| { | |
| "best_metric": 11.728843688964844, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-400", | |
| "epoch": 0.0657327143502732, | |
| "eval_steps": 50, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.000164331785875683, | |
| "grad_norm": 0.042701028287410736, | |
| "learning_rate": 1e-05, | |
| "loss": 11.7654, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.000164331785875683, | |
| "eval_loss": 11.765250205993652, | |
| "eval_runtime": 48.2849, | |
| "eval_samples_per_second": 212.261, | |
| "eval_steps_per_second": 53.081, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.000328663571751366, | |
| "grad_norm": 0.04443906247615814, | |
| "learning_rate": 2e-05, | |
| "loss": 11.7649, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.000492995357627049, | |
| "grad_norm": 0.05119364336133003, | |
| "learning_rate": 3e-05, | |
| "loss": 11.7635, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.000657327143502732, | |
| "grad_norm": 0.04371592029929161, | |
| "learning_rate": 4e-05, | |
| "loss": 11.7631, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.000821658929378415, | |
| "grad_norm": 0.050174057483673096, | |
| "learning_rate": 5e-05, | |
| "loss": 11.7675, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.000985990715254098, | |
| "grad_norm": 0.05489945784211159, | |
| "learning_rate": 6e-05, | |
| "loss": 11.7681, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.001150322501129781, | |
| "grad_norm": 0.055287834256887436, | |
| "learning_rate": 7e-05, | |
| "loss": 11.7656, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.001314654287005464, | |
| "grad_norm": 0.05079254135489464, | |
| "learning_rate": 8e-05, | |
| "loss": 11.7637, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0014789860728811471, | |
| "grad_norm": 0.038517486304044724, | |
| "learning_rate": 9e-05, | |
| "loss": 11.7672, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.00164331785875683, | |
| "grad_norm": 0.06290938705205917, | |
| "learning_rate": 0.0001, | |
| "loss": 11.7617, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.001807649644632513, | |
| "grad_norm": 0.05406084656715393, | |
| "learning_rate": 9.99983777858264e-05, | |
| "loss": 11.7618, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.001971981430508196, | |
| "grad_norm": 0.07149634510278702, | |
| "learning_rate": 9.999351124856874e-05, | |
| "loss": 11.7615, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.002136313216383879, | |
| "grad_norm": 0.06020309031009674, | |
| "learning_rate": 9.998540070400966e-05, | |
| "loss": 11.7648, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.002300645002259562, | |
| "grad_norm": 0.08084205538034439, | |
| "learning_rate": 9.997404667843075e-05, | |
| "loss": 11.7612, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0024649767881352452, | |
| "grad_norm": 0.06595081835985184, | |
| "learning_rate": 9.995944990857849e-05, | |
| "loss": 11.7615, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.002629308574010928, | |
| "grad_norm": 0.0957484170794487, | |
| "learning_rate": 9.994161134161634e-05, | |
| "loss": 11.7614, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.002793640359886611, | |
| "grad_norm": 0.09352919459342957, | |
| "learning_rate": 9.992053213506334e-05, | |
| "loss": 11.7624, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0029579721457622943, | |
| "grad_norm": 0.1028677374124527, | |
| "learning_rate": 9.989621365671902e-05, | |
| "loss": 11.7618, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.003122303931637977, | |
| "grad_norm": 0.12186416238546371, | |
| "learning_rate": 9.986865748457457e-05, | |
| "loss": 11.7574, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.00328663571751366, | |
| "grad_norm": 0.08856413513422012, | |
| "learning_rate": 9.983786540671051e-05, | |
| "loss": 11.758, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.003450967503389343, | |
| "grad_norm": 0.08863840252161026, | |
| "learning_rate": 9.980383942118066e-05, | |
| "loss": 11.7614, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.003615299289265026, | |
| "grad_norm": 0.13004614412784576, | |
| "learning_rate": 9.976658173588244e-05, | |
| "loss": 11.7599, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.003779631075140709, | |
| "grad_norm": 0.12614966928958893, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 11.76, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.003943962861016392, | |
| "grad_norm": 0.16564443707466125, | |
| "learning_rate": 9.968238114591566e-05, | |
| "loss": 11.7531, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.004108294646892075, | |
| "grad_norm": 0.12352258712053299, | |
| "learning_rate": 9.96354437049027e-05, | |
| "loss": 11.7589, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.004272626432767758, | |
| "grad_norm": 0.19589397311210632, | |
| "learning_rate": 9.95852854910781e-05, | |
| "loss": 11.755, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.004436958218643441, | |
| "grad_norm": 0.1362888664007187, | |
| "learning_rate": 9.953190975913647e-05, | |
| "loss": 11.7526, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.004601290004519124, | |
| "grad_norm": 0.16259880363941193, | |
| "learning_rate": 9.947531997255256e-05, | |
| "loss": 11.7523, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.004765621790394807, | |
| "grad_norm": 0.16003204882144928, | |
| "learning_rate": 9.941551980335652e-05, | |
| "loss": 11.7501, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0049299535762704905, | |
| "grad_norm": 0.15248532593250275, | |
| "learning_rate": 9.935251313189564e-05, | |
| "loss": 11.7525, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.005094285362146173, | |
| "grad_norm": 0.13439106941223145, | |
| "learning_rate": 9.928630404658255e-05, | |
| "loss": 11.7477, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.005258617148021856, | |
| "grad_norm": 0.19816677272319794, | |
| "learning_rate": 9.921689684362989e-05, | |
| "loss": 11.7489, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0054229489338975395, | |
| "grad_norm": 0.14417704939842224, | |
| "learning_rate": 9.914429602677162e-05, | |
| "loss": 11.7435, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.005587280719773222, | |
| "grad_norm": 0.1288655698299408, | |
| "learning_rate": 9.906850630697068e-05, | |
| "loss": 11.7456, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.005751612505648905, | |
| "grad_norm": 0.1479674130678177, | |
| "learning_rate": 9.898953260211338e-05, | |
| "loss": 11.7448, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.0059159442915245886, | |
| "grad_norm": 0.11588233709335327, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 11.7495, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.006080276077400271, | |
| "grad_norm": 0.11137348413467407, | |
| "learning_rate": 9.882205394146361e-05, | |
| "loss": 11.743, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.006244607863275954, | |
| "grad_norm": 0.11102559417486191, | |
| "learning_rate": 9.87335598531214e-05, | |
| "loss": 11.7388, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.006408939649151637, | |
| "grad_norm": 0.10916115343570709, | |
| "learning_rate": 9.864190351391822e-05, | |
| "loss": 11.7415, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.00657327143502732, | |
| "grad_norm": 0.10234220325946808, | |
| "learning_rate": 9.85470908713026e-05, | |
| "loss": 11.7427, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.006737603220903003, | |
| "grad_norm": 0.1283135861158371, | |
| "learning_rate": 9.844912807753104e-05, | |
| "loss": 11.7424, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.006901935006778686, | |
| "grad_norm": 0.11565033346414566, | |
| "learning_rate": 9.834802148926882e-05, | |
| "loss": 11.7331, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.007066266792654369, | |
| "grad_norm": 0.11571874469518661, | |
| "learning_rate": 9.824377766717759e-05, | |
| "loss": 11.7374, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.007230598578530052, | |
| "grad_norm": 0.10002451390028, | |
| "learning_rate": 9.813640337548954e-05, | |
| "loss": 11.7423, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.007394930364405735, | |
| "grad_norm": 0.09244619309902191, | |
| "learning_rate": 9.802590558156862e-05, | |
| "loss": 11.7401, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.007559262150281418, | |
| "grad_norm": 0.08112873882055283, | |
| "learning_rate": 9.791229145545831e-05, | |
| "loss": 11.7388, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.0077235939361571014, | |
| "grad_norm": 0.10474348813295364, | |
| "learning_rate": 9.779556836941645e-05, | |
| "loss": 11.7399, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.007887925722032785, | |
| "grad_norm": 0.11409953236579895, | |
| "learning_rate": 9.767574389743682e-05, | |
| "loss": 11.7385, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.008052257507908467, | |
| "grad_norm": 0.2044820338487625, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 11.7294, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.00821658929378415, | |
| "grad_norm": 0.2515033483505249, | |
| "learning_rate": 9.742682209735727e-05, | |
| "loss": 11.7291, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.00821658929378415, | |
| "eval_loss": 11.74016284942627, | |
| "eval_runtime": 48.5939, | |
| "eval_samples_per_second": 210.911, | |
| "eval_steps_per_second": 52.743, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.008380921079659834, | |
| "grad_norm": 0.10825006663799286, | |
| "learning_rate": 9.729774092143627e-05, | |
| "loss": 11.7462, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.008545252865535516, | |
| "grad_norm": 0.09035345911979675, | |
| "learning_rate": 9.716559066288715e-05, | |
| "loss": 11.7452, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.008709584651411199, | |
| "grad_norm": 0.109395332634449, | |
| "learning_rate": 9.703037989675087e-05, | |
| "loss": 11.7461, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.008873916437286883, | |
| "grad_norm": 0.09128301590681076, | |
| "learning_rate": 9.689211739666023e-05, | |
| "loss": 11.7422, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.009038248223162565, | |
| "grad_norm": 0.09604194760322571, | |
| "learning_rate": 9.675081213427076e-05, | |
| "loss": 11.741, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.009202580009038248, | |
| "grad_norm": 0.09056773781776428, | |
| "learning_rate": 9.66064732786784e-05, | |
| "loss": 11.7421, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.009366911794913932, | |
| "grad_norm": 0.08328656107187271, | |
| "learning_rate": 9.645911019582467e-05, | |
| "loss": 11.7412, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.009531243580789614, | |
| "grad_norm": 0.09039179235696793, | |
| "learning_rate": 9.630873244788883e-05, | |
| "loss": 11.7383, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.009695575366665297, | |
| "grad_norm": 0.07337184250354767, | |
| "learning_rate": 9.615534979266745e-05, | |
| "loss": 11.7419, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.009859907152540981, | |
| "grad_norm": 0.07708930224180222, | |
| "learning_rate": 9.599897218294122e-05, | |
| "loss": 11.7387, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.010024238938416663, | |
| "grad_norm": 0.08723892271518707, | |
| "learning_rate": 9.583960976582913e-05, | |
| "loss": 11.7393, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.010188570724292346, | |
| "grad_norm": 0.08119846135377884, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 11.7412, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.01035290251016803, | |
| "grad_norm": 0.0847805067896843, | |
| "learning_rate": 9.551197206565173e-05, | |
| "loss": 11.743, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.010517234296043712, | |
| "grad_norm": 0.07967144250869751, | |
| "learning_rate": 9.534371804252728e-05, | |
| "loss": 11.7405, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.010681566081919395, | |
| "grad_norm": 0.046992260962724686, | |
| "learning_rate": 9.517252173051911e-05, | |
| "loss": 11.7404, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.010845897867795079, | |
| "grad_norm": 0.054778508841991425, | |
| "learning_rate": 9.49983942383106e-05, | |
| "loss": 11.7398, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.011010229653670761, | |
| "grad_norm": 0.05465163663029671, | |
| "learning_rate": 9.482134686478519e-05, | |
| "loss": 11.7414, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.011174561439546444, | |
| "grad_norm": 0.06919052451848984, | |
| "learning_rate": 9.464139109829321e-05, | |
| "loss": 11.7432, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.011338893225422128, | |
| "grad_norm": 0.04978770390152931, | |
| "learning_rate": 9.445853861590647e-05, | |
| "loss": 11.7406, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.01150322501129781, | |
| "grad_norm": 0.05154619738459587, | |
| "learning_rate": 9.42728012826605e-05, | |
| "loss": 11.7389, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.011667556797173493, | |
| "grad_norm": 0.058466438204050064, | |
| "learning_rate": 9.408419115078471e-05, | |
| "loss": 11.7376, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.011831888583049177, | |
| "grad_norm": 0.07136223465204239, | |
| "learning_rate": 9.389272045892024e-05, | |
| "loss": 11.738, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.01199622036892486, | |
| "grad_norm": 0.0563649944961071, | |
| "learning_rate": 9.36984016313259e-05, | |
| "loss": 11.738, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.012160552154800542, | |
| "grad_norm": 0.062396273016929626, | |
| "learning_rate": 9.350124727707197e-05, | |
| "loss": 11.7354, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.012324883940676224, | |
| "grad_norm": 0.05039061978459358, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 11.7388, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.012489215726551909, | |
| "grad_norm": 0.058604415506124496, | |
| "learning_rate": 9.309848334400246e-05, | |
| "loss": 11.7386, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.012653547512427591, | |
| "grad_norm": 0.08999053388834, | |
| "learning_rate": 9.289289989996133e-05, | |
| "loss": 11.7338, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.012817879298303273, | |
| "grad_norm": 0.05026502162218094, | |
| "learning_rate": 9.268453319711363e-05, | |
| "loss": 11.7371, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.012982211084178958, | |
| "grad_norm": 0.06558049470186234, | |
| "learning_rate": 9.247339675607605e-05, | |
| "loss": 11.7358, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.01314654287005464, | |
| "grad_norm": 0.06828798353672028, | |
| "learning_rate": 9.225950427718975e-05, | |
| "loss": 11.7333, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.013310874655930323, | |
| "grad_norm": 0.05039716139435768, | |
| "learning_rate": 9.204286963963111e-05, | |
| "loss": 11.7367, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.013475206441806007, | |
| "grad_norm": 0.09033340215682983, | |
| "learning_rate": 9.182350690051133e-05, | |
| "loss": 11.737, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.01363953822768169, | |
| "grad_norm": 0.04329407215118408, | |
| "learning_rate": 9.160143029396422e-05, | |
| "loss": 11.7365, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.013803870013557372, | |
| "grad_norm": 0.05856756493449211, | |
| "learning_rate": 9.13766542302225e-05, | |
| "loss": 11.7351, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.013968201799433056, | |
| "grad_norm": 0.06545337289571762, | |
| "learning_rate": 9.114919329468282e-05, | |
| "loss": 11.7321, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.014132533585308738, | |
| "grad_norm": 0.06385994702577591, | |
| "learning_rate": 9.091906224695935e-05, | |
| "loss": 11.734, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.01429686537118442, | |
| "grad_norm": 0.0672338604927063, | |
| "learning_rate": 9.068627601992598e-05, | |
| "loss": 11.7357, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.014461197157060105, | |
| "grad_norm": 0.06891307979822159, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 11.7329, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.014625528942935787, | |
| "grad_norm": 0.05319250375032425, | |
| "learning_rate": 9.021279861989885e-05, | |
| "loss": 11.7367, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.01478986072881147, | |
| "grad_norm": 0.07806988060474396, | |
| "learning_rate": 8.997213817017507e-05, | |
| "loss": 11.7364, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.014954192514687154, | |
| "grad_norm": 0.07189088314771652, | |
| "learning_rate": 8.972888398568772e-05, | |
| "loss": 11.7356, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.015118524300562836, | |
| "grad_norm": 0.07356931269168854, | |
| "learning_rate": 8.948305185085225e-05, | |
| "loss": 11.7331, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.015282856086438519, | |
| "grad_norm": 0.07599375396966934, | |
| "learning_rate": 8.92346577173636e-05, | |
| "loss": 11.7298, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.015447187872314203, | |
| "grad_norm": 0.08366679400205612, | |
| "learning_rate": 8.898371770316111e-05, | |
| "loss": 11.728, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.015611519658189885, | |
| "grad_norm": 0.08361729979515076, | |
| "learning_rate": 8.873024809138272e-05, | |
| "loss": 11.7304, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.01577585144406557, | |
| "grad_norm": 0.07593885809183121, | |
| "learning_rate": 8.847426532930831e-05, | |
| "loss": 11.7272, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.01594018322994125, | |
| "grad_norm": 0.09948297590017319, | |
| "learning_rate": 8.821578602729242e-05, | |
| "loss": 11.7321, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.016104515015816934, | |
| "grad_norm": 0.13416649401187897, | |
| "learning_rate": 8.795482695768658e-05, | |
| "loss": 11.7273, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.01626884680169262, | |
| "grad_norm": 0.12499107420444489, | |
| "learning_rate": 8.769140505375085e-05, | |
| "loss": 11.7205, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.0164331785875683, | |
| "grad_norm": 0.2499997466802597, | |
| "learning_rate": 8.742553740855506e-05, | |
| "loss": 11.7202, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0164331785875683, | |
| "eval_loss": 11.734024047851562, | |
| "eval_runtime": 48.7146, | |
| "eval_samples_per_second": 210.389, | |
| "eval_steps_per_second": 52.613, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.016597510373443983, | |
| "grad_norm": 0.111448734998703, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 11.7424, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.016761842159319668, | |
| "grad_norm": 0.0742286741733551, | |
| "learning_rate": 8.688653405904652e-05, | |
| "loss": 11.7416, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.01692617394519535, | |
| "grad_norm": 0.08018925040960312, | |
| "learning_rate": 8.661343332988869e-05, | |
| "loss": 11.7371, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.017090505731071032, | |
| "grad_norm": 0.08447378873825073, | |
| "learning_rate": 8.633795680751116e-05, | |
| "loss": 11.7374, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.017254837516946717, | |
| "grad_norm": 0.08231153339147568, | |
| "learning_rate": 8.606012236719073e-05, | |
| "loss": 11.7377, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.017419169302822397, | |
| "grad_norm": 0.07651390880346298, | |
| "learning_rate": 8.577994803720606e-05, | |
| "loss": 11.7391, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.01758350108869808, | |
| "grad_norm": 0.08170609176158905, | |
| "learning_rate": 8.549745199766792e-05, | |
| "loss": 11.7371, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.017747832874573766, | |
| "grad_norm": 0.05312446877360344, | |
| "learning_rate": 8.521265257933948e-05, | |
| "loss": 11.7405, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.017912164660449446, | |
| "grad_norm": 0.05158795416355133, | |
| "learning_rate": 8.492556826244687e-05, | |
| "loss": 11.7399, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.01807649644632513, | |
| "grad_norm": 0.060141801834106445, | |
| "learning_rate": 8.463621767547998e-05, | |
| "loss": 11.7336, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.018240828232200815, | |
| "grad_norm": 0.07077991217374802, | |
| "learning_rate": 8.434461959398376e-05, | |
| "loss": 11.7364, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.018405160018076495, | |
| "grad_norm": 0.05313685163855553, | |
| "learning_rate": 8.405079293933986e-05, | |
| "loss": 11.7386, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.01856949180395218, | |
| "grad_norm": 0.056390322744846344, | |
| "learning_rate": 8.375475677753881e-05, | |
| "loss": 11.7361, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.018733823589827864, | |
| "grad_norm": 0.06804954260587692, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 11.7351, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.018898155375703544, | |
| "grad_norm": 0.06049950420856476, | |
| "learning_rate": 8.315613291203976e-05, | |
| "loss": 11.7321, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.01906248716157923, | |
| "grad_norm": 0.062231432646512985, | |
| "learning_rate": 8.285358405218655e-05, | |
| "loss": 11.7345, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.019226818947454913, | |
| "grad_norm": 0.07354050874710083, | |
| "learning_rate": 8.25489033703452e-05, | |
| "loss": 11.7314, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.019391150733330594, | |
| "grad_norm": 0.05630939453840256, | |
| "learning_rate": 8.224211063680853e-05, | |
| "loss": 11.7317, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.019555482519206278, | |
| "grad_norm": 0.06644688546657562, | |
| "learning_rate": 8.19332257589174e-05, | |
| "loss": 11.7328, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.019719814305081962, | |
| "grad_norm": 0.07483275234699249, | |
| "learning_rate": 8.162226877976887e-05, | |
| "loss": 11.7338, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.019884146090957643, | |
| "grad_norm": 0.06993226706981659, | |
| "learning_rate": 8.130925987691569e-05, | |
| "loss": 11.7337, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.020048477876833327, | |
| "grad_norm": 0.06282593309879303, | |
| "learning_rate": 8.099421936105702e-05, | |
| "loss": 11.7343, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.02021280966270901, | |
| "grad_norm": 0.04913927987217903, | |
| "learning_rate": 8.067716767472045e-05, | |
| "loss": 11.7289, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.02037714144858469, | |
| "grad_norm": 0.06600144505500793, | |
| "learning_rate": 8.035812539093557e-05, | |
| "loss": 11.7353, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.020541473234460376, | |
| "grad_norm": 0.05454692989587784, | |
| "learning_rate": 8.003711321189895e-05, | |
| "loss": 11.7329, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.02070580502033606, | |
| "grad_norm": 0.05435941740870476, | |
| "learning_rate": 7.971415196763088e-05, | |
| "loss": 11.7338, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.02087013680621174, | |
| "grad_norm": 0.05516434460878372, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 11.7278, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.021034468592087425, | |
| "grad_norm": 0.05917780473828316, | |
| "learning_rate": 7.906246623448183e-05, | |
| "loss": 11.7353, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.02119880037796311, | |
| "grad_norm": 0.05429854616522789, | |
| "learning_rate": 7.873378403255419e-05, | |
| "loss": 11.7334, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.02136313216383879, | |
| "grad_norm": 0.059085000306367874, | |
| "learning_rate": 7.840323733655778e-05, | |
| "loss": 11.7326, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.021527463949714474, | |
| "grad_norm": 0.06832055002450943, | |
| "learning_rate": 7.807084759519405e-05, | |
| "loss": 11.7345, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.021691795735590158, | |
| "grad_norm": 0.06523392349481583, | |
| "learning_rate": 7.773663637675694e-05, | |
| "loss": 11.7315, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.02185612752146584, | |
| "grad_norm": 0.05600828677415848, | |
| "learning_rate": 7.740062536773352e-05, | |
| "loss": 11.7334, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.022020459307341523, | |
| "grad_norm": 0.05971834808588028, | |
| "learning_rate": 7.706283637139658e-05, | |
| "loss": 11.7314, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.022184791093217207, | |
| "grad_norm": 0.06425528228282928, | |
| "learning_rate": 7.672329130639005e-05, | |
| "loss": 11.7293, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.022349122879092888, | |
| "grad_norm": 0.048994455486536026, | |
| "learning_rate": 7.638201220530665e-05, | |
| "loss": 11.7309, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.022513454664968572, | |
| "grad_norm": 0.05090711638331413, | |
| "learning_rate": 7.603902121325813e-05, | |
| "loss": 11.726, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.022677786450844256, | |
| "grad_norm": 0.0425398163497448, | |
| "learning_rate": 7.569434058643844e-05, | |
| "loss": 11.7301, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.022842118236719937, | |
| "grad_norm": 0.08484908938407898, | |
| "learning_rate": 7.534799269067953e-05, | |
| "loss": 11.7318, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.02300645002259562, | |
| "grad_norm": 0.0685524120926857, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 11.7285, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.023170781808471305, | |
| "grad_norm": 0.05587577447295189, | |
| "learning_rate": 7.465038509514688e-05, | |
| "loss": 11.7249, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.023335113594346986, | |
| "grad_norm": 0.078126922249794, | |
| "learning_rate": 7.42991706621303e-05, | |
| "loss": 11.7242, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.02349944538022267, | |
| "grad_norm": 0.07973043620586395, | |
| "learning_rate": 7.394637949075154e-05, | |
| "loss": 11.7341, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.023663777166098354, | |
| "grad_norm": 0.08124510943889618, | |
| "learning_rate": 7.35920344731241e-05, | |
| "loss": 11.7245, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.023828108951974035, | |
| "grad_norm": 0.06348835676908493, | |
| "learning_rate": 7.323615860218843e-05, | |
| "loss": 11.724, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.02399244073784972, | |
| "grad_norm": 0.10301569104194641, | |
| "learning_rate": 7.287877497021978e-05, | |
| "loss": 11.718, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.024156772523725403, | |
| "grad_norm": 0.08800891041755676, | |
| "learning_rate": 7.251990676732984e-05, | |
| "loss": 11.7249, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.024321104309601084, | |
| "grad_norm": 0.1236993744969368, | |
| "learning_rate": 7.215957727996207e-05, | |
| "loss": 11.7283, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.024485436095476768, | |
| "grad_norm": 0.13688653707504272, | |
| "learning_rate": 7.179780988938051e-05, | |
| "loss": 11.7278, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.02464976788135245, | |
| "grad_norm": 0.2236781120300293, | |
| "learning_rate": 7.143462807015271e-05, | |
| "loss": 11.716, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.02464976788135245, | |
| "eval_loss": 11.731405258178711, | |
| "eval_runtime": 48.5824, | |
| "eval_samples_per_second": 210.961, | |
| "eval_steps_per_second": 52.756, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.024814099667228133, | |
| "grad_norm": 0.097035713493824, | |
| "learning_rate": 7.107005538862646e-05, | |
| "loss": 11.7423, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.024978431453103817, | |
| "grad_norm": 0.08533385396003723, | |
| "learning_rate": 7.07041155014006e-05, | |
| "loss": 11.7377, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.025142763238979498, | |
| "grad_norm": 0.04586503654718399, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 11.7377, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.025307095024855182, | |
| "grad_norm": 0.08356229960918427, | |
| "learning_rate": 6.996822917828477e-05, | |
| "loss": 11.7346, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.025471426810730866, | |
| "grad_norm": 0.06619363278150558, | |
| "learning_rate": 6.959833049300377e-05, | |
| "loss": 11.7376, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.025635758596606547, | |
| "grad_norm": 0.07668127864599228, | |
| "learning_rate": 6.922716010014255e-05, | |
| "loss": 11.7374, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.02580009038248223, | |
| "grad_norm": 0.056997958570718765, | |
| "learning_rate": 6.885474208441603e-05, | |
| "loss": 11.7321, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.025964422168357915, | |
| "grad_norm": 0.07596030831336975, | |
| "learning_rate": 6.848110061149556e-05, | |
| "loss": 11.7388, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.026128753954233596, | |
| "grad_norm": 0.06272446364164352, | |
| "learning_rate": 6.810625992644085e-05, | |
| "loss": 11.7357, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.02629308574010928, | |
| "grad_norm": 0.06435201317071915, | |
| "learning_rate": 6.773024435212678e-05, | |
| "loss": 11.7327, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.026457417525984964, | |
| "grad_norm": 0.0636703222990036, | |
| "learning_rate": 6.735307828766515e-05, | |
| "loss": 11.7331, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.026621749311860645, | |
| "grad_norm": 0.043630119413137436, | |
| "learning_rate": 6.697478620682137e-05, | |
| "loss": 11.7361, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.02678608109773633, | |
| "grad_norm": 0.06924209743738174, | |
| "learning_rate": 6.659539265642643e-05, | |
| "loss": 11.7336, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.026950412883612013, | |
| "grad_norm": 0.04856458678841591, | |
| "learning_rate": 6.621492225478414e-05, | |
| "loss": 11.7375, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.027114744669487694, | |
| "grad_norm": 0.05992913618683815, | |
| "learning_rate": 6.583339969007363e-05, | |
| "loss": 11.732, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.02727907645536338, | |
| "grad_norm": 0.04800641909241676, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 11.7306, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.027443408241239062, | |
| "grad_norm": 0.07342834025621414, | |
| "learning_rate": 6.506729716392481e-05, | |
| "loss": 11.732, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.027607740027114743, | |
| "grad_norm": 0.055936265736818314, | |
| "learning_rate": 6.468276691378155e-05, | |
| "loss": 11.7306, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.027772071812990427, | |
| "grad_norm": 0.06329957395792007, | |
| "learning_rate": 6.429728391993446e-05, | |
| "loss": 11.7333, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.02793640359886611, | |
| "grad_norm": 0.05157172307372093, | |
| "learning_rate": 6.391087319582264e-05, | |
| "loss": 11.7335, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.028100735384741792, | |
| "grad_norm": 0.04706769809126854, | |
| "learning_rate": 6.35235598150842e-05, | |
| "loss": 11.73, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.028265067170617476, | |
| "grad_norm": 0.05760306119918823, | |
| "learning_rate": 6.313536890992935e-05, | |
| "loss": 11.7345, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.02842939895649316, | |
| "grad_norm": 0.049483705312013626, | |
| "learning_rate": 6.274632566950967e-05, | |
| "loss": 11.728, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.02859373074236884, | |
| "grad_norm": 0.047165218740701675, | |
| "learning_rate": 6.235645533828349e-05, | |
| "loss": 11.7328, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.028758062528244525, | |
| "grad_norm": 0.05180773138999939, | |
| "learning_rate": 6.19657832143779e-05, | |
| "loss": 11.7366, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.02892239431412021, | |
| "grad_norm": 0.04299752414226532, | |
| "learning_rate": 6.157433464794716e-05, | |
| "loss": 11.7347, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.02908672609999589, | |
| "grad_norm": 0.05092500150203705, | |
| "learning_rate": 6.118213503952779e-05, | |
| "loss": 11.7291, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.029251057885871574, | |
| "grad_norm": 0.037994395941495895, | |
| "learning_rate": 6.078920983839031e-05, | |
| "loss": 11.7313, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.02941538967174726, | |
| "grad_norm": 0.06346435844898224, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 11.7298, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.02957972145762294, | |
| "grad_norm": 0.06775764375925064, | |
| "learning_rate": 6.0001284688802226e-05, | |
| "loss": 11.7279, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.029744053243498624, | |
| "grad_norm": 0.05034025013446808, | |
| "learning_rate": 5.960633586768543e-05, | |
| "loss": 11.7286, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.029908385029374308, | |
| "grad_norm": 0.05579889938235283, | |
| "learning_rate": 5.921076370520058e-05, | |
| "loss": 11.7258, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.03007271681524999, | |
| "grad_norm": 0.04144003987312317, | |
| "learning_rate": 5.8814593869458455e-05, | |
| "loss": 11.73, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.030237048601125673, | |
| "grad_norm": 0.053334783762693405, | |
| "learning_rate": 5.841785206735192e-05, | |
| "loss": 11.7293, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.030401380387001357, | |
| "grad_norm": 0.05581595003604889, | |
| "learning_rate": 5.8020564042888015e-05, | |
| "loss": 11.7283, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.030565712172877037, | |
| "grad_norm": 0.0592731349170208, | |
| "learning_rate": 5.762275557551727e-05, | |
| "loss": 11.7291, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.03073004395875272, | |
| "grad_norm": 0.05535244196653366, | |
| "learning_rate": 5.7224452478461064e-05, | |
| "loss": 11.7224, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.030894375744628406, | |
| "grad_norm": 0.04812576249241829, | |
| "learning_rate": 5.682568059703659e-05, | |
| "loss": 11.7276, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.031058707530504086, | |
| "grad_norm": 0.045011844485998154, | |
| "learning_rate": 5.642646580697973e-05, | |
| "loss": 11.7279, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.03122303931637977, | |
| "grad_norm": 0.05568164959549904, | |
| "learning_rate": 5.602683401276615e-05, | |
| "loss": 11.7236, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.03138737110225545, | |
| "grad_norm": 0.07187269628047943, | |
| "learning_rate": 5.562681114593028e-05, | |
| "loss": 11.7256, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.03155170288813114, | |
| "grad_norm": 0.06249718740582466, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 11.721, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.03171603467400682, | |
| "grad_norm": 0.06274542212486267, | |
| "learning_rate": 5.482569604572576e-05, | |
| "loss": 11.723, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.0318803664598825, | |
| "grad_norm": 0.05509130284190178, | |
| "learning_rate": 5.442465579556793e-05, | |
| "loss": 11.7282, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.03204469824575819, | |
| "grad_norm": 0.07715385407209396, | |
| "learning_rate": 5.402332843583631e-05, | |
| "loss": 11.7274, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.03220903003163387, | |
| "grad_norm": 0.06635577231645584, | |
| "learning_rate": 5.3621740008088126e-05, | |
| "loss": 11.7224, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.03237336181750955, | |
| "grad_norm": 0.08863984048366547, | |
| "learning_rate": 5.321991657082097e-05, | |
| "loss": 11.7221, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.03253769360338524, | |
| "grad_norm": 0.07797987759113312, | |
| "learning_rate": 5.281788419778187e-05, | |
| "loss": 11.7189, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.03270202538926092, | |
| "grad_norm": 0.10929285734891891, | |
| "learning_rate": 5.2415668976275355e-05, | |
| "loss": 11.7152, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.0328663571751366, | |
| "grad_norm": 0.19686374068260193, | |
| "learning_rate": 5.201329700547076e-05, | |
| "loss": 11.7165, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.0328663571751366, | |
| "eval_loss": 11.729823112487793, | |
| "eval_runtime": 48.4047, | |
| "eval_samples_per_second": 211.735, | |
| "eval_steps_per_second": 52.949, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.033030688961012286, | |
| "grad_norm": 0.08583119511604309, | |
| "learning_rate": 5.161079439470866e-05, | |
| "loss": 11.7399, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.03319502074688797, | |
| "grad_norm": 0.061848536133766174, | |
| "learning_rate": 5.1208187261806615e-05, | |
| "loss": 11.7356, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.03335935253276365, | |
| "grad_norm": 0.05972637981176376, | |
| "learning_rate": 5.080550173136457e-05, | |
| "loss": 11.7329, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.033523684318639335, | |
| "grad_norm": 0.05366925522685051, | |
| "learning_rate": 5.0402763933069496e-05, | |
| "loss": 11.7334, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.033688016104515016, | |
| "grad_norm": 0.05121481791138649, | |
| "learning_rate": 5e-05, | |
| "loss": 11.735, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.0338523478903907, | |
| "grad_norm": 0.04870014265179634, | |
| "learning_rate": 4.9597236066930516e-05, | |
| "loss": 11.7347, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.034016679676266384, | |
| "grad_norm": 0.05482671409845352, | |
| "learning_rate": 4.919449826863544e-05, | |
| "loss": 11.7389, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.034181011462142065, | |
| "grad_norm": 0.0395878441631794, | |
| "learning_rate": 4.87918127381934e-05, | |
| "loss": 11.7373, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.034345343248017746, | |
| "grad_norm": 0.06222689151763916, | |
| "learning_rate": 4.8389205605291365e-05, | |
| "loss": 11.7367, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.03450967503389343, | |
| "grad_norm": 0.04613790661096573, | |
| "learning_rate": 4.798670299452926e-05, | |
| "loss": 11.7344, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.034674006819769114, | |
| "grad_norm": 0.03873327374458313, | |
| "learning_rate": 4.758433102372466e-05, | |
| "loss": 11.7324, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.034838338605644795, | |
| "grad_norm": 0.06393284350633621, | |
| "learning_rate": 4.7182115802218126e-05, | |
| "loss": 11.7293, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.03500267039152048, | |
| "grad_norm": 0.054058074951171875, | |
| "learning_rate": 4.678008342917903e-05, | |
| "loss": 11.7288, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.03516700217739616, | |
| "grad_norm": 0.04814436286687851, | |
| "learning_rate": 4.6378259991911886e-05, | |
| "loss": 11.7329, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.035331333963271844, | |
| "grad_norm": 0.047144170850515366, | |
| "learning_rate": 4.597667156416371e-05, | |
| "loss": 11.7317, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.03549566574914753, | |
| "grad_norm": 0.04114987701177597, | |
| "learning_rate": 4.5575344204432084e-05, | |
| "loss": 11.7334, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.03565999753502321, | |
| "grad_norm": 0.05130726471543312, | |
| "learning_rate": 4.5174303954274244e-05, | |
| "loss": 11.7313, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.03582432932089889, | |
| "grad_norm": 0.04466979578137398, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 11.7299, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.03598866110677458, | |
| "grad_norm": 0.06053079292178154, | |
| "learning_rate": 4.437318885406973e-05, | |
| "loss": 11.7288, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.03615299289265026, | |
| "grad_norm": 0.055950433015823364, | |
| "learning_rate": 4.397316598723385e-05, | |
| "loss": 11.7317, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.03631732467852594, | |
| "grad_norm": 0.05433724448084831, | |
| "learning_rate": 4.3573534193020274e-05, | |
| "loss": 11.7314, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.03648165646440163, | |
| "grad_norm": 0.059027209877967834, | |
| "learning_rate": 4.317431940296343e-05, | |
| "loss": 11.7279, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.03664598825027731, | |
| "grad_norm": 0.04436887800693512, | |
| "learning_rate": 4.277554752153895e-05, | |
| "loss": 11.7289, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.03681032003615299, | |
| "grad_norm": 0.05126571282744408, | |
| "learning_rate": 4.237724442448273e-05, | |
| "loss": 11.7301, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.03697465182202868, | |
| "grad_norm": 0.04554610699415207, | |
| "learning_rate": 4.197943595711198e-05, | |
| "loss": 11.7306, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.03713898360790436, | |
| "grad_norm": 0.06180788576602936, | |
| "learning_rate": 4.1582147932648074e-05, | |
| "loss": 11.725, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.03730331539378004, | |
| "grad_norm": 0.06206175312399864, | |
| "learning_rate": 4.118540613054156e-05, | |
| "loss": 11.7316, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.03746764717965573, | |
| "grad_norm": 0.06412436813116074, | |
| "learning_rate": 4.078923629479943e-05, | |
| "loss": 11.7288, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.03763197896553141, | |
| "grad_norm": 0.059079430997371674, | |
| "learning_rate": 4.039366413231458e-05, | |
| "loss": 11.7303, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.03779631075140709, | |
| "grad_norm": 0.05567527934908867, | |
| "learning_rate": 3.9998715311197785e-05, | |
| "loss": 11.729, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.03796064253728278, | |
| "grad_norm": 0.06493416428565979, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 11.7304, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.03812497432315846, | |
| "grad_norm": 0.048941027373075485, | |
| "learning_rate": 3.92107901616097e-05, | |
| "loss": 11.731, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.03828930610903414, | |
| "grad_norm": 0.041554663330316544, | |
| "learning_rate": 3.8817864960472236e-05, | |
| "loss": 11.7341, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.038453637894909826, | |
| "grad_norm": 0.042613379657268524, | |
| "learning_rate": 3.842566535205286e-05, | |
| "loss": 11.7274, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.038617969680785506, | |
| "grad_norm": 0.051606450229883194, | |
| "learning_rate": 3.803421678562213e-05, | |
| "loss": 11.7251, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.03878230146666119, | |
| "grad_norm": 0.050569821149110794, | |
| "learning_rate": 3.764354466171652e-05, | |
| "loss": 11.7301, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.038946633252536875, | |
| "grad_norm": 0.0642523244023323, | |
| "learning_rate": 3.725367433049033e-05, | |
| "loss": 11.7294, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.039110965038412555, | |
| "grad_norm": 0.05185132101178169, | |
| "learning_rate": 3.6864631090070655e-05, | |
| "loss": 11.7267, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.039275296824288236, | |
| "grad_norm": 0.04526262730360031, | |
| "learning_rate": 3.6476440184915815e-05, | |
| "loss": 11.7257, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.039439628610163924, | |
| "grad_norm": 0.06386663764715195, | |
| "learning_rate": 3.608912680417737e-05, | |
| "loss": 11.7228, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.039603960396039604, | |
| "grad_norm": 0.0651041641831398, | |
| "learning_rate": 3.570271608006555e-05, | |
| "loss": 11.7286, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.039768292181915285, | |
| "grad_norm": 0.05235549435019493, | |
| "learning_rate": 3.531723308621847e-05, | |
| "loss": 11.7248, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.03993262396779097, | |
| "grad_norm": 0.06960996240377426, | |
| "learning_rate": 3.493270283607522e-05, | |
| "loss": 11.7285, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.040096955753666653, | |
| "grad_norm": 0.05223159119486809, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 11.7229, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.040261287539542334, | |
| "grad_norm": 0.0649554505944252, | |
| "learning_rate": 3.4166600309926387e-05, | |
| "loss": 11.7265, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.04042561932541802, | |
| "grad_norm": 0.058237750083208084, | |
| "learning_rate": 3.3785077745215873e-05, | |
| "loss": 11.7186, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.0405899511112937, | |
| "grad_norm": 0.0660976842045784, | |
| "learning_rate": 3.340460734357359e-05, | |
| "loss": 11.7216, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.04075428289716938, | |
| "grad_norm": 0.08093632757663727, | |
| "learning_rate": 3.3025213793178646e-05, | |
| "loss": 11.7193, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.04091861468304507, | |
| "grad_norm": 0.11029274761676788, | |
| "learning_rate": 3.264692171233485e-05, | |
| "loss": 11.7199, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.04108294646892075, | |
| "grad_norm": 0.16421598196029663, | |
| "learning_rate": 3.226975564787322e-05, | |
| "loss": 11.7199, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.04108294646892075, | |
| "eval_loss": 11.729228019714355, | |
| "eval_runtime": 48.4124, | |
| "eval_samples_per_second": 211.702, | |
| "eval_steps_per_second": 52.941, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.04124727825479643, | |
| "grad_norm": 0.08251149952411652, | |
| "learning_rate": 3.189374007355917e-05, | |
| "loss": 11.738, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.04141161004067212, | |
| "grad_norm": 0.07333094626665115, | |
| "learning_rate": 3.151889938850445e-05, | |
| "loss": 11.7412, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.0415759418265478, | |
| "grad_norm": 0.0722256600856781, | |
| "learning_rate": 3.114525791558398e-05, | |
| "loss": 11.7378, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.04174027361242348, | |
| "grad_norm": 0.06710943579673767, | |
| "learning_rate": 3.0772839899857464e-05, | |
| "loss": 11.7372, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.04190460539829917, | |
| "grad_norm": 0.059952352195978165, | |
| "learning_rate": 3.0401669506996256e-05, | |
| "loss": 11.7369, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.04206893718417485, | |
| "grad_norm": 0.051765549927949905, | |
| "learning_rate": 3.003177082171523e-05, | |
| "loss": 11.7356, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.04223326897005053, | |
| "grad_norm": 0.060351401567459106, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 11.7339, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.04239760075592622, | |
| "grad_norm": 0.04224325716495514, | |
| "learning_rate": 2.9295884498599414e-05, | |
| "loss": 11.7363, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.0425619325418019, | |
| "grad_norm": 0.045162707567214966, | |
| "learning_rate": 2.8929944611373554e-05, | |
| "loss": 11.7312, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.04272626432767758, | |
| "grad_norm": 0.050782881677150726, | |
| "learning_rate": 2.8565371929847284e-05, | |
| "loss": 11.7289, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.04289059611355327, | |
| "grad_norm": 0.06226112321019173, | |
| "learning_rate": 2.8202190110619493e-05, | |
| "loss": 11.7317, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.04305492789942895, | |
| "grad_norm": 0.06022179499268532, | |
| "learning_rate": 2.784042272003794e-05, | |
| "loss": 11.7308, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.04321925968530463, | |
| "grad_norm": 0.055496398359537125, | |
| "learning_rate": 2.7480093232670158e-05, | |
| "loss": 11.7287, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.043383591471180316, | |
| "grad_norm": 0.04589381441473961, | |
| "learning_rate": 2.712122502978024e-05, | |
| "loss": 11.733, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.043547923257056, | |
| "grad_norm": 0.05781928077340126, | |
| "learning_rate": 2.6763841397811573e-05, | |
| "loss": 11.7306, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.04371225504293168, | |
| "grad_norm": 0.04345298931002617, | |
| "learning_rate": 2.64079655268759e-05, | |
| "loss": 11.73, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.043876586828807365, | |
| "grad_norm": 0.053832121193408966, | |
| "learning_rate": 2.605362050924848e-05, | |
| "loss": 11.7299, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.044040918614683046, | |
| "grad_norm": 0.061959072947502136, | |
| "learning_rate": 2.57008293378697e-05, | |
| "loss": 11.7283, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.04420525040055873, | |
| "grad_norm": 0.04928471893072128, | |
| "learning_rate": 2.534961490485313e-05, | |
| "loss": 11.7287, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.044369582186434414, | |
| "grad_norm": 0.04960151016712189, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 11.7284, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.044533913972310095, | |
| "grad_norm": 0.05671358108520508, | |
| "learning_rate": 2.4652007309320498e-05, | |
| "loss": 11.7317, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.044698245758185776, | |
| "grad_norm": 0.04368621110916138, | |
| "learning_rate": 2.430565941356157e-05, | |
| "loss": 11.7326, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.04486257754406146, | |
| "grad_norm": 0.046738844364881516, | |
| "learning_rate": 2.3960978786741877e-05, | |
| "loss": 11.7268, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.045026909329937144, | |
| "grad_norm": 0.0573868490755558, | |
| "learning_rate": 2.361798779469336e-05, | |
| "loss": 11.7263, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.045191241115812825, | |
| "grad_norm": 0.047170136123895645, | |
| "learning_rate": 2.3276708693609943e-05, | |
| "loss": 11.7294, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.04535557290168851, | |
| "grad_norm": 0.04488043114542961, | |
| "learning_rate": 2.2937163628603435e-05, | |
| "loss": 11.7306, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.04551990468756419, | |
| "grad_norm": 0.04915056750178337, | |
| "learning_rate": 2.259937463226651e-05, | |
| "loss": 11.732, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.045684236473439874, | |
| "grad_norm": 0.045971404761075974, | |
| "learning_rate": 2.2263363623243054e-05, | |
| "loss": 11.7338, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.04584856825931556, | |
| "grad_norm": 0.05000988021492958, | |
| "learning_rate": 2.192915240480596e-05, | |
| "loss": 11.7311, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.04601290004519124, | |
| "grad_norm": 0.038665771484375, | |
| "learning_rate": 2.1596762663442218e-05, | |
| "loss": 11.7213, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.04617723183106692, | |
| "grad_norm": 0.052090033888816833, | |
| "learning_rate": 2.1266215967445824e-05, | |
| "loss": 11.732, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.04634156361694261, | |
| "grad_norm": 0.04800020158290863, | |
| "learning_rate": 2.0937533765518187e-05, | |
| "loss": 11.7283, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.04650589540281829, | |
| "grad_norm": 0.044077180325984955, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 11.7275, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.04667022718869397, | |
| "grad_norm": 0.06349783390760422, | |
| "learning_rate": 2.0285848032369137e-05, | |
| "loss": 11.7311, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.04683455897456966, | |
| "grad_norm": 0.051240213215351105, | |
| "learning_rate": 1.996288678810105e-05, | |
| "loss": 11.723, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.04699889076044534, | |
| "grad_norm": 0.05683488771319389, | |
| "learning_rate": 1.9641874609064443e-05, | |
| "loss": 11.7307, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.04716322254632102, | |
| "grad_norm": 0.06246829405426979, | |
| "learning_rate": 1.932283232527956e-05, | |
| "loss": 11.7296, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.04732755433219671, | |
| "grad_norm": 0.05560317263007164, | |
| "learning_rate": 1.9005780638942982e-05, | |
| "loss": 11.7269, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.04749188611807239, | |
| "grad_norm": 0.0536818653345108, | |
| "learning_rate": 1.8690740123084316e-05, | |
| "loss": 11.7256, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.04765621790394807, | |
| "grad_norm": 0.05386808514595032, | |
| "learning_rate": 1.837773122023114e-05, | |
| "loss": 11.7246, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.04782054968982376, | |
| "grad_norm": 0.050899688154459, | |
| "learning_rate": 1.8066774241082612e-05, | |
| "loss": 11.7288, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.04798488147569944, | |
| "grad_norm": 0.05641918256878853, | |
| "learning_rate": 1.7757889363191483e-05, | |
| "loss": 11.7243, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.04814921326157512, | |
| "grad_norm": 0.06967247277498245, | |
| "learning_rate": 1.745109662965481e-05, | |
| "loss": 11.7252, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.04831354504745081, | |
| "grad_norm": 0.06244378536939621, | |
| "learning_rate": 1.714641594781347e-05, | |
| "loss": 11.726, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.04847787683332649, | |
| "grad_norm": 0.050528816878795624, | |
| "learning_rate": 1.684386708796025e-05, | |
| "loss": 11.7232, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.04864220861920217, | |
| "grad_norm": 0.051517076790332794, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 11.7235, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.048806540405077856, | |
| "grad_norm": 0.07252132147550583, | |
| "learning_rate": 1.62452432224612e-05, | |
| "loss": 11.7191, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.048970872190953536, | |
| "grad_norm": 0.11321794241666794, | |
| "learning_rate": 1.5949207060660138e-05, | |
| "loss": 11.7252, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.04913520397682922, | |
| "grad_norm": 0.11245723068714142, | |
| "learning_rate": 1.5655380406016235e-05, | |
| "loss": 11.7133, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.0492995357627049, | |
| "grad_norm": 0.2027348428964615, | |
| "learning_rate": 1.536378232452003e-05, | |
| "loss": 11.708, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.0492995357627049, | |
| "eval_loss": 11.728962898254395, | |
| "eval_runtime": 48.8448, | |
| "eval_samples_per_second": 209.828, | |
| "eval_steps_per_second": 52.472, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.049463867548580585, | |
| "grad_norm": 0.07305345684289932, | |
| "learning_rate": 1.5074431737553157e-05, | |
| "loss": 11.7376, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.049628199334456266, | |
| "grad_norm": 0.060168150812387466, | |
| "learning_rate": 1.4787347420660541e-05, | |
| "loss": 11.7355, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.04979253112033195, | |
| "grad_norm": 0.05955510959029198, | |
| "learning_rate": 1.4502548002332088e-05, | |
| "loss": 11.7376, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.049956862906207634, | |
| "grad_norm": 0.051338743418455124, | |
| "learning_rate": 1.422005196279395e-05, | |
| "loss": 11.7334, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.050121194692083315, | |
| "grad_norm": 0.040862489491701126, | |
| "learning_rate": 1.3939877632809278e-05, | |
| "loss": 11.7356, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.050285526477958996, | |
| "grad_norm": 0.05390212684869766, | |
| "learning_rate": 1.3662043192488849e-05, | |
| "loss": 11.7341, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.050449858263834683, | |
| "grad_norm": 0.049856122583150864, | |
| "learning_rate": 1.338656667011134e-05, | |
| "loss": 11.7368, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.050614190049710364, | |
| "grad_norm": 0.05018840357661247, | |
| "learning_rate": 1.3113465940953495e-05, | |
| "loss": 11.7294, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.050778521835586045, | |
| "grad_norm": 0.041660480201244354, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 11.7319, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.05094285362146173, | |
| "grad_norm": 0.034591760486364365, | |
| "learning_rate": 1.257446259144494e-05, | |
| "loss": 11.7332, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.05110718540733741, | |
| "grad_norm": 0.04666655510663986, | |
| "learning_rate": 1.2308594946249163e-05, | |
| "loss": 11.7297, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.051271517193213094, | |
| "grad_norm": 0.050986479967832565, | |
| "learning_rate": 1.204517304231343e-05, | |
| "loss": 11.7347, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.05143584897908878, | |
| "grad_norm": 0.05234253406524658, | |
| "learning_rate": 1.178421397270758e-05, | |
| "loss": 11.7342, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.05160018076496446, | |
| "grad_norm": 0.05376161262392998, | |
| "learning_rate": 1.1525734670691701e-05, | |
| "loss": 11.7288, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.05176451255084014, | |
| "grad_norm": 0.043203748762607574, | |
| "learning_rate": 1.1269751908617277e-05, | |
| "loss": 11.728, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.05192884433671583, | |
| "grad_norm": 0.040804892778396606, | |
| "learning_rate": 1.1016282296838887e-05, | |
| "loss": 11.7346, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.05209317612259151, | |
| "grad_norm": 0.05142979696393013, | |
| "learning_rate": 1.0765342282636416e-05, | |
| "loss": 11.7333, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.05225750790846719, | |
| "grad_norm": 0.059128034859895706, | |
| "learning_rate": 1.0516948149147754e-05, | |
| "loss": 11.7294, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.05242183969434288, | |
| "grad_norm": 0.05118521302938461, | |
| "learning_rate": 1.0271116014312293e-05, | |
| "loss": 11.7303, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.05258617148021856, | |
| "grad_norm": 0.049545012414455414, | |
| "learning_rate": 1.0027861829824952e-05, | |
| "loss": 11.7304, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.05275050326609424, | |
| "grad_norm": 0.044770702719688416, | |
| "learning_rate": 9.787201380101157e-06, | |
| "loss": 11.7317, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.05291483505196993, | |
| "grad_norm": 0.05450345575809479, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 11.7286, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.05307916683784561, | |
| "grad_norm": 0.04360397160053253, | |
| "learning_rate": 9.313723980074018e-06, | |
| "loss": 11.7299, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.05324349862372129, | |
| "grad_norm": 0.06776392459869385, | |
| "learning_rate": 9.080937753040646e-06, | |
| "loss": 11.7297, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.05340783040959698, | |
| "grad_norm": 0.047481000423431396, | |
| "learning_rate": 8.850806705317183e-06, | |
| "loss": 11.7295, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.05357216219547266, | |
| "grad_norm": 0.04181436076760292, | |
| "learning_rate": 8.623345769777514e-06, | |
| "loss": 11.7312, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.05373649398134834, | |
| "grad_norm": 0.056467074900865555, | |
| "learning_rate": 8.398569706035792e-06, | |
| "loss": 11.7301, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.05390082576722403, | |
| "grad_norm": 0.047711823135614395, | |
| "learning_rate": 8.176493099488663e-06, | |
| "loss": 11.733, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.05406515755309971, | |
| "grad_norm": 0.05818518251180649, | |
| "learning_rate": 7.957130360368898e-06, | |
| "loss": 11.7269, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.05422948933897539, | |
| "grad_norm": 0.058100625872612, | |
| "learning_rate": 7.740495722810271e-06, | |
| "loss": 11.7265, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.054393821124851076, | |
| "grad_norm": 0.046868883073329926, | |
| "learning_rate": 7.526603243923957e-06, | |
| "loss": 11.7262, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.05455815291072676, | |
| "grad_norm": 0.04526424780488014, | |
| "learning_rate": 7.315466802886401e-06, | |
| "loss": 11.7275, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.05472248469660244, | |
| "grad_norm": 0.06406047940254211, | |
| "learning_rate": 7.107100100038671e-06, | |
| "loss": 11.7313, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.054886816482478125, | |
| "grad_norm": 0.04994950816035271, | |
| "learning_rate": 6.901516655997536e-06, | |
| "loss": 11.7246, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.055051148268353806, | |
| "grad_norm": 0.054884202778339386, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 11.7282, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.055215480054229486, | |
| "grad_norm": 0.05075803026556969, | |
| "learning_rate": 6.498752722928042e-06, | |
| "loss": 11.7321, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.055379811840105174, | |
| "grad_norm": 0.04193416237831116, | |
| "learning_rate": 6.301598368674105e-06, | |
| "loss": 11.7267, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.055544143625980855, | |
| "grad_norm": 0.04416705667972565, | |
| "learning_rate": 6.107279541079769e-06, | |
| "loss": 11.7181, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.055708475411856535, | |
| "grad_norm": 0.052217379212379456, | |
| "learning_rate": 5.915808849215304e-06, | |
| "loss": 11.7238, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.05587280719773222, | |
| "grad_norm": 0.050664108246564865, | |
| "learning_rate": 5.727198717339511e-06, | |
| "loss": 11.7246, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.056037138983607904, | |
| "grad_norm": 0.04814159497618675, | |
| "learning_rate": 5.54146138409355e-06, | |
| "loss": 11.7278, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.056201470769483584, | |
| "grad_norm": 0.05136456713080406, | |
| "learning_rate": 5.358608901706802e-06, | |
| "loss": 11.7274, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.05636580255535927, | |
| "grad_norm": 0.048106636852025986, | |
| "learning_rate": 5.178653135214812e-06, | |
| "loss": 11.7243, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.05653013434123495, | |
| "grad_norm": 0.058061935007572174, | |
| "learning_rate": 5.001605761689398e-06, | |
| "loss": 11.7241, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.05669446612711063, | |
| "grad_norm": 0.06779101490974426, | |
| "learning_rate": 4.827478269480895e-06, | |
| "loss": 11.7196, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.05685879791298632, | |
| "grad_norm": 0.05850118398666382, | |
| "learning_rate": 4.65628195747273e-06, | |
| "loss": 11.7234, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.057023129698862, | |
| "grad_norm": 0.08406161516904831, | |
| "learning_rate": 4.488027934348271e-06, | |
| "loss": 11.7165, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.05718746148473768, | |
| "grad_norm": 0.08055326342582703, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 11.7168, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.05735179327061337, | |
| "grad_norm": 0.11211990565061569, | |
| "learning_rate": 4.16039023417088e-06, | |
| "loss": 11.7213, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.05751612505648905, | |
| "grad_norm": 0.202422633767128, | |
| "learning_rate": 4.001027817058789e-06, | |
| "loss": 11.705, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.05751612505648905, | |
| "eval_loss": 11.728860855102539, | |
| "eval_runtime": 48.8743, | |
| "eval_samples_per_second": 209.701, | |
| "eval_steps_per_second": 52.441, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.05768045684236473, | |
| "grad_norm": 0.08208485692739487, | |
| "learning_rate": 3.844650207332562e-06, | |
| "loss": 11.7381, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.05784478862824042, | |
| "grad_norm": 0.07359649240970612, | |
| "learning_rate": 3.691267552111183e-06, | |
| "loss": 11.7384, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.0580091204141161, | |
| "grad_norm": 0.06210344284772873, | |
| "learning_rate": 3.54088980417534e-06, | |
| "loss": 11.7368, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.05817345219999178, | |
| "grad_norm": 0.055169302970170975, | |
| "learning_rate": 3.393526721321616e-06, | |
| "loss": 11.7347, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.05833778398586747, | |
| "grad_norm": 0.05137772858142853, | |
| "learning_rate": 3.249187865729264e-06, | |
| "loss": 11.7372, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.05850211577174315, | |
| "grad_norm": 0.05137026309967041, | |
| "learning_rate": 3.1078826033397843e-06, | |
| "loss": 11.7329, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.05866644755761883, | |
| "grad_norm": 0.04226312413811684, | |
| "learning_rate": 2.9696201032491434e-06, | |
| "loss": 11.7324, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.05883077934349452, | |
| "grad_norm": 0.03976316750049591, | |
| "learning_rate": 2.8344093371128424e-06, | |
| "loss": 11.7344, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.0589951111293702, | |
| "grad_norm": 0.04595496505498886, | |
| "learning_rate": 2.70225907856374e-06, | |
| "loss": 11.7316, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.05915944291524588, | |
| "grad_norm": 0.051731742918491364, | |
| "learning_rate": 2.573177902642726e-06, | |
| "loss": 11.7312, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.059323774701121566, | |
| "grad_norm": 0.04073851928114891, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 11.7325, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.05948810648699725, | |
| "grad_norm": 0.038902051746845245, | |
| "learning_rate": 2.324256102563188e-06, | |
| "loss": 11.7311, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.05965243827287293, | |
| "grad_norm": 0.0553940124809742, | |
| "learning_rate": 2.204431630583548e-06, | |
| "loss": 11.7336, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.059816770058748615, | |
| "grad_norm": 0.05747877433896065, | |
| "learning_rate": 2.087708544541689e-06, | |
| "loss": 11.7305, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.059981101844624296, | |
| "grad_norm": 0.05037953704595566, | |
| "learning_rate": 1.974094418431388e-06, | |
| "loss": 11.7329, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.06014543363049998, | |
| "grad_norm": 0.047859761863946915, | |
| "learning_rate": 1.8635966245104664e-06, | |
| "loss": 11.7308, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.060309765416375664, | |
| "grad_norm": 0.04152829200029373, | |
| "learning_rate": 1.7562223328224325e-06, | |
| "loss": 11.7284, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.060474097202251345, | |
| "grad_norm": 0.042301371693611145, | |
| "learning_rate": 1.6519785107311891e-06, | |
| "loss": 11.732, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.060638428988127026, | |
| "grad_norm": 0.057195208966732025, | |
| "learning_rate": 1.5508719224689717e-06, | |
| "loss": 11.7318, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.06080276077400271, | |
| "grad_norm": 0.052724309265613556, | |
| "learning_rate": 1.4529091286973995e-06, | |
| "loss": 11.7308, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.060967092559878394, | |
| "grad_norm": 0.04090241715312004, | |
| "learning_rate": 1.358096486081778e-06, | |
| "loss": 11.733, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.061131424345754075, | |
| "grad_norm": 0.043258052319288254, | |
| "learning_rate": 1.2664401468786114e-06, | |
| "loss": 11.731, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.06129575613162976, | |
| "grad_norm": 0.04883791506290436, | |
| "learning_rate": 1.1779460585363944e-06, | |
| "loss": 11.7319, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.06146008791750544, | |
| "grad_norm": 0.05210534855723381, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 11.7308, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.061624419703381124, | |
| "grad_norm": 0.05686335638165474, | |
| "learning_rate": 1.0104673978866164e-06, | |
| "loss": 11.7305, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.06178875148925681, | |
| "grad_norm": 0.08011097460985184, | |
| "learning_rate": 9.314936930293283e-07, | |
| "loss": 11.7287, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.06195308327513249, | |
| "grad_norm": 0.059680283069610596, | |
| "learning_rate": 8.557039732283944e-07, | |
| "loss": 11.7319, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.06211741506100817, | |
| "grad_norm": 0.07159244269132614, | |
| "learning_rate": 7.83103156370113e-07, | |
| "loss": 11.7292, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.06228174684688386, | |
| "grad_norm": 0.04805372655391693, | |
| "learning_rate": 7.136959534174592e-07, | |
| "loss": 11.7277, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.06244607863275954, | |
| "grad_norm": 0.03767850250005722, | |
| "learning_rate": 6.474868681043578e-07, | |
| "loss": 11.7282, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.06261041041863523, | |
| "grad_norm": 0.04085582122206688, | |
| "learning_rate": 5.844801966434832e-07, | |
| "loss": 11.7256, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.0627747422045109, | |
| "grad_norm": 0.04595988988876343, | |
| "learning_rate": 5.246800274474439e-07, | |
| "loss": 11.7294, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.06293907399038659, | |
| "grad_norm": 0.040494050830602646, | |
| "learning_rate": 4.680902408635335e-07, | |
| "loss": 11.7289, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.06310340577626228, | |
| "grad_norm": 0.04093782976269722, | |
| "learning_rate": 4.1471450892189846e-07, | |
| "loss": 11.7281, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.06326773756213795, | |
| "grad_norm": 0.05765712633728981, | |
| "learning_rate": 3.6455629509730136e-07, | |
| "loss": 11.7256, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.06343206934801364, | |
| "grad_norm": 0.045128606259822845, | |
| "learning_rate": 3.1761885408435054e-07, | |
| "loss": 11.7282, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.06359640113388933, | |
| "grad_norm": 0.05226225033402443, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 11.7289, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.063760732919765, | |
| "grad_norm": 0.046483978629112244, | |
| "learning_rate": 2.334182641175686e-07, | |
| "loss": 11.7242, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.06392506470564069, | |
| "grad_norm": 0.062402740120887756, | |
| "learning_rate": 1.9616057881935436e-07, | |
| "loss": 11.7259, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.06408939649151638, | |
| "grad_norm": 0.04270698502659798, | |
| "learning_rate": 1.6213459328950352e-07, | |
| "loss": 11.7308, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.06425372827739205, | |
| "grad_norm": 0.06238847225904465, | |
| "learning_rate": 1.3134251542544774e-07, | |
| "loss": 11.7215, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.06441806006326774, | |
| "grad_norm": 0.04868527129292488, | |
| "learning_rate": 1.0378634328099269e-07, | |
| "loss": 11.7282, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.06458239184914343, | |
| "grad_norm": 0.05947161093354225, | |
| "learning_rate": 7.946786493666647e-08, | |
| "loss": 11.7183, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.0647467236350191, | |
| "grad_norm": 0.04062836617231369, | |
| "learning_rate": 5.838865838366792e-08, | |
| "loss": 11.728, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.06491105542089479, | |
| "grad_norm": 0.06852062791585922, | |
| "learning_rate": 4.055009142152067e-08, | |
| "loss": 11.7225, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.06507538720677047, | |
| "grad_norm": 0.07569117099046707, | |
| "learning_rate": 2.595332156925534e-08, | |
| "loss": 11.723, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.06523971899264615, | |
| "grad_norm": 0.08385792374610901, | |
| "learning_rate": 1.4599295990352924e-08, | |
| "loss": 11.7212, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.06540405077852184, | |
| "grad_norm": 0.08387003093957901, | |
| "learning_rate": 6.488751431266149e-09, | |
| "loss": 11.7231, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.06556838256439752, | |
| "grad_norm": 0.10819617658853531, | |
| "learning_rate": 1.622214173602199e-09, | |
| "loss": 11.7179, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.0657327143502732, | |
| "grad_norm": 0.1437562257051468, | |
| "learning_rate": 0.0, | |
| "loss": 11.7059, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.0657327143502732, | |
| "eval_loss": 11.728843688964844, | |
| "eval_runtime": 48.2858, | |
| "eval_samples_per_second": 212.257, | |
| "eval_steps_per_second": 53.08, | |
| "step": 400 | |
| } | |
| ], | |
| "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": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 165738341990400.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |