad_norm": 0.44873446226119995, | |
| "learning_rate": 1.1786634010042719e-05, | |
| "loss": 1.9578, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 0.846512686312436, | |
| "grad_norm": 0.43957433104515076, | |
| "learning_rate": 1.1651125465064516e-05, | |
| "loss": 2.0078, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.8474229150073956, | |
| "grad_norm": 0.4639342129230499, | |
| "learning_rate": 1.1516352186462586e-05, | |
| "loss": 2.0714, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.8483331437023552, | |
| "grad_norm": 0.44638022780418396, | |
| "learning_rate": 1.13823152958614e-05, | |
| "loss": 1.8991, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 0.8492433723973148, | |
| "grad_norm": 0.4596819579601288, | |
| "learning_rate": 1.1249015908756998e-05, | |
| "loss": 1.9595, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 0.8501536010922744, | |
| "grad_norm": 0.47656434774398804, | |
| "learning_rate": 1.1116455134507664e-05, | |
| "loss": 2.0788, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 0.851063829787234, | |
| "grad_norm": 0.4645254611968994, | |
| "learning_rate": 1.098463407632474e-05, | |
| "loss": 2.0703, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.8519740584821937, | |
| "grad_norm": 0.4659541845321655, | |
| "learning_rate": 1.0853553831263418e-05, | |
| "loss": 2.0804, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 0.8528842871771533, | |
| "grad_norm": 0.4771886467933655, | |
| "learning_rate": 1.0723215490213634e-05, | |
| "loss": 2.1124, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 0.8537945158721129, | |
| "grad_norm": 0.49211612343788147, | |
| "learning_rate": 1.0593620137890948e-05, | |
| "loss": 2.2221, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 0.8547047445670725, | |
| "grad_norm": 0.5174618363380432, | |
| "learning_rate": 1.0464768852827545e-05, | |
| "loss": 2.1684, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 0.8556149732620321, | |
| "grad_norm": 0.5098733305931091, | |
| "learning_rate": 1.0336662707363287e-05, | |
| "loss": 2.103, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.8565252019569917, | |
| "grad_norm": 0.5197715163230896, | |
| "learning_rate": 1.0209302767636664e-05, | |
| "loss": 2.2107, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 0.8574354306519513, | |
| "grad_norm": 0.547512412071228, | |
| "learning_rate": 1.0082690093576163e-05, | |
| "loss": 2.2448, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 0.8583456593469109, | |
| "grad_norm": 0.5418568849563599, | |
| "learning_rate": 9.95682573889114e-06, | |
| "loss": 2.2423, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 0.8592558880418705, | |
| "grad_norm": 0.5369839072227478, | |
| "learning_rate": 9.831710751063283e-06, | |
| "loss": 1.9788, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 0.8601661167368301, | |
| "grad_norm": 0.573844313621521, | |
| "learning_rate": 9.707346171337894e-06, | |
| "loss": 2.2906, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.8610763454317898, | |
| "grad_norm": 0.6142247915267944, | |
| "learning_rate": 9.583733034714981e-06, | |
| "loss": 2.3744, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 0.8619865741267494, | |
| "grad_norm": 0.6646602153778076, | |
| "learning_rate": 9.460872369940955e-06, | |
| "loss": 2.4641, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 0.862896802821709, | |
| "grad_norm": 0.727783739566803, | |
| "learning_rate": 9.338765199499854e-06, | |
| "loss": 2.4612, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 0.8638070315166686, | |
| "grad_norm": 0.851578950881958, | |
| "learning_rate": 9.217412539604942e-06, | |
| "loss": 2.6441, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 0.8647172602116282, | |
| "grad_norm": 1.461125373840332, | |
| "learning_rate": 9.096815400190172e-06, | |
| "loss": 2.4248, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.8656274889065878, | |
| "grad_norm": 0.4408392608165741, | |
| "learning_rate": 8.97697478490188e-06, | |
| "loss": 2.5431, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 0.8665377176015474, | |
| "grad_norm": 0.4102359414100647, | |
| "learning_rate": 8.857891691090337e-06, | |
| "loss": 2.3448, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 0.867447946296507, | |
| "grad_norm": 0.4374777674674988, | |
| "learning_rate": 8.739567109801494e-06, | |
| "loss": 2.3647, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 0.8683581749914666, | |
| "grad_norm": 0.40009114146232605, | |
| "learning_rate": 8.62200202576875e-06, | |
| "loss": 2.2401, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 0.8692684036864262, | |
| "grad_norm": 0.42013484239578247, | |
| "learning_rate": 8.505197417404687e-06, | |
| "loss": 2.1772, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.8701786323813858, | |
| "grad_norm": 0.43588119745254517, | |
| "learning_rate": 8.38915425679304e-06, | |
| "loss": 2.4605, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 0.8710888610763454, | |
| "grad_norm": 0.4295041561126709, | |
| "learning_rate": 8.273873509680519e-06, | |
| "loss": 2.4302, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 0.871999089771305, | |
| "grad_norm": 0.4430733621120453, | |
| "learning_rate": 8.15935613546872e-06, | |
| "loss": 2.3014, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 0.8729093184662646, | |
| "grad_norm": 0.4275224804878235, | |
| "learning_rate": 8.045603087206388e-06, | |
| "loss": 2.251, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 0.8738195471612242, | |
| "grad_norm": 0.4218734800815582, | |
| "learning_rate": 7.932615311581126e-06, | |
| "loss": 2.2841, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.8747297758561838, | |
| "grad_norm": 0.4275785982608795, | |
| "learning_rate": 7.820393748911791e-06, | |
| "loss": 2.2751, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 0.8756400045511434, | |
| "grad_norm": 0.40714067220687866, | |
| "learning_rate": 7.708939333140642e-06, | |
| "loss": 2.2023, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 0.876550233246103, | |
| "grad_norm": 0.4284750521183014, | |
| "learning_rate": 7.598252991825372e-06, | |
| "loss": 2.1991, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 0.8774604619410626, | |
| "grad_norm": 0.40348193049430847, | |
| "learning_rate": 7.488335646131628e-06, | |
| "loss": 2.1214, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 0.8783706906360224, | |
| "grad_norm": 0.4067203998565674, | |
| "learning_rate": 7.3791882108251945e-06, | |
| "loss": 2.0977, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.879280919330982, | |
| "grad_norm": 0.40969371795654297, | |
| "learning_rate": 7.270811594264437e-06, | |
| "loss": 2.1751, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 0.8801911480259416, | |
| "grad_norm": 0.39071908593177795, | |
| "learning_rate": 7.163206698392744e-06, | |
| "loss": 2.0464, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 0.8811013767209012, | |
| "grad_norm": 0.4038424789905548, | |
| "learning_rate": 7.056374418730971e-06, | |
| "loss": 2.1137, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 0.8820116054158608, | |
| "grad_norm": 0.38801443576812744, | |
| "learning_rate": 6.950315644370075e-06, | |
| "loss": 1.883, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 0.8829218341108204, | |
| "grad_norm": 0.3895006477832794, | |
| "learning_rate": 6.845031257963619e-06, | |
| "loss": 2.0169, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.88383206280578, | |
| "grad_norm": 0.413171648979187, | |
| "learning_rate": 6.740522135720517e-06, | |
| "loss": 2.2054, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 0.8847422915007396, | |
| "grad_norm": 0.4204188585281372, | |
| "learning_rate": 6.636789147397637e-06, | |
| "loss": 2.1765, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 0.8856525201956992, | |
| "grad_norm": 0.4209098517894745, | |
| "learning_rate": 6.533833156292679e-06, | |
| "loss": 1.9617, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 0.8865627488906588, | |
| "grad_norm": 0.4256611168384552, | |
| "learning_rate": 6.431655019236948e-06, | |
| "loss": 2.108, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 0.8874729775856184, | |
| "grad_norm": 0.43669816851615906, | |
| "learning_rate": 6.3302555865880965e-06, | |
| "loss": 2.0991, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.888383206280578, | |
| "grad_norm": 0.44833648204803467, | |
| "learning_rate": 6.229635702223324e-06, | |
| "loss": 2.2335, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 0.8892934349755376, | |
| "grad_norm": 0.45070621371269226, | |
| "learning_rate": 6.129796203532057e-06, | |
| "loss": 2.2487, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 0.8902036636704972, | |
| "grad_norm": 0.4609052538871765, | |
| "learning_rate": 6.030737921409169e-06, | |
| "loss": 2.1147, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 0.8911138923654568, | |
| "grad_norm": 0.4470416307449341, | |
| "learning_rate": 5.932461680248014e-06, | |
| "loss": 2.0615, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 0.8920241210604164, | |
| "grad_norm": 0.42171233892440796, | |
| "learning_rate": 5.834968297933541e-06, | |
| "loss": 2.0669, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.892934349755376, | |
| "grad_norm": 0.4385877251625061, | |
| "learning_rate": 5.738258585835532e-06, | |
| "loss": 1.9846, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 0.8938445784503356, | |
| "grad_norm": 0.4574371576309204, | |
| "learning_rate": 5.6423333488018095e-06, | |
| "loss": 2.114, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 0.8947548071452952, | |
| "grad_norm": 0.46896499395370483, | |
| "learning_rate": 5.547193385151561e-06, | |
| "loss": 2.0444, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 0.8956650358402549, | |
| "grad_norm": 0.45737412571907043, | |
| "learning_rate": 5.45283948666866e-06, | |
| "loss": 2.0976, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 0.8965752645352145, | |
| "grad_norm": 0.47739726305007935, | |
| "learning_rate": 5.359272438595153e-06, | |
| "loss": 2.1393, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.8974854932301741, | |
| "grad_norm": 0.47124338150024414, | |
| "learning_rate": 5.266493019624663e-06, | |
| "loss": 2.0509, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 0.8983957219251337, | |
| "grad_norm": 0.4660322070121765, | |
| "learning_rate": 5.1745020018958866e-06, | |
| "loss": 1.9704, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 0.8993059506200933, | |
| "grad_norm": 0.48330241441726685, | |
| "learning_rate": 5.083300150986259e-06, | |
| "loss": 2.021, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 0.9002161793150529, | |
| "grad_norm": 0.51470547914505, | |
| "learning_rate": 4.992888225905468e-06, | |
| "loss": 2.1097, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 0.9011264080100125, | |
| "grad_norm": 0.516373336315155, | |
| "learning_rate": 4.903266979089249e-06, | |
| "loss": 2.1694, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.9020366367049721, | |
| "grad_norm": 0.5256400108337402, | |
| "learning_rate": 4.8144371563930476e-06, | |
| "loss": 2.2843, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 0.9029468653999317, | |
| "grad_norm": 0.5559744238853455, | |
| "learning_rate": 4.726399497085832e-06, | |
| "loss": 2.2733, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 0.9038570940948913, | |
| "grad_norm": 0.5462202429771423, | |
| "learning_rate": 4.6391547338439536e-06, | |
| "loss": 2.1758, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 0.904767322789851, | |
| "grad_norm": 0.5769087672233582, | |
| "learning_rate": 4.552703592745033e-06, | |
| "loss": 2.2552, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 0.9056775514848106, | |
| "grad_norm": 0.5632253289222717, | |
| "learning_rate": 4.467046793261931e-06, | |
| "loss": 2.2402, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.9065877801797702, | |
| "grad_norm": 0.610163688659668, | |
| "learning_rate": 4.3821850482567595e-06, | |
| "loss": 2.4484, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 0.9074980088747298, | |
| "grad_norm": 0.6236492395401001, | |
| "learning_rate": 4.298119063974914e-06, | |
| "loss": 2.1914, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 0.9084082375696894, | |
| "grad_norm": 0.7362584471702576, | |
| "learning_rate": 4.214849540039267e-06, | |
| "loss": 2.5582, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 0.909318466264649, | |
| "grad_norm": 0.8780522346496582, | |
| "learning_rate": 4.132377169444279e-06, | |
| "loss": 2.5269, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 0.9102286949596086, | |
| "grad_norm": 1.4850975275039673, | |
| "learning_rate": 4.050702638550275e-06, | |
| "loss": 2.3857, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.9111389236545682, | |
| "grad_norm": 0.4350475072860718, | |
| "learning_rate": 3.969826627077655e-06, | |
| "loss": 2.5653, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 0.9120491523495278, | |
| "grad_norm": 0.44277626276016235, | |
| "learning_rate": 3.889749808101395e-06, | |
| "loss": 2.3969, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 0.9129593810444874, | |
| "grad_norm": 0.44005268812179565, | |
| "learning_rate": 3.810472848045266e-06, | |
| "loss": 2.6065, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 0.9138696097394471, | |
| "grad_norm": 0.41925248503685, | |
| "learning_rate": 3.7319964066763858e-06, | |
| "loss": 2.3878, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 0.9147798384344067, | |
| "grad_norm": 0.4320535957813263, | |
| "learning_rate": 3.6543211370997587e-06, | |
| "loss": 2.3829, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.9156900671293663, | |
| "grad_norm": 0.43817150592803955, | |
| "learning_rate": 3.5774476857527107e-06, | |
| "loss": 2.3854, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 0.9166002958243259, | |
| "grad_norm": 0.42210131883621216, | |
| "learning_rate": 3.5013766923996604e-06, | |
| "loss": 2.2874, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 0.9175105245192855, | |
| "grad_norm": 0.41610825061798096, | |
| "learning_rate": 3.426108790126681e-06, | |
| "loss": 2.3301, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 0.9184207532142451, | |
| "grad_norm": 0.42343541979789734, | |
| "learning_rate": 3.3516446053363015e-06, | |
| "loss": 2.2083, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 0.9193309819092047, | |
| "grad_norm": 0.4323045015335083, | |
| "learning_rate": 3.2779847577422697e-06, | |
| "loss": 2.2401, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.9202412106041643, | |
| "grad_norm": 0.4198078513145447, | |
| "learning_rate": 3.2051298603643753e-06, | |
| "loss": 2.0988, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 0.9211514392991239, | |
| "grad_norm": 0.4277539551258087, | |
| "learning_rate": 3.133080519523368e-06, | |
| "loss": 2.3482, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 0.9220616679940835, | |
| "grad_norm": 0.42749837040901184, | |
| "learning_rate": 3.0618373348359264e-06, | |
| "loss": 2.3242, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 0.9229718966890431, | |
| "grad_norm": 0.4157456159591675, | |
| "learning_rate": 2.991400899209651e-06, | |
| "loss": 2.11, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 0.9238821253840027, | |
| "grad_norm": 0.41514283418655396, | |
| "learning_rate": 2.921771798838069e-06, | |
| "loss": 2.0979, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.9247923540789623, | |
| "grad_norm": 0.4146190285682678, | |
| "learning_rate": 2.852950613195915e-06, | |
| "loss": 2.1057, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 0.9257025827739219, | |
| "grad_norm": 0.4031788110733032, | |
| "learning_rate": 2.784937915034169e-06, | |
| "loss": 2.1094, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 0.9266128114688815, | |
| "grad_norm": 0.4135347604751587, | |
| "learning_rate": 2.717734270375272e-06, | |
| "loss": 2.2154, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 0.9275230401638411, | |
| "grad_norm": 0.40153443813323975, | |
| "learning_rate": 2.6513402385085704e-06, | |
| "loss": 2.0342, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 0.9284332688588007, | |
| "grad_norm": 0.4011882543563843, | |
| "learning_rate": 2.585756371985493e-06, | |
| "loss": 1.9751, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.9293434975537603, | |
| "grad_norm": 0.4032374322414398, | |
| "learning_rate": 2.520983216615047e-06, | |
| "loss": 2.0868, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 0.9302537262487199, | |
| "grad_norm": 0.4052782952785492, | |
| "learning_rate": 2.4570213114592954e-06, | |
| "loss": 2.0716, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 0.9311639549436797, | |
| "grad_norm": 0.4261015057563782, | |
| "learning_rate": 2.393871188828767e-06, | |
| "loss": 2.1153, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 0.9320741836386393, | |
| "grad_norm": 0.4141393303871155, | |
| "learning_rate": 2.3315333742780942e-06, | |
| "loss": 2.0839, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 0.9329844123335989, | |
| "grad_norm": 0.41185298562049866, | |
| "learning_rate": 2.270008386601685e-06, | |
| "loss": 2.0686, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.9338946410285585, | |
| "grad_norm": 0.41932380199432373, | |
| "learning_rate": 2.2092967378292915e-06, | |
| "loss": 2.0688, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 0.9348048697235181, | |
| "grad_norm": 0.430480033159256, | |
| "learning_rate": 2.1493989332218468e-06, | |
| "loss": 2.1202, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 0.9357150984184777, | |
| "grad_norm": 0.4182969629764557, | |
| "learning_rate": 2.0903154712672237e-06, | |
| "loss": 1.8457, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 0.9366253271134373, | |
| "grad_norm": 0.44270989298820496, | |
| "learning_rate": 2.032046843676061e-06, | |
| "loss": 2.2296, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 0.9375355558083969, | |
| "grad_norm": 0.4312398433685303, | |
| "learning_rate": 1.974593535377722e-06, | |
| "loss": 1.9802, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.9384457845033565, | |
| "grad_norm": 0.4557861387729645, | |
| "learning_rate": 1.917956024516243e-06, | |
| "loss": 1.9306, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 0.9393560131983161, | |
| "grad_norm": 0.4993920624256134, | |
| "learning_rate": 1.8621347824462787e-06, | |
| "loss": 2.275, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 0.9402662418932757, | |
| "grad_norm": 0.4622466266155243, | |
| "learning_rate": 1.8071302737293295e-06, | |
| "loss": 2.1966, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 0.4710349142551422, | |
| "learning_rate": 1.752942956129744e-06, | |
| "loss": 2.1353, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 0.9420866992831949, | |
| "grad_norm": 0.4602985680103302, | |
| "learning_rate": 1.6995732806109554e-06, | |
| "loss": 2.0855, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.9429969279781545, | |
| "grad_norm": 0.4736422002315521, | |
| "learning_rate": 1.6470216913317626e-06, | |
| "loss": 2.081, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 0.9439071566731141, | |
| "grad_norm": 0.4875909984111786, | |
| "learning_rate": 1.5952886256425547e-06, | |
| "loss": 2.1607, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 0.9448173853680737, | |
| "grad_norm": 0.5093516111373901, | |
| "learning_rate": 1.5443745140817366e-06, | |
| "loss": 2.2025, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 0.9457276140630333, | |
| "grad_norm": 0.5034651160240173, | |
| "learning_rate": 1.4942797803721543e-06, | |
| "loss": 1.9985, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 0.9466378427579929, | |
| "grad_norm": 0.5074111819267273, | |
| "learning_rate": 1.4450048414174854e-06, | |
| "loss": 2.1175, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.9475480714529525, | |
| "grad_norm": 0.5183742046356201, | |
| "learning_rate": 1.3965501072988663e-06, | |
| "loss": 2.0718, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 0.9484583001479122, | |
| "grad_norm": 0.5284718871116638, | |
| "learning_rate": 1.348915981271437e-06, | |
| "loss": 2.1586, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 0.9493685288428718, | |
| "grad_norm": 0.545464813709259, | |
| "learning_rate": 1.3021028597609675e-06, | |
| "loss": 2.2445, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 0.9502787575378314, | |
| "grad_norm": 0.5713001489639282, | |
| "learning_rate": 1.2561111323605712e-06, | |
| "loss": 2.1888, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 0.951188986232791, | |
| "grad_norm": 0.5774447321891785, | |
| "learning_rate": 1.2109411818274852e-06, | |
| "loss": 2.2029, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.9520992149277506, | |
| "grad_norm": 0.6209971308708191, | |
| "learning_rate": 1.1665933840798838e-06, | |
| "loss": 2.1735, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 0.9530094436227102, | |
| "grad_norm": 0.6675162315368652, | |
| "learning_rate": 1.1230681081936923e-06, | |
| "loss": 2.4231, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 0.9539196723176698, | |
| "grad_norm": 0.7409051060676575, | |
| "learning_rate": 1.0803657163995895e-06, | |
| "loss": 2.45, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 0.9548299010126294, | |
| "grad_norm": 0.8856377601623535, | |
| "learning_rate": 1.0384865640799435e-06, | |
| "loss": 2.4769, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 0.955740129707589, | |
| "grad_norm": 1.4897712469100952, | |
| "learning_rate": 9.974309997658915e-07, | |
| "loss": 2.6822, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.9566503584025486, | |
| "grad_norm": 0.4162799119949341, | |
| "learning_rate": 9.57199365134387e-07, | |
| "loss": 2.493, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 0.9575605870975082, | |
| "grad_norm": 0.4336461126804352, | |
| "learning_rate": 9.177919950054237e-07, | |
| "loss": 2.4071, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 0.9584708157924678, | |
| "grad_norm": 0.42869681119918823, | |
| "learning_rate": 8.792092173391831e-07, | |
| "loss": 2.3585, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 0.9593810444874274, | |
| "grad_norm": 0.4060511887073517, | |
| "learning_rate": 8.41451353233369e-07, | |
| "loss": 2.3244, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 0.960291273182387, | |
| "grad_norm": 0.42351028323173523, | |
| "learning_rate": 8.04518716920466e-07, | |
| "loss": 2.4194, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.9612015018773467, | |
| "grad_norm": 0.42555585503578186, | |
| "learning_rate": 7.684116157651966e-07, | |
| "loss": 2.368, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 0.9621117305723063, | |
| "grad_norm": 0.4169003665447235, | |
| "learning_rate": 7.331303502618903e-07, | |
| "loss": 2.3947, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 0.9630219592672659, | |
| "grad_norm": 0.43154704570770264, | |
| "learning_rate": 6.986752140320518e-07, | |
| "loss": 2.2809, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 0.9639321879622255, | |
| "grad_norm": 0.40992870926856995, | |
| "learning_rate": 6.650464938218637e-07, | |
| "loss": 2.2827, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 0.9648424166571851, | |
| "grad_norm": 0.4168105125427246, | |
| "learning_rate": 6.322444694998319e-07, | |
| "loss": 2.3334, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.9657526453521448, | |
| "grad_norm": 0.43256238102912903, | |
| "learning_rate": 6.002694140544329e-07, | |
| "loss": 2.2669, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 0.9666628740471044, | |
| "grad_norm": 0.42290446162223816, | |
| "learning_rate": 5.691215935918815e-07, | |
| "loss": 2.1027, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 0.967573102742064, | |
| "grad_norm": 0.4109112024307251, | |
| "learning_rate": 5.388012673338661e-07, | |
| "loss": 2.15, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 0.9684833314370236, | |
| "grad_norm": 0.4292824864387512, | |
| "learning_rate": 5.093086876154174e-07, | |
| "loss": 2.2394, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 0.9693935601319832, | |
| "grad_norm": 0.4136911928653717, | |
| "learning_rate": 4.80644099882821e-07, | |
| "loss": 2.1904, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.9703037888269428, | |
| "grad_norm": 0.4183078110218048, | |
| "learning_rate": 4.5280774269154115e-07, | |
| "loss": 2.1499, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 0.9712140175219024, | |
| "grad_norm": 0.41571420431137085, | |
| "learning_rate": 4.2579984770426686e-07, | |
| "loss": 2.0975, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 0.972124246216862, | |
| "grad_norm": 0.39653652906417847, | |
| "learning_rate": 3.99620639688969e-07, | |
| "loss": 1.9686, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 0.9730344749118216, | |
| "grad_norm": 0.4133754372596741, | |
| "learning_rate": 3.742703365170241e-07, | |
| "loss": 2.0801, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 0.9739447036067812, | |
| "grad_norm": 0.41434839367866516, | |
| "learning_rate": 3.497491491614158e-07, | |
| "loss": 1.9311, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.9748549323017408, | |
| "grad_norm": 0.41751164197921753, | |
| "learning_rate": 3.260572816949692e-07, | |
| "loss": 2.1841, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 0.9757651609967004, | |
| "grad_norm": 0.41821038722991943, | |
| "learning_rate": 3.0319493128866396e-07, | |
| "loss": 2.0719, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 0.97667538969166, | |
| "grad_norm": 0.4422387480735779, | |
| "learning_rate": 2.8116228820997957e-07, | |
| "loss": 2.2655, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 0.9775856183866196, | |
| "grad_norm": 0.4252420961856842, | |
| "learning_rate": 2.5995953582130804e-07, | |
| "loss": 2.098, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 0.9784958470815792, | |
| "grad_norm": 0.4352913498878479, | |
| "learning_rate": 2.395868505784438e-07, | |
| "loss": 2.1038, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.9794060757765388, | |
| "grad_norm": 0.43940603733062744, | |
| "learning_rate": 2.2004440202911814e-07, | |
| "loss": 2.1152, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 0.9803163044714984, | |
| "grad_norm": 0.43396398425102234, | |
| "learning_rate": 2.0133235281156736e-07, | |
| "loss": 2.0175, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 0.981226533166458, | |
| "grad_norm": 0.43348294496536255, | |
| "learning_rate": 1.83450858653178e-07, | |
| "loss": 2.0593, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 0.9821367618614176, | |
| "grad_norm": 0.43314129114151, | |
| "learning_rate": 1.664000683692324e-07, | |
| "loss": 1.9414, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 0.9830469905563773, | |
| "grad_norm": 0.47113651037216187, | |
| "learning_rate": 1.5018012386162072e-07, | |
| "loss": 2.3204, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.983957219251337, | |
| "grad_norm": 0.43835383653640747, | |
| "learning_rate": 1.3479116011769767e-07, | |
| "loss": 2.0271, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 0.9848674479462965, | |
| "grad_norm": 0.42517396807670593, | |
| "learning_rate": 1.2023330520911646e-07, | |
| "loss": 1.8366, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 0.9857776766412562, | |
| "grad_norm": 0.4593140482902527, | |
| "learning_rate": 1.0650668029079658e-07, | |
| "loss": 2.0962, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 0.9866879053362158, | |
| "grad_norm": 0.4892318546772003, | |
| "learning_rate": 9.361139959993549e-08, | |
| "loss": 2.144, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 0.9875981340311754, | |
| "grad_norm": 0.48190560936927795, | |
| "learning_rate": 8.154757045497619e-08, | |
| "loss": 2.1081, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.988508362726135, | |
| "grad_norm": 0.4542143940925598, | |
| "learning_rate": 7.0315293254819e-08, | |
| "loss": 1.9731, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 0.9894185914210946, | |
| "grad_norm": 0.4879550337791443, | |
| "learning_rate": 5.991466147791113e-08, | |
| "loss": 2.0791, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 0.9903288201160542, | |
| "grad_norm": 0.5068708062171936, | |
| "learning_rate": 5.0345761681491746e-08, | |
| "loss": 2.181, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 0.9912390488110138, | |
| "grad_norm": 0.5021325945854187, | |
| "learning_rate": 4.1608673500859175e-08, | |
| "loss": 2.1546, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 0.9921492775059734, | |
| "grad_norm": 0.5045585036277771, | |
| "learning_rate": 3.370346964876036e-08, | |
| "loss": 2.0335, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.993059506200933, | |
| "grad_norm": 0.5066413879394531, | |
| "learning_rate": 2.6630215914702495e-08, | |
| "loss": 2.0639, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 0.9939697348958926, | |
| "grad_norm": 0.5761440992355347, | |
| "learning_rate": 2.038897116447558e-08, | |
| "loss": 2.3702, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 0.9948799635908522, | |
| "grad_norm": 0.5774243474006653, | |
| "learning_rate": 1.4979787339619578e-08, | |
| "loss": 2.1865, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 0.9957901922858118, | |
| "grad_norm": 0.6156805157661438, | |
| "learning_rate": 1.0402709457035808e-08, | |
| "loss": 2.3407, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 0.9967004209807714, | |
| "grad_norm": 0.6527782082557678, | |
| "learning_rate": 6.657775608553962e-09, | |
| "loss": 2.3115, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.997610649675731, | |
| "grad_norm": 0.7234971523284912, | |
| "learning_rate": 3.745016960665648e-09, | |
| "loss": 2.5377, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 0.9985208783706906, | |
| "grad_norm": 0.7837737798690796, | |
| "learning_rate": 1.6644577542357375e-09, | |
| "loss": 2.4761, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 0.9994311070656502, | |
| "grad_norm": 1.1138982772827148, | |
| "learning_rate": 4.1611530431362453e-10, | |
| "loss": 2.4967, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 1.0006826715212198, | |
| "grad_norm": 3.5543136596679688, | |
| "learning_rate": 0.0, | |
| "loss": 4.7799, | |
| "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 | |
| } | |