| { | |
| "best_metric": 1.7323445081710815, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-350", | |
| "epoch": 0.06973153359565672, | |
| "eval_steps": 50, | |
| "global_step": 350, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00019923295313044778, | |
| "grad_norm": 1.2284634113311768, | |
| "learning_rate": 1e-05, | |
| "loss": 2.9428, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00019923295313044778, | |
| "eval_loss": 4.057371616363525, | |
| "eval_runtime": 602.6822, | |
| "eval_samples_per_second": 14.027, | |
| "eval_steps_per_second": 3.508, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00039846590626089557, | |
| "grad_norm": 1.68904447555542, | |
| "learning_rate": 2e-05, | |
| "loss": 3.5104, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0005976988593913434, | |
| "grad_norm": 1.1529206037521362, | |
| "learning_rate": 3e-05, | |
| "loss": 2.5356, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0007969318125217911, | |
| "grad_norm": 1.5264250040054321, | |
| "learning_rate": 4e-05, | |
| "loss": 3.2004, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0009961647656522388, | |
| "grad_norm": 1.771896481513977, | |
| "learning_rate": 5e-05, | |
| "loss": 3.2337, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0011953977187826868, | |
| "grad_norm": 1.5935925245285034, | |
| "learning_rate": 6e-05, | |
| "loss": 3.3152, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0013946306719131345, | |
| "grad_norm": 1.6249325275421143, | |
| "learning_rate": 7e-05, | |
| "loss": 2.8173, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0015938636250435823, | |
| "grad_norm": 1.6689910888671875, | |
| "learning_rate": 8e-05, | |
| "loss": 2.8131, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.00179309657817403, | |
| "grad_norm": 1.5419508218765259, | |
| "learning_rate": 9e-05, | |
| "loss": 2.5659, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0019923295313044776, | |
| "grad_norm": 1.5755772590637207, | |
| "learning_rate": 0.0001, | |
| "loss": 2.3418, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0021915624844349255, | |
| "grad_norm": 1.8057647943496704, | |
| "learning_rate": 9.99983777858264e-05, | |
| "loss": 2.3202, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0023907954375653735, | |
| "grad_norm": 1.9575777053833008, | |
| "learning_rate": 9.999351124856874e-05, | |
| "loss": 2.4932, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.002590028390695821, | |
| "grad_norm": 1.9527984857559204, | |
| "learning_rate": 9.998540070400966e-05, | |
| "loss": 1.8685, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.002789261343826269, | |
| "grad_norm": 1.5637708902359009, | |
| "learning_rate": 9.997404667843075e-05, | |
| "loss": 1.831, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0029884942969567166, | |
| "grad_norm": 1.2870111465454102, | |
| "learning_rate": 9.995944990857849e-05, | |
| "loss": 2.0248, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0031877272500871645, | |
| "grad_norm": 1.454819679260254, | |
| "learning_rate": 9.994161134161634e-05, | |
| "loss": 1.9433, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.003386960203217612, | |
| "grad_norm": 1.470567226409912, | |
| "learning_rate": 9.992053213506334e-05, | |
| "loss": 1.8326, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.00358619315634806, | |
| "grad_norm": 1.2251739501953125, | |
| "learning_rate": 9.989621365671902e-05, | |
| "loss": 1.9384, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0037854261094785076, | |
| "grad_norm": 1.148119568824768, | |
| "learning_rate": 9.986865748457457e-05, | |
| "loss": 1.9556, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.003984659062608955, | |
| "grad_norm": 1.022918939590454, | |
| "learning_rate": 9.983786540671051e-05, | |
| "loss": 1.7457, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0041838920157394035, | |
| "grad_norm": 1.0495177507400513, | |
| "learning_rate": 9.980383942118066e-05, | |
| "loss": 1.7567, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.004383124968869851, | |
| "grad_norm": 1.1652737855911255, | |
| "learning_rate": 9.976658173588244e-05, | |
| "loss": 1.7853, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.004582357922000299, | |
| "grad_norm": 1.1292279958724976, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 1.6153, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.004781590875130747, | |
| "grad_norm": 1.0591461658477783, | |
| "learning_rate": 9.968238114591566e-05, | |
| "loss": 2.0229, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0049808238282611945, | |
| "grad_norm": 1.0844651460647583, | |
| "learning_rate": 9.96354437049027e-05, | |
| "loss": 1.8225, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.005180056781391642, | |
| "grad_norm": 1.2742100954055786, | |
| "learning_rate": 9.95852854910781e-05, | |
| "loss": 2.1446, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.00537928973452209, | |
| "grad_norm": 1.0586717128753662, | |
| "learning_rate": 9.953190975913647e-05, | |
| "loss": 1.8019, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.005578522687652538, | |
| "grad_norm": 0.8670190572738647, | |
| "learning_rate": 9.947531997255256e-05, | |
| "loss": 1.8777, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.005777755640782986, | |
| "grad_norm": 1.0841562747955322, | |
| "learning_rate": 9.941551980335652e-05, | |
| "loss": 1.8035, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.005976988593913433, | |
| "grad_norm": 1.0292096138000488, | |
| "learning_rate": 9.935251313189564e-05, | |
| "loss": 1.8991, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.006176221547043881, | |
| "grad_norm": 0.9672908782958984, | |
| "learning_rate": 9.928630404658255e-05, | |
| "loss": 1.8272, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.006375454500174329, | |
| "grad_norm": 0.9663026332855225, | |
| "learning_rate": 9.921689684362989e-05, | |
| "loss": 1.9222, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.006574687453304777, | |
| "grad_norm": 0.8640333414077759, | |
| "learning_rate": 9.914429602677162e-05, | |
| "loss": 1.9071, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.006773920406435224, | |
| "grad_norm": 0.9674398303031921, | |
| "learning_rate": 9.906850630697068e-05, | |
| "loss": 2.113, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0069731533595656725, | |
| "grad_norm": 0.9393993616104126, | |
| "learning_rate": 9.898953260211338e-05, | |
| "loss": 1.8711, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.00717238631269612, | |
| "grad_norm": 0.9489927291870117, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 2.0426, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.007371619265826568, | |
| "grad_norm": 1.1132601499557495, | |
| "learning_rate": 9.882205394146361e-05, | |
| "loss": 2.3052, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.007570852218957015, | |
| "grad_norm": 1.2090137004852295, | |
| "learning_rate": 9.87335598531214e-05, | |
| "loss": 2.0266, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.0077700851720874636, | |
| "grad_norm": 1.0615320205688477, | |
| "learning_rate": 9.864190351391822e-05, | |
| "loss": 2.1822, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.00796931812521791, | |
| "grad_norm": 0.9923949837684631, | |
| "learning_rate": 9.85470908713026e-05, | |
| "loss": 1.9442, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.008168551078348359, | |
| "grad_norm": 1.0683104991912842, | |
| "learning_rate": 9.844912807753104e-05, | |
| "loss": 2.1782, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.008367784031478807, | |
| "grad_norm": 1.1865333318710327, | |
| "learning_rate": 9.834802148926882e-05, | |
| "loss": 2.1076, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.008567016984609254, | |
| "grad_norm": 0.9799418449401855, | |
| "learning_rate": 9.824377766717759e-05, | |
| "loss": 2.0544, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.008766249937739702, | |
| "grad_norm": 1.0477815866470337, | |
| "learning_rate": 9.813640337548954e-05, | |
| "loss": 1.9788, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.00896548289087015, | |
| "grad_norm": 0.9999564290046692, | |
| "learning_rate": 9.802590558156862e-05, | |
| "loss": 2.12, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.009164715844000597, | |
| "grad_norm": 0.998177170753479, | |
| "learning_rate": 9.791229145545831e-05, | |
| "loss": 2.039, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.009363948797131046, | |
| "grad_norm": 1.1539019346237183, | |
| "learning_rate": 9.779556836941645e-05, | |
| "loss": 2.3537, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.009563181750261494, | |
| "grad_norm": 1.0150654315948486, | |
| "learning_rate": 9.767574389743682e-05, | |
| "loss": 2.0415, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.00976241470339194, | |
| "grad_norm": 1.0183194875717163, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 2.0481, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.009961647656522389, | |
| "grad_norm": 0.9892953038215637, | |
| "learning_rate": 9.742682209735727e-05, | |
| "loss": 2.1764, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.009961647656522389, | |
| "eval_loss": 1.82403564453125, | |
| "eval_runtime": 607.5197, | |
| "eval_samples_per_second": 13.916, | |
| "eval_steps_per_second": 3.48, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.010160880609652836, | |
| "grad_norm": 0.8542794585227966, | |
| "learning_rate": 9.729774092143627e-05, | |
| "loss": 1.6775, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.010360113562783284, | |
| "grad_norm": 0.8750085234642029, | |
| "learning_rate": 9.716559066288715e-05, | |
| "loss": 1.4926, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.010559346515913733, | |
| "grad_norm": 0.875607430934906, | |
| "learning_rate": 9.703037989675087e-05, | |
| "loss": 1.8167, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.01075857946904418, | |
| "grad_norm": 0.7322773933410645, | |
| "learning_rate": 9.689211739666023e-05, | |
| "loss": 1.4753, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.010957812422174628, | |
| "grad_norm": 0.648223876953125, | |
| "learning_rate": 9.675081213427076e-05, | |
| "loss": 1.3975, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.011157045375305076, | |
| "grad_norm": 0.6270776391029358, | |
| "learning_rate": 9.66064732786784e-05, | |
| "loss": 1.1602, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.011356278328435523, | |
| "grad_norm": 0.6656715273857117, | |
| "learning_rate": 9.645911019582467e-05, | |
| "loss": 1.3633, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.011555511281565971, | |
| "grad_norm": 0.7466765642166138, | |
| "learning_rate": 9.630873244788883e-05, | |
| "loss": 1.4871, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.01175474423469642, | |
| "grad_norm": 0.8152953386306763, | |
| "learning_rate": 9.615534979266745e-05, | |
| "loss": 1.4701, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.011953977187826866, | |
| "grad_norm": 0.8492268919944763, | |
| "learning_rate": 9.599897218294122e-05, | |
| "loss": 1.688, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.012153210140957315, | |
| "grad_norm": 0.7441631555557251, | |
| "learning_rate": 9.583960976582913e-05, | |
| "loss": 1.5078, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.012352443094087761, | |
| "grad_norm": 0.8003615736961365, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 1.86, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.01255167604721821, | |
| "grad_norm": 0.7498038411140442, | |
| "learning_rate": 9.551197206565173e-05, | |
| "loss": 1.4841, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.012750909000348658, | |
| "grad_norm": 0.7930290699005127, | |
| "learning_rate": 9.534371804252728e-05, | |
| "loss": 1.6008, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.012950141953479105, | |
| "grad_norm": 0.7177601456642151, | |
| "learning_rate": 9.517252173051911e-05, | |
| "loss": 1.4822, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.013149374906609553, | |
| "grad_norm": 0.7637808918952942, | |
| "learning_rate": 9.49983942383106e-05, | |
| "loss": 1.6999, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.013348607859740002, | |
| "grad_norm": 0.6996709108352661, | |
| "learning_rate": 9.482134686478519e-05, | |
| "loss": 1.5346, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.013547840812870448, | |
| "grad_norm": 0.8542125821113586, | |
| "learning_rate": 9.464139109829321e-05, | |
| "loss": 1.8176, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.013747073766000897, | |
| "grad_norm": 0.8093202710151672, | |
| "learning_rate": 9.445853861590647e-05, | |
| "loss": 1.5795, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.013946306719131345, | |
| "grad_norm": 0.8053815960884094, | |
| "learning_rate": 9.42728012826605e-05, | |
| "loss": 1.7278, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.014145539672261792, | |
| "grad_norm": 0.823073148727417, | |
| "learning_rate": 9.408419115078471e-05, | |
| "loss": 1.6882, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.01434477262539224, | |
| "grad_norm": 0.9014071822166443, | |
| "learning_rate": 9.389272045892024e-05, | |
| "loss": 1.7093, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.014544005578522687, | |
| "grad_norm": 0.9905192852020264, | |
| "learning_rate": 9.36984016313259e-05, | |
| "loss": 1.8529, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.014743238531653135, | |
| "grad_norm": 0.8374629020690918, | |
| "learning_rate": 9.350124727707197e-05, | |
| "loss": 1.7811, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.014942471484783584, | |
| "grad_norm": 1.0269732475280762, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 1.8239, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.01514170443791403, | |
| "grad_norm": 0.7490584254264832, | |
| "learning_rate": 9.309848334400246e-05, | |
| "loss": 1.5871, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.015340937391044479, | |
| "grad_norm": 0.8312053680419922, | |
| "learning_rate": 9.289289989996133e-05, | |
| "loss": 1.6725, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.015540170344174927, | |
| "grad_norm": 0.7715761661529541, | |
| "learning_rate": 9.268453319711363e-05, | |
| "loss": 1.6291, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.015739403297305374, | |
| "grad_norm": 0.8476051688194275, | |
| "learning_rate": 9.247339675607605e-05, | |
| "loss": 1.8967, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.01593863625043582, | |
| "grad_norm": 0.7571315765380859, | |
| "learning_rate": 9.225950427718975e-05, | |
| "loss": 1.6547, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.01613786920356627, | |
| "grad_norm": 0.831329882144928, | |
| "learning_rate": 9.204286963963111e-05, | |
| "loss": 1.9368, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.016337102156696717, | |
| "grad_norm": 0.7381522059440613, | |
| "learning_rate": 9.182350690051133e-05, | |
| "loss": 1.8917, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.016536335109827164, | |
| "grad_norm": 0.9195905923843384, | |
| "learning_rate": 9.160143029396422e-05, | |
| "loss": 1.907, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.016735568062957614, | |
| "grad_norm": 0.7933158874511719, | |
| "learning_rate": 9.13766542302225e-05, | |
| "loss": 1.7316, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.01693480101608806, | |
| "grad_norm": 0.8460677266120911, | |
| "learning_rate": 9.114919329468282e-05, | |
| "loss": 2.0394, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.017134033969218507, | |
| "grad_norm": 0.8626026511192322, | |
| "learning_rate": 9.091906224695935e-05, | |
| "loss": 1.7807, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.017333266922348958, | |
| "grad_norm": 1.0532417297363281, | |
| "learning_rate": 9.068627601992598e-05, | |
| "loss": 1.9692, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.017532499875479404, | |
| "grad_norm": 0.9541111588478088, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 2.1123, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.01773173282860985, | |
| "grad_norm": 0.8808003067970276, | |
| "learning_rate": 9.021279861989885e-05, | |
| "loss": 2.1134, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.0179309657817403, | |
| "grad_norm": 0.9053674936294556, | |
| "learning_rate": 8.997213817017507e-05, | |
| "loss": 1.8627, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.018130198734870748, | |
| "grad_norm": 0.9115805625915527, | |
| "learning_rate": 8.972888398568772e-05, | |
| "loss": 1.8644, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.018329431688001194, | |
| "grad_norm": 1.0219849348068237, | |
| "learning_rate": 8.948305185085225e-05, | |
| "loss": 2.2721, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.018528664641131645, | |
| "grad_norm": 0.9994537234306335, | |
| "learning_rate": 8.92346577173636e-05, | |
| "loss": 2.1956, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.01872789759426209, | |
| "grad_norm": 0.881737470626831, | |
| "learning_rate": 8.898371770316111e-05, | |
| "loss": 1.9247, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.018927130547392538, | |
| "grad_norm": 1.0032176971435547, | |
| "learning_rate": 8.873024809138272e-05, | |
| "loss": 2.1392, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.019126363500522988, | |
| "grad_norm": 1.0586591958999634, | |
| "learning_rate": 8.847426532930831e-05, | |
| "loss": 2.0194, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.019325596453653435, | |
| "grad_norm": 1.0651748180389404, | |
| "learning_rate": 8.821578602729242e-05, | |
| "loss": 2.0458, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.01952482940678388, | |
| "grad_norm": 1.0627988576889038, | |
| "learning_rate": 8.795482695768658e-05, | |
| "loss": 1.9618, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.01972406235991433, | |
| "grad_norm": 0.9913427233695984, | |
| "learning_rate": 8.769140505375085e-05, | |
| "loss": 1.8278, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.019923295313044778, | |
| "grad_norm": 1.2119276523590088, | |
| "learning_rate": 8.742553740855506e-05, | |
| "loss": 2.1889, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.019923295313044778, | |
| "eval_loss": 1.8152918815612793, | |
| "eval_runtime": 607.3335, | |
| "eval_samples_per_second": 13.92, | |
| "eval_steps_per_second": 3.481, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.020122528266175225, | |
| "grad_norm": 0.7810643315315247, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 1.3115, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.02032176121930567, | |
| "grad_norm": 0.9216941595077515, | |
| "learning_rate": 8.688653405904652e-05, | |
| "loss": 1.8541, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.02052099417243612, | |
| "grad_norm": 0.733933687210083, | |
| "learning_rate": 8.661343332988869e-05, | |
| "loss": 1.5391, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.02072022712556657, | |
| "grad_norm": 0.7394716739654541, | |
| "learning_rate": 8.633795680751116e-05, | |
| "loss": 1.3938, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.020919460078697015, | |
| "grad_norm": 0.8477760553359985, | |
| "learning_rate": 8.606012236719073e-05, | |
| "loss": 1.9247, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.021118693031827465, | |
| "grad_norm": 0.669335663318634, | |
| "learning_rate": 8.577994803720606e-05, | |
| "loss": 1.728, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.021317925984957912, | |
| "grad_norm": 0.6468355655670166, | |
| "learning_rate": 8.549745199766792e-05, | |
| "loss": 1.3566, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.02151715893808836, | |
| "grad_norm": 0.6598238945007324, | |
| "learning_rate": 8.521265257933948e-05, | |
| "loss": 1.4654, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.02171639189121881, | |
| "grad_norm": 0.6626983284950256, | |
| "learning_rate": 8.492556826244687e-05, | |
| "loss": 1.4336, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.021915624844349255, | |
| "grad_norm": 0.6764017939567566, | |
| "learning_rate": 8.463621767547998e-05, | |
| "loss": 1.3487, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.022114857797479702, | |
| "grad_norm": 0.8265367150306702, | |
| "learning_rate": 8.434461959398376e-05, | |
| "loss": 1.487, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.022314090750610152, | |
| "grad_norm": 0.6030274629592896, | |
| "learning_rate": 8.405079293933986e-05, | |
| "loss": 1.3341, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.0225133237037406, | |
| "grad_norm": 0.7178739309310913, | |
| "learning_rate": 8.375475677753881e-05, | |
| "loss": 1.537, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.022712556656871045, | |
| "grad_norm": 0.6907858848571777, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 1.6211, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.022911789610001496, | |
| "grad_norm": 0.8321515321731567, | |
| "learning_rate": 8.315613291203976e-05, | |
| "loss": 1.6012, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.023111022563131942, | |
| "grad_norm": 0.7017425894737244, | |
| "learning_rate": 8.285358405218655e-05, | |
| "loss": 1.6058, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.02331025551626239, | |
| "grad_norm": 0.6512852311134338, | |
| "learning_rate": 8.25489033703452e-05, | |
| "loss": 1.4762, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.02350948846939284, | |
| "grad_norm": 0.6753814816474915, | |
| "learning_rate": 8.224211063680853e-05, | |
| "loss": 1.3924, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.023708721422523286, | |
| "grad_norm": 0.6458118557929993, | |
| "learning_rate": 8.19332257589174e-05, | |
| "loss": 1.427, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.023907954375653732, | |
| "grad_norm": 0.7541099786758423, | |
| "learning_rate": 8.162226877976887e-05, | |
| "loss": 1.6374, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.024107187328784183, | |
| "grad_norm": 0.7807430624961853, | |
| "learning_rate": 8.130925987691569e-05, | |
| "loss": 1.8552, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.02430642028191463, | |
| "grad_norm": 0.7646523118019104, | |
| "learning_rate": 8.099421936105702e-05, | |
| "loss": 1.5886, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.024505653235045076, | |
| "grad_norm": 0.7880610823631287, | |
| "learning_rate": 8.067716767472045e-05, | |
| "loss": 1.6209, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.024704886188175523, | |
| "grad_norm": 0.7491058111190796, | |
| "learning_rate": 8.035812539093557e-05, | |
| "loss": 1.6557, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.024904119141305973, | |
| "grad_norm": 0.7884393930435181, | |
| "learning_rate": 8.003711321189895e-05, | |
| "loss": 1.768, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.02510335209443642, | |
| "grad_norm": 0.8286698460578918, | |
| "learning_rate": 7.971415196763088e-05, | |
| "loss": 1.7031, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.025302585047566866, | |
| "grad_norm": 0.9087210297584534, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 1.9608, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.025501818000697316, | |
| "grad_norm": 0.7920821309089661, | |
| "learning_rate": 7.906246623448183e-05, | |
| "loss": 1.6142, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.025701050953827763, | |
| "grad_norm": 0.8957509398460388, | |
| "learning_rate": 7.873378403255419e-05, | |
| "loss": 1.7978, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.02590028390695821, | |
| "grad_norm": 0.6929237246513367, | |
| "learning_rate": 7.840323733655778e-05, | |
| "loss": 1.5131, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.02609951686008866, | |
| "grad_norm": 0.8800172805786133, | |
| "learning_rate": 7.807084759519405e-05, | |
| "loss": 1.8921, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.026298749813219106, | |
| "grad_norm": 0.7827117443084717, | |
| "learning_rate": 7.773663637675694e-05, | |
| "loss": 1.7208, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.026497982766349553, | |
| "grad_norm": 0.7875890135765076, | |
| "learning_rate": 7.740062536773352e-05, | |
| "loss": 1.7275, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.026697215719480003, | |
| "grad_norm": 0.8130915760993958, | |
| "learning_rate": 7.706283637139658e-05, | |
| "loss": 1.8376, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.02689644867261045, | |
| "grad_norm": 0.8318361639976501, | |
| "learning_rate": 7.672329130639005e-05, | |
| "loss": 1.8565, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.027095681625740897, | |
| "grad_norm": 0.8131917715072632, | |
| "learning_rate": 7.638201220530665e-05, | |
| "loss": 1.7917, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.027294914578871347, | |
| "grad_norm": 0.9158076643943787, | |
| "learning_rate": 7.603902121325813e-05, | |
| "loss": 1.9982, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.027494147532001793, | |
| "grad_norm": 0.9296155571937561, | |
| "learning_rate": 7.569434058643844e-05, | |
| "loss": 2.0053, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.02769338048513224, | |
| "grad_norm": 0.9064381718635559, | |
| "learning_rate": 7.534799269067953e-05, | |
| "loss": 1.9478, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.02789261343826269, | |
| "grad_norm": 0.9871942400932312, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 2.1819, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.028091846391393137, | |
| "grad_norm": 0.8777363300323486, | |
| "learning_rate": 7.465038509514688e-05, | |
| "loss": 2.0469, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.028291079344523583, | |
| "grad_norm": 1.0275037288665771, | |
| "learning_rate": 7.42991706621303e-05, | |
| "loss": 1.9602, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.028490312297654034, | |
| "grad_norm": 0.8730279803276062, | |
| "learning_rate": 7.394637949075154e-05, | |
| "loss": 1.9302, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.02868954525078448, | |
| "grad_norm": 0.8100825548171997, | |
| "learning_rate": 7.35920344731241e-05, | |
| "loss": 1.6562, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.028888778203914927, | |
| "grad_norm": 1.087735652923584, | |
| "learning_rate": 7.323615860218843e-05, | |
| "loss": 2.1821, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.029088011157045374, | |
| "grad_norm": 1.0589756965637207, | |
| "learning_rate": 7.287877497021978e-05, | |
| "loss": 2.2272, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.029287244110175824, | |
| "grad_norm": 1.2219501733779907, | |
| "learning_rate": 7.251990676732984e-05, | |
| "loss": 2.2479, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.02948647706330627, | |
| "grad_norm": 1.0689516067504883, | |
| "learning_rate": 7.215957727996207e-05, | |
| "loss": 2.2357, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.029685710016436717, | |
| "grad_norm": 0.8567739129066467, | |
| "learning_rate": 7.179780988938051e-05, | |
| "loss": 1.8367, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.029884942969567167, | |
| "grad_norm": 1.024474024772644, | |
| "learning_rate": 7.143462807015271e-05, | |
| "loss": 2.0585, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.029884942969567167, | |
| "eval_loss": 1.7750892639160156, | |
| "eval_runtime": 607.0219, | |
| "eval_samples_per_second": 13.927, | |
| "eval_steps_per_second": 3.483, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.030084175922697614, | |
| "grad_norm": 0.8313390016555786, | |
| "learning_rate": 7.107005538862646e-05, | |
| "loss": 1.5251, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.03028340887582806, | |
| "grad_norm": 0.6984227299690247, | |
| "learning_rate": 7.07041155014006e-05, | |
| "loss": 1.4523, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.03048264182895851, | |
| "grad_norm": 0.7919133305549622, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 1.6288, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.030681874782088957, | |
| "grad_norm": 0.8622307181358337, | |
| "learning_rate": 6.996822917828477e-05, | |
| "loss": 1.6547, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.030881107735219404, | |
| "grad_norm": 0.6328975558280945, | |
| "learning_rate": 6.959833049300377e-05, | |
| "loss": 1.2618, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.031080340688349854, | |
| "grad_norm": 0.7241643667221069, | |
| "learning_rate": 6.922716010014255e-05, | |
| "loss": 1.5949, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.0312795736414803, | |
| "grad_norm": 0.63180011510849, | |
| "learning_rate": 6.885474208441603e-05, | |
| "loss": 1.4086, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.03147880659461075, | |
| "grad_norm": 0.7782214879989624, | |
| "learning_rate": 6.848110061149556e-05, | |
| "loss": 1.6755, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.0316780395477412, | |
| "grad_norm": 0.7412595152854919, | |
| "learning_rate": 6.810625992644085e-05, | |
| "loss": 1.5487, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.03187727250087164, | |
| "grad_norm": 0.7262195348739624, | |
| "learning_rate": 6.773024435212678e-05, | |
| "loss": 1.5054, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.03207650545400209, | |
| "grad_norm": 0.6977418065071106, | |
| "learning_rate": 6.735307828766515e-05, | |
| "loss": 1.4481, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.03227573840713254, | |
| "grad_norm": 0.8268071413040161, | |
| "learning_rate": 6.697478620682137e-05, | |
| "loss": 1.5609, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.032474971360262984, | |
| "grad_norm": 0.7182347178459167, | |
| "learning_rate": 6.659539265642643e-05, | |
| "loss": 1.6529, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.032674204313393435, | |
| "grad_norm": 0.6928174495697021, | |
| "learning_rate": 6.621492225478414e-05, | |
| "loss": 1.5177, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.032873437266523885, | |
| "grad_norm": 0.7981646060943604, | |
| "learning_rate": 6.583339969007363e-05, | |
| "loss": 1.6601, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.03307267021965433, | |
| "grad_norm": 0.8732507228851318, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 1.6293, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.03327190317278478, | |
| "grad_norm": 0.7294723391532898, | |
| "learning_rate": 6.506729716392481e-05, | |
| "loss": 1.5132, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.03347113612591523, | |
| "grad_norm": 0.7707034945487976, | |
| "learning_rate": 6.468276691378155e-05, | |
| "loss": 1.6842, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.03367036907904567, | |
| "grad_norm": 0.7112640142440796, | |
| "learning_rate": 6.429728391993446e-05, | |
| "loss": 1.5651, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.03386960203217612, | |
| "grad_norm": 0.8815944194793701, | |
| "learning_rate": 6.391087319582264e-05, | |
| "loss": 1.9648, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.03406883498530657, | |
| "grad_norm": 0.8880531191825867, | |
| "learning_rate": 6.35235598150842e-05, | |
| "loss": 1.805, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.034268067938437015, | |
| "grad_norm": 0.7516195178031921, | |
| "learning_rate": 6.313536890992935e-05, | |
| "loss": 1.6364, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.034467300891567465, | |
| "grad_norm": 0.7268112301826477, | |
| "learning_rate": 6.274632566950967e-05, | |
| "loss": 1.7021, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.034666533844697915, | |
| "grad_norm": 0.7526605725288391, | |
| "learning_rate": 6.235645533828349e-05, | |
| "loss": 1.7656, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.03486576679782836, | |
| "grad_norm": 0.7969704270362854, | |
| "learning_rate": 6.19657832143779e-05, | |
| "loss": 1.7097, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.03506499975095881, | |
| "grad_norm": 0.8148013949394226, | |
| "learning_rate": 6.157433464794716e-05, | |
| "loss": 1.8028, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.03526423270408926, | |
| "grad_norm": 0.7393292188644409, | |
| "learning_rate": 6.118213503952779e-05, | |
| "loss": 1.7528, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.0354634656572197, | |
| "grad_norm": 0.8155067563056946, | |
| "learning_rate": 6.078920983839031e-05, | |
| "loss": 1.8142, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.03566269861035015, | |
| "grad_norm": 0.8007950186729431, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 1.6058, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.0358619315634806, | |
| "grad_norm": 0.7950125336647034, | |
| "learning_rate": 6.0001284688802226e-05, | |
| "loss": 1.8424, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.036061164516611045, | |
| "grad_norm": 0.7428115010261536, | |
| "learning_rate": 5.960633586768543e-05, | |
| "loss": 1.6715, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.036260397469741495, | |
| "grad_norm": 0.8062713742256165, | |
| "learning_rate": 5.921076370520058e-05, | |
| "loss": 1.885, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.036459630422871946, | |
| "grad_norm": 0.8469149470329285, | |
| "learning_rate": 5.8814593869458455e-05, | |
| "loss": 1.7839, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.03665886337600239, | |
| "grad_norm": 0.7747473120689392, | |
| "learning_rate": 5.841785206735192e-05, | |
| "loss": 1.8037, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.03685809632913284, | |
| "grad_norm": 0.9092196822166443, | |
| "learning_rate": 5.8020564042888015e-05, | |
| "loss": 1.6015, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.03705732928226329, | |
| "grad_norm": 0.8225083351135254, | |
| "learning_rate": 5.762275557551727e-05, | |
| "loss": 1.7744, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.03725656223539373, | |
| "grad_norm": 0.8785688877105713, | |
| "learning_rate": 5.7224452478461064e-05, | |
| "loss": 1.9095, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.03745579518852418, | |
| "grad_norm": 0.8963973522186279, | |
| "learning_rate": 5.682568059703659e-05, | |
| "loss": 1.8677, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.03765502814165463, | |
| "grad_norm": 0.9210934638977051, | |
| "learning_rate": 5.642646580697973e-05, | |
| "loss": 1.9918, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.037854261094785076, | |
| "grad_norm": 1.1235376596450806, | |
| "learning_rate": 5.602683401276615e-05, | |
| "loss": 2.0277, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.038053494047915526, | |
| "grad_norm": 0.9675539135932922, | |
| "learning_rate": 5.562681114593028e-05, | |
| "loss": 2.0708, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.038252727001045976, | |
| "grad_norm": 0.9660516381263733, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 2.0641, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.03845195995417642, | |
| "grad_norm": 0.9473053812980652, | |
| "learning_rate": 5.482569604572576e-05, | |
| "loss": 2.0431, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.03865119290730687, | |
| "grad_norm": 0.8935657739639282, | |
| "learning_rate": 5.442465579556793e-05, | |
| "loss": 1.9439, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.03885042586043732, | |
| "grad_norm": 0.919715940952301, | |
| "learning_rate": 5.402332843583631e-05, | |
| "loss": 1.9152, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.03904965881356776, | |
| "grad_norm": 0.8705791234970093, | |
| "learning_rate": 5.3621740008088126e-05, | |
| "loss": 1.9256, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.03924889176669821, | |
| "grad_norm": 0.9605872631072998, | |
| "learning_rate": 5.321991657082097e-05, | |
| "loss": 1.9576, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.03944812471982866, | |
| "grad_norm": 0.9608198404312134, | |
| "learning_rate": 5.281788419778187e-05, | |
| "loss": 2.0148, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.039647357672959106, | |
| "grad_norm": 0.9603476524353027, | |
| "learning_rate": 5.2415668976275355e-05, | |
| "loss": 2.0684, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.039846590626089556, | |
| "grad_norm": 1.0739209651947021, | |
| "learning_rate": 5.201329700547076e-05, | |
| "loss": 1.8792, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.039846590626089556, | |
| "eval_loss": 1.760303020477295, | |
| "eval_runtime": 607.8397, | |
| "eval_samples_per_second": 13.908, | |
| "eval_steps_per_second": 3.478, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.04004582357922, | |
| "grad_norm": 0.6963627338409424, | |
| "learning_rate": 5.161079439470866e-05, | |
| "loss": 1.3263, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.04024505653235045, | |
| "grad_norm": 0.76730877161026, | |
| "learning_rate": 5.1208187261806615e-05, | |
| "loss": 1.3553, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.0404442894854809, | |
| "grad_norm": 0.8844528198242188, | |
| "learning_rate": 5.080550173136457e-05, | |
| "loss": 1.5542, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.04064352243861134, | |
| "grad_norm": 0.7753970623016357, | |
| "learning_rate": 5.0402763933069496e-05, | |
| "loss": 1.6801, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.04084275539174179, | |
| "grad_norm": 0.8743450045585632, | |
| "learning_rate": 5e-05, | |
| "loss": 1.6864, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.04104198834487224, | |
| "grad_norm": 0.7265819311141968, | |
| "learning_rate": 4.9597236066930516e-05, | |
| "loss": 1.518, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.04124122129800269, | |
| "grad_norm": 0.8164213299751282, | |
| "learning_rate": 4.919449826863544e-05, | |
| "loss": 1.812, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.04144045425113314, | |
| "grad_norm": 0.7274611592292786, | |
| "learning_rate": 4.87918127381934e-05, | |
| "loss": 1.5158, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.04163968720426359, | |
| "grad_norm": 0.6912136077880859, | |
| "learning_rate": 4.8389205605291365e-05, | |
| "loss": 1.5693, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.04183892015739403, | |
| "grad_norm": 0.7156378626823425, | |
| "learning_rate": 4.798670299452926e-05, | |
| "loss": 1.4035, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.04203815311052448, | |
| "grad_norm": 0.7882273197174072, | |
| "learning_rate": 4.758433102372466e-05, | |
| "loss": 1.7326, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.04223738606365493, | |
| "grad_norm": 0.793198823928833, | |
| "learning_rate": 4.7182115802218126e-05, | |
| "loss": 1.5803, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.042436619016785374, | |
| "grad_norm": 0.669638991355896, | |
| "learning_rate": 4.678008342917903e-05, | |
| "loss": 1.5177, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.042635851969915824, | |
| "grad_norm": 0.7197118997573853, | |
| "learning_rate": 4.6378259991911886e-05, | |
| "loss": 1.6084, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.042835084923046274, | |
| "grad_norm": 0.8170608878135681, | |
| "learning_rate": 4.597667156416371e-05, | |
| "loss": 1.4552, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.04303431787617672, | |
| "grad_norm": 0.6870235204696655, | |
| "learning_rate": 4.5575344204432084e-05, | |
| "loss": 1.6176, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.04323355082930717, | |
| "grad_norm": 0.683409571647644, | |
| "learning_rate": 4.5174303954274244e-05, | |
| "loss": 1.4193, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.04343278378243762, | |
| "grad_norm": 0.8911597728729248, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 1.4589, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.04363201673556806, | |
| "grad_norm": 0.6283190846443176, | |
| "learning_rate": 4.437318885406973e-05, | |
| "loss": 1.3721, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.04383124968869851, | |
| "grad_norm": 0.7931171655654907, | |
| "learning_rate": 4.397316598723385e-05, | |
| "loss": 1.6176, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.04403048264182896, | |
| "grad_norm": 0.6742306351661682, | |
| "learning_rate": 4.3573534193020274e-05, | |
| "loss": 1.4137, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.044229715594959404, | |
| "grad_norm": 0.742749035358429, | |
| "learning_rate": 4.317431940296343e-05, | |
| "loss": 1.5411, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.044428948548089854, | |
| "grad_norm": 0.6754933595657349, | |
| "learning_rate": 4.277554752153895e-05, | |
| "loss": 1.4061, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.044628181501220304, | |
| "grad_norm": 0.7746677994728088, | |
| "learning_rate": 4.237724442448273e-05, | |
| "loss": 1.6348, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.04482741445435075, | |
| "grad_norm": 0.8098635077476501, | |
| "learning_rate": 4.197943595711198e-05, | |
| "loss": 1.6149, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.0450266474074812, | |
| "grad_norm": 0.7978078126907349, | |
| "learning_rate": 4.1582147932648074e-05, | |
| "loss": 1.7439, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.04522588036061165, | |
| "grad_norm": 0.7287787795066833, | |
| "learning_rate": 4.118540613054156e-05, | |
| "loss": 1.4338, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.04542511331374209, | |
| "grad_norm": 0.7751352787017822, | |
| "learning_rate": 4.078923629479943e-05, | |
| "loss": 1.3109, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.04562434626687254, | |
| "grad_norm": 0.8360660672187805, | |
| "learning_rate": 4.039366413231458e-05, | |
| "loss": 1.7695, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.04582357922000299, | |
| "grad_norm": 0.8752302527427673, | |
| "learning_rate": 3.9998715311197785e-05, | |
| "loss": 1.9162, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.046022812173133434, | |
| "grad_norm": 0.8240324258804321, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 1.7368, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.046222045126263885, | |
| "grad_norm": 0.8629664778709412, | |
| "learning_rate": 3.92107901616097e-05, | |
| "loss": 1.8234, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.046421278079394335, | |
| "grad_norm": 0.8692692518234253, | |
| "learning_rate": 3.8817864960472236e-05, | |
| "loss": 1.7835, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.04662051103252478, | |
| "grad_norm": 0.8354206085205078, | |
| "learning_rate": 3.842566535205286e-05, | |
| "loss": 1.7645, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.04681974398565523, | |
| "grad_norm": 0.8872989416122437, | |
| "learning_rate": 3.803421678562213e-05, | |
| "loss": 1.838, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.04701897693878568, | |
| "grad_norm": 0.9834421873092651, | |
| "learning_rate": 3.764354466171652e-05, | |
| "loss": 1.9409, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.04721820989191612, | |
| "grad_norm": 0.8917065262794495, | |
| "learning_rate": 3.725367433049033e-05, | |
| "loss": 1.991, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.04741744284504657, | |
| "grad_norm": 0.8875999450683594, | |
| "learning_rate": 3.6864631090070655e-05, | |
| "loss": 1.9511, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.04761667579817702, | |
| "grad_norm": 0.9579639434814453, | |
| "learning_rate": 3.6476440184915815e-05, | |
| "loss": 1.9965, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.047815908751307465, | |
| "grad_norm": 0.8680365085601807, | |
| "learning_rate": 3.608912680417737e-05, | |
| "loss": 2.0093, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.048015141704437915, | |
| "grad_norm": 0.8048585057258606, | |
| "learning_rate": 3.570271608006555e-05, | |
| "loss": 1.7207, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.048214374657568365, | |
| "grad_norm": 0.9614121913909912, | |
| "learning_rate": 3.531723308621847e-05, | |
| "loss": 2.0035, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.04841360761069881, | |
| "grad_norm": 0.993566632270813, | |
| "learning_rate": 3.493270283607522e-05, | |
| "loss": 1.978, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.04861284056382926, | |
| "grad_norm": 0.9472538232803345, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 2.0894, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.0488120735169597, | |
| "grad_norm": 0.9993523955345154, | |
| "learning_rate": 3.4166600309926387e-05, | |
| "loss": 1.9909, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.04901130647009015, | |
| "grad_norm": 1.0084898471832275, | |
| "learning_rate": 3.3785077745215873e-05, | |
| "loss": 2.0606, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.0492105394232206, | |
| "grad_norm": 0.9558363556861877, | |
| "learning_rate": 3.340460734357359e-05, | |
| "loss": 2.0389, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.049409772376351045, | |
| "grad_norm": 1.0957932472229004, | |
| "learning_rate": 3.3025213793178646e-05, | |
| "loss": 2.042, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.049609005329481495, | |
| "grad_norm": 0.965751051902771, | |
| "learning_rate": 3.264692171233485e-05, | |
| "loss": 2.0706, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.049808238282611945, | |
| "grad_norm": 1.1002291440963745, | |
| "learning_rate": 3.226975564787322e-05, | |
| "loss": 2.1116, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.049808238282611945, | |
| "eval_loss": 1.750185489654541, | |
| "eval_runtime": 608.0539, | |
| "eval_samples_per_second": 13.903, | |
| "eval_steps_per_second": 3.477, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.05000747123574239, | |
| "grad_norm": 0.6731575727462769, | |
| "learning_rate": 3.189374007355917e-05, | |
| "loss": 1.2006, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.05020670418887284, | |
| "grad_norm": 0.8384291529655457, | |
| "learning_rate": 3.151889938850445e-05, | |
| "loss": 1.3281, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.05040593714200329, | |
| "grad_norm": 0.8505686521530151, | |
| "learning_rate": 3.114525791558398e-05, | |
| "loss": 1.6739, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.05060517009513373, | |
| "grad_norm": 0.8174691200256348, | |
| "learning_rate": 3.0772839899857464e-05, | |
| "loss": 1.6558, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.05080440304826418, | |
| "grad_norm": 0.7438680529594421, | |
| "learning_rate": 3.0401669506996256e-05, | |
| "loss": 1.5605, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.05100363600139463, | |
| "grad_norm": 0.9211633205413818, | |
| "learning_rate": 3.003177082171523e-05, | |
| "loss": 1.7153, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.051202868954525076, | |
| "grad_norm": 0.7322139143943787, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 1.4613, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.051402101907655526, | |
| "grad_norm": 0.7553418874740601, | |
| "learning_rate": 2.9295884498599414e-05, | |
| "loss": 1.6048, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.051601334860785976, | |
| "grad_norm": 0.7404634356498718, | |
| "learning_rate": 2.8929944611373554e-05, | |
| "loss": 1.6173, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.05180056781391642, | |
| "grad_norm": 0.6997928619384766, | |
| "learning_rate": 2.8565371929847284e-05, | |
| "loss": 1.4457, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.05199980076704687, | |
| "grad_norm": 0.7190682291984558, | |
| "learning_rate": 2.8202190110619493e-05, | |
| "loss": 1.4685, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.05219903372017732, | |
| "grad_norm": 0.8214508891105652, | |
| "learning_rate": 2.784042272003794e-05, | |
| "loss": 1.6993, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.05239826667330776, | |
| "grad_norm": 0.8008723258972168, | |
| "learning_rate": 2.7480093232670158e-05, | |
| "loss": 1.6652, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.05259749962643821, | |
| "grad_norm": 0.7961799502372742, | |
| "learning_rate": 2.712122502978024e-05, | |
| "loss": 1.6416, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.05279673257956866, | |
| "grad_norm": 0.6760292053222656, | |
| "learning_rate": 2.6763841397811573e-05, | |
| "loss": 1.2897, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.052995965532699106, | |
| "grad_norm": 0.8537053465843201, | |
| "learning_rate": 2.64079655268759e-05, | |
| "loss": 1.5309, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.053195198485829556, | |
| "grad_norm": 0.7754104137420654, | |
| "learning_rate": 2.605362050924848e-05, | |
| "loss": 1.6193, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.053394431438960006, | |
| "grad_norm": 0.7383844256401062, | |
| "learning_rate": 2.57008293378697e-05, | |
| "loss": 1.5956, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.05359366439209045, | |
| "grad_norm": 0.8079094886779785, | |
| "learning_rate": 2.534961490485313e-05, | |
| "loss": 1.6054, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.0537928973452209, | |
| "grad_norm": 0.7139529585838318, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 1.5294, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.05399213029835135, | |
| "grad_norm": 0.825450599193573, | |
| "learning_rate": 2.4652007309320498e-05, | |
| "loss": 1.7282, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.05419136325148179, | |
| "grad_norm": 0.792938232421875, | |
| "learning_rate": 2.430565941356157e-05, | |
| "loss": 1.6671, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.05439059620461224, | |
| "grad_norm": 0.7084431052207947, | |
| "learning_rate": 2.3960978786741877e-05, | |
| "loss": 1.582, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.05458982915774269, | |
| "grad_norm": 0.7082628607749939, | |
| "learning_rate": 2.361798779469336e-05, | |
| "loss": 1.5363, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.05478906211087314, | |
| "grad_norm": 0.808502197265625, | |
| "learning_rate": 2.3276708693609943e-05, | |
| "loss": 1.8669, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.05498829506400359, | |
| "grad_norm": 0.748428463935852, | |
| "learning_rate": 2.2937163628603435e-05, | |
| "loss": 1.7152, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.05518752801713404, | |
| "grad_norm": 0.6738754510879517, | |
| "learning_rate": 2.259937463226651e-05, | |
| "loss": 1.5148, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.05538676097026448, | |
| "grad_norm": 0.7960265874862671, | |
| "learning_rate": 2.2263363623243054e-05, | |
| "loss": 1.7718, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.05558599392339493, | |
| "grad_norm": 0.8715181350708008, | |
| "learning_rate": 2.192915240480596e-05, | |
| "loss": 1.5031, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.05578522687652538, | |
| "grad_norm": 0.7735579609870911, | |
| "learning_rate": 2.1596762663442218e-05, | |
| "loss": 1.6745, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.055984459829655824, | |
| "grad_norm": 0.8760421276092529, | |
| "learning_rate": 2.1266215967445824e-05, | |
| "loss": 1.8857, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.056183692782786274, | |
| "grad_norm": 0.8865538835525513, | |
| "learning_rate": 2.0937533765518187e-05, | |
| "loss": 1.9028, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.056382925735916724, | |
| "grad_norm": 0.9539476037025452, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 2.0445, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.05658215868904717, | |
| "grad_norm": 0.9166472554206848, | |
| "learning_rate": 2.0285848032369137e-05, | |
| "loss": 1.9541, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.05678139164217762, | |
| "grad_norm": 0.9191790223121643, | |
| "learning_rate": 1.996288678810105e-05, | |
| "loss": 1.6677, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.05698062459530807, | |
| "grad_norm": 0.7826664447784424, | |
| "learning_rate": 1.9641874609064443e-05, | |
| "loss": 1.6405, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.05717985754843851, | |
| "grad_norm": 0.9028159379959106, | |
| "learning_rate": 1.932283232527956e-05, | |
| "loss": 1.7956, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.05737909050156896, | |
| "grad_norm": 0.937842607498169, | |
| "learning_rate": 1.9005780638942982e-05, | |
| "loss": 2.145, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.057578323454699404, | |
| "grad_norm": 0.9173859357833862, | |
| "learning_rate": 1.8690740123084316e-05, | |
| "loss": 2.0756, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.057777556407829854, | |
| "grad_norm": 0.9420987367630005, | |
| "learning_rate": 1.837773122023114e-05, | |
| "loss": 1.9523, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.057976789360960304, | |
| "grad_norm": 0.8245025873184204, | |
| "learning_rate": 1.8066774241082612e-05, | |
| "loss": 1.6465, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.05817602231409075, | |
| "grad_norm": 1.0325926542282104, | |
| "learning_rate": 1.7757889363191483e-05, | |
| "loss": 2.0746, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.0583752552672212, | |
| "grad_norm": 0.9942927956581116, | |
| "learning_rate": 1.745109662965481e-05, | |
| "loss": 2.1365, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.05857448822035165, | |
| "grad_norm": 0.9741763472557068, | |
| "learning_rate": 1.714641594781347e-05, | |
| "loss": 1.8655, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.05877372117348209, | |
| "grad_norm": 0.8840393424034119, | |
| "learning_rate": 1.684386708796025e-05, | |
| "loss": 1.9521, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.05897295412661254, | |
| "grad_norm": 0.9799483418464661, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 1.991, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.05917218707974299, | |
| "grad_norm": 0.9580841660499573, | |
| "learning_rate": 1.62452432224612e-05, | |
| "loss": 1.8491, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.059371420032873434, | |
| "grad_norm": 0.9038205742835999, | |
| "learning_rate": 1.5949207060660138e-05, | |
| "loss": 1.9239, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.059570652986003884, | |
| "grad_norm": 0.9356576800346375, | |
| "learning_rate": 1.5655380406016235e-05, | |
| "loss": 1.7673, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.059769885939134335, | |
| "grad_norm": 1.2466371059417725, | |
| "learning_rate": 1.536378232452003e-05, | |
| "loss": 2.0537, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.059769885939134335, | |
| "eval_loss": 1.7397185564041138, | |
| "eval_runtime": 607.4946, | |
| "eval_samples_per_second": 13.916, | |
| "eval_steps_per_second": 3.48, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.05996911889226478, | |
| "grad_norm": 0.702065646648407, | |
| "learning_rate": 1.5074431737553157e-05, | |
| "loss": 1.4463, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.06016835184539523, | |
| "grad_norm": 0.6998178958892822, | |
| "learning_rate": 1.4787347420660541e-05, | |
| "loss": 1.4507, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.06036758479852568, | |
| "grad_norm": 0.7496803998947144, | |
| "learning_rate": 1.4502548002332088e-05, | |
| "loss": 1.4052, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.06056681775165612, | |
| "grad_norm": 0.7046330571174622, | |
| "learning_rate": 1.422005196279395e-05, | |
| "loss": 1.705, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.06076605070478657, | |
| "grad_norm": 0.6876763105392456, | |
| "learning_rate": 1.3939877632809278e-05, | |
| "loss": 1.3119, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.06096528365791702, | |
| "grad_norm": 0.5933532118797302, | |
| "learning_rate": 1.3662043192488849e-05, | |
| "loss": 1.2781, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.061164516611047465, | |
| "grad_norm": 0.7412620186805725, | |
| "learning_rate": 1.338656667011134e-05, | |
| "loss": 1.5097, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.061363749564177915, | |
| "grad_norm": 0.7208048105239868, | |
| "learning_rate": 1.3113465940953495e-05, | |
| "loss": 1.4867, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.061562982517308365, | |
| "grad_norm": 1.1274152994155884, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 1.7899, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.06176221547043881, | |
| "grad_norm": 0.7465068101882935, | |
| "learning_rate": 1.257446259144494e-05, | |
| "loss": 1.5564, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.06196144842356926, | |
| "grad_norm": 0.7269659042358398, | |
| "learning_rate": 1.2308594946249163e-05, | |
| "loss": 1.4898, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.06216068137669971, | |
| "grad_norm": 0.7326762080192566, | |
| "learning_rate": 1.204517304231343e-05, | |
| "loss": 1.5166, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.06235991432983015, | |
| "grad_norm": 0.6581925749778748, | |
| "learning_rate": 1.178421397270758e-05, | |
| "loss": 1.3075, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.0625591472829606, | |
| "grad_norm": 0.676794707775116, | |
| "learning_rate": 1.1525734670691701e-05, | |
| "loss": 1.5829, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.06275838023609105, | |
| "grad_norm": 0.7718315720558167, | |
| "learning_rate": 1.1269751908617277e-05, | |
| "loss": 1.5633, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.0629576131892215, | |
| "grad_norm": 0.7361578941345215, | |
| "learning_rate": 1.1016282296838887e-05, | |
| "loss": 1.479, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.06315684614235194, | |
| "grad_norm": 0.6821121573448181, | |
| "learning_rate": 1.0765342282636416e-05, | |
| "loss": 1.4351, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.0633560790954824, | |
| "grad_norm": 0.6935531497001648, | |
| "learning_rate": 1.0516948149147754e-05, | |
| "loss": 1.505, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.06355531204861284, | |
| "grad_norm": 0.8006517291069031, | |
| "learning_rate": 1.0271116014312293e-05, | |
| "loss": 1.6302, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.06375454500174328, | |
| "grad_norm": 0.7899420261383057, | |
| "learning_rate": 1.0027861829824952e-05, | |
| "loss": 1.7407, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.06395377795487374, | |
| "grad_norm": 0.8421038389205933, | |
| "learning_rate": 9.787201380101157e-06, | |
| "loss": 1.7289, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.06415301090800418, | |
| "grad_norm": 0.7998735308647156, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 1.7149, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.06435224386113463, | |
| "grad_norm": 0.8017531037330627, | |
| "learning_rate": 9.313723980074018e-06, | |
| "loss": 1.7455, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.06455147681426508, | |
| "grad_norm": 0.7217588424682617, | |
| "learning_rate": 9.080937753040646e-06, | |
| "loss": 1.5329, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.06475070976739553, | |
| "grad_norm": 0.8260882496833801, | |
| "learning_rate": 8.850806705317183e-06, | |
| "loss": 1.6931, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.06494994272052597, | |
| "grad_norm": 0.7838022112846375, | |
| "learning_rate": 8.623345769777514e-06, | |
| "loss": 1.7323, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.06514917567365643, | |
| "grad_norm": 0.7219420075416565, | |
| "learning_rate": 8.398569706035792e-06, | |
| "loss": 1.5608, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.06534840862678687, | |
| "grad_norm": 0.7160537242889404, | |
| "learning_rate": 8.176493099488663e-06, | |
| "loss": 1.6434, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.06554764157991731, | |
| "grad_norm": 0.7419732213020325, | |
| "learning_rate": 7.957130360368898e-06, | |
| "loss": 1.6096, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.06574687453304777, | |
| "grad_norm": 0.7772380709648132, | |
| "learning_rate": 7.740495722810271e-06, | |
| "loss": 1.6037, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.06594610748617821, | |
| "grad_norm": 0.8023072481155396, | |
| "learning_rate": 7.526603243923957e-06, | |
| "loss": 1.6626, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.06614534043930866, | |
| "grad_norm": 0.9004577994346619, | |
| "learning_rate": 7.315466802886401e-06, | |
| "loss": 1.569, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.06634457339243911, | |
| "grad_norm": 0.7904461622238159, | |
| "learning_rate": 7.107100100038671e-06, | |
| "loss": 1.7634, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.06654380634556956, | |
| "grad_norm": 0.8704246282577515, | |
| "learning_rate": 6.901516655997536e-06, | |
| "loss": 1.9085, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.0667430392987, | |
| "grad_norm": 0.830004096031189, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 1.6895, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.06694227225183046, | |
| "grad_norm": 0.8283828496932983, | |
| "learning_rate": 6.498752722928042e-06, | |
| "loss": 1.6123, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.0671415052049609, | |
| "grad_norm": 1.0017095804214478, | |
| "learning_rate": 6.301598368674105e-06, | |
| "loss": 1.9623, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.06734073815809134, | |
| "grad_norm": 0.9769595861434937, | |
| "learning_rate": 6.107279541079769e-06, | |
| "loss": 1.924, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.0675399711112218, | |
| "grad_norm": 0.9661926031112671, | |
| "learning_rate": 5.915808849215304e-06, | |
| "loss": 2.1755, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.06773920406435224, | |
| "grad_norm": 0.8978859782218933, | |
| "learning_rate": 5.727198717339511e-06, | |
| "loss": 1.9377, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.06793843701748269, | |
| "grad_norm": 1.0468230247497559, | |
| "learning_rate": 5.54146138409355e-06, | |
| "loss": 1.9326, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.06813766997061314, | |
| "grad_norm": 1.1197434663772583, | |
| "learning_rate": 5.358608901706802e-06, | |
| "loss": 2.0748, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.06833690292374359, | |
| "grad_norm": 1.0096327066421509, | |
| "learning_rate": 5.178653135214812e-06, | |
| "loss": 2.1963, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.06853613587687403, | |
| "grad_norm": 1.186009168624878, | |
| "learning_rate": 5.001605761689398e-06, | |
| "loss": 1.9852, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.06873536883000449, | |
| "grad_norm": 0.9862110018730164, | |
| "learning_rate": 4.827478269480895e-06, | |
| "loss": 2.0651, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.06893460178313493, | |
| "grad_norm": 0.9604895710945129, | |
| "learning_rate": 4.65628195747273e-06, | |
| "loss": 1.9107, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.06913383473626537, | |
| "grad_norm": 1.15689218044281, | |
| "learning_rate": 4.488027934348271e-06, | |
| "loss": 2.1757, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.06933306768939583, | |
| "grad_norm": 0.9657102823257446, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 2.0478, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.06953230064252627, | |
| "grad_norm": 1.027337908744812, | |
| "learning_rate": 4.16039023417088e-06, | |
| "loss": 1.9578, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.06973153359565672, | |
| "grad_norm": 0.9494063854217529, | |
| "learning_rate": 4.001027817058789e-06, | |
| "loss": 1.8879, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.06973153359565672, | |
| "eval_loss": 1.7323445081710815, | |
| "eval_runtime": 607.3247, | |
| "eval_samples_per_second": 13.92, | |
| "eval_steps_per_second": 3.481, | |
| "step": 350 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.97661039869952e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |