| { |
| "best_metric": 0.6484375, |
| "best_model_checkpoint": "./model/pythia/checkpoint-500", |
| "epoch": 10.0, |
| "eval_steps": 500, |
| "global_step": 31340, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0, |
| "grad_norm": 8.522640369319012, |
| "learning_rate": 3.362938826936834e-06, |
| "loss": 0.6119, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.01, |
| "grad_norm": 7.983384042896044, |
| "learning_rate": 4.375284287427863e-06, |
| "loss": 0.5694, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.01, |
| "grad_norm": 10.055095720636968, |
| "learning_rate": 4.967468419590406e-06, |
| "loss": 0.6249, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.01, |
| "grad_norm": 7.53553289299737, |
| "learning_rate": 5.387629747918892e-06, |
| "loss": 0.5568, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.02, |
| "grad_norm": 8.118836397881559, |
| "learning_rate": 5.7135321933826385e-06, |
| "loss": 0.5936, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.02, |
| "grad_norm": 7.6280603989768885, |
| "learning_rate": 5.979813880081434e-06, |
| "loss": 0.5694, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.02, |
| "grad_norm": 11.42073272545716, |
| "learning_rate": 6.204951838268996e-06, |
| "loss": 0.6012, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.03, |
| "grad_norm": 9.806669215814228, |
| "learning_rate": 6.399975208409921e-06, |
| "loss": 0.6094, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.03, |
| "grad_norm": 6.849330072987465, |
| "learning_rate": 6.571998012243978e-06, |
| "loss": 0.5436, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.03, |
| "grad_norm": 9.732692909945165, |
| "learning_rate": 6.725877653873668e-06, |
| "loss": 0.5344, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.04, |
| "grad_norm": 10.206521092661369, |
| "learning_rate": 6.8650787219434345e-06, |
| "loss": 0.5941, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.04, |
| "grad_norm": 8.625267135242646, |
| "learning_rate": 6.992159340572463e-06, |
| "loss": 0.6011, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.04, |
| "grad_norm": 7.280813572455723, |
| "learning_rate": 7.109062177417672e-06, |
| "loss": 0.5653, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.04, |
| "grad_norm": 7.812786127572885, |
| "learning_rate": 7.2172972987600245e-06, |
| "loss": 0.6978, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.05, |
| "grad_norm": 7.558609395171147, |
| "learning_rate": 7.31806178603621e-06, |
| "loss": 0.6191, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.05, |
| "grad_norm": 10.192926564688133, |
| "learning_rate": 7.412320668900949e-06, |
| "loss": 0.6166, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.05, |
| "grad_norm": 7.912136133394027, |
| "learning_rate": 7.50086327920238e-06, |
| "loss": 0.5991, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.06, |
| "grad_norm": 8.674835414589422, |
| "learning_rate": 7.584343472735007e-06, |
| "loss": 0.6216, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.06, |
| "grad_norm": 8.796134741416832, |
| "learning_rate": 7.663308961666392e-06, |
| "loss": 0.6302, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.06, |
| "grad_norm": 7.660618870687511, |
| "learning_rate": 7.738223114364696e-06, |
| "loss": 0.5724, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.07, |
| "grad_norm": 13.562169787346726, |
| "learning_rate": 7.809481430922567e-06, |
| "loss": 0.6123, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.07, |
| "grad_norm": 7.834728630736687, |
| "learning_rate": 7.877424182434465e-06, |
| "loss": 0.5944, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.07, |
| "grad_norm": 10.057517008509846, |
| "learning_rate": 7.942346238401071e-06, |
| "loss": 0.5877, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.08, |
| "grad_norm": 7.988689055910942, |
| "learning_rate": 8.004504801063494e-06, |
| "loss": 0.5889, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.08, |
| "grad_norm": 8.739957005026229, |
| "learning_rate": 8.064125559828444e-06, |
| "loss": 0.5899, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.08, |
| "grad_norm": 11.022972048443274, |
| "learning_rate": 8.1214076379087e-06, |
| "loss": 0.5856, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.09, |
| "grad_norm": 8.477231216422004, |
| "learning_rate": 8.17652760489755e-06, |
| "loss": 0.5563, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.09, |
| "grad_norm": 9.664780487551623, |
| "learning_rate": 8.229642759251054e-06, |
| "loss": 0.5973, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.09, |
| "grad_norm": 8.736563693238642, |
| "learning_rate": 8.280893834505923e-06, |
| "loss": 0.5888, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.1, |
| "grad_norm": 11.822335160304759, |
| "learning_rate": 8.330407246527238e-06, |
| "loss": 0.5849, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.1, |
| "grad_norm": 9.60260575216151, |
| "learning_rate": 8.378296972138392e-06, |
| "loss": 0.6269, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.1, |
| "grad_norm": 10.65816500731914, |
| "learning_rate": 8.424666129391978e-06, |
| "loss": 0.6654, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.11, |
| "grad_norm": 8.02095339466419, |
| "learning_rate": 8.469608314597006e-06, |
| "loss": 0.586, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.11, |
| "grad_norm": 7.846919540224174, |
| "learning_rate": 8.513208739693409e-06, |
| "loss": 0.6049, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.11, |
| "grad_norm": 8.94391535213588, |
| "learning_rate": 8.5555452047148e-06, |
| "loss": 0.6393, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.11, |
| "grad_norm": 9.815570715286853, |
| "learning_rate": 8.596688933226036e-06, |
| "loss": 0.5734, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.12, |
| "grad_norm": 9.360835246452663, |
| "learning_rate": 8.636705293271014e-06, |
| "loss": 0.6971, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.12, |
| "grad_norm": 9.263523011335534, |
| "learning_rate": 8.675654422157421e-06, |
| "loss": 0.576, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.12, |
| "grad_norm": 9.802057431734397, |
| "learning_rate": 8.713591770071243e-06, |
| "loss": 0.5941, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.13, |
| "grad_norm": 9.88665199935487, |
| "learning_rate": 8.750568574855726e-06, |
| "loss": 0.713, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.13, |
| "grad_norm": 9.65315152427361, |
| "learning_rate": 8.786632278156564e-06, |
| "loss": 0.6558, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.13, |
| "grad_norm": 8.88587478822165, |
| "learning_rate": 8.821826891413596e-06, |
| "loss": 0.617, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.14, |
| "grad_norm": 8.794520440746888, |
| "learning_rate": 8.856193318781969e-06, |
| "loss": 0.5749, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.14, |
| "grad_norm": 9.995773185070918, |
| "learning_rate": 8.889769642925493e-06, |
| "loss": 0.5345, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.14, |
| "grad_norm": 8.144031478221441, |
| "learning_rate": 8.922591378689783e-06, |
| "loss": 0.6443, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.15, |
| "grad_norm": 10.407281864545876, |
| "learning_rate": 8.9546916988921e-06, |
| "loss": 0.664, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.15, |
| "grad_norm": 7.917226778260077, |
| "learning_rate": 8.986101635826769e-06, |
| "loss": 0.5821, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.15, |
| "grad_norm": 9.393052280933166, |
| "learning_rate": 9.016850261554521e-06, |
| "loss": 0.5811, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.16, |
| "grad_norm": 12.426719402261366, |
| "learning_rate": 9.046964849601158e-06, |
| "loss": 0.6574, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.16, |
| "grad_norm": 7.861789582729573, |
| "learning_rate": 9.076471020319471e-06, |
| "loss": 0.5895, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_loss": 0.6484375, |
| "eval_runtime": 55.5873, |
| "eval_samples_per_second": 45.586, |
| "eval_steps_per_second": 2.86, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.16, |
| "grad_norm": 11.016033971242392, |
| "learning_rate": 9.10539287185595e-06, |
| "loss": 0.6262, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.17, |
| "grad_norm": 10.548365418418546, |
| "learning_rate": 9.13375309839973e-06, |
| "loss": 0.6008, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.17, |
| "grad_norm": 9.484333170212535, |
| "learning_rate": 9.1615730971676e-06, |
| "loss": 0.6789, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.17, |
| "grad_norm": 9.064504677926552, |
| "learning_rate": 9.18887306538858e-06, |
| "loss": 0.6376, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.18, |
| "grad_norm": 8.319771017444822, |
| "learning_rate": 9.215672088389239e-06, |
| "loss": 0.6686, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.18, |
| "grad_norm": 8.210879426382405, |
| "learning_rate": 9.241988219742083e-06, |
| "loss": 0.6765, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.18, |
| "grad_norm": 7.126705300226697, |
| "learning_rate": 9.267838554319965e-06, |
| "loss": 0.6247, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.19, |
| "grad_norm": 8.319283607158019, |
| "learning_rate": 9.293239294996953e-06, |
| "loss": 0.6771, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.19, |
| "grad_norm": 9.182003115517817, |
| "learning_rate": 9.318205813647398e-06, |
| "loss": 0.6074, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.19, |
| "grad_norm": 6.843244978155808, |
| "learning_rate": 9.342752707018269e-06, |
| "loss": 0.6308, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.19, |
| "grad_norm": 9.805132973570247, |
| "learning_rate": 9.366893847983273e-06, |
| "loss": 0.6909, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.2, |
| "grad_norm": 8.4618567533583, |
| "learning_rate": 9.390642432629423e-06, |
| "loss": 0.6282, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.2, |
| "grad_norm": 7.44587556485106, |
| "learning_rate": 9.414011023576139e-06, |
| "loss": 0.6961, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.2, |
| "grad_norm": 9.375859335615589, |
| "learning_rate": 9.437011589883009e-06, |
| "loss": 0.6884, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.21, |
| "grad_norm": 8.949246456880942, |
| "learning_rate": 9.459655543863476e-06, |
| "loss": 0.657, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.21, |
| "grad_norm": 9.88604659286654, |
| "learning_rate": 9.481953775088035e-06, |
| "loss": 0.6043, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.21, |
| "grad_norm": 7.62429130887891, |
| "learning_rate": 9.503916681830461e-06, |
| "loss": 0.686, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.22, |
| "grad_norm": 7.4562720606206545, |
| "learning_rate": 9.525554200184436e-06, |
| "loss": 0.6364, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.22, |
| "grad_norm": 7.557225681667983, |
| "learning_rate": 9.546875831054641e-06, |
| "loss": 0.6558, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.22, |
| "grad_norm": 9.76717761095224, |
| "learning_rate": 9.567890665205829e-06, |
| "loss": 0.643, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.23, |
| "grad_norm": 10.184604724822604, |
| "learning_rate": 9.588607406535182e-06, |
| "loss": 0.6346, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.23, |
| "grad_norm": 10.172235391681099, |
| "learning_rate": 9.609034393717065e-06, |
| "loss": 0.6084, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.23, |
| "grad_norm": 10.560056866579394, |
| "learning_rate": 9.629179620354906e-06, |
| "loss": 0.6348, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.24, |
| "grad_norm": 14.42320752832994, |
| "learning_rate": 9.649050753762043e-06, |
| "loss": 0.6698, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.24, |
| "grad_norm": 9.870287966097615, |
| "learning_rate": 9.668655152482015e-06, |
| "loss": 0.5985, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.24, |
| "grad_norm": 10.010728040623235, |
| "learning_rate": 9.687999882648452e-06, |
| "loss": 0.6298, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.25, |
| "grad_norm": 8.290163245716275, |
| "learning_rate": 9.707091733275597e-06, |
| "loss": 0.6169, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.25, |
| "grad_norm": 10.722770804260582, |
| "learning_rate": 9.725937230562272e-06, |
| "loss": 0.6366, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.25, |
| "grad_norm": 7.795721974096569, |
| "learning_rate": 9.744542651284668e-06, |
| "loss": 0.6858, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.26, |
| "grad_norm": 7.892398325253414, |
| "learning_rate": 9.762914035346754e-06, |
| "loss": 0.6279, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.26, |
| "grad_norm": 9.90915384662357, |
| "learning_rate": 9.78105719755112e-06, |
| "loss": 0.6497, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.26, |
| "grad_norm": 10.33199875940071, |
| "learning_rate": 9.798977738647593e-06, |
| "loss": 0.6161, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.26, |
| "grad_norm": 11.389416624004772, |
| "learning_rate": 9.816681055712203e-06, |
| "loss": 0.7049, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.27, |
| "grad_norm": 13.612260094412305, |
| "learning_rate": 9.834172351904627e-06, |
| "loss": 0.6611, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.27, |
| "grad_norm": 9.370678935822191, |
| "learning_rate": 9.851456645648184e-06, |
| "loss": 0.7209, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.27, |
| "grad_norm": 9.351381381391434, |
| "learning_rate": 9.868538779273e-06, |
| "loss": 0.6241, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.28, |
| "grad_norm": 9.346804524637141, |
| "learning_rate": 9.885423427159494e-06, |
| "loss": 0.6687, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.28, |
| "grad_norm": 11.280557796132891, |
| "learning_rate": 9.902115103416522e-06, |
| "loss": 0.6979, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.28, |
| "grad_norm": 7.306238607458104, |
| "learning_rate": 9.918618169125664e-06, |
| "loss": 0.6756, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.29, |
| "grad_norm": 7.435475819285521, |
| "learning_rate": 9.934936839180812e-06, |
| "loss": 0.6163, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.29, |
| "grad_norm": 10.331830028168147, |
| "learning_rate": 9.951075188749834e-06, |
| "loss": 0.63, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.29, |
| "grad_norm": 8.738029356980729, |
| "learning_rate": 9.96703715938313e-06, |
| "loss": 0.636, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.3, |
| "grad_norm": 9.286808905168302, |
| "learning_rate": 9.982826564791964e-06, |
| "loss": 0.6711, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.3, |
| "grad_norm": 10.751453390123588, |
| "learning_rate": 9.998447096317798e-06, |
| "loss": 0.6858, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.3, |
| "grad_norm": 9.095874027074377, |
| "learning_rate": 1e-05, |
| "loss": 0.6581, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.31, |
| "grad_norm": 6.055216710743101, |
| "learning_rate": 1e-05, |
| "loss": 0.6646, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.31, |
| "grad_norm": 9.795813495373052, |
| "learning_rate": 1e-05, |
| "loss": 0.6813, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.31, |
| "grad_norm": 9.8874865429349, |
| "learning_rate": 1e-05, |
| "loss": 0.7049, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.32, |
| "grad_norm": 9.23744586933882, |
| "learning_rate": 1e-05, |
| "loss": 0.6457, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.32, |
| "grad_norm": 5.095406105096481, |
| "learning_rate": 1e-05, |
| "loss": 0.6887, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_loss": 0.68994140625, |
| "eval_runtime": 54.8677, |
| "eval_samples_per_second": 46.184, |
| "eval_steps_per_second": 2.898, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.32, |
| "grad_norm": 9.311313630387364, |
| "learning_rate": 1e-05, |
| "loss": 0.7102, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.33, |
| "grad_norm": 12.312312245752787, |
| "learning_rate": 1e-05, |
| "loss": 0.6471, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.33, |
| "grad_norm": 9.734376959394867, |
| "learning_rate": 1e-05, |
| "loss": 0.6561, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.33, |
| "grad_norm": 11.807948290124989, |
| "learning_rate": 1e-05, |
| "loss": 0.6743, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.34, |
| "grad_norm": 7.6010719446833255, |
| "learning_rate": 1e-05, |
| "loss": 0.6501, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.34, |
| "grad_norm": 8.208591463577534, |
| "learning_rate": 1e-05, |
| "loss": 0.7381, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.34, |
| "grad_norm": 7.319791891339502, |
| "learning_rate": 1e-05, |
| "loss": 0.6332, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.34, |
| "grad_norm": 13.476409645872279, |
| "learning_rate": 1e-05, |
| "loss": 0.6588, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.35, |
| "grad_norm": 8.615353704753382, |
| "learning_rate": 1e-05, |
| "loss": 0.6379, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.35, |
| "grad_norm": 8.655988668635315, |
| "learning_rate": 1e-05, |
| "loss": 0.6632, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.35, |
| "grad_norm": 8.142050825079858, |
| "learning_rate": 1e-05, |
| "loss": 0.6962, |
| "step": 1110 |
| }, |
| { |
| "epoch": 0.36, |
| "grad_norm": 7.202374532997727, |
| "learning_rate": 1e-05, |
| "loss": 0.6677, |
| "step": 1120 |
| }, |
| { |
| "epoch": 0.36, |
| "grad_norm": 9.053955259164965, |
| "learning_rate": 1e-05, |
| "loss": 0.6208, |
| "step": 1130 |
| }, |
| { |
| "epoch": 0.36, |
| "grad_norm": 6.652696117447686, |
| "learning_rate": 1e-05, |
| "loss": 0.6093, |
| "step": 1140 |
| }, |
| { |
| "epoch": 0.37, |
| "grad_norm": 7.7282419259881845, |
| "learning_rate": 1e-05, |
| "loss": 0.6803, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.37, |
| "grad_norm": 7.029215200362542, |
| "learning_rate": 1e-05, |
| "loss": 0.6769, |
| "step": 1160 |
| }, |
| { |
| "epoch": 0.37, |
| "grad_norm": 9.107216005004846, |
| "learning_rate": 1e-05, |
| "loss": 0.677, |
| "step": 1170 |
| }, |
| { |
| "epoch": 0.38, |
| "grad_norm": 9.204699551612167, |
| "learning_rate": 1e-05, |
| "loss": 0.6461, |
| "step": 1180 |
| }, |
| { |
| "epoch": 0.38, |
| "grad_norm": 9.852893663782467, |
| "learning_rate": 1e-05, |
| "loss": 0.6962, |
| "step": 1190 |
| }, |
| { |
| "epoch": 0.38, |
| "grad_norm": 8.988809622263283, |
| "learning_rate": 1e-05, |
| "loss": 0.6701, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.39, |
| "grad_norm": 10.705998345494477, |
| "learning_rate": 1e-05, |
| "loss": 0.6802, |
| "step": 1210 |
| }, |
| { |
| "epoch": 0.39, |
| "grad_norm": 9.840482623662396, |
| "learning_rate": 1e-05, |
| "loss": 0.6993, |
| "step": 1220 |
| }, |
| { |
| "epoch": 0.39, |
| "grad_norm": 7.880177566928814, |
| "learning_rate": 1e-05, |
| "loss": 0.6837, |
| "step": 1230 |
| }, |
| { |
| "epoch": 0.4, |
| "grad_norm": 7.706460536237244, |
| "learning_rate": 1e-05, |
| "loss": 0.6909, |
| "step": 1240 |
| }, |
| { |
| "epoch": 0.4, |
| "grad_norm": 10.028489918546397, |
| "learning_rate": 1e-05, |
| "loss": 0.697, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.4, |
| "grad_norm": 9.191168020446483, |
| "learning_rate": 1e-05, |
| "loss": 0.713, |
| "step": 1260 |
| }, |
| { |
| "epoch": 0.41, |
| "grad_norm": 12.226322881038527, |
| "learning_rate": 1e-05, |
| "loss": 0.684, |
| "step": 1270 |
| }, |
| { |
| "epoch": 0.41, |
| "grad_norm": 10.89354623942131, |
| "learning_rate": 1e-05, |
| "loss": 0.6626, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.41, |
| "grad_norm": 10.895417088179352, |
| "learning_rate": 1e-05, |
| "loss": 0.674, |
| "step": 1290 |
| }, |
| { |
| "epoch": 0.41, |
| "grad_norm": 8.077790185125036, |
| "learning_rate": 1e-05, |
| "loss": 0.6431, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.42, |
| "grad_norm": 7.997282520329073, |
| "learning_rate": 1e-05, |
| "loss": 0.6328, |
| "step": 1310 |
| }, |
| { |
| "epoch": 0.42, |
| "grad_norm": 7.161558089977561, |
| "learning_rate": 1e-05, |
| "loss": 0.6764, |
| "step": 1320 |
| }, |
| { |
| "epoch": 0.42, |
| "grad_norm": 10.130792538518202, |
| "learning_rate": 1e-05, |
| "loss": 0.711, |
| "step": 1330 |
| }, |
| { |
| "epoch": 0.43, |
| "grad_norm": 7.58729611771741, |
| "learning_rate": 1e-05, |
| "loss": 0.6376, |
| "step": 1340 |
| }, |
| { |
| "epoch": 0.43, |
| "grad_norm": 8.722958441221943, |
| "learning_rate": 1e-05, |
| "loss": 0.6375, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.43, |
| "grad_norm": 9.37620638396922, |
| "learning_rate": 1e-05, |
| "loss": 0.7174, |
| "step": 1360 |
| }, |
| { |
| "epoch": 0.44, |
| "grad_norm": 7.129040827068791, |
| "learning_rate": 1e-05, |
| "loss": 0.6081, |
| "step": 1370 |
| }, |
| { |
| "epoch": 0.44, |
| "grad_norm": 8.9904853542118, |
| "learning_rate": 1e-05, |
| "loss": 0.692, |
| "step": 1380 |
| }, |
| { |
| "epoch": 0.44, |
| "grad_norm": 6.334602262299757, |
| "learning_rate": 1e-05, |
| "loss": 0.6464, |
| "step": 1390 |
| }, |
| { |
| "epoch": 0.45, |
| "grad_norm": 10.429152388543233, |
| "learning_rate": 1e-05, |
| "loss": 0.6696, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.45, |
| "grad_norm": 11.45156745676315, |
| "learning_rate": 1e-05, |
| "loss": 0.6881, |
| "step": 1410 |
| }, |
| { |
| "epoch": 0.45, |
| "grad_norm": 9.313614951344835, |
| "learning_rate": 1e-05, |
| "loss": 0.6359, |
| "step": 1420 |
| }, |
| { |
| "epoch": 0.46, |
| "grad_norm": 9.5075063663996, |
| "learning_rate": 1e-05, |
| "loss": 0.6559, |
| "step": 1430 |
| }, |
| { |
| "epoch": 0.46, |
| "grad_norm": 9.329695628476822, |
| "learning_rate": 1e-05, |
| "loss": 0.6861, |
| "step": 1440 |
| }, |
| { |
| "epoch": 0.46, |
| "grad_norm": 8.390670009044785, |
| "learning_rate": 1e-05, |
| "loss": 0.6146, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.47, |
| "grad_norm": 7.145525896967856, |
| "learning_rate": 1e-05, |
| "loss": 0.6202, |
| "step": 1460 |
| }, |
| { |
| "epoch": 0.47, |
| "grad_norm": 9.936903833704589, |
| "learning_rate": 1e-05, |
| "loss": 0.7294, |
| "step": 1470 |
| }, |
| { |
| "epoch": 0.47, |
| "grad_norm": 10.832362283077313, |
| "learning_rate": 1e-05, |
| "loss": 0.5821, |
| "step": 1480 |
| }, |
| { |
| "epoch": 0.48, |
| "grad_norm": 7.756103573261522, |
| "learning_rate": 1e-05, |
| "loss": 0.6458, |
| "step": 1490 |
| }, |
| { |
| "epoch": 0.48, |
| "grad_norm": 6.4251347271638615, |
| "learning_rate": 1e-05, |
| "loss": 0.6475, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_loss": 0.712890625, |
| "eval_runtime": 55.184, |
| "eval_samples_per_second": 45.919, |
| "eval_steps_per_second": 2.881, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.48, |
| "grad_norm": 8.426862779017274, |
| "learning_rate": 1e-05, |
| "loss": 0.639, |
| "step": 1510 |
| }, |
| { |
| "epoch": 0.49, |
| "grad_norm": 8.01830105792173, |
| "learning_rate": 1e-05, |
| "loss": 0.7363, |
| "step": 1520 |
| }, |
| { |
| "epoch": 0.49, |
| "grad_norm": 6.8450717433516255, |
| "learning_rate": 1e-05, |
| "loss": 0.6331, |
| "step": 1530 |
| }, |
| { |
| "epoch": 0.49, |
| "grad_norm": 6.791175434469226, |
| "learning_rate": 1e-05, |
| "loss": 0.6546, |
| "step": 1540 |
| }, |
| { |
| "epoch": 0.49, |
| "grad_norm": 8.34539115852979, |
| "learning_rate": 1e-05, |
| "loss": 0.6755, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.5, |
| "grad_norm": 11.508932873584408, |
| "learning_rate": 1e-05, |
| "loss": 0.5975, |
| "step": 1560 |
| }, |
| { |
| "epoch": 0.5, |
| "grad_norm": 9.415577873367791, |
| "learning_rate": 1e-05, |
| "loss": 0.6264, |
| "step": 1570 |
| }, |
| { |
| "epoch": 0.5, |
| "grad_norm": 8.233517451936788, |
| "learning_rate": 1e-05, |
| "loss": 0.6411, |
| "step": 1580 |
| }, |
| { |
| "epoch": 0.51, |
| "grad_norm": 12.658554357156506, |
| "learning_rate": 1e-05, |
| "loss": 0.6865, |
| "step": 1590 |
| }, |
| { |
| "epoch": 0.51, |
| "grad_norm": 9.40643614524597, |
| "learning_rate": 1e-05, |
| "loss": 0.6547, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.51, |
| "grad_norm": 7.827186634829597, |
| "learning_rate": 1e-05, |
| "loss": 0.6674, |
| "step": 1610 |
| }, |
| { |
| "epoch": 0.52, |
| "grad_norm": 6.877784303913023, |
| "learning_rate": 1e-05, |
| "loss": 0.6803, |
| "step": 1620 |
| }, |
| { |
| "epoch": 0.52, |
| "grad_norm": 9.495560662870103, |
| "learning_rate": 1e-05, |
| "loss": 0.6656, |
| "step": 1630 |
| }, |
| { |
| "epoch": 0.52, |
| "grad_norm": 7.487057772458736, |
| "learning_rate": 1e-05, |
| "loss": 0.6451, |
| "step": 1640 |
| }, |
| { |
| "epoch": 0.53, |
| "grad_norm": 7.7627038508138035, |
| "learning_rate": 1e-05, |
| "loss": 0.6156, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.53, |
| "grad_norm": 9.048687109372294, |
| "learning_rate": 1e-05, |
| "loss": 0.6755, |
| "step": 1660 |
| }, |
| { |
| "epoch": 0.53, |
| "grad_norm": 10.15649977156812, |
| "learning_rate": 1e-05, |
| "loss": 0.6679, |
| "step": 1670 |
| }, |
| { |
| "epoch": 0.54, |
| "grad_norm": 8.887542887609689, |
| "learning_rate": 1e-05, |
| "loss": 0.7209, |
| "step": 1680 |
| }, |
| { |
| "epoch": 0.54, |
| "grad_norm": 8.320048411301839, |
| "learning_rate": 1e-05, |
| "loss": 0.6945, |
| "step": 1690 |
| }, |
| { |
| "epoch": 0.54, |
| "grad_norm": 8.860738551785932, |
| "learning_rate": 1e-05, |
| "loss": 0.6693, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.55, |
| "grad_norm": 8.238461372645636, |
| "learning_rate": 1e-05, |
| "loss": 0.6592, |
| "step": 1710 |
| }, |
| { |
| "epoch": 0.55, |
| "grad_norm": 9.54022614600892, |
| "learning_rate": 1e-05, |
| "loss": 0.6621, |
| "step": 1720 |
| }, |
| { |
| "epoch": 0.55, |
| "grad_norm": 8.994644266911353, |
| "learning_rate": 1e-05, |
| "loss": 0.6627, |
| "step": 1730 |
| }, |
| { |
| "epoch": 0.56, |
| "grad_norm": 6.5315399835632, |
| "learning_rate": 1e-05, |
| "loss": 0.6738, |
| "step": 1740 |
| }, |
| { |
| "epoch": 0.56, |
| "grad_norm": 9.393299198057026, |
| "learning_rate": 1e-05, |
| "loss": 0.6261, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.56, |
| "grad_norm": 8.314715326864587, |
| "learning_rate": 1e-05, |
| "loss": 0.6368, |
| "step": 1760 |
| }, |
| { |
| "epoch": 0.56, |
| "grad_norm": 8.207725644410043, |
| "learning_rate": 1e-05, |
| "loss": 0.6963, |
| "step": 1770 |
| }, |
| { |
| "epoch": 0.57, |
| "grad_norm": 8.25053773919251, |
| "learning_rate": 1e-05, |
| "loss": 0.6327, |
| "step": 1780 |
| }, |
| { |
| "epoch": 0.57, |
| "grad_norm": 6.900186630157612, |
| "learning_rate": 1e-05, |
| "loss": 0.6374, |
| "step": 1790 |
| }, |
| { |
| "epoch": 0.57, |
| "grad_norm": 8.177742071810924, |
| "learning_rate": 1e-05, |
| "loss": 0.6747, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.58, |
| "grad_norm": 7.820129331411966, |
| "learning_rate": 1e-05, |
| "loss": 0.6412, |
| "step": 1810 |
| }, |
| { |
| "epoch": 0.58, |
| "grad_norm": 8.300866267946969, |
| "learning_rate": 1e-05, |
| "loss": 0.6145, |
| "step": 1820 |
| }, |
| { |
| "epoch": 0.58, |
| "grad_norm": 10.817917012586616, |
| "learning_rate": 1e-05, |
| "loss": 0.6804, |
| "step": 1830 |
| }, |
| { |
| "epoch": 0.59, |
| "grad_norm": 7.484555682269077, |
| "learning_rate": 1e-05, |
| "loss": 0.6534, |
| "step": 1840 |
| }, |
| { |
| "epoch": 0.59, |
| "grad_norm": 8.468819818525622, |
| "learning_rate": 1e-05, |
| "loss": 0.6776, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.59, |
| "grad_norm": 7.228109790012504, |
| "learning_rate": 1e-05, |
| "loss": 0.6723, |
| "step": 1860 |
| }, |
| { |
| "epoch": 0.6, |
| "grad_norm": 9.59844478089623, |
| "learning_rate": 1e-05, |
| "loss": 0.6253, |
| "step": 1870 |
| }, |
| { |
| "epoch": 0.6, |
| "grad_norm": 12.571247149062518, |
| "learning_rate": 1e-05, |
| "loss": 0.7084, |
| "step": 1880 |
| }, |
| { |
| "epoch": 0.6, |
| "grad_norm": 6.566215117496402, |
| "learning_rate": 1e-05, |
| "loss": 0.5636, |
| "step": 1890 |
| }, |
| { |
| "epoch": 0.61, |
| "grad_norm": 8.81532967230638, |
| "learning_rate": 1e-05, |
| "loss": 0.6462, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.61, |
| "grad_norm": 8.776395820391842, |
| "learning_rate": 1e-05, |
| "loss": 0.7267, |
| "step": 1910 |
| }, |
| { |
| "epoch": 0.61, |
| "grad_norm": 6.867640492294022, |
| "learning_rate": 1e-05, |
| "loss": 0.6579, |
| "step": 1920 |
| }, |
| { |
| "epoch": 0.62, |
| "grad_norm": 6.332752636170762, |
| "learning_rate": 1e-05, |
| "loss": 0.6889, |
| "step": 1930 |
| }, |
| { |
| "epoch": 0.62, |
| "grad_norm": 8.238399325700778, |
| "learning_rate": 1e-05, |
| "loss": 0.6697, |
| "step": 1940 |
| }, |
| { |
| "epoch": 0.62, |
| "grad_norm": 9.490568197141299, |
| "learning_rate": 1e-05, |
| "loss": 0.6693, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.63, |
| "grad_norm": 7.762493031570821, |
| "learning_rate": 1e-05, |
| "loss": 0.6493, |
| "step": 1960 |
| }, |
| { |
| "epoch": 0.63, |
| "grad_norm": 7.91508623376212, |
| "learning_rate": 1e-05, |
| "loss": 0.7038, |
| "step": 1970 |
| }, |
| { |
| "epoch": 0.63, |
| "grad_norm": 6.0630669082958875, |
| "learning_rate": 1e-05, |
| "loss": 0.6232, |
| "step": 1980 |
| }, |
| { |
| "epoch": 0.63, |
| "grad_norm": 10.910236318740383, |
| "learning_rate": 1e-05, |
| "loss": 0.6945, |
| "step": 1990 |
| }, |
| { |
| "epoch": 0.64, |
| "grad_norm": 4.735119248110151, |
| "learning_rate": 1e-05, |
| "loss": 0.625, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_loss": 0.70556640625, |
| "eval_runtime": 54.8359, |
| "eval_samples_per_second": 46.211, |
| "eval_steps_per_second": 2.9, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.64, |
| "grad_norm": 8.635693541453085, |
| "learning_rate": 1e-05, |
| "loss": 0.6386, |
| "step": 2010 |
| }, |
| { |
| "epoch": 0.64, |
| "grad_norm": 8.426901709662715, |
| "learning_rate": 1e-05, |
| "loss": 0.7006, |
| "step": 2020 |
| }, |
| { |
| "epoch": 0.65, |
| "grad_norm": 7.4361086273220875, |
| "learning_rate": 1e-05, |
| "loss": 0.6638, |
| "step": 2030 |
| }, |
| { |
| "epoch": 0.65, |
| "grad_norm": 8.108455774912443, |
| "learning_rate": 1e-05, |
| "loss": 0.6794, |
| "step": 2040 |
| }, |
| { |
| "epoch": 0.65, |
| "grad_norm": 7.08413353682222, |
| "learning_rate": 1e-05, |
| "loss": 0.6382, |
| "step": 2050 |
| }, |
| { |
| "epoch": 0.66, |
| "grad_norm": 8.311746434064245, |
| "learning_rate": 1e-05, |
| "loss": 0.6537, |
| "step": 2060 |
| }, |
| { |
| "epoch": 0.66, |
| "grad_norm": 9.662089437075144, |
| "learning_rate": 1e-05, |
| "loss": 0.619, |
| "step": 2070 |
| }, |
| { |
| "epoch": 0.66, |
| "grad_norm": 8.159454534662391, |
| "learning_rate": 1e-05, |
| "loss": 0.611, |
| "step": 2080 |
| }, |
| { |
| "epoch": 0.67, |
| "grad_norm": 8.053803240895968, |
| "learning_rate": 1e-05, |
| "loss": 0.6677, |
| "step": 2090 |
| }, |
| { |
| "epoch": 0.67, |
| "grad_norm": 7.334937209242176, |
| "learning_rate": 1e-05, |
| "loss": 0.6577, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.67, |
| "grad_norm": 9.179013439347747, |
| "learning_rate": 1e-05, |
| "loss": 0.6279, |
| "step": 2110 |
| }, |
| { |
| "epoch": 0.68, |
| "grad_norm": 10.335927904143317, |
| "learning_rate": 1e-05, |
| "loss": 0.645, |
| "step": 2120 |
| }, |
| { |
| "epoch": 0.68, |
| "grad_norm": 8.276397550183981, |
| "learning_rate": 1e-05, |
| "loss": 0.6933, |
| "step": 2130 |
| }, |
| { |
| "epoch": 0.68, |
| "grad_norm": 8.768957908596345, |
| "learning_rate": 1e-05, |
| "loss": 0.6417, |
| "step": 2140 |
| }, |
| { |
| "epoch": 0.69, |
| "grad_norm": 8.9794628215432, |
| "learning_rate": 1e-05, |
| "loss": 0.676, |
| "step": 2150 |
| }, |
| { |
| "epoch": 0.69, |
| "grad_norm": 8.305310345293584, |
| "learning_rate": 1e-05, |
| "loss": 0.6452, |
| "step": 2160 |
| }, |
| { |
| "epoch": 0.69, |
| "grad_norm": 7.373386788522239, |
| "learning_rate": 1e-05, |
| "loss": 0.666, |
| "step": 2170 |
| }, |
| { |
| "epoch": 0.7, |
| "grad_norm": 6.5495361163859345, |
| "learning_rate": 1e-05, |
| "loss": 0.6263, |
| "step": 2180 |
| }, |
| { |
| "epoch": 0.7, |
| "grad_norm": 6.421402442790468, |
| "learning_rate": 1e-05, |
| "loss": 0.6502, |
| "step": 2190 |
| }, |
| { |
| "epoch": 0.7, |
| "grad_norm": 8.171283532759949, |
| "learning_rate": 1e-05, |
| "loss": 0.6892, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.71, |
| "grad_norm": 7.418369340651381, |
| "learning_rate": 1e-05, |
| "loss": 0.662, |
| "step": 2210 |
| }, |
| { |
| "epoch": 0.71, |
| "grad_norm": 9.434575113939493, |
| "learning_rate": 1e-05, |
| "loss": 0.6087, |
| "step": 2220 |
| }, |
| { |
| "epoch": 0.71, |
| "grad_norm": 7.68028156082286, |
| "learning_rate": 1e-05, |
| "loss": 0.6835, |
| "step": 2230 |
| }, |
| { |
| "epoch": 0.71, |
| "grad_norm": 10.67556185170642, |
| "learning_rate": 1e-05, |
| "loss": 0.6708, |
| "step": 2240 |
| }, |
| { |
| "epoch": 0.72, |
| "grad_norm": 7.646206741846737, |
| "learning_rate": 1e-05, |
| "loss": 0.6806, |
| "step": 2250 |
| }, |
| { |
| "epoch": 0.72, |
| "grad_norm": 13.830428281870411, |
| "learning_rate": 1e-05, |
| "loss": 0.6532, |
| "step": 2260 |
| }, |
| { |
| "epoch": 0.72, |
| "grad_norm": 9.257034101637885, |
| "learning_rate": 1e-05, |
| "loss": 0.6893, |
| "step": 2270 |
| }, |
| { |
| "epoch": 0.73, |
| "grad_norm": 9.848724174170313, |
| "learning_rate": 1e-05, |
| "loss": 0.6346, |
| "step": 2280 |
| }, |
| { |
| "epoch": 0.73, |
| "grad_norm": 9.638603208658651, |
| "learning_rate": 1e-05, |
| "loss": 0.711, |
| "step": 2290 |
| }, |
| { |
| "epoch": 0.73, |
| "grad_norm": 6.717566381480693, |
| "learning_rate": 1e-05, |
| "loss": 0.7398, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.74, |
| "grad_norm": 9.206662079600251, |
| "learning_rate": 1e-05, |
| "loss": 0.6003, |
| "step": 2310 |
| }, |
| { |
| "epoch": 0.74, |
| "grad_norm": 8.052763034677959, |
| "learning_rate": 1e-05, |
| "loss": 0.6206, |
| "step": 2320 |
| }, |
| { |
| "epoch": 0.74, |
| "grad_norm": 9.061921199862617, |
| "learning_rate": 1e-05, |
| "loss": 0.6294, |
| "step": 2330 |
| }, |
| { |
| "epoch": 0.75, |
| "grad_norm": 7.655084513491109, |
| "learning_rate": 1e-05, |
| "loss": 0.6647, |
| "step": 2340 |
| }, |
| { |
| "epoch": 0.75, |
| "grad_norm": 9.281850056886054, |
| "learning_rate": 1e-05, |
| "loss": 0.6459, |
| "step": 2350 |
| }, |
| { |
| "epoch": 0.75, |
| "grad_norm": 9.1977415546987, |
| "learning_rate": 1e-05, |
| "loss": 0.6499, |
| "step": 2360 |
| }, |
| { |
| "epoch": 0.76, |
| "grad_norm": 9.558615699683624, |
| "learning_rate": 1e-05, |
| "loss": 0.6822, |
| "step": 2370 |
| }, |
| { |
| "epoch": 0.76, |
| "grad_norm": 5.832877586363355, |
| "learning_rate": 1e-05, |
| "loss": 0.6589, |
| "step": 2380 |
| }, |
| { |
| "epoch": 0.76, |
| "grad_norm": 7.410695990902614, |
| "learning_rate": 1e-05, |
| "loss": 0.6627, |
| "step": 2390 |
| }, |
| { |
| "epoch": 0.77, |
| "grad_norm": 10.172640770186364, |
| "learning_rate": 1e-05, |
| "loss": 0.6649, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.77, |
| "grad_norm": 10.83731741393915, |
| "learning_rate": 1e-05, |
| "loss": 0.6061, |
| "step": 2410 |
| }, |
| { |
| "epoch": 0.77, |
| "grad_norm": 10.652024312225455, |
| "learning_rate": 1e-05, |
| "loss": 0.6744, |
| "step": 2420 |
| }, |
| { |
| "epoch": 0.78, |
| "grad_norm": 8.907890302622429, |
| "learning_rate": 1e-05, |
| "loss": 0.7082, |
| "step": 2430 |
| }, |
| { |
| "epoch": 0.78, |
| "grad_norm": 7.929493868864356, |
| "learning_rate": 1e-05, |
| "loss": 0.6118, |
| "step": 2440 |
| }, |
| { |
| "epoch": 0.78, |
| "grad_norm": 8.305680080176353, |
| "learning_rate": 1e-05, |
| "loss": 0.6525, |
| "step": 2450 |
| }, |
| { |
| "epoch": 0.78, |
| "grad_norm": 6.918617739356601, |
| "learning_rate": 1e-05, |
| "loss": 0.6118, |
| "step": 2460 |
| }, |
| { |
| "epoch": 0.79, |
| "grad_norm": 7.558366891379199, |
| "learning_rate": 1e-05, |
| "loss": 0.6245, |
| "step": 2470 |
| }, |
| { |
| "epoch": 0.79, |
| "grad_norm": 8.360044420480213, |
| "learning_rate": 1e-05, |
| "loss": 0.693, |
| "step": 2480 |
| }, |
| { |
| "epoch": 0.79, |
| "grad_norm": 8.257374559697569, |
| "learning_rate": 1e-05, |
| "loss": 0.6342, |
| "step": 2490 |
| }, |
| { |
| "epoch": 0.8, |
| "grad_norm": 8.034386642014368, |
| "learning_rate": 1e-05, |
| "loss": 0.6422, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_loss": 0.6943359375, |
| "eval_runtime": 55.0682, |
| "eval_samples_per_second": 46.016, |
| "eval_steps_per_second": 2.887, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.8, |
| "grad_norm": 10.366981441056927, |
| "learning_rate": 1e-05, |
| "loss": 0.6302, |
| "step": 2510 |
| }, |
| { |
| "epoch": 0.8, |
| "grad_norm": 11.248580843105797, |
| "learning_rate": 1e-05, |
| "loss": 0.607, |
| "step": 2520 |
| }, |
| { |
| "epoch": 0.81, |
| "grad_norm": 7.97667082022742, |
| "learning_rate": 1e-05, |
| "loss": 0.6557, |
| "step": 2530 |
| }, |
| { |
| "epoch": 0.81, |
| "grad_norm": 9.962744551161553, |
| "learning_rate": 1e-05, |
| "loss": 0.7029, |
| "step": 2540 |
| }, |
| { |
| "epoch": 0.81, |
| "grad_norm": 8.371693428123857, |
| "learning_rate": 1e-05, |
| "loss": 0.6175, |
| "step": 2550 |
| }, |
| { |
| "epoch": 0.82, |
| "grad_norm": 7.977576778834976, |
| "learning_rate": 1e-05, |
| "loss": 0.6642, |
| "step": 2560 |
| }, |
| { |
| "epoch": 0.82, |
| "grad_norm": 7.476536733808135, |
| "learning_rate": 1e-05, |
| "loss": 0.6176, |
| "step": 2570 |
| }, |
| { |
| "epoch": 0.82, |
| "grad_norm": 6.728419243685609, |
| "learning_rate": 1e-05, |
| "loss": 0.6184, |
| "step": 2580 |
| }, |
| { |
| "epoch": 0.83, |
| "grad_norm": 9.926337348525804, |
| "learning_rate": 1e-05, |
| "loss": 0.6558, |
| "step": 2590 |
| }, |
| { |
| "epoch": 0.83, |
| "grad_norm": 7.866177172976108, |
| "learning_rate": 1e-05, |
| "loss": 0.6375, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.83, |
| "grad_norm": 8.224298381157185, |
| "learning_rate": 1e-05, |
| "loss": 0.676, |
| "step": 2610 |
| }, |
| { |
| "epoch": 0.84, |
| "grad_norm": 9.93880084658438, |
| "learning_rate": 1e-05, |
| "loss": 0.6721, |
| "step": 2620 |
| }, |
| { |
| "epoch": 0.84, |
| "grad_norm": 11.622124429084263, |
| "learning_rate": 1e-05, |
| "loss": 0.6784, |
| "step": 2630 |
| }, |
| { |
| "epoch": 0.84, |
| "grad_norm": 9.09088991683325, |
| "learning_rate": 1e-05, |
| "loss": 0.6199, |
| "step": 2640 |
| }, |
| { |
| "epoch": 0.85, |
| "grad_norm": 8.924002746089169, |
| "learning_rate": 1e-05, |
| "loss": 0.6715, |
| "step": 2650 |
| }, |
| { |
| "epoch": 0.85, |
| "grad_norm": 8.25737686957022, |
| "learning_rate": 1e-05, |
| "loss": 0.6779, |
| "step": 2660 |
| }, |
| { |
| "epoch": 0.85, |
| "grad_norm": 8.613755124436148, |
| "learning_rate": 1e-05, |
| "loss": 0.6244, |
| "step": 2670 |
| }, |
| { |
| "epoch": 0.86, |
| "grad_norm": 9.695249546532118, |
| "learning_rate": 1e-05, |
| "loss": 0.6451, |
| "step": 2680 |
| }, |
| { |
| "epoch": 0.86, |
| "grad_norm": 7.764986130897924, |
| "learning_rate": 1e-05, |
| "loss": 0.67, |
| "step": 2690 |
| }, |
| { |
| "epoch": 0.86, |
| "grad_norm": 10.223621198964771, |
| "learning_rate": 1e-05, |
| "loss": 0.6604, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.86, |
| "grad_norm": 8.991341558547363, |
| "learning_rate": 1e-05, |
| "loss": 0.7089, |
| "step": 2710 |
| }, |
| { |
| "epoch": 0.87, |
| "grad_norm": 11.401239669054016, |
| "learning_rate": 1e-05, |
| "loss": 0.7243, |
| "step": 2720 |
| }, |
| { |
| "epoch": 0.87, |
| "grad_norm": 8.463416866067044, |
| "learning_rate": 1e-05, |
| "loss": 0.6209, |
| "step": 2730 |
| }, |
| { |
| "epoch": 0.87, |
| "grad_norm": 8.763431376589168, |
| "learning_rate": 1e-05, |
| "loss": 0.6484, |
| "step": 2740 |
| }, |
| { |
| "epoch": 0.88, |
| "grad_norm": 7.276682456618737, |
| "learning_rate": 1e-05, |
| "loss": 0.6549, |
| "step": 2750 |
| }, |
| { |
| "epoch": 0.88, |
| "grad_norm": 6.196654666248102, |
| "learning_rate": 1e-05, |
| "loss": 0.6539, |
| "step": 2760 |
| }, |
| { |
| "epoch": 0.88, |
| "grad_norm": 10.038241317946298, |
| "learning_rate": 1e-05, |
| "loss": 0.6402, |
| "step": 2770 |
| }, |
| { |
| "epoch": 0.89, |
| "grad_norm": 7.229139901440531, |
| "learning_rate": 1e-05, |
| "loss": 0.6607, |
| "step": 2780 |
| }, |
| { |
| "epoch": 0.89, |
| "grad_norm": 10.851727368370502, |
| "learning_rate": 1e-05, |
| "loss": 0.6598, |
| "step": 2790 |
| }, |
| { |
| "epoch": 0.89, |
| "grad_norm": 9.22358437197775, |
| "learning_rate": 1e-05, |
| "loss": 0.6764, |
| "step": 2800 |
| }, |
| { |
| "epoch": 0.9, |
| "grad_norm": 7.940228091021767, |
| "learning_rate": 1e-05, |
| "loss": 0.6319, |
| "step": 2810 |
| }, |
| { |
| "epoch": 0.9, |
| "grad_norm": 8.941846950831234, |
| "learning_rate": 1e-05, |
| "loss": 0.6435, |
| "step": 2820 |
| }, |
| { |
| "epoch": 0.9, |
| "grad_norm": 7.762340933479933, |
| "learning_rate": 1e-05, |
| "loss": 0.6342, |
| "step": 2830 |
| }, |
| { |
| "epoch": 0.91, |
| "grad_norm": 7.967605568479971, |
| "learning_rate": 1e-05, |
| "loss": 0.6501, |
| "step": 2840 |
| }, |
| { |
| "epoch": 0.91, |
| "grad_norm": 8.596005563371152, |
| "learning_rate": 1e-05, |
| "loss": 0.6704, |
| "step": 2850 |
| }, |
| { |
| "epoch": 0.91, |
| "grad_norm": 8.197680484870782, |
| "learning_rate": 1e-05, |
| "loss": 0.6671, |
| "step": 2860 |
| }, |
| { |
| "epoch": 0.92, |
| "grad_norm": 9.458297751831704, |
| "learning_rate": 1e-05, |
| "loss": 0.6484, |
| "step": 2870 |
| }, |
| { |
| "epoch": 0.92, |
| "grad_norm": 6.367092539073313, |
| "learning_rate": 1e-05, |
| "loss": 0.6415, |
| "step": 2880 |
| }, |
| { |
| "epoch": 0.92, |
| "grad_norm": 10.943089392138582, |
| "learning_rate": 1e-05, |
| "loss": 0.6329, |
| "step": 2890 |
| }, |
| { |
| "epoch": 0.93, |
| "grad_norm": 6.580063341757659, |
| "learning_rate": 1e-05, |
| "loss": 0.6503, |
| "step": 2900 |
| }, |
| { |
| "epoch": 0.93, |
| "grad_norm": 7.416221119502801, |
| "learning_rate": 1e-05, |
| "loss": 0.6961, |
| "step": 2910 |
| }, |
| { |
| "epoch": 0.93, |
| "grad_norm": 7.3036680438346355, |
| "learning_rate": 1e-05, |
| "loss": 0.5741, |
| "step": 2920 |
| }, |
| { |
| "epoch": 0.93, |
| "grad_norm": 9.30604073191193, |
| "learning_rate": 1e-05, |
| "loss": 0.7246, |
| "step": 2930 |
| }, |
| { |
| "epoch": 0.94, |
| "grad_norm": 8.352372304196583, |
| "learning_rate": 1e-05, |
| "loss": 0.705, |
| "step": 2940 |
| }, |
| { |
| "epoch": 0.94, |
| "grad_norm": 8.912810255217364, |
| "learning_rate": 1e-05, |
| "loss": 0.664, |
| "step": 2950 |
| }, |
| { |
| "epoch": 0.94, |
| "grad_norm": 7.288662391467114, |
| "learning_rate": 1e-05, |
| "loss": 0.6804, |
| "step": 2960 |
| }, |
| { |
| "epoch": 0.95, |
| "grad_norm": 7.78822890003973, |
| "learning_rate": 1e-05, |
| "loss": 0.6649, |
| "step": 2970 |
| }, |
| { |
| "epoch": 0.95, |
| "grad_norm": 8.479080937082607, |
| "learning_rate": 1e-05, |
| "loss": 0.6673, |
| "step": 2980 |
| }, |
| { |
| "epoch": 0.95, |
| "grad_norm": 9.323045737511453, |
| "learning_rate": 1e-05, |
| "loss": 0.7003, |
| "step": 2990 |
| }, |
| { |
| "epoch": 0.96, |
| "grad_norm": 9.409615456388714, |
| "learning_rate": 1e-05, |
| "loss": 0.6443, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.96, |
| "eval_loss": 0.7080078125, |
| "eval_runtime": 55.165, |
| "eval_samples_per_second": 45.935, |
| "eval_steps_per_second": 2.882, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.96, |
| "grad_norm": 6.798238683242589, |
| "learning_rate": 1e-05, |
| "loss": 0.6665, |
| "step": 3010 |
| }, |
| { |
| "epoch": 0.96, |
| "grad_norm": 8.117431871923072, |
| "learning_rate": 1e-05, |
| "loss": 0.5636, |
| "step": 3020 |
| }, |
| { |
| "epoch": 0.97, |
| "grad_norm": 8.039360018638135, |
| "learning_rate": 1e-05, |
| "loss": 0.6572, |
| "step": 3030 |
| }, |
| { |
| "epoch": 0.97, |
| "grad_norm": 8.602996623463264, |
| "learning_rate": 1e-05, |
| "loss": 0.5762, |
| "step": 3040 |
| }, |
| { |
| "epoch": 0.97, |
| "grad_norm": 10.098253125492718, |
| "learning_rate": 1e-05, |
| "loss": 0.6521, |
| "step": 3050 |
| }, |
| { |
| "epoch": 0.98, |
| "grad_norm": 6.417205820923598, |
| "learning_rate": 1e-05, |
| "loss": 0.6716, |
| "step": 3060 |
| }, |
| { |
| "epoch": 0.98, |
| "grad_norm": 7.476064252843637, |
| "learning_rate": 1e-05, |
| "loss": 0.6243, |
| "step": 3070 |
| }, |
| { |
| "epoch": 0.98, |
| "grad_norm": 6.03448304801151, |
| "learning_rate": 1e-05, |
| "loss": 0.6327, |
| "step": 3080 |
| }, |
| { |
| "epoch": 0.99, |
| "grad_norm": 9.153041863446486, |
| "learning_rate": 1e-05, |
| "loss": 0.704, |
| "step": 3090 |
| }, |
| { |
| "epoch": 0.99, |
| "grad_norm": 8.089942773114382, |
| "learning_rate": 1e-05, |
| "loss": 0.6848, |
| "step": 3100 |
| }, |
| { |
| "epoch": 0.99, |
| "grad_norm": 8.909461797520901, |
| "learning_rate": 1e-05, |
| "loss": 0.6277, |
| "step": 3110 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 7.83869169140242, |
| "learning_rate": 1e-05, |
| "loss": 0.6683, |
| "step": 3120 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 5.387254274787465, |
| "learning_rate": 1e-05, |
| "loss": 0.6976, |
| "step": 3130 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 5.73882135870757, |
| "learning_rate": 1e-05, |
| "loss": 0.4927, |
| "step": 3140 |
| }, |
| { |
| "epoch": 1.01, |
| "grad_norm": 5.517724521011402, |
| "learning_rate": 1e-05, |
| "loss": 0.3586, |
| "step": 3150 |
| }, |
| { |
| "epoch": 1.01, |
| "grad_norm": 7.239433217493402, |
| "learning_rate": 1e-05, |
| "loss": 0.4244, |
| "step": 3160 |
| }, |
| { |
| "epoch": 1.01, |
| "grad_norm": 8.120115249642733, |
| "learning_rate": 1e-05, |
| "loss": 0.3748, |
| "step": 3170 |
| }, |
| { |
| "epoch": 1.01, |
| "grad_norm": 9.784457232392583, |
| "learning_rate": 1e-05, |
| "loss": 0.3674, |
| "step": 3180 |
| }, |
| { |
| "epoch": 1.02, |
| "grad_norm": 7.209051048696012, |
| "learning_rate": 1e-05, |
| "loss": 0.3861, |
| "step": 3190 |
| }, |
| { |
| "epoch": 1.02, |
| "grad_norm": 8.955200167857038, |
| "learning_rate": 1e-05, |
| "loss": 0.4034, |
| "step": 3200 |
| }, |
| { |
| "epoch": 1.02, |
| "grad_norm": 6.778097659346512, |
| "learning_rate": 1e-05, |
| "loss": 0.4087, |
| "step": 3210 |
| }, |
| { |
| "epoch": 1.03, |
| "grad_norm": 11.335891061026519, |
| "learning_rate": 1e-05, |
| "loss": 0.4021, |
| "step": 3220 |
| }, |
| { |
| "epoch": 1.03, |
| "grad_norm": 6.457527192369038, |
| "learning_rate": 1e-05, |
| "loss": 0.3841, |
| "step": 3230 |
| }, |
| { |
| "epoch": 1.03, |
| "grad_norm": 8.438014035991054, |
| "learning_rate": 1e-05, |
| "loss": 0.3843, |
| "step": 3240 |
| }, |
| { |
| "epoch": 1.04, |
| "grad_norm": 7.505006263725213, |
| "learning_rate": 1e-05, |
| "loss": 0.3785, |
| "step": 3250 |
| }, |
| { |
| "epoch": 1.04, |
| "grad_norm": 11.194001253461606, |
| "learning_rate": 1e-05, |
| "loss": 0.3688, |
| "step": 3260 |
| }, |
| { |
| "epoch": 1.04, |
| "grad_norm": 7.122818462285816, |
| "learning_rate": 1e-05, |
| "loss": 0.3725, |
| "step": 3270 |
| }, |
| { |
| "epoch": 1.05, |
| "grad_norm": 10.926223609542442, |
| "learning_rate": 1e-05, |
| "loss": 0.3699, |
| "step": 3280 |
| }, |
| { |
| "epoch": 1.05, |
| "grad_norm": 8.760142524291396, |
| "learning_rate": 1e-05, |
| "loss": 0.3757, |
| "step": 3290 |
| }, |
| { |
| "epoch": 1.05, |
| "grad_norm": 8.037980046627416, |
| "learning_rate": 1e-05, |
| "loss": 0.4222, |
| "step": 3300 |
| }, |
| { |
| "epoch": 1.06, |
| "grad_norm": 8.210537945202457, |
| "learning_rate": 1e-05, |
| "loss": 0.3721, |
| "step": 3310 |
| }, |
| { |
| "epoch": 1.06, |
| "grad_norm": 6.59223754314082, |
| "learning_rate": 1e-05, |
| "loss": 0.3695, |
| "step": 3320 |
| }, |
| { |
| "epoch": 1.06, |
| "grad_norm": 9.186006385686388, |
| "learning_rate": 1e-05, |
| "loss": 0.3852, |
| "step": 3330 |
| }, |
| { |
| "epoch": 1.07, |
| "grad_norm": 8.871062358899765, |
| "learning_rate": 1e-05, |
| "loss": 0.4016, |
| "step": 3340 |
| }, |
| { |
| "epoch": 1.07, |
| "grad_norm": 7.746646770948554, |
| "learning_rate": 1e-05, |
| "loss": 0.4084, |
| "step": 3350 |
| }, |
| { |
| "epoch": 1.07, |
| "grad_norm": 6.827997341752163, |
| "learning_rate": 1e-05, |
| "loss": 0.3779, |
| "step": 3360 |
| }, |
| { |
| "epoch": 1.08, |
| "grad_norm": 8.036294147217701, |
| "learning_rate": 1e-05, |
| "loss": 0.3971, |
| "step": 3370 |
| }, |
| { |
| "epoch": 1.08, |
| "grad_norm": 6.877378849805353, |
| "learning_rate": 1e-05, |
| "loss": 0.3955, |
| "step": 3380 |
| }, |
| { |
| "epoch": 1.08, |
| "grad_norm": 8.410377595206134, |
| "learning_rate": 1e-05, |
| "loss": 0.3951, |
| "step": 3390 |
| }, |
| { |
| "epoch": 1.08, |
| "grad_norm": 8.475525120133367, |
| "learning_rate": 1e-05, |
| "loss": 0.3402, |
| "step": 3400 |
| }, |
| { |
| "epoch": 1.09, |
| "grad_norm": 7.558362853789239, |
| "learning_rate": 1e-05, |
| "loss": 0.3886, |
| "step": 3410 |
| }, |
| { |
| "epoch": 1.09, |
| "grad_norm": 7.298297359531505, |
| "learning_rate": 1e-05, |
| "loss": 0.3598, |
| "step": 3420 |
| }, |
| { |
| "epoch": 1.09, |
| "grad_norm": 7.023643889487004, |
| "learning_rate": 1e-05, |
| "loss": 0.3944, |
| "step": 3430 |
| }, |
| { |
| "epoch": 1.1, |
| "grad_norm": 8.270964832045273, |
| "learning_rate": 1e-05, |
| "loss": 0.3674, |
| "step": 3440 |
| }, |
| { |
| "epoch": 1.1, |
| "grad_norm": 8.580283305152557, |
| "learning_rate": 1e-05, |
| "loss": 0.3683, |
| "step": 3450 |
| }, |
| { |
| "epoch": 1.1, |
| "grad_norm": 7.517469532698408, |
| "learning_rate": 1e-05, |
| "loss": 0.3868, |
| "step": 3460 |
| }, |
| { |
| "epoch": 1.11, |
| "grad_norm": 10.772748785432759, |
| "learning_rate": 1e-05, |
| "loss": 0.3633, |
| "step": 3470 |
| }, |
| { |
| "epoch": 1.11, |
| "grad_norm": 8.50104426532848, |
| "learning_rate": 1e-05, |
| "loss": 0.4104, |
| "step": 3480 |
| }, |
| { |
| "epoch": 1.11, |
| "grad_norm": 8.474235531697222, |
| "learning_rate": 1e-05, |
| "loss": 0.3352, |
| "step": 3490 |
| }, |
| { |
| "epoch": 1.12, |
| "grad_norm": 8.13886568531234, |
| "learning_rate": 1e-05, |
| "loss": 0.3609, |
| "step": 3500 |
| }, |
| { |
| "epoch": 1.12, |
| "eval_loss": 0.7509765625, |
| "eval_runtime": 55.2186, |
| "eval_samples_per_second": 45.89, |
| "eval_steps_per_second": 2.879, |
| "step": 3500 |
| }, |
| { |
| "epoch": 1.12, |
| "grad_norm": 9.038599950351824, |
| "learning_rate": 1e-05, |
| "loss": 0.4083, |
| "step": 3510 |
| }, |
| { |
| "epoch": 1.12, |
| "grad_norm": 10.583307301246487, |
| "learning_rate": 1e-05, |
| "loss": 0.3868, |
| "step": 3520 |
| }, |
| { |
| "epoch": 1.13, |
| "grad_norm": 9.339088527470215, |
| "learning_rate": 1e-05, |
| "loss": 0.3934, |
| "step": 3530 |
| }, |
| { |
| "epoch": 1.13, |
| "grad_norm": 7.195662399880005, |
| "learning_rate": 1e-05, |
| "loss": 0.376, |
| "step": 3540 |
| }, |
| { |
| "epoch": 1.13, |
| "grad_norm": 8.863179246190255, |
| "learning_rate": 1e-05, |
| "loss": 0.3509, |
| "step": 3550 |
| }, |
| { |
| "epoch": 1.14, |
| "grad_norm": 12.440528079289944, |
| "learning_rate": 1e-05, |
| "loss": 0.3923, |
| "step": 3560 |
| }, |
| { |
| "epoch": 1.14, |
| "grad_norm": 8.86800786059233, |
| "learning_rate": 1e-05, |
| "loss": 0.3993, |
| "step": 3570 |
| }, |
| { |
| "epoch": 1.14, |
| "grad_norm": 8.053857710652547, |
| "learning_rate": 1e-05, |
| "loss": 0.3816, |
| "step": 3580 |
| }, |
| { |
| "epoch": 1.15, |
| "grad_norm": 8.379874106396976, |
| "learning_rate": 1e-05, |
| "loss": 0.3694, |
| "step": 3590 |
| }, |
| { |
| "epoch": 1.15, |
| "grad_norm": 7.3932055752547905, |
| "learning_rate": 1e-05, |
| "loss": 0.4247, |
| "step": 3600 |
| }, |
| { |
| "epoch": 1.15, |
| "grad_norm": 10.560532606738558, |
| "learning_rate": 1e-05, |
| "loss": 0.4149, |
| "step": 3610 |
| }, |
| { |
| "epoch": 1.16, |
| "grad_norm": 9.519644653997275, |
| "learning_rate": 1e-05, |
| "loss": 0.3792, |
| "step": 3620 |
| }, |
| { |
| "epoch": 1.16, |
| "grad_norm": 7.841714420098881, |
| "learning_rate": 1e-05, |
| "loss": 0.4241, |
| "step": 3630 |
| }, |
| { |
| "epoch": 1.16, |
| "grad_norm": 6.017007881837754, |
| "learning_rate": 1e-05, |
| "loss": 0.4025, |
| "step": 3640 |
| }, |
| { |
| "epoch": 1.16, |
| "grad_norm": 7.964142787709019, |
| "learning_rate": 1e-05, |
| "loss": 0.4306, |
| "step": 3650 |
| }, |
| { |
| "epoch": 1.17, |
| "grad_norm": 6.069637855146463, |
| "learning_rate": 1e-05, |
| "loss": 0.3917, |
| "step": 3660 |
| }, |
| { |
| "epoch": 1.17, |
| "grad_norm": 11.797525592573296, |
| "learning_rate": 1e-05, |
| "loss": 0.4125, |
| "step": 3670 |
| }, |
| { |
| "epoch": 1.17, |
| "grad_norm": 7.869678455264727, |
| "learning_rate": 1e-05, |
| "loss": 0.4022, |
| "step": 3680 |
| }, |
| { |
| "epoch": 1.18, |
| "grad_norm": 8.347246330453507, |
| "learning_rate": 1e-05, |
| "loss": 0.4167, |
| "step": 3690 |
| }, |
| { |
| "epoch": 1.18, |
| "grad_norm": 9.52343469608757, |
| "learning_rate": 1e-05, |
| "loss": 0.3782, |
| "step": 3700 |
| }, |
| { |
| "epoch": 1.18, |
| "grad_norm": 7.197033472341361, |
| "learning_rate": 1e-05, |
| "loss": 0.3946, |
| "step": 3710 |
| }, |
| { |
| "epoch": 1.19, |
| "grad_norm": 10.087204077255075, |
| "learning_rate": 1e-05, |
| "loss": 0.3942, |
| "step": 3720 |
| }, |
| { |
| "epoch": 1.19, |
| "grad_norm": 6.250359486731855, |
| "learning_rate": 1e-05, |
| "loss": 0.4251, |
| "step": 3730 |
| }, |
| { |
| "epoch": 1.19, |
| "grad_norm": 6.860469374544463, |
| "learning_rate": 1e-05, |
| "loss": 0.3844, |
| "step": 3740 |
| }, |
| { |
| "epoch": 1.2, |
| "grad_norm": 7.1905540280443345, |
| "learning_rate": 1e-05, |
| "loss": 0.4052, |
| "step": 3750 |
| }, |
| { |
| "epoch": 1.2, |
| "grad_norm": 11.816091808832802, |
| "learning_rate": 1e-05, |
| "loss": 0.3628, |
| "step": 3760 |
| }, |
| { |
| "epoch": 1.2, |
| "grad_norm": 7.293356330923659, |
| "learning_rate": 1e-05, |
| "loss": 0.3589, |
| "step": 3770 |
| }, |
| { |
| "epoch": 1.21, |
| "grad_norm": 8.450456631754065, |
| "learning_rate": 1e-05, |
| "loss": 0.3834, |
| "step": 3780 |
| }, |
| { |
| "epoch": 1.21, |
| "grad_norm": 6.6732882358629055, |
| "learning_rate": 1e-05, |
| "loss": 0.3584, |
| "step": 3790 |
| }, |
| { |
| "epoch": 1.21, |
| "grad_norm": 10.615176338100316, |
| "learning_rate": 1e-05, |
| "loss": 0.3785, |
| "step": 3800 |
| }, |
| { |
| "epoch": 1.22, |
| "grad_norm": 8.65379326176578, |
| "learning_rate": 1e-05, |
| "loss": 0.4138, |
| "step": 3810 |
| }, |
| { |
| "epoch": 1.22, |
| "grad_norm": 6.246174366271501, |
| "learning_rate": 1e-05, |
| "loss": 0.3775, |
| "step": 3820 |
| }, |
| { |
| "epoch": 1.22, |
| "grad_norm": 8.797562729833277, |
| "learning_rate": 1e-05, |
| "loss": 0.4164, |
| "step": 3830 |
| }, |
| { |
| "epoch": 1.23, |
| "grad_norm": 8.170263419603014, |
| "learning_rate": 1e-05, |
| "loss": 0.3956, |
| "step": 3840 |
| }, |
| { |
| "epoch": 1.23, |
| "grad_norm": 6.646593710072652, |
| "learning_rate": 1e-05, |
| "loss": 0.4062, |
| "step": 3850 |
| }, |
| { |
| "epoch": 1.23, |
| "grad_norm": 9.899295712475071, |
| "learning_rate": 1e-05, |
| "loss": 0.3752, |
| "step": 3860 |
| }, |
| { |
| "epoch": 1.23, |
| "grad_norm": 10.625580895195874, |
| "learning_rate": 1e-05, |
| "loss": 0.3512, |
| "step": 3870 |
| }, |
| { |
| "epoch": 1.24, |
| "grad_norm": 7.377536402179152, |
| "learning_rate": 1e-05, |
| "loss": 0.3968, |
| "step": 3880 |
| }, |
| { |
| "epoch": 1.24, |
| "grad_norm": 8.886396371213058, |
| "learning_rate": 1e-05, |
| "loss": 0.4097, |
| "step": 3890 |
| }, |
| { |
| "epoch": 1.24, |
| "grad_norm": 7.167305348643475, |
| "learning_rate": 1e-05, |
| "loss": 0.4044, |
| "step": 3900 |
| }, |
| { |
| "epoch": 1.25, |
| "grad_norm": 8.742613808445006, |
| "learning_rate": 1e-05, |
| "loss": 0.398, |
| "step": 3910 |
| }, |
| { |
| "epoch": 1.25, |
| "grad_norm": 9.22190466746435, |
| "learning_rate": 1e-05, |
| "loss": 0.42, |
| "step": 3920 |
| }, |
| { |
| "epoch": 1.25, |
| "grad_norm": 6.2960952768254135, |
| "learning_rate": 1e-05, |
| "loss": 0.3566, |
| "step": 3930 |
| }, |
| { |
| "epoch": 1.26, |
| "grad_norm": 7.953360503481521, |
| "learning_rate": 1e-05, |
| "loss": 0.3881, |
| "step": 3940 |
| }, |
| { |
| "epoch": 1.26, |
| "grad_norm": 7.944290260413401, |
| "learning_rate": 1e-05, |
| "loss": 0.3802, |
| "step": 3950 |
| }, |
| { |
| "epoch": 1.26, |
| "grad_norm": 7.8157654308467785, |
| "learning_rate": 1e-05, |
| "loss": 0.3825, |
| "step": 3960 |
| }, |
| { |
| "epoch": 1.27, |
| "grad_norm": 8.772331297136397, |
| "learning_rate": 1e-05, |
| "loss": 0.4162, |
| "step": 3970 |
| }, |
| { |
| "epoch": 1.27, |
| "grad_norm": 8.477322564570109, |
| "learning_rate": 1e-05, |
| "loss": 0.4047, |
| "step": 3980 |
| }, |
| { |
| "epoch": 1.27, |
| "grad_norm": 6.486540724553347, |
| "learning_rate": 1e-05, |
| "loss": 0.3856, |
| "step": 3990 |
| }, |
| { |
| "epoch": 1.28, |
| "grad_norm": 8.016970753867035, |
| "learning_rate": 1e-05, |
| "loss": 0.3997, |
| "step": 4000 |
| }, |
| { |
| "epoch": 1.28, |
| "eval_loss": 0.7392578125, |
| "eval_runtime": 55.0598, |
| "eval_samples_per_second": 46.023, |
| "eval_steps_per_second": 2.888, |
| "step": 4000 |
| }, |
| { |
| "epoch": 1.28, |
| "grad_norm": 7.822671357299029, |
| "learning_rate": 1e-05, |
| "loss": 0.4168, |
| "step": 4010 |
| }, |
| { |
| "epoch": 1.28, |
| "grad_norm": 7.165015303920845, |
| "learning_rate": 1e-05, |
| "loss": 0.3804, |
| "step": 4020 |
| }, |
| { |
| "epoch": 1.29, |
| "grad_norm": 7.27207427561212, |
| "learning_rate": 1e-05, |
| "loss": 0.3992, |
| "step": 4030 |
| }, |
| { |
| "epoch": 1.29, |
| "grad_norm": 6.755668238207569, |
| "learning_rate": 1e-05, |
| "loss": 0.4013, |
| "step": 4040 |
| }, |
| { |
| "epoch": 1.29, |
| "grad_norm": 9.061990657831048, |
| "learning_rate": 1e-05, |
| "loss": 0.3626, |
| "step": 4050 |
| }, |
| { |
| "epoch": 1.3, |
| "grad_norm": 8.14455794734833, |
| "learning_rate": 1e-05, |
| "loss": 0.3784, |
| "step": 4060 |
| }, |
| { |
| "epoch": 1.3, |
| "grad_norm": 8.328390206700128, |
| "learning_rate": 1e-05, |
| "loss": 0.401, |
| "step": 4070 |
| }, |
| { |
| "epoch": 1.3, |
| "grad_norm": 7.755286105969864, |
| "learning_rate": 1e-05, |
| "loss": 0.4068, |
| "step": 4080 |
| }, |
| { |
| "epoch": 1.31, |
| "grad_norm": 7.1705842362944825, |
| "learning_rate": 1e-05, |
| "loss": 0.3821, |
| "step": 4090 |
| }, |
| { |
| "epoch": 1.31, |
| "grad_norm": 9.14027401502954, |
| "learning_rate": 1e-05, |
| "loss": 0.4083, |
| "step": 4100 |
| }, |
| { |
| "epoch": 1.31, |
| "grad_norm": 8.780590100454924, |
| "learning_rate": 1e-05, |
| "loss": 0.3607, |
| "step": 4110 |
| }, |
| { |
| "epoch": 1.31, |
| "grad_norm": 7.52398255018759, |
| "learning_rate": 1e-05, |
| "loss": 0.4399, |
| "step": 4120 |
| }, |
| { |
| "epoch": 1.32, |
| "grad_norm": 6.419696373340324, |
| "learning_rate": 1e-05, |
| "loss": 0.4188, |
| "step": 4130 |
| }, |
| { |
| "epoch": 1.32, |
| "grad_norm": 9.26967322702585, |
| "learning_rate": 1e-05, |
| "loss": 0.4794, |
| "step": 4140 |
| }, |
| { |
| "epoch": 1.32, |
| "grad_norm": 8.641791321883245, |
| "learning_rate": 1e-05, |
| "loss": 0.3987, |
| "step": 4150 |
| }, |
| { |
| "epoch": 1.33, |
| "grad_norm": 7.96433294210171, |
| "learning_rate": 1e-05, |
| "loss": 0.393, |
| "step": 4160 |
| }, |
| { |
| "epoch": 1.33, |
| "grad_norm": 9.222887875638603, |
| "learning_rate": 1e-05, |
| "loss": 0.4267, |
| "step": 4170 |
| }, |
| { |
| "epoch": 1.33, |
| "grad_norm": 8.521340452242672, |
| "learning_rate": 1e-05, |
| "loss": 0.4458, |
| "step": 4180 |
| }, |
| { |
| "epoch": 1.34, |
| "grad_norm": 8.375327032735703, |
| "learning_rate": 1e-05, |
| "loss": 0.4657, |
| "step": 4190 |
| }, |
| { |
| "epoch": 1.34, |
| "grad_norm": 6.930817556005128, |
| "learning_rate": 1e-05, |
| "loss": 0.3854, |
| "step": 4200 |
| }, |
| { |
| "epoch": 1.34, |
| "grad_norm": 11.053128340798548, |
| "learning_rate": 1e-05, |
| "loss": 0.4132, |
| "step": 4210 |
| }, |
| { |
| "epoch": 1.35, |
| "grad_norm": 8.378082163728612, |
| "learning_rate": 1e-05, |
| "loss": 0.4289, |
| "step": 4220 |
| }, |
| { |
| "epoch": 1.35, |
| "grad_norm": 7.747261055594125, |
| "learning_rate": 1e-05, |
| "loss": 0.388, |
| "step": 4230 |
| }, |
| { |
| "epoch": 1.35, |
| "grad_norm": 6.913409220549674, |
| "learning_rate": 1e-05, |
| "loss": 0.4101, |
| "step": 4240 |
| }, |
| { |
| "epoch": 1.36, |
| "grad_norm": 10.085387175754002, |
| "learning_rate": 1e-05, |
| "loss": 0.4347, |
| "step": 4250 |
| }, |
| { |
| "epoch": 1.36, |
| "grad_norm": 8.01527424838698, |
| "learning_rate": 1e-05, |
| "loss": 0.4352, |
| "step": 4260 |
| }, |
| { |
| "epoch": 1.36, |
| "grad_norm": 10.324755605248265, |
| "learning_rate": 1e-05, |
| "loss": 0.4821, |
| "step": 4270 |
| }, |
| { |
| "epoch": 1.37, |
| "grad_norm": 9.818370836538467, |
| "learning_rate": 1e-05, |
| "loss": 0.3774, |
| "step": 4280 |
| }, |
| { |
| "epoch": 1.37, |
| "grad_norm": 8.916486886847325, |
| "learning_rate": 1e-05, |
| "loss": 0.411, |
| "step": 4290 |
| }, |
| { |
| "epoch": 1.37, |
| "grad_norm": 8.644568317814787, |
| "learning_rate": 1e-05, |
| "loss": 0.4566, |
| "step": 4300 |
| }, |
| { |
| "epoch": 1.38, |
| "grad_norm": 6.515524884010509, |
| "learning_rate": 1e-05, |
| "loss": 0.4292, |
| "step": 4310 |
| }, |
| { |
| "epoch": 1.38, |
| "grad_norm": 8.41833897460233, |
| "learning_rate": 1e-05, |
| "loss": 0.4385, |
| "step": 4320 |
| }, |
| { |
| "epoch": 1.38, |
| "grad_norm": 7.882942191078825, |
| "learning_rate": 1e-05, |
| "loss": 0.394, |
| "step": 4330 |
| }, |
| { |
| "epoch": 1.38, |
| "grad_norm": 6.733660666689486, |
| "learning_rate": 1e-05, |
| "loss": 0.3977, |
| "step": 4340 |
| }, |
| { |
| "epoch": 1.39, |
| "grad_norm": 6.729004313712028, |
| "learning_rate": 1e-05, |
| "loss": 0.3933, |
| "step": 4350 |
| }, |
| { |
| "epoch": 1.39, |
| "grad_norm": 8.881101418312019, |
| "learning_rate": 1e-05, |
| "loss": 0.4159, |
| "step": 4360 |
| }, |
| { |
| "epoch": 1.39, |
| "grad_norm": 5.959339012513422, |
| "learning_rate": 1e-05, |
| "loss": 0.4214, |
| "step": 4370 |
| }, |
| { |
| "epoch": 1.4, |
| "grad_norm": 8.257988039170202, |
| "learning_rate": 1e-05, |
| "loss": 0.4363, |
| "step": 4380 |
| }, |
| { |
| "epoch": 1.4, |
| "grad_norm": 7.456228317907102, |
| "learning_rate": 1e-05, |
| "loss": 0.4154, |
| "step": 4390 |
| }, |
| { |
| "epoch": 1.4, |
| "grad_norm": 9.655091250130154, |
| "learning_rate": 1e-05, |
| "loss": 0.4046, |
| "step": 4400 |
| }, |
| { |
| "epoch": 1.41, |
| "grad_norm": 7.644969867175379, |
| "learning_rate": 1e-05, |
| "loss": 0.4347, |
| "step": 4410 |
| }, |
| { |
| "epoch": 1.41, |
| "grad_norm": 9.097634446088083, |
| "learning_rate": 1e-05, |
| "loss": 0.4181, |
| "step": 4420 |
| }, |
| { |
| "epoch": 1.41, |
| "grad_norm": 6.619329076691769, |
| "learning_rate": 1e-05, |
| "loss": 0.427, |
| "step": 4430 |
| }, |
| { |
| "epoch": 1.42, |
| "grad_norm": 10.207755891207071, |
| "learning_rate": 1e-05, |
| "loss": 0.4007, |
| "step": 4440 |
| }, |
| { |
| "epoch": 1.42, |
| "grad_norm": 10.724742937897597, |
| "learning_rate": 1e-05, |
| "loss": 0.414, |
| "step": 4450 |
| }, |
| { |
| "epoch": 1.42, |
| "grad_norm": 8.200657646045025, |
| "learning_rate": 1e-05, |
| "loss": 0.3774, |
| "step": 4460 |
| }, |
| { |
| "epoch": 1.43, |
| "grad_norm": 10.865329696069828, |
| "learning_rate": 1e-05, |
| "loss": 0.4418, |
| "step": 4470 |
| }, |
| { |
| "epoch": 1.43, |
| "grad_norm": 7.250540614525296, |
| "learning_rate": 1e-05, |
| "loss": 0.3698, |
| "step": 4480 |
| }, |
| { |
| "epoch": 1.43, |
| "grad_norm": 11.18975534561968, |
| "learning_rate": 1e-05, |
| "loss": 0.4282, |
| "step": 4490 |
| }, |
| { |
| "epoch": 1.44, |
| "grad_norm": 9.417912448634077, |
| "learning_rate": 1e-05, |
| "loss": 0.3973, |
| "step": 4500 |
| }, |
| { |
| "epoch": 1.44, |
| "eval_loss": 0.75830078125, |
| "eval_runtime": 55.1884, |
| "eval_samples_per_second": 45.915, |
| "eval_steps_per_second": 2.881, |
| "step": 4500 |
| }, |
| { |
| "epoch": 1.44, |
| "grad_norm": 7.127268379914297, |
| "learning_rate": 1e-05, |
| "loss": 0.418, |
| "step": 4510 |
| }, |
| { |
| "epoch": 1.44, |
| "grad_norm": 6.7295102562203, |
| "learning_rate": 1e-05, |
| "loss": 0.3991, |
| "step": 4520 |
| }, |
| { |
| "epoch": 1.45, |
| "grad_norm": 6.803979987974524, |
| "learning_rate": 1e-05, |
| "loss": 0.3518, |
| "step": 4530 |
| }, |
| { |
| "epoch": 1.45, |
| "grad_norm": 10.577087441614474, |
| "learning_rate": 1e-05, |
| "loss": 0.4393, |
| "step": 4540 |
| }, |
| { |
| "epoch": 1.45, |
| "grad_norm": 8.686078950210899, |
| "learning_rate": 1e-05, |
| "loss": 0.3963, |
| "step": 4550 |
| }, |
| { |
| "epoch": 1.46, |
| "grad_norm": 8.80400552621726, |
| "learning_rate": 1e-05, |
| "loss": 0.4637, |
| "step": 4560 |
| }, |
| { |
| "epoch": 1.46, |
| "grad_norm": 6.998345315730221, |
| "learning_rate": 1e-05, |
| "loss": 0.395, |
| "step": 4570 |
| }, |
| { |
| "epoch": 1.46, |
| "grad_norm": 7.820146648466103, |
| "learning_rate": 1e-05, |
| "loss": 0.4571, |
| "step": 4580 |
| }, |
| { |
| "epoch": 1.46, |
| "grad_norm": 7.398401665122429, |
| "learning_rate": 1e-05, |
| "loss": 0.4455, |
| "step": 4590 |
| }, |
| { |
| "epoch": 1.47, |
| "grad_norm": 7.506575944034966, |
| "learning_rate": 1e-05, |
| "loss": 0.3974, |
| "step": 4600 |
| }, |
| { |
| "epoch": 1.47, |
| "grad_norm": 9.467260463878553, |
| "learning_rate": 1e-05, |
| "loss": 0.386, |
| "step": 4610 |
| }, |
| { |
| "epoch": 1.47, |
| "grad_norm": 7.700985377291013, |
| "learning_rate": 1e-05, |
| "loss": 0.3997, |
| "step": 4620 |
| }, |
| { |
| "epoch": 1.48, |
| "grad_norm": 8.492564371093573, |
| "learning_rate": 1e-05, |
| "loss": 0.4035, |
| "step": 4630 |
| }, |
| { |
| "epoch": 1.48, |
| "grad_norm": 6.813078969448784, |
| "learning_rate": 1e-05, |
| "loss": 0.3979, |
| "step": 4640 |
| }, |
| { |
| "epoch": 1.48, |
| "grad_norm": 7.857530138453749, |
| "learning_rate": 1e-05, |
| "loss": 0.412, |
| "step": 4650 |
| }, |
| { |
| "epoch": 1.49, |
| "grad_norm": 7.821510428978855, |
| "learning_rate": 1e-05, |
| "loss": 0.3598, |
| "step": 4660 |
| }, |
| { |
| "epoch": 1.49, |
| "grad_norm": 6.752394639979467, |
| "learning_rate": 1e-05, |
| "loss": 0.3905, |
| "step": 4670 |
| }, |
| { |
| "epoch": 1.49, |
| "grad_norm": 7.312708925050263, |
| "learning_rate": 1e-05, |
| "loss": 0.4246, |
| "step": 4680 |
| }, |
| { |
| "epoch": 1.5, |
| "grad_norm": 9.355257125085597, |
| "learning_rate": 1e-05, |
| "loss": 0.4233, |
| "step": 4690 |
| }, |
| { |
| "epoch": 1.5, |
| "grad_norm": 7.076742748846365, |
| "learning_rate": 1e-05, |
| "loss": 0.3627, |
| "step": 4700 |
| }, |
| { |
| "epoch": 1.5, |
| "grad_norm": 7.5700340889046025, |
| "learning_rate": 1e-05, |
| "loss": 0.4384, |
| "step": 4710 |
| }, |
| { |
| "epoch": 1.51, |
| "grad_norm": 9.204263561654585, |
| "learning_rate": 1e-05, |
| "loss": 0.4126, |
| "step": 4720 |
| }, |
| { |
| "epoch": 1.51, |
| "grad_norm": 7.633574370641786, |
| "learning_rate": 1e-05, |
| "loss": 0.3986, |
| "step": 4730 |
| }, |
| { |
| "epoch": 1.51, |
| "grad_norm": 11.410936537210695, |
| "learning_rate": 1e-05, |
| "loss": 0.4508, |
| "step": 4740 |
| }, |
| { |
| "epoch": 1.52, |
| "grad_norm": 8.10875498141563, |
| "learning_rate": 1e-05, |
| "loss": 0.3888, |
| "step": 4750 |
| }, |
| { |
| "epoch": 1.52, |
| "grad_norm": 6.455355575069892, |
| "learning_rate": 1e-05, |
| "loss": 0.4133, |
| "step": 4760 |
| }, |
| { |
| "epoch": 1.52, |
| "grad_norm": 7.74928329599341, |
| "learning_rate": 1e-05, |
| "loss": 0.422, |
| "step": 4770 |
| }, |
| { |
| "epoch": 1.53, |
| "grad_norm": 9.313844724508032, |
| "learning_rate": 1e-05, |
| "loss": 0.4221, |
| "step": 4780 |
| }, |
| { |
| "epoch": 1.53, |
| "grad_norm": 6.791703144123747, |
| "learning_rate": 1e-05, |
| "loss": 0.3867, |
| "step": 4790 |
| }, |
| { |
| "epoch": 1.53, |
| "grad_norm": 8.429396517365443, |
| "learning_rate": 1e-05, |
| "loss": 0.4063, |
| "step": 4800 |
| }, |
| { |
| "epoch": 1.53, |
| "grad_norm": 7.467867874979078, |
| "learning_rate": 1e-05, |
| "loss": 0.4264, |
| "step": 4810 |
| }, |
| { |
| "epoch": 1.54, |
| "grad_norm": 9.298873866000186, |
| "learning_rate": 1e-05, |
| "loss": 0.4239, |
| "step": 4820 |
| }, |
| { |
| "epoch": 1.54, |
| "grad_norm": 8.46600319773404, |
| "learning_rate": 1e-05, |
| "loss": 0.4164, |
| "step": 4830 |
| }, |
| { |
| "epoch": 1.54, |
| "grad_norm": 6.931017347030935, |
| "learning_rate": 1e-05, |
| "loss": 0.426, |
| "step": 4840 |
| }, |
| { |
| "epoch": 1.55, |
| "grad_norm": 7.599904019602993, |
| "learning_rate": 1e-05, |
| "loss": 0.4258, |
| "step": 4850 |
| }, |
| { |
| "epoch": 1.55, |
| "grad_norm": 12.063378494767036, |
| "learning_rate": 1e-05, |
| "loss": 0.4938, |
| "step": 4860 |
| }, |
| { |
| "epoch": 1.55, |
| "grad_norm": 7.147039491456, |
| "learning_rate": 1e-05, |
| "loss": 0.4249, |
| "step": 4870 |
| }, |
| { |
| "epoch": 1.56, |
| "grad_norm": 9.323328877860423, |
| "learning_rate": 1e-05, |
| "loss": 0.4637, |
| "step": 4880 |
| }, |
| { |
| "epoch": 1.56, |
| "grad_norm": 10.03119676047009, |
| "learning_rate": 1e-05, |
| "loss": 0.4247, |
| "step": 4890 |
| }, |
| { |
| "epoch": 1.56, |
| "grad_norm": 8.520948289563666, |
| "learning_rate": 1e-05, |
| "loss": 0.4148, |
| "step": 4900 |
| }, |
| { |
| "epoch": 1.57, |
| "grad_norm": 9.86972909603902, |
| "learning_rate": 1e-05, |
| "loss": 0.4392, |
| "step": 4910 |
| }, |
| { |
| "epoch": 1.57, |
| "grad_norm": 8.832037728740437, |
| "learning_rate": 1e-05, |
| "loss": 0.4035, |
| "step": 4920 |
| }, |
| { |
| "epoch": 1.57, |
| "grad_norm": 7.974004232489212, |
| "learning_rate": 1e-05, |
| "loss": 0.3781, |
| "step": 4930 |
| }, |
| { |
| "epoch": 1.58, |
| "grad_norm": 10.338442821014144, |
| "learning_rate": 1e-05, |
| "loss": 0.434, |
| "step": 4940 |
| }, |
| { |
| "epoch": 1.58, |
| "grad_norm": 7.955541097579312, |
| "learning_rate": 1e-05, |
| "loss": 0.429, |
| "step": 4950 |
| }, |
| { |
| "epoch": 1.58, |
| "grad_norm": 10.944679214704877, |
| "learning_rate": 1e-05, |
| "loss": 0.4005, |
| "step": 4960 |
| }, |
| { |
| "epoch": 1.59, |
| "grad_norm": 11.384183090313906, |
| "learning_rate": 1e-05, |
| "loss": 0.399, |
| "step": 4970 |
| }, |
| { |
| "epoch": 1.59, |
| "grad_norm": 7.709262963676569, |
| "learning_rate": 1e-05, |
| "loss": 0.4203, |
| "step": 4980 |
| }, |
| { |
| "epoch": 1.59, |
| "grad_norm": 8.785523593086953, |
| "learning_rate": 1e-05, |
| "loss": 0.4399, |
| "step": 4990 |
| }, |
| { |
| "epoch": 1.6, |
| "grad_norm": 7.71861458478129, |
| "learning_rate": 1e-05, |
| "loss": 0.3716, |
| "step": 5000 |
| }, |
| { |
| "epoch": 1.6, |
| "eval_loss": 0.7392578125, |
| "eval_runtime": 55.2179, |
| "eval_samples_per_second": 45.891, |
| "eval_steps_per_second": 2.88, |
| "step": 5000 |
| }, |
| { |
| "epoch": 1.6, |
| "grad_norm": 11.664023254287626, |
| "learning_rate": 1e-05, |
| "loss": 0.4278, |
| "step": 5010 |
| }, |
| { |
| "epoch": 1.6, |
| "grad_norm": 8.25578105556357, |
| "learning_rate": 1e-05, |
| "loss": 0.4283, |
| "step": 5020 |
| }, |
| { |
| "epoch": 1.6, |
| "grad_norm": 7.336549776016435, |
| "learning_rate": 1e-05, |
| "loss": 0.3569, |
| "step": 5030 |
| }, |
| { |
| "epoch": 1.61, |
| "grad_norm": 6.916179716441526, |
| "learning_rate": 1e-05, |
| "loss": 0.442, |
| "step": 5040 |
| }, |
| { |
| "epoch": 1.61, |
| "grad_norm": 9.284481218231809, |
| "learning_rate": 1e-05, |
| "loss": 0.4165, |
| "step": 5050 |
| }, |
| { |
| "epoch": 1.61, |
| "grad_norm": 8.102183539811534, |
| "learning_rate": 1e-05, |
| "loss": 0.3643, |
| "step": 5060 |
| }, |
| { |
| "epoch": 1.62, |
| "grad_norm": 7.930941057012382, |
| "learning_rate": 1e-05, |
| "loss": 0.445, |
| "step": 5070 |
| }, |
| { |
| "epoch": 1.62, |
| "grad_norm": 7.661420796551131, |
| "learning_rate": 1e-05, |
| "loss": 0.4208, |
| "step": 5080 |
| }, |
| { |
| "epoch": 1.62, |
| "grad_norm": 5.676998966068444, |
| "learning_rate": 1e-05, |
| "loss": 0.3938, |
| "step": 5090 |
| }, |
| { |
| "epoch": 1.63, |
| "grad_norm": 8.590274954359954, |
| "learning_rate": 1e-05, |
| "loss": 0.3963, |
| "step": 5100 |
| }, |
| { |
| "epoch": 1.63, |
| "grad_norm": 9.281228216383317, |
| "learning_rate": 1e-05, |
| "loss": 0.4476, |
| "step": 5110 |
| }, |
| { |
| "epoch": 1.63, |
| "grad_norm": 6.6096390475292255, |
| "learning_rate": 1e-05, |
| "loss": 0.3925, |
| "step": 5120 |
| }, |
| { |
| "epoch": 1.64, |
| "grad_norm": 8.927518806533852, |
| "learning_rate": 1e-05, |
| "loss": 0.4285, |
| "step": 5130 |
| }, |
| { |
| "epoch": 1.64, |
| "grad_norm": 10.86465558436147, |
| "learning_rate": 1e-05, |
| "loss": 0.429, |
| "step": 5140 |
| }, |
| { |
| "epoch": 1.64, |
| "grad_norm": 8.764038159898899, |
| "learning_rate": 1e-05, |
| "loss": 0.4201, |
| "step": 5150 |
| }, |
| { |
| "epoch": 1.65, |
| "grad_norm": 6.642729158823972, |
| "learning_rate": 1e-05, |
| "loss": 0.3832, |
| "step": 5160 |
| }, |
| { |
| "epoch": 1.65, |
| "grad_norm": 9.980344911169075, |
| "learning_rate": 1e-05, |
| "loss": 0.3453, |
| "step": 5170 |
| }, |
| { |
| "epoch": 1.65, |
| "grad_norm": 8.710756199066205, |
| "learning_rate": 1e-05, |
| "loss": 0.4088, |
| "step": 5180 |
| }, |
| { |
| "epoch": 1.66, |
| "grad_norm": 7.424904809206732, |
| "learning_rate": 1e-05, |
| "loss": 0.4335, |
| "step": 5190 |
| }, |
| { |
| "epoch": 1.66, |
| "grad_norm": 8.741338748314984, |
| "learning_rate": 1e-05, |
| "loss": 0.4974, |
| "step": 5200 |
| }, |
| { |
| "epoch": 1.66, |
| "grad_norm": 8.326968343546984, |
| "learning_rate": 1e-05, |
| "loss": 0.3962, |
| "step": 5210 |
| }, |
| { |
| "epoch": 1.67, |
| "grad_norm": 8.214027353334545, |
| "learning_rate": 1e-05, |
| "loss": 0.3908, |
| "step": 5220 |
| }, |
| { |
| "epoch": 1.67, |
| "grad_norm": 8.33796497060729, |
| "learning_rate": 1e-05, |
| "loss": 0.46, |
| "step": 5230 |
| }, |
| { |
| "epoch": 1.67, |
| "grad_norm": 7.94653598963845, |
| "learning_rate": 1e-05, |
| "loss": 0.469, |
| "step": 5240 |
| }, |
| { |
| "epoch": 1.68, |
| "grad_norm": 10.084073084238417, |
| "learning_rate": 1e-05, |
| "loss": 0.4251, |
| "step": 5250 |
| }, |
| { |
| "epoch": 1.68, |
| "grad_norm": 6.9028538427462305, |
| "learning_rate": 1e-05, |
| "loss": 0.4042, |
| "step": 5260 |
| }, |
| { |
| "epoch": 1.68, |
| "grad_norm": 6.179915278079327, |
| "learning_rate": 1e-05, |
| "loss": 0.3968, |
| "step": 5270 |
| }, |
| { |
| "epoch": 1.68, |
| "grad_norm": 8.133768837113916, |
| "learning_rate": 1e-05, |
| "loss": 0.4063, |
| "step": 5280 |
| }, |
| { |
| "epoch": 1.69, |
| "grad_norm": 7.178440654027076, |
| "learning_rate": 1e-05, |
| "loss": 0.3934, |
| "step": 5290 |
| }, |
| { |
| "epoch": 1.69, |
| "grad_norm": 7.026488730815509, |
| "learning_rate": 1e-05, |
| "loss": 0.3874, |
| "step": 5300 |
| }, |
| { |
| "epoch": 1.69, |
| "grad_norm": 8.64506430518711, |
| "learning_rate": 1e-05, |
| "loss": 0.377, |
| "step": 5310 |
| }, |
| { |
| "epoch": 1.7, |
| "grad_norm": 7.481211716530859, |
| "learning_rate": 1e-05, |
| "loss": 0.416, |
| "step": 5320 |
| }, |
| { |
| "epoch": 1.7, |
| "grad_norm": 6.036761201946469, |
| "learning_rate": 1e-05, |
| "loss": 0.422, |
| "step": 5330 |
| }, |
| { |
| "epoch": 1.7, |
| "grad_norm": 8.273041885902575, |
| "learning_rate": 1e-05, |
| "loss": 0.4542, |
| "step": 5340 |
| }, |
| { |
| "epoch": 1.71, |
| "grad_norm": 8.01927105109423, |
| "learning_rate": 1e-05, |
| "loss": 0.4313, |
| "step": 5350 |
| }, |
| { |
| "epoch": 1.71, |
| "grad_norm": 7.791361294506449, |
| "learning_rate": 1e-05, |
| "loss": 0.4283, |
| "step": 5360 |
| }, |
| { |
| "epoch": 1.71, |
| "grad_norm": 8.223848451279327, |
| "learning_rate": 1e-05, |
| "loss": 0.4291, |
| "step": 5370 |
| }, |
| { |
| "epoch": 1.72, |
| "grad_norm": 8.742343713695913, |
| "learning_rate": 1e-05, |
| "loss": 0.4054, |
| "step": 5380 |
| }, |
| { |
| "epoch": 1.72, |
| "grad_norm": 8.511110113335869, |
| "learning_rate": 1e-05, |
| "loss": 0.4481, |
| "step": 5390 |
| }, |
| { |
| "epoch": 1.72, |
| "grad_norm": 7.184316709729142, |
| "learning_rate": 1e-05, |
| "loss": 0.4186, |
| "step": 5400 |
| }, |
| { |
| "epoch": 1.73, |
| "grad_norm": 7.193613653274557, |
| "learning_rate": 1e-05, |
| "loss": 0.4389, |
| "step": 5410 |
| }, |
| { |
| "epoch": 1.73, |
| "grad_norm": 8.73637113468901, |
| "learning_rate": 1e-05, |
| "loss": 0.4335, |
| "step": 5420 |
| }, |
| { |
| "epoch": 1.73, |
| "grad_norm": 7.535880157360104, |
| "learning_rate": 1e-05, |
| "loss": 0.3888, |
| "step": 5430 |
| }, |
| { |
| "epoch": 1.74, |
| "grad_norm": 8.127316658160412, |
| "learning_rate": 1e-05, |
| "loss": 0.4313, |
| "step": 5440 |
| }, |
| { |
| "epoch": 1.74, |
| "grad_norm": 6.150759344596138, |
| "learning_rate": 1e-05, |
| "loss": 0.428, |
| "step": 5450 |
| }, |
| { |
| "epoch": 1.74, |
| "grad_norm": 7.999205072962964, |
| "learning_rate": 1e-05, |
| "loss": 0.3686, |
| "step": 5460 |
| }, |
| { |
| "epoch": 1.75, |
| "grad_norm": 9.581660204801329, |
| "learning_rate": 1e-05, |
| "loss": 0.4257, |
| "step": 5470 |
| }, |
| { |
| "epoch": 1.75, |
| "grad_norm": 8.782425879761018, |
| "learning_rate": 1e-05, |
| "loss": 0.4105, |
| "step": 5480 |
| }, |
| { |
| "epoch": 1.75, |
| "grad_norm": 10.301186384002998, |
| "learning_rate": 1e-05, |
| "loss": 0.4623, |
| "step": 5490 |
| }, |
| { |
| "epoch": 1.75, |
| "grad_norm": 6.577598713823145, |
| "learning_rate": 1e-05, |
| "loss": 0.4124, |
| "step": 5500 |
| }, |
| { |
| "epoch": 1.75, |
| "eval_loss": 0.73291015625, |
| "eval_runtime": 55.0248, |
| "eval_samples_per_second": 46.052, |
| "eval_steps_per_second": 2.89, |
| "step": 5500 |
| }, |
| { |
| "epoch": 1.76, |
| "grad_norm": 6.3826188579867535, |
| "learning_rate": 1e-05, |
| "loss": 0.3979, |
| "step": 5510 |
| }, |
| { |
| "epoch": 1.76, |
| "grad_norm": 9.766809303188209, |
| "learning_rate": 1e-05, |
| "loss": 0.4341, |
| "step": 5520 |
| }, |
| { |
| "epoch": 1.76, |
| "grad_norm": 9.796088033458934, |
| "learning_rate": 1e-05, |
| "loss": 0.3967, |
| "step": 5530 |
| }, |
| { |
| "epoch": 1.77, |
| "grad_norm": 9.124190203124492, |
| "learning_rate": 1e-05, |
| "loss": 0.4452, |
| "step": 5540 |
| }, |
| { |
| "epoch": 1.77, |
| "grad_norm": 7.624788875471098, |
| "learning_rate": 1e-05, |
| "loss": 0.427, |
| "step": 5550 |
| }, |
| { |
| "epoch": 1.77, |
| "grad_norm": 7.422716594965076, |
| "learning_rate": 1e-05, |
| "loss": 0.4071, |
| "step": 5560 |
| }, |
| { |
| "epoch": 1.78, |
| "grad_norm": 7.996312722178823, |
| "learning_rate": 1e-05, |
| "loss": 0.3914, |
| "step": 5570 |
| }, |
| { |
| "epoch": 1.78, |
| "grad_norm": 7.791227875728153, |
| "learning_rate": 1e-05, |
| "loss": 0.3933, |
| "step": 5580 |
| }, |
| { |
| "epoch": 1.78, |
| "grad_norm": 7.777954138920993, |
| "learning_rate": 1e-05, |
| "loss": 0.4492, |
| "step": 5590 |
| }, |
| { |
| "epoch": 1.79, |
| "grad_norm": 9.643134929551243, |
| "learning_rate": 1e-05, |
| "loss": 0.3855, |
| "step": 5600 |
| }, |
| { |
| "epoch": 1.79, |
| "grad_norm": 8.545610463326653, |
| "learning_rate": 1e-05, |
| "loss": 0.4428, |
| "step": 5610 |
| }, |
| { |
| "epoch": 1.79, |
| "grad_norm": 7.924564424211375, |
| "learning_rate": 1e-05, |
| "loss": 0.4053, |
| "step": 5620 |
| }, |
| { |
| "epoch": 1.8, |
| "grad_norm": 7.231481117505046, |
| "learning_rate": 1e-05, |
| "loss": 0.4351, |
| "step": 5630 |
| }, |
| { |
| "epoch": 1.8, |
| "grad_norm": 8.857487553160576, |
| "learning_rate": 1e-05, |
| "loss": 0.4486, |
| "step": 5640 |
| }, |
| { |
| "epoch": 1.8, |
| "grad_norm": 9.310047210671996, |
| "learning_rate": 1e-05, |
| "loss": 0.4145, |
| "step": 5650 |
| }, |
| { |
| "epoch": 1.81, |
| "grad_norm": 6.91071848202082, |
| "learning_rate": 1e-05, |
| "loss": 0.447, |
| "step": 5660 |
| }, |
| { |
| "epoch": 1.81, |
| "grad_norm": 8.67256340524913, |
| "learning_rate": 1e-05, |
| "loss": 0.4491, |
| "step": 5670 |
| }, |
| { |
| "epoch": 1.81, |
| "grad_norm": 7.675978299395016, |
| "learning_rate": 1e-05, |
| "loss": 0.4276, |
| "step": 5680 |
| }, |
| { |
| "epoch": 1.82, |
| "grad_norm": 9.647604025235196, |
| "learning_rate": 1e-05, |
| "loss": 0.4858, |
| "step": 5690 |
| }, |
| { |
| "epoch": 1.82, |
| "grad_norm": 11.557466308227783, |
| "learning_rate": 1e-05, |
| "loss": 0.4611, |
| "step": 5700 |
| }, |
| { |
| "epoch": 1.82, |
| "grad_norm": 7.43098869923079, |
| "learning_rate": 1e-05, |
| "loss": 0.4196, |
| "step": 5710 |
| }, |
| { |
| "epoch": 1.83, |
| "grad_norm": 12.718492659981063, |
| "learning_rate": 1e-05, |
| "loss": 0.4306, |
| "step": 5720 |
| }, |
| { |
| "epoch": 1.83, |
| "grad_norm": 8.064427345033268, |
| "learning_rate": 1e-05, |
| "loss": 0.4603, |
| "step": 5730 |
| }, |
| { |
| "epoch": 1.83, |
| "grad_norm": 8.481041803243242, |
| "learning_rate": 1e-05, |
| "loss": 0.4377, |
| "step": 5740 |
| }, |
| { |
| "epoch": 1.83, |
| "grad_norm": 8.141807660613328, |
| "learning_rate": 1e-05, |
| "loss": 0.4309, |
| "step": 5750 |
| }, |
| { |
| "epoch": 1.84, |
| "grad_norm": 7.244437616081114, |
| "learning_rate": 1e-05, |
| "loss": 0.4408, |
| "step": 5760 |
| }, |
| { |
| "epoch": 1.84, |
| "grad_norm": 8.221312612904436, |
| "learning_rate": 1e-05, |
| "loss": 0.4352, |
| "step": 5770 |
| }, |
| { |
| "epoch": 1.84, |
| "grad_norm": 8.803014967030649, |
| "learning_rate": 1e-05, |
| "loss": 0.4463, |
| "step": 5780 |
| }, |
| { |
| "epoch": 1.85, |
| "grad_norm": 9.406835594832256, |
| "learning_rate": 1e-05, |
| "loss": 0.4395, |
| "step": 5790 |
| }, |
| { |
| "epoch": 1.85, |
| "grad_norm": 6.894193732885848, |
| "learning_rate": 1e-05, |
| "loss": 0.4255, |
| "step": 5800 |
| }, |
| { |
| "epoch": 1.85, |
| "grad_norm": 6.655614974843375, |
| "learning_rate": 1e-05, |
| "loss": 0.3886, |
| "step": 5810 |
| }, |
| { |
| "epoch": 1.86, |
| "grad_norm": 6.780449745317732, |
| "learning_rate": 1e-05, |
| "loss": 0.4533, |
| "step": 5820 |
| }, |
| { |
| "epoch": 1.86, |
| "grad_norm": 8.837821779773291, |
| "learning_rate": 1e-05, |
| "loss": 0.4842, |
| "step": 5830 |
| }, |
| { |
| "epoch": 1.86, |
| "grad_norm": 6.567471900534226, |
| "learning_rate": 1e-05, |
| "loss": 0.4423, |
| "step": 5840 |
| }, |
| { |
| "epoch": 1.87, |
| "grad_norm": 8.943236317467974, |
| "learning_rate": 1e-05, |
| "loss": 0.441, |
| "step": 5850 |
| }, |
| { |
| "epoch": 1.87, |
| "grad_norm": 9.058851046771752, |
| "learning_rate": 1e-05, |
| "loss": 0.4292, |
| "step": 5860 |
| }, |
| { |
| "epoch": 1.87, |
| "grad_norm": 6.995553102801188, |
| "learning_rate": 1e-05, |
| "loss": 0.4026, |
| "step": 5870 |
| }, |
| { |
| "epoch": 1.88, |
| "grad_norm": 7.952199945818231, |
| "learning_rate": 1e-05, |
| "loss": 0.4549, |
| "step": 5880 |
| }, |
| { |
| "epoch": 1.88, |
| "grad_norm": 9.570229591477613, |
| "learning_rate": 1e-05, |
| "loss": 0.4527, |
| "step": 5890 |
| }, |
| { |
| "epoch": 1.88, |
| "grad_norm": 8.389298261319764, |
| "learning_rate": 1e-05, |
| "loss": 0.4399, |
| "step": 5900 |
| }, |
| { |
| "epoch": 1.89, |
| "grad_norm": 7.360753410173503, |
| "learning_rate": 1e-05, |
| "loss": 0.4612, |
| "step": 5910 |
| }, |
| { |
| "epoch": 1.89, |
| "grad_norm": 7.273992892161419, |
| "learning_rate": 1e-05, |
| "loss": 0.409, |
| "step": 5920 |
| }, |
| { |
| "epoch": 1.89, |
| "grad_norm": 8.453611462146204, |
| "learning_rate": 1e-05, |
| "loss": 0.4284, |
| "step": 5930 |
| }, |
| { |
| "epoch": 1.9, |
| "grad_norm": 8.663702775867499, |
| "learning_rate": 1e-05, |
| "loss": 0.4521, |
| "step": 5940 |
| }, |
| { |
| "epoch": 1.9, |
| "grad_norm": 8.600353020807898, |
| "learning_rate": 1e-05, |
| "loss": 0.4301, |
| "step": 5950 |
| }, |
| { |
| "epoch": 1.9, |
| "grad_norm": 8.18416521721882, |
| "learning_rate": 1e-05, |
| "loss": 0.4583, |
| "step": 5960 |
| }, |
| { |
| "epoch": 1.9, |
| "grad_norm": 8.058526057792523, |
| "learning_rate": 1e-05, |
| "loss": 0.4401, |
| "step": 5970 |
| }, |
| { |
| "epoch": 1.91, |
| "grad_norm": 11.90241554170997, |
| "learning_rate": 1e-05, |
| "loss": 0.4428, |
| "step": 5980 |
| }, |
| { |
| "epoch": 1.91, |
| "grad_norm": 9.137309094064223, |
| "learning_rate": 1e-05, |
| "loss": 0.398, |
| "step": 5990 |
| }, |
| { |
| "epoch": 1.91, |
| "grad_norm": 6.653980710037472, |
| "learning_rate": 1e-05, |
| "loss": 0.4068, |
| "step": 6000 |
| }, |
| { |
| "epoch": 1.91, |
| "eval_loss": 0.74755859375, |
| "eval_runtime": 54.8611, |
| "eval_samples_per_second": 46.189, |
| "eval_steps_per_second": 2.898, |
| "step": 6000 |
| }, |
| { |
| "epoch": 1.92, |
| "grad_norm": 8.280856141135153, |
| "learning_rate": 1e-05, |
| "loss": 0.4405, |
| "step": 6010 |
| }, |
| { |
| "epoch": 1.92, |
| "grad_norm": 7.993258496788677, |
| "learning_rate": 1e-05, |
| "loss": 0.4275, |
| "step": 6020 |
| }, |
| { |
| "epoch": 1.92, |
| "grad_norm": 8.657884794579502, |
| "learning_rate": 1e-05, |
| "loss": 0.4073, |
| "step": 6030 |
| }, |
| { |
| "epoch": 1.93, |
| "grad_norm": 6.7067246808575804, |
| "learning_rate": 1e-05, |
| "loss": 0.4305, |
| "step": 6040 |
| }, |
| { |
| "epoch": 1.93, |
| "grad_norm": 8.81401233006241, |
| "learning_rate": 1e-05, |
| "loss": 0.4131, |
| "step": 6050 |
| }, |
| { |
| "epoch": 1.93, |
| "grad_norm": 6.837470531225773, |
| "learning_rate": 1e-05, |
| "loss": 0.4278, |
| "step": 6060 |
| }, |
| { |
| "epoch": 1.94, |
| "grad_norm": 10.7781659566533, |
| "learning_rate": 1e-05, |
| "loss": 0.444, |
| "step": 6070 |
| }, |
| { |
| "epoch": 1.94, |
| "grad_norm": 5.353941362244413, |
| "learning_rate": 1e-05, |
| "loss": 0.4241, |
| "step": 6080 |
| }, |
| { |
| "epoch": 1.94, |
| "grad_norm": 7.710442029500203, |
| "learning_rate": 1e-05, |
| "loss": 0.4036, |
| "step": 6090 |
| }, |
| { |
| "epoch": 1.95, |
| "grad_norm": 7.745989346245477, |
| "learning_rate": 1e-05, |
| "loss": 0.4206, |
| "step": 6100 |
| }, |
| { |
| "epoch": 1.95, |
| "grad_norm": 6.6274582242854185, |
| "learning_rate": 1e-05, |
| "loss": 0.402, |
| "step": 6110 |
| }, |
| { |
| "epoch": 1.95, |
| "grad_norm": 10.129076101906959, |
| "learning_rate": 1e-05, |
| "loss": 0.4235, |
| "step": 6120 |
| }, |
| { |
| "epoch": 1.96, |
| "grad_norm": 7.47254649439131, |
| "learning_rate": 1e-05, |
| "loss": 0.4472, |
| "step": 6130 |
| }, |
| { |
| "epoch": 1.96, |
| "grad_norm": 6.41533511659392, |
| "learning_rate": 1e-05, |
| "loss": 0.4428, |
| "step": 6140 |
| }, |
| { |
| "epoch": 1.96, |
| "grad_norm": 8.27967629504809, |
| "learning_rate": 1e-05, |
| "loss": 0.4234, |
| "step": 6150 |
| }, |
| { |
| "epoch": 1.97, |
| "grad_norm": 8.784096687437918, |
| "learning_rate": 1e-05, |
| "loss": 0.4227, |
| "step": 6160 |
| }, |
| { |
| "epoch": 1.97, |
| "grad_norm": 9.333154494979093, |
| "learning_rate": 1e-05, |
| "loss": 0.4361, |
| "step": 6170 |
| }, |
| { |
| "epoch": 1.97, |
| "grad_norm": 7.4631214081203145, |
| "learning_rate": 1e-05, |
| "loss": 0.4187, |
| "step": 6180 |
| }, |
| { |
| "epoch": 1.98, |
| "grad_norm": 6.327157104316683, |
| "learning_rate": 1e-05, |
| "loss": 0.4218, |
| "step": 6190 |
| }, |
| { |
| "epoch": 1.98, |
| "grad_norm": 7.012548371010479, |
| "learning_rate": 1e-05, |
| "loss": 0.404, |
| "step": 6200 |
| }, |
| { |
| "epoch": 1.98, |
| "grad_norm": 6.910500163678967, |
| "learning_rate": 1e-05, |
| "loss": 0.4416, |
| "step": 6210 |
| }, |
| { |
| "epoch": 1.98, |
| "grad_norm": 7.761624386064528, |
| "learning_rate": 1e-05, |
| "loss": 0.3872, |
| "step": 6220 |
| }, |
| { |
| "epoch": 1.99, |
| "grad_norm": 6.944781906405633, |
| "learning_rate": 1e-05, |
| "loss": 0.402, |
| "step": 6230 |
| }, |
| { |
| "epoch": 1.99, |
| "grad_norm": 10.150859822653457, |
| "learning_rate": 1e-05, |
| "loss": 0.3978, |
| "step": 6240 |
| }, |
| { |
| "epoch": 1.99, |
| "grad_norm": 8.425735523027287, |
| "learning_rate": 1e-05, |
| "loss": 0.4106, |
| "step": 6250 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 9.320122588174137, |
| "learning_rate": 1e-05, |
| "loss": 0.4002, |
| "step": 6260 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 5.7611840986048675, |
| "learning_rate": 1e-05, |
| "loss": 0.4024, |
| "step": 6270 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 6.066436433440687, |
| "learning_rate": 1e-05, |
| "loss": 0.2228, |
| "step": 6280 |
| }, |
| { |
| "epoch": 2.01, |
| "grad_norm": 6.4625550612680955, |
| "learning_rate": 1e-05, |
| "loss": 0.2191, |
| "step": 6290 |
| }, |
| { |
| "epoch": 2.01, |
| "grad_norm": 7.225711256874621, |
| "learning_rate": 1e-05, |
| "loss": 0.2114, |
| "step": 6300 |
| }, |
| { |
| "epoch": 2.01, |
| "grad_norm": 8.399684682104981, |
| "learning_rate": 1e-05, |
| "loss": 0.2707, |
| "step": 6310 |
| }, |
| { |
| "epoch": 2.02, |
| "grad_norm": 6.03356256769587, |
| "learning_rate": 1e-05, |
| "loss": 0.212, |
| "step": 6320 |
| }, |
| { |
| "epoch": 2.02, |
| "grad_norm": 7.638432911836879, |
| "learning_rate": 1e-05, |
| "loss": 0.1895, |
| "step": 6330 |
| }, |
| { |
| "epoch": 2.02, |
| "grad_norm": 7.701199118643164, |
| "learning_rate": 1e-05, |
| "loss": 0.2253, |
| "step": 6340 |
| }, |
| { |
| "epoch": 2.03, |
| "grad_norm": 9.166073681696064, |
| "learning_rate": 1e-05, |
| "loss": 0.2082, |
| "step": 6350 |
| }, |
| { |
| "epoch": 2.03, |
| "grad_norm": 9.099329125124761, |
| "learning_rate": 1e-05, |
| "loss": 0.226, |
| "step": 6360 |
| }, |
| { |
| "epoch": 2.03, |
| "grad_norm": 6.80502188875937, |
| "learning_rate": 1e-05, |
| "loss": 0.2242, |
| "step": 6370 |
| }, |
| { |
| "epoch": 2.04, |
| "grad_norm": 6.797448565957373, |
| "learning_rate": 1e-05, |
| "loss": 0.2276, |
| "step": 6380 |
| }, |
| { |
| "epoch": 2.04, |
| "grad_norm": 8.057308921146145, |
| "learning_rate": 1e-05, |
| "loss": 0.2343, |
| "step": 6390 |
| }, |
| { |
| "epoch": 2.04, |
| "grad_norm": 6.3957537750697036, |
| "learning_rate": 1e-05, |
| "loss": 0.2312, |
| "step": 6400 |
| }, |
| { |
| "epoch": 2.05, |
| "grad_norm": 5.343040910041943, |
| "learning_rate": 1e-05, |
| "loss": 0.2376, |
| "step": 6410 |
| }, |
| { |
| "epoch": 2.05, |
| "grad_norm": 6.84211579063316, |
| "learning_rate": 1e-05, |
| "loss": 0.2477, |
| "step": 6420 |
| }, |
| { |
| "epoch": 2.05, |
| "grad_norm": 7.251568854702559, |
| "learning_rate": 1e-05, |
| "loss": 0.2678, |
| "step": 6430 |
| }, |
| { |
| "epoch": 2.05, |
| "grad_norm": 7.680294474678262, |
| "learning_rate": 1e-05, |
| "loss": 0.2595, |
| "step": 6440 |
| }, |
| { |
| "epoch": 2.06, |
| "grad_norm": 7.396260796313895, |
| "learning_rate": 1e-05, |
| "loss": 0.2055, |
| "step": 6450 |
| }, |
| { |
| "epoch": 2.06, |
| "grad_norm": 7.7650784889277835, |
| "learning_rate": 1e-05, |
| "loss": 0.2233, |
| "step": 6460 |
| }, |
| { |
| "epoch": 2.06, |
| "grad_norm": 5.059420555621224, |
| "learning_rate": 1e-05, |
| "loss": 0.2435, |
| "step": 6470 |
| }, |
| { |
| "epoch": 2.07, |
| "grad_norm": 7.89806657140887, |
| "learning_rate": 1e-05, |
| "loss": 0.2071, |
| "step": 6480 |
| }, |
| { |
| "epoch": 2.07, |
| "grad_norm": 5.983635520565131, |
| "learning_rate": 1e-05, |
| "loss": 0.2574, |
| "step": 6490 |
| }, |
| { |
| "epoch": 2.07, |
| "grad_norm": 7.50663565503361, |
| "learning_rate": 1e-05, |
| "loss": 0.2555, |
| "step": 6500 |
| }, |
| { |
| "epoch": 2.07, |
| "eval_loss": 0.806640625, |
| "eval_runtime": 55.0114, |
| "eval_samples_per_second": 46.063, |
| "eval_steps_per_second": 2.89, |
| "step": 6500 |
| }, |
| { |
| "epoch": 2.08, |
| "grad_norm": 7.731173875907314, |
| "learning_rate": 1e-05, |
| "loss": 0.2338, |
| "step": 6510 |
| }, |
| { |
| "epoch": 2.08, |
| "grad_norm": 7.030329800896612, |
| "learning_rate": 1e-05, |
| "loss": 0.2237, |
| "step": 6520 |
| }, |
| { |
| "epoch": 2.08, |
| "grad_norm": 7.458496205631136, |
| "learning_rate": 1e-05, |
| "loss": 0.249, |
| "step": 6530 |
| }, |
| { |
| "epoch": 2.09, |
| "grad_norm": 11.795743167972319, |
| "learning_rate": 1e-05, |
| "loss": 0.2451, |
| "step": 6540 |
| }, |
| { |
| "epoch": 2.09, |
| "grad_norm": 6.331272726969019, |
| "learning_rate": 1e-05, |
| "loss": 0.2397, |
| "step": 6550 |
| }, |
| { |
| "epoch": 2.09, |
| "grad_norm": 7.908645689101402, |
| "learning_rate": 1e-05, |
| "loss": 0.2315, |
| "step": 6560 |
| }, |
| { |
| "epoch": 2.1, |
| "grad_norm": 9.183013858765184, |
| "learning_rate": 1e-05, |
| "loss": 0.2312, |
| "step": 6570 |
| }, |
| { |
| "epoch": 2.1, |
| "grad_norm": 7.291830994707097, |
| "learning_rate": 1e-05, |
| "loss": 0.2138, |
| "step": 6580 |
| }, |
| { |
| "epoch": 2.1, |
| "grad_norm": 9.691183066883003, |
| "learning_rate": 1e-05, |
| "loss": 0.2139, |
| "step": 6590 |
| }, |
| { |
| "epoch": 2.11, |
| "grad_norm": 5.765645841235198, |
| "learning_rate": 1e-05, |
| "loss": 0.209, |
| "step": 6600 |
| }, |
| { |
| "epoch": 2.11, |
| "grad_norm": 7.656247508768727, |
| "learning_rate": 1e-05, |
| "loss": 0.2497, |
| "step": 6610 |
| }, |
| { |
| "epoch": 2.11, |
| "grad_norm": 6.537736363997612, |
| "learning_rate": 1e-05, |
| "loss": 0.2323, |
| "step": 6620 |
| }, |
| { |
| "epoch": 2.12, |
| "grad_norm": 5.941147175996038, |
| "learning_rate": 1e-05, |
| "loss": 0.2446, |
| "step": 6630 |
| }, |
| { |
| "epoch": 2.12, |
| "grad_norm": 5.501817229640507, |
| "learning_rate": 1e-05, |
| "loss": 0.2355, |
| "step": 6640 |
| }, |
| { |
| "epoch": 2.12, |
| "grad_norm": 9.564046896685115, |
| "learning_rate": 1e-05, |
| "loss": 0.223, |
| "step": 6650 |
| }, |
| { |
| "epoch": 2.13, |
| "grad_norm": 5.818963179898432, |
| "learning_rate": 1e-05, |
| "loss": 0.2201, |
| "step": 6660 |
| }, |
| { |
| "epoch": 2.13, |
| "grad_norm": 9.872891357887436, |
| "learning_rate": 1e-05, |
| "loss": 0.2243, |
| "step": 6670 |
| }, |
| { |
| "epoch": 2.13, |
| "grad_norm": 6.523098910561809, |
| "learning_rate": 1e-05, |
| "loss": 0.225, |
| "step": 6680 |
| }, |
| { |
| "epoch": 2.13, |
| "grad_norm": 8.683549819951857, |
| "learning_rate": 1e-05, |
| "loss": 0.2327, |
| "step": 6690 |
| }, |
| { |
| "epoch": 2.14, |
| "grad_norm": 11.103443305888057, |
| "learning_rate": 1e-05, |
| "loss": 0.2505, |
| "step": 6700 |
| }, |
| { |
| "epoch": 2.14, |
| "grad_norm": 6.58140558274447, |
| "learning_rate": 1e-05, |
| "loss": 0.2489, |
| "step": 6710 |
| }, |
| { |
| "epoch": 2.14, |
| "grad_norm": 6.284850953003169, |
| "learning_rate": 1e-05, |
| "loss": 0.2484, |
| "step": 6720 |
| }, |
| { |
| "epoch": 2.15, |
| "grad_norm": 11.080968924893398, |
| "learning_rate": 1e-05, |
| "loss": 0.2258, |
| "step": 6730 |
| }, |
| { |
| "epoch": 2.15, |
| "grad_norm": 12.650599746660049, |
| "learning_rate": 1e-05, |
| "loss": 0.2403, |
| "step": 6740 |
| }, |
| { |
| "epoch": 2.15, |
| "grad_norm": 8.298590703608392, |
| "learning_rate": 1e-05, |
| "loss": 0.2447, |
| "step": 6750 |
| }, |
| { |
| "epoch": 2.16, |
| "grad_norm": 5.45315840309209, |
| "learning_rate": 1e-05, |
| "loss": 0.2244, |
| "step": 6760 |
| }, |
| { |
| "epoch": 2.16, |
| "grad_norm": 5.707295938420483, |
| "learning_rate": 1e-05, |
| "loss": 0.2436, |
| "step": 6770 |
| }, |
| { |
| "epoch": 2.16, |
| "grad_norm": 11.167938349796854, |
| "learning_rate": 1e-05, |
| "loss": 0.2634, |
| "step": 6780 |
| }, |
| { |
| "epoch": 2.17, |
| "grad_norm": 7.483693515914232, |
| "learning_rate": 1e-05, |
| "loss": 0.2255, |
| "step": 6790 |
| }, |
| { |
| "epoch": 2.17, |
| "grad_norm": 7.881702779761242, |
| "learning_rate": 1e-05, |
| "loss": 0.2263, |
| "step": 6800 |
| }, |
| { |
| "epoch": 2.17, |
| "grad_norm": 7.432125938900638, |
| "learning_rate": 1e-05, |
| "loss": 0.2474, |
| "step": 6810 |
| }, |
| { |
| "epoch": 2.18, |
| "grad_norm": 6.019516040701097, |
| "learning_rate": 1e-05, |
| "loss": 0.2601, |
| "step": 6820 |
| }, |
| { |
| "epoch": 2.18, |
| "grad_norm": 6.939751062243386, |
| "learning_rate": 1e-05, |
| "loss": 0.2555, |
| "step": 6830 |
| }, |
| { |
| "epoch": 2.18, |
| "grad_norm": 7.880377250587199, |
| "learning_rate": 1e-05, |
| "loss": 0.2544, |
| "step": 6840 |
| }, |
| { |
| "epoch": 2.19, |
| "grad_norm": 8.322999217770233, |
| "learning_rate": 1e-05, |
| "loss": 0.2249, |
| "step": 6850 |
| }, |
| { |
| "epoch": 2.19, |
| "grad_norm": 6.258224568985949, |
| "learning_rate": 1e-05, |
| "loss": 0.2308, |
| "step": 6860 |
| }, |
| { |
| "epoch": 2.19, |
| "grad_norm": 7.4144578995098165, |
| "learning_rate": 1e-05, |
| "loss": 0.2413, |
| "step": 6870 |
| }, |
| { |
| "epoch": 2.2, |
| "grad_norm": 8.718185570607837, |
| "learning_rate": 1e-05, |
| "loss": 0.2213, |
| "step": 6880 |
| }, |
| { |
| "epoch": 2.2, |
| "grad_norm": 5.1125478429817495, |
| "learning_rate": 1e-05, |
| "loss": 0.2245, |
| "step": 6890 |
| }, |
| { |
| "epoch": 2.2, |
| "grad_norm": 9.22954752002049, |
| "learning_rate": 1e-05, |
| "loss": 0.2416, |
| "step": 6900 |
| }, |
| { |
| "epoch": 2.2, |
| "grad_norm": 9.696731598777522, |
| "learning_rate": 1e-05, |
| "loss": 0.2544, |
| "step": 6910 |
| }, |
| { |
| "epoch": 2.21, |
| "grad_norm": 8.350665823150955, |
| "learning_rate": 1e-05, |
| "loss": 0.2638, |
| "step": 6920 |
| }, |
| { |
| "epoch": 2.21, |
| "grad_norm": 7.282471124635672, |
| "learning_rate": 1e-05, |
| "loss": 0.2451, |
| "step": 6930 |
| }, |
| { |
| "epoch": 2.21, |
| "grad_norm": 5.690944089956408, |
| "learning_rate": 1e-05, |
| "loss": 0.2518, |
| "step": 6940 |
| }, |
| { |
| "epoch": 2.22, |
| "grad_norm": 5.548051908045313, |
| "learning_rate": 1e-05, |
| "loss": 0.2447, |
| "step": 6950 |
| }, |
| { |
| "epoch": 2.22, |
| "grad_norm": 8.826141404641474, |
| "learning_rate": 1e-05, |
| "loss": 0.2698, |
| "step": 6960 |
| }, |
| { |
| "epoch": 2.22, |
| "grad_norm": 7.068685877717427, |
| "learning_rate": 1e-05, |
| "loss": 0.2209, |
| "step": 6970 |
| }, |
| { |
| "epoch": 2.23, |
| "grad_norm": 7.307551869884653, |
| "learning_rate": 1e-05, |
| "loss": 0.2255, |
| "step": 6980 |
| }, |
| { |
| "epoch": 2.23, |
| "grad_norm": 7.027315522455814, |
| "learning_rate": 1e-05, |
| "loss": 0.2453, |
| "step": 6990 |
| }, |
| { |
| "epoch": 2.23, |
| "grad_norm": 5.635998294545789, |
| "learning_rate": 1e-05, |
| "loss": 0.2432, |
| "step": 7000 |
| }, |
| { |
| "epoch": 2.23, |
| "eval_loss": 0.841796875, |
| "eval_runtime": 54.9724, |
| "eval_samples_per_second": 46.096, |
| "eval_steps_per_second": 2.892, |
| "step": 7000 |
| }, |
| { |
| "epoch": 2.24, |
| "grad_norm": 9.93859089555747, |
| "learning_rate": 1e-05, |
| "loss": 0.2717, |
| "step": 7010 |
| }, |
| { |
| "epoch": 2.24, |
| "grad_norm": 9.175032479958901, |
| "learning_rate": 1e-05, |
| "loss": 0.2542, |
| "step": 7020 |
| }, |
| { |
| "epoch": 2.24, |
| "grad_norm": 9.843332239399182, |
| "learning_rate": 1e-05, |
| "loss": 0.2512, |
| "step": 7030 |
| }, |
| { |
| "epoch": 2.25, |
| "grad_norm": 7.119180276810014, |
| "learning_rate": 1e-05, |
| "loss": 0.2489, |
| "step": 7040 |
| }, |
| { |
| "epoch": 2.25, |
| "grad_norm": 5.0653748061599435, |
| "learning_rate": 1e-05, |
| "loss": 0.2574, |
| "step": 7050 |
| }, |
| { |
| "epoch": 2.25, |
| "grad_norm": 6.355788660702283, |
| "learning_rate": 1e-05, |
| "loss": 0.2451, |
| "step": 7060 |
| }, |
| { |
| "epoch": 2.26, |
| "grad_norm": 7.947721373098963, |
| "learning_rate": 1e-05, |
| "loss": 0.2501, |
| "step": 7070 |
| }, |
| { |
| "epoch": 2.26, |
| "grad_norm": 4.989478102857212, |
| "learning_rate": 1e-05, |
| "loss": 0.2367, |
| "step": 7080 |
| }, |
| { |
| "epoch": 2.26, |
| "grad_norm": 7.533712783754466, |
| "learning_rate": 1e-05, |
| "loss": 0.2421, |
| "step": 7090 |
| }, |
| { |
| "epoch": 2.27, |
| "grad_norm": 9.238389022230782, |
| "learning_rate": 1e-05, |
| "loss": 0.252, |
| "step": 7100 |
| }, |
| { |
| "epoch": 2.27, |
| "grad_norm": 7.050750140599653, |
| "learning_rate": 1e-05, |
| "loss": 0.2435, |
| "step": 7110 |
| }, |
| { |
| "epoch": 2.27, |
| "grad_norm": 8.674718881938714, |
| "learning_rate": 1e-05, |
| "loss": 0.2786, |
| "step": 7120 |
| }, |
| { |
| "epoch": 2.28, |
| "grad_norm": 7.603477997273965, |
| "learning_rate": 1e-05, |
| "loss": 0.2772, |
| "step": 7130 |
| }, |
| { |
| "epoch": 2.28, |
| "grad_norm": 7.202818890933322, |
| "learning_rate": 1e-05, |
| "loss": 0.2564, |
| "step": 7140 |
| }, |
| { |
| "epoch": 2.28, |
| "grad_norm": 8.656715353014965, |
| "learning_rate": 1e-05, |
| "loss": 0.267, |
| "step": 7150 |
| }, |
| { |
| "epoch": 2.28, |
| "grad_norm": 9.089675042751116, |
| "learning_rate": 1e-05, |
| "loss": 0.2265, |
| "step": 7160 |
| }, |
| { |
| "epoch": 2.29, |
| "grad_norm": 6.123940512439874, |
| "learning_rate": 1e-05, |
| "loss": 0.2282, |
| "step": 7170 |
| }, |
| { |
| "epoch": 2.29, |
| "grad_norm": 8.811517755404491, |
| "learning_rate": 1e-05, |
| "loss": 0.275, |
| "step": 7180 |
| }, |
| { |
| "epoch": 2.29, |
| "grad_norm": 5.597997150568971, |
| "learning_rate": 1e-05, |
| "loss": 0.2514, |
| "step": 7190 |
| }, |
| { |
| "epoch": 2.3, |
| "grad_norm": 8.029069062612866, |
| "learning_rate": 1e-05, |
| "loss": 0.2388, |
| "step": 7200 |
| }, |
| { |
| "epoch": 2.3, |
| "grad_norm": 5.803097246521713, |
| "learning_rate": 1e-05, |
| "loss": 0.2459, |
| "step": 7210 |
| }, |
| { |
| "epoch": 2.3, |
| "grad_norm": 5.88869033734668, |
| "learning_rate": 1e-05, |
| "loss": 0.2359, |
| "step": 7220 |
| }, |
| { |
| "epoch": 2.31, |
| "grad_norm": 9.740458588192897, |
| "learning_rate": 1e-05, |
| "loss": 0.2667, |
| "step": 7230 |
| }, |
| { |
| "epoch": 2.31, |
| "grad_norm": 5.9005752024465, |
| "learning_rate": 1e-05, |
| "loss": 0.2498, |
| "step": 7240 |
| }, |
| { |
| "epoch": 2.31, |
| "grad_norm": 8.447991524085214, |
| "learning_rate": 1e-05, |
| "loss": 0.2427, |
| "step": 7250 |
| }, |
| { |
| "epoch": 2.32, |
| "grad_norm": 5.993866328442137, |
| "learning_rate": 1e-05, |
| "loss": 0.2295, |
| "step": 7260 |
| }, |
| { |
| "epoch": 2.32, |
| "grad_norm": 9.490085447547763, |
| "learning_rate": 1e-05, |
| "loss": 0.2574, |
| "step": 7270 |
| }, |
| { |
| "epoch": 2.32, |
| "grad_norm": 6.001069609352888, |
| "learning_rate": 1e-05, |
| "loss": 0.2398, |
| "step": 7280 |
| }, |
| { |
| "epoch": 2.33, |
| "grad_norm": 6.130129476303109, |
| "learning_rate": 1e-05, |
| "loss": 0.2545, |
| "step": 7290 |
| }, |
| { |
| "epoch": 2.33, |
| "grad_norm": 7.006387248809006, |
| "learning_rate": 1e-05, |
| "loss": 0.2752, |
| "step": 7300 |
| }, |
| { |
| "epoch": 2.33, |
| "grad_norm": 6.53107214872606, |
| "learning_rate": 1e-05, |
| "loss": 0.2624, |
| "step": 7310 |
| }, |
| { |
| "epoch": 2.34, |
| "grad_norm": 6.999598900338862, |
| "learning_rate": 1e-05, |
| "loss": 0.2569, |
| "step": 7320 |
| }, |
| { |
| "epoch": 2.34, |
| "grad_norm": 10.578770492445516, |
| "learning_rate": 1e-05, |
| "loss": 0.2588, |
| "step": 7330 |
| }, |
| { |
| "epoch": 2.34, |
| "grad_norm": 7.92866565152497, |
| "learning_rate": 1e-05, |
| "loss": 0.2727, |
| "step": 7340 |
| }, |
| { |
| "epoch": 2.35, |
| "grad_norm": 6.155813753578807, |
| "learning_rate": 1e-05, |
| "loss": 0.2472, |
| "step": 7350 |
| }, |
| { |
| "epoch": 2.35, |
| "grad_norm": 7.9715705218277195, |
| "learning_rate": 1e-05, |
| "loss": 0.2525, |
| "step": 7360 |
| }, |
| { |
| "epoch": 2.35, |
| "grad_norm": 7.861666578506211, |
| "learning_rate": 1e-05, |
| "loss": 0.2581, |
| "step": 7370 |
| }, |
| { |
| "epoch": 2.35, |
| "grad_norm": 8.327965597090811, |
| "learning_rate": 1e-05, |
| "loss": 0.2357, |
| "step": 7380 |
| }, |
| { |
| "epoch": 2.36, |
| "grad_norm": 6.758671876806129, |
| "learning_rate": 1e-05, |
| "loss": 0.291, |
| "step": 7390 |
| }, |
| { |
| "epoch": 2.36, |
| "grad_norm": 7.085728618062374, |
| "learning_rate": 1e-05, |
| "loss": 0.2296, |
| "step": 7400 |
| }, |
| { |
| "epoch": 2.36, |
| "grad_norm": 6.00437831077671, |
| "learning_rate": 1e-05, |
| "loss": 0.2686, |
| "step": 7410 |
| }, |
| { |
| "epoch": 2.37, |
| "grad_norm": 5.658477850273097, |
| "learning_rate": 1e-05, |
| "loss": 0.2321, |
| "step": 7420 |
| }, |
| { |
| "epoch": 2.37, |
| "grad_norm": 6.0920811642144885, |
| "learning_rate": 1e-05, |
| "loss": 0.2349, |
| "step": 7430 |
| }, |
| { |
| "epoch": 2.37, |
| "grad_norm": 8.84736349198321, |
| "learning_rate": 1e-05, |
| "loss": 0.2799, |
| "step": 7440 |
| }, |
| { |
| "epoch": 2.38, |
| "grad_norm": 5.67459621253292, |
| "learning_rate": 1e-05, |
| "loss": 0.2548, |
| "step": 7450 |
| }, |
| { |
| "epoch": 2.38, |
| "grad_norm": 5.667919955316328, |
| "learning_rate": 1e-05, |
| "loss": 0.2389, |
| "step": 7460 |
| }, |
| { |
| "epoch": 2.38, |
| "grad_norm": 6.665293106994503, |
| "learning_rate": 1e-05, |
| "loss": 0.2255, |
| "step": 7470 |
| }, |
| { |
| "epoch": 2.39, |
| "grad_norm": 10.238343100013486, |
| "learning_rate": 1e-05, |
| "loss": 0.2514, |
| "step": 7480 |
| }, |
| { |
| "epoch": 2.39, |
| "grad_norm": 5.360844091025791, |
| "learning_rate": 1e-05, |
| "loss": 0.2692, |
| "step": 7490 |
| }, |
| { |
| "epoch": 2.39, |
| "grad_norm": 10.005247265278129, |
| "learning_rate": 1e-05, |
| "loss": 0.2429, |
| "step": 7500 |
| }, |
| { |
| "epoch": 2.39, |
| "eval_loss": 0.833984375, |
| "eval_runtime": 54.7853, |
| "eval_samples_per_second": 46.253, |
| "eval_steps_per_second": 2.902, |
| "step": 7500 |
| }, |
| { |
| "epoch": 2.4, |
| "grad_norm": 8.642315718120907, |
| "learning_rate": 1e-05, |
| "loss": 0.2448, |
| "step": 7510 |
| }, |
| { |
| "epoch": 2.4, |
| "grad_norm": 10.98664947446694, |
| "learning_rate": 1e-05, |
| "loss": 0.2193, |
| "step": 7520 |
| }, |
| { |
| "epoch": 2.4, |
| "grad_norm": 8.953089635755855, |
| "learning_rate": 1e-05, |
| "loss": 0.245, |
| "step": 7530 |
| }, |
| { |
| "epoch": 2.41, |
| "grad_norm": 6.273867999705953, |
| "learning_rate": 1e-05, |
| "loss": 0.2636, |
| "step": 7540 |
| }, |
| { |
| "epoch": 2.41, |
| "grad_norm": 7.613204587704069, |
| "learning_rate": 1e-05, |
| "loss": 0.2643, |
| "step": 7550 |
| }, |
| { |
| "epoch": 2.41, |
| "grad_norm": 9.414876942207693, |
| "learning_rate": 1e-05, |
| "loss": 0.2495, |
| "step": 7560 |
| }, |
| { |
| "epoch": 2.42, |
| "grad_norm": 6.753200478685811, |
| "learning_rate": 1e-05, |
| "loss": 0.2346, |
| "step": 7570 |
| }, |
| { |
| "epoch": 2.42, |
| "grad_norm": 6.937737813086826, |
| "learning_rate": 1e-05, |
| "loss": 0.2975, |
| "step": 7580 |
| }, |
| { |
| "epoch": 2.42, |
| "grad_norm": 7.843785259273964, |
| "learning_rate": 1e-05, |
| "loss": 0.2881, |
| "step": 7590 |
| }, |
| { |
| "epoch": 2.43, |
| "grad_norm": 8.212977249468853, |
| "learning_rate": 1e-05, |
| "loss": 0.264, |
| "step": 7600 |
| }, |
| { |
| "epoch": 2.43, |
| "grad_norm": 8.061252704512842, |
| "learning_rate": 1e-05, |
| "loss": 0.2662, |
| "step": 7610 |
| }, |
| { |
| "epoch": 2.43, |
| "grad_norm": 8.53461722411201, |
| "learning_rate": 1e-05, |
| "loss": 0.2614, |
| "step": 7620 |
| }, |
| { |
| "epoch": 2.43, |
| "grad_norm": 6.09200320516641, |
| "learning_rate": 1e-05, |
| "loss": 0.2306, |
| "step": 7630 |
| }, |
| { |
| "epoch": 2.44, |
| "grad_norm": 5.944506556740013, |
| "learning_rate": 1e-05, |
| "loss": 0.2376, |
| "step": 7640 |
| }, |
| { |
| "epoch": 2.44, |
| "grad_norm": 6.702393107260854, |
| "learning_rate": 1e-05, |
| "loss": 0.2492, |
| "step": 7650 |
| }, |
| { |
| "epoch": 2.44, |
| "grad_norm": 6.063490010076286, |
| "learning_rate": 1e-05, |
| "loss": 0.2423, |
| "step": 7660 |
| }, |
| { |
| "epoch": 2.45, |
| "grad_norm": 8.188895303190845, |
| "learning_rate": 1e-05, |
| "loss": 0.2769, |
| "step": 7670 |
| }, |
| { |
| "epoch": 2.45, |
| "grad_norm": 6.302655692683647, |
| "learning_rate": 1e-05, |
| "loss": 0.2607, |
| "step": 7680 |
| }, |
| { |
| "epoch": 2.45, |
| "grad_norm": 6.361067862921245, |
| "learning_rate": 1e-05, |
| "loss": 0.297, |
| "step": 7690 |
| }, |
| { |
| "epoch": 2.46, |
| "grad_norm": 8.834032954694527, |
| "learning_rate": 1e-05, |
| "loss": 0.2955, |
| "step": 7700 |
| }, |
| { |
| "epoch": 2.46, |
| "grad_norm": 7.760248114955092, |
| "learning_rate": 1e-05, |
| "loss": 0.2561, |
| "step": 7710 |
| }, |
| { |
| "epoch": 2.46, |
| "grad_norm": 8.817076881062254, |
| "learning_rate": 1e-05, |
| "loss": 0.2613, |
| "step": 7720 |
| }, |
| { |
| "epoch": 2.47, |
| "grad_norm": 6.277451614564095, |
| "learning_rate": 1e-05, |
| "loss": 0.2699, |
| "step": 7730 |
| }, |
| { |
| "epoch": 2.47, |
| "grad_norm": 7.588536360474151, |
| "learning_rate": 1e-05, |
| "loss": 0.2544, |
| "step": 7740 |
| }, |
| { |
| "epoch": 2.47, |
| "grad_norm": 8.442495266114124, |
| "learning_rate": 1e-05, |
| "loss": 0.2297, |
| "step": 7750 |
| }, |
| { |
| "epoch": 2.48, |
| "grad_norm": 6.727583786617379, |
| "learning_rate": 1e-05, |
| "loss": 0.2823, |
| "step": 7760 |
| }, |
| { |
| "epoch": 2.48, |
| "grad_norm": 7.212077702633157, |
| "learning_rate": 1e-05, |
| "loss": 0.2737, |
| "step": 7770 |
| }, |
| { |
| "epoch": 2.48, |
| "grad_norm": 7.561510509985882, |
| "learning_rate": 1e-05, |
| "loss": 0.3093, |
| "step": 7780 |
| }, |
| { |
| "epoch": 2.49, |
| "grad_norm": 7.1531930582751, |
| "learning_rate": 1e-05, |
| "loss": 0.2451, |
| "step": 7790 |
| }, |
| { |
| "epoch": 2.49, |
| "grad_norm": 7.668515217489572, |
| "learning_rate": 1e-05, |
| "loss": 0.2417, |
| "step": 7800 |
| }, |
| { |
| "epoch": 2.49, |
| "grad_norm": 7.3324927223916, |
| "learning_rate": 1e-05, |
| "loss": 0.2722, |
| "step": 7810 |
| }, |
| { |
| "epoch": 2.5, |
| "grad_norm": 10.852012806194102, |
| "learning_rate": 1e-05, |
| "loss": 0.2855, |
| "step": 7820 |
| }, |
| { |
| "epoch": 2.5, |
| "grad_norm": 9.783381513209418, |
| "learning_rate": 1e-05, |
| "loss": 0.2793, |
| "step": 7830 |
| }, |
| { |
| "epoch": 2.5, |
| "grad_norm": 9.927387202209736, |
| "learning_rate": 1e-05, |
| "loss": 0.2765, |
| "step": 7840 |
| }, |
| { |
| "epoch": 2.5, |
| "grad_norm": 7.600466673727822, |
| "learning_rate": 1e-05, |
| "loss": 0.2759, |
| "step": 7850 |
| }, |
| { |
| "epoch": 2.51, |
| "grad_norm": 8.425658103527526, |
| "learning_rate": 1e-05, |
| "loss": 0.2671, |
| "step": 7860 |
| }, |
| { |
| "epoch": 2.51, |
| "grad_norm": 13.246545881217607, |
| "learning_rate": 1e-05, |
| "loss": 0.2558, |
| "step": 7870 |
| }, |
| { |
| "epoch": 2.51, |
| "grad_norm": 6.657352839701032, |
| "learning_rate": 1e-05, |
| "loss": 0.2779, |
| "step": 7880 |
| }, |
| { |
| "epoch": 2.52, |
| "grad_norm": 7.64235076449122, |
| "learning_rate": 1e-05, |
| "loss": 0.2444, |
| "step": 7890 |
| }, |
| { |
| "epoch": 2.52, |
| "grad_norm": 7.03652200357096, |
| "learning_rate": 1e-05, |
| "loss": 0.252, |
| "step": 7900 |
| }, |
| { |
| "epoch": 2.52, |
| "grad_norm": 8.318879167913883, |
| "learning_rate": 1e-05, |
| "loss": 0.2495, |
| "step": 7910 |
| }, |
| { |
| "epoch": 2.53, |
| "grad_norm": 11.989489720981926, |
| "learning_rate": 1e-05, |
| "loss": 0.2932, |
| "step": 7920 |
| }, |
| { |
| "epoch": 2.53, |
| "grad_norm": 6.8422766366724685, |
| "learning_rate": 1e-05, |
| "loss": 0.2719, |
| "step": 7930 |
| }, |
| { |
| "epoch": 2.53, |
| "grad_norm": 5.4963731077946205, |
| "learning_rate": 1e-05, |
| "loss": 0.295, |
| "step": 7940 |
| }, |
| { |
| "epoch": 2.54, |
| "grad_norm": 8.192891216884258, |
| "learning_rate": 1e-05, |
| "loss": 0.2734, |
| "step": 7950 |
| }, |
| { |
| "epoch": 2.54, |
| "grad_norm": 7.780041704237692, |
| "learning_rate": 1e-05, |
| "loss": 0.2765, |
| "step": 7960 |
| }, |
| { |
| "epoch": 2.54, |
| "grad_norm": 7.704404279786756, |
| "learning_rate": 1e-05, |
| "loss": 0.2682, |
| "step": 7970 |
| }, |
| { |
| "epoch": 2.55, |
| "grad_norm": 7.9138699326430055, |
| "learning_rate": 1e-05, |
| "loss": 0.3001, |
| "step": 7980 |
| }, |
| { |
| "epoch": 2.55, |
| "grad_norm": 6.240966371875658, |
| "learning_rate": 1e-05, |
| "loss": 0.2455, |
| "step": 7990 |
| }, |
| { |
| "epoch": 2.55, |
| "grad_norm": 7.157592381005428, |
| "learning_rate": 1e-05, |
| "loss": 0.254, |
| "step": 8000 |
| }, |
| { |
| "epoch": 2.55, |
| "eval_loss": 0.8203125, |
| "eval_runtime": 54.8211, |
| "eval_samples_per_second": 46.223, |
| "eval_steps_per_second": 2.9, |
| "step": 8000 |
| }, |
| { |
| "epoch": 2.56, |
| "grad_norm": 7.902980445031531, |
| "learning_rate": 1e-05, |
| "loss": 0.2613, |
| "step": 8010 |
| }, |
| { |
| "epoch": 2.56, |
| "grad_norm": 7.11778027340742, |
| "learning_rate": 1e-05, |
| "loss": 0.28, |
| "step": 8020 |
| }, |
| { |
| "epoch": 2.56, |
| "grad_norm": 8.948948923120149, |
| "learning_rate": 1e-05, |
| "loss": 0.2668, |
| "step": 8030 |
| }, |
| { |
| "epoch": 2.57, |
| "grad_norm": 10.890877894526069, |
| "learning_rate": 1e-05, |
| "loss": 0.2638, |
| "step": 8040 |
| }, |
| { |
| "epoch": 2.57, |
| "grad_norm": 6.702584625017246, |
| "learning_rate": 1e-05, |
| "loss": 0.2568, |
| "step": 8050 |
| }, |
| { |
| "epoch": 2.57, |
| "grad_norm": 6.826152315780276, |
| "learning_rate": 1e-05, |
| "loss": 0.2603, |
| "step": 8060 |
| }, |
| { |
| "epoch": 2.57, |
| "grad_norm": 9.151373804904368, |
| "learning_rate": 1e-05, |
| "loss": 0.2782, |
| "step": 8070 |
| }, |
| { |
| "epoch": 2.58, |
| "grad_norm": 7.241740816861216, |
| "learning_rate": 1e-05, |
| "loss": 0.2458, |
| "step": 8080 |
| }, |
| { |
| "epoch": 2.58, |
| "grad_norm": 8.625157451574616, |
| "learning_rate": 1e-05, |
| "loss": 0.2319, |
| "step": 8090 |
| }, |
| { |
| "epoch": 2.58, |
| "grad_norm": 8.643523297540025, |
| "learning_rate": 1e-05, |
| "loss": 0.283, |
| "step": 8100 |
| }, |
| { |
| "epoch": 2.59, |
| "grad_norm": 7.045022048428365, |
| "learning_rate": 1e-05, |
| "loss": 0.2612, |
| "step": 8110 |
| }, |
| { |
| "epoch": 2.59, |
| "grad_norm": 8.620600808573684, |
| "learning_rate": 1e-05, |
| "loss": 0.2887, |
| "step": 8120 |
| }, |
| { |
| "epoch": 2.59, |
| "grad_norm": 7.577563707467108, |
| "learning_rate": 1e-05, |
| "loss": 0.295, |
| "step": 8130 |
| }, |
| { |
| "epoch": 2.6, |
| "grad_norm": 7.791884420959403, |
| "learning_rate": 1e-05, |
| "loss": 0.2907, |
| "step": 8140 |
| }, |
| { |
| "epoch": 2.6, |
| "grad_norm": 7.622334999508127, |
| "learning_rate": 1e-05, |
| "loss": 0.2698, |
| "step": 8150 |
| }, |
| { |
| "epoch": 2.6, |
| "grad_norm": 8.083735452724124, |
| "learning_rate": 1e-05, |
| "loss": 0.2662, |
| "step": 8160 |
| }, |
| { |
| "epoch": 2.61, |
| "grad_norm": 7.183093503177487, |
| "learning_rate": 1e-05, |
| "loss": 0.3025, |
| "step": 8170 |
| }, |
| { |
| "epoch": 2.61, |
| "grad_norm": 8.200078024144085, |
| "learning_rate": 1e-05, |
| "loss": 0.2577, |
| "step": 8180 |
| }, |
| { |
| "epoch": 2.61, |
| "grad_norm": 8.422570251189981, |
| "learning_rate": 1e-05, |
| "loss": 0.2847, |
| "step": 8190 |
| }, |
| { |
| "epoch": 2.62, |
| "grad_norm": 9.517442852196393, |
| "learning_rate": 1e-05, |
| "loss": 0.2576, |
| "step": 8200 |
| }, |
| { |
| "epoch": 2.62, |
| "grad_norm": 6.076056042471135, |
| "learning_rate": 1e-05, |
| "loss": 0.2755, |
| "step": 8210 |
| }, |
| { |
| "epoch": 2.62, |
| "grad_norm": 8.955966891215207, |
| "learning_rate": 1e-05, |
| "loss": 0.2402, |
| "step": 8220 |
| }, |
| { |
| "epoch": 2.63, |
| "grad_norm": 8.094532814611131, |
| "learning_rate": 1e-05, |
| "loss": 0.2423, |
| "step": 8230 |
| }, |
| { |
| "epoch": 2.63, |
| "grad_norm": 6.946463627889644, |
| "learning_rate": 1e-05, |
| "loss": 0.3125, |
| "step": 8240 |
| }, |
| { |
| "epoch": 2.63, |
| "grad_norm": 7.399958131001565, |
| "learning_rate": 1e-05, |
| "loss": 0.2721, |
| "step": 8250 |
| }, |
| { |
| "epoch": 2.64, |
| "grad_norm": 8.7854806069235, |
| "learning_rate": 1e-05, |
| "loss": 0.2822, |
| "step": 8260 |
| }, |
| { |
| "epoch": 2.64, |
| "grad_norm": 6.691988855620577, |
| "learning_rate": 1e-05, |
| "loss": 0.256, |
| "step": 8270 |
| }, |
| { |
| "epoch": 2.64, |
| "grad_norm": 8.044413780917369, |
| "learning_rate": 1e-05, |
| "loss": 0.258, |
| "step": 8280 |
| }, |
| { |
| "epoch": 2.65, |
| "grad_norm": 8.320253355863734, |
| "learning_rate": 1e-05, |
| "loss": 0.2965, |
| "step": 8290 |
| }, |
| { |
| "epoch": 2.65, |
| "grad_norm": 7.03445483602588, |
| "learning_rate": 1e-05, |
| "loss": 0.2412, |
| "step": 8300 |
| }, |
| { |
| "epoch": 2.65, |
| "grad_norm": 7.9766521691672665, |
| "learning_rate": 1e-05, |
| "loss": 0.3015, |
| "step": 8310 |
| }, |
| { |
| "epoch": 2.65, |
| "grad_norm": 7.280158882084145, |
| "learning_rate": 1e-05, |
| "loss": 0.2634, |
| "step": 8320 |
| }, |
| { |
| "epoch": 2.66, |
| "grad_norm": 8.51747220504252, |
| "learning_rate": 1e-05, |
| "loss": 0.2591, |
| "step": 8330 |
| }, |
| { |
| "epoch": 2.66, |
| "grad_norm": 9.183716702175667, |
| "learning_rate": 1e-05, |
| "loss": 0.2995, |
| "step": 8340 |
| }, |
| { |
| "epoch": 2.66, |
| "grad_norm": 8.704665973878427, |
| "learning_rate": 1e-05, |
| "loss": 0.2537, |
| "step": 8350 |
| }, |
| { |
| "epoch": 2.67, |
| "grad_norm": 5.6645396485444115, |
| "learning_rate": 1e-05, |
| "loss": 0.2777, |
| "step": 8360 |
| }, |
| { |
| "epoch": 2.67, |
| "grad_norm": 6.806610921742937, |
| "learning_rate": 1e-05, |
| "loss": 0.2645, |
| "step": 8370 |
| }, |
| { |
| "epoch": 2.67, |
| "grad_norm": 7.946734485891536, |
| "learning_rate": 1e-05, |
| "loss": 0.2608, |
| "step": 8380 |
| }, |
| { |
| "epoch": 2.68, |
| "grad_norm": 6.345685391153275, |
| "learning_rate": 1e-05, |
| "loss": 0.2323, |
| "step": 8390 |
| }, |
| { |
| "epoch": 2.68, |
| "grad_norm": 8.054206781999518, |
| "learning_rate": 1e-05, |
| "loss": 0.2701, |
| "step": 8400 |
| }, |
| { |
| "epoch": 2.68, |
| "grad_norm": 8.35214896502597, |
| "learning_rate": 1e-05, |
| "loss": 0.2555, |
| "step": 8410 |
| }, |
| { |
| "epoch": 2.69, |
| "grad_norm": 8.045889844124705, |
| "learning_rate": 1e-05, |
| "loss": 0.2548, |
| "step": 8420 |
| }, |
| { |
| "epoch": 2.69, |
| "grad_norm": 7.113160586394211, |
| "learning_rate": 1e-05, |
| "loss": 0.2882, |
| "step": 8430 |
| }, |
| { |
| "epoch": 2.69, |
| "grad_norm": 8.83937250180045, |
| "learning_rate": 1e-05, |
| "loss": 0.2989, |
| "step": 8440 |
| }, |
| { |
| "epoch": 2.7, |
| "grad_norm": 4.963570636399205, |
| "learning_rate": 1e-05, |
| "loss": 0.2694, |
| "step": 8450 |
| }, |
| { |
| "epoch": 2.7, |
| "grad_norm": 8.91857998959261, |
| "learning_rate": 1e-05, |
| "loss": 0.2791, |
| "step": 8460 |
| }, |
| { |
| "epoch": 2.7, |
| "grad_norm": 5.672085172263902, |
| "learning_rate": 1e-05, |
| "loss": 0.2761, |
| "step": 8470 |
| }, |
| { |
| "epoch": 2.71, |
| "grad_norm": 7.393285292712138, |
| "learning_rate": 1e-05, |
| "loss": 0.2652, |
| "step": 8480 |
| }, |
| { |
| "epoch": 2.71, |
| "grad_norm": 8.148153327529236, |
| "learning_rate": 1e-05, |
| "loss": 0.267, |
| "step": 8490 |
| }, |
| { |
| "epoch": 2.71, |
| "grad_norm": 9.461041526317032, |
| "learning_rate": 1e-05, |
| "loss": 0.2738, |
| "step": 8500 |
| }, |
| { |
| "epoch": 2.71, |
| "eval_loss": 0.8212890625, |
| "eval_runtime": 55.0068, |
| "eval_samples_per_second": 46.067, |
| "eval_steps_per_second": 2.891, |
| "step": 8500 |
| }, |
| { |
| "epoch": 2.72, |
| "grad_norm": 8.428314408269808, |
| "learning_rate": 1e-05, |
| "loss": 0.2894, |
| "step": 8510 |
| }, |
| { |
| "epoch": 2.72, |
| "grad_norm": 7.146408043115168, |
| "learning_rate": 1e-05, |
| "loss": 0.2902, |
| "step": 8520 |
| }, |
| { |
| "epoch": 2.72, |
| "grad_norm": 7.0913972019710645, |
| "learning_rate": 1e-05, |
| "loss": 0.2852, |
| "step": 8530 |
| }, |
| { |
| "epoch": 2.72, |
| "grad_norm": 8.323640857902728, |
| "learning_rate": 1e-05, |
| "loss": 0.3116, |
| "step": 8540 |
| }, |
| { |
| "epoch": 2.73, |
| "grad_norm": 7.127825260508, |
| "learning_rate": 1e-05, |
| "loss": 0.2573, |
| "step": 8550 |
| }, |
| { |
| "epoch": 2.73, |
| "grad_norm": 7.362289855755885, |
| "learning_rate": 1e-05, |
| "loss": 0.2741, |
| "step": 8560 |
| }, |
| { |
| "epoch": 2.73, |
| "grad_norm": 6.411656921765378, |
| "learning_rate": 1e-05, |
| "loss": 0.2769, |
| "step": 8570 |
| }, |
| { |
| "epoch": 2.74, |
| "grad_norm": 12.408799377939557, |
| "learning_rate": 1e-05, |
| "loss": 0.2824, |
| "step": 8580 |
| }, |
| { |
| "epoch": 2.74, |
| "grad_norm": 7.594353294740246, |
| "learning_rate": 1e-05, |
| "loss": 0.2913, |
| "step": 8590 |
| }, |
| { |
| "epoch": 2.74, |
| "grad_norm": 6.802492683403906, |
| "learning_rate": 1e-05, |
| "loss": 0.2924, |
| "step": 8600 |
| }, |
| { |
| "epoch": 2.75, |
| "grad_norm": 6.759303992291298, |
| "learning_rate": 1e-05, |
| "loss": 0.2608, |
| "step": 8610 |
| }, |
| { |
| "epoch": 2.75, |
| "grad_norm": 5.31710382575628, |
| "learning_rate": 1e-05, |
| "loss": 0.2811, |
| "step": 8620 |
| }, |
| { |
| "epoch": 2.75, |
| "grad_norm": 7.608321161946328, |
| "learning_rate": 1e-05, |
| "loss": 0.2716, |
| "step": 8630 |
| }, |
| { |
| "epoch": 2.76, |
| "grad_norm": 7.014472985785215, |
| "learning_rate": 1e-05, |
| "loss": 0.253, |
| "step": 8640 |
| }, |
| { |
| "epoch": 2.76, |
| "grad_norm": 7.460089734230643, |
| "learning_rate": 1e-05, |
| "loss": 0.2397, |
| "step": 8650 |
| }, |
| { |
| "epoch": 2.76, |
| "grad_norm": 20.654451185150517, |
| "learning_rate": 1e-05, |
| "loss": 0.3041, |
| "step": 8660 |
| }, |
| { |
| "epoch": 2.77, |
| "grad_norm": 8.984193697219586, |
| "learning_rate": 1e-05, |
| "loss": 0.2848, |
| "step": 8670 |
| }, |
| { |
| "epoch": 2.77, |
| "grad_norm": 6.495205504798178, |
| "learning_rate": 1e-05, |
| "loss": 0.2831, |
| "step": 8680 |
| }, |
| { |
| "epoch": 2.77, |
| "grad_norm": 6.124251962253673, |
| "learning_rate": 1e-05, |
| "loss": 0.2689, |
| "step": 8690 |
| }, |
| { |
| "epoch": 2.78, |
| "grad_norm": 5.719079847151612, |
| "learning_rate": 1e-05, |
| "loss": 0.2662, |
| "step": 8700 |
| }, |
| { |
| "epoch": 2.78, |
| "grad_norm": 8.772510455808282, |
| "learning_rate": 1e-05, |
| "loss": 0.2933, |
| "step": 8710 |
| }, |
| { |
| "epoch": 2.78, |
| "grad_norm": 5.482748023953412, |
| "learning_rate": 1e-05, |
| "loss": 0.2779, |
| "step": 8720 |
| }, |
| { |
| "epoch": 2.79, |
| "grad_norm": 6.790797109656677, |
| "learning_rate": 1e-05, |
| "loss": 0.2713, |
| "step": 8730 |
| }, |
| { |
| "epoch": 2.79, |
| "grad_norm": 7.383311356740698, |
| "learning_rate": 1e-05, |
| "loss": 0.2838, |
| "step": 8740 |
| }, |
| { |
| "epoch": 2.79, |
| "grad_norm": 9.27542578858622, |
| "learning_rate": 1e-05, |
| "loss": 0.2477, |
| "step": 8750 |
| }, |
| { |
| "epoch": 2.8, |
| "grad_norm": 7.231409902956014, |
| "learning_rate": 1e-05, |
| "loss": 0.2766, |
| "step": 8760 |
| }, |
| { |
| "epoch": 2.8, |
| "grad_norm": 5.82061135977678, |
| "learning_rate": 1e-05, |
| "loss": 0.2914, |
| "step": 8770 |
| }, |
| { |
| "epoch": 2.8, |
| "grad_norm": 8.377829458053121, |
| "learning_rate": 1e-05, |
| "loss": 0.2501, |
| "step": 8780 |
| }, |
| { |
| "epoch": 2.8, |
| "grad_norm": 7.669341933606825, |
| "learning_rate": 1e-05, |
| "loss": 0.2787, |
| "step": 8790 |
| }, |
| { |
| "epoch": 2.81, |
| "grad_norm": 7.9010933807497885, |
| "learning_rate": 1e-05, |
| "loss": 0.2848, |
| "step": 8800 |
| }, |
| { |
| "epoch": 2.81, |
| "grad_norm": 10.234888241543945, |
| "learning_rate": 1e-05, |
| "loss": 0.2686, |
| "step": 8810 |
| }, |
| { |
| "epoch": 2.81, |
| "grad_norm": 6.69997063246027, |
| "learning_rate": 1e-05, |
| "loss": 0.2738, |
| "step": 8820 |
| }, |
| { |
| "epoch": 2.82, |
| "grad_norm": 12.074802588197677, |
| "learning_rate": 1e-05, |
| "loss": 0.2683, |
| "step": 8830 |
| }, |
| { |
| "epoch": 2.82, |
| "grad_norm": 6.1248589324276415, |
| "learning_rate": 1e-05, |
| "loss": 0.2689, |
| "step": 8840 |
| }, |
| { |
| "epoch": 2.82, |
| "grad_norm": 7.038914550852228, |
| "learning_rate": 1e-05, |
| "loss": 0.2892, |
| "step": 8850 |
| }, |
| { |
| "epoch": 2.83, |
| "grad_norm": 7.795187360409416, |
| "learning_rate": 1e-05, |
| "loss": 0.2428, |
| "step": 8860 |
| }, |
| { |
| "epoch": 2.83, |
| "grad_norm": 6.47005313149854, |
| "learning_rate": 1e-05, |
| "loss": 0.2426, |
| "step": 8870 |
| }, |
| { |
| "epoch": 2.83, |
| "grad_norm": 7.636542916773699, |
| "learning_rate": 1e-05, |
| "loss": 0.2803, |
| "step": 8880 |
| }, |
| { |
| "epoch": 2.84, |
| "grad_norm": 8.724479734107895, |
| "learning_rate": 1e-05, |
| "loss": 0.2622, |
| "step": 8890 |
| }, |
| { |
| "epoch": 2.84, |
| "grad_norm": 8.134316604879476, |
| "learning_rate": 1e-05, |
| "loss": 0.242, |
| "step": 8900 |
| }, |
| { |
| "epoch": 2.84, |
| "grad_norm": 8.259195462778612, |
| "learning_rate": 1e-05, |
| "loss": 0.3029, |
| "step": 8910 |
| }, |
| { |
| "epoch": 2.85, |
| "grad_norm": 7.026164882098872, |
| "learning_rate": 1e-05, |
| "loss": 0.2926, |
| "step": 8920 |
| }, |
| { |
| "epoch": 2.85, |
| "grad_norm": 6.835588997812953, |
| "learning_rate": 1e-05, |
| "loss": 0.302, |
| "step": 8930 |
| }, |
| { |
| "epoch": 2.85, |
| "grad_norm": 7.42106338078716, |
| "learning_rate": 1e-05, |
| "loss": 0.3303, |
| "step": 8940 |
| }, |
| { |
| "epoch": 2.86, |
| "grad_norm": 7.297721339936946, |
| "learning_rate": 1e-05, |
| "loss": 0.2801, |
| "step": 8950 |
| }, |
| { |
| "epoch": 2.86, |
| "grad_norm": 6.07650083989753, |
| "learning_rate": 1e-05, |
| "loss": 0.2792, |
| "step": 8960 |
| }, |
| { |
| "epoch": 2.86, |
| "grad_norm": 6.972211722087004, |
| "learning_rate": 1e-05, |
| "loss": 0.2608, |
| "step": 8970 |
| }, |
| { |
| "epoch": 2.87, |
| "grad_norm": 6.608233847075685, |
| "learning_rate": 1e-05, |
| "loss": 0.2808, |
| "step": 8980 |
| }, |
| { |
| "epoch": 2.87, |
| "grad_norm": 8.701640305502616, |
| "learning_rate": 1e-05, |
| "loss": 0.277, |
| "step": 8990 |
| }, |
| { |
| "epoch": 2.87, |
| "grad_norm": 8.249723487612801, |
| "learning_rate": 1e-05, |
| "loss": 0.2855, |
| "step": 9000 |
| }, |
| { |
| "epoch": 2.87, |
| "eval_loss": 0.83056640625, |
| "eval_runtime": 55.0008, |
| "eval_samples_per_second": 46.072, |
| "eval_steps_per_second": 2.891, |
| "step": 9000 |
| }, |
| { |
| "epoch": 2.87, |
| "grad_norm": 5.667885294009988, |
| "learning_rate": 1e-05, |
| "loss": 0.2664, |
| "step": 9010 |
| }, |
| { |
| "epoch": 2.88, |
| "grad_norm": 8.420109664674014, |
| "learning_rate": 1e-05, |
| "loss": 0.2784, |
| "step": 9020 |
| }, |
| { |
| "epoch": 2.88, |
| "grad_norm": 6.696858070768674, |
| "learning_rate": 1e-05, |
| "loss": 0.2552, |
| "step": 9030 |
| }, |
| { |
| "epoch": 2.88, |
| "grad_norm": 7.552002990996852, |
| "learning_rate": 1e-05, |
| "loss": 0.2777, |
| "step": 9040 |
| }, |
| { |
| "epoch": 2.89, |
| "grad_norm": 7.434429656922612, |
| "learning_rate": 1e-05, |
| "loss": 0.2425, |
| "step": 9050 |
| }, |
| { |
| "epoch": 2.89, |
| "grad_norm": 7.316812401133359, |
| "learning_rate": 1e-05, |
| "loss": 0.3112, |
| "step": 9060 |
| }, |
| { |
| "epoch": 2.89, |
| "grad_norm": 11.263658497423563, |
| "learning_rate": 1e-05, |
| "loss": 0.2755, |
| "step": 9070 |
| }, |
| { |
| "epoch": 2.9, |
| "grad_norm": 6.830732951791587, |
| "learning_rate": 1e-05, |
| "loss": 0.3023, |
| "step": 9080 |
| }, |
| { |
| "epoch": 2.9, |
| "grad_norm": 8.678333303872666, |
| "learning_rate": 1e-05, |
| "loss": 0.2679, |
| "step": 9090 |
| }, |
| { |
| "epoch": 2.9, |
| "grad_norm": 6.25324622727318, |
| "learning_rate": 1e-05, |
| "loss": 0.2426, |
| "step": 9100 |
| }, |
| { |
| "epoch": 2.91, |
| "grad_norm": 9.624314320898966, |
| "learning_rate": 1e-05, |
| "loss": 0.2902, |
| "step": 9110 |
| }, |
| { |
| "epoch": 2.91, |
| "grad_norm": 9.19992740436605, |
| "learning_rate": 1e-05, |
| "loss": 0.3094, |
| "step": 9120 |
| }, |
| { |
| "epoch": 2.91, |
| "grad_norm": 6.002421843986314, |
| "learning_rate": 1e-05, |
| "loss": 0.281, |
| "step": 9130 |
| }, |
| { |
| "epoch": 2.92, |
| "grad_norm": 7.030888120027772, |
| "learning_rate": 1e-05, |
| "loss": 0.3191, |
| "step": 9140 |
| }, |
| { |
| "epoch": 2.92, |
| "grad_norm": 8.221804439252049, |
| "learning_rate": 1e-05, |
| "loss": 0.2978, |
| "step": 9150 |
| }, |
| { |
| "epoch": 2.92, |
| "grad_norm": 8.09771042477657, |
| "learning_rate": 1e-05, |
| "loss": 0.2877, |
| "step": 9160 |
| }, |
| { |
| "epoch": 2.93, |
| "grad_norm": 10.335370960120434, |
| "learning_rate": 1e-05, |
| "loss": 0.2894, |
| "step": 9170 |
| }, |
| { |
| "epoch": 2.93, |
| "grad_norm": 7.941265504624651, |
| "learning_rate": 1e-05, |
| "loss": 0.2927, |
| "step": 9180 |
| }, |
| { |
| "epoch": 2.93, |
| "grad_norm": 6.939518815214694, |
| "learning_rate": 1e-05, |
| "loss": 0.2914, |
| "step": 9190 |
| }, |
| { |
| "epoch": 2.94, |
| "grad_norm": 10.699928240669129, |
| "learning_rate": 1e-05, |
| "loss": 0.2907, |
| "step": 9200 |
| }, |
| { |
| "epoch": 2.94, |
| "grad_norm": 6.469878314847625, |
| "learning_rate": 1e-05, |
| "loss": 0.2733, |
| "step": 9210 |
| }, |
| { |
| "epoch": 2.94, |
| "grad_norm": 8.391758790658008, |
| "learning_rate": 1e-05, |
| "loss": 0.2775, |
| "step": 9220 |
| }, |
| { |
| "epoch": 2.95, |
| "grad_norm": 8.354247843537442, |
| "learning_rate": 1e-05, |
| "loss": 0.2703, |
| "step": 9230 |
| }, |
| { |
| "epoch": 2.95, |
| "grad_norm": 8.898978065368526, |
| "learning_rate": 1e-05, |
| "loss": 0.2807, |
| "step": 9240 |
| }, |
| { |
| "epoch": 2.95, |
| "grad_norm": 9.638294105122334, |
| "learning_rate": 1e-05, |
| "loss": 0.3279, |
| "step": 9250 |
| }, |
| { |
| "epoch": 2.95, |
| "grad_norm": 8.363216480073808, |
| "learning_rate": 1e-05, |
| "loss": 0.2672, |
| "step": 9260 |
| }, |
| { |
| "epoch": 2.96, |
| "grad_norm": 8.395529565608815, |
| "learning_rate": 1e-05, |
| "loss": 0.278, |
| "step": 9270 |
| }, |
| { |
| "epoch": 2.96, |
| "grad_norm": 9.08585269065105, |
| "learning_rate": 1e-05, |
| "loss": 0.2961, |
| "step": 9280 |
| }, |
| { |
| "epoch": 2.96, |
| "grad_norm": 5.523148460325408, |
| "learning_rate": 1e-05, |
| "loss": 0.2875, |
| "step": 9290 |
| }, |
| { |
| "epoch": 2.97, |
| "grad_norm": 5.7479735410926835, |
| "learning_rate": 1e-05, |
| "loss": 0.2898, |
| "step": 9300 |
| }, |
| { |
| "epoch": 2.97, |
| "grad_norm": 6.186353172602731, |
| "learning_rate": 1e-05, |
| "loss": 0.3198, |
| "step": 9310 |
| }, |
| { |
| "epoch": 2.97, |
| "grad_norm": 5.454119135351055, |
| "learning_rate": 1e-05, |
| "loss": 0.2543, |
| "step": 9320 |
| }, |
| { |
| "epoch": 2.98, |
| "grad_norm": 5.655476554130632, |
| "learning_rate": 1e-05, |
| "loss": 0.3023, |
| "step": 9330 |
| }, |
| { |
| "epoch": 2.98, |
| "grad_norm": 7.015675024181342, |
| "learning_rate": 1e-05, |
| "loss": 0.2717, |
| "step": 9340 |
| }, |
| { |
| "epoch": 2.98, |
| "grad_norm": 9.268673991781023, |
| "learning_rate": 1e-05, |
| "loss": 0.2564, |
| "step": 9350 |
| }, |
| { |
| "epoch": 2.99, |
| "grad_norm": 5.851098647023247, |
| "learning_rate": 1e-05, |
| "loss": 0.2934, |
| "step": 9360 |
| }, |
| { |
| "epoch": 2.99, |
| "grad_norm": 10.314328303247283, |
| "learning_rate": 1e-05, |
| "loss": 0.3368, |
| "step": 9370 |
| }, |
| { |
| "epoch": 2.99, |
| "grad_norm": 5.346623412232383, |
| "learning_rate": 1e-05, |
| "loss": 0.2825, |
| "step": 9380 |
| }, |
| { |
| "epoch": 3.0, |
| "grad_norm": 6.474678897628251, |
| "learning_rate": 1e-05, |
| "loss": 0.2561, |
| "step": 9390 |
| }, |
| { |
| "epoch": 3.0, |
| "grad_norm": 10.212206877630488, |
| "learning_rate": 1e-05, |
| "loss": 0.3085, |
| "step": 9400 |
| }, |
| { |
| "epoch": 3.0, |
| "grad_norm": 6.705019525304665, |
| "learning_rate": 1e-05, |
| "loss": 0.172, |
| "step": 9410 |
| }, |
| { |
| "epoch": 3.01, |
| "grad_norm": 7.719380341402061, |
| "learning_rate": 1e-05, |
| "loss": 0.1293, |
| "step": 9420 |
| }, |
| { |
| "epoch": 3.01, |
| "grad_norm": 6.126627803066275, |
| "learning_rate": 1e-05, |
| "loss": 0.1157, |
| "step": 9430 |
| }, |
| { |
| "epoch": 3.01, |
| "grad_norm": 5.774757822746937, |
| "learning_rate": 1e-05, |
| "loss": 0.1486, |
| "step": 9440 |
| }, |
| { |
| "epoch": 3.02, |
| "grad_norm": 4.320949514378818, |
| "learning_rate": 1e-05, |
| "loss": 0.1612, |
| "step": 9450 |
| }, |
| { |
| "epoch": 3.02, |
| "grad_norm": 4.477665896683697, |
| "learning_rate": 1e-05, |
| "loss": 0.1524, |
| "step": 9460 |
| }, |
| { |
| "epoch": 3.02, |
| "grad_norm": 7.314700586184464, |
| "learning_rate": 1e-05, |
| "loss": 0.1559, |
| "step": 9470 |
| }, |
| { |
| "epoch": 3.02, |
| "grad_norm": 6.57195422042718, |
| "learning_rate": 1e-05, |
| "loss": 0.1518, |
| "step": 9480 |
| }, |
| { |
| "epoch": 3.03, |
| "grad_norm": 8.578212605569277, |
| "learning_rate": 1e-05, |
| "loss": 0.1613, |
| "step": 9490 |
| }, |
| { |
| "epoch": 3.03, |
| "grad_norm": 6.453321846460465, |
| "learning_rate": 1e-05, |
| "loss": 0.1577, |
| "step": 9500 |
| }, |
| { |
| "epoch": 3.03, |
| "eval_loss": 0.8720703125, |
| "eval_runtime": 54.7661, |
| "eval_samples_per_second": 46.269, |
| "eval_steps_per_second": 2.903, |
| "step": 9500 |
| }, |
| { |
| "epoch": 3.03, |
| "grad_norm": 7.252586429558278, |
| "learning_rate": 1e-05, |
| "loss": 0.1627, |
| "step": 9510 |
| }, |
| { |
| "epoch": 3.04, |
| "grad_norm": 8.600184469684221, |
| "learning_rate": 1e-05, |
| "loss": 0.1324, |
| "step": 9520 |
| }, |
| { |
| "epoch": 3.04, |
| "grad_norm": 6.506099186889074, |
| "learning_rate": 1e-05, |
| "loss": 0.1564, |
| "step": 9530 |
| }, |
| { |
| "epoch": 3.04, |
| "grad_norm": 5.142893745655413, |
| "learning_rate": 1e-05, |
| "loss": 0.1552, |
| "step": 9540 |
| }, |
| { |
| "epoch": 3.05, |
| "grad_norm": 6.524544371851103, |
| "learning_rate": 1e-05, |
| "loss": 0.1593, |
| "step": 9550 |
| }, |
| { |
| "epoch": 3.05, |
| "grad_norm": 7.224177446367948, |
| "learning_rate": 1e-05, |
| "loss": 0.1726, |
| "step": 9560 |
| }, |
| { |
| "epoch": 3.05, |
| "grad_norm": 6.169890342986853, |
| "learning_rate": 1e-05, |
| "loss": 0.1538, |
| "step": 9570 |
| }, |
| { |
| "epoch": 3.06, |
| "grad_norm": 6.538513525156336, |
| "learning_rate": 1e-05, |
| "loss": 0.1494, |
| "step": 9580 |
| }, |
| { |
| "epoch": 3.06, |
| "grad_norm": 6.556330805626917, |
| "learning_rate": 1e-05, |
| "loss": 0.1568, |
| "step": 9590 |
| }, |
| { |
| "epoch": 3.06, |
| "grad_norm": 5.515248747790076, |
| "learning_rate": 1e-05, |
| "loss": 0.1552, |
| "step": 9600 |
| }, |
| { |
| "epoch": 3.07, |
| "grad_norm": 5.349545553995904, |
| "learning_rate": 1e-05, |
| "loss": 0.1708, |
| "step": 9610 |
| }, |
| { |
| "epoch": 3.07, |
| "grad_norm": 12.46186749566421, |
| "learning_rate": 1e-05, |
| "loss": 0.1768, |
| "step": 9620 |
| }, |
| { |
| "epoch": 3.07, |
| "grad_norm": 7.084233155784582, |
| "learning_rate": 1e-05, |
| "loss": 0.1725, |
| "step": 9630 |
| }, |
| { |
| "epoch": 3.08, |
| "grad_norm": 6.338881555334394, |
| "learning_rate": 1e-05, |
| "loss": 0.1587, |
| "step": 9640 |
| }, |
| { |
| "epoch": 3.08, |
| "grad_norm": 5.237533117658478, |
| "learning_rate": 1e-05, |
| "loss": 0.1551, |
| "step": 9650 |
| }, |
| { |
| "epoch": 3.08, |
| "grad_norm": 5.499362388411692, |
| "learning_rate": 1e-05, |
| "loss": 0.1593, |
| "step": 9660 |
| }, |
| { |
| "epoch": 3.09, |
| "grad_norm": 6.81638526366755, |
| "learning_rate": 1e-05, |
| "loss": 0.1527, |
| "step": 9670 |
| }, |
| { |
| "epoch": 3.09, |
| "grad_norm": 7.0113645671339695, |
| "learning_rate": 1e-05, |
| "loss": 0.1502, |
| "step": 9680 |
| }, |
| { |
| "epoch": 3.09, |
| "grad_norm": 5.37548488825087, |
| "learning_rate": 1e-05, |
| "loss": 0.156, |
| "step": 9690 |
| }, |
| { |
| "epoch": 3.1, |
| "grad_norm": 6.808004313269212, |
| "learning_rate": 1e-05, |
| "loss": 0.1425, |
| "step": 9700 |
| }, |
| { |
| "epoch": 3.1, |
| "grad_norm": 4.85942641841265, |
| "learning_rate": 1e-05, |
| "loss": 0.1553, |
| "step": 9710 |
| }, |
| { |
| "epoch": 3.1, |
| "grad_norm": 6.245966105442138, |
| "learning_rate": 1e-05, |
| "loss": 0.1702, |
| "step": 9720 |
| }, |
| { |
| "epoch": 3.1, |
| "grad_norm": 4.121722682310522, |
| "learning_rate": 1e-05, |
| "loss": 0.1503, |
| "step": 9730 |
| }, |
| { |
| "epoch": 3.11, |
| "grad_norm": 6.256522780348953, |
| "learning_rate": 1e-05, |
| "loss": 0.1489, |
| "step": 9740 |
| }, |
| { |
| "epoch": 3.11, |
| "grad_norm": 7.245202154125887, |
| "learning_rate": 1e-05, |
| "loss": 0.1518, |
| "step": 9750 |
| }, |
| { |
| "epoch": 3.11, |
| "grad_norm": 5.3823295121892345, |
| "learning_rate": 1e-05, |
| "loss": 0.1586, |
| "step": 9760 |
| }, |
| { |
| "epoch": 3.12, |
| "grad_norm": 6.625378004122464, |
| "learning_rate": 1e-05, |
| "loss": 0.1434, |
| "step": 9770 |
| }, |
| { |
| "epoch": 3.12, |
| "grad_norm": 5.027027134977083, |
| "learning_rate": 1e-05, |
| "loss": 0.167, |
| "step": 9780 |
| }, |
| { |
| "epoch": 3.12, |
| "grad_norm": 6.57671654634787, |
| "learning_rate": 1e-05, |
| "loss": 0.1429, |
| "step": 9790 |
| }, |
| { |
| "epoch": 3.13, |
| "grad_norm": 5.2123565800464675, |
| "learning_rate": 1e-05, |
| "loss": 0.1424, |
| "step": 9800 |
| }, |
| { |
| "epoch": 3.13, |
| "grad_norm": 7.0161246820591225, |
| "learning_rate": 1e-05, |
| "loss": 0.1639, |
| "step": 9810 |
| }, |
| { |
| "epoch": 3.13, |
| "grad_norm": 6.140100354794159, |
| "learning_rate": 1e-05, |
| "loss": 0.1635, |
| "step": 9820 |
| }, |
| { |
| "epoch": 3.14, |
| "grad_norm": 5.551535612170586, |
| "learning_rate": 1e-05, |
| "loss": 0.1523, |
| "step": 9830 |
| }, |
| { |
| "epoch": 3.14, |
| "grad_norm": 5.981723925271381, |
| "learning_rate": 1e-05, |
| "loss": 0.1743, |
| "step": 9840 |
| }, |
| { |
| "epoch": 3.14, |
| "grad_norm": 7.845039413674071, |
| "learning_rate": 1e-05, |
| "loss": 0.1588, |
| "step": 9850 |
| }, |
| { |
| "epoch": 3.15, |
| "grad_norm": 6.003410006126531, |
| "learning_rate": 1e-05, |
| "loss": 0.1827, |
| "step": 9860 |
| }, |
| { |
| "epoch": 3.15, |
| "grad_norm": 6.985217971622932, |
| "learning_rate": 1e-05, |
| "loss": 0.1505, |
| "step": 9870 |
| }, |
| { |
| "epoch": 3.15, |
| "grad_norm": 5.134001293733536, |
| "learning_rate": 1e-05, |
| "loss": 0.1899, |
| "step": 9880 |
| }, |
| { |
| "epoch": 3.16, |
| "grad_norm": 8.357919612646729, |
| "learning_rate": 1e-05, |
| "loss": 0.1559, |
| "step": 9890 |
| }, |
| { |
| "epoch": 3.16, |
| "grad_norm": 7.59002543971607, |
| "learning_rate": 1e-05, |
| "loss": 0.1469, |
| "step": 9900 |
| }, |
| { |
| "epoch": 3.16, |
| "grad_norm": 5.581092365768485, |
| "learning_rate": 1e-05, |
| "loss": 0.1531, |
| "step": 9910 |
| }, |
| { |
| "epoch": 3.17, |
| "grad_norm": 8.274152600558212, |
| "learning_rate": 1e-05, |
| "loss": 0.1755, |
| "step": 9920 |
| }, |
| { |
| "epoch": 3.17, |
| "grad_norm": 9.466415469996335, |
| "learning_rate": 1e-05, |
| "loss": 0.1606, |
| "step": 9930 |
| }, |
| { |
| "epoch": 3.17, |
| "grad_norm": 5.8246330047388035, |
| "learning_rate": 1e-05, |
| "loss": 0.1584, |
| "step": 9940 |
| }, |
| { |
| "epoch": 3.17, |
| "grad_norm": 6.754430376568705, |
| "learning_rate": 1e-05, |
| "loss": 0.1516, |
| "step": 9950 |
| }, |
| { |
| "epoch": 3.18, |
| "grad_norm": 7.188200012616978, |
| "learning_rate": 1e-05, |
| "loss": 0.185, |
| "step": 9960 |
| }, |
| { |
| "epoch": 3.18, |
| "grad_norm": 8.893207624119833, |
| "learning_rate": 1e-05, |
| "loss": 0.1548, |
| "step": 9970 |
| }, |
| { |
| "epoch": 3.18, |
| "grad_norm": 6.209871174187111, |
| "learning_rate": 1e-05, |
| "loss": 0.1706, |
| "step": 9980 |
| }, |
| { |
| "epoch": 3.19, |
| "grad_norm": 6.879373442485997, |
| "learning_rate": 1e-05, |
| "loss": 0.1769, |
| "step": 9990 |
| }, |
| { |
| "epoch": 3.19, |
| "grad_norm": 6.088040694833858, |
| "learning_rate": 1e-05, |
| "loss": 0.1542, |
| "step": 10000 |
| }, |
| { |
| "epoch": 3.19, |
| "eval_loss": 0.91748046875, |
| "eval_runtime": 55.0033, |
| "eval_samples_per_second": 46.07, |
| "eval_steps_per_second": 2.891, |
| "step": 10000 |
| }, |
| { |
| "epoch": 3.19, |
| "grad_norm": 6.086018984130375, |
| "learning_rate": 1e-05, |
| "loss": 0.1499, |
| "step": 10010 |
| }, |
| { |
| "epoch": 3.2, |
| "grad_norm": 6.93000594894352, |
| "learning_rate": 1e-05, |
| "loss": 0.1593, |
| "step": 10020 |
| }, |
| { |
| "epoch": 3.2, |
| "grad_norm": 8.25060061956074, |
| "learning_rate": 1e-05, |
| "loss": 0.175, |
| "step": 10030 |
| }, |
| { |
| "epoch": 3.2, |
| "grad_norm": 9.67088300458541, |
| "learning_rate": 1e-05, |
| "loss": 0.1786, |
| "step": 10040 |
| }, |
| { |
| "epoch": 3.21, |
| "grad_norm": 5.079767612939085, |
| "learning_rate": 1e-05, |
| "loss": 0.1628, |
| "step": 10050 |
| }, |
| { |
| "epoch": 3.21, |
| "grad_norm": 5.652054384492314, |
| "learning_rate": 1e-05, |
| "loss": 0.1741, |
| "step": 10060 |
| }, |
| { |
| "epoch": 3.21, |
| "grad_norm": 5.362922228572023, |
| "learning_rate": 1e-05, |
| "loss": 0.1598, |
| "step": 10070 |
| }, |
| { |
| "epoch": 3.22, |
| "grad_norm": 7.233566865212891, |
| "learning_rate": 1e-05, |
| "loss": 0.1672, |
| "step": 10080 |
| }, |
| { |
| "epoch": 3.22, |
| "grad_norm": 7.450572602457625, |
| "learning_rate": 1e-05, |
| "loss": 0.1694, |
| "step": 10090 |
| }, |
| { |
| "epoch": 3.22, |
| "grad_norm": 6.821253201520803, |
| "learning_rate": 1e-05, |
| "loss": 0.1653, |
| "step": 10100 |
| }, |
| { |
| "epoch": 3.23, |
| "grad_norm": 8.676207743625488, |
| "learning_rate": 1e-05, |
| "loss": 0.1396, |
| "step": 10110 |
| }, |
| { |
| "epoch": 3.23, |
| "grad_norm": 7.161732801321849, |
| "learning_rate": 1e-05, |
| "loss": 0.1506, |
| "step": 10120 |
| }, |
| { |
| "epoch": 3.23, |
| "grad_norm": 6.474710418234935, |
| "learning_rate": 1e-05, |
| "loss": 0.1906, |
| "step": 10130 |
| }, |
| { |
| "epoch": 3.24, |
| "grad_norm": 6.460973811018732, |
| "learning_rate": 1e-05, |
| "loss": 0.172, |
| "step": 10140 |
| }, |
| { |
| "epoch": 3.24, |
| "grad_norm": 6.973954109340658, |
| "learning_rate": 1e-05, |
| "loss": 0.1777, |
| "step": 10150 |
| }, |
| { |
| "epoch": 3.24, |
| "grad_norm": 7.04872313495515, |
| "learning_rate": 1e-05, |
| "loss": 0.1549, |
| "step": 10160 |
| }, |
| { |
| "epoch": 3.25, |
| "grad_norm": 7.353496171538013, |
| "learning_rate": 1e-05, |
| "loss": 0.1756, |
| "step": 10170 |
| }, |
| { |
| "epoch": 3.25, |
| "grad_norm": 8.877539472865505, |
| "learning_rate": 1e-05, |
| "loss": 0.1753, |
| "step": 10180 |
| }, |
| { |
| "epoch": 3.25, |
| "grad_norm": 8.335942076205399, |
| "learning_rate": 1e-05, |
| "loss": 0.1703, |
| "step": 10190 |
| }, |
| { |
| "epoch": 3.25, |
| "grad_norm": 6.640739314273428, |
| "learning_rate": 1e-05, |
| "loss": 0.1664, |
| "step": 10200 |
| }, |
| { |
| "epoch": 3.26, |
| "grad_norm": 8.143290429311316, |
| "learning_rate": 1e-05, |
| "loss": 0.1898, |
| "step": 10210 |
| }, |
| { |
| "epoch": 3.26, |
| "grad_norm": 7.633435195620788, |
| "learning_rate": 1e-05, |
| "loss": 0.1942, |
| "step": 10220 |
| }, |
| { |
| "epoch": 3.26, |
| "grad_norm": 6.565268087150161, |
| "learning_rate": 1e-05, |
| "loss": 0.1631, |
| "step": 10230 |
| }, |
| { |
| "epoch": 3.27, |
| "grad_norm": 5.643317174525972, |
| "learning_rate": 1e-05, |
| "loss": 0.1943, |
| "step": 10240 |
| }, |
| { |
| "epoch": 3.27, |
| "grad_norm": 6.874312487818384, |
| "learning_rate": 1e-05, |
| "loss": 0.1501, |
| "step": 10250 |
| }, |
| { |
| "epoch": 3.27, |
| "grad_norm": 6.25704132166907, |
| "learning_rate": 1e-05, |
| "loss": 0.1703, |
| "step": 10260 |
| }, |
| { |
| "epoch": 3.28, |
| "grad_norm": 6.8064722114208145, |
| "learning_rate": 1e-05, |
| "loss": 0.1747, |
| "step": 10270 |
| }, |
| { |
| "epoch": 3.28, |
| "grad_norm": 6.390233407929048, |
| "learning_rate": 1e-05, |
| "loss": 0.1521, |
| "step": 10280 |
| }, |
| { |
| "epoch": 3.28, |
| "grad_norm": 6.677748085135092, |
| "learning_rate": 1e-05, |
| "loss": 0.1569, |
| "step": 10290 |
| }, |
| { |
| "epoch": 3.29, |
| "grad_norm": 7.230734384892937, |
| "learning_rate": 1e-05, |
| "loss": 0.1738, |
| "step": 10300 |
| }, |
| { |
| "epoch": 3.29, |
| "grad_norm": 6.7386950733434965, |
| "learning_rate": 1e-05, |
| "loss": 0.156, |
| "step": 10310 |
| }, |
| { |
| "epoch": 3.29, |
| "grad_norm": 7.543509607664574, |
| "learning_rate": 1e-05, |
| "loss": 0.1688, |
| "step": 10320 |
| }, |
| { |
| "epoch": 3.3, |
| "grad_norm": 6.157559483094077, |
| "learning_rate": 1e-05, |
| "loss": 0.1689, |
| "step": 10330 |
| }, |
| { |
| "epoch": 3.3, |
| "grad_norm": 6.799005143935812, |
| "learning_rate": 1e-05, |
| "loss": 0.1811, |
| "step": 10340 |
| }, |
| { |
| "epoch": 3.3, |
| "grad_norm": 6.291535870874611, |
| "learning_rate": 1e-05, |
| "loss": 0.1883, |
| "step": 10350 |
| }, |
| { |
| "epoch": 3.31, |
| "grad_norm": 6.558362546643247, |
| "learning_rate": 1e-05, |
| "loss": 0.1566, |
| "step": 10360 |
| }, |
| { |
| "epoch": 3.31, |
| "grad_norm": 5.925483831939644, |
| "learning_rate": 1e-05, |
| "loss": 0.1751, |
| "step": 10370 |
| }, |
| { |
| "epoch": 3.31, |
| "grad_norm": 5.850659859766053, |
| "learning_rate": 1e-05, |
| "loss": 0.1613, |
| "step": 10380 |
| }, |
| { |
| "epoch": 3.32, |
| "grad_norm": 8.207228791203054, |
| "learning_rate": 1e-05, |
| "loss": 0.1726, |
| "step": 10390 |
| }, |
| { |
| "epoch": 3.32, |
| "grad_norm": 4.696226618132108, |
| "learning_rate": 1e-05, |
| "loss": 0.1527, |
| "step": 10400 |
| }, |
| { |
| "epoch": 3.32, |
| "grad_norm": 5.164094448712158, |
| "learning_rate": 1e-05, |
| "loss": 0.1724, |
| "step": 10410 |
| }, |
| { |
| "epoch": 3.32, |
| "grad_norm": 5.4946403398210935, |
| "learning_rate": 1e-05, |
| "loss": 0.1574, |
| "step": 10420 |
| }, |
| { |
| "epoch": 3.33, |
| "grad_norm": 6.401860217116737, |
| "learning_rate": 1e-05, |
| "loss": 0.1658, |
| "step": 10430 |
| }, |
| { |
| "epoch": 3.33, |
| "grad_norm": 7.350646850771424, |
| "learning_rate": 1e-05, |
| "loss": 0.1789, |
| "step": 10440 |
| }, |
| { |
| "epoch": 3.33, |
| "grad_norm": 8.244645201554262, |
| "learning_rate": 1e-05, |
| "loss": 0.1789, |
| "step": 10450 |
| }, |
| { |
| "epoch": 3.34, |
| "grad_norm": 6.1399077561994915, |
| "learning_rate": 1e-05, |
| "loss": 0.1679, |
| "step": 10460 |
| }, |
| { |
| "epoch": 3.34, |
| "grad_norm": 8.42712804706218, |
| "learning_rate": 1e-05, |
| "loss": 0.1754, |
| "step": 10470 |
| }, |
| { |
| "epoch": 3.34, |
| "grad_norm": 6.926249135968606, |
| "learning_rate": 1e-05, |
| "loss": 0.1738, |
| "step": 10480 |
| }, |
| { |
| "epoch": 3.35, |
| "grad_norm": 8.439420806392064, |
| "learning_rate": 1e-05, |
| "loss": 0.1776, |
| "step": 10490 |
| }, |
| { |
| "epoch": 3.35, |
| "grad_norm": 7.862112004784799, |
| "learning_rate": 1e-05, |
| "loss": 0.1882, |
| "step": 10500 |
| }, |
| { |
| "epoch": 3.35, |
| "eval_loss": 0.8896484375, |
| "eval_runtime": 54.827, |
| "eval_samples_per_second": 46.218, |
| "eval_steps_per_second": 2.9, |
| "step": 10500 |
| }, |
| { |
| "epoch": 3.35, |
| "grad_norm": 5.077865640912832, |
| "learning_rate": 1e-05, |
| "loss": 0.1653, |
| "step": 10510 |
| }, |
| { |
| "epoch": 3.36, |
| "grad_norm": 5.203584667899245, |
| "learning_rate": 1e-05, |
| "loss": 0.164, |
| "step": 10520 |
| }, |
| { |
| "epoch": 3.36, |
| "grad_norm": 5.507376665768146, |
| "learning_rate": 1e-05, |
| "loss": 0.1636, |
| "step": 10530 |
| }, |
| { |
| "epoch": 3.36, |
| "grad_norm": 6.706669223903727, |
| "learning_rate": 1e-05, |
| "loss": 0.1524, |
| "step": 10540 |
| }, |
| { |
| "epoch": 3.37, |
| "grad_norm": 7.88267506384155, |
| "learning_rate": 1e-05, |
| "loss": 0.1594, |
| "step": 10550 |
| }, |
| { |
| "epoch": 3.37, |
| "grad_norm": 6.62324961445077, |
| "learning_rate": 1e-05, |
| "loss": 0.2056, |
| "step": 10560 |
| }, |
| { |
| "epoch": 3.37, |
| "grad_norm": 9.253895944353696, |
| "learning_rate": 1e-05, |
| "loss": 0.1741, |
| "step": 10570 |
| }, |
| { |
| "epoch": 3.38, |
| "grad_norm": 8.787936125740654, |
| "learning_rate": 1e-05, |
| "loss": 0.1838, |
| "step": 10580 |
| }, |
| { |
| "epoch": 3.38, |
| "grad_norm": 5.69882717447641, |
| "learning_rate": 1e-05, |
| "loss": 0.1568, |
| "step": 10590 |
| }, |
| { |
| "epoch": 3.38, |
| "grad_norm": 6.443980094286947, |
| "learning_rate": 1e-05, |
| "loss": 0.1779, |
| "step": 10600 |
| }, |
| { |
| "epoch": 3.39, |
| "grad_norm": 7.869494011983151, |
| "learning_rate": 1e-05, |
| "loss": 0.1783, |
| "step": 10610 |
| }, |
| { |
| "epoch": 3.39, |
| "grad_norm": 6.769339935716274, |
| "learning_rate": 1e-05, |
| "loss": 0.1858, |
| "step": 10620 |
| }, |
| { |
| "epoch": 3.39, |
| "grad_norm": 6.2666358798197415, |
| "learning_rate": 1e-05, |
| "loss": 0.1606, |
| "step": 10630 |
| }, |
| { |
| "epoch": 3.4, |
| "grad_norm": 7.824568315487473, |
| "learning_rate": 1e-05, |
| "loss": 0.2003, |
| "step": 10640 |
| }, |
| { |
| "epoch": 3.4, |
| "grad_norm": 4.091542354852597, |
| "learning_rate": 1e-05, |
| "loss": 0.1558, |
| "step": 10650 |
| }, |
| { |
| "epoch": 3.4, |
| "grad_norm": 6.48837767086694, |
| "learning_rate": 1e-05, |
| "loss": 0.1764, |
| "step": 10660 |
| }, |
| { |
| "epoch": 3.4, |
| "grad_norm": 5.800327804772957, |
| "learning_rate": 1e-05, |
| "loss": 0.1925, |
| "step": 10670 |
| }, |
| { |
| "epoch": 3.41, |
| "grad_norm": 6.451581054424306, |
| "learning_rate": 1e-05, |
| "loss": 0.188, |
| "step": 10680 |
| }, |
| { |
| "epoch": 3.41, |
| "grad_norm": 7.121115177559643, |
| "learning_rate": 1e-05, |
| "loss": 0.1653, |
| "step": 10690 |
| }, |
| { |
| "epoch": 3.41, |
| "grad_norm": 5.847576010847291, |
| "learning_rate": 1e-05, |
| "loss": 0.1492, |
| "step": 10700 |
| }, |
| { |
| "epoch": 3.42, |
| "grad_norm": 5.897086972872708, |
| "learning_rate": 1e-05, |
| "loss": 0.172, |
| "step": 10710 |
| }, |
| { |
| "epoch": 3.42, |
| "grad_norm": 5.814529556833051, |
| "learning_rate": 1e-05, |
| "loss": 0.2091, |
| "step": 10720 |
| }, |
| { |
| "epoch": 3.42, |
| "grad_norm": 5.781827330759268, |
| "learning_rate": 1e-05, |
| "loss": 0.164, |
| "step": 10730 |
| }, |
| { |
| "epoch": 3.43, |
| "grad_norm": 6.801940854163973, |
| "learning_rate": 1e-05, |
| "loss": 0.1681, |
| "step": 10740 |
| }, |
| { |
| "epoch": 3.43, |
| "grad_norm": 8.05068411437641, |
| "learning_rate": 1e-05, |
| "loss": 0.212, |
| "step": 10750 |
| }, |
| { |
| "epoch": 3.43, |
| "grad_norm": 6.97683151427033, |
| "learning_rate": 1e-05, |
| "loss": 0.1932, |
| "step": 10760 |
| }, |
| { |
| "epoch": 3.44, |
| "grad_norm": 7.261958027509561, |
| "learning_rate": 1e-05, |
| "loss": 0.188, |
| "step": 10770 |
| }, |
| { |
| "epoch": 3.44, |
| "grad_norm": 7.712018371085291, |
| "learning_rate": 1e-05, |
| "loss": 0.1831, |
| "step": 10780 |
| }, |
| { |
| "epoch": 3.44, |
| "grad_norm": 6.92484290430242, |
| "learning_rate": 1e-05, |
| "loss": 0.1949, |
| "step": 10790 |
| }, |
| { |
| "epoch": 3.45, |
| "grad_norm": 8.478248590228695, |
| "learning_rate": 1e-05, |
| "loss": 0.1956, |
| "step": 10800 |
| }, |
| { |
| "epoch": 3.45, |
| "grad_norm": 7.982848139004048, |
| "learning_rate": 1e-05, |
| "loss": 0.177, |
| "step": 10810 |
| }, |
| { |
| "epoch": 3.45, |
| "grad_norm": 7.423326081419567, |
| "learning_rate": 1e-05, |
| "loss": 0.1968, |
| "step": 10820 |
| }, |
| { |
| "epoch": 3.46, |
| "grad_norm": 7.2307562788926925, |
| "learning_rate": 1e-05, |
| "loss": 0.1898, |
| "step": 10830 |
| }, |
| { |
| "epoch": 3.46, |
| "grad_norm": 7.931763746428054, |
| "learning_rate": 1e-05, |
| "loss": 0.18, |
| "step": 10840 |
| }, |
| { |
| "epoch": 3.46, |
| "grad_norm": 7.65132948349437, |
| "learning_rate": 1e-05, |
| "loss": 0.1869, |
| "step": 10850 |
| }, |
| { |
| "epoch": 3.47, |
| "grad_norm": 5.637183156382151, |
| "learning_rate": 1e-05, |
| "loss": 0.1579, |
| "step": 10860 |
| }, |
| { |
| "epoch": 3.47, |
| "grad_norm": 5.637582735186184, |
| "learning_rate": 1e-05, |
| "loss": 0.2086, |
| "step": 10870 |
| }, |
| { |
| "epoch": 3.47, |
| "grad_norm": 5.279428021054192, |
| "learning_rate": 1e-05, |
| "loss": 0.1764, |
| "step": 10880 |
| }, |
| { |
| "epoch": 3.47, |
| "grad_norm": 7.975876437730914, |
| "learning_rate": 1e-05, |
| "loss": 0.1637, |
| "step": 10890 |
| }, |
| { |
| "epoch": 3.48, |
| "grad_norm": 7.044970878936164, |
| "learning_rate": 1e-05, |
| "loss": 0.1833, |
| "step": 10900 |
| }, |
| { |
| "epoch": 3.48, |
| "grad_norm": 6.541038078025811, |
| "learning_rate": 1e-05, |
| "loss": 0.1564, |
| "step": 10910 |
| }, |
| { |
| "epoch": 3.48, |
| "grad_norm": 8.781341660859466, |
| "learning_rate": 1e-05, |
| "loss": 0.2181, |
| "step": 10920 |
| }, |
| { |
| "epoch": 3.49, |
| "grad_norm": 7.483402706694104, |
| "learning_rate": 1e-05, |
| "loss": 0.1936, |
| "step": 10930 |
| }, |
| { |
| "epoch": 3.49, |
| "grad_norm": 6.083480240950982, |
| "learning_rate": 1e-05, |
| "loss": 0.1784, |
| "step": 10940 |
| }, |
| { |
| "epoch": 3.49, |
| "grad_norm": 5.079035375127533, |
| "learning_rate": 1e-05, |
| "loss": 0.1593, |
| "step": 10950 |
| }, |
| { |
| "epoch": 3.5, |
| "grad_norm": 7.50637622635008, |
| "learning_rate": 1e-05, |
| "loss": 0.1869, |
| "step": 10960 |
| }, |
| { |
| "epoch": 3.5, |
| "grad_norm": 9.631578451468735, |
| "learning_rate": 1e-05, |
| "loss": 0.1796, |
| "step": 10970 |
| }, |
| { |
| "epoch": 3.5, |
| "grad_norm": 9.01708803255874, |
| "learning_rate": 1e-05, |
| "loss": 0.1934, |
| "step": 10980 |
| }, |
| { |
| "epoch": 3.51, |
| "grad_norm": 6.2994461300506535, |
| "learning_rate": 1e-05, |
| "loss": 0.187, |
| "step": 10990 |
| }, |
| { |
| "epoch": 3.51, |
| "grad_norm": 9.865903890182212, |
| "learning_rate": 1e-05, |
| "loss": 0.1982, |
| "step": 11000 |
| }, |
| { |
| "epoch": 3.51, |
| "eval_loss": 0.89794921875, |
| "eval_runtime": 55.0515, |
| "eval_samples_per_second": 46.03, |
| "eval_steps_per_second": 2.888, |
| "step": 11000 |
| }, |
| { |
| "epoch": 3.51, |
| "grad_norm": 9.354520681998698, |
| "learning_rate": 1e-05, |
| "loss": 0.1967, |
| "step": 11010 |
| }, |
| { |
| "epoch": 3.52, |
| "grad_norm": 6.451774696044153, |
| "learning_rate": 1e-05, |
| "loss": 0.1572, |
| "step": 11020 |
| }, |
| { |
| "epoch": 3.52, |
| "grad_norm": 6.054697895656398, |
| "learning_rate": 1e-05, |
| "loss": 0.1974, |
| "step": 11030 |
| }, |
| { |
| "epoch": 3.52, |
| "grad_norm": 8.858506040662892, |
| "learning_rate": 1e-05, |
| "loss": 0.1699, |
| "step": 11040 |
| }, |
| { |
| "epoch": 3.53, |
| "grad_norm": 6.2187981531780485, |
| "learning_rate": 1e-05, |
| "loss": 0.1716, |
| "step": 11050 |
| }, |
| { |
| "epoch": 3.53, |
| "grad_norm": 5.982959390641755, |
| "learning_rate": 1e-05, |
| "loss": 0.1783, |
| "step": 11060 |
| }, |
| { |
| "epoch": 3.53, |
| "grad_norm": 6.171453147667688, |
| "learning_rate": 1e-05, |
| "loss": 0.1907, |
| "step": 11070 |
| }, |
| { |
| "epoch": 3.54, |
| "grad_norm": 6.545835426114932, |
| "learning_rate": 1e-05, |
| "loss": 0.1938, |
| "step": 11080 |
| }, |
| { |
| "epoch": 3.54, |
| "grad_norm": 6.083653933642689, |
| "learning_rate": 1e-05, |
| "loss": 0.1795, |
| "step": 11090 |
| }, |
| { |
| "epoch": 3.54, |
| "grad_norm": 4.734948655427137, |
| "learning_rate": 1e-05, |
| "loss": 0.1798, |
| "step": 11100 |
| }, |
| { |
| "epoch": 3.54, |
| "grad_norm": 5.879635220066371, |
| "learning_rate": 1e-05, |
| "loss": 0.1573, |
| "step": 11110 |
| }, |
| { |
| "epoch": 3.55, |
| "grad_norm": 7.059286010775707, |
| "learning_rate": 1e-05, |
| "loss": 0.1757, |
| "step": 11120 |
| }, |
| { |
| "epoch": 3.55, |
| "grad_norm": 7.26219309488279, |
| "learning_rate": 1e-05, |
| "loss": 0.2039, |
| "step": 11130 |
| }, |
| { |
| "epoch": 3.55, |
| "grad_norm": 7.881575488297652, |
| "learning_rate": 1e-05, |
| "loss": 0.1583, |
| "step": 11140 |
| }, |
| { |
| "epoch": 3.56, |
| "grad_norm": 6.406704025927066, |
| "learning_rate": 1e-05, |
| "loss": 0.1621, |
| "step": 11150 |
| }, |
| { |
| "epoch": 3.56, |
| "grad_norm": 5.975263466871997, |
| "learning_rate": 1e-05, |
| "loss": 0.1888, |
| "step": 11160 |
| }, |
| { |
| "epoch": 3.56, |
| "grad_norm": 7.751371939034783, |
| "learning_rate": 1e-05, |
| "loss": 0.1952, |
| "step": 11170 |
| }, |
| { |
| "epoch": 3.57, |
| "grad_norm": 5.730444119536365, |
| "learning_rate": 1e-05, |
| "loss": 0.1961, |
| "step": 11180 |
| }, |
| { |
| "epoch": 3.57, |
| "grad_norm": 7.6488599339100345, |
| "learning_rate": 1e-05, |
| "loss": 0.1762, |
| "step": 11190 |
| }, |
| { |
| "epoch": 3.57, |
| "grad_norm": 7.155878182292824, |
| "learning_rate": 1e-05, |
| "loss": 0.1843, |
| "step": 11200 |
| }, |
| { |
| "epoch": 3.58, |
| "grad_norm": 6.2852982717795, |
| "learning_rate": 1e-05, |
| "loss": 0.2111, |
| "step": 11210 |
| }, |
| { |
| "epoch": 3.58, |
| "grad_norm": 8.995201102993795, |
| "learning_rate": 1e-05, |
| "loss": 0.1768, |
| "step": 11220 |
| }, |
| { |
| "epoch": 3.58, |
| "grad_norm": 5.710522679659222, |
| "learning_rate": 1e-05, |
| "loss": 0.1652, |
| "step": 11230 |
| }, |
| { |
| "epoch": 3.59, |
| "grad_norm": 5.673904770351506, |
| "learning_rate": 1e-05, |
| "loss": 0.2101, |
| "step": 11240 |
| }, |
| { |
| "epoch": 3.59, |
| "grad_norm": 6.850758486890567, |
| "learning_rate": 1e-05, |
| "loss": 0.2022, |
| "step": 11250 |
| }, |
| { |
| "epoch": 3.59, |
| "grad_norm": 6.711906034591627, |
| "learning_rate": 1e-05, |
| "loss": 0.173, |
| "step": 11260 |
| }, |
| { |
| "epoch": 3.6, |
| "grad_norm": 7.318092750597498, |
| "learning_rate": 1e-05, |
| "loss": 0.2023, |
| "step": 11270 |
| }, |
| { |
| "epoch": 3.6, |
| "grad_norm": 6.7624422476342, |
| "learning_rate": 1e-05, |
| "loss": 0.1819, |
| "step": 11280 |
| }, |
| { |
| "epoch": 3.6, |
| "grad_norm": 8.19313053725426, |
| "learning_rate": 1e-05, |
| "loss": 0.1826, |
| "step": 11290 |
| }, |
| { |
| "epoch": 3.61, |
| "grad_norm": 6.622160969023495, |
| "learning_rate": 1e-05, |
| "loss": 0.1953, |
| "step": 11300 |
| }, |
| { |
| "epoch": 3.61, |
| "grad_norm": 6.449237677894695, |
| "learning_rate": 1e-05, |
| "loss": 0.2093, |
| "step": 11310 |
| }, |
| { |
| "epoch": 3.61, |
| "grad_norm": 8.708538241614608, |
| "learning_rate": 1e-05, |
| "loss": 0.1886, |
| "step": 11320 |
| }, |
| { |
| "epoch": 3.62, |
| "grad_norm": 5.967406116605275, |
| "learning_rate": 1e-05, |
| "loss": 0.1834, |
| "step": 11330 |
| }, |
| { |
| "epoch": 3.62, |
| "grad_norm": 6.160982278559057, |
| "learning_rate": 1e-05, |
| "loss": 0.194, |
| "step": 11340 |
| }, |
| { |
| "epoch": 3.62, |
| "grad_norm": 5.2724094773431, |
| "learning_rate": 1e-05, |
| "loss": 0.1998, |
| "step": 11350 |
| }, |
| { |
| "epoch": 3.62, |
| "grad_norm": 5.610212258252053, |
| "learning_rate": 1e-05, |
| "loss": 0.1822, |
| "step": 11360 |
| }, |
| { |
| "epoch": 3.63, |
| "grad_norm": 6.732770334759328, |
| "learning_rate": 1e-05, |
| "loss": 0.2042, |
| "step": 11370 |
| }, |
| { |
| "epoch": 3.63, |
| "grad_norm": 7.9177208717212455, |
| "learning_rate": 1e-05, |
| "loss": 0.1928, |
| "step": 11380 |
| }, |
| { |
| "epoch": 3.63, |
| "grad_norm": 7.894730087411233, |
| "learning_rate": 1e-05, |
| "loss": 0.1799, |
| "step": 11390 |
| }, |
| { |
| "epoch": 3.64, |
| "grad_norm": 5.123559516740489, |
| "learning_rate": 1e-05, |
| "loss": 0.2035, |
| "step": 11400 |
| }, |
| { |
| "epoch": 3.64, |
| "grad_norm": 5.890775562254389, |
| "learning_rate": 1e-05, |
| "loss": 0.1908, |
| "step": 11410 |
| }, |
| { |
| "epoch": 3.64, |
| "grad_norm": 7.7491036173585, |
| "learning_rate": 1e-05, |
| "loss": 0.2042, |
| "step": 11420 |
| }, |
| { |
| "epoch": 3.65, |
| "grad_norm": 10.026381597659972, |
| "learning_rate": 1e-05, |
| "loss": 0.2032, |
| "step": 11430 |
| }, |
| { |
| "epoch": 3.65, |
| "grad_norm": 5.926082193557133, |
| "learning_rate": 1e-05, |
| "loss": 0.1677, |
| "step": 11440 |
| }, |
| { |
| "epoch": 3.65, |
| "grad_norm": 6.375550788572389, |
| "learning_rate": 1e-05, |
| "loss": 0.1901, |
| "step": 11450 |
| }, |
| { |
| "epoch": 3.66, |
| "grad_norm": 4.97300240749452, |
| "learning_rate": 1e-05, |
| "loss": 0.1703, |
| "step": 11460 |
| }, |
| { |
| "epoch": 3.66, |
| "grad_norm": 7.863902674308482, |
| "learning_rate": 1e-05, |
| "loss": 0.1665, |
| "step": 11470 |
| }, |
| { |
| "epoch": 3.66, |
| "grad_norm": 8.071873620033516, |
| "learning_rate": 1e-05, |
| "loss": 0.2083, |
| "step": 11480 |
| }, |
| { |
| "epoch": 3.67, |
| "grad_norm": 6.236884126091724, |
| "learning_rate": 1e-05, |
| "loss": 0.1918, |
| "step": 11490 |
| }, |
| { |
| "epoch": 3.67, |
| "grad_norm": 7.187589694583003, |
| "learning_rate": 1e-05, |
| "loss": 0.1838, |
| "step": 11500 |
| }, |
| { |
| "epoch": 3.67, |
| "eval_loss": 0.900390625, |
| "eval_runtime": 55.1406, |
| "eval_samples_per_second": 45.955, |
| "eval_steps_per_second": 2.884, |
| "step": 11500 |
| }, |
| { |
| "epoch": 3.67, |
| "grad_norm": 9.040291771020213, |
| "learning_rate": 1e-05, |
| "loss": 0.1839, |
| "step": 11510 |
| }, |
| { |
| "epoch": 3.68, |
| "grad_norm": 4.932405958560088, |
| "learning_rate": 1e-05, |
| "loss": 0.2175, |
| "step": 11520 |
| }, |
| { |
| "epoch": 3.68, |
| "grad_norm": 7.525123985511099, |
| "learning_rate": 1e-05, |
| "loss": 0.181, |
| "step": 11530 |
| }, |
| { |
| "epoch": 3.68, |
| "grad_norm": 9.810218886053724, |
| "learning_rate": 1e-05, |
| "loss": 0.2116, |
| "step": 11540 |
| }, |
| { |
| "epoch": 3.69, |
| "grad_norm": 7.655027455405864, |
| "learning_rate": 1e-05, |
| "loss": 0.2048, |
| "step": 11550 |
| }, |
| { |
| "epoch": 3.69, |
| "grad_norm": 5.3819908987795495, |
| "learning_rate": 1e-05, |
| "loss": 0.1841, |
| "step": 11560 |
| }, |
| { |
| "epoch": 3.69, |
| "grad_norm": 7.350074674265913, |
| "learning_rate": 1e-05, |
| "loss": 0.2036, |
| "step": 11570 |
| }, |
| { |
| "epoch": 3.69, |
| "grad_norm": 5.742598372311425, |
| "learning_rate": 1e-05, |
| "loss": 0.203, |
| "step": 11580 |
| }, |
| { |
| "epoch": 3.7, |
| "grad_norm": 5.994758541790089, |
| "learning_rate": 1e-05, |
| "loss": 0.192, |
| "step": 11590 |
| }, |
| { |
| "epoch": 3.7, |
| "grad_norm": 8.881899019240665, |
| "learning_rate": 1e-05, |
| "loss": 0.1927, |
| "step": 11600 |
| }, |
| { |
| "epoch": 3.7, |
| "grad_norm": 7.6631660209312615, |
| "learning_rate": 1e-05, |
| "loss": 0.2052, |
| "step": 11610 |
| }, |
| { |
| "epoch": 3.71, |
| "grad_norm": 7.808984804880935, |
| "learning_rate": 1e-05, |
| "loss": 0.2032, |
| "step": 11620 |
| }, |
| { |
| "epoch": 3.71, |
| "grad_norm": 5.904660288497434, |
| "learning_rate": 1e-05, |
| "loss": 0.1685, |
| "step": 11630 |
| }, |
| { |
| "epoch": 3.71, |
| "grad_norm": 10.595268388258896, |
| "learning_rate": 1e-05, |
| "loss": 0.1998, |
| "step": 11640 |
| }, |
| { |
| "epoch": 3.72, |
| "grad_norm": 5.419760167914256, |
| "learning_rate": 1e-05, |
| "loss": 0.2038, |
| "step": 11650 |
| }, |
| { |
| "epoch": 3.72, |
| "grad_norm": 7.755148377367158, |
| "learning_rate": 1e-05, |
| "loss": 0.1789, |
| "step": 11660 |
| }, |
| { |
| "epoch": 3.72, |
| "grad_norm": 8.18166370438676, |
| "learning_rate": 1e-05, |
| "loss": 0.2109, |
| "step": 11670 |
| }, |
| { |
| "epoch": 3.73, |
| "grad_norm": 5.370140540416807, |
| "learning_rate": 1e-05, |
| "loss": 0.2087, |
| "step": 11680 |
| }, |
| { |
| "epoch": 3.73, |
| "grad_norm": 7.844198145735518, |
| "learning_rate": 1e-05, |
| "loss": 0.1831, |
| "step": 11690 |
| }, |
| { |
| "epoch": 3.73, |
| "grad_norm": 6.527419139488332, |
| "learning_rate": 1e-05, |
| "loss": 0.1797, |
| "step": 11700 |
| }, |
| { |
| "epoch": 3.74, |
| "grad_norm": 4.591245214894432, |
| "learning_rate": 1e-05, |
| "loss": 0.1768, |
| "step": 11710 |
| }, |
| { |
| "epoch": 3.74, |
| "grad_norm": 6.564088465678529, |
| "learning_rate": 1e-05, |
| "loss": 0.1756, |
| "step": 11720 |
| }, |
| { |
| "epoch": 3.74, |
| "grad_norm": 8.083220595789484, |
| "learning_rate": 1e-05, |
| "loss": 0.1866, |
| "step": 11730 |
| }, |
| { |
| "epoch": 3.75, |
| "grad_norm": 8.214150421734368, |
| "learning_rate": 1e-05, |
| "loss": 0.2091, |
| "step": 11740 |
| }, |
| { |
| "epoch": 3.75, |
| "grad_norm": 6.873139285663918, |
| "learning_rate": 1e-05, |
| "loss": 0.1693, |
| "step": 11750 |
| }, |
| { |
| "epoch": 3.75, |
| "grad_norm": 7.7240372437458795, |
| "learning_rate": 1e-05, |
| "loss": 0.1718, |
| "step": 11760 |
| }, |
| { |
| "epoch": 3.76, |
| "grad_norm": 6.1530195561369, |
| "learning_rate": 1e-05, |
| "loss": 0.1789, |
| "step": 11770 |
| }, |
| { |
| "epoch": 3.76, |
| "grad_norm": 8.067568583024062, |
| "learning_rate": 1e-05, |
| "loss": 0.2035, |
| "step": 11780 |
| }, |
| { |
| "epoch": 3.76, |
| "grad_norm": 7.463593430660151, |
| "learning_rate": 1e-05, |
| "loss": 0.199, |
| "step": 11790 |
| }, |
| { |
| "epoch": 3.77, |
| "grad_norm": 6.7755699452884475, |
| "learning_rate": 1e-05, |
| "loss": 0.2058, |
| "step": 11800 |
| }, |
| { |
| "epoch": 3.77, |
| "grad_norm": 7.072183637504606, |
| "learning_rate": 1e-05, |
| "loss": 0.1982, |
| "step": 11810 |
| }, |
| { |
| "epoch": 3.77, |
| "grad_norm": 2.6171252798399234, |
| "learning_rate": 1e-05, |
| "loss": 0.1963, |
| "step": 11820 |
| }, |
| { |
| "epoch": 3.77, |
| "grad_norm": 5.399221448934899, |
| "learning_rate": 1e-05, |
| "loss": 0.18, |
| "step": 11830 |
| }, |
| { |
| "epoch": 3.78, |
| "grad_norm": 7.546585109031855, |
| "learning_rate": 1e-05, |
| "loss": 0.1836, |
| "step": 11840 |
| }, |
| { |
| "epoch": 3.78, |
| "grad_norm": 8.128754672571652, |
| "learning_rate": 1e-05, |
| "loss": 0.1982, |
| "step": 11850 |
| }, |
| { |
| "epoch": 3.78, |
| "grad_norm": 7.335605211623699, |
| "learning_rate": 1e-05, |
| "loss": 0.2331, |
| "step": 11860 |
| }, |
| { |
| "epoch": 3.79, |
| "grad_norm": 8.862206492655034, |
| "learning_rate": 1e-05, |
| "loss": 0.185, |
| "step": 11870 |
| }, |
| { |
| "epoch": 3.79, |
| "grad_norm": 8.069128815662197, |
| "learning_rate": 1e-05, |
| "loss": 0.2141, |
| "step": 11880 |
| }, |
| { |
| "epoch": 3.79, |
| "grad_norm": 5.3759241418595485, |
| "learning_rate": 1e-05, |
| "loss": 0.1749, |
| "step": 11890 |
| }, |
| { |
| "epoch": 3.8, |
| "grad_norm": 7.765814857541437, |
| "learning_rate": 1e-05, |
| "loss": 0.2052, |
| "step": 11900 |
| }, |
| { |
| "epoch": 3.8, |
| "grad_norm": 5.928555467226271, |
| "learning_rate": 1e-05, |
| "loss": 0.1987, |
| "step": 11910 |
| }, |
| { |
| "epoch": 3.8, |
| "grad_norm": 6.260937694971298, |
| "learning_rate": 1e-05, |
| "loss": 0.2001, |
| "step": 11920 |
| }, |
| { |
| "epoch": 3.81, |
| "grad_norm": 6.3122585741464015, |
| "learning_rate": 1e-05, |
| "loss": 0.1962, |
| "step": 11930 |
| }, |
| { |
| "epoch": 3.81, |
| "grad_norm": 6.547411038616237, |
| "learning_rate": 1e-05, |
| "loss": 0.1863, |
| "step": 11940 |
| }, |
| { |
| "epoch": 3.81, |
| "grad_norm": 6.696663256211249, |
| "learning_rate": 1e-05, |
| "loss": 0.1796, |
| "step": 11950 |
| }, |
| { |
| "epoch": 3.82, |
| "grad_norm": 5.265393169970785, |
| "learning_rate": 1e-05, |
| "loss": 0.1683, |
| "step": 11960 |
| }, |
| { |
| "epoch": 3.82, |
| "grad_norm": 8.240414686416065, |
| "learning_rate": 1e-05, |
| "loss": 0.2024, |
| "step": 11970 |
| }, |
| { |
| "epoch": 3.82, |
| "grad_norm": 6.494009044719633, |
| "learning_rate": 1e-05, |
| "loss": 0.2114, |
| "step": 11980 |
| }, |
| { |
| "epoch": 3.83, |
| "grad_norm": 7.326396551223613, |
| "learning_rate": 1e-05, |
| "loss": 0.224, |
| "step": 11990 |
| }, |
| { |
| "epoch": 3.83, |
| "grad_norm": 7.287416916335993, |
| "learning_rate": 1e-05, |
| "loss": 0.1933, |
| "step": 12000 |
| }, |
| { |
| "epoch": 3.83, |
| "eval_loss": 0.8740234375, |
| "eval_runtime": 54.9459, |
| "eval_samples_per_second": 46.118, |
| "eval_steps_per_second": 2.894, |
| "step": 12000 |
| }, |
| { |
| "epoch": 3.83, |
| "grad_norm": 7.712310205037818, |
| "learning_rate": 1e-05, |
| "loss": 0.205, |
| "step": 12010 |
| }, |
| { |
| "epoch": 3.84, |
| "grad_norm": 5.058254204036346, |
| "learning_rate": 1e-05, |
| "loss": 0.1951, |
| "step": 12020 |
| }, |
| { |
| "epoch": 3.84, |
| "grad_norm": 6.096824144457301, |
| "learning_rate": 1e-05, |
| "loss": 0.196, |
| "step": 12030 |
| }, |
| { |
| "epoch": 3.84, |
| "grad_norm": 7.165039528353706, |
| "learning_rate": 1e-05, |
| "loss": 0.1735, |
| "step": 12040 |
| }, |
| { |
| "epoch": 3.84, |
| "grad_norm": 5.632963689079253, |
| "learning_rate": 1e-05, |
| "loss": 0.1845, |
| "step": 12050 |
| }, |
| { |
| "epoch": 3.85, |
| "grad_norm": 6.979472169784677, |
| "learning_rate": 1e-05, |
| "loss": 0.1738, |
| "step": 12060 |
| }, |
| { |
| "epoch": 3.85, |
| "grad_norm": 7.519353437426875, |
| "learning_rate": 1e-05, |
| "loss": 0.1936, |
| "step": 12070 |
| }, |
| { |
| "epoch": 3.85, |
| "grad_norm": 7.098673855144282, |
| "learning_rate": 1e-05, |
| "loss": 0.1786, |
| "step": 12080 |
| }, |
| { |
| "epoch": 3.86, |
| "grad_norm": 8.260766142627668, |
| "learning_rate": 1e-05, |
| "loss": 0.2214, |
| "step": 12090 |
| }, |
| { |
| "epoch": 3.86, |
| "grad_norm": 7.52119781653554, |
| "learning_rate": 1e-05, |
| "loss": 0.1968, |
| "step": 12100 |
| }, |
| { |
| "epoch": 3.86, |
| "grad_norm": 6.854507978882829, |
| "learning_rate": 1e-05, |
| "loss": 0.1934, |
| "step": 12110 |
| }, |
| { |
| "epoch": 3.87, |
| "grad_norm": 6.897165252427051, |
| "learning_rate": 1e-05, |
| "loss": 0.1902, |
| "step": 12120 |
| }, |
| { |
| "epoch": 3.87, |
| "grad_norm": 6.437532295220114, |
| "learning_rate": 1e-05, |
| "loss": 0.1929, |
| "step": 12130 |
| }, |
| { |
| "epoch": 3.87, |
| "grad_norm": 6.002100894597306, |
| "learning_rate": 1e-05, |
| "loss": 0.1946, |
| "step": 12140 |
| }, |
| { |
| "epoch": 3.88, |
| "grad_norm": 6.22542777066238, |
| "learning_rate": 1e-05, |
| "loss": 0.1825, |
| "step": 12150 |
| }, |
| { |
| "epoch": 3.88, |
| "grad_norm": 5.5509004841829706, |
| "learning_rate": 1e-05, |
| "loss": 0.197, |
| "step": 12160 |
| }, |
| { |
| "epoch": 3.88, |
| "grad_norm": 8.425406372049938, |
| "learning_rate": 1e-05, |
| "loss": 0.185, |
| "step": 12170 |
| }, |
| { |
| "epoch": 3.89, |
| "grad_norm": 7.660520024588353, |
| "learning_rate": 1e-05, |
| "loss": 0.1949, |
| "step": 12180 |
| }, |
| { |
| "epoch": 3.89, |
| "grad_norm": 6.461355802378603, |
| "learning_rate": 1e-05, |
| "loss": 0.1776, |
| "step": 12190 |
| }, |
| { |
| "epoch": 3.89, |
| "grad_norm": 7.079542274980932, |
| "learning_rate": 1e-05, |
| "loss": 0.2009, |
| "step": 12200 |
| }, |
| { |
| "epoch": 3.9, |
| "grad_norm": 8.203861109085647, |
| "learning_rate": 1e-05, |
| "loss": 0.1741, |
| "step": 12210 |
| }, |
| { |
| "epoch": 3.9, |
| "grad_norm": 7.573865940322843, |
| "learning_rate": 1e-05, |
| "loss": 0.1963, |
| "step": 12220 |
| }, |
| { |
| "epoch": 3.9, |
| "grad_norm": 6.683087756178809, |
| "learning_rate": 1e-05, |
| "loss": 0.1974, |
| "step": 12230 |
| }, |
| { |
| "epoch": 3.91, |
| "grad_norm": 5.533866823668124, |
| "learning_rate": 1e-05, |
| "loss": 0.19, |
| "step": 12240 |
| }, |
| { |
| "epoch": 3.91, |
| "grad_norm": 9.282690071360646, |
| "learning_rate": 1e-05, |
| "loss": 0.1956, |
| "step": 12250 |
| }, |
| { |
| "epoch": 3.91, |
| "grad_norm": 9.520020520484671, |
| "learning_rate": 1e-05, |
| "loss": 0.1631, |
| "step": 12260 |
| }, |
| { |
| "epoch": 3.92, |
| "grad_norm": 5.44369609819217, |
| "learning_rate": 1e-05, |
| "loss": 0.2114, |
| "step": 12270 |
| }, |
| { |
| "epoch": 3.92, |
| "grad_norm": 6.643214109237476, |
| "learning_rate": 1e-05, |
| "loss": 0.2044, |
| "step": 12280 |
| }, |
| { |
| "epoch": 3.92, |
| "grad_norm": 8.522049970128043, |
| "learning_rate": 1e-05, |
| "loss": 0.1977, |
| "step": 12290 |
| }, |
| { |
| "epoch": 3.92, |
| "grad_norm": 5.923204096706463, |
| "learning_rate": 1e-05, |
| "loss": 0.1773, |
| "step": 12300 |
| }, |
| { |
| "epoch": 3.93, |
| "grad_norm": 6.676873150719637, |
| "learning_rate": 1e-05, |
| "loss": 0.1929, |
| "step": 12310 |
| }, |
| { |
| "epoch": 3.93, |
| "grad_norm": 6.325884839312532, |
| "learning_rate": 1e-05, |
| "loss": 0.1964, |
| "step": 12320 |
| }, |
| { |
| "epoch": 3.93, |
| "grad_norm": 6.2989316851951855, |
| "learning_rate": 1e-05, |
| "loss": 0.199, |
| "step": 12330 |
| }, |
| { |
| "epoch": 3.94, |
| "grad_norm": 7.941990820989633, |
| "learning_rate": 1e-05, |
| "loss": 0.2005, |
| "step": 12340 |
| }, |
| { |
| "epoch": 3.94, |
| "grad_norm": 6.751519773842908, |
| "learning_rate": 1e-05, |
| "loss": 0.2182, |
| "step": 12350 |
| }, |
| { |
| "epoch": 3.94, |
| "grad_norm": 5.335205027851225, |
| "learning_rate": 1e-05, |
| "loss": 0.1904, |
| "step": 12360 |
| }, |
| { |
| "epoch": 3.95, |
| "grad_norm": 5.411452243521246, |
| "learning_rate": 1e-05, |
| "loss": 0.2097, |
| "step": 12370 |
| }, |
| { |
| "epoch": 3.95, |
| "grad_norm": 7.297107112876779, |
| "learning_rate": 1e-05, |
| "loss": 0.1833, |
| "step": 12380 |
| }, |
| { |
| "epoch": 3.95, |
| "grad_norm": 6.297857846780281, |
| "learning_rate": 1e-05, |
| "loss": 0.2137, |
| "step": 12390 |
| }, |
| { |
| "epoch": 3.96, |
| "grad_norm": 7.066418666206616, |
| "learning_rate": 1e-05, |
| "loss": 0.1862, |
| "step": 12400 |
| }, |
| { |
| "epoch": 3.96, |
| "grad_norm": 7.966299600271701, |
| "learning_rate": 1e-05, |
| "loss": 0.2083, |
| "step": 12410 |
| }, |
| { |
| "epoch": 3.96, |
| "grad_norm": 6.899947522834583, |
| "learning_rate": 1e-05, |
| "loss": 0.192, |
| "step": 12420 |
| }, |
| { |
| "epoch": 3.97, |
| "grad_norm": 7.093925574730429, |
| "learning_rate": 1e-05, |
| "loss": 0.1648, |
| "step": 12430 |
| }, |
| { |
| "epoch": 3.97, |
| "grad_norm": 14.427568198455822, |
| "learning_rate": 1e-05, |
| "loss": 0.2155, |
| "step": 12440 |
| }, |
| { |
| "epoch": 3.97, |
| "grad_norm": 11.976047771065534, |
| "learning_rate": 1e-05, |
| "loss": 0.1855, |
| "step": 12450 |
| }, |
| { |
| "epoch": 3.98, |
| "grad_norm": 7.398838890329268, |
| "learning_rate": 1e-05, |
| "loss": 0.2072, |
| "step": 12460 |
| }, |
| { |
| "epoch": 3.98, |
| "grad_norm": 8.668086253569262, |
| "learning_rate": 1e-05, |
| "loss": 0.1963, |
| "step": 12470 |
| }, |
| { |
| "epoch": 3.98, |
| "grad_norm": 11.485418612233609, |
| "learning_rate": 1e-05, |
| "loss": 0.2131, |
| "step": 12480 |
| }, |
| { |
| "epoch": 3.99, |
| "grad_norm": 6.621675628930008, |
| "learning_rate": 1e-05, |
| "loss": 0.1874, |
| "step": 12490 |
| }, |
| { |
| "epoch": 3.99, |
| "grad_norm": 5.025746433369317, |
| "learning_rate": 1e-05, |
| "loss": 0.2203, |
| "step": 12500 |
| }, |
| { |
| "epoch": 3.99, |
| "eval_loss": 0.873046875, |
| "eval_runtime": 54.8674, |
| "eval_samples_per_second": 46.184, |
| "eval_steps_per_second": 2.898, |
| "step": 12500 |
| }, |
| { |
| "epoch": 3.99, |
| "grad_norm": 6.496194752672401, |
| "learning_rate": 1e-05, |
| "loss": 0.1566, |
| "step": 12510 |
| }, |
| { |
| "epoch": 3.99, |
| "grad_norm": 12.342774635119907, |
| "learning_rate": 1e-05, |
| "loss": 0.2148, |
| "step": 12520 |
| }, |
| { |
| "epoch": 4.0, |
| "grad_norm": 5.898876290073873, |
| "learning_rate": 1e-05, |
| "loss": 0.2018, |
| "step": 12530 |
| }, |
| { |
| "epoch": 4.0, |
| "grad_norm": 4.9942469878314855, |
| "learning_rate": 1e-05, |
| "loss": 0.1612, |
| "step": 12540 |
| }, |
| { |
| "epoch": 4.0, |
| "grad_norm": 4.096670259412572, |
| "learning_rate": 1e-05, |
| "loss": 0.108, |
| "step": 12550 |
| }, |
| { |
| "epoch": 4.01, |
| "grad_norm": 7.1357832839970605, |
| "learning_rate": 1e-05, |
| "loss": 0.1112, |
| "step": 12560 |
| }, |
| { |
| "epoch": 4.01, |
| "grad_norm": 7.827885241463088, |
| "learning_rate": 1e-05, |
| "loss": 0.0954, |
| "step": 12570 |
| }, |
| { |
| "epoch": 4.01, |
| "grad_norm": 5.126994605486959, |
| "learning_rate": 1e-05, |
| "loss": 0.1124, |
| "step": 12580 |
| }, |
| { |
| "epoch": 4.02, |
| "grad_norm": 4.328408793155293, |
| "learning_rate": 1e-05, |
| "loss": 0.1198, |
| "step": 12590 |
| }, |
| { |
| "epoch": 4.02, |
| "grad_norm": 5.7996802932494695, |
| "learning_rate": 1e-05, |
| "loss": 0.0997, |
| "step": 12600 |
| }, |
| { |
| "epoch": 4.02, |
| "grad_norm": 3.9374384799571494, |
| "learning_rate": 1e-05, |
| "loss": 0.1031, |
| "step": 12610 |
| }, |
| { |
| "epoch": 4.03, |
| "grad_norm": 3.2741088178816793, |
| "learning_rate": 1e-05, |
| "loss": 0.1104, |
| "step": 12620 |
| }, |
| { |
| "epoch": 4.03, |
| "grad_norm": 6.15698671531762, |
| "learning_rate": 1e-05, |
| "loss": 0.1203, |
| "step": 12630 |
| }, |
| { |
| "epoch": 4.03, |
| "grad_norm": 5.066877756406233, |
| "learning_rate": 1e-05, |
| "loss": 0.0926, |
| "step": 12640 |
| }, |
| { |
| "epoch": 4.04, |
| "grad_norm": 5.433466500520664, |
| "learning_rate": 1e-05, |
| "loss": 0.0841, |
| "step": 12650 |
| }, |
| { |
| "epoch": 4.04, |
| "grad_norm": 7.396482569893495, |
| "learning_rate": 1e-05, |
| "loss": 0.1038, |
| "step": 12660 |
| }, |
| { |
| "epoch": 4.04, |
| "grad_norm": 6.5471095222948446, |
| "learning_rate": 1e-05, |
| "loss": 0.1023, |
| "step": 12670 |
| }, |
| { |
| "epoch": 4.05, |
| "grad_norm": 6.314799277165401, |
| "learning_rate": 1e-05, |
| "loss": 0.1085, |
| "step": 12680 |
| }, |
| { |
| "epoch": 4.05, |
| "grad_norm": 7.218961109856129, |
| "learning_rate": 1e-05, |
| "loss": 0.1158, |
| "step": 12690 |
| }, |
| { |
| "epoch": 4.05, |
| "grad_norm": 6.64911652626377, |
| "learning_rate": 1e-05, |
| "loss": 0.1105, |
| "step": 12700 |
| }, |
| { |
| "epoch": 4.06, |
| "grad_norm": 4.813356001387572, |
| "learning_rate": 1e-05, |
| "loss": 0.097, |
| "step": 12710 |
| }, |
| { |
| "epoch": 4.06, |
| "grad_norm": 6.953702409750445, |
| "learning_rate": 1e-05, |
| "loss": 0.1045, |
| "step": 12720 |
| }, |
| { |
| "epoch": 4.06, |
| "grad_norm": 5.822555867472118, |
| "learning_rate": 1e-05, |
| "loss": 0.1226, |
| "step": 12730 |
| }, |
| { |
| "epoch": 4.07, |
| "grad_norm": 5.052990871859817, |
| "learning_rate": 1e-05, |
| "loss": 0.1023, |
| "step": 12740 |
| }, |
| { |
| "epoch": 4.07, |
| "grad_norm": 4.098450735509271, |
| "learning_rate": 1e-05, |
| "loss": 0.1315, |
| "step": 12750 |
| }, |
| { |
| "epoch": 4.07, |
| "grad_norm": 4.961470159334607, |
| "learning_rate": 1e-05, |
| "loss": 0.1099, |
| "step": 12760 |
| }, |
| { |
| "epoch": 4.07, |
| "grad_norm": 5.571473159891082, |
| "learning_rate": 1e-05, |
| "loss": 0.1239, |
| "step": 12770 |
| }, |
| { |
| "epoch": 4.08, |
| "grad_norm": 4.62424235968694, |
| "learning_rate": 1e-05, |
| "loss": 0.0954, |
| "step": 12780 |
| }, |
| { |
| "epoch": 4.08, |
| "grad_norm": 5.715416775889501, |
| "learning_rate": 1e-05, |
| "loss": 0.1032, |
| "step": 12790 |
| }, |
| { |
| "epoch": 4.08, |
| "grad_norm": 5.671854823052954, |
| "learning_rate": 1e-05, |
| "loss": 0.1181, |
| "step": 12800 |
| }, |
| { |
| "epoch": 4.09, |
| "grad_norm": 5.062830407934079, |
| "learning_rate": 1e-05, |
| "loss": 0.1257, |
| "step": 12810 |
| }, |
| { |
| "epoch": 4.09, |
| "grad_norm": 3.1973923906795654, |
| "learning_rate": 1e-05, |
| "loss": 0.121, |
| "step": 12820 |
| }, |
| { |
| "epoch": 4.09, |
| "grad_norm": 5.884653479681512, |
| "learning_rate": 1e-05, |
| "loss": 0.1132, |
| "step": 12830 |
| }, |
| { |
| "epoch": 4.1, |
| "grad_norm": 5.100224411458893, |
| "learning_rate": 1e-05, |
| "loss": 0.1112, |
| "step": 12840 |
| }, |
| { |
| "epoch": 4.1, |
| "grad_norm": 5.640553481898148, |
| "learning_rate": 1e-05, |
| "loss": 0.0992, |
| "step": 12850 |
| }, |
| { |
| "epoch": 4.1, |
| "grad_norm": 8.528333675505534, |
| "learning_rate": 1e-05, |
| "loss": 0.1132, |
| "step": 12860 |
| }, |
| { |
| "epoch": 4.11, |
| "grad_norm": 4.882237466140013, |
| "learning_rate": 1e-05, |
| "loss": 0.1093, |
| "step": 12870 |
| }, |
| { |
| "epoch": 4.11, |
| "grad_norm": 7.421303174928847, |
| "learning_rate": 1e-05, |
| "loss": 0.1236, |
| "step": 12880 |
| }, |
| { |
| "epoch": 4.11, |
| "grad_norm": 6.793898638912735, |
| "learning_rate": 1e-05, |
| "loss": 0.1219, |
| "step": 12890 |
| }, |
| { |
| "epoch": 4.12, |
| "grad_norm": 5.551713407501616, |
| "learning_rate": 1e-05, |
| "loss": 0.1317, |
| "step": 12900 |
| }, |
| { |
| "epoch": 4.12, |
| "grad_norm": 4.41058482680067, |
| "learning_rate": 1e-05, |
| "loss": 0.1304, |
| "step": 12910 |
| }, |
| { |
| "epoch": 4.12, |
| "grad_norm": 7.435297535713018, |
| "learning_rate": 1e-05, |
| "loss": 0.1399, |
| "step": 12920 |
| }, |
| { |
| "epoch": 4.13, |
| "grad_norm": 5.1631255576726, |
| "learning_rate": 1e-05, |
| "loss": 0.1181, |
| "step": 12930 |
| }, |
| { |
| "epoch": 4.13, |
| "grad_norm": 6.819817489016557, |
| "learning_rate": 1e-05, |
| "loss": 0.1137, |
| "step": 12940 |
| }, |
| { |
| "epoch": 4.13, |
| "grad_norm": 5.83394187296202, |
| "learning_rate": 1e-05, |
| "loss": 0.1214, |
| "step": 12950 |
| }, |
| { |
| "epoch": 4.14, |
| "grad_norm": 6.828746217450444, |
| "learning_rate": 1e-05, |
| "loss": 0.1243, |
| "step": 12960 |
| }, |
| { |
| "epoch": 4.14, |
| "grad_norm": 5.791962467388132, |
| "learning_rate": 1e-05, |
| "loss": 0.1254, |
| "step": 12970 |
| }, |
| { |
| "epoch": 4.14, |
| "grad_norm": 4.798426076017722, |
| "learning_rate": 1e-05, |
| "loss": 0.1237, |
| "step": 12980 |
| }, |
| { |
| "epoch": 4.14, |
| "grad_norm": 4.330752900456739, |
| "learning_rate": 1e-05, |
| "loss": 0.1164, |
| "step": 12990 |
| }, |
| { |
| "epoch": 4.15, |
| "grad_norm": 7.22705709878385, |
| "learning_rate": 1e-05, |
| "loss": 0.1171, |
| "step": 13000 |
| }, |
| { |
| "epoch": 4.15, |
| "eval_loss": 0.97021484375, |
| "eval_runtime": 54.8169, |
| "eval_samples_per_second": 46.227, |
| "eval_steps_per_second": 2.901, |
| "step": 13000 |
| }, |
| { |
| "epoch": 4.15, |
| "grad_norm": 6.927638012087068, |
| "learning_rate": 1e-05, |
| "loss": 0.1386, |
| "step": 13010 |
| }, |
| { |
| "epoch": 4.15, |
| "grad_norm": 5.138909421770928, |
| "learning_rate": 1e-05, |
| "loss": 0.1272, |
| "step": 13020 |
| }, |
| { |
| "epoch": 4.16, |
| "grad_norm": 6.013054950188329, |
| "learning_rate": 1e-05, |
| "loss": 0.117, |
| "step": 13030 |
| }, |
| { |
| "epoch": 4.16, |
| "grad_norm": 5.557220868485247, |
| "learning_rate": 1e-05, |
| "loss": 0.1186, |
| "step": 13040 |
| }, |
| { |
| "epoch": 4.16, |
| "grad_norm": 8.20176793090263, |
| "learning_rate": 1e-05, |
| "loss": 0.1286, |
| "step": 13050 |
| }, |
| { |
| "epoch": 4.17, |
| "grad_norm": 7.366805098025529, |
| "learning_rate": 1e-05, |
| "loss": 0.1241, |
| "step": 13060 |
| }, |
| { |
| "epoch": 4.17, |
| "grad_norm": 9.128332914173988, |
| "learning_rate": 1e-05, |
| "loss": 0.1117, |
| "step": 13070 |
| }, |
| { |
| "epoch": 4.17, |
| "grad_norm": 5.245989766514141, |
| "learning_rate": 1e-05, |
| "loss": 0.1095, |
| "step": 13080 |
| }, |
| { |
| "epoch": 4.18, |
| "grad_norm": 4.732537747175651, |
| "learning_rate": 1e-05, |
| "loss": 0.1298, |
| "step": 13090 |
| }, |
| { |
| "epoch": 4.18, |
| "grad_norm": 6.995441042102717, |
| "learning_rate": 1e-05, |
| "loss": 0.1043, |
| "step": 13100 |
| }, |
| { |
| "epoch": 4.18, |
| "grad_norm": 7.1091504407357755, |
| "learning_rate": 1e-05, |
| "loss": 0.1371, |
| "step": 13110 |
| }, |
| { |
| "epoch": 4.19, |
| "grad_norm": 5.316328935250181, |
| "learning_rate": 1e-05, |
| "loss": 0.1152, |
| "step": 13120 |
| }, |
| { |
| "epoch": 4.19, |
| "grad_norm": 5.888410156912084, |
| "learning_rate": 1e-05, |
| "loss": 0.1325, |
| "step": 13130 |
| }, |
| { |
| "epoch": 4.19, |
| "grad_norm": 4.302385271882327, |
| "learning_rate": 1e-05, |
| "loss": 0.1374, |
| "step": 13140 |
| }, |
| { |
| "epoch": 4.2, |
| "grad_norm": 4.107033181947425, |
| "learning_rate": 1e-05, |
| "loss": 0.1088, |
| "step": 13150 |
| }, |
| { |
| "epoch": 4.2, |
| "grad_norm": 4.8114833005543804, |
| "learning_rate": 1e-05, |
| "loss": 0.1245, |
| "step": 13160 |
| }, |
| { |
| "epoch": 4.2, |
| "grad_norm": 6.704428663838261, |
| "learning_rate": 1e-05, |
| "loss": 0.127, |
| "step": 13170 |
| }, |
| { |
| "epoch": 4.21, |
| "grad_norm": 4.782864167246858, |
| "learning_rate": 1e-05, |
| "loss": 0.1193, |
| "step": 13180 |
| }, |
| { |
| "epoch": 4.21, |
| "grad_norm": 7.842228836950852, |
| "learning_rate": 1e-05, |
| "loss": 0.1169, |
| "step": 13190 |
| }, |
| { |
| "epoch": 4.21, |
| "grad_norm": 4.670092596733018, |
| "learning_rate": 1e-05, |
| "loss": 0.1219, |
| "step": 13200 |
| }, |
| { |
| "epoch": 4.22, |
| "grad_norm": 9.313399508256405, |
| "learning_rate": 1e-05, |
| "loss": 0.13, |
| "step": 13210 |
| }, |
| { |
| "epoch": 4.22, |
| "grad_norm": 4.439192328321902, |
| "learning_rate": 1e-05, |
| "loss": 0.1343, |
| "step": 13220 |
| }, |
| { |
| "epoch": 4.22, |
| "grad_norm": 6.509472327344879, |
| "learning_rate": 1e-05, |
| "loss": 0.1456, |
| "step": 13230 |
| }, |
| { |
| "epoch": 4.22, |
| "grad_norm": 5.627531541611389, |
| "learning_rate": 1e-05, |
| "loss": 0.0957, |
| "step": 13240 |
| }, |
| { |
| "epoch": 4.23, |
| "grad_norm": 9.002575823616313, |
| "learning_rate": 1e-05, |
| "loss": 0.1158, |
| "step": 13250 |
| }, |
| { |
| "epoch": 4.23, |
| "grad_norm": 7.217429564620889, |
| "learning_rate": 1e-05, |
| "loss": 0.1181, |
| "step": 13260 |
| }, |
| { |
| "epoch": 4.23, |
| "grad_norm": 5.851143958245873, |
| "learning_rate": 1e-05, |
| "loss": 0.1204, |
| "step": 13270 |
| }, |
| { |
| "epoch": 4.24, |
| "grad_norm": 7.513496462324014, |
| "learning_rate": 1e-05, |
| "loss": 0.1257, |
| "step": 13280 |
| }, |
| { |
| "epoch": 4.24, |
| "grad_norm": 6.42568982647729, |
| "learning_rate": 1e-05, |
| "loss": 0.134, |
| "step": 13290 |
| }, |
| { |
| "epoch": 4.24, |
| "grad_norm": 5.7063012880191, |
| "learning_rate": 1e-05, |
| "loss": 0.1334, |
| "step": 13300 |
| }, |
| { |
| "epoch": 4.25, |
| "grad_norm": 6.142744868066731, |
| "learning_rate": 1e-05, |
| "loss": 0.1411, |
| "step": 13310 |
| }, |
| { |
| "epoch": 4.25, |
| "grad_norm": 6.662616389476458, |
| "learning_rate": 1e-05, |
| "loss": 0.119, |
| "step": 13320 |
| }, |
| { |
| "epoch": 4.25, |
| "grad_norm": 6.939250828215159, |
| "learning_rate": 1e-05, |
| "loss": 0.146, |
| "step": 13330 |
| }, |
| { |
| "epoch": 4.26, |
| "grad_norm": 6.1031974917089755, |
| "learning_rate": 1e-05, |
| "loss": 0.1338, |
| "step": 13340 |
| }, |
| { |
| "epoch": 4.26, |
| "grad_norm": 5.284915532327635, |
| "learning_rate": 1e-05, |
| "loss": 0.129, |
| "step": 13350 |
| }, |
| { |
| "epoch": 4.26, |
| "grad_norm": 9.78920941193131, |
| "learning_rate": 1e-05, |
| "loss": 0.1385, |
| "step": 13360 |
| }, |
| { |
| "epoch": 4.27, |
| "grad_norm": 6.449504732737811, |
| "learning_rate": 1e-05, |
| "loss": 0.1328, |
| "step": 13370 |
| }, |
| { |
| "epoch": 4.27, |
| "grad_norm": 3.8383752725562403, |
| "learning_rate": 1e-05, |
| "loss": 0.1323, |
| "step": 13380 |
| }, |
| { |
| "epoch": 4.27, |
| "grad_norm": 7.594109923070751, |
| "learning_rate": 1e-05, |
| "loss": 0.1339, |
| "step": 13390 |
| }, |
| { |
| "epoch": 4.28, |
| "grad_norm": 6.029121299982473, |
| "learning_rate": 1e-05, |
| "loss": 0.1528, |
| "step": 13400 |
| }, |
| { |
| "epoch": 4.28, |
| "grad_norm": 5.685231521441548, |
| "learning_rate": 1e-05, |
| "loss": 0.1072, |
| "step": 13410 |
| }, |
| { |
| "epoch": 4.28, |
| "grad_norm": 7.489460469716187, |
| "learning_rate": 1e-05, |
| "loss": 0.127, |
| "step": 13420 |
| }, |
| { |
| "epoch": 4.29, |
| "grad_norm": 4.550053170700445, |
| "learning_rate": 1e-05, |
| "loss": 0.1285, |
| "step": 13430 |
| }, |
| { |
| "epoch": 4.29, |
| "grad_norm": 4.662021209432894, |
| "learning_rate": 1e-05, |
| "loss": 0.1197, |
| "step": 13440 |
| }, |
| { |
| "epoch": 4.29, |
| "grad_norm": 7.282948570068827, |
| "learning_rate": 1e-05, |
| "loss": 0.1101, |
| "step": 13450 |
| }, |
| { |
| "epoch": 4.29, |
| "grad_norm": 6.306675102491698, |
| "learning_rate": 1e-05, |
| "loss": 0.1246, |
| "step": 13460 |
| }, |
| { |
| "epoch": 4.3, |
| "grad_norm": 6.272306993600732, |
| "learning_rate": 1e-05, |
| "loss": 0.1211, |
| "step": 13470 |
| }, |
| { |
| "epoch": 4.3, |
| "grad_norm": 6.563054524562691, |
| "learning_rate": 1e-05, |
| "loss": 0.1036, |
| "step": 13480 |
| }, |
| { |
| "epoch": 4.3, |
| "grad_norm": 5.004822693998685, |
| "learning_rate": 1e-05, |
| "loss": 0.1406, |
| "step": 13490 |
| }, |
| { |
| "epoch": 4.31, |
| "grad_norm": 5.508069447685581, |
| "learning_rate": 1e-05, |
| "loss": 0.155, |
| "step": 13500 |
| }, |
| { |
| "epoch": 4.31, |
| "eval_loss": 0.95068359375, |
| "eval_runtime": 55.0449, |
| "eval_samples_per_second": 46.035, |
| "eval_steps_per_second": 2.889, |
| "step": 13500 |
| }, |
| { |
| "epoch": 4.31, |
| "grad_norm": 6.718039399649584, |
| "learning_rate": 1e-05, |
| "loss": 0.1386, |
| "step": 13510 |
| }, |
| { |
| "epoch": 4.31, |
| "grad_norm": 7.273174735955825, |
| "learning_rate": 1e-05, |
| "loss": 0.1212, |
| "step": 13520 |
| }, |
| { |
| "epoch": 4.32, |
| "grad_norm": 5.084425365946469, |
| "learning_rate": 1e-05, |
| "loss": 0.152, |
| "step": 13530 |
| }, |
| { |
| "epoch": 4.32, |
| "grad_norm": 7.224834041603374, |
| "learning_rate": 1e-05, |
| "loss": 0.1329, |
| "step": 13540 |
| }, |
| { |
| "epoch": 4.32, |
| "grad_norm": 8.703953811382709, |
| "learning_rate": 1e-05, |
| "loss": 0.1373, |
| "step": 13550 |
| }, |
| { |
| "epoch": 4.33, |
| "grad_norm": 5.215743060915696, |
| "learning_rate": 1e-05, |
| "loss": 0.1151, |
| "step": 13560 |
| }, |
| { |
| "epoch": 4.33, |
| "grad_norm": 7.792982945632434, |
| "learning_rate": 1e-05, |
| "loss": 0.1404, |
| "step": 13570 |
| }, |
| { |
| "epoch": 4.33, |
| "grad_norm": 5.909935825470666, |
| "learning_rate": 1e-05, |
| "loss": 0.1281, |
| "step": 13580 |
| }, |
| { |
| "epoch": 4.34, |
| "grad_norm": 6.861261964170061, |
| "learning_rate": 1e-05, |
| "loss": 0.1288, |
| "step": 13590 |
| }, |
| { |
| "epoch": 4.34, |
| "grad_norm": 5.798218604818019, |
| "learning_rate": 1e-05, |
| "loss": 0.1164, |
| "step": 13600 |
| }, |
| { |
| "epoch": 4.34, |
| "grad_norm": 8.107822512158975, |
| "learning_rate": 1e-05, |
| "loss": 0.1441, |
| "step": 13610 |
| }, |
| { |
| "epoch": 4.35, |
| "grad_norm": 5.559605048717893, |
| "learning_rate": 1e-05, |
| "loss": 0.1184, |
| "step": 13620 |
| }, |
| { |
| "epoch": 4.35, |
| "grad_norm": 4.885314007665342, |
| "learning_rate": 1e-05, |
| "loss": 0.1349, |
| "step": 13630 |
| }, |
| { |
| "epoch": 4.35, |
| "grad_norm": 6.254609201299976, |
| "learning_rate": 1e-05, |
| "loss": 0.1024, |
| "step": 13640 |
| }, |
| { |
| "epoch": 4.36, |
| "grad_norm": 6.913565097309515, |
| "learning_rate": 1e-05, |
| "loss": 0.1474, |
| "step": 13650 |
| }, |
| { |
| "epoch": 4.36, |
| "grad_norm": 6.203859223313768, |
| "learning_rate": 1e-05, |
| "loss": 0.1343, |
| "step": 13660 |
| }, |
| { |
| "epoch": 4.36, |
| "grad_norm": 8.1173914570743, |
| "learning_rate": 1e-05, |
| "loss": 0.1385, |
| "step": 13670 |
| }, |
| { |
| "epoch": 4.37, |
| "grad_norm": 8.171768100250848, |
| "learning_rate": 1e-05, |
| "loss": 0.1397, |
| "step": 13680 |
| }, |
| { |
| "epoch": 4.37, |
| "grad_norm": 6.4440632667206215, |
| "learning_rate": 1e-05, |
| "loss": 0.1532, |
| "step": 13690 |
| }, |
| { |
| "epoch": 4.37, |
| "grad_norm": 6.67971748211473, |
| "learning_rate": 1e-05, |
| "loss": 0.1329, |
| "step": 13700 |
| }, |
| { |
| "epoch": 4.37, |
| "grad_norm": 3.378923607930487, |
| "learning_rate": 1e-05, |
| "loss": 0.116, |
| "step": 13710 |
| }, |
| { |
| "epoch": 4.38, |
| "grad_norm": 5.9150528475415145, |
| "learning_rate": 1e-05, |
| "loss": 0.1441, |
| "step": 13720 |
| }, |
| { |
| "epoch": 4.38, |
| "grad_norm": 5.315527916974269, |
| "learning_rate": 1e-05, |
| "loss": 0.114, |
| "step": 13730 |
| }, |
| { |
| "epoch": 4.38, |
| "grad_norm": 7.546474406665604, |
| "learning_rate": 1e-05, |
| "loss": 0.1392, |
| "step": 13740 |
| }, |
| { |
| "epoch": 4.39, |
| "grad_norm": 7.181791742273966, |
| "learning_rate": 1e-05, |
| "loss": 0.1369, |
| "step": 13750 |
| }, |
| { |
| "epoch": 4.39, |
| "grad_norm": 9.17781646490474, |
| "learning_rate": 1e-05, |
| "loss": 0.1448, |
| "step": 13760 |
| }, |
| { |
| "epoch": 4.39, |
| "grad_norm": 5.363566458412592, |
| "learning_rate": 1e-05, |
| "loss": 0.1316, |
| "step": 13770 |
| }, |
| { |
| "epoch": 4.4, |
| "grad_norm": 5.671349367108222, |
| "learning_rate": 1e-05, |
| "loss": 0.1456, |
| "step": 13780 |
| }, |
| { |
| "epoch": 4.4, |
| "grad_norm": 5.639312337793366, |
| "learning_rate": 1e-05, |
| "loss": 0.1197, |
| "step": 13790 |
| }, |
| { |
| "epoch": 4.4, |
| "grad_norm": 3.9482615109748505, |
| "learning_rate": 1e-05, |
| "loss": 0.1313, |
| "step": 13800 |
| }, |
| { |
| "epoch": 4.41, |
| "grad_norm": 7.447992697018691, |
| "learning_rate": 1e-05, |
| "loss": 0.1532, |
| "step": 13810 |
| }, |
| { |
| "epoch": 4.41, |
| "grad_norm": 9.049496496536165, |
| "learning_rate": 1e-05, |
| "loss": 0.1251, |
| "step": 13820 |
| }, |
| { |
| "epoch": 4.41, |
| "grad_norm": 5.480195542348197, |
| "learning_rate": 1e-05, |
| "loss": 0.1483, |
| "step": 13830 |
| }, |
| { |
| "epoch": 4.42, |
| "grad_norm": 5.077700177542741, |
| "learning_rate": 1e-05, |
| "loss": 0.134, |
| "step": 13840 |
| }, |
| { |
| "epoch": 4.42, |
| "grad_norm": 4.7630644376193425, |
| "learning_rate": 1e-05, |
| "loss": 0.1165, |
| "step": 13850 |
| }, |
| { |
| "epoch": 4.42, |
| "grad_norm": 6.2738862405665525, |
| "learning_rate": 1e-05, |
| "loss": 0.1165, |
| "step": 13860 |
| }, |
| { |
| "epoch": 4.43, |
| "grad_norm": 8.003157469397967, |
| "learning_rate": 1e-05, |
| "loss": 0.1637, |
| "step": 13870 |
| }, |
| { |
| "epoch": 4.43, |
| "grad_norm": 7.224790481644579, |
| "learning_rate": 1e-05, |
| "loss": 0.1281, |
| "step": 13880 |
| }, |
| { |
| "epoch": 4.43, |
| "grad_norm": 3.7071117717973694, |
| "learning_rate": 1e-05, |
| "loss": 0.1254, |
| "step": 13890 |
| }, |
| { |
| "epoch": 4.44, |
| "grad_norm": 7.690870337078116, |
| "learning_rate": 1e-05, |
| "loss": 0.1255, |
| "step": 13900 |
| }, |
| { |
| "epoch": 4.44, |
| "grad_norm": 6.644319973860309, |
| "learning_rate": 1e-05, |
| "loss": 0.1088, |
| "step": 13910 |
| }, |
| { |
| "epoch": 4.44, |
| "grad_norm": 5.3057671460915925, |
| "learning_rate": 1e-05, |
| "loss": 0.1624, |
| "step": 13920 |
| }, |
| { |
| "epoch": 4.44, |
| "grad_norm": 7.176535029476055, |
| "learning_rate": 1e-05, |
| "loss": 0.1443, |
| "step": 13930 |
| }, |
| { |
| "epoch": 4.45, |
| "grad_norm": 5.294844606765154, |
| "learning_rate": 1e-05, |
| "loss": 0.1301, |
| "step": 13940 |
| }, |
| { |
| "epoch": 4.45, |
| "grad_norm": 6.298106790092106, |
| "learning_rate": 1e-05, |
| "loss": 0.1497, |
| "step": 13950 |
| }, |
| { |
| "epoch": 4.45, |
| "grad_norm": 7.630230969774088, |
| "learning_rate": 1e-05, |
| "loss": 0.1577, |
| "step": 13960 |
| }, |
| { |
| "epoch": 4.46, |
| "grad_norm": 7.311672587737599, |
| "learning_rate": 1e-05, |
| "loss": 0.1233, |
| "step": 13970 |
| }, |
| { |
| "epoch": 4.46, |
| "grad_norm": 5.206259210218913, |
| "learning_rate": 1e-05, |
| "loss": 0.1319, |
| "step": 13980 |
| }, |
| { |
| "epoch": 4.46, |
| "grad_norm": 5.521746213282215, |
| "learning_rate": 1e-05, |
| "loss": 0.156, |
| "step": 13990 |
| }, |
| { |
| "epoch": 4.47, |
| "grad_norm": 6.335964292194432, |
| "learning_rate": 1e-05, |
| "loss": 0.1211, |
| "step": 14000 |
| }, |
| { |
| "epoch": 4.47, |
| "eval_loss": 0.9658203125, |
| "eval_runtime": 54.856, |
| "eval_samples_per_second": 46.194, |
| "eval_steps_per_second": 2.898, |
| "step": 14000 |
| }, |
| { |
| "epoch": 4.47, |
| "grad_norm": 5.887553337487297, |
| "learning_rate": 1e-05, |
| "loss": 0.1369, |
| "step": 14010 |
| }, |
| { |
| "epoch": 4.47, |
| "grad_norm": 7.53121277099525, |
| "learning_rate": 1e-05, |
| "loss": 0.1453, |
| "step": 14020 |
| }, |
| { |
| "epoch": 4.48, |
| "grad_norm": 6.255973707685319, |
| "learning_rate": 1e-05, |
| "loss": 0.1317, |
| "step": 14030 |
| }, |
| { |
| "epoch": 4.48, |
| "grad_norm": 5.306134707722658, |
| "learning_rate": 1e-05, |
| "loss": 0.115, |
| "step": 14040 |
| }, |
| { |
| "epoch": 4.48, |
| "grad_norm": 6.110613685169027, |
| "learning_rate": 1e-05, |
| "loss": 0.1296, |
| "step": 14050 |
| }, |
| { |
| "epoch": 4.49, |
| "grad_norm": 5.561268873810761, |
| "learning_rate": 1e-05, |
| "loss": 0.147, |
| "step": 14060 |
| }, |
| { |
| "epoch": 4.49, |
| "grad_norm": 7.603416538254314, |
| "learning_rate": 1e-05, |
| "loss": 0.1116, |
| "step": 14070 |
| }, |
| { |
| "epoch": 4.49, |
| "grad_norm": 6.088717999500503, |
| "learning_rate": 1e-05, |
| "loss": 0.1223, |
| "step": 14080 |
| }, |
| { |
| "epoch": 4.5, |
| "grad_norm": 5.499361521334452, |
| "learning_rate": 1e-05, |
| "loss": 0.1409, |
| "step": 14090 |
| }, |
| { |
| "epoch": 4.5, |
| "grad_norm": 7.803625574887426, |
| "learning_rate": 1e-05, |
| "loss": 0.1325, |
| "step": 14100 |
| }, |
| { |
| "epoch": 4.5, |
| "grad_norm": 6.844993164861595, |
| "learning_rate": 1e-05, |
| "loss": 0.1587, |
| "step": 14110 |
| }, |
| { |
| "epoch": 4.51, |
| "grad_norm": 6.512036842673115, |
| "learning_rate": 1e-05, |
| "loss": 0.1619, |
| "step": 14120 |
| }, |
| { |
| "epoch": 4.51, |
| "grad_norm": 5.03839154175248, |
| "learning_rate": 1e-05, |
| "loss": 0.1178, |
| "step": 14130 |
| }, |
| { |
| "epoch": 4.51, |
| "grad_norm": 4.347831729387686, |
| "learning_rate": 1e-05, |
| "loss": 0.12, |
| "step": 14140 |
| }, |
| { |
| "epoch": 4.51, |
| "grad_norm": 5.042417086825984, |
| "learning_rate": 1e-05, |
| "loss": 0.1443, |
| "step": 14150 |
| }, |
| { |
| "epoch": 4.52, |
| "grad_norm": 6.681496751045271, |
| "learning_rate": 1e-05, |
| "loss": 0.1436, |
| "step": 14160 |
| }, |
| { |
| "epoch": 4.52, |
| "grad_norm": 7.813387400773147, |
| "learning_rate": 1e-05, |
| "loss": 0.1438, |
| "step": 14170 |
| }, |
| { |
| "epoch": 4.52, |
| "grad_norm": 5.374998048294068, |
| "learning_rate": 1e-05, |
| "loss": 0.1369, |
| "step": 14180 |
| }, |
| { |
| "epoch": 4.53, |
| "grad_norm": 6.032110755652372, |
| "learning_rate": 1e-05, |
| "loss": 0.1355, |
| "step": 14190 |
| }, |
| { |
| "epoch": 4.53, |
| "grad_norm": 7.000496437997663, |
| "learning_rate": 1e-05, |
| "loss": 0.1403, |
| "step": 14200 |
| }, |
| { |
| "epoch": 4.53, |
| "grad_norm": 5.339291145624496, |
| "learning_rate": 1e-05, |
| "loss": 0.1213, |
| "step": 14210 |
| }, |
| { |
| "epoch": 4.54, |
| "grad_norm": 6.017527727697919, |
| "learning_rate": 1e-05, |
| "loss": 0.1382, |
| "step": 14220 |
| }, |
| { |
| "epoch": 4.54, |
| "grad_norm": 4.89425417274897, |
| "learning_rate": 1e-05, |
| "loss": 0.1238, |
| "step": 14230 |
| }, |
| { |
| "epoch": 4.54, |
| "grad_norm": 7.128386913874449, |
| "learning_rate": 1e-05, |
| "loss": 0.1422, |
| "step": 14240 |
| }, |
| { |
| "epoch": 4.55, |
| "grad_norm": 5.343512457313984, |
| "learning_rate": 1e-05, |
| "loss": 0.1333, |
| "step": 14250 |
| }, |
| { |
| "epoch": 4.55, |
| "grad_norm": 5.102499532592094, |
| "learning_rate": 1e-05, |
| "loss": 0.1462, |
| "step": 14260 |
| }, |
| { |
| "epoch": 4.55, |
| "grad_norm": 4.830084245788176, |
| "learning_rate": 1e-05, |
| "loss": 0.1498, |
| "step": 14270 |
| }, |
| { |
| "epoch": 4.56, |
| "grad_norm": 5.138553654272466, |
| "learning_rate": 1e-05, |
| "loss": 0.1677, |
| "step": 14280 |
| }, |
| { |
| "epoch": 4.56, |
| "grad_norm": 6.913729798669015, |
| "learning_rate": 1e-05, |
| "loss": 0.1543, |
| "step": 14290 |
| }, |
| { |
| "epoch": 4.56, |
| "grad_norm": 5.142639136613487, |
| "learning_rate": 1e-05, |
| "loss": 0.1364, |
| "step": 14300 |
| }, |
| { |
| "epoch": 4.57, |
| "grad_norm": 6.1977952790327615, |
| "learning_rate": 1e-05, |
| "loss": 0.135, |
| "step": 14310 |
| }, |
| { |
| "epoch": 4.57, |
| "grad_norm": 5.28671938497484, |
| "learning_rate": 1e-05, |
| "loss": 0.157, |
| "step": 14320 |
| }, |
| { |
| "epoch": 4.57, |
| "grad_norm": 5.836241115449739, |
| "learning_rate": 1e-05, |
| "loss": 0.1462, |
| "step": 14330 |
| }, |
| { |
| "epoch": 4.58, |
| "grad_norm": 7.420825121055941, |
| "learning_rate": 1e-05, |
| "loss": 0.1577, |
| "step": 14340 |
| }, |
| { |
| "epoch": 4.58, |
| "grad_norm": 6.116583822384028, |
| "learning_rate": 1e-05, |
| "loss": 0.1473, |
| "step": 14350 |
| }, |
| { |
| "epoch": 4.58, |
| "grad_norm": 6.7071927761563135, |
| "learning_rate": 1e-05, |
| "loss": 0.1604, |
| "step": 14360 |
| }, |
| { |
| "epoch": 4.59, |
| "grad_norm": 7.675662968236237, |
| "learning_rate": 1e-05, |
| "loss": 0.1317, |
| "step": 14370 |
| }, |
| { |
| "epoch": 4.59, |
| "grad_norm": 4.245920579567305, |
| "learning_rate": 1e-05, |
| "loss": 0.1157, |
| "step": 14380 |
| }, |
| { |
| "epoch": 4.59, |
| "grad_norm": 5.29186865153258, |
| "learning_rate": 1e-05, |
| "loss": 0.127, |
| "step": 14390 |
| }, |
| { |
| "epoch": 4.59, |
| "grad_norm": 8.742926654129736, |
| "learning_rate": 1e-05, |
| "loss": 0.1413, |
| "step": 14400 |
| }, |
| { |
| "epoch": 4.6, |
| "grad_norm": 5.93663869935091, |
| "learning_rate": 1e-05, |
| "loss": 0.141, |
| "step": 14410 |
| }, |
| { |
| "epoch": 4.6, |
| "grad_norm": 5.8012198645202115, |
| "learning_rate": 1e-05, |
| "loss": 0.1349, |
| "step": 14420 |
| }, |
| { |
| "epoch": 4.6, |
| "grad_norm": 6.96655266165044, |
| "learning_rate": 1e-05, |
| "loss": 0.1325, |
| "step": 14430 |
| }, |
| { |
| "epoch": 4.61, |
| "grad_norm": 6.631032824038542, |
| "learning_rate": 1e-05, |
| "loss": 0.1642, |
| "step": 14440 |
| }, |
| { |
| "epoch": 4.61, |
| "grad_norm": 9.600318871606476, |
| "learning_rate": 1e-05, |
| "loss": 0.1461, |
| "step": 14450 |
| }, |
| { |
| "epoch": 4.61, |
| "grad_norm": 5.930933805022658, |
| "learning_rate": 1e-05, |
| "loss": 0.1561, |
| "step": 14460 |
| }, |
| { |
| "epoch": 4.62, |
| "grad_norm": 4.974193542138597, |
| "learning_rate": 1e-05, |
| "loss": 0.1262, |
| "step": 14470 |
| }, |
| { |
| "epoch": 4.62, |
| "grad_norm": 4.227799944848056, |
| "learning_rate": 1e-05, |
| "loss": 0.1307, |
| "step": 14480 |
| }, |
| { |
| "epoch": 4.62, |
| "grad_norm": 6.509472913367355, |
| "learning_rate": 1e-05, |
| "loss": 0.1287, |
| "step": 14490 |
| }, |
| { |
| "epoch": 4.63, |
| "grad_norm": 7.448552998566965, |
| "learning_rate": 1e-05, |
| "loss": 0.1401, |
| "step": 14500 |
| }, |
| { |
| "epoch": 4.63, |
| "eval_loss": 0.9736328125, |
| "eval_runtime": 54.8242, |
| "eval_samples_per_second": 46.22, |
| "eval_steps_per_second": 2.9, |
| "step": 14500 |
| }, |
| { |
| "epoch": 4.63, |
| "grad_norm": 5.432508962301699, |
| "learning_rate": 1e-05, |
| "loss": 0.1462, |
| "step": 14510 |
| }, |
| { |
| "epoch": 4.63, |
| "grad_norm": 3.0144560925235653, |
| "learning_rate": 1e-05, |
| "loss": 0.1595, |
| "step": 14520 |
| }, |
| { |
| "epoch": 4.64, |
| "grad_norm": 5.897889047349378, |
| "learning_rate": 1e-05, |
| "loss": 0.1472, |
| "step": 14530 |
| }, |
| { |
| "epoch": 4.64, |
| "grad_norm": 8.820103604839211, |
| "learning_rate": 1e-05, |
| "loss": 0.1341, |
| "step": 14540 |
| }, |
| { |
| "epoch": 4.64, |
| "grad_norm": 7.3733992051578925, |
| "learning_rate": 1e-05, |
| "loss": 0.1445, |
| "step": 14550 |
| }, |
| { |
| "epoch": 4.65, |
| "grad_norm": 7.860892359816852, |
| "learning_rate": 1e-05, |
| "loss": 0.1406, |
| "step": 14560 |
| }, |
| { |
| "epoch": 4.65, |
| "grad_norm": 5.308770699601028, |
| "learning_rate": 1e-05, |
| "loss": 0.1637, |
| "step": 14570 |
| }, |
| { |
| "epoch": 4.65, |
| "grad_norm": 7.5074580939536855, |
| "learning_rate": 1e-05, |
| "loss": 0.1476, |
| "step": 14580 |
| }, |
| { |
| "epoch": 4.66, |
| "grad_norm": 8.265944361150796, |
| "learning_rate": 1e-05, |
| "loss": 0.143, |
| "step": 14590 |
| }, |
| { |
| "epoch": 4.66, |
| "grad_norm": 6.296158288344928, |
| "learning_rate": 1e-05, |
| "loss": 0.1371, |
| "step": 14600 |
| }, |
| { |
| "epoch": 4.66, |
| "grad_norm": 6.582382165627064, |
| "learning_rate": 1e-05, |
| "loss": 0.139, |
| "step": 14610 |
| }, |
| { |
| "epoch": 4.66, |
| "grad_norm": 8.020146275277993, |
| "learning_rate": 1e-05, |
| "loss": 0.1278, |
| "step": 14620 |
| }, |
| { |
| "epoch": 4.67, |
| "grad_norm": 10.449338127238647, |
| "learning_rate": 1e-05, |
| "loss": 0.147, |
| "step": 14630 |
| }, |
| { |
| "epoch": 4.67, |
| "grad_norm": 7.605049172668194, |
| "learning_rate": 1e-05, |
| "loss": 0.1132, |
| "step": 14640 |
| }, |
| { |
| "epoch": 4.67, |
| "grad_norm": 7.364097165504288, |
| "learning_rate": 1e-05, |
| "loss": 0.1349, |
| "step": 14650 |
| }, |
| { |
| "epoch": 4.68, |
| "grad_norm": 7.2340446380258925, |
| "learning_rate": 1e-05, |
| "loss": 0.1628, |
| "step": 14660 |
| }, |
| { |
| "epoch": 4.68, |
| "grad_norm": 6.9366130089995, |
| "learning_rate": 1e-05, |
| "loss": 0.1469, |
| "step": 14670 |
| }, |
| { |
| "epoch": 4.68, |
| "grad_norm": 5.691889148718449, |
| "learning_rate": 1e-05, |
| "loss": 0.1588, |
| "step": 14680 |
| }, |
| { |
| "epoch": 4.69, |
| "grad_norm": 7.1650720049373895, |
| "learning_rate": 1e-05, |
| "loss": 0.1712, |
| "step": 14690 |
| }, |
| { |
| "epoch": 4.69, |
| "grad_norm": 3.904990153281952, |
| "learning_rate": 1e-05, |
| "loss": 0.163, |
| "step": 14700 |
| }, |
| { |
| "epoch": 4.69, |
| "grad_norm": 5.407540757108995, |
| "learning_rate": 1e-05, |
| "loss": 0.1463, |
| "step": 14710 |
| }, |
| { |
| "epoch": 4.7, |
| "grad_norm": 6.11323226567835, |
| "learning_rate": 1e-05, |
| "loss": 0.142, |
| "step": 14720 |
| }, |
| { |
| "epoch": 4.7, |
| "grad_norm": 5.379007309440076, |
| "learning_rate": 1e-05, |
| "loss": 0.1368, |
| "step": 14730 |
| }, |
| { |
| "epoch": 4.7, |
| "grad_norm": 6.546218292876299, |
| "learning_rate": 1e-05, |
| "loss": 0.1329, |
| "step": 14740 |
| }, |
| { |
| "epoch": 4.71, |
| "grad_norm": 6.847071290260256, |
| "learning_rate": 1e-05, |
| "loss": 0.1511, |
| "step": 14750 |
| }, |
| { |
| "epoch": 4.71, |
| "grad_norm": 6.885505556426349, |
| "learning_rate": 1e-05, |
| "loss": 0.1519, |
| "step": 14760 |
| }, |
| { |
| "epoch": 4.71, |
| "grad_norm": 6.693170442967464, |
| "learning_rate": 1e-05, |
| "loss": 0.1375, |
| "step": 14770 |
| }, |
| { |
| "epoch": 4.72, |
| "grad_norm": 4.2076849611079, |
| "learning_rate": 1e-05, |
| "loss": 0.1108, |
| "step": 14780 |
| }, |
| { |
| "epoch": 4.72, |
| "grad_norm": 9.832814175969188, |
| "learning_rate": 1e-05, |
| "loss": 0.1467, |
| "step": 14790 |
| }, |
| { |
| "epoch": 4.72, |
| "grad_norm": 7.095558776143224, |
| "learning_rate": 1e-05, |
| "loss": 0.136, |
| "step": 14800 |
| }, |
| { |
| "epoch": 4.73, |
| "grad_norm": 6.541569638190732, |
| "learning_rate": 1e-05, |
| "loss": 0.1462, |
| "step": 14810 |
| }, |
| { |
| "epoch": 4.73, |
| "grad_norm": 5.283334715609365, |
| "learning_rate": 1e-05, |
| "loss": 0.1532, |
| "step": 14820 |
| }, |
| { |
| "epoch": 4.73, |
| "grad_norm": 10.01764114725984, |
| "learning_rate": 1e-05, |
| "loss": 0.1435, |
| "step": 14830 |
| }, |
| { |
| "epoch": 4.74, |
| "grad_norm": 6.341890410170249, |
| "learning_rate": 1e-05, |
| "loss": 0.1405, |
| "step": 14840 |
| }, |
| { |
| "epoch": 4.74, |
| "grad_norm": 6.249421054728758, |
| "learning_rate": 1e-05, |
| "loss": 0.1381, |
| "step": 14850 |
| }, |
| { |
| "epoch": 4.74, |
| "grad_norm": 4.9708972825538424, |
| "learning_rate": 1e-05, |
| "loss": 0.1505, |
| "step": 14860 |
| }, |
| { |
| "epoch": 4.74, |
| "grad_norm": 5.405950990026397, |
| "learning_rate": 1e-05, |
| "loss": 0.1546, |
| "step": 14870 |
| }, |
| { |
| "epoch": 4.75, |
| "grad_norm": 5.691629775873975, |
| "learning_rate": 1e-05, |
| "loss": 0.178, |
| "step": 14880 |
| }, |
| { |
| "epoch": 4.75, |
| "grad_norm": 7.669063635695029, |
| "learning_rate": 1e-05, |
| "loss": 0.1728, |
| "step": 14890 |
| }, |
| { |
| "epoch": 4.75, |
| "grad_norm": 5.200730331192265, |
| "learning_rate": 1e-05, |
| "loss": 0.1409, |
| "step": 14900 |
| }, |
| { |
| "epoch": 4.76, |
| "grad_norm": 7.607347911626966, |
| "learning_rate": 1e-05, |
| "loss": 0.1472, |
| "step": 14910 |
| }, |
| { |
| "epoch": 4.76, |
| "grad_norm": 6.424182041215959, |
| "learning_rate": 1e-05, |
| "loss": 0.16, |
| "step": 14920 |
| }, |
| { |
| "epoch": 4.76, |
| "grad_norm": 7.432878869822717, |
| "learning_rate": 1e-05, |
| "loss": 0.1462, |
| "step": 14930 |
| }, |
| { |
| "epoch": 4.77, |
| "grad_norm": 5.547673729354768, |
| "learning_rate": 1e-05, |
| "loss": 0.1489, |
| "step": 14940 |
| }, |
| { |
| "epoch": 4.77, |
| "grad_norm": 7.3450952474652045, |
| "learning_rate": 1e-05, |
| "loss": 0.1455, |
| "step": 14950 |
| }, |
| { |
| "epoch": 4.77, |
| "grad_norm": 8.817306614554049, |
| "learning_rate": 1e-05, |
| "loss": 0.1458, |
| "step": 14960 |
| }, |
| { |
| "epoch": 4.78, |
| "grad_norm": 4.553560251601577, |
| "learning_rate": 1e-05, |
| "loss": 0.1529, |
| "step": 14970 |
| }, |
| { |
| "epoch": 4.78, |
| "grad_norm": 5.003466548853221, |
| "learning_rate": 1e-05, |
| "loss": 0.138, |
| "step": 14980 |
| }, |
| { |
| "epoch": 4.78, |
| "grad_norm": 7.807576085042686, |
| "learning_rate": 1e-05, |
| "loss": 0.1583, |
| "step": 14990 |
| }, |
| { |
| "epoch": 4.79, |
| "grad_norm": 5.807199337714284, |
| "learning_rate": 1e-05, |
| "loss": 0.1269, |
| "step": 15000 |
| }, |
| { |
| "epoch": 4.79, |
| "eval_loss": 0.95166015625, |
| "eval_runtime": 54.7753, |
| "eval_samples_per_second": 46.262, |
| "eval_steps_per_second": 2.903, |
| "step": 15000 |
| }, |
| { |
| "epoch": 4.79, |
| "grad_norm": 7.050138474856252, |
| "learning_rate": 1e-05, |
| "loss": 0.1792, |
| "step": 15010 |
| }, |
| { |
| "epoch": 4.79, |
| "grad_norm": 7.693630100434674, |
| "learning_rate": 1e-05, |
| "loss": 0.1525, |
| "step": 15020 |
| }, |
| { |
| "epoch": 4.8, |
| "grad_norm": 7.097207189030993, |
| "learning_rate": 1e-05, |
| "loss": 0.1581, |
| "step": 15030 |
| }, |
| { |
| "epoch": 4.8, |
| "grad_norm": 6.118873490518394, |
| "learning_rate": 1e-05, |
| "loss": 0.1379, |
| "step": 15040 |
| }, |
| { |
| "epoch": 4.8, |
| "grad_norm": 5.591432059189683, |
| "learning_rate": 1e-05, |
| "loss": 0.151, |
| "step": 15050 |
| }, |
| { |
| "epoch": 4.81, |
| "grad_norm": 8.445967882612454, |
| "learning_rate": 1e-05, |
| "loss": 0.1471, |
| "step": 15060 |
| }, |
| { |
| "epoch": 4.81, |
| "grad_norm": 4.1204309301097, |
| "learning_rate": 1e-05, |
| "loss": 0.1276, |
| "step": 15070 |
| }, |
| { |
| "epoch": 4.81, |
| "grad_norm": 6.040150138555255, |
| "learning_rate": 1e-05, |
| "loss": 0.1459, |
| "step": 15080 |
| }, |
| { |
| "epoch": 4.81, |
| "grad_norm": 6.832961126595728, |
| "learning_rate": 1e-05, |
| "loss": 0.1353, |
| "step": 15090 |
| }, |
| { |
| "epoch": 4.82, |
| "grad_norm": 8.11100573304352, |
| "learning_rate": 1e-05, |
| "loss": 0.1137, |
| "step": 15100 |
| }, |
| { |
| "epoch": 4.82, |
| "grad_norm": 5.4558387277993186, |
| "learning_rate": 1e-05, |
| "loss": 0.1503, |
| "step": 15110 |
| }, |
| { |
| "epoch": 4.82, |
| "grad_norm": 5.766300151548143, |
| "learning_rate": 1e-05, |
| "loss": 0.1584, |
| "step": 15120 |
| }, |
| { |
| "epoch": 4.83, |
| "grad_norm": 4.4174590689506665, |
| "learning_rate": 1e-05, |
| "loss": 0.1405, |
| "step": 15130 |
| }, |
| { |
| "epoch": 4.83, |
| "grad_norm": 5.813219907744828, |
| "learning_rate": 1e-05, |
| "loss": 0.1677, |
| "step": 15140 |
| }, |
| { |
| "epoch": 4.83, |
| "grad_norm": 6.986621333307631, |
| "learning_rate": 1e-05, |
| "loss": 0.1489, |
| "step": 15150 |
| }, |
| { |
| "epoch": 4.84, |
| "grad_norm": 4.961457665274051, |
| "learning_rate": 1e-05, |
| "loss": 0.1306, |
| "step": 15160 |
| }, |
| { |
| "epoch": 4.84, |
| "grad_norm": 9.831220712393796, |
| "learning_rate": 1e-05, |
| "loss": 0.1628, |
| "step": 15170 |
| }, |
| { |
| "epoch": 4.84, |
| "grad_norm": 5.391448469429226, |
| "learning_rate": 1e-05, |
| "loss": 0.1528, |
| "step": 15180 |
| }, |
| { |
| "epoch": 4.85, |
| "grad_norm": 6.230465076409575, |
| "learning_rate": 1e-05, |
| "loss": 0.1315, |
| "step": 15190 |
| }, |
| { |
| "epoch": 4.85, |
| "grad_norm": 7.502348977043113, |
| "learning_rate": 1e-05, |
| "loss": 0.1407, |
| "step": 15200 |
| }, |
| { |
| "epoch": 4.85, |
| "grad_norm": 8.135560200915167, |
| "learning_rate": 1e-05, |
| "loss": 0.1579, |
| "step": 15210 |
| }, |
| { |
| "epoch": 4.86, |
| "grad_norm": 4.540557772617215, |
| "learning_rate": 1e-05, |
| "loss": 0.1539, |
| "step": 15220 |
| }, |
| { |
| "epoch": 4.86, |
| "grad_norm": 5.36965410393657, |
| "learning_rate": 1e-05, |
| "loss": 0.146, |
| "step": 15230 |
| }, |
| { |
| "epoch": 4.86, |
| "grad_norm": 5.074668858440632, |
| "learning_rate": 1e-05, |
| "loss": 0.1534, |
| "step": 15240 |
| }, |
| { |
| "epoch": 4.87, |
| "grad_norm": 7.2992906513202485, |
| "learning_rate": 1e-05, |
| "loss": 0.152, |
| "step": 15250 |
| }, |
| { |
| "epoch": 4.87, |
| "grad_norm": 4.321826524662616, |
| "learning_rate": 1e-05, |
| "loss": 0.1395, |
| "step": 15260 |
| }, |
| { |
| "epoch": 4.87, |
| "grad_norm": 8.047435792807264, |
| "learning_rate": 1e-05, |
| "loss": 0.1511, |
| "step": 15270 |
| }, |
| { |
| "epoch": 4.88, |
| "grad_norm": 5.111855001264715, |
| "learning_rate": 1e-05, |
| "loss": 0.1329, |
| "step": 15280 |
| }, |
| { |
| "epoch": 4.88, |
| "grad_norm": 8.398322128568022, |
| "learning_rate": 1e-05, |
| "loss": 0.1611, |
| "step": 15290 |
| }, |
| { |
| "epoch": 4.88, |
| "grad_norm": 6.306933980092708, |
| "learning_rate": 1e-05, |
| "loss": 0.139, |
| "step": 15300 |
| }, |
| { |
| "epoch": 4.89, |
| "grad_norm": 7.2652968465880985, |
| "learning_rate": 1e-05, |
| "loss": 0.1582, |
| "step": 15310 |
| }, |
| { |
| "epoch": 4.89, |
| "grad_norm": 7.938229024253704, |
| "learning_rate": 1e-05, |
| "loss": 0.1639, |
| "step": 15320 |
| }, |
| { |
| "epoch": 4.89, |
| "grad_norm": 7.858067549751343, |
| "learning_rate": 1e-05, |
| "loss": 0.166, |
| "step": 15330 |
| }, |
| { |
| "epoch": 4.89, |
| "grad_norm": 5.3753619404846535, |
| "learning_rate": 1e-05, |
| "loss": 0.152, |
| "step": 15340 |
| }, |
| { |
| "epoch": 4.9, |
| "grad_norm": 7.962743312884984, |
| "learning_rate": 1e-05, |
| "loss": 0.1645, |
| "step": 15350 |
| }, |
| { |
| "epoch": 4.9, |
| "grad_norm": 6.049810283618275, |
| "learning_rate": 1e-05, |
| "loss": 0.1584, |
| "step": 15360 |
| }, |
| { |
| "epoch": 4.9, |
| "grad_norm": 5.304064707089478, |
| "learning_rate": 1e-05, |
| "loss": 0.1374, |
| "step": 15370 |
| }, |
| { |
| "epoch": 4.91, |
| "grad_norm": 6.028921043315562, |
| "learning_rate": 1e-05, |
| "loss": 0.1639, |
| "step": 15380 |
| }, |
| { |
| "epoch": 4.91, |
| "grad_norm": 6.742233612150394, |
| "learning_rate": 1e-05, |
| "loss": 0.1413, |
| "step": 15390 |
| }, |
| { |
| "epoch": 4.91, |
| "grad_norm": 8.04787899531561, |
| "learning_rate": 1e-05, |
| "loss": 0.1361, |
| "step": 15400 |
| }, |
| { |
| "epoch": 4.92, |
| "grad_norm": 6.684485781172057, |
| "learning_rate": 1e-05, |
| "loss": 0.1608, |
| "step": 15410 |
| }, |
| { |
| "epoch": 4.92, |
| "grad_norm": 6.125532477452406, |
| "learning_rate": 1e-05, |
| "loss": 0.1567, |
| "step": 15420 |
| }, |
| { |
| "epoch": 4.92, |
| "grad_norm": 5.956607310734599, |
| "learning_rate": 1e-05, |
| "loss": 0.1459, |
| "step": 15430 |
| }, |
| { |
| "epoch": 4.93, |
| "grad_norm": 8.348638695404555, |
| "learning_rate": 1e-05, |
| "loss": 0.1436, |
| "step": 15440 |
| }, |
| { |
| "epoch": 4.93, |
| "grad_norm": 6.140620651437709, |
| "learning_rate": 1e-05, |
| "loss": 0.1582, |
| "step": 15450 |
| }, |
| { |
| "epoch": 4.93, |
| "grad_norm": 6.005530034161093, |
| "learning_rate": 1e-05, |
| "loss": 0.1668, |
| "step": 15460 |
| }, |
| { |
| "epoch": 4.94, |
| "grad_norm": 8.502179932148744, |
| "learning_rate": 1e-05, |
| "loss": 0.1744, |
| "step": 15470 |
| }, |
| { |
| "epoch": 4.94, |
| "grad_norm": 4.729763895154414, |
| "learning_rate": 1e-05, |
| "loss": 0.1317, |
| "step": 15480 |
| }, |
| { |
| "epoch": 4.94, |
| "grad_norm": 4.824889645278048, |
| "learning_rate": 1e-05, |
| "loss": 0.1638, |
| "step": 15490 |
| }, |
| { |
| "epoch": 4.95, |
| "grad_norm": 5.9628004210340615, |
| "learning_rate": 1e-05, |
| "loss": 0.1527, |
| "step": 15500 |
| }, |
| { |
| "epoch": 4.95, |
| "eval_loss": 0.94091796875, |
| "eval_runtime": 55.0992, |
| "eval_samples_per_second": 45.99, |
| "eval_steps_per_second": 2.886, |
| "step": 15500 |
| }, |
| { |
| "epoch": 4.95, |
| "grad_norm": 7.478788006759439, |
| "learning_rate": 1e-05, |
| "loss": 0.1731, |
| "step": 15510 |
| }, |
| { |
| "epoch": 4.95, |
| "grad_norm": 5.328589455505758, |
| "learning_rate": 1e-05, |
| "loss": 0.1626, |
| "step": 15520 |
| }, |
| { |
| "epoch": 4.96, |
| "grad_norm": 4.620148666103659, |
| "learning_rate": 1e-05, |
| "loss": 0.1569, |
| "step": 15530 |
| }, |
| { |
| "epoch": 4.96, |
| "grad_norm": 5.926292040341255, |
| "learning_rate": 1e-05, |
| "loss": 0.1277, |
| "step": 15540 |
| }, |
| { |
| "epoch": 4.96, |
| "grad_norm": 5.289842105809448, |
| "learning_rate": 1e-05, |
| "loss": 0.161, |
| "step": 15550 |
| }, |
| { |
| "epoch": 4.96, |
| "grad_norm": 6.546270447269867, |
| "learning_rate": 1e-05, |
| "loss": 0.1397, |
| "step": 15560 |
| }, |
| { |
| "epoch": 4.97, |
| "grad_norm": 5.985913907977687, |
| "learning_rate": 1e-05, |
| "loss": 0.1521, |
| "step": 15570 |
| }, |
| { |
| "epoch": 4.97, |
| "grad_norm": 4.679673234666279, |
| "learning_rate": 1e-05, |
| "loss": 0.149, |
| "step": 15580 |
| }, |
| { |
| "epoch": 4.97, |
| "grad_norm": 7.1294335742469475, |
| "learning_rate": 1e-05, |
| "loss": 0.1629, |
| "step": 15590 |
| }, |
| { |
| "epoch": 4.98, |
| "grad_norm": 6.101817263707406, |
| "learning_rate": 1e-05, |
| "loss": 0.132, |
| "step": 15600 |
| }, |
| { |
| "epoch": 4.98, |
| "grad_norm": 5.929020337756259, |
| "learning_rate": 1e-05, |
| "loss": 0.1585, |
| "step": 15610 |
| }, |
| { |
| "epoch": 4.98, |
| "grad_norm": 8.005232053749285, |
| "learning_rate": 1e-05, |
| "loss": 0.1549, |
| "step": 15620 |
| }, |
| { |
| "epoch": 4.99, |
| "grad_norm": 5.992767425286436, |
| "learning_rate": 1e-05, |
| "loss": 0.13, |
| "step": 15630 |
| }, |
| { |
| "epoch": 4.99, |
| "grad_norm": 6.9720429308341565, |
| "learning_rate": 1e-05, |
| "loss": 0.1344, |
| "step": 15640 |
| }, |
| { |
| "epoch": 4.99, |
| "grad_norm": 6.287020789023313, |
| "learning_rate": 1e-05, |
| "loss": 0.1872, |
| "step": 15650 |
| }, |
| { |
| "epoch": 5.0, |
| "grad_norm": 5.4267983193951155, |
| "learning_rate": 1e-05, |
| "loss": 0.152, |
| "step": 15660 |
| }, |
| { |
| "epoch": 5.0, |
| "grad_norm": 4.746809038177318, |
| "learning_rate": 1e-05, |
| "loss": 0.1561, |
| "step": 15670 |
| }, |
| { |
| "epoch": 5.0, |
| "grad_norm": 5.980871544547086, |
| "learning_rate": 1e-05, |
| "loss": 0.1045, |
| "step": 15680 |
| }, |
| { |
| "epoch": 5.01, |
| "grad_norm": 6.9494193843120975, |
| "learning_rate": 1e-05, |
| "loss": 0.1005, |
| "step": 15690 |
| }, |
| { |
| "epoch": 5.01, |
| "grad_norm": 5.629657258464171, |
| "learning_rate": 1e-05, |
| "loss": 0.0785, |
| "step": 15700 |
| }, |
| { |
| "epoch": 5.01, |
| "grad_norm": 4.215259153243594, |
| "learning_rate": 1e-05, |
| "loss": 0.0889, |
| "step": 15710 |
| }, |
| { |
| "epoch": 5.02, |
| "grad_norm": 4.437596279765375, |
| "learning_rate": 1e-05, |
| "loss": 0.0952, |
| "step": 15720 |
| }, |
| { |
| "epoch": 5.02, |
| "grad_norm": 4.773175171481911, |
| "learning_rate": 1e-05, |
| "loss": 0.0874, |
| "step": 15730 |
| }, |
| { |
| "epoch": 5.02, |
| "grad_norm": 5.573087239254628, |
| "learning_rate": 1e-05, |
| "loss": 0.095, |
| "step": 15740 |
| }, |
| { |
| "epoch": 5.03, |
| "grad_norm": 4.520637027445483, |
| "learning_rate": 1e-05, |
| "loss": 0.0847, |
| "step": 15750 |
| }, |
| { |
| "epoch": 5.03, |
| "grad_norm": 4.207799871392464, |
| "learning_rate": 1e-05, |
| "loss": 0.088, |
| "step": 15760 |
| }, |
| { |
| "epoch": 5.03, |
| "grad_norm": 6.985681057538089, |
| "learning_rate": 1e-05, |
| "loss": 0.0835, |
| "step": 15770 |
| }, |
| { |
| "epoch": 5.04, |
| "grad_norm": 3.562196952913701, |
| "learning_rate": 1e-05, |
| "loss": 0.0978, |
| "step": 15780 |
| }, |
| { |
| "epoch": 5.04, |
| "grad_norm": 4.422153423696419, |
| "learning_rate": 1e-05, |
| "loss": 0.0863, |
| "step": 15790 |
| }, |
| { |
| "epoch": 5.04, |
| "grad_norm": 5.348166442118508, |
| "learning_rate": 1e-05, |
| "loss": 0.0826, |
| "step": 15800 |
| }, |
| { |
| "epoch": 5.04, |
| "grad_norm": 4.028999350912497, |
| "learning_rate": 1e-05, |
| "loss": 0.0987, |
| "step": 15810 |
| }, |
| { |
| "epoch": 5.05, |
| "grad_norm": 3.0513180933210515, |
| "learning_rate": 1e-05, |
| "loss": 0.0822, |
| "step": 15820 |
| }, |
| { |
| "epoch": 5.05, |
| "grad_norm": 4.367896935123842, |
| "learning_rate": 1e-05, |
| "loss": 0.0841, |
| "step": 15830 |
| }, |
| { |
| "epoch": 5.05, |
| "grad_norm": 4.890577224644343, |
| "learning_rate": 1e-05, |
| "loss": 0.0785, |
| "step": 15840 |
| }, |
| { |
| "epoch": 5.06, |
| "grad_norm": 3.5663073088342965, |
| "learning_rate": 1e-05, |
| "loss": 0.0969, |
| "step": 15850 |
| }, |
| { |
| "epoch": 5.06, |
| "grad_norm": 6.368534474567895, |
| "learning_rate": 1e-05, |
| "loss": 0.0983, |
| "step": 15860 |
| }, |
| { |
| "epoch": 5.06, |
| "grad_norm": 3.7369467845758724, |
| "learning_rate": 1e-05, |
| "loss": 0.0811, |
| "step": 15870 |
| }, |
| { |
| "epoch": 5.07, |
| "grad_norm": 4.761201400128229, |
| "learning_rate": 1e-05, |
| "loss": 0.097, |
| "step": 15880 |
| }, |
| { |
| "epoch": 5.07, |
| "grad_norm": 8.253865116881652, |
| "learning_rate": 1e-05, |
| "loss": 0.1201, |
| "step": 15890 |
| }, |
| { |
| "epoch": 5.07, |
| "grad_norm": 7.111673306388369, |
| "learning_rate": 1e-05, |
| "loss": 0.1104, |
| "step": 15900 |
| }, |
| { |
| "epoch": 5.08, |
| "grad_norm": 5.774187712907893, |
| "learning_rate": 1e-05, |
| "loss": 0.0998, |
| "step": 15910 |
| }, |
| { |
| "epoch": 5.08, |
| "grad_norm": 7.228675261257407, |
| "learning_rate": 1e-05, |
| "loss": 0.0728, |
| "step": 15920 |
| }, |
| { |
| "epoch": 5.08, |
| "grad_norm": 4.0287628781512, |
| "learning_rate": 1e-05, |
| "loss": 0.0994, |
| "step": 15930 |
| }, |
| { |
| "epoch": 5.09, |
| "grad_norm": 7.13706029816948, |
| "learning_rate": 1e-05, |
| "loss": 0.0861, |
| "step": 15940 |
| }, |
| { |
| "epoch": 5.09, |
| "grad_norm": 5.171609462594188, |
| "learning_rate": 1e-05, |
| "loss": 0.0972, |
| "step": 15950 |
| }, |
| { |
| "epoch": 5.09, |
| "grad_norm": 5.935357922091142, |
| "learning_rate": 1e-05, |
| "loss": 0.1081, |
| "step": 15960 |
| }, |
| { |
| "epoch": 5.1, |
| "grad_norm": 6.564673944701606, |
| "learning_rate": 1e-05, |
| "loss": 0.0963, |
| "step": 15970 |
| }, |
| { |
| "epoch": 5.1, |
| "grad_norm": 5.220529036040093, |
| "learning_rate": 1e-05, |
| "loss": 0.0823, |
| "step": 15980 |
| }, |
| { |
| "epoch": 5.1, |
| "grad_norm": 4.473134057788811, |
| "learning_rate": 1e-05, |
| "loss": 0.0793, |
| "step": 15990 |
| }, |
| { |
| "epoch": 5.11, |
| "grad_norm": 5.398258719678952, |
| "learning_rate": 1e-05, |
| "loss": 0.0932, |
| "step": 16000 |
| }, |
| { |
| "epoch": 5.11, |
| "eval_loss": 1.0537109375, |
| "eval_runtime": 54.9158, |
| "eval_samples_per_second": 46.143, |
| "eval_steps_per_second": 2.895, |
| "step": 16000 |
| }, |
| { |
| "epoch": 5.11, |
| "grad_norm": 6.557810242993119, |
| "learning_rate": 1e-05, |
| "loss": 0.1067, |
| "step": 16010 |
| }, |
| { |
| "epoch": 5.11, |
| "grad_norm": 4.207734597296291, |
| "learning_rate": 1e-05, |
| "loss": 0.0819, |
| "step": 16020 |
| }, |
| { |
| "epoch": 5.11, |
| "grad_norm": 4.315216977707757, |
| "learning_rate": 1e-05, |
| "loss": 0.0912, |
| "step": 16030 |
| }, |
| { |
| "epoch": 5.12, |
| "grad_norm": 13.007695194666262, |
| "learning_rate": 1e-05, |
| "loss": 0.0883, |
| "step": 16040 |
| }, |
| { |
| "epoch": 5.12, |
| "grad_norm": 4.677123932637449, |
| "learning_rate": 1e-05, |
| "loss": 0.0901, |
| "step": 16050 |
| }, |
| { |
| "epoch": 5.12, |
| "grad_norm": 4.052914623250701, |
| "learning_rate": 1e-05, |
| "loss": 0.0993, |
| "step": 16060 |
| }, |
| { |
| "epoch": 5.13, |
| "grad_norm": 7.83725837687943, |
| "learning_rate": 1e-05, |
| "loss": 0.1058, |
| "step": 16070 |
| }, |
| { |
| "epoch": 5.13, |
| "grad_norm": 4.778665080711692, |
| "learning_rate": 1e-05, |
| "loss": 0.0999, |
| "step": 16080 |
| }, |
| { |
| "epoch": 5.13, |
| "grad_norm": 7.179563436173066, |
| "learning_rate": 1e-05, |
| "loss": 0.088, |
| "step": 16090 |
| }, |
| { |
| "epoch": 5.14, |
| "grad_norm": 8.802779226990793, |
| "learning_rate": 1e-05, |
| "loss": 0.0769, |
| "step": 16100 |
| }, |
| { |
| "epoch": 5.14, |
| "grad_norm": 4.927661025823997, |
| "learning_rate": 1e-05, |
| "loss": 0.0989, |
| "step": 16110 |
| }, |
| { |
| "epoch": 5.14, |
| "grad_norm": 4.854800312559632, |
| "learning_rate": 1e-05, |
| "loss": 0.0765, |
| "step": 16120 |
| }, |
| { |
| "epoch": 5.15, |
| "grad_norm": 4.431428866414952, |
| "learning_rate": 1e-05, |
| "loss": 0.1048, |
| "step": 16130 |
| }, |
| { |
| "epoch": 5.15, |
| "grad_norm": 5.641388479536823, |
| "learning_rate": 1e-05, |
| "loss": 0.0961, |
| "step": 16140 |
| }, |
| { |
| "epoch": 5.15, |
| "grad_norm": 4.892459415663162, |
| "learning_rate": 1e-05, |
| "loss": 0.1121, |
| "step": 16150 |
| }, |
| { |
| "epoch": 5.16, |
| "grad_norm": 4.3892813693509165, |
| "learning_rate": 1e-05, |
| "loss": 0.0957, |
| "step": 16160 |
| }, |
| { |
| "epoch": 5.16, |
| "grad_norm": 2.4572033844586847, |
| "learning_rate": 1e-05, |
| "loss": 0.0961, |
| "step": 16170 |
| }, |
| { |
| "epoch": 5.16, |
| "grad_norm": 5.05403182480653, |
| "learning_rate": 1e-05, |
| "loss": 0.0845, |
| "step": 16180 |
| }, |
| { |
| "epoch": 5.17, |
| "grad_norm": 6.659401017938821, |
| "learning_rate": 1e-05, |
| "loss": 0.1193, |
| "step": 16190 |
| }, |
| { |
| "epoch": 5.17, |
| "grad_norm": 5.315928889698125, |
| "learning_rate": 1e-05, |
| "loss": 0.0843, |
| "step": 16200 |
| }, |
| { |
| "epoch": 5.17, |
| "grad_norm": 4.776262661623352, |
| "learning_rate": 1e-05, |
| "loss": 0.0804, |
| "step": 16210 |
| }, |
| { |
| "epoch": 5.18, |
| "grad_norm": 6.633442519677489, |
| "learning_rate": 1e-05, |
| "loss": 0.097, |
| "step": 16220 |
| }, |
| { |
| "epoch": 5.18, |
| "grad_norm": 8.553765289903952, |
| "learning_rate": 1e-05, |
| "loss": 0.0939, |
| "step": 16230 |
| }, |
| { |
| "epoch": 5.18, |
| "grad_norm": 5.904198022659088, |
| "learning_rate": 1e-05, |
| "loss": 0.1101, |
| "step": 16240 |
| }, |
| { |
| "epoch": 5.19, |
| "grad_norm": 4.2581457725073095, |
| "learning_rate": 1e-05, |
| "loss": 0.0775, |
| "step": 16250 |
| }, |
| { |
| "epoch": 5.19, |
| "grad_norm": 4.047482006841182, |
| "learning_rate": 1e-05, |
| "loss": 0.1188, |
| "step": 16260 |
| }, |
| { |
| "epoch": 5.19, |
| "grad_norm": 5.142992952612333, |
| "learning_rate": 1e-05, |
| "loss": 0.1076, |
| "step": 16270 |
| }, |
| { |
| "epoch": 5.19, |
| "grad_norm": 5.873440799372244, |
| "learning_rate": 1e-05, |
| "loss": 0.1089, |
| "step": 16280 |
| }, |
| { |
| "epoch": 5.2, |
| "grad_norm": 7.607770370708342, |
| "learning_rate": 1e-05, |
| "loss": 0.0925, |
| "step": 16290 |
| }, |
| { |
| "epoch": 5.2, |
| "grad_norm": 7.013374902338528, |
| "learning_rate": 1e-05, |
| "loss": 0.1001, |
| "step": 16300 |
| }, |
| { |
| "epoch": 5.2, |
| "grad_norm": 6.3484521736578206, |
| "learning_rate": 1e-05, |
| "loss": 0.1178, |
| "step": 16310 |
| }, |
| { |
| "epoch": 5.21, |
| "grad_norm": 6.710624287837871, |
| "learning_rate": 1e-05, |
| "loss": 0.0901, |
| "step": 16320 |
| }, |
| { |
| "epoch": 5.21, |
| "grad_norm": 4.825812813876559, |
| "learning_rate": 1e-05, |
| "loss": 0.1059, |
| "step": 16330 |
| }, |
| { |
| "epoch": 5.21, |
| "grad_norm": 5.7337556343391896, |
| "learning_rate": 1e-05, |
| "loss": 0.1108, |
| "step": 16340 |
| }, |
| { |
| "epoch": 5.22, |
| "grad_norm": 7.300993384090363, |
| "learning_rate": 1e-05, |
| "loss": 0.1107, |
| "step": 16350 |
| }, |
| { |
| "epoch": 5.22, |
| "grad_norm": 2.3867383291187516, |
| "learning_rate": 1e-05, |
| "loss": 0.0855, |
| "step": 16360 |
| }, |
| { |
| "epoch": 5.22, |
| "grad_norm": 4.43055956025509, |
| "learning_rate": 1e-05, |
| "loss": 0.0759, |
| "step": 16370 |
| }, |
| { |
| "epoch": 5.23, |
| "grad_norm": 5.1085452651368515, |
| "learning_rate": 1e-05, |
| "loss": 0.1032, |
| "step": 16380 |
| }, |
| { |
| "epoch": 5.23, |
| "grad_norm": 4.182063202973742, |
| "learning_rate": 1e-05, |
| "loss": 0.1063, |
| "step": 16390 |
| }, |
| { |
| "epoch": 5.23, |
| "grad_norm": 7.029044521727525, |
| "learning_rate": 1e-05, |
| "loss": 0.1218, |
| "step": 16400 |
| }, |
| { |
| "epoch": 5.24, |
| "grad_norm": 6.216137514613552, |
| "learning_rate": 1e-05, |
| "loss": 0.1112, |
| "step": 16410 |
| }, |
| { |
| "epoch": 5.24, |
| "grad_norm": 5.689778971745252, |
| "learning_rate": 1e-05, |
| "loss": 0.108, |
| "step": 16420 |
| }, |
| { |
| "epoch": 5.24, |
| "grad_norm": 4.6202258651810615, |
| "learning_rate": 1e-05, |
| "loss": 0.0916, |
| "step": 16430 |
| }, |
| { |
| "epoch": 5.25, |
| "grad_norm": 4.689736811689441, |
| "learning_rate": 1e-05, |
| "loss": 0.0985, |
| "step": 16440 |
| }, |
| { |
| "epoch": 5.25, |
| "grad_norm": 5.143558117612061, |
| "learning_rate": 1e-05, |
| "loss": 0.089, |
| "step": 16450 |
| }, |
| { |
| "epoch": 5.25, |
| "grad_norm": 6.184544946186926, |
| "learning_rate": 1e-05, |
| "loss": 0.1012, |
| "step": 16460 |
| }, |
| { |
| "epoch": 5.26, |
| "grad_norm": 7.1814193878249535, |
| "learning_rate": 1e-05, |
| "loss": 0.1044, |
| "step": 16470 |
| }, |
| { |
| "epoch": 5.26, |
| "grad_norm": 5.137202738308788, |
| "learning_rate": 1e-05, |
| "loss": 0.0972, |
| "step": 16480 |
| }, |
| { |
| "epoch": 5.26, |
| "grad_norm": 6.101393068430216, |
| "learning_rate": 1e-05, |
| "loss": 0.1206, |
| "step": 16490 |
| }, |
| { |
| "epoch": 5.26, |
| "grad_norm": 6.902435215774139, |
| "learning_rate": 1e-05, |
| "loss": 0.0855, |
| "step": 16500 |
| }, |
| { |
| "epoch": 5.26, |
| "eval_loss": 0.99755859375, |
| "eval_runtime": 54.9581, |
| "eval_samples_per_second": 46.108, |
| "eval_steps_per_second": 2.893, |
| "step": 16500 |
| }, |
| { |
| "epoch": 5.27, |
| "grad_norm": 4.52446749594095, |
| "learning_rate": 1e-05, |
| "loss": 0.1017, |
| "step": 16510 |
| }, |
| { |
| "epoch": 5.27, |
| "grad_norm": 6.165196767669466, |
| "learning_rate": 1e-05, |
| "loss": 0.1039, |
| "step": 16520 |
| }, |
| { |
| "epoch": 5.27, |
| "grad_norm": 5.399004893979523, |
| "learning_rate": 1e-05, |
| "loss": 0.1351, |
| "step": 16530 |
| }, |
| { |
| "epoch": 5.28, |
| "grad_norm": 7.611352432521489, |
| "learning_rate": 1e-05, |
| "loss": 0.1107, |
| "step": 16540 |
| }, |
| { |
| "epoch": 5.28, |
| "grad_norm": 4.9857024813749256, |
| "learning_rate": 1e-05, |
| "loss": 0.0842, |
| "step": 16550 |
| }, |
| { |
| "epoch": 5.28, |
| "grad_norm": 5.140880636334694, |
| "learning_rate": 1e-05, |
| "loss": 0.0872, |
| "step": 16560 |
| }, |
| { |
| "epoch": 5.29, |
| "grad_norm": 5.667143221993245, |
| "learning_rate": 1e-05, |
| "loss": 0.0787, |
| "step": 16570 |
| }, |
| { |
| "epoch": 5.29, |
| "grad_norm": 4.330797162404359, |
| "learning_rate": 1e-05, |
| "loss": 0.1063, |
| "step": 16580 |
| }, |
| { |
| "epoch": 5.29, |
| "grad_norm": 6.096131783630392, |
| "learning_rate": 1e-05, |
| "loss": 0.0956, |
| "step": 16590 |
| }, |
| { |
| "epoch": 5.3, |
| "grad_norm": 5.242645743931427, |
| "learning_rate": 1e-05, |
| "loss": 0.0931, |
| "step": 16600 |
| }, |
| { |
| "epoch": 5.3, |
| "grad_norm": 5.267543364039922, |
| "learning_rate": 1e-05, |
| "loss": 0.1106, |
| "step": 16610 |
| }, |
| { |
| "epoch": 5.3, |
| "grad_norm": 8.086948438464704, |
| "learning_rate": 1e-05, |
| "loss": 0.115, |
| "step": 16620 |
| }, |
| { |
| "epoch": 5.31, |
| "grad_norm": 8.040339332365882, |
| "learning_rate": 1e-05, |
| "loss": 0.0954, |
| "step": 16630 |
| }, |
| { |
| "epoch": 5.31, |
| "grad_norm": 9.202426955169727, |
| "learning_rate": 1e-05, |
| "loss": 0.1096, |
| "step": 16640 |
| }, |
| { |
| "epoch": 5.31, |
| "grad_norm": 5.565617577078237, |
| "learning_rate": 1e-05, |
| "loss": 0.0853, |
| "step": 16650 |
| }, |
| { |
| "epoch": 5.32, |
| "grad_norm": 5.982445786864552, |
| "learning_rate": 1e-05, |
| "loss": 0.0914, |
| "step": 16660 |
| }, |
| { |
| "epoch": 5.32, |
| "grad_norm": 7.306594420212769, |
| "learning_rate": 1e-05, |
| "loss": 0.0811, |
| "step": 16670 |
| }, |
| { |
| "epoch": 5.32, |
| "grad_norm": 5.361582843448211, |
| "learning_rate": 1e-05, |
| "loss": 0.1007, |
| "step": 16680 |
| }, |
| { |
| "epoch": 5.33, |
| "grad_norm": 4.733569589727293, |
| "learning_rate": 1e-05, |
| "loss": 0.1105, |
| "step": 16690 |
| }, |
| { |
| "epoch": 5.33, |
| "grad_norm": 2.6131141050340174, |
| "learning_rate": 1e-05, |
| "loss": 0.0912, |
| "step": 16700 |
| }, |
| { |
| "epoch": 5.33, |
| "grad_norm": 4.542559591352788, |
| "learning_rate": 1e-05, |
| "loss": 0.0955, |
| "step": 16710 |
| }, |
| { |
| "epoch": 5.34, |
| "grad_norm": 5.01049618515815, |
| "learning_rate": 1e-05, |
| "loss": 0.1122, |
| "step": 16720 |
| }, |
| { |
| "epoch": 5.34, |
| "grad_norm": 5.212753780050957, |
| "learning_rate": 1e-05, |
| "loss": 0.1199, |
| "step": 16730 |
| }, |
| { |
| "epoch": 5.34, |
| "grad_norm": 8.360808233588882, |
| "learning_rate": 1e-05, |
| "loss": 0.093, |
| "step": 16740 |
| }, |
| { |
| "epoch": 5.34, |
| "grad_norm": 7.870608846338417, |
| "learning_rate": 1e-05, |
| "loss": 0.1189, |
| "step": 16750 |
| }, |
| { |
| "epoch": 5.35, |
| "grad_norm": 5.843585822915189, |
| "learning_rate": 1e-05, |
| "loss": 0.0986, |
| "step": 16760 |
| }, |
| { |
| "epoch": 5.35, |
| "grad_norm": 3.942939996737345, |
| "learning_rate": 1e-05, |
| "loss": 0.0953, |
| "step": 16770 |
| }, |
| { |
| "epoch": 5.35, |
| "grad_norm": 6.09347580512923, |
| "learning_rate": 1e-05, |
| "loss": 0.086, |
| "step": 16780 |
| }, |
| { |
| "epoch": 5.36, |
| "grad_norm": 4.431151886406657, |
| "learning_rate": 1e-05, |
| "loss": 0.1075, |
| "step": 16790 |
| }, |
| { |
| "epoch": 5.36, |
| "grad_norm": 6.212205980431633, |
| "learning_rate": 1e-05, |
| "loss": 0.096, |
| "step": 16800 |
| }, |
| { |
| "epoch": 5.36, |
| "grad_norm": 8.693654905741152, |
| "learning_rate": 1e-05, |
| "loss": 0.0964, |
| "step": 16810 |
| }, |
| { |
| "epoch": 5.37, |
| "grad_norm": 7.685111566452263, |
| "learning_rate": 1e-05, |
| "loss": 0.0897, |
| "step": 16820 |
| }, |
| { |
| "epoch": 5.37, |
| "grad_norm": 5.0302801678330535, |
| "learning_rate": 1e-05, |
| "loss": 0.1039, |
| "step": 16830 |
| }, |
| { |
| "epoch": 5.37, |
| "grad_norm": 4.482939699800878, |
| "learning_rate": 1e-05, |
| "loss": 0.0905, |
| "step": 16840 |
| }, |
| { |
| "epoch": 5.38, |
| "grad_norm": 5.068272631422925, |
| "learning_rate": 1e-05, |
| "loss": 0.0931, |
| "step": 16850 |
| }, |
| { |
| "epoch": 5.38, |
| "grad_norm": 6.080490946272766, |
| "learning_rate": 1e-05, |
| "loss": 0.0984, |
| "step": 16860 |
| }, |
| { |
| "epoch": 5.38, |
| "grad_norm": 6.050314070691199, |
| "learning_rate": 1e-05, |
| "loss": 0.0983, |
| "step": 16870 |
| }, |
| { |
| "epoch": 5.39, |
| "grad_norm": 7.823633668214154, |
| "learning_rate": 1e-05, |
| "loss": 0.0972, |
| "step": 16880 |
| }, |
| { |
| "epoch": 5.39, |
| "grad_norm": 6.812992183159538, |
| "learning_rate": 1e-05, |
| "loss": 0.128, |
| "step": 16890 |
| }, |
| { |
| "epoch": 5.39, |
| "grad_norm": 3.6811613787879076, |
| "learning_rate": 1e-05, |
| "loss": 0.1063, |
| "step": 16900 |
| }, |
| { |
| "epoch": 5.4, |
| "grad_norm": 5.930015585044207, |
| "learning_rate": 1e-05, |
| "loss": 0.1055, |
| "step": 16910 |
| }, |
| { |
| "epoch": 5.4, |
| "grad_norm": 3.6779234073332687, |
| "learning_rate": 1e-05, |
| "loss": 0.0932, |
| "step": 16920 |
| }, |
| { |
| "epoch": 5.4, |
| "grad_norm": 8.98968295988453, |
| "learning_rate": 1e-05, |
| "loss": 0.0988, |
| "step": 16930 |
| }, |
| { |
| "epoch": 5.41, |
| "grad_norm": 5.94382372949802, |
| "learning_rate": 1e-05, |
| "loss": 0.0856, |
| "step": 16940 |
| }, |
| { |
| "epoch": 5.41, |
| "grad_norm": 8.013339841631607, |
| "learning_rate": 1e-05, |
| "loss": 0.1131, |
| "step": 16950 |
| }, |
| { |
| "epoch": 5.41, |
| "grad_norm": 8.500628279859304, |
| "learning_rate": 1e-05, |
| "loss": 0.1076, |
| "step": 16960 |
| }, |
| { |
| "epoch": 5.41, |
| "grad_norm": 6.086961613528568, |
| "learning_rate": 1e-05, |
| "loss": 0.1034, |
| "step": 16970 |
| }, |
| { |
| "epoch": 5.42, |
| "grad_norm": 11.389921022698985, |
| "learning_rate": 1e-05, |
| "loss": 0.1109, |
| "step": 16980 |
| }, |
| { |
| "epoch": 5.42, |
| "grad_norm": 4.50133092060145, |
| "learning_rate": 1e-05, |
| "loss": 0.1194, |
| "step": 16990 |
| }, |
| { |
| "epoch": 5.42, |
| "grad_norm": 8.025977871934439, |
| "learning_rate": 1e-05, |
| "loss": 0.1098, |
| "step": 17000 |
| }, |
| { |
| "epoch": 5.42, |
| "eval_loss": 1.005859375, |
| "eval_runtime": 54.8987, |
| "eval_samples_per_second": 46.158, |
| "eval_steps_per_second": 2.896, |
| "step": 17000 |
| }, |
| { |
| "epoch": 5.43, |
| "grad_norm": 5.57084663901194, |
| "learning_rate": 1e-05, |
| "loss": 0.1108, |
| "step": 17010 |
| }, |
| { |
| "epoch": 5.43, |
| "grad_norm": 7.361111770086039, |
| "learning_rate": 1e-05, |
| "loss": 0.1226, |
| "step": 17020 |
| }, |
| { |
| "epoch": 5.43, |
| "grad_norm": 3.9135088805076834, |
| "learning_rate": 1e-05, |
| "loss": 0.1164, |
| "step": 17030 |
| }, |
| { |
| "epoch": 5.44, |
| "grad_norm": 7.811966046206299, |
| "learning_rate": 1e-05, |
| "loss": 0.1218, |
| "step": 17040 |
| }, |
| { |
| "epoch": 5.44, |
| "grad_norm": 5.720863493719242, |
| "learning_rate": 1e-05, |
| "loss": 0.1142, |
| "step": 17050 |
| }, |
| { |
| "epoch": 5.44, |
| "grad_norm": 5.498207493768423, |
| "learning_rate": 1e-05, |
| "loss": 0.1232, |
| "step": 17060 |
| }, |
| { |
| "epoch": 5.45, |
| "grad_norm": 4.626592207391009, |
| "learning_rate": 1e-05, |
| "loss": 0.0999, |
| "step": 17070 |
| }, |
| { |
| "epoch": 5.45, |
| "grad_norm": 6.862419964544242, |
| "learning_rate": 1e-05, |
| "loss": 0.1143, |
| "step": 17080 |
| }, |
| { |
| "epoch": 5.45, |
| "grad_norm": 5.415979879806763, |
| "learning_rate": 1e-05, |
| "loss": 0.1041, |
| "step": 17090 |
| }, |
| { |
| "epoch": 5.46, |
| "grad_norm": 6.081369823439729, |
| "learning_rate": 1e-05, |
| "loss": 0.122, |
| "step": 17100 |
| }, |
| { |
| "epoch": 5.46, |
| "grad_norm": 5.53368948897131, |
| "learning_rate": 1e-05, |
| "loss": 0.1156, |
| "step": 17110 |
| }, |
| { |
| "epoch": 5.46, |
| "grad_norm": 5.599904005045324, |
| "learning_rate": 1e-05, |
| "loss": 0.1013, |
| "step": 17120 |
| }, |
| { |
| "epoch": 5.47, |
| "grad_norm": 5.494763916331239, |
| "learning_rate": 1e-05, |
| "loss": 0.132, |
| "step": 17130 |
| }, |
| { |
| "epoch": 5.47, |
| "grad_norm": 5.001556917024766, |
| "learning_rate": 1e-05, |
| "loss": 0.1203, |
| "step": 17140 |
| }, |
| { |
| "epoch": 5.47, |
| "grad_norm": 4.662294292258582, |
| "learning_rate": 1e-05, |
| "loss": 0.095, |
| "step": 17150 |
| }, |
| { |
| "epoch": 5.48, |
| "grad_norm": 5.70130904040203, |
| "learning_rate": 1e-05, |
| "loss": 0.1005, |
| "step": 17160 |
| }, |
| { |
| "epoch": 5.48, |
| "grad_norm": 3.9507417392848327, |
| "learning_rate": 1e-05, |
| "loss": 0.1156, |
| "step": 17170 |
| }, |
| { |
| "epoch": 5.48, |
| "grad_norm": 4.669714386597169, |
| "learning_rate": 1e-05, |
| "loss": 0.1029, |
| "step": 17180 |
| }, |
| { |
| "epoch": 5.49, |
| "grad_norm": 5.0179282155746465, |
| "learning_rate": 1e-05, |
| "loss": 0.1053, |
| "step": 17190 |
| }, |
| { |
| "epoch": 5.49, |
| "grad_norm": 5.487093433795008, |
| "learning_rate": 1e-05, |
| "loss": 0.111, |
| "step": 17200 |
| }, |
| { |
| "epoch": 5.49, |
| "grad_norm": 6.805017123905025, |
| "learning_rate": 1e-05, |
| "loss": 0.1095, |
| "step": 17210 |
| }, |
| { |
| "epoch": 5.49, |
| "grad_norm": 6.519227566511606, |
| "learning_rate": 1e-05, |
| "loss": 0.1081, |
| "step": 17220 |
| }, |
| { |
| "epoch": 5.5, |
| "grad_norm": 3.5754749862869235, |
| "learning_rate": 1e-05, |
| "loss": 0.111, |
| "step": 17230 |
| }, |
| { |
| "epoch": 5.5, |
| "grad_norm": 7.579754274758692, |
| "learning_rate": 1e-05, |
| "loss": 0.1242, |
| "step": 17240 |
| }, |
| { |
| "epoch": 5.5, |
| "grad_norm": 5.715096062013908, |
| "learning_rate": 1e-05, |
| "loss": 0.133, |
| "step": 17250 |
| }, |
| { |
| "epoch": 5.51, |
| "grad_norm": 7.419932414361428, |
| "learning_rate": 1e-05, |
| "loss": 0.1116, |
| "step": 17260 |
| }, |
| { |
| "epoch": 5.51, |
| "grad_norm": 4.493638946901211, |
| "learning_rate": 1e-05, |
| "loss": 0.1256, |
| "step": 17270 |
| }, |
| { |
| "epoch": 5.51, |
| "grad_norm": 5.4807033148751625, |
| "learning_rate": 1e-05, |
| "loss": 0.0964, |
| "step": 17280 |
| }, |
| { |
| "epoch": 5.52, |
| "grad_norm": 8.37203198122349, |
| "learning_rate": 1e-05, |
| "loss": 0.1044, |
| "step": 17290 |
| }, |
| { |
| "epoch": 5.52, |
| "grad_norm": 6.560299458995696, |
| "learning_rate": 1e-05, |
| "loss": 0.1261, |
| "step": 17300 |
| }, |
| { |
| "epoch": 5.52, |
| "grad_norm": 7.68892823884826, |
| "learning_rate": 1e-05, |
| "loss": 0.1255, |
| "step": 17310 |
| }, |
| { |
| "epoch": 5.53, |
| "grad_norm": 5.542694493633567, |
| "learning_rate": 1e-05, |
| "loss": 0.0973, |
| "step": 17320 |
| }, |
| { |
| "epoch": 5.53, |
| "grad_norm": 6.121744964760801, |
| "learning_rate": 1e-05, |
| "loss": 0.097, |
| "step": 17330 |
| }, |
| { |
| "epoch": 5.53, |
| "grad_norm": 4.042210073743501, |
| "learning_rate": 1e-05, |
| "loss": 0.1228, |
| "step": 17340 |
| }, |
| { |
| "epoch": 5.54, |
| "grad_norm": 7.139989244468987, |
| "learning_rate": 1e-05, |
| "loss": 0.1104, |
| "step": 17350 |
| }, |
| { |
| "epoch": 5.54, |
| "grad_norm": 9.114654529111819, |
| "learning_rate": 1e-05, |
| "loss": 0.1337, |
| "step": 17360 |
| }, |
| { |
| "epoch": 5.54, |
| "grad_norm": 5.654153846420713, |
| "learning_rate": 1e-05, |
| "loss": 0.1108, |
| "step": 17370 |
| }, |
| { |
| "epoch": 5.55, |
| "grad_norm": 6.84119273444223, |
| "learning_rate": 1e-05, |
| "loss": 0.1022, |
| "step": 17380 |
| }, |
| { |
| "epoch": 5.55, |
| "grad_norm": 5.544400719058471, |
| "learning_rate": 1e-05, |
| "loss": 0.1268, |
| "step": 17390 |
| }, |
| { |
| "epoch": 5.55, |
| "grad_norm": 6.248197982406107, |
| "learning_rate": 1e-05, |
| "loss": 0.1208, |
| "step": 17400 |
| }, |
| { |
| "epoch": 5.56, |
| "grad_norm": 5.464365569456363, |
| "learning_rate": 1e-05, |
| "loss": 0.0997, |
| "step": 17410 |
| }, |
| { |
| "epoch": 5.56, |
| "grad_norm": 8.34009257532787, |
| "learning_rate": 1e-05, |
| "loss": 0.1174, |
| "step": 17420 |
| }, |
| { |
| "epoch": 5.56, |
| "grad_norm": 6.822402895931466, |
| "learning_rate": 1e-05, |
| "loss": 0.1369, |
| "step": 17430 |
| }, |
| { |
| "epoch": 5.56, |
| "grad_norm": 7.004262988509714, |
| "learning_rate": 1e-05, |
| "loss": 0.1353, |
| "step": 17440 |
| }, |
| { |
| "epoch": 5.57, |
| "grad_norm": 5.342309054798992, |
| "learning_rate": 1e-05, |
| "loss": 0.1159, |
| "step": 17450 |
| }, |
| { |
| "epoch": 5.57, |
| "grad_norm": 4.8060660020706925, |
| "learning_rate": 1e-05, |
| "loss": 0.1221, |
| "step": 17460 |
| }, |
| { |
| "epoch": 5.57, |
| "grad_norm": 6.217668142096693, |
| "learning_rate": 1e-05, |
| "loss": 0.1064, |
| "step": 17470 |
| }, |
| { |
| "epoch": 5.58, |
| "grad_norm": 4.288810792079071, |
| "learning_rate": 1e-05, |
| "loss": 0.0856, |
| "step": 17480 |
| }, |
| { |
| "epoch": 5.58, |
| "grad_norm": 5.883290386673582, |
| "learning_rate": 1e-05, |
| "loss": 0.1273, |
| "step": 17490 |
| }, |
| { |
| "epoch": 5.58, |
| "grad_norm": 6.110551881730193, |
| "learning_rate": 1e-05, |
| "loss": 0.123, |
| "step": 17500 |
| }, |
| { |
| "epoch": 5.58, |
| "eval_loss": 1.0126953125, |
| "eval_runtime": 54.9485, |
| "eval_samples_per_second": 46.116, |
| "eval_steps_per_second": 2.894, |
| "step": 17500 |
| }, |
| { |
| "epoch": 5.59, |
| "grad_norm": 6.601119853872926, |
| "learning_rate": 1e-05, |
| "loss": 0.1137, |
| "step": 17510 |
| }, |
| { |
| "epoch": 5.59, |
| "grad_norm": 4.812262591166228, |
| "learning_rate": 1e-05, |
| "loss": 0.1147, |
| "step": 17520 |
| }, |
| { |
| "epoch": 5.59, |
| "grad_norm": 6.156179050094805, |
| "learning_rate": 1e-05, |
| "loss": 0.1078, |
| "step": 17530 |
| }, |
| { |
| "epoch": 5.6, |
| "grad_norm": 4.782291473452925, |
| "learning_rate": 1e-05, |
| "loss": 0.1043, |
| "step": 17540 |
| }, |
| { |
| "epoch": 5.6, |
| "grad_norm": 6.0934704838738485, |
| "learning_rate": 1e-05, |
| "loss": 0.1077, |
| "step": 17550 |
| }, |
| { |
| "epoch": 5.6, |
| "grad_norm": 6.459771010648114, |
| "learning_rate": 1e-05, |
| "loss": 0.1072, |
| "step": 17560 |
| }, |
| { |
| "epoch": 5.61, |
| "grad_norm": 4.735185711128136, |
| "learning_rate": 1e-05, |
| "loss": 0.0962, |
| "step": 17570 |
| }, |
| { |
| "epoch": 5.61, |
| "grad_norm": 4.242093529547378, |
| "learning_rate": 1e-05, |
| "loss": 0.1122, |
| "step": 17580 |
| }, |
| { |
| "epoch": 5.61, |
| "grad_norm": 5.335518905730558, |
| "learning_rate": 1e-05, |
| "loss": 0.0886, |
| "step": 17590 |
| }, |
| { |
| "epoch": 5.62, |
| "grad_norm": 5.911366986032477, |
| "learning_rate": 1e-05, |
| "loss": 0.1273, |
| "step": 17600 |
| }, |
| { |
| "epoch": 5.62, |
| "grad_norm": 5.045974320390564, |
| "learning_rate": 1e-05, |
| "loss": 0.1158, |
| "step": 17610 |
| }, |
| { |
| "epoch": 5.62, |
| "grad_norm": 6.901363801935104, |
| "learning_rate": 1e-05, |
| "loss": 0.111, |
| "step": 17620 |
| }, |
| { |
| "epoch": 5.63, |
| "grad_norm": 7.091250344714676, |
| "learning_rate": 1e-05, |
| "loss": 0.123, |
| "step": 17630 |
| }, |
| { |
| "epoch": 5.63, |
| "grad_norm": 3.863740868608538, |
| "learning_rate": 1e-05, |
| "loss": 0.0891, |
| "step": 17640 |
| }, |
| { |
| "epoch": 5.63, |
| "grad_norm": 5.199000284490268, |
| "learning_rate": 1e-05, |
| "loss": 0.1292, |
| "step": 17650 |
| }, |
| { |
| "epoch": 5.63, |
| "grad_norm": 5.756362049118119, |
| "learning_rate": 1e-05, |
| "loss": 0.1062, |
| "step": 17660 |
| }, |
| { |
| "epoch": 5.64, |
| "grad_norm": 5.325110872215864, |
| "learning_rate": 1e-05, |
| "loss": 0.0924, |
| "step": 17670 |
| }, |
| { |
| "epoch": 5.64, |
| "grad_norm": 5.357869195290491, |
| "learning_rate": 1e-05, |
| "loss": 0.1232, |
| "step": 17680 |
| }, |
| { |
| "epoch": 5.64, |
| "grad_norm": 5.995387529664732, |
| "learning_rate": 1e-05, |
| "loss": 0.1004, |
| "step": 17690 |
| }, |
| { |
| "epoch": 5.65, |
| "grad_norm": 8.158047649718537, |
| "learning_rate": 1e-05, |
| "loss": 0.1035, |
| "step": 17700 |
| }, |
| { |
| "epoch": 5.65, |
| "grad_norm": 7.074385642616536, |
| "learning_rate": 1e-05, |
| "loss": 0.1317, |
| "step": 17710 |
| }, |
| { |
| "epoch": 5.65, |
| "grad_norm": 6.695530428283068, |
| "learning_rate": 1e-05, |
| "loss": 0.1247, |
| "step": 17720 |
| }, |
| { |
| "epoch": 5.66, |
| "grad_norm": 5.877160182460806, |
| "learning_rate": 1e-05, |
| "loss": 0.1234, |
| "step": 17730 |
| }, |
| { |
| "epoch": 5.66, |
| "grad_norm": 8.1385296195121, |
| "learning_rate": 1e-05, |
| "loss": 0.1018, |
| "step": 17740 |
| }, |
| { |
| "epoch": 5.66, |
| "grad_norm": 4.703713820594817, |
| "learning_rate": 1e-05, |
| "loss": 0.0984, |
| "step": 17750 |
| }, |
| { |
| "epoch": 5.67, |
| "grad_norm": 3.814888753006428, |
| "learning_rate": 1e-05, |
| "loss": 0.1153, |
| "step": 17760 |
| }, |
| { |
| "epoch": 5.67, |
| "grad_norm": 5.210936584930111, |
| "learning_rate": 1e-05, |
| "loss": 0.107, |
| "step": 17770 |
| }, |
| { |
| "epoch": 5.67, |
| "grad_norm": 7.768862749725487, |
| "learning_rate": 1e-05, |
| "loss": 0.1349, |
| "step": 17780 |
| }, |
| { |
| "epoch": 5.68, |
| "grad_norm": 3.08993094185013, |
| "learning_rate": 1e-05, |
| "loss": 0.0947, |
| "step": 17790 |
| }, |
| { |
| "epoch": 5.68, |
| "grad_norm": 7.9249459057234315, |
| "learning_rate": 1e-05, |
| "loss": 0.1173, |
| "step": 17800 |
| }, |
| { |
| "epoch": 5.68, |
| "grad_norm": 4.615350709692809, |
| "learning_rate": 1e-05, |
| "loss": 0.1418, |
| "step": 17810 |
| }, |
| { |
| "epoch": 5.69, |
| "grad_norm": 6.983570163446717, |
| "learning_rate": 1e-05, |
| "loss": 0.1093, |
| "step": 17820 |
| }, |
| { |
| "epoch": 5.69, |
| "grad_norm": 4.514130975475249, |
| "learning_rate": 1e-05, |
| "loss": 0.0985, |
| "step": 17830 |
| }, |
| { |
| "epoch": 5.69, |
| "grad_norm": 5.357609672189281, |
| "learning_rate": 1e-05, |
| "loss": 0.1331, |
| "step": 17840 |
| }, |
| { |
| "epoch": 5.7, |
| "grad_norm": 4.821224676594335, |
| "learning_rate": 1e-05, |
| "loss": 0.0952, |
| "step": 17850 |
| }, |
| { |
| "epoch": 5.7, |
| "grad_norm": 6.426645850362872, |
| "learning_rate": 1e-05, |
| "loss": 0.1315, |
| "step": 17860 |
| }, |
| { |
| "epoch": 5.7, |
| "grad_norm": 3.972074062782502, |
| "learning_rate": 1e-05, |
| "loss": 0.1019, |
| "step": 17870 |
| }, |
| { |
| "epoch": 5.71, |
| "grad_norm": 4.95562603980679, |
| "learning_rate": 1e-05, |
| "loss": 0.1114, |
| "step": 17880 |
| }, |
| { |
| "epoch": 5.71, |
| "grad_norm": 5.486396786405079, |
| "learning_rate": 1e-05, |
| "loss": 0.1201, |
| "step": 17890 |
| }, |
| { |
| "epoch": 5.71, |
| "grad_norm": 5.243942217208061, |
| "learning_rate": 1e-05, |
| "loss": 0.1125, |
| "step": 17900 |
| }, |
| { |
| "epoch": 5.71, |
| "grad_norm": 5.346946251203282, |
| "learning_rate": 1e-05, |
| "loss": 0.1175, |
| "step": 17910 |
| }, |
| { |
| "epoch": 5.72, |
| "grad_norm": 7.116616118323788, |
| "learning_rate": 1e-05, |
| "loss": 0.1148, |
| "step": 17920 |
| }, |
| { |
| "epoch": 5.72, |
| "grad_norm": 6.165306903659265, |
| "learning_rate": 1e-05, |
| "loss": 0.1025, |
| "step": 17930 |
| }, |
| { |
| "epoch": 5.72, |
| "grad_norm": 4.6783724610464, |
| "learning_rate": 1e-05, |
| "loss": 0.0913, |
| "step": 17940 |
| }, |
| { |
| "epoch": 5.73, |
| "grad_norm": 6.586160784971625, |
| "learning_rate": 1e-05, |
| "loss": 0.1132, |
| "step": 17950 |
| }, |
| { |
| "epoch": 5.73, |
| "grad_norm": 5.854945591413977, |
| "learning_rate": 1e-05, |
| "loss": 0.1317, |
| "step": 17960 |
| }, |
| { |
| "epoch": 5.73, |
| "grad_norm": 5.502309834349897, |
| "learning_rate": 1e-05, |
| "loss": 0.1203, |
| "step": 17970 |
| }, |
| { |
| "epoch": 5.74, |
| "grad_norm": 6.896389511436911, |
| "learning_rate": 1e-05, |
| "loss": 0.1219, |
| "step": 17980 |
| }, |
| { |
| "epoch": 5.74, |
| "grad_norm": 3.6912494282991695, |
| "learning_rate": 1e-05, |
| "loss": 0.117, |
| "step": 17990 |
| }, |
| { |
| "epoch": 5.74, |
| "grad_norm": 5.228562318852013, |
| "learning_rate": 1e-05, |
| "loss": 0.1155, |
| "step": 18000 |
| }, |
| { |
| "epoch": 5.74, |
| "eval_loss": 1.0224609375, |
| "eval_runtime": 54.6994, |
| "eval_samples_per_second": 46.326, |
| "eval_steps_per_second": 2.907, |
| "step": 18000 |
| }, |
| { |
| "epoch": 5.75, |
| "grad_norm": 7.418249525689374, |
| "learning_rate": 1e-05, |
| "loss": 0.1368, |
| "step": 18010 |
| }, |
| { |
| "epoch": 5.75, |
| "grad_norm": 5.551146503970592, |
| "learning_rate": 1e-05, |
| "loss": 0.1236, |
| "step": 18020 |
| }, |
| { |
| "epoch": 5.75, |
| "grad_norm": 7.779823264343095, |
| "learning_rate": 1e-05, |
| "loss": 0.109, |
| "step": 18030 |
| }, |
| { |
| "epoch": 5.76, |
| "grad_norm": 4.731558890696893, |
| "learning_rate": 1e-05, |
| "loss": 0.1119, |
| "step": 18040 |
| }, |
| { |
| "epoch": 5.76, |
| "grad_norm": 4.157729036281194, |
| "learning_rate": 1e-05, |
| "loss": 0.118, |
| "step": 18050 |
| }, |
| { |
| "epoch": 5.76, |
| "grad_norm": 6.340132861264528, |
| "learning_rate": 1e-05, |
| "loss": 0.1335, |
| "step": 18060 |
| }, |
| { |
| "epoch": 5.77, |
| "grad_norm": 7.227799461280745, |
| "learning_rate": 1e-05, |
| "loss": 0.1257, |
| "step": 18070 |
| }, |
| { |
| "epoch": 5.77, |
| "grad_norm": 5.099051869839478, |
| "learning_rate": 1e-05, |
| "loss": 0.106, |
| "step": 18080 |
| }, |
| { |
| "epoch": 5.77, |
| "grad_norm": 6.262240516022806, |
| "learning_rate": 1e-05, |
| "loss": 0.1049, |
| "step": 18090 |
| }, |
| { |
| "epoch": 5.78, |
| "grad_norm": 6.6508517874247515, |
| "learning_rate": 1e-05, |
| "loss": 0.1051, |
| "step": 18100 |
| }, |
| { |
| "epoch": 5.78, |
| "grad_norm": 9.859684136769062, |
| "learning_rate": 1e-05, |
| "loss": 0.1499, |
| "step": 18110 |
| }, |
| { |
| "epoch": 5.78, |
| "grad_norm": 7.586306090588335, |
| "learning_rate": 1e-05, |
| "loss": 0.1299, |
| "step": 18120 |
| }, |
| { |
| "epoch": 5.78, |
| "grad_norm": 4.608702115875042, |
| "learning_rate": 1e-05, |
| "loss": 0.1047, |
| "step": 18130 |
| }, |
| { |
| "epoch": 5.79, |
| "grad_norm": 5.697723255210694, |
| "learning_rate": 1e-05, |
| "loss": 0.1141, |
| "step": 18140 |
| }, |
| { |
| "epoch": 5.79, |
| "grad_norm": 7.552848773619658, |
| "learning_rate": 1e-05, |
| "loss": 0.1495, |
| "step": 18150 |
| }, |
| { |
| "epoch": 5.79, |
| "grad_norm": 6.707804720252302, |
| "learning_rate": 1e-05, |
| "loss": 0.1446, |
| "step": 18160 |
| }, |
| { |
| "epoch": 5.8, |
| "grad_norm": 7.3562803463889255, |
| "learning_rate": 1e-05, |
| "loss": 0.1103, |
| "step": 18170 |
| }, |
| { |
| "epoch": 5.8, |
| "grad_norm": 6.4042271839758005, |
| "learning_rate": 1e-05, |
| "loss": 0.1092, |
| "step": 18180 |
| }, |
| { |
| "epoch": 5.8, |
| "grad_norm": 8.64542303988499, |
| "learning_rate": 1e-05, |
| "loss": 0.1334, |
| "step": 18190 |
| }, |
| { |
| "epoch": 5.81, |
| "grad_norm": 5.757060153208929, |
| "learning_rate": 1e-05, |
| "loss": 0.1203, |
| "step": 18200 |
| }, |
| { |
| "epoch": 5.81, |
| "grad_norm": 5.016374668651075, |
| "learning_rate": 1e-05, |
| "loss": 0.1158, |
| "step": 18210 |
| }, |
| { |
| "epoch": 5.81, |
| "grad_norm": 4.3072639412162514, |
| "learning_rate": 1e-05, |
| "loss": 0.1, |
| "step": 18220 |
| }, |
| { |
| "epoch": 5.82, |
| "grad_norm": 5.6202053404522205, |
| "learning_rate": 1e-05, |
| "loss": 0.1105, |
| "step": 18230 |
| }, |
| { |
| "epoch": 5.82, |
| "grad_norm": 6.532983869172132, |
| "learning_rate": 1e-05, |
| "loss": 0.1078, |
| "step": 18240 |
| }, |
| { |
| "epoch": 5.82, |
| "grad_norm": 5.5816755348302936, |
| "learning_rate": 1e-05, |
| "loss": 0.1151, |
| "step": 18250 |
| }, |
| { |
| "epoch": 5.83, |
| "grad_norm": 10.36314100392758, |
| "learning_rate": 1e-05, |
| "loss": 0.1202, |
| "step": 18260 |
| }, |
| { |
| "epoch": 5.83, |
| "grad_norm": 9.54014857402405, |
| "learning_rate": 1e-05, |
| "loss": 0.1325, |
| "step": 18270 |
| }, |
| { |
| "epoch": 5.83, |
| "grad_norm": 5.818845177352466, |
| "learning_rate": 1e-05, |
| "loss": 0.1096, |
| "step": 18280 |
| }, |
| { |
| "epoch": 5.84, |
| "grad_norm": 4.0734282007319775, |
| "learning_rate": 1e-05, |
| "loss": 0.1091, |
| "step": 18290 |
| }, |
| { |
| "epoch": 5.84, |
| "grad_norm": 8.052337629309065, |
| "learning_rate": 1e-05, |
| "loss": 0.1176, |
| "step": 18300 |
| }, |
| { |
| "epoch": 5.84, |
| "grad_norm": 4.995041868475405, |
| "learning_rate": 1e-05, |
| "loss": 0.0994, |
| "step": 18310 |
| }, |
| { |
| "epoch": 5.85, |
| "grad_norm": 8.043949995540855, |
| "learning_rate": 1e-05, |
| "loss": 0.1038, |
| "step": 18320 |
| }, |
| { |
| "epoch": 5.85, |
| "grad_norm": 7.352980246944969, |
| "learning_rate": 1e-05, |
| "loss": 0.1136, |
| "step": 18330 |
| }, |
| { |
| "epoch": 5.85, |
| "grad_norm": 7.017383516470817, |
| "learning_rate": 1e-05, |
| "loss": 0.1122, |
| "step": 18340 |
| }, |
| { |
| "epoch": 5.86, |
| "grad_norm": 3.176666549276995, |
| "learning_rate": 1e-05, |
| "loss": 0.1111, |
| "step": 18350 |
| }, |
| { |
| "epoch": 5.86, |
| "grad_norm": 4.4638796856024445, |
| "learning_rate": 1e-05, |
| "loss": 0.0988, |
| "step": 18360 |
| }, |
| { |
| "epoch": 5.86, |
| "grad_norm": 4.672084841911535, |
| "learning_rate": 1e-05, |
| "loss": 0.0941, |
| "step": 18370 |
| }, |
| { |
| "epoch": 5.86, |
| "grad_norm": 5.560676018488469, |
| "learning_rate": 1e-05, |
| "loss": 0.1181, |
| "step": 18380 |
| }, |
| { |
| "epoch": 5.87, |
| "grad_norm": 4.8747104656608515, |
| "learning_rate": 1e-05, |
| "loss": 0.1137, |
| "step": 18390 |
| }, |
| { |
| "epoch": 5.87, |
| "grad_norm": 5.643080243051361, |
| "learning_rate": 1e-05, |
| "loss": 0.151, |
| "step": 18400 |
| }, |
| { |
| "epoch": 5.87, |
| "grad_norm": 4.757604635173464, |
| "learning_rate": 1e-05, |
| "loss": 0.1198, |
| "step": 18410 |
| }, |
| { |
| "epoch": 5.88, |
| "grad_norm": 7.478485019687962, |
| "learning_rate": 1e-05, |
| "loss": 0.1125, |
| "step": 18420 |
| }, |
| { |
| "epoch": 5.88, |
| "grad_norm": 6.371145841103724, |
| "learning_rate": 1e-05, |
| "loss": 0.1287, |
| "step": 18430 |
| }, |
| { |
| "epoch": 5.88, |
| "grad_norm": 8.34884156740192, |
| "learning_rate": 1e-05, |
| "loss": 0.1237, |
| "step": 18440 |
| }, |
| { |
| "epoch": 5.89, |
| "grad_norm": 7.008135835977198, |
| "learning_rate": 1e-05, |
| "loss": 0.1454, |
| "step": 18450 |
| }, |
| { |
| "epoch": 5.89, |
| "grad_norm": 5.335062024972412, |
| "learning_rate": 1e-05, |
| "loss": 0.1093, |
| "step": 18460 |
| }, |
| { |
| "epoch": 5.89, |
| "grad_norm": 5.044848906387708, |
| "learning_rate": 1e-05, |
| "loss": 0.1222, |
| "step": 18470 |
| }, |
| { |
| "epoch": 5.9, |
| "grad_norm": 4.278243073978947, |
| "learning_rate": 1e-05, |
| "loss": 0.1095, |
| "step": 18480 |
| }, |
| { |
| "epoch": 5.9, |
| "grad_norm": 4.139339837776374, |
| "learning_rate": 1e-05, |
| "loss": 0.0965, |
| "step": 18490 |
| }, |
| { |
| "epoch": 5.9, |
| "grad_norm": 9.994280896330942, |
| "learning_rate": 1e-05, |
| "loss": 0.1179, |
| "step": 18500 |
| }, |
| { |
| "epoch": 5.9, |
| "eval_loss": 1.037109375, |
| "eval_runtime": 54.9596, |
| "eval_samples_per_second": 46.107, |
| "eval_steps_per_second": 2.893, |
| "step": 18500 |
| }, |
| { |
| "epoch": 5.91, |
| "grad_norm": 3.711907896354389, |
| "learning_rate": 1e-05, |
| "loss": 0.1114, |
| "step": 18510 |
| }, |
| { |
| "epoch": 5.91, |
| "grad_norm": 7.346509569199634, |
| "learning_rate": 1e-05, |
| "loss": 0.1347, |
| "step": 18520 |
| }, |
| { |
| "epoch": 5.91, |
| "grad_norm": 6.487517475552992, |
| "learning_rate": 1e-05, |
| "loss": 0.1347, |
| "step": 18530 |
| }, |
| { |
| "epoch": 5.92, |
| "grad_norm": 5.272025997439294, |
| "learning_rate": 1e-05, |
| "loss": 0.1217, |
| "step": 18540 |
| }, |
| { |
| "epoch": 5.92, |
| "grad_norm": 4.623665281964995, |
| "learning_rate": 1e-05, |
| "loss": 0.1285, |
| "step": 18550 |
| }, |
| { |
| "epoch": 5.92, |
| "grad_norm": 9.36746034710007, |
| "learning_rate": 1e-05, |
| "loss": 0.1046, |
| "step": 18560 |
| }, |
| { |
| "epoch": 5.93, |
| "grad_norm": 8.1584661403467, |
| "learning_rate": 1e-05, |
| "loss": 0.1107, |
| "step": 18570 |
| }, |
| { |
| "epoch": 5.93, |
| "grad_norm": 6.1841992137843915, |
| "learning_rate": 1e-05, |
| "loss": 0.1252, |
| "step": 18580 |
| }, |
| { |
| "epoch": 5.93, |
| "grad_norm": 4.692883871043434, |
| "learning_rate": 1e-05, |
| "loss": 0.1337, |
| "step": 18590 |
| }, |
| { |
| "epoch": 5.93, |
| "grad_norm": 5.656348296096425, |
| "learning_rate": 1e-05, |
| "loss": 0.1202, |
| "step": 18600 |
| }, |
| { |
| "epoch": 5.94, |
| "grad_norm": 6.350219734775454, |
| "learning_rate": 1e-05, |
| "loss": 0.129, |
| "step": 18610 |
| }, |
| { |
| "epoch": 5.94, |
| "grad_norm": 9.301340264988571, |
| "learning_rate": 1e-05, |
| "loss": 0.1318, |
| "step": 18620 |
| }, |
| { |
| "epoch": 5.94, |
| "grad_norm": 5.131117356475852, |
| "learning_rate": 1e-05, |
| "loss": 0.1213, |
| "step": 18630 |
| }, |
| { |
| "epoch": 5.95, |
| "grad_norm": 9.206217895368962, |
| "learning_rate": 1e-05, |
| "loss": 0.126, |
| "step": 18640 |
| }, |
| { |
| "epoch": 5.95, |
| "grad_norm": 7.360022643925207, |
| "learning_rate": 1e-05, |
| "loss": 0.129, |
| "step": 18650 |
| }, |
| { |
| "epoch": 5.95, |
| "grad_norm": 4.964931916445232, |
| "learning_rate": 1e-05, |
| "loss": 0.1081, |
| "step": 18660 |
| }, |
| { |
| "epoch": 5.96, |
| "grad_norm": 5.172153983990272, |
| "learning_rate": 1e-05, |
| "loss": 0.1284, |
| "step": 18670 |
| }, |
| { |
| "epoch": 5.96, |
| "grad_norm": 4.759949753760558, |
| "learning_rate": 1e-05, |
| "loss": 0.1286, |
| "step": 18680 |
| }, |
| { |
| "epoch": 5.96, |
| "grad_norm": 9.129772035859265, |
| "learning_rate": 1e-05, |
| "loss": 0.1081, |
| "step": 18690 |
| }, |
| { |
| "epoch": 5.97, |
| "grad_norm": 6.5790690199998565, |
| "learning_rate": 1e-05, |
| "loss": 0.1163, |
| "step": 18700 |
| }, |
| { |
| "epoch": 5.97, |
| "grad_norm": 7.2764381586468065, |
| "learning_rate": 1e-05, |
| "loss": 0.1148, |
| "step": 18710 |
| }, |
| { |
| "epoch": 5.97, |
| "grad_norm": 7.377896273604845, |
| "learning_rate": 1e-05, |
| "loss": 0.0974, |
| "step": 18720 |
| }, |
| { |
| "epoch": 5.98, |
| "grad_norm": 4.926584660764264, |
| "learning_rate": 1e-05, |
| "loss": 0.1106, |
| "step": 18730 |
| }, |
| { |
| "epoch": 5.98, |
| "grad_norm": 5.971840263197802, |
| "learning_rate": 1e-05, |
| "loss": 0.1066, |
| "step": 18740 |
| }, |
| { |
| "epoch": 5.98, |
| "grad_norm": 7.406824017263072, |
| "learning_rate": 1e-05, |
| "loss": 0.1078, |
| "step": 18750 |
| }, |
| { |
| "epoch": 5.99, |
| "grad_norm": 7.729484969694198, |
| "learning_rate": 1e-05, |
| "loss": 0.1137, |
| "step": 18760 |
| }, |
| { |
| "epoch": 5.99, |
| "grad_norm": 6.29362521308299, |
| "learning_rate": 1e-05, |
| "loss": 0.1082, |
| "step": 18770 |
| }, |
| { |
| "epoch": 5.99, |
| "grad_norm": 8.409513499445927, |
| "learning_rate": 1e-05, |
| "loss": 0.1188, |
| "step": 18780 |
| }, |
| { |
| "epoch": 6.0, |
| "grad_norm": 6.271085343308337, |
| "learning_rate": 1e-05, |
| "loss": 0.1138, |
| "step": 18790 |
| }, |
| { |
| "epoch": 6.0, |
| "grad_norm": 4.781575709200617, |
| "learning_rate": 1e-05, |
| "loss": 0.1287, |
| "step": 18800 |
| }, |
| { |
| "epoch": 6.0, |
| "grad_norm": 5.751312106153065, |
| "learning_rate": 1e-05, |
| "loss": 0.0811, |
| "step": 18810 |
| }, |
| { |
| "epoch": 6.01, |
| "grad_norm": 5.924849354563167, |
| "learning_rate": 1e-05, |
| "loss": 0.084, |
| "step": 18820 |
| }, |
| { |
| "epoch": 6.01, |
| "grad_norm": 2.7484598181654962, |
| "learning_rate": 1e-05, |
| "loss": 0.0626, |
| "step": 18830 |
| }, |
| { |
| "epoch": 6.01, |
| "grad_norm": 7.667390899727789, |
| "learning_rate": 1e-05, |
| "loss": 0.0611, |
| "step": 18840 |
| }, |
| { |
| "epoch": 6.01, |
| "grad_norm": 5.154565339434643, |
| "learning_rate": 1e-05, |
| "loss": 0.0582, |
| "step": 18850 |
| }, |
| { |
| "epoch": 6.02, |
| "grad_norm": 5.682151207614191, |
| "learning_rate": 1e-05, |
| "loss": 0.0969, |
| "step": 18860 |
| }, |
| { |
| "epoch": 6.02, |
| "grad_norm": 5.434024269934332, |
| "learning_rate": 1e-05, |
| "loss": 0.0823, |
| "step": 18870 |
| }, |
| { |
| "epoch": 6.02, |
| "grad_norm": 19.953021760940825, |
| "learning_rate": 1e-05, |
| "loss": 0.0825, |
| "step": 18880 |
| }, |
| { |
| "epoch": 6.03, |
| "grad_norm": 3.529126364257514, |
| "learning_rate": 1e-05, |
| "loss": 0.0707, |
| "step": 18890 |
| }, |
| { |
| "epoch": 6.03, |
| "grad_norm": 10.229295614534328, |
| "learning_rate": 1e-05, |
| "loss": 0.0954, |
| "step": 18900 |
| }, |
| { |
| "epoch": 6.03, |
| "grad_norm": 3.076544496281121, |
| "learning_rate": 1e-05, |
| "loss": 0.0723, |
| "step": 18910 |
| }, |
| { |
| "epoch": 6.04, |
| "grad_norm": 3.6377725191051784, |
| "learning_rate": 1e-05, |
| "loss": 0.0732, |
| "step": 18920 |
| }, |
| { |
| "epoch": 6.04, |
| "grad_norm": 2.5917002840304764, |
| "learning_rate": 1e-05, |
| "loss": 0.078, |
| "step": 18930 |
| }, |
| { |
| "epoch": 6.04, |
| "grad_norm": 4.113084402563641, |
| "learning_rate": 1e-05, |
| "loss": 0.0746, |
| "step": 18940 |
| }, |
| { |
| "epoch": 6.05, |
| "grad_norm": 6.97080764841781, |
| "learning_rate": 1e-05, |
| "loss": 0.0847, |
| "step": 18950 |
| }, |
| { |
| "epoch": 6.05, |
| "grad_norm": 4.844896734532753, |
| "learning_rate": 1e-05, |
| "loss": 0.0695, |
| "step": 18960 |
| }, |
| { |
| "epoch": 6.05, |
| "grad_norm": 5.494899812571227, |
| "learning_rate": 1e-05, |
| "loss": 0.0705, |
| "step": 18970 |
| }, |
| { |
| "epoch": 6.06, |
| "grad_norm": 3.6033237478556313, |
| "learning_rate": 1e-05, |
| "loss": 0.0697, |
| "step": 18980 |
| }, |
| { |
| "epoch": 6.06, |
| "grad_norm": 5.162673001473326, |
| "learning_rate": 1e-05, |
| "loss": 0.0909, |
| "step": 18990 |
| }, |
| { |
| "epoch": 6.06, |
| "grad_norm": 7.546639954123157, |
| "learning_rate": 1e-05, |
| "loss": 0.0891, |
| "step": 19000 |
| }, |
| { |
| "epoch": 6.06, |
| "eval_loss": 1.06640625, |
| "eval_runtime": 54.7024, |
| "eval_samples_per_second": 46.323, |
| "eval_steps_per_second": 2.907, |
| "step": 19000 |
| }, |
| { |
| "epoch": 6.07, |
| "grad_norm": 3.4223201932411174, |
| "learning_rate": 1e-05, |
| "loss": 0.088, |
| "step": 19010 |
| }, |
| { |
| "epoch": 6.07, |
| "grad_norm": 4.1743650398986585, |
| "learning_rate": 1e-05, |
| "loss": 0.0752, |
| "step": 19020 |
| }, |
| { |
| "epoch": 6.07, |
| "grad_norm": 5.727312192464383, |
| "learning_rate": 1e-05, |
| "loss": 0.0754, |
| "step": 19030 |
| }, |
| { |
| "epoch": 6.08, |
| "grad_norm": 6.720112791023099, |
| "learning_rate": 1e-05, |
| "loss": 0.0686, |
| "step": 19040 |
| }, |
| { |
| "epoch": 6.08, |
| "grad_norm": 3.657102566948412, |
| "learning_rate": 1e-05, |
| "loss": 0.061, |
| "step": 19050 |
| }, |
| { |
| "epoch": 6.08, |
| "grad_norm": 5.288001986723972, |
| "learning_rate": 1e-05, |
| "loss": 0.0654, |
| "step": 19060 |
| }, |
| { |
| "epoch": 6.08, |
| "grad_norm": 5.7018693769754005, |
| "learning_rate": 1e-05, |
| "loss": 0.0894, |
| "step": 19070 |
| }, |
| { |
| "epoch": 6.09, |
| "grad_norm": 6.256174013512853, |
| "learning_rate": 1e-05, |
| "loss": 0.0825, |
| "step": 19080 |
| }, |
| { |
| "epoch": 6.09, |
| "grad_norm": 9.44544012269032, |
| "learning_rate": 1e-05, |
| "loss": 0.0801, |
| "step": 19090 |
| }, |
| { |
| "epoch": 6.09, |
| "grad_norm": 3.6800463070236464, |
| "learning_rate": 1e-05, |
| "loss": 0.074, |
| "step": 19100 |
| }, |
| { |
| "epoch": 6.1, |
| "grad_norm": 3.3638922549489116, |
| "learning_rate": 1e-05, |
| "loss": 0.0649, |
| "step": 19110 |
| }, |
| { |
| "epoch": 6.1, |
| "grad_norm": 4.169454608260735, |
| "learning_rate": 1e-05, |
| "loss": 0.0845, |
| "step": 19120 |
| }, |
| { |
| "epoch": 6.1, |
| "grad_norm": 7.346063257673609, |
| "learning_rate": 1e-05, |
| "loss": 0.0892, |
| "step": 19130 |
| }, |
| { |
| "epoch": 6.11, |
| "grad_norm": 4.440802621208493, |
| "learning_rate": 1e-05, |
| "loss": 0.1018, |
| "step": 19140 |
| }, |
| { |
| "epoch": 6.11, |
| "grad_norm": 4.383749577783586, |
| "learning_rate": 1e-05, |
| "loss": 0.0665, |
| "step": 19150 |
| }, |
| { |
| "epoch": 6.11, |
| "grad_norm": 5.041705528088487, |
| "learning_rate": 1e-05, |
| "loss": 0.0886, |
| "step": 19160 |
| }, |
| { |
| "epoch": 6.12, |
| "grad_norm": 6.83876838705317, |
| "learning_rate": 1e-05, |
| "loss": 0.0886, |
| "step": 19170 |
| }, |
| { |
| "epoch": 6.12, |
| "grad_norm": 5.770606889195356, |
| "learning_rate": 1e-05, |
| "loss": 0.0854, |
| "step": 19180 |
| }, |
| { |
| "epoch": 6.12, |
| "grad_norm": 10.33207112468676, |
| "learning_rate": 1e-05, |
| "loss": 0.0864, |
| "step": 19190 |
| }, |
| { |
| "epoch": 6.13, |
| "grad_norm": 3.88049603447443, |
| "learning_rate": 1e-05, |
| "loss": 0.0795, |
| "step": 19200 |
| }, |
| { |
| "epoch": 6.13, |
| "grad_norm": 6.0914775034509345, |
| "learning_rate": 1e-05, |
| "loss": 0.0705, |
| "step": 19210 |
| }, |
| { |
| "epoch": 6.13, |
| "grad_norm": 4.62243457473495, |
| "learning_rate": 1e-05, |
| "loss": 0.076, |
| "step": 19220 |
| }, |
| { |
| "epoch": 6.14, |
| "grad_norm": 5.462649883247895, |
| "learning_rate": 1e-05, |
| "loss": 0.0703, |
| "step": 19230 |
| }, |
| { |
| "epoch": 6.14, |
| "grad_norm": 5.985450588209419, |
| "learning_rate": 1e-05, |
| "loss": 0.0867, |
| "step": 19240 |
| }, |
| { |
| "epoch": 6.14, |
| "grad_norm": 4.296728513128022, |
| "learning_rate": 1e-05, |
| "loss": 0.0817, |
| "step": 19250 |
| }, |
| { |
| "epoch": 6.15, |
| "grad_norm": 4.530980174482419, |
| "learning_rate": 1e-05, |
| "loss": 0.0875, |
| "step": 19260 |
| }, |
| { |
| "epoch": 6.15, |
| "grad_norm": 6.013099041002426, |
| "learning_rate": 1e-05, |
| "loss": 0.0637, |
| "step": 19270 |
| }, |
| { |
| "epoch": 6.15, |
| "grad_norm": 4.180840962335658, |
| "learning_rate": 1e-05, |
| "loss": 0.0822, |
| "step": 19280 |
| }, |
| { |
| "epoch": 6.16, |
| "grad_norm": 4.729255753852985, |
| "learning_rate": 1e-05, |
| "loss": 0.1036, |
| "step": 19290 |
| }, |
| { |
| "epoch": 6.16, |
| "grad_norm": 3.806093430078448, |
| "learning_rate": 1e-05, |
| "loss": 0.0665, |
| "step": 19300 |
| }, |
| { |
| "epoch": 6.16, |
| "grad_norm": 6.367634725953886, |
| "learning_rate": 1e-05, |
| "loss": 0.0676, |
| "step": 19310 |
| }, |
| { |
| "epoch": 6.16, |
| "grad_norm": 4.957073767609719, |
| "learning_rate": 1e-05, |
| "loss": 0.0873, |
| "step": 19320 |
| }, |
| { |
| "epoch": 6.17, |
| "grad_norm": 3.6738719947706344, |
| "learning_rate": 1e-05, |
| "loss": 0.0804, |
| "step": 19330 |
| }, |
| { |
| "epoch": 6.17, |
| "grad_norm": 3.9678001168189923, |
| "learning_rate": 1e-05, |
| "loss": 0.09, |
| "step": 19340 |
| }, |
| { |
| "epoch": 6.17, |
| "grad_norm": 5.7192889517212935, |
| "learning_rate": 1e-05, |
| "loss": 0.0887, |
| "step": 19350 |
| }, |
| { |
| "epoch": 6.18, |
| "grad_norm": 6.1691480575803705, |
| "learning_rate": 1e-05, |
| "loss": 0.0575, |
| "step": 19360 |
| }, |
| { |
| "epoch": 6.18, |
| "grad_norm": 5.082447454253132, |
| "learning_rate": 1e-05, |
| "loss": 0.0781, |
| "step": 19370 |
| }, |
| { |
| "epoch": 6.18, |
| "grad_norm": 5.178177327160451, |
| "learning_rate": 1e-05, |
| "loss": 0.1, |
| "step": 19380 |
| }, |
| { |
| "epoch": 6.19, |
| "grad_norm": 6.093726211892672, |
| "learning_rate": 1e-05, |
| "loss": 0.1005, |
| "step": 19390 |
| }, |
| { |
| "epoch": 6.19, |
| "grad_norm": 4.216505569778215, |
| "learning_rate": 1e-05, |
| "loss": 0.0828, |
| "step": 19400 |
| }, |
| { |
| "epoch": 6.19, |
| "grad_norm": 3.668021139759976, |
| "learning_rate": 1e-05, |
| "loss": 0.0791, |
| "step": 19410 |
| }, |
| { |
| "epoch": 6.2, |
| "grad_norm": 6.528254501935492, |
| "learning_rate": 1e-05, |
| "loss": 0.0993, |
| "step": 19420 |
| }, |
| { |
| "epoch": 6.2, |
| "grad_norm": 5.889020058632553, |
| "learning_rate": 1e-05, |
| "loss": 0.0776, |
| "step": 19430 |
| }, |
| { |
| "epoch": 6.2, |
| "grad_norm": 5.025537505747986, |
| "learning_rate": 1e-05, |
| "loss": 0.0678, |
| "step": 19440 |
| }, |
| { |
| "epoch": 6.21, |
| "grad_norm": 3.939698226331934, |
| "learning_rate": 1e-05, |
| "loss": 0.0894, |
| "step": 19450 |
| }, |
| { |
| "epoch": 6.21, |
| "grad_norm": 6.872651964979624, |
| "learning_rate": 1e-05, |
| "loss": 0.0808, |
| "step": 19460 |
| }, |
| { |
| "epoch": 6.21, |
| "grad_norm": 6.005591012805146, |
| "learning_rate": 1e-05, |
| "loss": 0.0932, |
| "step": 19470 |
| }, |
| { |
| "epoch": 6.22, |
| "grad_norm": 4.689955614933253, |
| "learning_rate": 1e-05, |
| "loss": 0.0875, |
| "step": 19480 |
| }, |
| { |
| "epoch": 6.22, |
| "grad_norm": 5.006036738164488, |
| "learning_rate": 1e-05, |
| "loss": 0.0893, |
| "step": 19490 |
| }, |
| { |
| "epoch": 6.22, |
| "grad_norm": 7.405384898463061, |
| "learning_rate": 1e-05, |
| "loss": 0.0908, |
| "step": 19500 |
| }, |
| { |
| "epoch": 6.22, |
| "eval_loss": 1.0458984375, |
| "eval_runtime": 54.9194, |
| "eval_samples_per_second": 46.14, |
| "eval_steps_per_second": 2.895, |
| "step": 19500 |
| }, |
| { |
| "epoch": 6.23, |
| "grad_norm": 4.285429536351654, |
| "learning_rate": 1e-05, |
| "loss": 0.0835, |
| "step": 19510 |
| }, |
| { |
| "epoch": 6.23, |
| "grad_norm": 3.63321859079401, |
| "learning_rate": 1e-05, |
| "loss": 0.0856, |
| "step": 19520 |
| }, |
| { |
| "epoch": 6.23, |
| "grad_norm": 4.551864141545301, |
| "learning_rate": 1e-05, |
| "loss": 0.0701, |
| "step": 19530 |
| }, |
| { |
| "epoch": 6.23, |
| "grad_norm": 4.854904620788458, |
| "learning_rate": 1e-05, |
| "loss": 0.0953, |
| "step": 19540 |
| }, |
| { |
| "epoch": 6.24, |
| "grad_norm": 3.792570837738583, |
| "learning_rate": 1e-05, |
| "loss": 0.0539, |
| "step": 19550 |
| }, |
| { |
| "epoch": 6.24, |
| "grad_norm": 4.119952726370692, |
| "learning_rate": 1e-05, |
| "loss": 0.0879, |
| "step": 19560 |
| }, |
| { |
| "epoch": 6.24, |
| "grad_norm": 3.3093332952621832, |
| "learning_rate": 1e-05, |
| "loss": 0.0777, |
| "step": 19570 |
| }, |
| { |
| "epoch": 6.25, |
| "grad_norm": 7.2998628890865405, |
| "learning_rate": 1e-05, |
| "loss": 0.0893, |
| "step": 19580 |
| }, |
| { |
| "epoch": 6.25, |
| "grad_norm": 3.080240674709647, |
| "learning_rate": 1e-05, |
| "loss": 0.0871, |
| "step": 19590 |
| }, |
| { |
| "epoch": 6.25, |
| "grad_norm": 5.9183438928127305, |
| "learning_rate": 1e-05, |
| "loss": 0.0973, |
| "step": 19600 |
| }, |
| { |
| "epoch": 6.26, |
| "grad_norm": 3.8880719613531016, |
| "learning_rate": 1e-05, |
| "loss": 0.0855, |
| "step": 19610 |
| }, |
| { |
| "epoch": 6.26, |
| "grad_norm": 6.664293184569714, |
| "learning_rate": 1e-05, |
| "loss": 0.09, |
| "step": 19620 |
| }, |
| { |
| "epoch": 6.26, |
| "grad_norm": 3.1237876066132393, |
| "learning_rate": 1e-05, |
| "loss": 0.0772, |
| "step": 19630 |
| }, |
| { |
| "epoch": 6.27, |
| "grad_norm": 5.997982639675578, |
| "learning_rate": 1e-05, |
| "loss": 0.0814, |
| "step": 19640 |
| }, |
| { |
| "epoch": 6.27, |
| "grad_norm": 6.1754822824536095, |
| "learning_rate": 1e-05, |
| "loss": 0.0886, |
| "step": 19650 |
| }, |
| { |
| "epoch": 6.27, |
| "grad_norm": 4.092018984499807, |
| "learning_rate": 1e-05, |
| "loss": 0.0898, |
| "step": 19660 |
| }, |
| { |
| "epoch": 6.28, |
| "grad_norm": 5.568332657336815, |
| "learning_rate": 1e-05, |
| "loss": 0.0842, |
| "step": 19670 |
| }, |
| { |
| "epoch": 6.28, |
| "grad_norm": 4.62601748434903, |
| "learning_rate": 1e-05, |
| "loss": 0.0843, |
| "step": 19680 |
| }, |
| { |
| "epoch": 6.28, |
| "grad_norm": 4.186432887513578, |
| "learning_rate": 1e-05, |
| "loss": 0.0738, |
| "step": 19690 |
| }, |
| { |
| "epoch": 6.29, |
| "grad_norm": 4.84224219695985, |
| "learning_rate": 1e-05, |
| "loss": 0.0969, |
| "step": 19700 |
| }, |
| { |
| "epoch": 6.29, |
| "grad_norm": 5.1047067577044345, |
| "learning_rate": 1e-05, |
| "loss": 0.0801, |
| "step": 19710 |
| }, |
| { |
| "epoch": 6.29, |
| "grad_norm": 4.291822535576219, |
| "learning_rate": 1e-05, |
| "loss": 0.0868, |
| "step": 19720 |
| }, |
| { |
| "epoch": 6.3, |
| "grad_norm": 6.861174119393186, |
| "learning_rate": 1e-05, |
| "loss": 0.0889, |
| "step": 19730 |
| }, |
| { |
| "epoch": 6.3, |
| "grad_norm": 3.9300227543606514, |
| "learning_rate": 1e-05, |
| "loss": 0.0685, |
| "step": 19740 |
| }, |
| { |
| "epoch": 6.3, |
| "grad_norm": 5.458856669179322, |
| "learning_rate": 1e-05, |
| "loss": 0.0732, |
| "step": 19750 |
| }, |
| { |
| "epoch": 6.31, |
| "grad_norm": 4.37939167309703, |
| "learning_rate": 1e-05, |
| "loss": 0.0747, |
| "step": 19760 |
| }, |
| { |
| "epoch": 6.31, |
| "grad_norm": 4.928893108413888, |
| "learning_rate": 1e-05, |
| "loss": 0.0759, |
| "step": 19770 |
| }, |
| { |
| "epoch": 6.31, |
| "grad_norm": 3.7510909718785923, |
| "learning_rate": 1e-05, |
| "loss": 0.0867, |
| "step": 19780 |
| }, |
| { |
| "epoch": 6.31, |
| "grad_norm": 6.828539244353064, |
| "learning_rate": 1e-05, |
| "loss": 0.0848, |
| "step": 19790 |
| }, |
| { |
| "epoch": 6.32, |
| "grad_norm": 5.203498345804488, |
| "learning_rate": 1e-05, |
| "loss": 0.078, |
| "step": 19800 |
| }, |
| { |
| "epoch": 6.32, |
| "grad_norm": 3.6948894879613206, |
| "learning_rate": 1e-05, |
| "loss": 0.0824, |
| "step": 19810 |
| }, |
| { |
| "epoch": 6.32, |
| "grad_norm": 7.188610488686714, |
| "learning_rate": 1e-05, |
| "loss": 0.0968, |
| "step": 19820 |
| }, |
| { |
| "epoch": 6.33, |
| "grad_norm": 6.323132917130187, |
| "learning_rate": 1e-05, |
| "loss": 0.0897, |
| "step": 19830 |
| }, |
| { |
| "epoch": 6.33, |
| "grad_norm": 4.86343773279874, |
| "learning_rate": 1e-05, |
| "loss": 0.1034, |
| "step": 19840 |
| }, |
| { |
| "epoch": 6.33, |
| "grad_norm": 3.6473104627177264, |
| "learning_rate": 1e-05, |
| "loss": 0.0726, |
| "step": 19850 |
| }, |
| { |
| "epoch": 6.34, |
| "grad_norm": 5.12842705747701, |
| "learning_rate": 1e-05, |
| "loss": 0.0738, |
| "step": 19860 |
| }, |
| { |
| "epoch": 6.34, |
| "grad_norm": 4.279538223097687, |
| "learning_rate": 1e-05, |
| "loss": 0.08, |
| "step": 19870 |
| }, |
| { |
| "epoch": 6.34, |
| "grad_norm": 5.406675288941782, |
| "learning_rate": 1e-05, |
| "loss": 0.0794, |
| "step": 19880 |
| }, |
| { |
| "epoch": 6.35, |
| "grad_norm": 4.7496524482867795, |
| "learning_rate": 1e-05, |
| "loss": 0.1084, |
| "step": 19890 |
| }, |
| { |
| "epoch": 6.35, |
| "grad_norm": 5.604905071560628, |
| "learning_rate": 1e-05, |
| "loss": 0.1027, |
| "step": 19900 |
| }, |
| { |
| "epoch": 6.35, |
| "grad_norm": 7.4852167026541965, |
| "learning_rate": 1e-05, |
| "loss": 0.1091, |
| "step": 19910 |
| }, |
| { |
| "epoch": 6.36, |
| "grad_norm": 5.387743901442809, |
| "learning_rate": 1e-05, |
| "loss": 0.0974, |
| "step": 19920 |
| }, |
| { |
| "epoch": 6.36, |
| "grad_norm": 5.785185922653477, |
| "learning_rate": 1e-05, |
| "loss": 0.1093, |
| "step": 19930 |
| }, |
| { |
| "epoch": 6.36, |
| "grad_norm": 5.019915590526472, |
| "learning_rate": 1e-05, |
| "loss": 0.072, |
| "step": 19940 |
| }, |
| { |
| "epoch": 6.37, |
| "grad_norm": 5.997453785090542, |
| "learning_rate": 1e-05, |
| "loss": 0.0943, |
| "step": 19950 |
| }, |
| { |
| "epoch": 6.37, |
| "grad_norm": 2.818716257912607, |
| "learning_rate": 1e-05, |
| "loss": 0.0718, |
| "step": 19960 |
| }, |
| { |
| "epoch": 6.37, |
| "grad_norm": 4.729682031633979, |
| "learning_rate": 1e-05, |
| "loss": 0.0951, |
| "step": 19970 |
| }, |
| { |
| "epoch": 6.38, |
| "grad_norm": 6.604283312741146, |
| "learning_rate": 1e-05, |
| "loss": 0.0776, |
| "step": 19980 |
| }, |
| { |
| "epoch": 6.38, |
| "grad_norm": 5.919398935064388, |
| "learning_rate": 1e-05, |
| "loss": 0.0874, |
| "step": 19990 |
| }, |
| { |
| "epoch": 6.38, |
| "grad_norm": 6.303698456247326, |
| "learning_rate": 1e-05, |
| "loss": 0.0971, |
| "step": 20000 |
| }, |
| { |
| "epoch": 6.38, |
| "eval_loss": 1.0830078125, |
| "eval_runtime": 54.7664, |
| "eval_samples_per_second": 46.269, |
| "eval_steps_per_second": 2.903, |
| "step": 20000 |
| }, |
| { |
| "epoch": 6.38, |
| "grad_norm": 9.819105124167534, |
| "learning_rate": 1e-05, |
| "loss": 0.0853, |
| "step": 20010 |
| }, |
| { |
| "epoch": 6.39, |
| "grad_norm": 6.899940059235654, |
| "learning_rate": 1e-05, |
| "loss": 0.0846, |
| "step": 20020 |
| }, |
| { |
| "epoch": 6.39, |
| "grad_norm": 9.088081402131989, |
| "learning_rate": 1e-05, |
| "loss": 0.0828, |
| "step": 20030 |
| }, |
| { |
| "epoch": 6.39, |
| "grad_norm": 4.693976848621594, |
| "learning_rate": 1e-05, |
| "loss": 0.0922, |
| "step": 20040 |
| }, |
| { |
| "epoch": 6.4, |
| "grad_norm": 15.401053335653904, |
| "learning_rate": 1e-05, |
| "loss": 0.1068, |
| "step": 20050 |
| }, |
| { |
| "epoch": 6.4, |
| "grad_norm": 4.4273081045497955, |
| "learning_rate": 1e-05, |
| "loss": 0.073, |
| "step": 20060 |
| }, |
| { |
| "epoch": 6.4, |
| "grad_norm": 6.13409247192293, |
| "learning_rate": 1e-05, |
| "loss": 0.1027, |
| "step": 20070 |
| }, |
| { |
| "epoch": 6.41, |
| "grad_norm": 5.970899903733137, |
| "learning_rate": 1e-05, |
| "loss": 0.0846, |
| "step": 20080 |
| }, |
| { |
| "epoch": 6.41, |
| "grad_norm": 5.450989750574134, |
| "learning_rate": 1e-05, |
| "loss": 0.0989, |
| "step": 20090 |
| }, |
| { |
| "epoch": 6.41, |
| "grad_norm": 3.3573088213586573, |
| "learning_rate": 1e-05, |
| "loss": 0.081, |
| "step": 20100 |
| }, |
| { |
| "epoch": 6.42, |
| "grad_norm": 7.333823129885221, |
| "learning_rate": 1e-05, |
| "loss": 0.0809, |
| "step": 20110 |
| }, |
| { |
| "epoch": 6.42, |
| "grad_norm": 5.0271906618227025, |
| "learning_rate": 1e-05, |
| "loss": 0.0982, |
| "step": 20120 |
| }, |
| { |
| "epoch": 6.42, |
| "grad_norm": 6.9282794885269015, |
| "learning_rate": 1e-05, |
| "loss": 0.1011, |
| "step": 20130 |
| }, |
| { |
| "epoch": 6.43, |
| "grad_norm": 13.309255858383152, |
| "learning_rate": 1e-05, |
| "loss": 0.0797, |
| "step": 20140 |
| }, |
| { |
| "epoch": 6.43, |
| "grad_norm": 4.103379666283507, |
| "learning_rate": 1e-05, |
| "loss": 0.0731, |
| "step": 20150 |
| }, |
| { |
| "epoch": 6.43, |
| "grad_norm": 7.414471276352841, |
| "learning_rate": 1e-05, |
| "loss": 0.0817, |
| "step": 20160 |
| }, |
| { |
| "epoch": 6.44, |
| "grad_norm": 7.046932651596904, |
| "learning_rate": 1e-05, |
| "loss": 0.0867, |
| "step": 20170 |
| }, |
| { |
| "epoch": 6.44, |
| "grad_norm": 6.261339201106149, |
| "learning_rate": 1e-05, |
| "loss": 0.0748, |
| "step": 20180 |
| }, |
| { |
| "epoch": 6.44, |
| "grad_norm": 5.699562444121773, |
| "learning_rate": 1e-05, |
| "loss": 0.0854, |
| "step": 20190 |
| }, |
| { |
| "epoch": 6.45, |
| "grad_norm": 5.416472367934471, |
| "learning_rate": 1e-05, |
| "loss": 0.0966, |
| "step": 20200 |
| }, |
| { |
| "epoch": 6.45, |
| "grad_norm": 5.995259637626931, |
| "learning_rate": 1e-05, |
| "loss": 0.0768, |
| "step": 20210 |
| }, |
| { |
| "epoch": 6.45, |
| "grad_norm": 5.928803509947881, |
| "learning_rate": 1e-05, |
| "loss": 0.078, |
| "step": 20220 |
| }, |
| { |
| "epoch": 6.46, |
| "grad_norm": 5.743846752194501, |
| "learning_rate": 1e-05, |
| "loss": 0.1065, |
| "step": 20230 |
| }, |
| { |
| "epoch": 6.46, |
| "grad_norm": 9.475017424796395, |
| "learning_rate": 1e-05, |
| "loss": 0.1006, |
| "step": 20240 |
| }, |
| { |
| "epoch": 6.46, |
| "grad_norm": 6.74642454042064, |
| "learning_rate": 1e-05, |
| "loss": 0.1136, |
| "step": 20250 |
| }, |
| { |
| "epoch": 6.46, |
| "grad_norm": 6.482275933314993, |
| "learning_rate": 1e-05, |
| "loss": 0.0816, |
| "step": 20260 |
| }, |
| { |
| "epoch": 6.47, |
| "grad_norm": 3.9065521123406386, |
| "learning_rate": 1e-05, |
| "loss": 0.0771, |
| "step": 20270 |
| }, |
| { |
| "epoch": 6.47, |
| "grad_norm": 6.720504744511092, |
| "learning_rate": 1e-05, |
| "loss": 0.0945, |
| "step": 20280 |
| }, |
| { |
| "epoch": 6.47, |
| "grad_norm": 5.443427527212846, |
| "learning_rate": 1e-05, |
| "loss": 0.1139, |
| "step": 20290 |
| }, |
| { |
| "epoch": 6.48, |
| "grad_norm": 3.5958783025718124, |
| "learning_rate": 1e-05, |
| "loss": 0.1017, |
| "step": 20300 |
| }, |
| { |
| "epoch": 6.48, |
| "grad_norm": 9.725805144334654, |
| "learning_rate": 1e-05, |
| "loss": 0.09, |
| "step": 20310 |
| }, |
| { |
| "epoch": 6.48, |
| "grad_norm": 4.1546076921514015, |
| "learning_rate": 1e-05, |
| "loss": 0.105, |
| "step": 20320 |
| }, |
| { |
| "epoch": 6.49, |
| "grad_norm": 4.390050015523052, |
| "learning_rate": 1e-05, |
| "loss": 0.08, |
| "step": 20330 |
| }, |
| { |
| "epoch": 6.49, |
| "grad_norm": 7.275337143259803, |
| "learning_rate": 1e-05, |
| "loss": 0.097, |
| "step": 20340 |
| }, |
| { |
| "epoch": 6.49, |
| "grad_norm": 5.963137559372746, |
| "learning_rate": 1e-05, |
| "loss": 0.0992, |
| "step": 20350 |
| }, |
| { |
| "epoch": 6.5, |
| "grad_norm": 6.020195351913035, |
| "learning_rate": 1e-05, |
| "loss": 0.1101, |
| "step": 20360 |
| }, |
| { |
| "epoch": 6.5, |
| "grad_norm": 4.64954707903481, |
| "learning_rate": 1e-05, |
| "loss": 0.0815, |
| "step": 20370 |
| }, |
| { |
| "epoch": 6.5, |
| "grad_norm": 5.288023988989638, |
| "learning_rate": 1e-05, |
| "loss": 0.0919, |
| "step": 20380 |
| }, |
| { |
| "epoch": 6.51, |
| "grad_norm": 4.704282700510075, |
| "learning_rate": 1e-05, |
| "loss": 0.1127, |
| "step": 20390 |
| }, |
| { |
| "epoch": 6.51, |
| "grad_norm": 3.2801623767109964, |
| "learning_rate": 1e-05, |
| "loss": 0.1324, |
| "step": 20400 |
| }, |
| { |
| "epoch": 6.51, |
| "grad_norm": 4.751346196598564, |
| "learning_rate": 1e-05, |
| "loss": 0.0863, |
| "step": 20410 |
| }, |
| { |
| "epoch": 6.52, |
| "grad_norm": 5.4045013951520255, |
| "learning_rate": 1e-05, |
| "loss": 0.0784, |
| "step": 20420 |
| }, |
| { |
| "epoch": 6.52, |
| "grad_norm": 4.62514351287054, |
| "learning_rate": 1e-05, |
| "loss": 0.0966, |
| "step": 20430 |
| }, |
| { |
| "epoch": 6.52, |
| "grad_norm": 5.6115407349101245, |
| "learning_rate": 1e-05, |
| "loss": 0.0776, |
| "step": 20440 |
| }, |
| { |
| "epoch": 6.53, |
| "grad_norm": 4.390851541016733, |
| "learning_rate": 1e-05, |
| "loss": 0.0784, |
| "step": 20450 |
| }, |
| { |
| "epoch": 6.53, |
| "grad_norm": 5.542640638687826, |
| "learning_rate": 1e-05, |
| "loss": 0.0863, |
| "step": 20460 |
| }, |
| { |
| "epoch": 6.53, |
| "grad_norm": 4.08714378741307, |
| "learning_rate": 1e-05, |
| "loss": 0.0931, |
| "step": 20470 |
| }, |
| { |
| "epoch": 6.53, |
| "grad_norm": 7.606133054337349, |
| "learning_rate": 1e-05, |
| "loss": 0.1032, |
| "step": 20480 |
| }, |
| { |
| "epoch": 6.54, |
| "grad_norm": 5.941312830564872, |
| "learning_rate": 1e-05, |
| "loss": 0.1107, |
| "step": 20490 |
| }, |
| { |
| "epoch": 6.54, |
| "grad_norm": 3.858198225870447, |
| "learning_rate": 1e-05, |
| "loss": 0.1119, |
| "step": 20500 |
| }, |
| { |
| "epoch": 6.54, |
| "eval_loss": 1.060546875, |
| "eval_runtime": 54.9574, |
| "eval_samples_per_second": 46.108, |
| "eval_steps_per_second": 2.893, |
| "step": 20500 |
| }, |
| { |
| "epoch": 6.54, |
| "grad_norm": 3.4250506989005376, |
| "learning_rate": 1e-05, |
| "loss": 0.0888, |
| "step": 20510 |
| }, |
| { |
| "epoch": 6.55, |
| "grad_norm": 6.104948894214968, |
| "learning_rate": 1e-05, |
| "loss": 0.1165, |
| "step": 20520 |
| }, |
| { |
| "epoch": 6.55, |
| "grad_norm": 5.1628946666091275, |
| "learning_rate": 1e-05, |
| "loss": 0.0717, |
| "step": 20530 |
| }, |
| { |
| "epoch": 6.55, |
| "grad_norm": 3.5812625372169564, |
| "learning_rate": 1e-05, |
| "loss": 0.0936, |
| "step": 20540 |
| }, |
| { |
| "epoch": 6.56, |
| "grad_norm": 6.262284679860945, |
| "learning_rate": 1e-05, |
| "loss": 0.094, |
| "step": 20550 |
| }, |
| { |
| "epoch": 6.56, |
| "grad_norm": 4.378007781306323, |
| "learning_rate": 1e-05, |
| "loss": 0.1093, |
| "step": 20560 |
| }, |
| { |
| "epoch": 6.56, |
| "grad_norm": 5.813185784686841, |
| "learning_rate": 1e-05, |
| "loss": 0.1271, |
| "step": 20570 |
| }, |
| { |
| "epoch": 6.57, |
| "grad_norm": 6.018002840213226, |
| "learning_rate": 1e-05, |
| "loss": 0.0982, |
| "step": 20580 |
| }, |
| { |
| "epoch": 6.57, |
| "grad_norm": 8.577697632085224, |
| "learning_rate": 1e-05, |
| "loss": 0.1016, |
| "step": 20590 |
| }, |
| { |
| "epoch": 6.57, |
| "grad_norm": 8.524266329431414, |
| "learning_rate": 1e-05, |
| "loss": 0.1125, |
| "step": 20600 |
| }, |
| { |
| "epoch": 6.58, |
| "grad_norm": 6.913732005695769, |
| "learning_rate": 1e-05, |
| "loss": 0.0869, |
| "step": 20610 |
| }, |
| { |
| "epoch": 6.58, |
| "grad_norm": 6.0649824270825645, |
| "learning_rate": 1e-05, |
| "loss": 0.1157, |
| "step": 20620 |
| }, |
| { |
| "epoch": 6.58, |
| "grad_norm": 3.708450801289004, |
| "learning_rate": 1e-05, |
| "loss": 0.068, |
| "step": 20630 |
| }, |
| { |
| "epoch": 6.59, |
| "grad_norm": 4.606171506629592, |
| "learning_rate": 1e-05, |
| "loss": 0.1156, |
| "step": 20640 |
| }, |
| { |
| "epoch": 6.59, |
| "grad_norm": 4.264041369124768, |
| "learning_rate": 1e-05, |
| "loss": 0.0859, |
| "step": 20650 |
| }, |
| { |
| "epoch": 6.59, |
| "grad_norm": 4.632536855799089, |
| "learning_rate": 1e-05, |
| "loss": 0.0808, |
| "step": 20660 |
| }, |
| { |
| "epoch": 6.6, |
| "grad_norm": 5.18590732622003, |
| "learning_rate": 1e-05, |
| "loss": 0.0864, |
| "step": 20670 |
| }, |
| { |
| "epoch": 6.6, |
| "grad_norm": 5.063489464032503, |
| "learning_rate": 1e-05, |
| "loss": 0.0978, |
| "step": 20680 |
| }, |
| { |
| "epoch": 6.6, |
| "grad_norm": 6.098891373246945, |
| "learning_rate": 1e-05, |
| "loss": 0.1043, |
| "step": 20690 |
| }, |
| { |
| "epoch": 6.6, |
| "grad_norm": 6.70778936546063, |
| "learning_rate": 1e-05, |
| "loss": 0.1073, |
| "step": 20700 |
| }, |
| { |
| "epoch": 6.61, |
| "grad_norm": 5.653455180768397, |
| "learning_rate": 1e-05, |
| "loss": 0.1231, |
| "step": 20710 |
| }, |
| { |
| "epoch": 6.61, |
| "grad_norm": 5.239044202163376, |
| "learning_rate": 1e-05, |
| "loss": 0.0807, |
| "step": 20720 |
| }, |
| { |
| "epoch": 6.61, |
| "grad_norm": 3.7121246045344227, |
| "learning_rate": 1e-05, |
| "loss": 0.0764, |
| "step": 20730 |
| }, |
| { |
| "epoch": 6.62, |
| "grad_norm": 5.62368877810841, |
| "learning_rate": 1e-05, |
| "loss": 0.0899, |
| "step": 20740 |
| }, |
| { |
| "epoch": 6.62, |
| "grad_norm": 5.9834251347178595, |
| "learning_rate": 1e-05, |
| "loss": 0.0818, |
| "step": 20750 |
| }, |
| { |
| "epoch": 6.62, |
| "grad_norm": 8.69487290415081, |
| "learning_rate": 1e-05, |
| "loss": 0.1228, |
| "step": 20760 |
| }, |
| { |
| "epoch": 6.63, |
| "grad_norm": 6.6252812739708276, |
| "learning_rate": 1e-05, |
| "loss": 0.086, |
| "step": 20770 |
| }, |
| { |
| "epoch": 6.63, |
| "grad_norm": 4.520294626136672, |
| "learning_rate": 1e-05, |
| "loss": 0.088, |
| "step": 20780 |
| }, |
| { |
| "epoch": 6.63, |
| "grad_norm": 3.625127066654305, |
| "learning_rate": 1e-05, |
| "loss": 0.0879, |
| "step": 20790 |
| }, |
| { |
| "epoch": 6.64, |
| "grad_norm": 4.4107283972034015, |
| "learning_rate": 1e-05, |
| "loss": 0.0879, |
| "step": 20800 |
| }, |
| { |
| "epoch": 6.64, |
| "grad_norm": 4.40585758274467, |
| "learning_rate": 1e-05, |
| "loss": 0.0923, |
| "step": 20810 |
| }, |
| { |
| "epoch": 6.64, |
| "grad_norm": 5.6106176673456245, |
| "learning_rate": 1e-05, |
| "loss": 0.0983, |
| "step": 20820 |
| }, |
| { |
| "epoch": 6.65, |
| "grad_norm": 4.842340473468004, |
| "learning_rate": 1e-05, |
| "loss": 0.0965, |
| "step": 20830 |
| }, |
| { |
| "epoch": 6.65, |
| "grad_norm": 3.9221474602273774, |
| "learning_rate": 1e-05, |
| "loss": 0.1056, |
| "step": 20840 |
| }, |
| { |
| "epoch": 6.65, |
| "grad_norm": 5.673879222016044, |
| "learning_rate": 1e-05, |
| "loss": 0.0961, |
| "step": 20850 |
| }, |
| { |
| "epoch": 6.66, |
| "grad_norm": 3.892887790304312, |
| "learning_rate": 1e-05, |
| "loss": 0.0841, |
| "step": 20860 |
| }, |
| { |
| "epoch": 6.66, |
| "grad_norm": 5.673215933517806, |
| "learning_rate": 1e-05, |
| "loss": 0.0905, |
| "step": 20870 |
| }, |
| { |
| "epoch": 6.66, |
| "grad_norm": 6.629473345568229, |
| "learning_rate": 1e-05, |
| "loss": 0.0989, |
| "step": 20880 |
| }, |
| { |
| "epoch": 6.67, |
| "grad_norm": 5.383119172992518, |
| "learning_rate": 1e-05, |
| "loss": 0.1172, |
| "step": 20890 |
| }, |
| { |
| "epoch": 6.67, |
| "grad_norm": 6.50666086326924, |
| "learning_rate": 1e-05, |
| "loss": 0.0968, |
| "step": 20900 |
| }, |
| { |
| "epoch": 6.67, |
| "grad_norm": 6.800190575117006, |
| "learning_rate": 1e-05, |
| "loss": 0.0822, |
| "step": 20910 |
| }, |
| { |
| "epoch": 6.68, |
| "grad_norm": 5.056464659633841, |
| "learning_rate": 1e-05, |
| "loss": 0.0785, |
| "step": 20920 |
| }, |
| { |
| "epoch": 6.68, |
| "grad_norm": 3.849873325196058, |
| "learning_rate": 1e-05, |
| "loss": 0.0887, |
| "step": 20930 |
| }, |
| { |
| "epoch": 6.68, |
| "grad_norm": 6.881896565314925, |
| "learning_rate": 1e-05, |
| "loss": 0.0747, |
| "step": 20940 |
| }, |
| { |
| "epoch": 6.68, |
| "grad_norm": 5.5352964955547925, |
| "learning_rate": 1e-05, |
| "loss": 0.1054, |
| "step": 20950 |
| }, |
| { |
| "epoch": 6.69, |
| "grad_norm": 5.716367798639812, |
| "learning_rate": 1e-05, |
| "loss": 0.0973, |
| "step": 20960 |
| }, |
| { |
| "epoch": 6.69, |
| "grad_norm": 3.9847964793157735, |
| "learning_rate": 1e-05, |
| "loss": 0.0789, |
| "step": 20970 |
| }, |
| { |
| "epoch": 6.69, |
| "grad_norm": 6.957442205629719, |
| "learning_rate": 1e-05, |
| "loss": 0.0888, |
| "step": 20980 |
| }, |
| { |
| "epoch": 6.7, |
| "grad_norm": 4.017805049497983, |
| "learning_rate": 1e-05, |
| "loss": 0.0899, |
| "step": 20990 |
| }, |
| { |
| "epoch": 6.7, |
| "grad_norm": 5.214827108178034, |
| "learning_rate": 1e-05, |
| "loss": 0.1098, |
| "step": 21000 |
| }, |
| { |
| "epoch": 6.7, |
| "eval_loss": 1.0546875, |
| "eval_runtime": 55.0295, |
| "eval_samples_per_second": 46.048, |
| "eval_steps_per_second": 2.889, |
| "step": 21000 |
| }, |
| { |
| "epoch": 6.7, |
| "grad_norm": 6.230904055130225, |
| "learning_rate": 1e-05, |
| "loss": 0.1194, |
| "step": 21010 |
| }, |
| { |
| "epoch": 6.71, |
| "grad_norm": 3.973509570273326, |
| "learning_rate": 1e-05, |
| "loss": 0.0976, |
| "step": 21020 |
| }, |
| { |
| "epoch": 6.71, |
| "grad_norm": 4.754559587205332, |
| "learning_rate": 1e-05, |
| "loss": 0.0973, |
| "step": 21030 |
| }, |
| { |
| "epoch": 6.71, |
| "grad_norm": 4.4165449065692926, |
| "learning_rate": 1e-05, |
| "loss": 0.0856, |
| "step": 21040 |
| }, |
| { |
| "epoch": 6.72, |
| "grad_norm": 1.9255436563072184, |
| "learning_rate": 1e-05, |
| "loss": 0.0912, |
| "step": 21050 |
| }, |
| { |
| "epoch": 6.72, |
| "grad_norm": 6.991253974868406, |
| "learning_rate": 1e-05, |
| "loss": 0.1174, |
| "step": 21060 |
| }, |
| { |
| "epoch": 6.72, |
| "grad_norm": 4.340354855001372, |
| "learning_rate": 1e-05, |
| "loss": 0.094, |
| "step": 21070 |
| }, |
| { |
| "epoch": 6.73, |
| "grad_norm": 4.1291499214846485, |
| "learning_rate": 1e-05, |
| "loss": 0.1212, |
| "step": 21080 |
| }, |
| { |
| "epoch": 6.73, |
| "grad_norm": 3.4621964139089347, |
| "learning_rate": 1e-05, |
| "loss": 0.0853, |
| "step": 21090 |
| }, |
| { |
| "epoch": 6.73, |
| "grad_norm": 4.835213109303242, |
| "learning_rate": 1e-05, |
| "loss": 0.0951, |
| "step": 21100 |
| }, |
| { |
| "epoch": 6.74, |
| "grad_norm": 6.829811149879444, |
| "learning_rate": 1e-05, |
| "loss": 0.0901, |
| "step": 21110 |
| }, |
| { |
| "epoch": 6.74, |
| "grad_norm": 5.87059960027492, |
| "learning_rate": 1e-05, |
| "loss": 0.0854, |
| "step": 21120 |
| }, |
| { |
| "epoch": 6.74, |
| "grad_norm": 5.430804284598397, |
| "learning_rate": 1e-05, |
| "loss": 0.0861, |
| "step": 21130 |
| }, |
| { |
| "epoch": 6.75, |
| "grad_norm": 8.296726031501635, |
| "learning_rate": 1e-05, |
| "loss": 0.1037, |
| "step": 21140 |
| }, |
| { |
| "epoch": 6.75, |
| "grad_norm": 5.159455753574761, |
| "learning_rate": 1e-05, |
| "loss": 0.0863, |
| "step": 21150 |
| }, |
| { |
| "epoch": 6.75, |
| "grad_norm": 5.399255715399432, |
| "learning_rate": 1e-05, |
| "loss": 0.1007, |
| "step": 21160 |
| }, |
| { |
| "epoch": 6.75, |
| "grad_norm": 6.5540350679094415, |
| "learning_rate": 1e-05, |
| "loss": 0.1111, |
| "step": 21170 |
| }, |
| { |
| "epoch": 6.76, |
| "grad_norm": 3.9303374767275696, |
| "learning_rate": 1e-05, |
| "loss": 0.0967, |
| "step": 21180 |
| }, |
| { |
| "epoch": 6.76, |
| "grad_norm": 7.178121003617738, |
| "learning_rate": 1e-05, |
| "loss": 0.0962, |
| "step": 21190 |
| }, |
| { |
| "epoch": 6.76, |
| "grad_norm": 4.040999341058901, |
| "learning_rate": 1e-05, |
| "loss": 0.0978, |
| "step": 21200 |
| }, |
| { |
| "epoch": 6.77, |
| "grad_norm": 4.198760876200002, |
| "learning_rate": 1e-05, |
| "loss": 0.0989, |
| "step": 21210 |
| }, |
| { |
| "epoch": 6.77, |
| "grad_norm": 8.302426307578612, |
| "learning_rate": 1e-05, |
| "loss": 0.0957, |
| "step": 21220 |
| }, |
| { |
| "epoch": 6.77, |
| "grad_norm": 4.246847217460519, |
| "learning_rate": 1e-05, |
| "loss": 0.11, |
| "step": 21230 |
| }, |
| { |
| "epoch": 6.78, |
| "grad_norm": 5.426861759087483, |
| "learning_rate": 1e-05, |
| "loss": 0.0994, |
| "step": 21240 |
| }, |
| { |
| "epoch": 6.78, |
| "grad_norm": 5.984913933692043, |
| "learning_rate": 1e-05, |
| "loss": 0.105, |
| "step": 21250 |
| }, |
| { |
| "epoch": 6.78, |
| "grad_norm": 5.315593761103369, |
| "learning_rate": 1e-05, |
| "loss": 0.1104, |
| "step": 21260 |
| }, |
| { |
| "epoch": 6.79, |
| "grad_norm": 4.890313574287276, |
| "learning_rate": 1e-05, |
| "loss": 0.1133, |
| "step": 21270 |
| }, |
| { |
| "epoch": 6.79, |
| "grad_norm": 3.6241917202322504, |
| "learning_rate": 1e-05, |
| "loss": 0.0992, |
| "step": 21280 |
| }, |
| { |
| "epoch": 6.79, |
| "grad_norm": 4.867214735515191, |
| "learning_rate": 1e-05, |
| "loss": 0.0803, |
| "step": 21290 |
| }, |
| { |
| "epoch": 6.8, |
| "grad_norm": 4.3226852661460615, |
| "learning_rate": 1e-05, |
| "loss": 0.1038, |
| "step": 21300 |
| }, |
| { |
| "epoch": 6.8, |
| "grad_norm": 5.527159649374428, |
| "learning_rate": 1e-05, |
| "loss": 0.1023, |
| "step": 21310 |
| }, |
| { |
| "epoch": 6.8, |
| "grad_norm": 3.906572130272395, |
| "learning_rate": 1e-05, |
| "loss": 0.0823, |
| "step": 21320 |
| }, |
| { |
| "epoch": 6.81, |
| "grad_norm": 8.172203159490131, |
| "learning_rate": 1e-05, |
| "loss": 0.094, |
| "step": 21330 |
| }, |
| { |
| "epoch": 6.81, |
| "grad_norm": 8.222535164636962, |
| "learning_rate": 1e-05, |
| "loss": 0.0956, |
| "step": 21340 |
| }, |
| { |
| "epoch": 6.81, |
| "grad_norm": 5.280458470004511, |
| "learning_rate": 1e-05, |
| "loss": 0.1094, |
| "step": 21350 |
| }, |
| { |
| "epoch": 6.82, |
| "grad_norm": 4.4470829178542965, |
| "learning_rate": 1e-05, |
| "loss": 0.0969, |
| "step": 21360 |
| }, |
| { |
| "epoch": 6.82, |
| "grad_norm": 6.14368997041979, |
| "learning_rate": 1e-05, |
| "loss": 0.0962, |
| "step": 21370 |
| }, |
| { |
| "epoch": 6.82, |
| "grad_norm": 5.121172894899957, |
| "learning_rate": 1e-05, |
| "loss": 0.1062, |
| "step": 21380 |
| }, |
| { |
| "epoch": 6.83, |
| "grad_norm": 4.175425415293828, |
| "learning_rate": 1e-05, |
| "loss": 0.0961, |
| "step": 21390 |
| }, |
| { |
| "epoch": 6.83, |
| "grad_norm": 6.776386820051003, |
| "learning_rate": 1e-05, |
| "loss": 0.0935, |
| "step": 21400 |
| }, |
| { |
| "epoch": 6.83, |
| "grad_norm": 5.85951562331254, |
| "learning_rate": 1e-05, |
| "loss": 0.0884, |
| "step": 21410 |
| }, |
| { |
| "epoch": 6.83, |
| "grad_norm": 5.214089877039866, |
| "learning_rate": 1e-05, |
| "loss": 0.1034, |
| "step": 21420 |
| }, |
| { |
| "epoch": 6.84, |
| "grad_norm": 2.815638507637029, |
| "learning_rate": 1e-05, |
| "loss": 0.0825, |
| "step": 21430 |
| }, |
| { |
| "epoch": 6.84, |
| "grad_norm": 8.37810993812172, |
| "learning_rate": 1e-05, |
| "loss": 0.0963, |
| "step": 21440 |
| }, |
| { |
| "epoch": 6.84, |
| "grad_norm": 4.152219266272565, |
| "learning_rate": 1e-05, |
| "loss": 0.0978, |
| "step": 21450 |
| }, |
| { |
| "epoch": 6.85, |
| "grad_norm": 6.383520080118903, |
| "learning_rate": 1e-05, |
| "loss": 0.083, |
| "step": 21460 |
| }, |
| { |
| "epoch": 6.85, |
| "grad_norm": 6.144757848054752, |
| "learning_rate": 1e-05, |
| "loss": 0.0967, |
| "step": 21470 |
| }, |
| { |
| "epoch": 6.85, |
| "grad_norm": 6.84478863364012, |
| "learning_rate": 1e-05, |
| "loss": 0.1066, |
| "step": 21480 |
| }, |
| { |
| "epoch": 6.86, |
| "grad_norm": 7.295160324918141, |
| "learning_rate": 1e-05, |
| "loss": 0.1029, |
| "step": 21490 |
| }, |
| { |
| "epoch": 6.86, |
| "grad_norm": 5.8686966866213615, |
| "learning_rate": 1e-05, |
| "loss": 0.1041, |
| "step": 21500 |
| }, |
| { |
| "epoch": 6.86, |
| "eval_loss": 1.048828125, |
| "eval_runtime": 54.8626, |
| "eval_samples_per_second": 46.188, |
| "eval_steps_per_second": 2.898, |
| "step": 21500 |
| }, |
| { |
| "epoch": 6.86, |
| "grad_norm": 5.571051549905467, |
| "learning_rate": 1e-05, |
| "loss": 0.0964, |
| "step": 21510 |
| }, |
| { |
| "epoch": 6.87, |
| "grad_norm": 5.449645060609094, |
| "learning_rate": 1e-05, |
| "loss": 0.1024, |
| "step": 21520 |
| }, |
| { |
| "epoch": 6.87, |
| "grad_norm": 3.834080263980149, |
| "learning_rate": 1e-05, |
| "loss": 0.0906, |
| "step": 21530 |
| }, |
| { |
| "epoch": 6.87, |
| "grad_norm": 4.652368553473449, |
| "learning_rate": 1e-05, |
| "loss": 0.1002, |
| "step": 21540 |
| }, |
| { |
| "epoch": 6.88, |
| "grad_norm": 7.338406050985502, |
| "learning_rate": 1e-05, |
| "loss": 0.1122, |
| "step": 21550 |
| }, |
| { |
| "epoch": 6.88, |
| "grad_norm": 4.402936406425084, |
| "learning_rate": 1e-05, |
| "loss": 0.1027, |
| "step": 21560 |
| }, |
| { |
| "epoch": 6.88, |
| "grad_norm": 5.562542861601895, |
| "learning_rate": 1e-05, |
| "loss": 0.0854, |
| "step": 21570 |
| }, |
| { |
| "epoch": 6.89, |
| "grad_norm": 3.8810274560984794, |
| "learning_rate": 1e-05, |
| "loss": 0.0835, |
| "step": 21580 |
| }, |
| { |
| "epoch": 6.89, |
| "grad_norm": 3.3601913524817255, |
| "learning_rate": 1e-05, |
| "loss": 0.0958, |
| "step": 21590 |
| }, |
| { |
| "epoch": 6.89, |
| "grad_norm": 5.030882827913829, |
| "learning_rate": 1e-05, |
| "loss": 0.1189, |
| "step": 21600 |
| }, |
| { |
| "epoch": 6.9, |
| "grad_norm": 4.707202449233652, |
| "learning_rate": 1e-05, |
| "loss": 0.0929, |
| "step": 21610 |
| }, |
| { |
| "epoch": 6.9, |
| "grad_norm": 4.961064374607924, |
| "learning_rate": 1e-05, |
| "loss": 0.105, |
| "step": 21620 |
| }, |
| { |
| "epoch": 6.9, |
| "grad_norm": 4.871809869297845, |
| "learning_rate": 1e-05, |
| "loss": 0.109, |
| "step": 21630 |
| }, |
| { |
| "epoch": 6.9, |
| "grad_norm": 7.229648042715025, |
| "learning_rate": 1e-05, |
| "loss": 0.1075, |
| "step": 21640 |
| }, |
| { |
| "epoch": 6.91, |
| "grad_norm": 5.260986277141993, |
| "learning_rate": 1e-05, |
| "loss": 0.1031, |
| "step": 21650 |
| }, |
| { |
| "epoch": 6.91, |
| "grad_norm": 4.934247454588717, |
| "learning_rate": 1e-05, |
| "loss": 0.0787, |
| "step": 21660 |
| }, |
| { |
| "epoch": 6.91, |
| "grad_norm": 5.073955996011681, |
| "learning_rate": 1e-05, |
| "loss": 0.0993, |
| "step": 21670 |
| }, |
| { |
| "epoch": 6.92, |
| "grad_norm": 5.6503529868799145, |
| "learning_rate": 1e-05, |
| "loss": 0.1136, |
| "step": 21680 |
| }, |
| { |
| "epoch": 6.92, |
| "grad_norm": 7.501774895777269, |
| "learning_rate": 1e-05, |
| "loss": 0.0964, |
| "step": 21690 |
| }, |
| { |
| "epoch": 6.92, |
| "grad_norm": 7.832613025851854, |
| "learning_rate": 1e-05, |
| "loss": 0.0958, |
| "step": 21700 |
| }, |
| { |
| "epoch": 6.93, |
| "grad_norm": 5.195454576887189, |
| "learning_rate": 1e-05, |
| "loss": 0.1196, |
| "step": 21710 |
| }, |
| { |
| "epoch": 6.93, |
| "grad_norm": 8.966363994874078, |
| "learning_rate": 1e-05, |
| "loss": 0.1151, |
| "step": 21720 |
| }, |
| { |
| "epoch": 6.93, |
| "grad_norm": 5.4313043835269035, |
| "learning_rate": 1e-05, |
| "loss": 0.1117, |
| "step": 21730 |
| }, |
| { |
| "epoch": 6.94, |
| "grad_norm": 6.86987030084553, |
| "learning_rate": 1e-05, |
| "loss": 0.1043, |
| "step": 21740 |
| }, |
| { |
| "epoch": 6.94, |
| "grad_norm": 3.7594756889018894, |
| "learning_rate": 1e-05, |
| "loss": 0.0995, |
| "step": 21750 |
| }, |
| { |
| "epoch": 6.94, |
| "grad_norm": 4.20703125, |
| "learning_rate": 1e-05, |
| "loss": 0.0811, |
| "step": 21760 |
| }, |
| { |
| "epoch": 6.95, |
| "grad_norm": 5.414241286899629, |
| "learning_rate": 1e-05, |
| "loss": 0.1251, |
| "step": 21770 |
| }, |
| { |
| "epoch": 6.95, |
| "grad_norm": 4.56401392236002, |
| "learning_rate": 1e-05, |
| "loss": 0.0854, |
| "step": 21780 |
| }, |
| { |
| "epoch": 6.95, |
| "grad_norm": 4.662288564842939, |
| "learning_rate": 1e-05, |
| "loss": 0.103, |
| "step": 21790 |
| }, |
| { |
| "epoch": 6.96, |
| "grad_norm": 6.148212278475054, |
| "learning_rate": 1e-05, |
| "loss": 0.1148, |
| "step": 21800 |
| }, |
| { |
| "epoch": 6.96, |
| "grad_norm": 5.259140413779204, |
| "learning_rate": 1e-05, |
| "loss": 0.1182, |
| "step": 21810 |
| }, |
| { |
| "epoch": 6.96, |
| "grad_norm": 6.82262906543379, |
| "learning_rate": 1e-05, |
| "loss": 0.1058, |
| "step": 21820 |
| }, |
| { |
| "epoch": 6.97, |
| "grad_norm": 6.31165783995938, |
| "learning_rate": 1e-05, |
| "loss": 0.1098, |
| "step": 21830 |
| }, |
| { |
| "epoch": 6.97, |
| "grad_norm": 4.813306864685942, |
| "learning_rate": 1e-05, |
| "loss": 0.0968, |
| "step": 21840 |
| }, |
| { |
| "epoch": 6.97, |
| "grad_norm": 3.413266855228287, |
| "learning_rate": 1e-05, |
| "loss": 0.0957, |
| "step": 21850 |
| }, |
| { |
| "epoch": 6.98, |
| "grad_norm": 5.991383086998027, |
| "learning_rate": 1e-05, |
| "loss": 0.0851, |
| "step": 21860 |
| }, |
| { |
| "epoch": 6.98, |
| "grad_norm": 5.167397949805615, |
| "learning_rate": 1e-05, |
| "loss": 0.1018, |
| "step": 21870 |
| }, |
| { |
| "epoch": 6.98, |
| "grad_norm": 6.070263797363347, |
| "learning_rate": 1e-05, |
| "loss": 0.1165, |
| "step": 21880 |
| }, |
| { |
| "epoch": 6.98, |
| "grad_norm": 6.181187039045973, |
| "learning_rate": 1e-05, |
| "loss": 0.0991, |
| "step": 21890 |
| }, |
| { |
| "epoch": 6.99, |
| "grad_norm": 3.6041807641137695, |
| "learning_rate": 1e-05, |
| "loss": 0.0935, |
| "step": 21900 |
| }, |
| { |
| "epoch": 6.99, |
| "grad_norm": 7.708620660598994, |
| "learning_rate": 1e-05, |
| "loss": 0.1027, |
| "step": 21910 |
| }, |
| { |
| "epoch": 6.99, |
| "grad_norm": 5.767085688955038, |
| "learning_rate": 1e-05, |
| "loss": 0.1074, |
| "step": 21920 |
| }, |
| { |
| "epoch": 7.0, |
| "grad_norm": 5.34729475272173, |
| "learning_rate": 1e-05, |
| "loss": 0.115, |
| "step": 21930 |
| }, |
| { |
| "epoch": 7.0, |
| "grad_norm": 2.3991358234330513, |
| "learning_rate": 1e-05, |
| "loss": 0.0851, |
| "step": 21940 |
| }, |
| { |
| "epoch": 7.0, |
| "grad_norm": 4.2241287416576965, |
| "learning_rate": 1e-05, |
| "loss": 0.055, |
| "step": 21950 |
| }, |
| { |
| "epoch": 7.01, |
| "grad_norm": 6.403793831741706, |
| "learning_rate": 1e-05, |
| "loss": 0.0736, |
| "step": 21960 |
| }, |
| { |
| "epoch": 7.01, |
| "grad_norm": 3.6835685073780513, |
| "learning_rate": 1e-05, |
| "loss": 0.0597, |
| "step": 21970 |
| }, |
| { |
| "epoch": 7.01, |
| "grad_norm": 5.23061844125091, |
| "learning_rate": 1e-05, |
| "loss": 0.0616, |
| "step": 21980 |
| }, |
| { |
| "epoch": 7.02, |
| "grad_norm": 3.179154934909384, |
| "learning_rate": 1e-05, |
| "loss": 0.068, |
| "step": 21990 |
| }, |
| { |
| "epoch": 7.02, |
| "grad_norm": 2.3966525018067397, |
| "learning_rate": 1e-05, |
| "loss": 0.0566, |
| "step": 22000 |
| }, |
| { |
| "epoch": 7.02, |
| "eval_loss": 1.126953125, |
| "eval_runtime": 54.809, |
| "eval_samples_per_second": 46.233, |
| "eval_steps_per_second": 2.901, |
| "step": 22000 |
| }, |
| { |
| "epoch": 7.02, |
| "grad_norm": 6.093421027278012, |
| "learning_rate": 1e-05, |
| "loss": 0.064, |
| "step": 22010 |
| }, |
| { |
| "epoch": 7.03, |
| "grad_norm": 4.859789275636225, |
| "learning_rate": 1e-05, |
| "loss": 0.049, |
| "step": 22020 |
| }, |
| { |
| "epoch": 7.03, |
| "grad_norm": 1.5535580160003615, |
| "learning_rate": 1e-05, |
| "loss": 0.0513, |
| "step": 22030 |
| }, |
| { |
| "epoch": 7.03, |
| "grad_norm": 5.144397777311834, |
| "learning_rate": 1e-05, |
| "loss": 0.065, |
| "step": 22040 |
| }, |
| { |
| "epoch": 7.04, |
| "grad_norm": 3.078480501510632, |
| "learning_rate": 1e-05, |
| "loss": 0.0525, |
| "step": 22050 |
| }, |
| { |
| "epoch": 7.04, |
| "grad_norm": 4.458842786160914, |
| "learning_rate": 1e-05, |
| "loss": 0.0676, |
| "step": 22060 |
| }, |
| { |
| "epoch": 7.04, |
| "grad_norm": 3.8444926467132596, |
| "learning_rate": 1e-05, |
| "loss": 0.0632, |
| "step": 22070 |
| }, |
| { |
| "epoch": 7.05, |
| "grad_norm": 4.581979840730888, |
| "learning_rate": 1e-05, |
| "loss": 0.067, |
| "step": 22080 |
| }, |
| { |
| "epoch": 7.05, |
| "grad_norm": 4.180952048412888, |
| "learning_rate": 1e-05, |
| "loss": 0.0693, |
| "step": 22090 |
| }, |
| { |
| "epoch": 7.05, |
| "grad_norm": 5.4892049100424005, |
| "learning_rate": 1e-05, |
| "loss": 0.0617, |
| "step": 22100 |
| }, |
| { |
| "epoch": 7.05, |
| "grad_norm": 3.69180770386466, |
| "learning_rate": 1e-05, |
| "loss": 0.0648, |
| "step": 22110 |
| }, |
| { |
| "epoch": 7.06, |
| "grad_norm": 5.825349448531802, |
| "learning_rate": 1e-05, |
| "loss": 0.0747, |
| "step": 22120 |
| }, |
| { |
| "epoch": 7.06, |
| "grad_norm": 3.9133067362681375, |
| "learning_rate": 1e-05, |
| "loss": 0.0677, |
| "step": 22130 |
| }, |
| { |
| "epoch": 7.06, |
| "grad_norm": 5.294997701239154, |
| "learning_rate": 1e-05, |
| "loss": 0.0587, |
| "step": 22140 |
| }, |
| { |
| "epoch": 7.07, |
| "grad_norm": 7.253030866926722, |
| "learning_rate": 1e-05, |
| "loss": 0.0665, |
| "step": 22150 |
| }, |
| { |
| "epoch": 7.07, |
| "grad_norm": 3.2793295917162637, |
| "learning_rate": 1e-05, |
| "loss": 0.0529, |
| "step": 22160 |
| }, |
| { |
| "epoch": 7.07, |
| "grad_norm": 1.4419318173990863, |
| "learning_rate": 1e-05, |
| "loss": 0.0669, |
| "step": 22170 |
| }, |
| { |
| "epoch": 7.08, |
| "grad_norm": 2.859494232860285, |
| "learning_rate": 1e-05, |
| "loss": 0.0864, |
| "step": 22180 |
| }, |
| { |
| "epoch": 7.08, |
| "grad_norm": 4.570001157931592, |
| "learning_rate": 1e-05, |
| "loss": 0.0652, |
| "step": 22190 |
| }, |
| { |
| "epoch": 7.08, |
| "grad_norm": 6.5313454944985, |
| "learning_rate": 1e-05, |
| "loss": 0.0554, |
| "step": 22200 |
| }, |
| { |
| "epoch": 7.09, |
| "grad_norm": 6.609731678097479, |
| "learning_rate": 1e-05, |
| "loss": 0.0786, |
| "step": 22210 |
| }, |
| { |
| "epoch": 7.09, |
| "grad_norm": 1.997784161938792, |
| "learning_rate": 1e-05, |
| "loss": 0.0679, |
| "step": 22220 |
| }, |
| { |
| "epoch": 7.09, |
| "grad_norm": 5.048359844398007, |
| "learning_rate": 1e-05, |
| "loss": 0.0735, |
| "step": 22230 |
| }, |
| { |
| "epoch": 7.1, |
| "grad_norm": 4.952681560702609, |
| "learning_rate": 1e-05, |
| "loss": 0.0636, |
| "step": 22240 |
| }, |
| { |
| "epoch": 7.1, |
| "grad_norm": 2.9666862944335852, |
| "learning_rate": 1e-05, |
| "loss": 0.0641, |
| "step": 22250 |
| }, |
| { |
| "epoch": 7.1, |
| "grad_norm": 7.267202029790893, |
| "learning_rate": 1e-05, |
| "loss": 0.0789, |
| "step": 22260 |
| }, |
| { |
| "epoch": 7.11, |
| "grad_norm": 6.875389504236174, |
| "learning_rate": 1e-05, |
| "loss": 0.0608, |
| "step": 22270 |
| }, |
| { |
| "epoch": 7.11, |
| "grad_norm": 3.7681469841330184, |
| "learning_rate": 1e-05, |
| "loss": 0.074, |
| "step": 22280 |
| }, |
| { |
| "epoch": 7.11, |
| "grad_norm": 6.172198864129329, |
| "learning_rate": 1e-05, |
| "loss": 0.0914, |
| "step": 22290 |
| }, |
| { |
| "epoch": 7.12, |
| "grad_norm": 2.17630004176714, |
| "learning_rate": 1e-05, |
| "loss": 0.0731, |
| "step": 22300 |
| }, |
| { |
| "epoch": 7.12, |
| "grad_norm": 4.751078633763506, |
| "learning_rate": 1e-05, |
| "loss": 0.0752, |
| "step": 22310 |
| }, |
| { |
| "epoch": 7.12, |
| "grad_norm": 3.856148425082926, |
| "learning_rate": 1e-05, |
| "loss": 0.0726, |
| "step": 22320 |
| }, |
| { |
| "epoch": 7.13, |
| "grad_norm": 6.556389861528784, |
| "learning_rate": 1e-05, |
| "loss": 0.0706, |
| "step": 22330 |
| }, |
| { |
| "epoch": 7.13, |
| "grad_norm": 4.327210036975506, |
| "learning_rate": 1e-05, |
| "loss": 0.0656, |
| "step": 22340 |
| }, |
| { |
| "epoch": 7.13, |
| "grad_norm": 2.9971304521092255, |
| "learning_rate": 1e-05, |
| "loss": 0.0964, |
| "step": 22350 |
| }, |
| { |
| "epoch": 7.13, |
| "grad_norm": 8.64756410732146, |
| "learning_rate": 1e-05, |
| "loss": 0.0635, |
| "step": 22360 |
| }, |
| { |
| "epoch": 7.14, |
| "grad_norm": 3.613161817071748, |
| "learning_rate": 1e-05, |
| "loss": 0.0702, |
| "step": 22370 |
| }, |
| { |
| "epoch": 7.14, |
| "grad_norm": 3.0409000597722624, |
| "learning_rate": 1e-05, |
| "loss": 0.064, |
| "step": 22380 |
| }, |
| { |
| "epoch": 7.14, |
| "grad_norm": 3.6304731158432837, |
| "learning_rate": 1e-05, |
| "loss": 0.0696, |
| "step": 22390 |
| }, |
| { |
| "epoch": 7.15, |
| "grad_norm": 4.349008574357377, |
| "learning_rate": 1e-05, |
| "loss": 0.066, |
| "step": 22400 |
| }, |
| { |
| "epoch": 7.15, |
| "grad_norm": 3.4910669769269878, |
| "learning_rate": 1e-05, |
| "loss": 0.0604, |
| "step": 22410 |
| }, |
| { |
| "epoch": 7.15, |
| "grad_norm": 5.565360544594843, |
| "learning_rate": 1e-05, |
| "loss": 0.0619, |
| "step": 22420 |
| }, |
| { |
| "epoch": 7.16, |
| "grad_norm": 4.814619117954873, |
| "learning_rate": 1e-05, |
| "loss": 0.0533, |
| "step": 22430 |
| }, |
| { |
| "epoch": 7.16, |
| "grad_norm": 4.831286928891951, |
| "learning_rate": 1e-05, |
| "loss": 0.0716, |
| "step": 22440 |
| }, |
| { |
| "epoch": 7.16, |
| "grad_norm": 5.838938318187113, |
| "learning_rate": 1e-05, |
| "loss": 0.0822, |
| "step": 22450 |
| }, |
| { |
| "epoch": 7.17, |
| "grad_norm": 4.901321949040334, |
| "learning_rate": 1e-05, |
| "loss": 0.0883, |
| "step": 22460 |
| }, |
| { |
| "epoch": 7.17, |
| "grad_norm": 4.108754158261629, |
| "learning_rate": 1e-05, |
| "loss": 0.0754, |
| "step": 22470 |
| }, |
| { |
| "epoch": 7.17, |
| "grad_norm": 4.304805802452205, |
| "learning_rate": 1e-05, |
| "loss": 0.072, |
| "step": 22480 |
| }, |
| { |
| "epoch": 7.18, |
| "grad_norm": 3.637084286979847, |
| "learning_rate": 1e-05, |
| "loss": 0.0719, |
| "step": 22490 |
| }, |
| { |
| "epoch": 7.18, |
| "grad_norm": 1.8769301334683801, |
| "learning_rate": 1e-05, |
| "loss": 0.0635, |
| "step": 22500 |
| }, |
| { |
| "epoch": 7.18, |
| "eval_loss": 1.099609375, |
| "eval_runtime": 55.1145, |
| "eval_samples_per_second": 45.977, |
| "eval_steps_per_second": 2.885, |
| "step": 22500 |
| }, |
| { |
| "epoch": 7.18, |
| "grad_norm": 1.8440383427040754, |
| "learning_rate": 1e-05, |
| "loss": 0.0751, |
| "step": 22510 |
| }, |
| { |
| "epoch": 7.19, |
| "grad_norm": 4.246586719241477, |
| "learning_rate": 1e-05, |
| "loss": 0.06, |
| "step": 22520 |
| }, |
| { |
| "epoch": 7.19, |
| "grad_norm": 4.401052019757916, |
| "learning_rate": 1e-05, |
| "loss": 0.0603, |
| "step": 22530 |
| }, |
| { |
| "epoch": 7.19, |
| "grad_norm": 5.552557471131195, |
| "learning_rate": 1e-05, |
| "loss": 0.0743, |
| "step": 22540 |
| }, |
| { |
| "epoch": 7.2, |
| "grad_norm": 6.234863873963526, |
| "learning_rate": 1e-05, |
| "loss": 0.0711, |
| "step": 22550 |
| }, |
| { |
| "epoch": 7.2, |
| "grad_norm": 6.287290486964991, |
| "learning_rate": 1e-05, |
| "loss": 0.0759, |
| "step": 22560 |
| }, |
| { |
| "epoch": 7.2, |
| "grad_norm": 3.634999000727583, |
| "learning_rate": 1e-05, |
| "loss": 0.0922, |
| "step": 22570 |
| }, |
| { |
| "epoch": 7.2, |
| "grad_norm": 3.912961032017306, |
| "learning_rate": 1e-05, |
| "loss": 0.082, |
| "step": 22580 |
| }, |
| { |
| "epoch": 7.21, |
| "grad_norm": 3.486127609342229, |
| "learning_rate": 1e-05, |
| "loss": 0.0723, |
| "step": 22590 |
| }, |
| { |
| "epoch": 7.21, |
| "grad_norm": 8.03375703741596, |
| "learning_rate": 1e-05, |
| "loss": 0.0571, |
| "step": 22600 |
| }, |
| { |
| "epoch": 7.21, |
| "grad_norm": 5.059596983650268, |
| "learning_rate": 1e-05, |
| "loss": 0.0719, |
| "step": 22610 |
| }, |
| { |
| "epoch": 7.22, |
| "grad_norm": 4.8742127394176755, |
| "learning_rate": 1e-05, |
| "loss": 0.0785, |
| "step": 22620 |
| }, |
| { |
| "epoch": 7.22, |
| "grad_norm": 2.4039070315355104, |
| "learning_rate": 1e-05, |
| "loss": 0.0962, |
| "step": 22630 |
| }, |
| { |
| "epoch": 7.22, |
| "grad_norm": 3.277311740941721, |
| "learning_rate": 1e-05, |
| "loss": 0.0855, |
| "step": 22640 |
| }, |
| { |
| "epoch": 7.23, |
| "grad_norm": 4.337743324688415, |
| "learning_rate": 1e-05, |
| "loss": 0.0688, |
| "step": 22650 |
| }, |
| { |
| "epoch": 7.23, |
| "grad_norm": 3.489937349647952, |
| "learning_rate": 1e-05, |
| "loss": 0.071, |
| "step": 22660 |
| }, |
| { |
| "epoch": 7.23, |
| "grad_norm": 6.691509150277832, |
| "learning_rate": 1e-05, |
| "loss": 0.0697, |
| "step": 22670 |
| }, |
| { |
| "epoch": 7.24, |
| "grad_norm": 4.201684414196753, |
| "learning_rate": 1e-05, |
| "loss": 0.0822, |
| "step": 22680 |
| }, |
| { |
| "epoch": 7.24, |
| "grad_norm": 4.651401120001412, |
| "learning_rate": 1e-05, |
| "loss": 0.0716, |
| "step": 22690 |
| }, |
| { |
| "epoch": 7.24, |
| "grad_norm": 5.369002789018304, |
| "learning_rate": 1e-05, |
| "loss": 0.0661, |
| "step": 22700 |
| }, |
| { |
| "epoch": 7.25, |
| "grad_norm": 4.41101703741, |
| "learning_rate": 1e-05, |
| "loss": 0.0699, |
| "step": 22710 |
| }, |
| { |
| "epoch": 7.25, |
| "grad_norm": 5.9253293231592306, |
| "learning_rate": 1e-05, |
| "loss": 0.0833, |
| "step": 22720 |
| }, |
| { |
| "epoch": 7.25, |
| "grad_norm": 4.3713470467410565, |
| "learning_rate": 1e-05, |
| "loss": 0.0938, |
| "step": 22730 |
| }, |
| { |
| "epoch": 7.26, |
| "grad_norm": 4.208813303150469, |
| "learning_rate": 1e-05, |
| "loss": 0.0643, |
| "step": 22740 |
| }, |
| { |
| "epoch": 7.26, |
| "grad_norm": 3.9615695907987907, |
| "learning_rate": 1e-05, |
| "loss": 0.0574, |
| "step": 22750 |
| }, |
| { |
| "epoch": 7.26, |
| "grad_norm": 6.135974005971108, |
| "learning_rate": 1e-05, |
| "loss": 0.0713, |
| "step": 22760 |
| }, |
| { |
| "epoch": 7.27, |
| "grad_norm": 4.1514573537241155, |
| "learning_rate": 1e-05, |
| "loss": 0.0889, |
| "step": 22770 |
| }, |
| { |
| "epoch": 7.27, |
| "grad_norm": 3.3414616820772505, |
| "learning_rate": 1e-05, |
| "loss": 0.0852, |
| "step": 22780 |
| }, |
| { |
| "epoch": 7.27, |
| "grad_norm": 3.462400932134196, |
| "learning_rate": 1e-05, |
| "loss": 0.0704, |
| "step": 22790 |
| }, |
| { |
| "epoch": 7.28, |
| "grad_norm": 6.335708708107119, |
| "learning_rate": 1e-05, |
| "loss": 0.0791, |
| "step": 22800 |
| }, |
| { |
| "epoch": 7.28, |
| "grad_norm": 4.694130848800359, |
| "learning_rate": 1e-05, |
| "loss": 0.085, |
| "step": 22810 |
| }, |
| { |
| "epoch": 7.28, |
| "grad_norm": 5.197647648375192, |
| "learning_rate": 1e-05, |
| "loss": 0.0734, |
| "step": 22820 |
| }, |
| { |
| "epoch": 7.28, |
| "grad_norm": 4.1780242748002685, |
| "learning_rate": 1e-05, |
| "loss": 0.074, |
| "step": 22830 |
| }, |
| { |
| "epoch": 7.29, |
| "grad_norm": 2.8413667858069322, |
| "learning_rate": 1e-05, |
| "loss": 0.0645, |
| "step": 22840 |
| }, |
| { |
| "epoch": 7.29, |
| "grad_norm": 4.431916283364232, |
| "learning_rate": 1e-05, |
| "loss": 0.075, |
| "step": 22850 |
| }, |
| { |
| "epoch": 7.29, |
| "grad_norm": 5.868808161860691, |
| "learning_rate": 1e-05, |
| "loss": 0.0617, |
| "step": 22860 |
| }, |
| { |
| "epoch": 7.3, |
| "grad_norm": 5.0906915200174465, |
| "learning_rate": 1e-05, |
| "loss": 0.058, |
| "step": 22870 |
| }, |
| { |
| "epoch": 7.3, |
| "grad_norm": 5.197899379459396, |
| "learning_rate": 1e-05, |
| "loss": 0.0719, |
| "step": 22880 |
| }, |
| { |
| "epoch": 7.3, |
| "grad_norm": 3.6465353734884487, |
| "learning_rate": 1e-05, |
| "loss": 0.0593, |
| "step": 22890 |
| }, |
| { |
| "epoch": 7.31, |
| "grad_norm": 4.336593110977863, |
| "learning_rate": 1e-05, |
| "loss": 0.0805, |
| "step": 22900 |
| }, |
| { |
| "epoch": 7.31, |
| "grad_norm": 3.773088711558869, |
| "learning_rate": 1e-05, |
| "loss": 0.0701, |
| "step": 22910 |
| }, |
| { |
| "epoch": 7.31, |
| "grad_norm": 3.318447316190515, |
| "learning_rate": 1e-05, |
| "loss": 0.0739, |
| "step": 22920 |
| }, |
| { |
| "epoch": 7.32, |
| "grad_norm": 4.760211008027101, |
| "learning_rate": 1e-05, |
| "loss": 0.0851, |
| "step": 22930 |
| }, |
| { |
| "epoch": 7.32, |
| "grad_norm": 4.682746409680747, |
| "learning_rate": 1e-05, |
| "loss": 0.0515, |
| "step": 22940 |
| }, |
| { |
| "epoch": 7.32, |
| "grad_norm": 4.4895706117572205, |
| "learning_rate": 1e-05, |
| "loss": 0.0672, |
| "step": 22950 |
| }, |
| { |
| "epoch": 7.33, |
| "grad_norm": 2.791943768844984, |
| "learning_rate": 1e-05, |
| "loss": 0.0673, |
| "step": 22960 |
| }, |
| { |
| "epoch": 7.33, |
| "grad_norm": 11.311482589240864, |
| "learning_rate": 1e-05, |
| "loss": 0.0821, |
| "step": 22970 |
| }, |
| { |
| "epoch": 7.33, |
| "grad_norm": 7.406839725499553, |
| "learning_rate": 1e-05, |
| "loss": 0.0758, |
| "step": 22980 |
| }, |
| { |
| "epoch": 7.34, |
| "grad_norm": 4.730211699831174, |
| "learning_rate": 1e-05, |
| "loss": 0.0661, |
| "step": 22990 |
| }, |
| { |
| "epoch": 7.34, |
| "grad_norm": 5.811439991760393, |
| "learning_rate": 1e-05, |
| "loss": 0.0674, |
| "step": 23000 |
| }, |
| { |
| "epoch": 7.34, |
| "eval_loss": 1.15625, |
| "eval_runtime": 54.9155, |
| "eval_samples_per_second": 46.144, |
| "eval_steps_per_second": 2.895, |
| "step": 23000 |
| }, |
| { |
| "epoch": 7.34, |
| "grad_norm": 3.625134695773386, |
| "learning_rate": 1e-05, |
| "loss": 0.0721, |
| "step": 23010 |
| }, |
| { |
| "epoch": 7.35, |
| "grad_norm": 5.73543993964452, |
| "learning_rate": 1e-05, |
| "loss": 0.0682, |
| "step": 23020 |
| }, |
| { |
| "epoch": 7.35, |
| "grad_norm": 4.285367892662643, |
| "learning_rate": 1e-05, |
| "loss": 0.0605, |
| "step": 23030 |
| }, |
| { |
| "epoch": 7.35, |
| "grad_norm": 6.4568938917445475, |
| "learning_rate": 1e-05, |
| "loss": 0.0852, |
| "step": 23040 |
| }, |
| { |
| "epoch": 7.35, |
| "grad_norm": 2.864001590728318, |
| "learning_rate": 1e-05, |
| "loss": 0.0694, |
| "step": 23050 |
| }, |
| { |
| "epoch": 7.36, |
| "grad_norm": 4.815812593305297, |
| "learning_rate": 1e-05, |
| "loss": 0.0704, |
| "step": 23060 |
| }, |
| { |
| "epoch": 7.36, |
| "grad_norm": 5.602481442122426, |
| "learning_rate": 1e-05, |
| "loss": 0.088, |
| "step": 23070 |
| }, |
| { |
| "epoch": 7.36, |
| "grad_norm": 5.384720816994837, |
| "learning_rate": 1e-05, |
| "loss": 0.0645, |
| "step": 23080 |
| }, |
| { |
| "epoch": 7.37, |
| "grad_norm": 3.585052979528731, |
| "learning_rate": 1e-05, |
| "loss": 0.0603, |
| "step": 23090 |
| }, |
| { |
| "epoch": 7.37, |
| "grad_norm": 5.3153781050856965, |
| "learning_rate": 1e-05, |
| "loss": 0.0797, |
| "step": 23100 |
| }, |
| { |
| "epoch": 7.37, |
| "grad_norm": 7.373240891487032, |
| "learning_rate": 1e-05, |
| "loss": 0.0709, |
| "step": 23110 |
| }, |
| { |
| "epoch": 7.38, |
| "grad_norm": 5.8145013460588615, |
| "learning_rate": 1e-05, |
| "loss": 0.0769, |
| "step": 23120 |
| }, |
| { |
| "epoch": 7.38, |
| "grad_norm": 4.962284127639986, |
| "learning_rate": 1e-05, |
| "loss": 0.0796, |
| "step": 23130 |
| }, |
| { |
| "epoch": 7.38, |
| "grad_norm": 5.392880352190469, |
| "learning_rate": 1e-05, |
| "loss": 0.0896, |
| "step": 23140 |
| }, |
| { |
| "epoch": 7.39, |
| "grad_norm": 3.229568965011905, |
| "learning_rate": 1e-05, |
| "loss": 0.0826, |
| "step": 23150 |
| }, |
| { |
| "epoch": 7.39, |
| "grad_norm": 7.49637312614485, |
| "learning_rate": 1e-05, |
| "loss": 0.0905, |
| "step": 23160 |
| }, |
| { |
| "epoch": 7.39, |
| "grad_norm": 5.081016116607158, |
| "learning_rate": 1e-05, |
| "loss": 0.0648, |
| "step": 23170 |
| }, |
| { |
| "epoch": 7.4, |
| "grad_norm": 7.507726631110144, |
| "learning_rate": 1e-05, |
| "loss": 0.0707, |
| "step": 23180 |
| }, |
| { |
| "epoch": 7.4, |
| "grad_norm": 4.000419594691755, |
| "learning_rate": 1e-05, |
| "loss": 0.06, |
| "step": 23190 |
| }, |
| { |
| "epoch": 7.4, |
| "grad_norm": 6.772937655017929, |
| "learning_rate": 1e-05, |
| "loss": 0.0898, |
| "step": 23200 |
| }, |
| { |
| "epoch": 7.41, |
| "grad_norm": 3.619521638234594, |
| "learning_rate": 1e-05, |
| "loss": 0.0915, |
| "step": 23210 |
| }, |
| { |
| "epoch": 7.41, |
| "grad_norm": 6.807695286835935, |
| "learning_rate": 1e-05, |
| "loss": 0.0761, |
| "step": 23220 |
| }, |
| { |
| "epoch": 7.41, |
| "grad_norm": 6.216067861945093, |
| "learning_rate": 1e-05, |
| "loss": 0.0957, |
| "step": 23230 |
| }, |
| { |
| "epoch": 7.42, |
| "grad_norm": 4.718941185722348, |
| "learning_rate": 1e-05, |
| "loss": 0.1101, |
| "step": 23240 |
| }, |
| { |
| "epoch": 7.42, |
| "grad_norm": 4.514056609860622, |
| "learning_rate": 1e-05, |
| "loss": 0.0743, |
| "step": 23250 |
| }, |
| { |
| "epoch": 7.42, |
| "grad_norm": 3.3069937664442715, |
| "learning_rate": 1e-05, |
| "loss": 0.0689, |
| "step": 23260 |
| }, |
| { |
| "epoch": 7.43, |
| "grad_norm": 6.393837119186682, |
| "learning_rate": 1e-05, |
| "loss": 0.0855, |
| "step": 23270 |
| }, |
| { |
| "epoch": 7.43, |
| "grad_norm": 4.40384906259991, |
| "learning_rate": 1e-05, |
| "loss": 0.078, |
| "step": 23280 |
| }, |
| { |
| "epoch": 7.43, |
| "grad_norm": 3.390747068115173, |
| "learning_rate": 1e-05, |
| "loss": 0.078, |
| "step": 23290 |
| }, |
| { |
| "epoch": 7.43, |
| "grad_norm": 5.385176672344037, |
| "learning_rate": 1e-05, |
| "loss": 0.0731, |
| "step": 23300 |
| }, |
| { |
| "epoch": 7.44, |
| "grad_norm": 2.8458795643333086, |
| "learning_rate": 1e-05, |
| "loss": 0.0779, |
| "step": 23310 |
| }, |
| { |
| "epoch": 7.44, |
| "grad_norm": 4.68981164517424, |
| "learning_rate": 1e-05, |
| "loss": 0.0819, |
| "step": 23320 |
| }, |
| { |
| "epoch": 7.44, |
| "grad_norm": 5.192863908412107, |
| "learning_rate": 1e-05, |
| "loss": 0.0795, |
| "step": 23330 |
| }, |
| { |
| "epoch": 7.45, |
| "grad_norm": 3.748136820466799, |
| "learning_rate": 1e-05, |
| "loss": 0.0853, |
| "step": 23340 |
| }, |
| { |
| "epoch": 7.45, |
| "grad_norm": 6.541125265018704, |
| "learning_rate": 1e-05, |
| "loss": 0.0799, |
| "step": 23350 |
| }, |
| { |
| "epoch": 7.45, |
| "grad_norm": 3.8935804057653245, |
| "learning_rate": 1e-05, |
| "loss": 0.0778, |
| "step": 23360 |
| }, |
| { |
| "epoch": 7.46, |
| "grad_norm": 7.4133907612742025, |
| "learning_rate": 1e-05, |
| "loss": 0.0877, |
| "step": 23370 |
| }, |
| { |
| "epoch": 7.46, |
| "grad_norm": 3.734228410596996, |
| "learning_rate": 1e-05, |
| "loss": 0.075, |
| "step": 23380 |
| }, |
| { |
| "epoch": 7.46, |
| "grad_norm": 5.571012006284848, |
| "learning_rate": 1e-05, |
| "loss": 0.0722, |
| "step": 23390 |
| }, |
| { |
| "epoch": 7.47, |
| "grad_norm": 3.2396989578923154, |
| "learning_rate": 1e-05, |
| "loss": 0.0647, |
| "step": 23400 |
| }, |
| { |
| "epoch": 7.47, |
| "grad_norm": 5.136342592572217, |
| "learning_rate": 1e-05, |
| "loss": 0.0729, |
| "step": 23410 |
| }, |
| { |
| "epoch": 7.47, |
| "grad_norm": 5.376350942771395, |
| "learning_rate": 1e-05, |
| "loss": 0.0778, |
| "step": 23420 |
| }, |
| { |
| "epoch": 7.48, |
| "grad_norm": 3.9241884255077712, |
| "learning_rate": 1e-05, |
| "loss": 0.0875, |
| "step": 23430 |
| }, |
| { |
| "epoch": 7.48, |
| "grad_norm": 5.673048502267721, |
| "learning_rate": 1e-05, |
| "loss": 0.0866, |
| "step": 23440 |
| }, |
| { |
| "epoch": 7.48, |
| "grad_norm": 4.960027175644291, |
| "learning_rate": 1e-05, |
| "loss": 0.0605, |
| "step": 23450 |
| }, |
| { |
| "epoch": 7.49, |
| "grad_norm": 5.847147397699984, |
| "learning_rate": 1e-05, |
| "loss": 0.086, |
| "step": 23460 |
| }, |
| { |
| "epoch": 7.49, |
| "grad_norm": 5.841710025265471, |
| "learning_rate": 1e-05, |
| "loss": 0.0689, |
| "step": 23470 |
| }, |
| { |
| "epoch": 7.49, |
| "grad_norm": 4.117259537068444, |
| "learning_rate": 1e-05, |
| "loss": 0.0903, |
| "step": 23480 |
| }, |
| { |
| "epoch": 7.5, |
| "grad_norm": 6.686327109470771, |
| "learning_rate": 1e-05, |
| "loss": 0.0831, |
| "step": 23490 |
| }, |
| { |
| "epoch": 7.5, |
| "grad_norm": 3.0951647365526322, |
| "learning_rate": 1e-05, |
| "loss": 0.0856, |
| "step": 23500 |
| }, |
| { |
| "epoch": 7.5, |
| "eval_loss": 1.0693359375, |
| "eval_runtime": 54.6809, |
| "eval_samples_per_second": 46.342, |
| "eval_steps_per_second": 2.908, |
| "step": 23500 |
| }, |
| { |
| "epoch": 7.5, |
| "grad_norm": 4.606971243263287, |
| "learning_rate": 1e-05, |
| "loss": 0.0756, |
| "step": 23510 |
| }, |
| { |
| "epoch": 7.5, |
| "grad_norm": 6.084116985412715, |
| "learning_rate": 1e-05, |
| "loss": 0.061, |
| "step": 23520 |
| }, |
| { |
| "epoch": 7.51, |
| "grad_norm": 4.642995928691695, |
| "learning_rate": 1e-05, |
| "loss": 0.0777, |
| "step": 23530 |
| }, |
| { |
| "epoch": 7.51, |
| "grad_norm": 4.489693176389143, |
| "learning_rate": 1e-05, |
| "loss": 0.063, |
| "step": 23540 |
| }, |
| { |
| "epoch": 7.51, |
| "grad_norm": 4.296458165150986, |
| "learning_rate": 1e-05, |
| "loss": 0.0793, |
| "step": 23550 |
| }, |
| { |
| "epoch": 7.52, |
| "grad_norm": 6.794719205083293, |
| "learning_rate": 1e-05, |
| "loss": 0.0799, |
| "step": 23560 |
| }, |
| { |
| "epoch": 7.52, |
| "grad_norm": 2.3833417461042266, |
| "learning_rate": 1e-05, |
| "loss": 0.0751, |
| "step": 23570 |
| }, |
| { |
| "epoch": 7.52, |
| "grad_norm": 8.140233240933387, |
| "learning_rate": 1e-05, |
| "loss": 0.0824, |
| "step": 23580 |
| }, |
| { |
| "epoch": 7.53, |
| "grad_norm": 4.62344313528034, |
| "learning_rate": 1e-05, |
| "loss": 0.0906, |
| "step": 23590 |
| }, |
| { |
| "epoch": 7.53, |
| "grad_norm": 8.218438088187742, |
| "learning_rate": 1e-05, |
| "loss": 0.1042, |
| "step": 23600 |
| }, |
| { |
| "epoch": 7.53, |
| "grad_norm": 3.7272608200695045, |
| "learning_rate": 1e-05, |
| "loss": 0.0773, |
| "step": 23610 |
| }, |
| { |
| "epoch": 7.54, |
| "grad_norm": 4.830125314101774, |
| "learning_rate": 1e-05, |
| "loss": 0.0703, |
| "step": 23620 |
| }, |
| { |
| "epoch": 7.54, |
| "grad_norm": 7.192554587328428, |
| "learning_rate": 1e-05, |
| "loss": 0.0863, |
| "step": 23630 |
| }, |
| { |
| "epoch": 7.54, |
| "grad_norm": 5.731929413471994, |
| "learning_rate": 1e-05, |
| "loss": 0.0696, |
| "step": 23640 |
| }, |
| { |
| "epoch": 7.55, |
| "grad_norm": 4.412456064843017, |
| "learning_rate": 1e-05, |
| "loss": 0.0947, |
| "step": 23650 |
| }, |
| { |
| "epoch": 7.55, |
| "grad_norm": 4.990446118306138, |
| "learning_rate": 1e-05, |
| "loss": 0.0725, |
| "step": 23660 |
| }, |
| { |
| "epoch": 7.55, |
| "grad_norm": 5.741705509639433, |
| "learning_rate": 1e-05, |
| "loss": 0.0614, |
| "step": 23670 |
| }, |
| { |
| "epoch": 7.56, |
| "grad_norm": 6.819409427522705, |
| "learning_rate": 1e-05, |
| "loss": 0.0962, |
| "step": 23680 |
| }, |
| { |
| "epoch": 7.56, |
| "grad_norm": 4.59295232166117, |
| "learning_rate": 1e-05, |
| "loss": 0.0907, |
| "step": 23690 |
| }, |
| { |
| "epoch": 7.56, |
| "grad_norm": 3.7076428374172012, |
| "learning_rate": 1e-05, |
| "loss": 0.0751, |
| "step": 23700 |
| }, |
| { |
| "epoch": 7.57, |
| "grad_norm": 10.329686126224194, |
| "learning_rate": 1e-05, |
| "loss": 0.0792, |
| "step": 23710 |
| }, |
| { |
| "epoch": 7.57, |
| "grad_norm": 5.195531854802977, |
| "learning_rate": 1e-05, |
| "loss": 0.0891, |
| "step": 23720 |
| }, |
| { |
| "epoch": 7.57, |
| "grad_norm": 2.6851356219203284, |
| "learning_rate": 1e-05, |
| "loss": 0.086, |
| "step": 23730 |
| }, |
| { |
| "epoch": 7.57, |
| "grad_norm": 5.189599956463878, |
| "learning_rate": 1e-05, |
| "loss": 0.0765, |
| "step": 23740 |
| }, |
| { |
| "epoch": 7.58, |
| "grad_norm": 4.359096066975144, |
| "learning_rate": 1e-05, |
| "loss": 0.0734, |
| "step": 23750 |
| }, |
| { |
| "epoch": 7.58, |
| "grad_norm": 5.497258630218317, |
| "learning_rate": 1e-05, |
| "loss": 0.0758, |
| "step": 23760 |
| }, |
| { |
| "epoch": 7.58, |
| "grad_norm": 4.087457611144168, |
| "learning_rate": 1e-05, |
| "loss": 0.0754, |
| "step": 23770 |
| }, |
| { |
| "epoch": 7.59, |
| "grad_norm": 3.674639241683264, |
| "learning_rate": 1e-05, |
| "loss": 0.0733, |
| "step": 23780 |
| }, |
| { |
| "epoch": 7.59, |
| "grad_norm": 4.21942720982409, |
| "learning_rate": 1e-05, |
| "loss": 0.0705, |
| "step": 23790 |
| }, |
| { |
| "epoch": 7.59, |
| "grad_norm": 6.156709140754309, |
| "learning_rate": 1e-05, |
| "loss": 0.0849, |
| "step": 23800 |
| }, |
| { |
| "epoch": 7.6, |
| "grad_norm": 3.4904612940800805, |
| "learning_rate": 1e-05, |
| "loss": 0.0773, |
| "step": 23810 |
| }, |
| { |
| "epoch": 7.6, |
| "grad_norm": 6.42641346149058, |
| "learning_rate": 1e-05, |
| "loss": 0.0811, |
| "step": 23820 |
| }, |
| { |
| "epoch": 7.6, |
| "grad_norm": 9.746588868310344, |
| "learning_rate": 1e-05, |
| "loss": 0.0964, |
| "step": 23830 |
| }, |
| { |
| "epoch": 7.61, |
| "grad_norm": 5.592392399741415, |
| "learning_rate": 1e-05, |
| "loss": 0.0777, |
| "step": 23840 |
| }, |
| { |
| "epoch": 7.61, |
| "grad_norm": 5.125964190336803, |
| "learning_rate": 1e-05, |
| "loss": 0.0813, |
| "step": 23850 |
| }, |
| { |
| "epoch": 7.61, |
| "grad_norm": 3.906204955794978, |
| "learning_rate": 1e-05, |
| "loss": 0.0748, |
| "step": 23860 |
| }, |
| { |
| "epoch": 7.62, |
| "grad_norm": 4.04063753744188, |
| "learning_rate": 1e-05, |
| "loss": 0.0883, |
| "step": 23870 |
| }, |
| { |
| "epoch": 7.62, |
| "grad_norm": 5.048221184202787, |
| "learning_rate": 1e-05, |
| "loss": 0.0843, |
| "step": 23880 |
| }, |
| { |
| "epoch": 7.62, |
| "grad_norm": 9.790064730717598, |
| "learning_rate": 1e-05, |
| "loss": 0.0672, |
| "step": 23890 |
| }, |
| { |
| "epoch": 7.63, |
| "grad_norm": 3.1575267966395124, |
| "learning_rate": 1e-05, |
| "loss": 0.0737, |
| "step": 23900 |
| }, |
| { |
| "epoch": 7.63, |
| "grad_norm": 2.976003359044138, |
| "learning_rate": 1e-05, |
| "loss": 0.0888, |
| "step": 23910 |
| }, |
| { |
| "epoch": 7.63, |
| "grad_norm": 10.490453831080487, |
| "learning_rate": 1e-05, |
| "loss": 0.0809, |
| "step": 23920 |
| }, |
| { |
| "epoch": 7.64, |
| "grad_norm": 5.212174344581676, |
| "learning_rate": 1e-05, |
| "loss": 0.0684, |
| "step": 23930 |
| }, |
| { |
| "epoch": 7.64, |
| "grad_norm": 5.313081776795737, |
| "learning_rate": 1e-05, |
| "loss": 0.0712, |
| "step": 23940 |
| }, |
| { |
| "epoch": 7.64, |
| "grad_norm": 4.970319200239609, |
| "learning_rate": 1e-05, |
| "loss": 0.0857, |
| "step": 23950 |
| }, |
| { |
| "epoch": 7.65, |
| "grad_norm": 2.7754345218588266, |
| "learning_rate": 1e-05, |
| "loss": 0.0852, |
| "step": 23960 |
| }, |
| { |
| "epoch": 7.65, |
| "grad_norm": 5.700394988177594, |
| "learning_rate": 1e-05, |
| "loss": 0.0687, |
| "step": 23970 |
| }, |
| { |
| "epoch": 7.65, |
| "grad_norm": 5.411806107018785, |
| "learning_rate": 1e-05, |
| "loss": 0.0942, |
| "step": 23980 |
| }, |
| { |
| "epoch": 7.65, |
| "grad_norm": 4.394528428818539, |
| "learning_rate": 1e-05, |
| "loss": 0.0775, |
| "step": 23990 |
| }, |
| { |
| "epoch": 7.66, |
| "grad_norm": 4.780095578119951, |
| "learning_rate": 1e-05, |
| "loss": 0.0707, |
| "step": 24000 |
| }, |
| { |
| "epoch": 7.66, |
| "eval_loss": 1.10546875, |
| "eval_runtime": 54.7166, |
| "eval_samples_per_second": 46.311, |
| "eval_steps_per_second": 2.906, |
| "step": 24000 |
| }, |
| { |
| "epoch": 7.66, |
| "grad_norm": 3.8207336068875315, |
| "learning_rate": 1e-05, |
| "loss": 0.078, |
| "step": 24010 |
| }, |
| { |
| "epoch": 7.66, |
| "grad_norm": 5.126781991191461, |
| "learning_rate": 1e-05, |
| "loss": 0.0888, |
| "step": 24020 |
| }, |
| { |
| "epoch": 7.67, |
| "grad_norm": 4.780263362636217, |
| "learning_rate": 1e-05, |
| "loss": 0.0849, |
| "step": 24030 |
| }, |
| { |
| "epoch": 7.67, |
| "grad_norm": 2.8350724511417087, |
| "learning_rate": 1e-05, |
| "loss": 0.0731, |
| "step": 24040 |
| }, |
| { |
| "epoch": 7.67, |
| "grad_norm": 3.8174197862912673, |
| "learning_rate": 1e-05, |
| "loss": 0.0857, |
| "step": 24050 |
| }, |
| { |
| "epoch": 7.68, |
| "grad_norm": 3.5042514864846366, |
| "learning_rate": 1e-05, |
| "loss": 0.0783, |
| "step": 24060 |
| }, |
| { |
| "epoch": 7.68, |
| "grad_norm": 5.471257272619199, |
| "learning_rate": 1e-05, |
| "loss": 0.0801, |
| "step": 24070 |
| }, |
| { |
| "epoch": 7.68, |
| "grad_norm": 5.850624651041378, |
| "learning_rate": 1e-05, |
| "loss": 0.0851, |
| "step": 24080 |
| }, |
| { |
| "epoch": 7.69, |
| "grad_norm": 3.9046644731432623, |
| "learning_rate": 1e-05, |
| "loss": 0.0972, |
| "step": 24090 |
| }, |
| { |
| "epoch": 7.69, |
| "grad_norm": 4.901484027267218, |
| "learning_rate": 1e-05, |
| "loss": 0.0711, |
| "step": 24100 |
| }, |
| { |
| "epoch": 7.69, |
| "grad_norm": 4.883400355238245, |
| "learning_rate": 1e-05, |
| "loss": 0.0735, |
| "step": 24110 |
| }, |
| { |
| "epoch": 7.7, |
| "grad_norm": 4.2645512721141285, |
| "learning_rate": 1e-05, |
| "loss": 0.0811, |
| "step": 24120 |
| }, |
| { |
| "epoch": 7.7, |
| "grad_norm": 4.697765249221377, |
| "learning_rate": 1e-05, |
| "loss": 0.0776, |
| "step": 24130 |
| }, |
| { |
| "epoch": 7.7, |
| "grad_norm": 4.2160654074066795, |
| "learning_rate": 1e-05, |
| "loss": 0.0712, |
| "step": 24140 |
| }, |
| { |
| "epoch": 7.71, |
| "grad_norm": 4.2418621915008705, |
| "learning_rate": 1e-05, |
| "loss": 0.0865, |
| "step": 24150 |
| }, |
| { |
| "epoch": 7.71, |
| "grad_norm": 6.613978304688683, |
| "learning_rate": 1e-05, |
| "loss": 0.1041, |
| "step": 24160 |
| }, |
| { |
| "epoch": 7.71, |
| "grad_norm": 6.404773672305449, |
| "learning_rate": 1e-05, |
| "loss": 0.0988, |
| "step": 24170 |
| }, |
| { |
| "epoch": 7.72, |
| "grad_norm": 5.641920622496109, |
| "learning_rate": 1e-05, |
| "loss": 0.0946, |
| "step": 24180 |
| }, |
| { |
| "epoch": 7.72, |
| "grad_norm": 4.594788103167837, |
| "learning_rate": 1e-05, |
| "loss": 0.0835, |
| "step": 24190 |
| }, |
| { |
| "epoch": 7.72, |
| "grad_norm": 6.493859838854658, |
| "learning_rate": 1e-05, |
| "loss": 0.078, |
| "step": 24200 |
| }, |
| { |
| "epoch": 7.72, |
| "grad_norm": 6.623358757125114, |
| "learning_rate": 1e-05, |
| "loss": 0.0878, |
| "step": 24210 |
| }, |
| { |
| "epoch": 7.73, |
| "grad_norm": 4.186094361285578, |
| "learning_rate": 1e-05, |
| "loss": 0.0811, |
| "step": 24220 |
| }, |
| { |
| "epoch": 7.73, |
| "grad_norm": 7.157152144026571, |
| "learning_rate": 1e-05, |
| "loss": 0.0757, |
| "step": 24230 |
| }, |
| { |
| "epoch": 7.73, |
| "grad_norm": 4.731026148067277, |
| "learning_rate": 1e-05, |
| "loss": 0.0643, |
| "step": 24240 |
| }, |
| { |
| "epoch": 7.74, |
| "grad_norm": 4.854644337128586, |
| "learning_rate": 1e-05, |
| "loss": 0.0781, |
| "step": 24250 |
| }, |
| { |
| "epoch": 7.74, |
| "grad_norm": 6.759985425583517, |
| "learning_rate": 1e-05, |
| "loss": 0.0889, |
| "step": 24260 |
| }, |
| { |
| "epoch": 7.74, |
| "grad_norm": 3.507996824047778, |
| "learning_rate": 1e-05, |
| "loss": 0.0733, |
| "step": 24270 |
| }, |
| { |
| "epoch": 7.75, |
| "grad_norm": 3.0636351680830463, |
| "learning_rate": 1e-05, |
| "loss": 0.073, |
| "step": 24280 |
| }, |
| { |
| "epoch": 7.75, |
| "grad_norm": 3.73482724567733, |
| "learning_rate": 1e-05, |
| "loss": 0.0776, |
| "step": 24290 |
| }, |
| { |
| "epoch": 7.75, |
| "grad_norm": 6.593703138718255, |
| "learning_rate": 1e-05, |
| "loss": 0.1116, |
| "step": 24300 |
| }, |
| { |
| "epoch": 7.76, |
| "grad_norm": 5.862702505685071, |
| "learning_rate": 1e-05, |
| "loss": 0.077, |
| "step": 24310 |
| }, |
| { |
| "epoch": 7.76, |
| "grad_norm": 4.133679823555156, |
| "learning_rate": 1e-05, |
| "loss": 0.0758, |
| "step": 24320 |
| }, |
| { |
| "epoch": 7.76, |
| "grad_norm": 5.907885451170358, |
| "learning_rate": 1e-05, |
| "loss": 0.0952, |
| "step": 24330 |
| }, |
| { |
| "epoch": 7.77, |
| "grad_norm": 5.2450761866241, |
| "learning_rate": 1e-05, |
| "loss": 0.0696, |
| "step": 24340 |
| }, |
| { |
| "epoch": 7.77, |
| "grad_norm": 8.95913787304463, |
| "learning_rate": 1e-05, |
| "loss": 0.0782, |
| "step": 24350 |
| }, |
| { |
| "epoch": 7.77, |
| "grad_norm": 7.565375419826826, |
| "learning_rate": 1e-05, |
| "loss": 0.0734, |
| "step": 24360 |
| }, |
| { |
| "epoch": 7.78, |
| "grad_norm": 7.147864346899687, |
| "learning_rate": 1e-05, |
| "loss": 0.0838, |
| "step": 24370 |
| }, |
| { |
| "epoch": 7.78, |
| "grad_norm": 3.1076893286752476, |
| "learning_rate": 1e-05, |
| "loss": 0.0794, |
| "step": 24380 |
| }, |
| { |
| "epoch": 7.78, |
| "grad_norm": 5.22362998267707, |
| "learning_rate": 1e-05, |
| "loss": 0.0901, |
| "step": 24390 |
| }, |
| { |
| "epoch": 7.79, |
| "grad_norm": 4.401657633215462, |
| "learning_rate": 1e-05, |
| "loss": 0.0736, |
| "step": 24400 |
| }, |
| { |
| "epoch": 7.79, |
| "grad_norm": 3.395805093534723, |
| "learning_rate": 1e-05, |
| "loss": 0.0898, |
| "step": 24410 |
| }, |
| { |
| "epoch": 7.79, |
| "grad_norm": 4.4543575254300665, |
| "learning_rate": 1e-05, |
| "loss": 0.0875, |
| "step": 24420 |
| }, |
| { |
| "epoch": 7.8, |
| "grad_norm": 7.99135360767562, |
| "learning_rate": 1e-05, |
| "loss": 0.0903, |
| "step": 24430 |
| }, |
| { |
| "epoch": 7.8, |
| "grad_norm": 7.047029548523525, |
| "learning_rate": 1e-05, |
| "loss": 0.0928, |
| "step": 24440 |
| }, |
| { |
| "epoch": 7.8, |
| "grad_norm": 6.162713545064819, |
| "learning_rate": 1e-05, |
| "loss": 0.0942, |
| "step": 24450 |
| }, |
| { |
| "epoch": 7.8, |
| "grad_norm": 4.619851623398397, |
| "learning_rate": 1e-05, |
| "loss": 0.096, |
| "step": 24460 |
| }, |
| { |
| "epoch": 7.81, |
| "grad_norm": 4.176696734085503, |
| "learning_rate": 1e-05, |
| "loss": 0.0855, |
| "step": 24470 |
| }, |
| { |
| "epoch": 7.81, |
| "grad_norm": 5.181134650369433, |
| "learning_rate": 1e-05, |
| "loss": 0.1131, |
| "step": 24480 |
| }, |
| { |
| "epoch": 7.81, |
| "grad_norm": 5.860484920917448, |
| "learning_rate": 1e-05, |
| "loss": 0.0922, |
| "step": 24490 |
| }, |
| { |
| "epoch": 7.82, |
| "grad_norm": 7.1941566491547935, |
| "learning_rate": 1e-05, |
| "loss": 0.0778, |
| "step": 24500 |
| }, |
| { |
| "epoch": 7.82, |
| "eval_loss": 1.0830078125, |
| "eval_runtime": 54.8151, |
| "eval_samples_per_second": 46.228, |
| "eval_steps_per_second": 2.901, |
| "step": 24500 |
| }, |
| { |
| "epoch": 7.82, |
| "grad_norm": 4.7172875569697466, |
| "learning_rate": 1e-05, |
| "loss": 0.0847, |
| "step": 24510 |
| }, |
| { |
| "epoch": 7.82, |
| "grad_norm": 2.747549091903665, |
| "learning_rate": 1e-05, |
| "loss": 0.0853, |
| "step": 24520 |
| }, |
| { |
| "epoch": 7.83, |
| "grad_norm": 6.509808988566942, |
| "learning_rate": 1e-05, |
| "loss": 0.0964, |
| "step": 24530 |
| }, |
| { |
| "epoch": 7.83, |
| "grad_norm": 5.526905487262897, |
| "learning_rate": 1e-05, |
| "loss": 0.0907, |
| "step": 24540 |
| }, |
| { |
| "epoch": 7.83, |
| "grad_norm": 5.628814251774653, |
| "learning_rate": 1e-05, |
| "loss": 0.0986, |
| "step": 24550 |
| }, |
| { |
| "epoch": 7.84, |
| "grad_norm": 3.4663638594010466, |
| "learning_rate": 1e-05, |
| "loss": 0.0812, |
| "step": 24560 |
| }, |
| { |
| "epoch": 7.84, |
| "grad_norm": 7.24523874672673, |
| "learning_rate": 1e-05, |
| "loss": 0.0767, |
| "step": 24570 |
| }, |
| { |
| "epoch": 7.84, |
| "grad_norm": 4.16183898992695, |
| "learning_rate": 1e-05, |
| "loss": 0.0573, |
| "step": 24580 |
| }, |
| { |
| "epoch": 7.85, |
| "grad_norm": 5.867571740705368, |
| "learning_rate": 1e-05, |
| "loss": 0.1066, |
| "step": 24590 |
| }, |
| { |
| "epoch": 7.85, |
| "grad_norm": 7.395239262608121, |
| "learning_rate": 1e-05, |
| "loss": 0.0835, |
| "step": 24600 |
| }, |
| { |
| "epoch": 7.85, |
| "grad_norm": 5.573842004503008, |
| "learning_rate": 1e-05, |
| "loss": 0.0903, |
| "step": 24610 |
| }, |
| { |
| "epoch": 7.86, |
| "grad_norm": 4.932529506836774, |
| "learning_rate": 1e-05, |
| "loss": 0.0609, |
| "step": 24620 |
| }, |
| { |
| "epoch": 7.86, |
| "grad_norm": 6.498922405417734, |
| "learning_rate": 1e-05, |
| "loss": 0.0942, |
| "step": 24630 |
| }, |
| { |
| "epoch": 7.86, |
| "grad_norm": 7.01225951083284, |
| "learning_rate": 1e-05, |
| "loss": 0.0825, |
| "step": 24640 |
| }, |
| { |
| "epoch": 7.87, |
| "grad_norm": 3.785194420400613, |
| "learning_rate": 1e-05, |
| "loss": 0.0781, |
| "step": 24650 |
| }, |
| { |
| "epoch": 7.87, |
| "grad_norm": 3.407506028727672, |
| "learning_rate": 1e-05, |
| "loss": 0.0967, |
| "step": 24660 |
| }, |
| { |
| "epoch": 7.87, |
| "grad_norm": 3.9571657299149945, |
| "learning_rate": 1e-05, |
| "loss": 0.0672, |
| "step": 24670 |
| }, |
| { |
| "epoch": 7.87, |
| "grad_norm": 5.030512406203738, |
| "learning_rate": 1e-05, |
| "loss": 0.0774, |
| "step": 24680 |
| }, |
| { |
| "epoch": 7.88, |
| "grad_norm": 2.8735147039549345, |
| "learning_rate": 1e-05, |
| "loss": 0.0706, |
| "step": 24690 |
| }, |
| { |
| "epoch": 7.88, |
| "grad_norm": 6.201623384153294, |
| "learning_rate": 1e-05, |
| "loss": 0.0841, |
| "step": 24700 |
| }, |
| { |
| "epoch": 7.88, |
| "grad_norm": 5.645180170581378, |
| "learning_rate": 1e-05, |
| "loss": 0.0809, |
| "step": 24710 |
| }, |
| { |
| "epoch": 7.89, |
| "grad_norm": 4.3132834344853865, |
| "learning_rate": 1e-05, |
| "loss": 0.0707, |
| "step": 24720 |
| }, |
| { |
| "epoch": 7.89, |
| "grad_norm": 6.9367955855427565, |
| "learning_rate": 1e-05, |
| "loss": 0.0984, |
| "step": 24730 |
| }, |
| { |
| "epoch": 7.89, |
| "grad_norm": 6.809911113500623, |
| "learning_rate": 1e-05, |
| "loss": 0.0976, |
| "step": 24740 |
| }, |
| { |
| "epoch": 7.9, |
| "grad_norm": 6.1410471064964955, |
| "learning_rate": 1e-05, |
| "loss": 0.0958, |
| "step": 24750 |
| }, |
| { |
| "epoch": 7.9, |
| "grad_norm": 4.704553127491976, |
| "learning_rate": 1e-05, |
| "loss": 0.0845, |
| "step": 24760 |
| }, |
| { |
| "epoch": 7.9, |
| "grad_norm": 5.385367928831032, |
| "learning_rate": 1e-05, |
| "loss": 0.0863, |
| "step": 24770 |
| }, |
| { |
| "epoch": 7.91, |
| "grad_norm": 5.025296687387188, |
| "learning_rate": 1e-05, |
| "loss": 0.0766, |
| "step": 24780 |
| }, |
| { |
| "epoch": 7.91, |
| "grad_norm": 4.749472137032107, |
| "learning_rate": 1e-05, |
| "loss": 0.0824, |
| "step": 24790 |
| }, |
| { |
| "epoch": 7.91, |
| "grad_norm": 8.035680832566156, |
| "learning_rate": 1e-05, |
| "loss": 0.0915, |
| "step": 24800 |
| }, |
| { |
| "epoch": 7.92, |
| "grad_norm": 5.596393152535749, |
| "learning_rate": 1e-05, |
| "loss": 0.0771, |
| "step": 24810 |
| }, |
| { |
| "epoch": 7.92, |
| "grad_norm": 6.6928797678878365, |
| "learning_rate": 1e-05, |
| "loss": 0.081, |
| "step": 24820 |
| }, |
| { |
| "epoch": 7.92, |
| "grad_norm": 4.7432642912961285, |
| "learning_rate": 1e-05, |
| "loss": 0.0763, |
| "step": 24830 |
| }, |
| { |
| "epoch": 7.93, |
| "grad_norm": 4.61967615487048, |
| "learning_rate": 1e-05, |
| "loss": 0.0831, |
| "step": 24840 |
| }, |
| { |
| "epoch": 7.93, |
| "grad_norm": 4.397576852618112, |
| "learning_rate": 1e-05, |
| "loss": 0.0722, |
| "step": 24850 |
| }, |
| { |
| "epoch": 7.93, |
| "grad_norm": 6.741710234863936, |
| "learning_rate": 1e-05, |
| "loss": 0.0883, |
| "step": 24860 |
| }, |
| { |
| "epoch": 7.94, |
| "grad_norm": 5.570438335436572, |
| "learning_rate": 1e-05, |
| "loss": 0.0873, |
| "step": 24870 |
| }, |
| { |
| "epoch": 7.94, |
| "grad_norm": 5.493268488590753, |
| "learning_rate": 1e-05, |
| "loss": 0.0838, |
| "step": 24880 |
| }, |
| { |
| "epoch": 7.94, |
| "grad_norm": 6.031881190660355, |
| "learning_rate": 1e-05, |
| "loss": 0.0823, |
| "step": 24890 |
| }, |
| { |
| "epoch": 7.95, |
| "grad_norm": 3.033035224546501, |
| "learning_rate": 1e-05, |
| "loss": 0.0793, |
| "step": 24900 |
| }, |
| { |
| "epoch": 7.95, |
| "grad_norm": 4.934342932279823, |
| "learning_rate": 1e-05, |
| "loss": 0.0823, |
| "step": 24910 |
| }, |
| { |
| "epoch": 7.95, |
| "grad_norm": 6.143363361474989, |
| "learning_rate": 1e-05, |
| "loss": 0.0952, |
| "step": 24920 |
| }, |
| { |
| "epoch": 7.95, |
| "grad_norm": 6.589083064053754, |
| "learning_rate": 1e-05, |
| "loss": 0.0998, |
| "step": 24930 |
| }, |
| { |
| "epoch": 7.96, |
| "grad_norm": 4.264577883797825, |
| "learning_rate": 1e-05, |
| "loss": 0.0807, |
| "step": 24940 |
| }, |
| { |
| "epoch": 7.96, |
| "grad_norm": 7.486218060272307, |
| "learning_rate": 1e-05, |
| "loss": 0.0887, |
| "step": 24950 |
| }, |
| { |
| "epoch": 7.96, |
| "grad_norm": 7.956261996060158, |
| "learning_rate": 1e-05, |
| "loss": 0.0889, |
| "step": 24960 |
| }, |
| { |
| "epoch": 7.97, |
| "grad_norm": 3.4512346021287317, |
| "learning_rate": 1e-05, |
| "loss": 0.0725, |
| "step": 24970 |
| }, |
| { |
| "epoch": 7.97, |
| "grad_norm": 4.880450209815512, |
| "learning_rate": 1e-05, |
| "loss": 0.0802, |
| "step": 24980 |
| }, |
| { |
| "epoch": 7.97, |
| "grad_norm": 4.695055962534739, |
| "learning_rate": 1e-05, |
| "loss": 0.0832, |
| "step": 24990 |
| }, |
| { |
| "epoch": 7.98, |
| "grad_norm": 5.23515746687747, |
| "learning_rate": 1e-05, |
| "loss": 0.0953, |
| "step": 25000 |
| }, |
| { |
| "epoch": 7.98, |
| "eval_loss": 1.12109375, |
| "eval_runtime": 54.7577, |
| "eval_samples_per_second": 46.277, |
| "eval_steps_per_second": 2.904, |
| "step": 25000 |
| }, |
| { |
| "epoch": 7.98, |
| "grad_norm": 3.8134839398536196, |
| "learning_rate": 1e-05, |
| "loss": 0.1248, |
| "step": 25010 |
| }, |
| { |
| "epoch": 7.98, |
| "grad_norm": 3.7783105948198243, |
| "learning_rate": 1e-05, |
| "loss": 0.0877, |
| "step": 25020 |
| }, |
| { |
| "epoch": 7.99, |
| "grad_norm": 6.230651508389323, |
| "learning_rate": 1e-05, |
| "loss": 0.1062, |
| "step": 25030 |
| }, |
| { |
| "epoch": 7.99, |
| "grad_norm": 3.767387136124327, |
| "learning_rate": 1e-05, |
| "loss": 0.0914, |
| "step": 25040 |
| }, |
| { |
| "epoch": 7.99, |
| "grad_norm": 3.6706960530470214, |
| "learning_rate": 1e-05, |
| "loss": 0.0717, |
| "step": 25050 |
| }, |
| { |
| "epoch": 8.0, |
| "grad_norm": 6.369213694574724, |
| "learning_rate": 1e-05, |
| "loss": 0.0802, |
| "step": 25060 |
| }, |
| { |
| "epoch": 8.0, |
| "grad_norm": 5.410651735828217, |
| "learning_rate": 1e-05, |
| "loss": 0.0861, |
| "step": 25070 |
| }, |
| { |
| "epoch": 8.0, |
| "grad_norm": 3.933464525481039, |
| "learning_rate": 1e-05, |
| "loss": 0.0735, |
| "step": 25080 |
| }, |
| { |
| "epoch": 8.01, |
| "grad_norm": 4.86449474224097, |
| "learning_rate": 1e-05, |
| "loss": 0.0609, |
| "step": 25090 |
| }, |
| { |
| "epoch": 8.01, |
| "grad_norm": 1.901923037150177, |
| "learning_rate": 1e-05, |
| "loss": 0.066, |
| "step": 25100 |
| }, |
| { |
| "epoch": 8.01, |
| "grad_norm": 3.6889884660167733, |
| "learning_rate": 1e-05, |
| "loss": 0.0499, |
| "step": 25110 |
| }, |
| { |
| "epoch": 8.02, |
| "grad_norm": 5.971362116581421, |
| "learning_rate": 1e-05, |
| "loss": 0.0418, |
| "step": 25120 |
| }, |
| { |
| "epoch": 8.02, |
| "grad_norm": 6.028769818233255, |
| "learning_rate": 1e-05, |
| "loss": 0.0525, |
| "step": 25130 |
| }, |
| { |
| "epoch": 8.02, |
| "grad_norm": 4.250481073416132, |
| "learning_rate": 1e-05, |
| "loss": 0.0518, |
| "step": 25140 |
| }, |
| { |
| "epoch": 8.02, |
| "grad_norm": 3.0410164090096132, |
| "learning_rate": 1e-05, |
| "loss": 0.062, |
| "step": 25150 |
| }, |
| { |
| "epoch": 8.03, |
| "grad_norm": 3.736694764625156, |
| "learning_rate": 1e-05, |
| "loss": 0.0604, |
| "step": 25160 |
| }, |
| { |
| "epoch": 8.03, |
| "grad_norm": 3.5119585373327427, |
| "learning_rate": 1e-05, |
| "loss": 0.0683, |
| "step": 25170 |
| }, |
| { |
| "epoch": 8.03, |
| "grad_norm": 4.819902350042003, |
| "learning_rate": 1e-05, |
| "loss": 0.0635, |
| "step": 25180 |
| }, |
| { |
| "epoch": 8.04, |
| "grad_norm": 4.158250829047641, |
| "learning_rate": 1e-05, |
| "loss": 0.0539, |
| "step": 25190 |
| }, |
| { |
| "epoch": 8.04, |
| "grad_norm": 3.186988639952537, |
| "learning_rate": 1e-05, |
| "loss": 0.0475, |
| "step": 25200 |
| }, |
| { |
| "epoch": 8.04, |
| "grad_norm": 2.4459331078354065, |
| "learning_rate": 1e-05, |
| "loss": 0.0433, |
| "step": 25210 |
| }, |
| { |
| "epoch": 8.05, |
| "grad_norm": 7.7173752274728304, |
| "learning_rate": 1e-05, |
| "loss": 0.0726, |
| "step": 25220 |
| }, |
| { |
| "epoch": 8.05, |
| "grad_norm": 5.010782250954385, |
| "learning_rate": 1e-05, |
| "loss": 0.0715, |
| "step": 25230 |
| }, |
| { |
| "epoch": 8.05, |
| "grad_norm": 3.4027884156629575, |
| "learning_rate": 1e-05, |
| "loss": 0.0587, |
| "step": 25240 |
| }, |
| { |
| "epoch": 8.06, |
| "grad_norm": 7.578989510530571, |
| "learning_rate": 1e-05, |
| "loss": 0.0513, |
| "step": 25250 |
| }, |
| { |
| "epoch": 8.06, |
| "grad_norm": 4.273771785514313, |
| "learning_rate": 1e-05, |
| "loss": 0.0624, |
| "step": 25260 |
| }, |
| { |
| "epoch": 8.06, |
| "grad_norm": 3.0908520930435635, |
| "learning_rate": 1e-05, |
| "loss": 0.0607, |
| "step": 25270 |
| }, |
| { |
| "epoch": 8.07, |
| "grad_norm": 3.078159390175116, |
| "learning_rate": 1e-05, |
| "loss": 0.0545, |
| "step": 25280 |
| }, |
| { |
| "epoch": 8.07, |
| "grad_norm": 8.806752163368298, |
| "learning_rate": 1e-05, |
| "loss": 0.0713, |
| "step": 25290 |
| }, |
| { |
| "epoch": 8.07, |
| "grad_norm": 9.163331089604634, |
| "learning_rate": 1e-05, |
| "loss": 0.0614, |
| "step": 25300 |
| }, |
| { |
| "epoch": 8.08, |
| "grad_norm": 6.780802206557357, |
| "learning_rate": 1e-05, |
| "loss": 0.0652, |
| "step": 25310 |
| }, |
| { |
| "epoch": 8.08, |
| "grad_norm": 4.118618971458232, |
| "learning_rate": 1e-05, |
| "loss": 0.0537, |
| "step": 25320 |
| }, |
| { |
| "epoch": 8.08, |
| "grad_norm": 5.540234526479763, |
| "learning_rate": 1e-05, |
| "loss": 0.0502, |
| "step": 25330 |
| }, |
| { |
| "epoch": 8.09, |
| "grad_norm": 8.232922536303226, |
| "learning_rate": 1e-05, |
| "loss": 0.0586, |
| "step": 25340 |
| }, |
| { |
| "epoch": 8.09, |
| "grad_norm": 3.3409177960510648, |
| "learning_rate": 1e-05, |
| "loss": 0.0524, |
| "step": 25350 |
| }, |
| { |
| "epoch": 8.09, |
| "grad_norm": 5.242721598630065, |
| "learning_rate": 1e-05, |
| "loss": 0.0654, |
| "step": 25360 |
| }, |
| { |
| "epoch": 8.1, |
| "grad_norm": 3.3868219520481864, |
| "learning_rate": 1e-05, |
| "loss": 0.0621, |
| "step": 25370 |
| }, |
| { |
| "epoch": 8.1, |
| "grad_norm": 5.0286413028432575, |
| "learning_rate": 1e-05, |
| "loss": 0.0808, |
| "step": 25380 |
| }, |
| { |
| "epoch": 8.1, |
| "grad_norm": 3.926932550397459, |
| "learning_rate": 1e-05, |
| "loss": 0.0583, |
| "step": 25390 |
| }, |
| { |
| "epoch": 8.1, |
| "grad_norm": 2.2920047741785305, |
| "learning_rate": 1e-05, |
| "loss": 0.0508, |
| "step": 25400 |
| }, |
| { |
| "epoch": 8.11, |
| "grad_norm": 3.5608856826768216, |
| "learning_rate": 1e-05, |
| "loss": 0.0429, |
| "step": 25410 |
| }, |
| { |
| "epoch": 8.11, |
| "grad_norm": 3.2625598302741, |
| "learning_rate": 1e-05, |
| "loss": 0.0604, |
| "step": 25420 |
| }, |
| { |
| "epoch": 8.11, |
| "grad_norm": 7.146675468031365, |
| "learning_rate": 1e-05, |
| "loss": 0.0611, |
| "step": 25430 |
| }, |
| { |
| "epoch": 8.12, |
| "grad_norm": 4.334892848185411, |
| "learning_rate": 1e-05, |
| "loss": 0.0652, |
| "step": 25440 |
| }, |
| { |
| "epoch": 8.12, |
| "grad_norm": 3.319745327247372, |
| "learning_rate": 1e-05, |
| "loss": 0.0659, |
| "step": 25450 |
| }, |
| { |
| "epoch": 8.12, |
| "grad_norm": 6.0459242277807625, |
| "learning_rate": 1e-05, |
| "loss": 0.0507, |
| "step": 25460 |
| }, |
| { |
| "epoch": 8.13, |
| "grad_norm": 5.352300573533157, |
| "learning_rate": 1e-05, |
| "loss": 0.0573, |
| "step": 25470 |
| }, |
| { |
| "epoch": 8.13, |
| "grad_norm": 3.233570090864235, |
| "learning_rate": 1e-05, |
| "loss": 0.0566, |
| "step": 25480 |
| }, |
| { |
| "epoch": 8.13, |
| "grad_norm": 2.954272334939976, |
| "learning_rate": 1e-05, |
| "loss": 0.0548, |
| "step": 25490 |
| }, |
| { |
| "epoch": 8.14, |
| "grad_norm": 5.074574517515087, |
| "learning_rate": 1e-05, |
| "loss": 0.0605, |
| "step": 25500 |
| }, |
| { |
| "epoch": 8.14, |
| "eval_loss": 1.146484375, |
| "eval_runtime": 54.7594, |
| "eval_samples_per_second": 46.275, |
| "eval_steps_per_second": 2.904, |
| "step": 25500 |
| }, |
| { |
| "epoch": 8.14, |
| "grad_norm": 7.768263923003737, |
| "learning_rate": 1e-05, |
| "loss": 0.0638, |
| "step": 25510 |
| }, |
| { |
| "epoch": 8.14, |
| "grad_norm": 2.1671597457681746, |
| "learning_rate": 1e-05, |
| "loss": 0.0629, |
| "step": 25520 |
| }, |
| { |
| "epoch": 8.15, |
| "grad_norm": 6.264807593092288, |
| "learning_rate": 1e-05, |
| "loss": 0.0614, |
| "step": 25530 |
| }, |
| { |
| "epoch": 8.15, |
| "grad_norm": 2.3573584623112205, |
| "learning_rate": 1e-05, |
| "loss": 0.0549, |
| "step": 25540 |
| }, |
| { |
| "epoch": 8.15, |
| "grad_norm": 5.565347178581014, |
| "learning_rate": 1e-05, |
| "loss": 0.063, |
| "step": 25550 |
| }, |
| { |
| "epoch": 8.16, |
| "grad_norm": 3.423570604753169, |
| "learning_rate": 1e-05, |
| "loss": 0.0575, |
| "step": 25560 |
| }, |
| { |
| "epoch": 8.16, |
| "grad_norm": 5.480950745576853, |
| "learning_rate": 1e-05, |
| "loss": 0.055, |
| "step": 25570 |
| }, |
| { |
| "epoch": 8.16, |
| "grad_norm": 2.704710478880623, |
| "learning_rate": 1e-05, |
| "loss": 0.0783, |
| "step": 25580 |
| }, |
| { |
| "epoch": 8.17, |
| "grad_norm": 4.829428685046524, |
| "learning_rate": 1e-05, |
| "loss": 0.0739, |
| "step": 25590 |
| }, |
| { |
| "epoch": 8.17, |
| "grad_norm": 4.500349031263935, |
| "learning_rate": 1e-05, |
| "loss": 0.0509, |
| "step": 25600 |
| }, |
| { |
| "epoch": 8.17, |
| "grad_norm": 4.7283758546954155, |
| "learning_rate": 1e-05, |
| "loss": 0.0513, |
| "step": 25610 |
| }, |
| { |
| "epoch": 8.17, |
| "grad_norm": 6.478892460588394, |
| "learning_rate": 1e-05, |
| "loss": 0.0723, |
| "step": 25620 |
| }, |
| { |
| "epoch": 8.18, |
| "grad_norm": 4.514874353022904, |
| "learning_rate": 1e-05, |
| "loss": 0.0731, |
| "step": 25630 |
| }, |
| { |
| "epoch": 8.18, |
| "grad_norm": 2.268551849875808, |
| "learning_rate": 1e-05, |
| "loss": 0.0762, |
| "step": 25640 |
| }, |
| { |
| "epoch": 8.18, |
| "grad_norm": 5.207997832618649, |
| "learning_rate": 1e-05, |
| "loss": 0.0867, |
| "step": 25650 |
| }, |
| { |
| "epoch": 8.19, |
| "grad_norm": 5.731155364774676, |
| "learning_rate": 1e-05, |
| "loss": 0.0708, |
| "step": 25660 |
| }, |
| { |
| "epoch": 8.19, |
| "grad_norm": 5.934243684574316, |
| "learning_rate": 1e-05, |
| "loss": 0.059, |
| "step": 25670 |
| }, |
| { |
| "epoch": 8.19, |
| "grad_norm": 4.029305395652648, |
| "learning_rate": 1e-05, |
| "loss": 0.0598, |
| "step": 25680 |
| }, |
| { |
| "epoch": 8.2, |
| "grad_norm": 3.1186263126775864, |
| "learning_rate": 1e-05, |
| "loss": 0.0676, |
| "step": 25690 |
| }, |
| { |
| "epoch": 8.2, |
| "grad_norm": 3.3886451390335783, |
| "learning_rate": 1e-05, |
| "loss": 0.0648, |
| "step": 25700 |
| }, |
| { |
| "epoch": 8.2, |
| "grad_norm": 3.776517444321313, |
| "learning_rate": 1e-05, |
| "loss": 0.0575, |
| "step": 25710 |
| }, |
| { |
| "epoch": 8.21, |
| "grad_norm": 2.3000821928182225, |
| "learning_rate": 1e-05, |
| "loss": 0.0649, |
| "step": 25720 |
| }, |
| { |
| "epoch": 8.21, |
| "grad_norm": 4.998857939943418, |
| "learning_rate": 1e-05, |
| "loss": 0.0635, |
| "step": 25730 |
| }, |
| { |
| "epoch": 8.21, |
| "grad_norm": 4.941808245444294, |
| "learning_rate": 1e-05, |
| "loss": 0.0468, |
| "step": 25740 |
| }, |
| { |
| "epoch": 8.22, |
| "grad_norm": 3.774236937952644, |
| "learning_rate": 1e-05, |
| "loss": 0.0768, |
| "step": 25750 |
| }, |
| { |
| "epoch": 8.22, |
| "grad_norm": 4.569122315981847, |
| "learning_rate": 1e-05, |
| "loss": 0.0588, |
| "step": 25760 |
| }, |
| { |
| "epoch": 8.22, |
| "grad_norm": 4.738614541555105, |
| "learning_rate": 1e-05, |
| "loss": 0.0726, |
| "step": 25770 |
| }, |
| { |
| "epoch": 8.23, |
| "grad_norm": 6.266921368285971, |
| "learning_rate": 1e-05, |
| "loss": 0.0582, |
| "step": 25780 |
| }, |
| { |
| "epoch": 8.23, |
| "grad_norm": 3.739194816164301, |
| "learning_rate": 1e-05, |
| "loss": 0.0586, |
| "step": 25790 |
| }, |
| { |
| "epoch": 8.23, |
| "grad_norm": 6.243113576779487, |
| "learning_rate": 1e-05, |
| "loss": 0.0624, |
| "step": 25800 |
| }, |
| { |
| "epoch": 8.24, |
| "grad_norm": 5.586622758994772, |
| "learning_rate": 1e-05, |
| "loss": 0.0558, |
| "step": 25810 |
| }, |
| { |
| "epoch": 8.24, |
| "grad_norm": 4.564569082488891, |
| "learning_rate": 1e-05, |
| "loss": 0.0985, |
| "step": 25820 |
| }, |
| { |
| "epoch": 8.24, |
| "grad_norm": 3.047619537938169, |
| "learning_rate": 1e-05, |
| "loss": 0.0718, |
| "step": 25830 |
| }, |
| { |
| "epoch": 8.25, |
| "grad_norm": 4.883071868111366, |
| "learning_rate": 1e-05, |
| "loss": 0.0822, |
| "step": 25840 |
| }, |
| { |
| "epoch": 8.25, |
| "grad_norm": 6.34349142214692, |
| "learning_rate": 1e-05, |
| "loss": 0.0835, |
| "step": 25850 |
| }, |
| { |
| "epoch": 8.25, |
| "grad_norm": 2.2638352263577817, |
| "learning_rate": 1e-05, |
| "loss": 0.0569, |
| "step": 25860 |
| }, |
| { |
| "epoch": 8.25, |
| "grad_norm": 5.435214910811622, |
| "learning_rate": 1e-05, |
| "loss": 0.0573, |
| "step": 25870 |
| }, |
| { |
| "epoch": 8.26, |
| "grad_norm": 3.404398143529268, |
| "learning_rate": 1e-05, |
| "loss": 0.0593, |
| "step": 25880 |
| }, |
| { |
| "epoch": 8.26, |
| "grad_norm": 2.407906098622859, |
| "learning_rate": 1e-05, |
| "loss": 0.0759, |
| "step": 25890 |
| }, |
| { |
| "epoch": 8.26, |
| "grad_norm": 2.041776407627112, |
| "learning_rate": 1e-05, |
| "loss": 0.0718, |
| "step": 25900 |
| }, |
| { |
| "epoch": 8.27, |
| "grad_norm": 4.24520738434343, |
| "learning_rate": 1e-05, |
| "loss": 0.0695, |
| "step": 25910 |
| }, |
| { |
| "epoch": 8.27, |
| "grad_norm": 3.9381247130834165, |
| "learning_rate": 1e-05, |
| "loss": 0.0626, |
| "step": 25920 |
| }, |
| { |
| "epoch": 8.27, |
| "grad_norm": 4.760950616431001, |
| "learning_rate": 1e-05, |
| "loss": 0.0855, |
| "step": 25930 |
| }, |
| { |
| "epoch": 8.28, |
| "grad_norm": 4.518958105674343, |
| "learning_rate": 1e-05, |
| "loss": 0.0713, |
| "step": 25940 |
| }, |
| { |
| "epoch": 8.28, |
| "grad_norm": 4.892783179335687, |
| "learning_rate": 1e-05, |
| "loss": 0.0728, |
| "step": 25950 |
| }, |
| { |
| "epoch": 8.28, |
| "grad_norm": 2.9134717245279766, |
| "learning_rate": 1e-05, |
| "loss": 0.0894, |
| "step": 25960 |
| }, |
| { |
| "epoch": 8.29, |
| "grad_norm": 3.878096051227193, |
| "learning_rate": 1e-05, |
| "loss": 0.0627, |
| "step": 25970 |
| }, |
| { |
| "epoch": 8.29, |
| "grad_norm": 3.7991549706758394, |
| "learning_rate": 1e-05, |
| "loss": 0.072, |
| "step": 25980 |
| }, |
| { |
| "epoch": 8.29, |
| "grad_norm": 3.4816796522201168, |
| "learning_rate": 1e-05, |
| "loss": 0.0632, |
| "step": 25990 |
| }, |
| { |
| "epoch": 8.3, |
| "grad_norm": 4.8861365638538885, |
| "learning_rate": 1e-05, |
| "loss": 0.0607, |
| "step": 26000 |
| }, |
| { |
| "epoch": 8.3, |
| "eval_loss": 1.1435546875, |
| "eval_runtime": 54.8642, |
| "eval_samples_per_second": 46.187, |
| "eval_steps_per_second": 2.898, |
| "step": 26000 |
| }, |
| { |
| "epoch": 8.3, |
| "grad_norm": 2.691204091135424, |
| "learning_rate": 1e-05, |
| "loss": 0.0565, |
| "step": 26010 |
| }, |
| { |
| "epoch": 8.3, |
| "grad_norm": 4.778520590434149, |
| "learning_rate": 1e-05, |
| "loss": 0.0454, |
| "step": 26020 |
| }, |
| { |
| "epoch": 8.31, |
| "grad_norm": 3.1614301115700467, |
| "learning_rate": 1e-05, |
| "loss": 0.0582, |
| "step": 26030 |
| }, |
| { |
| "epoch": 8.31, |
| "grad_norm": 4.864973467826045, |
| "learning_rate": 1e-05, |
| "loss": 0.0738, |
| "step": 26040 |
| }, |
| { |
| "epoch": 8.31, |
| "grad_norm": 3.8243288486050657, |
| "learning_rate": 1e-05, |
| "loss": 0.0794, |
| "step": 26050 |
| }, |
| { |
| "epoch": 8.32, |
| "grad_norm": 3.4430677233086833, |
| "learning_rate": 1e-05, |
| "loss": 0.0705, |
| "step": 26060 |
| }, |
| { |
| "epoch": 8.32, |
| "grad_norm": 4.551292973509055, |
| "learning_rate": 1e-05, |
| "loss": 0.0655, |
| "step": 26070 |
| }, |
| { |
| "epoch": 8.32, |
| "grad_norm": 3.184485262682908, |
| "learning_rate": 1e-05, |
| "loss": 0.064, |
| "step": 26080 |
| }, |
| { |
| "epoch": 8.32, |
| "grad_norm": 4.276590303575243, |
| "learning_rate": 1e-05, |
| "loss": 0.0576, |
| "step": 26090 |
| }, |
| { |
| "epoch": 8.33, |
| "grad_norm": 4.521018005443242, |
| "learning_rate": 1e-05, |
| "loss": 0.0692, |
| "step": 26100 |
| }, |
| { |
| "epoch": 8.33, |
| "grad_norm": 2.7026295290526545, |
| "learning_rate": 1e-05, |
| "loss": 0.0644, |
| "step": 26110 |
| }, |
| { |
| "epoch": 8.33, |
| "grad_norm": 3.9383299422172144, |
| "learning_rate": 1e-05, |
| "loss": 0.0535, |
| "step": 26120 |
| }, |
| { |
| "epoch": 8.34, |
| "grad_norm": 4.839483338515037, |
| "learning_rate": 1e-05, |
| "loss": 0.0792, |
| "step": 26130 |
| }, |
| { |
| "epoch": 8.34, |
| "grad_norm": 7.05993497737519, |
| "learning_rate": 1e-05, |
| "loss": 0.0938, |
| "step": 26140 |
| }, |
| { |
| "epoch": 8.34, |
| "grad_norm": 3.7262804736376047, |
| "learning_rate": 1e-05, |
| "loss": 0.077, |
| "step": 26150 |
| }, |
| { |
| "epoch": 8.35, |
| "grad_norm": 4.141795097421928, |
| "learning_rate": 1e-05, |
| "loss": 0.078, |
| "step": 26160 |
| }, |
| { |
| "epoch": 8.35, |
| "grad_norm": 3.5245628843440717, |
| "learning_rate": 1e-05, |
| "loss": 0.0639, |
| "step": 26170 |
| }, |
| { |
| "epoch": 8.35, |
| "grad_norm": 2.8339197636501003, |
| "learning_rate": 1e-05, |
| "loss": 0.0684, |
| "step": 26180 |
| }, |
| { |
| "epoch": 8.36, |
| "grad_norm": 3.049721195422954, |
| "learning_rate": 1e-05, |
| "loss": 0.0612, |
| "step": 26190 |
| }, |
| { |
| "epoch": 8.36, |
| "grad_norm": 4.7012213722986305, |
| "learning_rate": 1e-05, |
| "loss": 0.0766, |
| "step": 26200 |
| }, |
| { |
| "epoch": 8.36, |
| "grad_norm": 4.544221818399103, |
| "learning_rate": 1e-05, |
| "loss": 0.0572, |
| "step": 26210 |
| }, |
| { |
| "epoch": 8.37, |
| "grad_norm": 4.628415676470118, |
| "learning_rate": 1e-05, |
| "loss": 0.0732, |
| "step": 26220 |
| }, |
| { |
| "epoch": 8.37, |
| "grad_norm": 5.232540526545253, |
| "learning_rate": 1e-05, |
| "loss": 0.0829, |
| "step": 26230 |
| }, |
| { |
| "epoch": 8.37, |
| "grad_norm": 4.577079783828787, |
| "learning_rate": 1e-05, |
| "loss": 0.0579, |
| "step": 26240 |
| }, |
| { |
| "epoch": 8.38, |
| "grad_norm": 6.914023327042984, |
| "learning_rate": 1e-05, |
| "loss": 0.0524, |
| "step": 26250 |
| }, |
| { |
| "epoch": 8.38, |
| "grad_norm": 4.541222063519928, |
| "learning_rate": 1e-05, |
| "loss": 0.0739, |
| "step": 26260 |
| }, |
| { |
| "epoch": 8.38, |
| "grad_norm": 6.403896885858409, |
| "learning_rate": 1e-05, |
| "loss": 0.0797, |
| "step": 26270 |
| }, |
| { |
| "epoch": 8.39, |
| "grad_norm": 4.021622153798739, |
| "learning_rate": 1e-05, |
| "loss": 0.0491, |
| "step": 26280 |
| }, |
| { |
| "epoch": 8.39, |
| "grad_norm": 5.60861260378674, |
| "learning_rate": 1e-05, |
| "loss": 0.0665, |
| "step": 26290 |
| }, |
| { |
| "epoch": 8.39, |
| "grad_norm": 5.080843435540943, |
| "learning_rate": 1e-05, |
| "loss": 0.0587, |
| "step": 26300 |
| }, |
| { |
| "epoch": 8.4, |
| "grad_norm": 3.3775181736093622, |
| "learning_rate": 1e-05, |
| "loss": 0.0677, |
| "step": 26310 |
| }, |
| { |
| "epoch": 8.4, |
| "grad_norm": 3.8602482969839746, |
| "learning_rate": 1e-05, |
| "loss": 0.063, |
| "step": 26320 |
| }, |
| { |
| "epoch": 8.4, |
| "grad_norm": 3.9510729141401026, |
| "learning_rate": 1e-05, |
| "loss": 0.0696, |
| "step": 26330 |
| }, |
| { |
| "epoch": 8.4, |
| "grad_norm": 5.6661092072588914, |
| "learning_rate": 1e-05, |
| "loss": 0.0682, |
| "step": 26340 |
| }, |
| { |
| "epoch": 8.41, |
| "grad_norm": 3.3723576409768263, |
| "learning_rate": 1e-05, |
| "loss": 0.0707, |
| "step": 26350 |
| }, |
| { |
| "epoch": 8.41, |
| "grad_norm": 6.776892602509447, |
| "learning_rate": 1e-05, |
| "loss": 0.0717, |
| "step": 26360 |
| }, |
| { |
| "epoch": 8.41, |
| "grad_norm": 4.934671871635731, |
| "learning_rate": 1e-05, |
| "loss": 0.0694, |
| "step": 26370 |
| }, |
| { |
| "epoch": 8.42, |
| "grad_norm": 3.5835001854899824, |
| "learning_rate": 1e-05, |
| "loss": 0.0725, |
| "step": 26380 |
| }, |
| { |
| "epoch": 8.42, |
| "grad_norm": 7.088323593055221, |
| "learning_rate": 1e-05, |
| "loss": 0.0629, |
| "step": 26390 |
| }, |
| { |
| "epoch": 8.42, |
| "grad_norm": 3.9420668744937544, |
| "learning_rate": 1e-05, |
| "loss": 0.069, |
| "step": 26400 |
| }, |
| { |
| "epoch": 8.43, |
| "grad_norm": 4.871672056814825, |
| "learning_rate": 1e-05, |
| "loss": 0.0743, |
| "step": 26410 |
| }, |
| { |
| "epoch": 8.43, |
| "grad_norm": 3.6922414006493325, |
| "learning_rate": 1e-05, |
| "loss": 0.0744, |
| "step": 26420 |
| }, |
| { |
| "epoch": 8.43, |
| "grad_norm": 4.961826322936382, |
| "learning_rate": 1e-05, |
| "loss": 0.0625, |
| "step": 26430 |
| }, |
| { |
| "epoch": 8.44, |
| "grad_norm": 6.862912180600362, |
| "learning_rate": 1e-05, |
| "loss": 0.0759, |
| "step": 26440 |
| }, |
| { |
| "epoch": 8.44, |
| "grad_norm": 3.013767441169462, |
| "learning_rate": 1e-05, |
| "loss": 0.054, |
| "step": 26450 |
| }, |
| { |
| "epoch": 8.44, |
| "grad_norm": 4.396627972027433, |
| "learning_rate": 1e-05, |
| "loss": 0.0613, |
| "step": 26460 |
| }, |
| { |
| "epoch": 8.45, |
| "grad_norm": 4.683412524512, |
| "learning_rate": 1e-05, |
| "loss": 0.0508, |
| "step": 26470 |
| }, |
| { |
| "epoch": 8.45, |
| "grad_norm": 6.218027044090828, |
| "learning_rate": 1e-05, |
| "loss": 0.0574, |
| "step": 26480 |
| }, |
| { |
| "epoch": 8.45, |
| "grad_norm": 3.7434644332262046, |
| "learning_rate": 1e-05, |
| "loss": 0.0628, |
| "step": 26490 |
| }, |
| { |
| "epoch": 8.46, |
| "grad_norm": 7.155016930304862, |
| "learning_rate": 1e-05, |
| "loss": 0.0569, |
| "step": 26500 |
| }, |
| { |
| "epoch": 8.46, |
| "eval_loss": 1.1630859375, |
| "eval_runtime": 54.8222, |
| "eval_samples_per_second": 46.222, |
| "eval_steps_per_second": 2.9, |
| "step": 26500 |
| }, |
| { |
| "epoch": 8.46, |
| "grad_norm": 4.382702395982192, |
| "learning_rate": 1e-05, |
| "loss": 0.0675, |
| "step": 26510 |
| }, |
| { |
| "epoch": 8.46, |
| "grad_norm": 4.578545977405044, |
| "learning_rate": 1e-05, |
| "loss": 0.0693, |
| "step": 26520 |
| }, |
| { |
| "epoch": 8.47, |
| "grad_norm": 5.971009151463866, |
| "learning_rate": 1e-05, |
| "loss": 0.0692, |
| "step": 26530 |
| }, |
| { |
| "epoch": 8.47, |
| "grad_norm": 5.0910954023304145, |
| "learning_rate": 1e-05, |
| "loss": 0.0718, |
| "step": 26540 |
| }, |
| { |
| "epoch": 8.47, |
| "grad_norm": 6.06719600058959, |
| "learning_rate": 1e-05, |
| "loss": 0.0815, |
| "step": 26550 |
| }, |
| { |
| "epoch": 8.47, |
| "grad_norm": 2.6864462272506366, |
| "learning_rate": 1e-05, |
| "loss": 0.0792, |
| "step": 26560 |
| }, |
| { |
| "epoch": 8.48, |
| "grad_norm": 4.509172416255183, |
| "learning_rate": 1e-05, |
| "loss": 0.0756, |
| "step": 26570 |
| }, |
| { |
| "epoch": 8.48, |
| "grad_norm": 5.578374947897117, |
| "learning_rate": 1e-05, |
| "loss": 0.0555, |
| "step": 26580 |
| }, |
| { |
| "epoch": 8.48, |
| "grad_norm": 4.829199810205078, |
| "learning_rate": 1e-05, |
| "loss": 0.0709, |
| "step": 26590 |
| }, |
| { |
| "epoch": 8.49, |
| "grad_norm": 2.7329900231345037, |
| "learning_rate": 1e-05, |
| "loss": 0.0787, |
| "step": 26600 |
| }, |
| { |
| "epoch": 8.49, |
| "grad_norm": 4.745121910530938, |
| "learning_rate": 1e-05, |
| "loss": 0.0718, |
| "step": 26610 |
| }, |
| { |
| "epoch": 8.49, |
| "grad_norm": 5.128253927243185, |
| "learning_rate": 1e-05, |
| "loss": 0.0833, |
| "step": 26620 |
| }, |
| { |
| "epoch": 8.5, |
| "grad_norm": 4.058239859902079, |
| "learning_rate": 1e-05, |
| "loss": 0.0682, |
| "step": 26630 |
| }, |
| { |
| "epoch": 8.5, |
| "grad_norm": 5.200638277054994, |
| "learning_rate": 1e-05, |
| "loss": 0.0828, |
| "step": 26640 |
| }, |
| { |
| "epoch": 8.5, |
| "grad_norm": 6.189624768348176, |
| "learning_rate": 1e-05, |
| "loss": 0.0853, |
| "step": 26650 |
| }, |
| { |
| "epoch": 8.51, |
| "grad_norm": 4.682560059772742, |
| "learning_rate": 1e-05, |
| "loss": 0.0557, |
| "step": 26660 |
| }, |
| { |
| "epoch": 8.51, |
| "grad_norm": 5.584490736758405, |
| "learning_rate": 1e-05, |
| "loss": 0.0763, |
| "step": 26670 |
| }, |
| { |
| "epoch": 8.51, |
| "grad_norm": 5.930886852213568, |
| "learning_rate": 1e-05, |
| "loss": 0.0707, |
| "step": 26680 |
| }, |
| { |
| "epoch": 8.52, |
| "grad_norm": 7.066801128667671, |
| "learning_rate": 1e-05, |
| "loss": 0.0701, |
| "step": 26690 |
| }, |
| { |
| "epoch": 8.52, |
| "grad_norm": 4.170700079243409, |
| "learning_rate": 1e-05, |
| "loss": 0.0714, |
| "step": 26700 |
| }, |
| { |
| "epoch": 8.52, |
| "grad_norm": 4.641294077924113, |
| "learning_rate": 1e-05, |
| "loss": 0.0559, |
| "step": 26710 |
| }, |
| { |
| "epoch": 8.53, |
| "grad_norm": 6.06212654143294, |
| "learning_rate": 1e-05, |
| "loss": 0.0748, |
| "step": 26720 |
| }, |
| { |
| "epoch": 8.53, |
| "grad_norm": 4.465552406441754, |
| "learning_rate": 1e-05, |
| "loss": 0.061, |
| "step": 26730 |
| }, |
| { |
| "epoch": 8.53, |
| "grad_norm": 6.132863193958475, |
| "learning_rate": 1e-05, |
| "loss": 0.0698, |
| "step": 26740 |
| }, |
| { |
| "epoch": 8.54, |
| "grad_norm": 3.2393462819072685, |
| "learning_rate": 1e-05, |
| "loss": 0.0776, |
| "step": 26750 |
| }, |
| { |
| "epoch": 8.54, |
| "grad_norm": 4.168247660463562, |
| "learning_rate": 1e-05, |
| "loss": 0.0777, |
| "step": 26760 |
| }, |
| { |
| "epoch": 8.54, |
| "grad_norm": 4.734894676769237, |
| "learning_rate": 1e-05, |
| "loss": 0.0525, |
| "step": 26770 |
| }, |
| { |
| "epoch": 8.54, |
| "grad_norm": 3.217387976011534, |
| "learning_rate": 1e-05, |
| "loss": 0.0683, |
| "step": 26780 |
| }, |
| { |
| "epoch": 8.55, |
| "grad_norm": 6.24623574387147, |
| "learning_rate": 1e-05, |
| "loss": 0.0582, |
| "step": 26790 |
| }, |
| { |
| "epoch": 8.55, |
| "grad_norm": 3.791712261107469, |
| "learning_rate": 1e-05, |
| "loss": 0.0634, |
| "step": 26800 |
| }, |
| { |
| "epoch": 8.55, |
| "grad_norm": 4.215111640003775, |
| "learning_rate": 1e-05, |
| "loss": 0.0647, |
| "step": 26810 |
| }, |
| { |
| "epoch": 8.56, |
| "grad_norm": 10.219587285914907, |
| "learning_rate": 1e-05, |
| "loss": 0.0691, |
| "step": 26820 |
| }, |
| { |
| "epoch": 8.56, |
| "grad_norm": 4.387856476042226, |
| "learning_rate": 1e-05, |
| "loss": 0.0635, |
| "step": 26830 |
| }, |
| { |
| "epoch": 8.56, |
| "grad_norm": 3.5239413085189475, |
| "learning_rate": 1e-05, |
| "loss": 0.0617, |
| "step": 26840 |
| }, |
| { |
| "epoch": 8.57, |
| "grad_norm": 4.953679483948018, |
| "learning_rate": 1e-05, |
| "loss": 0.0817, |
| "step": 26850 |
| }, |
| { |
| "epoch": 8.57, |
| "grad_norm": 5.226253939474063, |
| "learning_rate": 1e-05, |
| "loss": 0.0823, |
| "step": 26860 |
| }, |
| { |
| "epoch": 8.57, |
| "grad_norm": 4.503364576734414, |
| "learning_rate": 1e-05, |
| "loss": 0.0746, |
| "step": 26870 |
| }, |
| { |
| "epoch": 8.58, |
| "grad_norm": 4.348107875731584, |
| "learning_rate": 1e-05, |
| "loss": 0.0476, |
| "step": 26880 |
| }, |
| { |
| "epoch": 8.58, |
| "grad_norm": 3.211994229293713, |
| "learning_rate": 1e-05, |
| "loss": 0.06, |
| "step": 26890 |
| }, |
| { |
| "epoch": 8.58, |
| "grad_norm": 6.261073259481613, |
| "learning_rate": 1e-05, |
| "loss": 0.0749, |
| "step": 26900 |
| }, |
| { |
| "epoch": 8.59, |
| "grad_norm": 6.233610713451352, |
| "learning_rate": 1e-05, |
| "loss": 0.0692, |
| "step": 26910 |
| }, |
| { |
| "epoch": 8.59, |
| "grad_norm": 4.424414670345773, |
| "learning_rate": 1e-05, |
| "loss": 0.0814, |
| "step": 26920 |
| }, |
| { |
| "epoch": 8.59, |
| "grad_norm": 5.608851501348968, |
| "learning_rate": 1e-05, |
| "loss": 0.0717, |
| "step": 26930 |
| }, |
| { |
| "epoch": 8.6, |
| "grad_norm": 5.527179319274615, |
| "learning_rate": 1e-05, |
| "loss": 0.0633, |
| "step": 26940 |
| }, |
| { |
| "epoch": 8.6, |
| "grad_norm": 6.481761580374155, |
| "learning_rate": 1e-05, |
| "loss": 0.079, |
| "step": 26950 |
| }, |
| { |
| "epoch": 8.6, |
| "grad_norm": 4.413825050651508, |
| "learning_rate": 1e-05, |
| "loss": 0.0798, |
| "step": 26960 |
| }, |
| { |
| "epoch": 8.61, |
| "grad_norm": 7.632244483326742, |
| "learning_rate": 1e-05, |
| "loss": 0.0684, |
| "step": 26970 |
| }, |
| { |
| "epoch": 8.61, |
| "grad_norm": 7.612573722895788, |
| "learning_rate": 1e-05, |
| "loss": 0.1003, |
| "step": 26980 |
| }, |
| { |
| "epoch": 8.61, |
| "grad_norm": 4.342976014044901, |
| "learning_rate": 1e-05, |
| "loss": 0.0693, |
| "step": 26990 |
| }, |
| { |
| "epoch": 8.62, |
| "grad_norm": 5.497253425761885, |
| "learning_rate": 1e-05, |
| "loss": 0.0638, |
| "step": 27000 |
| }, |
| { |
| "epoch": 8.62, |
| "eval_loss": 1.12890625, |
| "eval_runtime": 54.6877, |
| "eval_samples_per_second": 46.336, |
| "eval_steps_per_second": 2.907, |
| "step": 27000 |
| }, |
| { |
| "epoch": 8.62, |
| "grad_norm": 4.4527373563460495, |
| "learning_rate": 1e-05, |
| "loss": 0.0823, |
| "step": 27010 |
| }, |
| { |
| "epoch": 8.62, |
| "grad_norm": 3.6040489897875383, |
| "learning_rate": 1e-05, |
| "loss": 0.0527, |
| "step": 27020 |
| }, |
| { |
| "epoch": 8.62, |
| "grad_norm": 4.369260238081061, |
| "learning_rate": 1e-05, |
| "loss": 0.1067, |
| "step": 27030 |
| }, |
| { |
| "epoch": 8.63, |
| "grad_norm": 5.6501521334801, |
| "learning_rate": 1e-05, |
| "loss": 0.0637, |
| "step": 27040 |
| }, |
| { |
| "epoch": 8.63, |
| "grad_norm": 3.840851722276538, |
| "learning_rate": 1e-05, |
| "loss": 0.0662, |
| "step": 27050 |
| }, |
| { |
| "epoch": 8.63, |
| "grad_norm": 3.427669548443256, |
| "learning_rate": 1e-05, |
| "loss": 0.0838, |
| "step": 27060 |
| }, |
| { |
| "epoch": 8.64, |
| "grad_norm": 5.342862596501772, |
| "learning_rate": 1e-05, |
| "loss": 0.0649, |
| "step": 27070 |
| }, |
| { |
| "epoch": 8.64, |
| "grad_norm": 3.3855941647679426, |
| "learning_rate": 1e-05, |
| "loss": 0.078, |
| "step": 27080 |
| }, |
| { |
| "epoch": 8.64, |
| "grad_norm": 4.952908965492803, |
| "learning_rate": 1e-05, |
| "loss": 0.0793, |
| "step": 27090 |
| }, |
| { |
| "epoch": 8.65, |
| "grad_norm": 4.8997674380796505, |
| "learning_rate": 1e-05, |
| "loss": 0.0657, |
| "step": 27100 |
| }, |
| { |
| "epoch": 8.65, |
| "grad_norm": 3.5948426285680326, |
| "learning_rate": 1e-05, |
| "loss": 0.0855, |
| "step": 27110 |
| }, |
| { |
| "epoch": 8.65, |
| "grad_norm": 4.1226515587221035, |
| "learning_rate": 1e-05, |
| "loss": 0.0674, |
| "step": 27120 |
| }, |
| { |
| "epoch": 8.66, |
| "grad_norm": 6.101606263847532, |
| "learning_rate": 1e-05, |
| "loss": 0.0858, |
| "step": 27130 |
| }, |
| { |
| "epoch": 8.66, |
| "grad_norm": 3.193081176958852, |
| "learning_rate": 1e-05, |
| "loss": 0.0884, |
| "step": 27140 |
| }, |
| { |
| "epoch": 8.66, |
| "grad_norm": 4.940193696954375, |
| "learning_rate": 1e-05, |
| "loss": 0.0719, |
| "step": 27150 |
| }, |
| { |
| "epoch": 8.67, |
| "grad_norm": 6.857697509858447, |
| "learning_rate": 1e-05, |
| "loss": 0.0695, |
| "step": 27160 |
| }, |
| { |
| "epoch": 8.67, |
| "grad_norm": 4.5575666537200075, |
| "learning_rate": 1e-05, |
| "loss": 0.0713, |
| "step": 27170 |
| }, |
| { |
| "epoch": 8.67, |
| "grad_norm": 3.4730463313058424, |
| "learning_rate": 1e-05, |
| "loss": 0.056, |
| "step": 27180 |
| }, |
| { |
| "epoch": 8.68, |
| "grad_norm": 4.026156971190398, |
| "learning_rate": 1e-05, |
| "loss": 0.0588, |
| "step": 27190 |
| }, |
| { |
| "epoch": 8.68, |
| "grad_norm": 3.2437630008150324, |
| "learning_rate": 1e-05, |
| "loss": 0.0645, |
| "step": 27200 |
| }, |
| { |
| "epoch": 8.68, |
| "grad_norm": 4.573612215271285, |
| "learning_rate": 1e-05, |
| "loss": 0.0609, |
| "step": 27210 |
| }, |
| { |
| "epoch": 8.69, |
| "grad_norm": 6.301647261568083, |
| "learning_rate": 1e-05, |
| "loss": 0.0908, |
| "step": 27220 |
| }, |
| { |
| "epoch": 8.69, |
| "grad_norm": 6.642423641984875, |
| "learning_rate": 1e-05, |
| "loss": 0.0742, |
| "step": 27230 |
| }, |
| { |
| "epoch": 8.69, |
| "grad_norm": 4.35848563361549, |
| "learning_rate": 1e-05, |
| "loss": 0.0891, |
| "step": 27240 |
| }, |
| { |
| "epoch": 8.69, |
| "grad_norm": 4.109276367860411, |
| "learning_rate": 1e-05, |
| "loss": 0.072, |
| "step": 27250 |
| }, |
| { |
| "epoch": 8.7, |
| "grad_norm": 7.6246768069953506, |
| "learning_rate": 1e-05, |
| "loss": 0.1073, |
| "step": 27260 |
| }, |
| { |
| "epoch": 8.7, |
| "grad_norm": 6.50555270779782, |
| "learning_rate": 1e-05, |
| "loss": 0.0686, |
| "step": 27270 |
| }, |
| { |
| "epoch": 8.7, |
| "grad_norm": 4.24241589797219, |
| "learning_rate": 1e-05, |
| "loss": 0.0619, |
| "step": 27280 |
| }, |
| { |
| "epoch": 8.71, |
| "grad_norm": 3.043335722316154, |
| "learning_rate": 1e-05, |
| "loss": 0.0711, |
| "step": 27290 |
| }, |
| { |
| "epoch": 8.71, |
| "grad_norm": 5.815797916160029, |
| "learning_rate": 1e-05, |
| "loss": 0.0672, |
| "step": 27300 |
| }, |
| { |
| "epoch": 8.71, |
| "grad_norm": 2.8779409542322725, |
| "learning_rate": 1e-05, |
| "loss": 0.0657, |
| "step": 27310 |
| }, |
| { |
| "epoch": 8.72, |
| "grad_norm": 4.5035949758275144, |
| "learning_rate": 1e-05, |
| "loss": 0.0672, |
| "step": 27320 |
| }, |
| { |
| "epoch": 8.72, |
| "grad_norm": 5.442166310262814, |
| "learning_rate": 1e-05, |
| "loss": 0.0672, |
| "step": 27330 |
| }, |
| { |
| "epoch": 8.72, |
| "grad_norm": 4.9517383258932535, |
| "learning_rate": 1e-05, |
| "loss": 0.0662, |
| "step": 27340 |
| }, |
| { |
| "epoch": 8.73, |
| "grad_norm": 5.002640789744534, |
| "learning_rate": 1e-05, |
| "loss": 0.073, |
| "step": 27350 |
| }, |
| { |
| "epoch": 8.73, |
| "grad_norm": 6.261132358596585, |
| "learning_rate": 1e-05, |
| "loss": 0.0962, |
| "step": 27360 |
| }, |
| { |
| "epoch": 8.73, |
| "grad_norm": 4.785519557317855, |
| "learning_rate": 1e-05, |
| "loss": 0.0942, |
| "step": 27370 |
| }, |
| { |
| "epoch": 8.74, |
| "grad_norm": 4.351750313285974, |
| "learning_rate": 1e-05, |
| "loss": 0.0572, |
| "step": 27380 |
| }, |
| { |
| "epoch": 8.74, |
| "grad_norm": 5.342657144794273, |
| "learning_rate": 1e-05, |
| "loss": 0.0836, |
| "step": 27390 |
| }, |
| { |
| "epoch": 8.74, |
| "grad_norm": 4.593096835872413, |
| "learning_rate": 1e-05, |
| "loss": 0.065, |
| "step": 27400 |
| }, |
| { |
| "epoch": 8.75, |
| "grad_norm": 6.799062372526657, |
| "learning_rate": 1e-05, |
| "loss": 0.089, |
| "step": 27410 |
| }, |
| { |
| "epoch": 8.75, |
| "grad_norm": 7.482868719935233, |
| "learning_rate": 1e-05, |
| "loss": 0.098, |
| "step": 27420 |
| }, |
| { |
| "epoch": 8.75, |
| "grad_norm": 5.3602189269833405, |
| "learning_rate": 1e-05, |
| "loss": 0.0921, |
| "step": 27430 |
| }, |
| { |
| "epoch": 8.76, |
| "grad_norm": 6.649699681893536, |
| "learning_rate": 1e-05, |
| "loss": 0.0832, |
| "step": 27440 |
| }, |
| { |
| "epoch": 8.76, |
| "grad_norm": 3.9149081293193513, |
| "learning_rate": 1e-05, |
| "loss": 0.0822, |
| "step": 27450 |
| }, |
| { |
| "epoch": 8.76, |
| "grad_norm": 6.320433512921386, |
| "learning_rate": 1e-05, |
| "loss": 0.0796, |
| "step": 27460 |
| }, |
| { |
| "epoch": 8.77, |
| "grad_norm": 3.981503278260281, |
| "learning_rate": 1e-05, |
| "loss": 0.0903, |
| "step": 27470 |
| }, |
| { |
| "epoch": 8.77, |
| "grad_norm": 4.238675681891088, |
| "learning_rate": 1e-05, |
| "loss": 0.0809, |
| "step": 27480 |
| }, |
| { |
| "epoch": 8.77, |
| "grad_norm": 5.900831855709453, |
| "learning_rate": 1e-05, |
| "loss": 0.0776, |
| "step": 27490 |
| }, |
| { |
| "epoch": 8.77, |
| "grad_norm": 4.207498650190802, |
| "learning_rate": 1e-05, |
| "loss": 0.0938, |
| "step": 27500 |
| }, |
| { |
| "epoch": 8.77, |
| "eval_loss": 1.130859375, |
| "eval_runtime": 54.7444, |
| "eval_samples_per_second": 46.288, |
| "eval_steps_per_second": 2.904, |
| "step": 27500 |
| }, |
| { |
| "epoch": 8.78, |
| "grad_norm": 5.217222618402974, |
| "learning_rate": 1e-05, |
| "loss": 0.0873, |
| "step": 27510 |
| }, |
| { |
| "epoch": 8.78, |
| "grad_norm": 2.9206826128053787, |
| "learning_rate": 1e-05, |
| "loss": 0.073, |
| "step": 27520 |
| }, |
| { |
| "epoch": 8.78, |
| "grad_norm": 3.7268931244187624, |
| "learning_rate": 1e-05, |
| "loss": 0.0703, |
| "step": 27530 |
| }, |
| { |
| "epoch": 8.79, |
| "grad_norm": 4.834863201363956, |
| "learning_rate": 1e-05, |
| "loss": 0.0964, |
| "step": 27540 |
| }, |
| { |
| "epoch": 8.79, |
| "grad_norm": 7.105783347557247, |
| "learning_rate": 1e-05, |
| "loss": 0.0687, |
| "step": 27550 |
| }, |
| { |
| "epoch": 8.79, |
| "grad_norm": 5.168567164368939, |
| "learning_rate": 1e-05, |
| "loss": 0.0904, |
| "step": 27560 |
| }, |
| { |
| "epoch": 8.8, |
| "grad_norm": 7.613886754798273, |
| "learning_rate": 1e-05, |
| "loss": 0.0691, |
| "step": 27570 |
| }, |
| { |
| "epoch": 8.8, |
| "grad_norm": 5.0216526402428325, |
| "learning_rate": 1e-05, |
| "loss": 0.0719, |
| "step": 27580 |
| }, |
| { |
| "epoch": 8.8, |
| "grad_norm": 5.902207701836257, |
| "learning_rate": 1e-05, |
| "loss": 0.0822, |
| "step": 27590 |
| }, |
| { |
| "epoch": 8.81, |
| "grad_norm": 4.332267679103889, |
| "learning_rate": 1e-05, |
| "loss": 0.0805, |
| "step": 27600 |
| }, |
| { |
| "epoch": 8.81, |
| "grad_norm": 6.51842791601916, |
| "learning_rate": 1e-05, |
| "loss": 0.0788, |
| "step": 27610 |
| }, |
| { |
| "epoch": 8.81, |
| "grad_norm": 4.582395052959393, |
| "learning_rate": 1e-05, |
| "loss": 0.0721, |
| "step": 27620 |
| }, |
| { |
| "epoch": 8.82, |
| "grad_norm": 7.940893694262266, |
| "learning_rate": 1e-05, |
| "loss": 0.08, |
| "step": 27630 |
| }, |
| { |
| "epoch": 8.82, |
| "grad_norm": 4.901112971123283, |
| "learning_rate": 1e-05, |
| "loss": 0.0759, |
| "step": 27640 |
| }, |
| { |
| "epoch": 8.82, |
| "grad_norm": 5.971786605364408, |
| "learning_rate": 1e-05, |
| "loss": 0.0697, |
| "step": 27650 |
| }, |
| { |
| "epoch": 8.83, |
| "grad_norm": 3.1376559518909213, |
| "learning_rate": 1e-05, |
| "loss": 0.0814, |
| "step": 27660 |
| }, |
| { |
| "epoch": 8.83, |
| "grad_norm": 4.442278371251125, |
| "learning_rate": 1e-05, |
| "loss": 0.105, |
| "step": 27670 |
| }, |
| { |
| "epoch": 8.83, |
| "grad_norm": 3.238172577620911, |
| "learning_rate": 1e-05, |
| "loss": 0.0565, |
| "step": 27680 |
| }, |
| { |
| "epoch": 8.84, |
| "grad_norm": 3.1469417076275903, |
| "learning_rate": 1e-05, |
| "loss": 0.0942, |
| "step": 27690 |
| }, |
| { |
| "epoch": 8.84, |
| "grad_norm": 2.859295119540381, |
| "learning_rate": 1e-05, |
| "loss": 0.0674, |
| "step": 27700 |
| }, |
| { |
| "epoch": 8.84, |
| "grad_norm": 5.596056414730876, |
| "learning_rate": 1e-05, |
| "loss": 0.0779, |
| "step": 27710 |
| }, |
| { |
| "epoch": 8.84, |
| "grad_norm": 3.9885802572553173, |
| "learning_rate": 1e-05, |
| "loss": 0.0684, |
| "step": 27720 |
| }, |
| { |
| "epoch": 8.85, |
| "grad_norm": 4.956167159284138, |
| "learning_rate": 1e-05, |
| "loss": 0.0618, |
| "step": 27730 |
| }, |
| { |
| "epoch": 8.85, |
| "grad_norm": 5.020025206726743, |
| "learning_rate": 1e-05, |
| "loss": 0.0735, |
| "step": 27740 |
| }, |
| { |
| "epoch": 8.85, |
| "grad_norm": 6.025775856303498, |
| "learning_rate": 1e-05, |
| "loss": 0.0669, |
| "step": 27750 |
| }, |
| { |
| "epoch": 8.86, |
| "grad_norm": 4.061364587427693, |
| "learning_rate": 1e-05, |
| "loss": 0.0781, |
| "step": 27760 |
| }, |
| { |
| "epoch": 8.86, |
| "grad_norm": 5.773126286647646, |
| "learning_rate": 1e-05, |
| "loss": 0.0804, |
| "step": 27770 |
| }, |
| { |
| "epoch": 8.86, |
| "grad_norm": 6.006484024934793, |
| "learning_rate": 1e-05, |
| "loss": 0.0791, |
| "step": 27780 |
| }, |
| { |
| "epoch": 8.87, |
| "grad_norm": 5.150718172237667, |
| "learning_rate": 1e-05, |
| "loss": 0.0713, |
| "step": 27790 |
| }, |
| { |
| "epoch": 8.87, |
| "grad_norm": 3.8327841296197644, |
| "learning_rate": 1e-05, |
| "loss": 0.0929, |
| "step": 27800 |
| }, |
| { |
| "epoch": 8.87, |
| "grad_norm": 4.932270032545886, |
| "learning_rate": 1e-05, |
| "loss": 0.0655, |
| "step": 27810 |
| }, |
| { |
| "epoch": 8.88, |
| "grad_norm": 6.634470378605202, |
| "learning_rate": 1e-05, |
| "loss": 0.0804, |
| "step": 27820 |
| }, |
| { |
| "epoch": 8.88, |
| "grad_norm": 5.8831130479127705, |
| "learning_rate": 1e-05, |
| "loss": 0.0913, |
| "step": 27830 |
| }, |
| { |
| "epoch": 8.88, |
| "grad_norm": 4.854702484726531, |
| "learning_rate": 1e-05, |
| "loss": 0.0804, |
| "step": 27840 |
| }, |
| { |
| "epoch": 8.89, |
| "grad_norm": 4.077105983124775, |
| "learning_rate": 1e-05, |
| "loss": 0.0699, |
| "step": 27850 |
| }, |
| { |
| "epoch": 8.89, |
| "grad_norm": 3.2385557125094118, |
| "learning_rate": 1e-05, |
| "loss": 0.0747, |
| "step": 27860 |
| }, |
| { |
| "epoch": 8.89, |
| "grad_norm": 4.58973882027665, |
| "learning_rate": 1e-05, |
| "loss": 0.0916, |
| "step": 27870 |
| }, |
| { |
| "epoch": 8.9, |
| "grad_norm": 8.903881092576551, |
| "learning_rate": 1e-05, |
| "loss": 0.0719, |
| "step": 27880 |
| }, |
| { |
| "epoch": 8.9, |
| "grad_norm": 7.244454729557862, |
| "learning_rate": 1e-05, |
| "loss": 0.07, |
| "step": 27890 |
| }, |
| { |
| "epoch": 8.9, |
| "grad_norm": 3.894134042087424, |
| "learning_rate": 1e-05, |
| "loss": 0.0748, |
| "step": 27900 |
| }, |
| { |
| "epoch": 8.91, |
| "grad_norm": 4.084009127585199, |
| "learning_rate": 1e-05, |
| "loss": 0.0838, |
| "step": 27910 |
| }, |
| { |
| "epoch": 8.91, |
| "grad_norm": 4.841204558931682, |
| "learning_rate": 1e-05, |
| "loss": 0.075, |
| "step": 27920 |
| }, |
| { |
| "epoch": 8.91, |
| "grad_norm": 5.755220779704679, |
| "learning_rate": 1e-05, |
| "loss": 0.0689, |
| "step": 27930 |
| }, |
| { |
| "epoch": 8.92, |
| "grad_norm": 4.26032685400476, |
| "learning_rate": 1e-05, |
| "loss": 0.0702, |
| "step": 27940 |
| }, |
| { |
| "epoch": 8.92, |
| "grad_norm": 5.959462234506711, |
| "learning_rate": 1e-05, |
| "loss": 0.0925, |
| "step": 27950 |
| }, |
| { |
| "epoch": 8.92, |
| "grad_norm": 4.936883429165009, |
| "learning_rate": 1e-05, |
| "loss": 0.0935, |
| "step": 27960 |
| }, |
| { |
| "epoch": 8.92, |
| "grad_norm": 5.129530973948331, |
| "learning_rate": 1e-05, |
| "loss": 0.0852, |
| "step": 27970 |
| }, |
| { |
| "epoch": 8.93, |
| "grad_norm": 5.438072612910399, |
| "learning_rate": 1e-05, |
| "loss": 0.0747, |
| "step": 27980 |
| }, |
| { |
| "epoch": 8.93, |
| "grad_norm": 3.2563832026966937, |
| "learning_rate": 1e-05, |
| "loss": 0.0937, |
| "step": 27990 |
| }, |
| { |
| "epoch": 8.93, |
| "grad_norm": 3.1980932396136756, |
| "learning_rate": 1e-05, |
| "loss": 0.071, |
| "step": 28000 |
| }, |
| { |
| "epoch": 8.93, |
| "eval_loss": 1.087890625, |
| "eval_runtime": 54.7321, |
| "eval_samples_per_second": 46.298, |
| "eval_steps_per_second": 2.905, |
| "step": 28000 |
| }, |
| { |
| "epoch": 8.94, |
| "grad_norm": 5.596973364224968, |
| "learning_rate": 1e-05, |
| "loss": 0.0699, |
| "step": 28010 |
| }, |
| { |
| "epoch": 8.94, |
| "grad_norm": 3.883597895133324, |
| "learning_rate": 1e-05, |
| "loss": 0.0742, |
| "step": 28020 |
| }, |
| { |
| "epoch": 8.94, |
| "grad_norm": 5.402544107535107, |
| "learning_rate": 1e-05, |
| "loss": 0.0808, |
| "step": 28030 |
| }, |
| { |
| "epoch": 8.95, |
| "grad_norm": 4.748911281095162, |
| "learning_rate": 1e-05, |
| "loss": 0.0619, |
| "step": 28040 |
| }, |
| { |
| "epoch": 8.95, |
| "grad_norm": 6.07548061400595, |
| "learning_rate": 1e-05, |
| "loss": 0.0878, |
| "step": 28050 |
| }, |
| { |
| "epoch": 8.95, |
| "grad_norm": 7.439756932543699, |
| "learning_rate": 1e-05, |
| "loss": 0.0836, |
| "step": 28060 |
| }, |
| { |
| "epoch": 8.96, |
| "grad_norm": 4.224146351547493, |
| "learning_rate": 1e-05, |
| "loss": 0.0701, |
| "step": 28070 |
| }, |
| { |
| "epoch": 8.96, |
| "grad_norm": 4.016651067004375, |
| "learning_rate": 1e-05, |
| "loss": 0.074, |
| "step": 28080 |
| }, |
| { |
| "epoch": 8.96, |
| "grad_norm": 4.5908646637203, |
| "learning_rate": 1e-05, |
| "loss": 0.0794, |
| "step": 28090 |
| }, |
| { |
| "epoch": 8.97, |
| "grad_norm": 3.2843360557388714, |
| "learning_rate": 1e-05, |
| "loss": 0.0676, |
| "step": 28100 |
| }, |
| { |
| "epoch": 8.97, |
| "grad_norm": 4.552499340377835, |
| "learning_rate": 1e-05, |
| "loss": 0.0785, |
| "step": 28110 |
| }, |
| { |
| "epoch": 8.97, |
| "grad_norm": 2.624794724930401, |
| "learning_rate": 1e-05, |
| "loss": 0.0743, |
| "step": 28120 |
| }, |
| { |
| "epoch": 8.98, |
| "grad_norm": 2.46827364804548, |
| "learning_rate": 1e-05, |
| "loss": 0.0765, |
| "step": 28130 |
| }, |
| { |
| "epoch": 8.98, |
| "grad_norm": 4.688693492267191, |
| "learning_rate": 1e-05, |
| "loss": 0.0914, |
| "step": 28140 |
| }, |
| { |
| "epoch": 8.98, |
| "grad_norm": 4.926314806798647, |
| "learning_rate": 1e-05, |
| "loss": 0.0788, |
| "step": 28150 |
| }, |
| { |
| "epoch": 8.99, |
| "grad_norm": 4.318560115542867, |
| "learning_rate": 1e-05, |
| "loss": 0.0789, |
| "step": 28160 |
| }, |
| { |
| "epoch": 8.99, |
| "grad_norm": 6.849670914590247, |
| "learning_rate": 1e-05, |
| "loss": 0.0784, |
| "step": 28170 |
| }, |
| { |
| "epoch": 8.99, |
| "grad_norm": 3.7800419666846903, |
| "learning_rate": 1e-05, |
| "loss": 0.0717, |
| "step": 28180 |
| }, |
| { |
| "epoch": 8.99, |
| "grad_norm": 8.023192642472363, |
| "learning_rate": 1e-05, |
| "loss": 0.0813, |
| "step": 28190 |
| }, |
| { |
| "epoch": 9.0, |
| "grad_norm": 4.187002123941266, |
| "learning_rate": 1e-05, |
| "loss": 0.0764, |
| "step": 28200 |
| }, |
| { |
| "epoch": 9.0, |
| "grad_norm": 2.8980998950534786, |
| "learning_rate": 1e-05, |
| "loss": 0.0528, |
| "step": 28210 |
| }, |
| { |
| "epoch": 9.0, |
| "grad_norm": 3.2436814141051116, |
| "learning_rate": 1e-05, |
| "loss": 0.0396, |
| "step": 28220 |
| }, |
| { |
| "epoch": 9.01, |
| "grad_norm": 4.661360221412794, |
| "learning_rate": 1e-05, |
| "loss": 0.0472, |
| "step": 28230 |
| }, |
| { |
| "epoch": 9.01, |
| "grad_norm": 2.5824579991399235, |
| "learning_rate": 1e-05, |
| "loss": 0.0378, |
| "step": 28240 |
| }, |
| { |
| "epoch": 9.01, |
| "grad_norm": 4.238656107414688, |
| "learning_rate": 1e-05, |
| "loss": 0.0495, |
| "step": 28250 |
| }, |
| { |
| "epoch": 9.02, |
| "grad_norm": 3.738676012455591, |
| "learning_rate": 1e-05, |
| "loss": 0.0649, |
| "step": 28260 |
| }, |
| { |
| "epoch": 9.02, |
| "grad_norm": 4.5315894690323315, |
| "learning_rate": 1e-05, |
| "loss": 0.0462, |
| "step": 28270 |
| }, |
| { |
| "epoch": 9.02, |
| "grad_norm": 6.705044273782158, |
| "learning_rate": 1e-05, |
| "loss": 0.0688, |
| "step": 28280 |
| }, |
| { |
| "epoch": 9.03, |
| "grad_norm": 5.213058383224686, |
| "learning_rate": 1e-05, |
| "loss": 0.0386, |
| "step": 28290 |
| }, |
| { |
| "epoch": 9.03, |
| "grad_norm": 4.6483568713464285, |
| "learning_rate": 1e-05, |
| "loss": 0.0745, |
| "step": 28300 |
| }, |
| { |
| "epoch": 9.03, |
| "grad_norm": 3.7259683515235573, |
| "learning_rate": 1e-05, |
| "loss": 0.047, |
| "step": 28310 |
| }, |
| { |
| "epoch": 9.04, |
| "grad_norm": 5.642031169406345, |
| "learning_rate": 1e-05, |
| "loss": 0.0476, |
| "step": 28320 |
| }, |
| { |
| "epoch": 9.04, |
| "grad_norm": 3.848054414082872, |
| "learning_rate": 1e-05, |
| "loss": 0.0365, |
| "step": 28330 |
| }, |
| { |
| "epoch": 9.04, |
| "grad_norm": 3.8905961981630326, |
| "learning_rate": 1e-05, |
| "loss": 0.0499, |
| "step": 28340 |
| }, |
| { |
| "epoch": 9.05, |
| "grad_norm": 4.878231444776744, |
| "learning_rate": 1e-05, |
| "loss": 0.0478, |
| "step": 28350 |
| }, |
| { |
| "epoch": 9.05, |
| "grad_norm": 5.806049007554535, |
| "learning_rate": 1e-05, |
| "loss": 0.0682, |
| "step": 28360 |
| }, |
| { |
| "epoch": 9.05, |
| "grad_norm": 5.342783879622176, |
| "learning_rate": 1e-05, |
| "loss": 0.0537, |
| "step": 28370 |
| }, |
| { |
| "epoch": 9.06, |
| "grad_norm": 2.3353700490808302, |
| "learning_rate": 1e-05, |
| "loss": 0.0493, |
| "step": 28380 |
| }, |
| { |
| "epoch": 9.06, |
| "grad_norm": 3.191959328509434, |
| "learning_rate": 1e-05, |
| "loss": 0.0479, |
| "step": 28390 |
| }, |
| { |
| "epoch": 9.06, |
| "grad_norm": 4.536363124030792, |
| "learning_rate": 1e-05, |
| "loss": 0.051, |
| "step": 28400 |
| }, |
| { |
| "epoch": 9.07, |
| "grad_norm": 2.493544062856761, |
| "learning_rate": 1e-05, |
| "loss": 0.0493, |
| "step": 28410 |
| }, |
| { |
| "epoch": 9.07, |
| "grad_norm": 2.750529325000551, |
| "learning_rate": 1e-05, |
| "loss": 0.0395, |
| "step": 28420 |
| }, |
| { |
| "epoch": 9.07, |
| "grad_norm": 4.140715054666335, |
| "learning_rate": 1e-05, |
| "loss": 0.034, |
| "step": 28430 |
| }, |
| { |
| "epoch": 9.07, |
| "grad_norm": 3.958126839053499, |
| "learning_rate": 1e-05, |
| "loss": 0.0575, |
| "step": 28440 |
| }, |
| { |
| "epoch": 9.08, |
| "grad_norm": 3.4199708143182415, |
| "learning_rate": 1e-05, |
| "loss": 0.0536, |
| "step": 28450 |
| }, |
| { |
| "epoch": 9.08, |
| "grad_norm": 2.7201084820768378, |
| "learning_rate": 1e-05, |
| "loss": 0.0369, |
| "step": 28460 |
| }, |
| { |
| "epoch": 9.08, |
| "grad_norm": 5.0209250329640955, |
| "learning_rate": 1e-05, |
| "loss": 0.0467, |
| "step": 28470 |
| }, |
| { |
| "epoch": 9.09, |
| "grad_norm": 4.624875041201493, |
| "learning_rate": 1e-05, |
| "loss": 0.0503, |
| "step": 28480 |
| }, |
| { |
| "epoch": 9.09, |
| "grad_norm": 5.771978748919517, |
| "learning_rate": 1e-05, |
| "loss": 0.0579, |
| "step": 28490 |
| }, |
| { |
| "epoch": 9.09, |
| "grad_norm": 3.6154554269601964, |
| "learning_rate": 1e-05, |
| "loss": 0.0669, |
| "step": 28500 |
| }, |
| { |
| "epoch": 9.09, |
| "eval_loss": 1.1923828125, |
| "eval_runtime": 54.8192, |
| "eval_samples_per_second": 46.225, |
| "eval_steps_per_second": 2.9, |
| "step": 28500 |
| }, |
| { |
| "epoch": 9.1, |
| "grad_norm": 5.494365233255453, |
| "learning_rate": 1e-05, |
| "loss": 0.0611, |
| "step": 28510 |
| }, |
| { |
| "epoch": 9.1, |
| "grad_norm": 6.102364578963621, |
| "learning_rate": 1e-05, |
| "loss": 0.0581, |
| "step": 28520 |
| }, |
| { |
| "epoch": 9.1, |
| "grad_norm": 2.371197417002467, |
| "learning_rate": 1e-05, |
| "loss": 0.0557, |
| "step": 28530 |
| }, |
| { |
| "epoch": 9.11, |
| "grad_norm": 3.6795259023118305, |
| "learning_rate": 1e-05, |
| "loss": 0.0514, |
| "step": 28540 |
| }, |
| { |
| "epoch": 9.11, |
| "grad_norm": 4.721151385096752, |
| "learning_rate": 1e-05, |
| "loss": 0.0546, |
| "step": 28550 |
| }, |
| { |
| "epoch": 9.11, |
| "grad_norm": 3.3934979199388637, |
| "learning_rate": 1e-05, |
| "loss": 0.0524, |
| "step": 28560 |
| }, |
| { |
| "epoch": 9.12, |
| "grad_norm": 3.9418213156000173, |
| "learning_rate": 1e-05, |
| "loss": 0.0425, |
| "step": 28570 |
| }, |
| { |
| "epoch": 9.12, |
| "grad_norm": 4.505432029271526, |
| "learning_rate": 1e-05, |
| "loss": 0.0418, |
| "step": 28580 |
| }, |
| { |
| "epoch": 9.12, |
| "grad_norm": 3.425883553694451, |
| "learning_rate": 1e-05, |
| "loss": 0.07, |
| "step": 28590 |
| }, |
| { |
| "epoch": 9.13, |
| "grad_norm": 2.0156116337296037, |
| "learning_rate": 1e-05, |
| "loss": 0.0475, |
| "step": 28600 |
| }, |
| { |
| "epoch": 9.13, |
| "grad_norm": 4.136224675773965, |
| "learning_rate": 1e-05, |
| "loss": 0.064, |
| "step": 28610 |
| }, |
| { |
| "epoch": 9.13, |
| "grad_norm": 4.381720476561324, |
| "learning_rate": 1e-05, |
| "loss": 0.0546, |
| "step": 28620 |
| }, |
| { |
| "epoch": 9.14, |
| "grad_norm": 4.732541172918535, |
| "learning_rate": 1e-05, |
| "loss": 0.048, |
| "step": 28630 |
| }, |
| { |
| "epoch": 9.14, |
| "grad_norm": 8.795768708883594, |
| "learning_rate": 1e-05, |
| "loss": 0.0705, |
| "step": 28640 |
| }, |
| { |
| "epoch": 9.14, |
| "grad_norm": 3.8855983633693354, |
| "learning_rate": 1e-05, |
| "loss": 0.0384, |
| "step": 28650 |
| }, |
| { |
| "epoch": 9.14, |
| "grad_norm": 3.7129750008906295, |
| "learning_rate": 1e-05, |
| "loss": 0.0642, |
| "step": 28660 |
| }, |
| { |
| "epoch": 9.15, |
| "grad_norm": 3.7169615989181315, |
| "learning_rate": 1e-05, |
| "loss": 0.0612, |
| "step": 28670 |
| }, |
| { |
| "epoch": 9.15, |
| "grad_norm": 4.40422585170007, |
| "learning_rate": 1e-05, |
| "loss": 0.0416, |
| "step": 28680 |
| }, |
| { |
| "epoch": 9.15, |
| "grad_norm": 4.515250114534272, |
| "learning_rate": 1e-05, |
| "loss": 0.0589, |
| "step": 28690 |
| }, |
| { |
| "epoch": 9.16, |
| "grad_norm": 3.4513983228534495, |
| "learning_rate": 1e-05, |
| "loss": 0.047, |
| "step": 28700 |
| }, |
| { |
| "epoch": 9.16, |
| "grad_norm": 3.0981474170716465, |
| "learning_rate": 1e-05, |
| "loss": 0.0632, |
| "step": 28710 |
| }, |
| { |
| "epoch": 9.16, |
| "grad_norm": 3.5126593394130357, |
| "learning_rate": 1e-05, |
| "loss": 0.0518, |
| "step": 28720 |
| }, |
| { |
| "epoch": 9.17, |
| "grad_norm": 4.93543673514326, |
| "learning_rate": 1e-05, |
| "loss": 0.0631, |
| "step": 28730 |
| }, |
| { |
| "epoch": 9.17, |
| "grad_norm": 4.35670526169773, |
| "learning_rate": 1e-05, |
| "loss": 0.0493, |
| "step": 28740 |
| }, |
| { |
| "epoch": 9.17, |
| "grad_norm": 4.14601561690323, |
| "learning_rate": 1e-05, |
| "loss": 0.0499, |
| "step": 28750 |
| }, |
| { |
| "epoch": 9.18, |
| "grad_norm": 2.375784593784898, |
| "learning_rate": 1e-05, |
| "loss": 0.0522, |
| "step": 28760 |
| }, |
| { |
| "epoch": 9.18, |
| "grad_norm": 3.4352147222092397, |
| "learning_rate": 1e-05, |
| "loss": 0.057, |
| "step": 28770 |
| }, |
| { |
| "epoch": 9.18, |
| "grad_norm": 3.1194324488270957, |
| "learning_rate": 1e-05, |
| "loss": 0.0606, |
| "step": 28780 |
| }, |
| { |
| "epoch": 9.19, |
| "grad_norm": 3.2727141753329594, |
| "learning_rate": 1e-05, |
| "loss": 0.0571, |
| "step": 28790 |
| }, |
| { |
| "epoch": 9.19, |
| "grad_norm": 4.607931654648443, |
| "learning_rate": 1e-05, |
| "loss": 0.057, |
| "step": 28800 |
| }, |
| { |
| "epoch": 9.19, |
| "grad_norm": 3.5680625721101995, |
| "learning_rate": 1e-05, |
| "loss": 0.0409, |
| "step": 28810 |
| }, |
| { |
| "epoch": 9.2, |
| "grad_norm": 4.558216540265952, |
| "learning_rate": 1e-05, |
| "loss": 0.0624, |
| "step": 28820 |
| }, |
| { |
| "epoch": 9.2, |
| "grad_norm": 3.111519385599725, |
| "learning_rate": 1e-05, |
| "loss": 0.0402, |
| "step": 28830 |
| }, |
| { |
| "epoch": 9.2, |
| "grad_norm": 3.8843504778332703, |
| "learning_rate": 1e-05, |
| "loss": 0.037, |
| "step": 28840 |
| }, |
| { |
| "epoch": 9.21, |
| "grad_norm": 5.88228658526524, |
| "learning_rate": 1e-05, |
| "loss": 0.0594, |
| "step": 28850 |
| }, |
| { |
| "epoch": 9.21, |
| "grad_norm": 4.053332269526225, |
| "learning_rate": 1e-05, |
| "loss": 0.0411, |
| "step": 28860 |
| }, |
| { |
| "epoch": 9.21, |
| "grad_norm": 2.6861531630963995, |
| "learning_rate": 1e-05, |
| "loss": 0.0434, |
| "step": 28870 |
| }, |
| { |
| "epoch": 9.22, |
| "grad_norm": 2.285059948057609, |
| "learning_rate": 1e-05, |
| "loss": 0.0536, |
| "step": 28880 |
| }, |
| { |
| "epoch": 9.22, |
| "grad_norm": 2.9654624002015737, |
| "learning_rate": 1e-05, |
| "loss": 0.0527, |
| "step": 28890 |
| }, |
| { |
| "epoch": 9.22, |
| "grad_norm": 2.951126310348947, |
| "learning_rate": 1e-05, |
| "loss": 0.0556, |
| "step": 28900 |
| }, |
| { |
| "epoch": 9.22, |
| "grad_norm": 5.320237572320218, |
| "learning_rate": 1e-05, |
| "loss": 0.05, |
| "step": 28910 |
| }, |
| { |
| "epoch": 9.23, |
| "grad_norm": 4.535328890515174, |
| "learning_rate": 1e-05, |
| "loss": 0.0744, |
| "step": 28920 |
| }, |
| { |
| "epoch": 9.23, |
| "grad_norm": 3.2217285300556457, |
| "learning_rate": 1e-05, |
| "loss": 0.0512, |
| "step": 28930 |
| }, |
| { |
| "epoch": 9.23, |
| "grad_norm": 5.575203950854886, |
| "learning_rate": 1e-05, |
| "loss": 0.0696, |
| "step": 28940 |
| }, |
| { |
| "epoch": 9.24, |
| "grad_norm": 3.52450727982675, |
| "learning_rate": 1e-05, |
| "loss": 0.0583, |
| "step": 28950 |
| }, |
| { |
| "epoch": 9.24, |
| "grad_norm": 3.6780912982754668, |
| "learning_rate": 1e-05, |
| "loss": 0.0455, |
| "step": 28960 |
| }, |
| { |
| "epoch": 9.24, |
| "grad_norm": 2.715367108951744, |
| "learning_rate": 1e-05, |
| "loss": 0.053, |
| "step": 28970 |
| }, |
| { |
| "epoch": 9.25, |
| "grad_norm": 4.173900098452328, |
| "learning_rate": 1e-05, |
| "loss": 0.0583, |
| "step": 28980 |
| }, |
| { |
| "epoch": 9.25, |
| "grad_norm": 4.046610348564629, |
| "learning_rate": 1e-05, |
| "loss": 0.0545, |
| "step": 28990 |
| }, |
| { |
| "epoch": 9.25, |
| "grad_norm": 4.3019975679852696, |
| "learning_rate": 1e-05, |
| "loss": 0.0678, |
| "step": 29000 |
| }, |
| { |
| "epoch": 9.25, |
| "eval_loss": 1.1708984375, |
| "eval_runtime": 54.7495, |
| "eval_samples_per_second": 46.284, |
| "eval_steps_per_second": 2.904, |
| "step": 29000 |
| }, |
| { |
| "epoch": 9.26, |
| "grad_norm": 5.725900983918314, |
| "learning_rate": 1e-05, |
| "loss": 0.0633, |
| "step": 29010 |
| }, |
| { |
| "epoch": 9.26, |
| "grad_norm": 4.090546497498025, |
| "learning_rate": 1e-05, |
| "loss": 0.0589, |
| "step": 29020 |
| }, |
| { |
| "epoch": 9.26, |
| "grad_norm": 3.345316867388993, |
| "learning_rate": 1e-05, |
| "loss": 0.0577, |
| "step": 29030 |
| }, |
| { |
| "epoch": 9.27, |
| "grad_norm": 5.367234764279701, |
| "learning_rate": 1e-05, |
| "loss": 0.0703, |
| "step": 29040 |
| }, |
| { |
| "epoch": 9.27, |
| "grad_norm": 2.4534084860893994, |
| "learning_rate": 1e-05, |
| "loss": 0.0544, |
| "step": 29050 |
| }, |
| { |
| "epoch": 9.27, |
| "grad_norm": 5.258499168686829, |
| "learning_rate": 1e-05, |
| "loss": 0.0669, |
| "step": 29060 |
| }, |
| { |
| "epoch": 9.28, |
| "grad_norm": 4.148058398035247, |
| "learning_rate": 1e-05, |
| "loss": 0.0484, |
| "step": 29070 |
| }, |
| { |
| "epoch": 9.28, |
| "grad_norm": 5.74639812936605, |
| "learning_rate": 1e-05, |
| "loss": 0.0451, |
| "step": 29080 |
| }, |
| { |
| "epoch": 9.28, |
| "grad_norm": 3.6366038893314174, |
| "learning_rate": 1e-05, |
| "loss": 0.0689, |
| "step": 29090 |
| }, |
| { |
| "epoch": 9.29, |
| "grad_norm": 1.4255387609504584, |
| "learning_rate": 1e-05, |
| "loss": 0.0585, |
| "step": 29100 |
| }, |
| { |
| "epoch": 9.29, |
| "grad_norm": 4.093059452433148, |
| "learning_rate": 1e-05, |
| "loss": 0.0576, |
| "step": 29110 |
| }, |
| { |
| "epoch": 9.29, |
| "grad_norm": 4.233515690958381, |
| "learning_rate": 1e-05, |
| "loss": 0.0637, |
| "step": 29120 |
| }, |
| { |
| "epoch": 9.29, |
| "grad_norm": 3.8067989561094935, |
| "learning_rate": 1e-05, |
| "loss": 0.052, |
| "step": 29130 |
| }, |
| { |
| "epoch": 9.3, |
| "grad_norm": 3.336604388935944, |
| "learning_rate": 1e-05, |
| "loss": 0.0589, |
| "step": 29140 |
| }, |
| { |
| "epoch": 9.3, |
| "grad_norm": 2.5591966690330374, |
| "learning_rate": 1e-05, |
| "loss": 0.0646, |
| "step": 29150 |
| }, |
| { |
| "epoch": 9.3, |
| "grad_norm": 4.363257426006467, |
| "learning_rate": 1e-05, |
| "loss": 0.0642, |
| "step": 29160 |
| }, |
| { |
| "epoch": 9.31, |
| "grad_norm": 3.8150834882607616, |
| "learning_rate": 1e-05, |
| "loss": 0.041, |
| "step": 29170 |
| }, |
| { |
| "epoch": 9.31, |
| "grad_norm": 2.090809365483953, |
| "learning_rate": 1e-05, |
| "loss": 0.0509, |
| "step": 29180 |
| }, |
| { |
| "epoch": 9.31, |
| "grad_norm": 3.9237685780542635, |
| "learning_rate": 1e-05, |
| "loss": 0.0637, |
| "step": 29190 |
| }, |
| { |
| "epoch": 9.32, |
| "grad_norm": 4.671928482004706, |
| "learning_rate": 1e-05, |
| "loss": 0.0608, |
| "step": 29200 |
| }, |
| { |
| "epoch": 9.32, |
| "grad_norm": 2.4234669037344196, |
| "learning_rate": 1e-05, |
| "loss": 0.0674, |
| "step": 29210 |
| }, |
| { |
| "epoch": 9.32, |
| "grad_norm": 6.837659381354828, |
| "learning_rate": 1e-05, |
| "loss": 0.0556, |
| "step": 29220 |
| }, |
| { |
| "epoch": 9.33, |
| "grad_norm": 4.854628621442018, |
| "learning_rate": 1e-05, |
| "loss": 0.0435, |
| "step": 29230 |
| }, |
| { |
| "epoch": 9.33, |
| "grad_norm": 2.7077448181040773, |
| "learning_rate": 1e-05, |
| "loss": 0.0605, |
| "step": 29240 |
| }, |
| { |
| "epoch": 9.33, |
| "grad_norm": 7.182299249952562, |
| "learning_rate": 1e-05, |
| "loss": 0.0494, |
| "step": 29250 |
| }, |
| { |
| "epoch": 9.34, |
| "grad_norm": 4.786547748891365, |
| "learning_rate": 1e-05, |
| "loss": 0.0525, |
| "step": 29260 |
| }, |
| { |
| "epoch": 9.34, |
| "grad_norm": 5.8188792672226555, |
| "learning_rate": 1e-05, |
| "loss": 0.0764, |
| "step": 29270 |
| }, |
| { |
| "epoch": 9.34, |
| "grad_norm": 3.9681186924840848, |
| "learning_rate": 1e-05, |
| "loss": 0.0558, |
| "step": 29280 |
| }, |
| { |
| "epoch": 9.35, |
| "grad_norm": 7.800400787229221, |
| "learning_rate": 1e-05, |
| "loss": 0.0541, |
| "step": 29290 |
| }, |
| { |
| "epoch": 9.35, |
| "grad_norm": 3.3098673885552845, |
| "learning_rate": 1e-05, |
| "loss": 0.0538, |
| "step": 29300 |
| }, |
| { |
| "epoch": 9.35, |
| "grad_norm": 2.97662001026094, |
| "learning_rate": 1e-05, |
| "loss": 0.0701, |
| "step": 29310 |
| }, |
| { |
| "epoch": 9.36, |
| "grad_norm": 4.413146121346368, |
| "learning_rate": 1e-05, |
| "loss": 0.0654, |
| "step": 29320 |
| }, |
| { |
| "epoch": 9.36, |
| "grad_norm": 5.045875473959504, |
| "learning_rate": 1e-05, |
| "loss": 0.0833, |
| "step": 29330 |
| }, |
| { |
| "epoch": 9.36, |
| "grad_norm": 4.3024300473260215, |
| "learning_rate": 1e-05, |
| "loss": 0.0697, |
| "step": 29340 |
| }, |
| { |
| "epoch": 9.37, |
| "grad_norm": 4.510875699102549, |
| "learning_rate": 1e-05, |
| "loss": 0.0645, |
| "step": 29350 |
| }, |
| { |
| "epoch": 9.37, |
| "grad_norm": 6.575421210311745, |
| "learning_rate": 1e-05, |
| "loss": 0.0699, |
| "step": 29360 |
| }, |
| { |
| "epoch": 9.37, |
| "grad_norm": 5.086900657948119, |
| "learning_rate": 1e-05, |
| "loss": 0.0642, |
| "step": 29370 |
| }, |
| { |
| "epoch": 9.37, |
| "grad_norm": 4.420859284224401, |
| "learning_rate": 1e-05, |
| "loss": 0.0696, |
| "step": 29380 |
| }, |
| { |
| "epoch": 9.38, |
| "grad_norm": 3.3483259316858334, |
| "learning_rate": 1e-05, |
| "loss": 0.0502, |
| "step": 29390 |
| }, |
| { |
| "epoch": 9.38, |
| "grad_norm": 4.317008495174883, |
| "learning_rate": 1e-05, |
| "loss": 0.0672, |
| "step": 29400 |
| }, |
| { |
| "epoch": 9.38, |
| "grad_norm": 4.981269849787928, |
| "learning_rate": 1e-05, |
| "loss": 0.05, |
| "step": 29410 |
| }, |
| { |
| "epoch": 9.39, |
| "grad_norm": 6.335421502377957, |
| "learning_rate": 1e-05, |
| "loss": 0.0536, |
| "step": 29420 |
| }, |
| { |
| "epoch": 9.39, |
| "grad_norm": 3.736520829044552, |
| "learning_rate": 1e-05, |
| "loss": 0.0557, |
| "step": 29430 |
| }, |
| { |
| "epoch": 9.39, |
| "grad_norm": 3.4726035218156652, |
| "learning_rate": 1e-05, |
| "loss": 0.0564, |
| "step": 29440 |
| }, |
| { |
| "epoch": 9.4, |
| "grad_norm": 2.978565893530726, |
| "learning_rate": 1e-05, |
| "loss": 0.074, |
| "step": 29450 |
| }, |
| { |
| "epoch": 9.4, |
| "grad_norm": 5.450251043294843, |
| "learning_rate": 1e-05, |
| "loss": 0.0725, |
| "step": 29460 |
| }, |
| { |
| "epoch": 9.4, |
| "grad_norm": 5.30955028547699, |
| "learning_rate": 1e-05, |
| "loss": 0.0574, |
| "step": 29470 |
| }, |
| { |
| "epoch": 9.41, |
| "grad_norm": 2.6338691741674713, |
| "learning_rate": 1e-05, |
| "loss": 0.0494, |
| "step": 29480 |
| }, |
| { |
| "epoch": 9.41, |
| "grad_norm": 3.4814207956272085, |
| "learning_rate": 1e-05, |
| "loss": 0.0614, |
| "step": 29490 |
| }, |
| { |
| "epoch": 9.41, |
| "grad_norm": 3.294995101901916, |
| "learning_rate": 1e-05, |
| "loss": 0.0531, |
| "step": 29500 |
| }, |
| { |
| "epoch": 9.41, |
| "eval_loss": 1.197265625, |
| "eval_runtime": 54.6853, |
| "eval_samples_per_second": 46.338, |
| "eval_steps_per_second": 2.908, |
| "step": 29500 |
| }, |
| { |
| "epoch": 9.42, |
| "grad_norm": 3.681857301523477, |
| "learning_rate": 1e-05, |
| "loss": 0.0786, |
| "step": 29510 |
| }, |
| { |
| "epoch": 9.42, |
| "grad_norm": 7.505845652797761, |
| "learning_rate": 1e-05, |
| "loss": 0.0885, |
| "step": 29520 |
| }, |
| { |
| "epoch": 9.42, |
| "grad_norm": 2.0678872928637935, |
| "learning_rate": 1e-05, |
| "loss": 0.0615, |
| "step": 29530 |
| }, |
| { |
| "epoch": 9.43, |
| "grad_norm": 3.678763174279046, |
| "learning_rate": 1e-05, |
| "loss": 0.047, |
| "step": 29540 |
| }, |
| { |
| "epoch": 9.43, |
| "grad_norm": 3.2786967198081856, |
| "learning_rate": 1e-05, |
| "loss": 0.0658, |
| "step": 29550 |
| }, |
| { |
| "epoch": 9.43, |
| "grad_norm": 4.013911137561376, |
| "learning_rate": 1e-05, |
| "loss": 0.045, |
| "step": 29560 |
| }, |
| { |
| "epoch": 9.44, |
| "grad_norm": 2.9887978898778162, |
| "learning_rate": 1e-05, |
| "loss": 0.0576, |
| "step": 29570 |
| }, |
| { |
| "epoch": 9.44, |
| "grad_norm": 5.317696038610858, |
| "learning_rate": 1e-05, |
| "loss": 0.0525, |
| "step": 29580 |
| }, |
| { |
| "epoch": 9.44, |
| "grad_norm": 3.7604385051286893, |
| "learning_rate": 1e-05, |
| "loss": 0.0452, |
| "step": 29590 |
| }, |
| { |
| "epoch": 9.44, |
| "grad_norm": 4.801019623862304, |
| "learning_rate": 1e-05, |
| "loss": 0.0697, |
| "step": 29600 |
| }, |
| { |
| "epoch": 9.45, |
| "grad_norm": 4.088178264560775, |
| "learning_rate": 1e-05, |
| "loss": 0.0549, |
| "step": 29610 |
| }, |
| { |
| "epoch": 9.45, |
| "grad_norm": 5.152970814523601, |
| "learning_rate": 1e-05, |
| "loss": 0.0597, |
| "step": 29620 |
| }, |
| { |
| "epoch": 9.45, |
| "grad_norm": 4.668908647857242, |
| "learning_rate": 1e-05, |
| "loss": 0.0551, |
| "step": 29630 |
| }, |
| { |
| "epoch": 9.46, |
| "grad_norm": 4.14137117392239, |
| "learning_rate": 1e-05, |
| "loss": 0.0538, |
| "step": 29640 |
| }, |
| { |
| "epoch": 9.46, |
| "grad_norm": 4.091420683034978, |
| "learning_rate": 1e-05, |
| "loss": 0.0803, |
| "step": 29650 |
| }, |
| { |
| "epoch": 9.46, |
| "grad_norm": 2.801329831675696, |
| "learning_rate": 1e-05, |
| "loss": 0.0634, |
| "step": 29660 |
| }, |
| { |
| "epoch": 9.47, |
| "grad_norm": 6.007510570781931, |
| "learning_rate": 1e-05, |
| "loss": 0.064, |
| "step": 29670 |
| }, |
| { |
| "epoch": 9.47, |
| "grad_norm": 1.5214332111653759, |
| "learning_rate": 1e-05, |
| "loss": 0.0491, |
| "step": 29680 |
| }, |
| { |
| "epoch": 9.47, |
| "grad_norm": 7.664190168804759, |
| "learning_rate": 1e-05, |
| "loss": 0.0646, |
| "step": 29690 |
| }, |
| { |
| "epoch": 9.48, |
| "grad_norm": 3.398370711448597, |
| "learning_rate": 1e-05, |
| "loss": 0.0834, |
| "step": 29700 |
| }, |
| { |
| "epoch": 9.48, |
| "grad_norm": 4.525827886392828, |
| "learning_rate": 1e-05, |
| "loss": 0.073, |
| "step": 29710 |
| }, |
| { |
| "epoch": 9.48, |
| "grad_norm": 6.754686494748219, |
| "learning_rate": 1e-05, |
| "loss": 0.0624, |
| "step": 29720 |
| }, |
| { |
| "epoch": 9.49, |
| "grad_norm": 3.4588069514861517, |
| "learning_rate": 1e-05, |
| "loss": 0.0756, |
| "step": 29730 |
| }, |
| { |
| "epoch": 9.49, |
| "grad_norm": 9.719346987919613, |
| "learning_rate": 1e-05, |
| "loss": 0.0665, |
| "step": 29740 |
| }, |
| { |
| "epoch": 9.49, |
| "grad_norm": 3.652380175077389, |
| "learning_rate": 1e-05, |
| "loss": 0.0637, |
| "step": 29750 |
| }, |
| { |
| "epoch": 9.5, |
| "grad_norm": 4.144207480652989, |
| "learning_rate": 1e-05, |
| "loss": 0.0635, |
| "step": 29760 |
| }, |
| { |
| "epoch": 9.5, |
| "grad_norm": 4.957611651218937, |
| "learning_rate": 1e-05, |
| "loss": 0.069, |
| "step": 29770 |
| }, |
| { |
| "epoch": 9.5, |
| "grad_norm": 3.4620504204457356, |
| "learning_rate": 1e-05, |
| "loss": 0.0606, |
| "step": 29780 |
| }, |
| { |
| "epoch": 9.51, |
| "grad_norm": 6.70576762858104, |
| "learning_rate": 1e-05, |
| "loss": 0.0586, |
| "step": 29790 |
| }, |
| { |
| "epoch": 9.51, |
| "grad_norm": 4.859546719672393, |
| "learning_rate": 1e-05, |
| "loss": 0.054, |
| "step": 29800 |
| }, |
| { |
| "epoch": 9.51, |
| "grad_norm": 6.718315074949154, |
| "learning_rate": 1e-05, |
| "loss": 0.0582, |
| "step": 29810 |
| }, |
| { |
| "epoch": 9.51, |
| "grad_norm": 3.7458414543722944, |
| "learning_rate": 1e-05, |
| "loss": 0.0678, |
| "step": 29820 |
| }, |
| { |
| "epoch": 9.52, |
| "grad_norm": 3.0767524873688283, |
| "learning_rate": 1e-05, |
| "loss": 0.0752, |
| "step": 29830 |
| }, |
| { |
| "epoch": 9.52, |
| "grad_norm": 4.558692911437555, |
| "learning_rate": 1e-05, |
| "loss": 0.0881, |
| "step": 29840 |
| }, |
| { |
| "epoch": 9.52, |
| "grad_norm": 2.9980603941572226, |
| "learning_rate": 1e-05, |
| "loss": 0.0632, |
| "step": 29850 |
| }, |
| { |
| "epoch": 9.53, |
| "grad_norm": 4.4235168212763964, |
| "learning_rate": 1e-05, |
| "loss": 0.0651, |
| "step": 29860 |
| }, |
| { |
| "epoch": 9.53, |
| "grad_norm": 9.823648327748971, |
| "learning_rate": 1e-05, |
| "loss": 0.0633, |
| "step": 29870 |
| }, |
| { |
| "epoch": 9.53, |
| "grad_norm": 7.325838622952393, |
| "learning_rate": 1e-05, |
| "loss": 0.0595, |
| "step": 29880 |
| }, |
| { |
| "epoch": 9.54, |
| "grad_norm": 2.8466865552554377, |
| "learning_rate": 1e-05, |
| "loss": 0.0571, |
| "step": 29890 |
| }, |
| { |
| "epoch": 9.54, |
| "grad_norm": 6.147075807659071, |
| "learning_rate": 1e-05, |
| "loss": 0.0612, |
| "step": 29900 |
| }, |
| { |
| "epoch": 9.54, |
| "grad_norm": 2.852841443180248, |
| "learning_rate": 1e-05, |
| "loss": 0.0567, |
| "step": 29910 |
| }, |
| { |
| "epoch": 9.55, |
| "grad_norm": 5.120567265616584, |
| "learning_rate": 1e-05, |
| "loss": 0.0616, |
| "step": 29920 |
| }, |
| { |
| "epoch": 9.55, |
| "grad_norm": 2.5653221592841526, |
| "learning_rate": 1e-05, |
| "loss": 0.0619, |
| "step": 29930 |
| }, |
| { |
| "epoch": 9.55, |
| "grad_norm": 4.79034476359495, |
| "learning_rate": 1e-05, |
| "loss": 0.0631, |
| "step": 29940 |
| }, |
| { |
| "epoch": 9.56, |
| "grad_norm": 4.496948691182492, |
| "learning_rate": 1e-05, |
| "loss": 0.0691, |
| "step": 29950 |
| }, |
| { |
| "epoch": 9.56, |
| "grad_norm": 8.096082667897463, |
| "learning_rate": 1e-05, |
| "loss": 0.0628, |
| "step": 29960 |
| }, |
| { |
| "epoch": 9.56, |
| "grad_norm": 5.447490765448661, |
| "learning_rate": 1e-05, |
| "loss": 0.0614, |
| "step": 29970 |
| }, |
| { |
| "epoch": 9.57, |
| "grad_norm": 5.374414988773905, |
| "learning_rate": 1e-05, |
| "loss": 0.0522, |
| "step": 29980 |
| }, |
| { |
| "epoch": 9.57, |
| "grad_norm": 6.647629864741532, |
| "learning_rate": 1e-05, |
| "loss": 0.0796, |
| "step": 29990 |
| }, |
| { |
| "epoch": 9.57, |
| "grad_norm": 4.308504548104834, |
| "learning_rate": 1e-05, |
| "loss": 0.0654, |
| "step": 30000 |
| }, |
| { |
| "epoch": 9.57, |
| "eval_loss": 1.1982421875, |
| "eval_runtime": 54.7538, |
| "eval_samples_per_second": 46.28, |
| "eval_steps_per_second": 2.904, |
| "step": 30000 |
| }, |
| { |
| "epoch": 9.58, |
| "grad_norm": 3.8896174217709563, |
| "learning_rate": 1e-05, |
| "loss": 0.0709, |
| "step": 30010 |
| }, |
| { |
| "epoch": 9.58, |
| "grad_norm": 6.798568339262543, |
| "learning_rate": 1e-05, |
| "loss": 0.0591, |
| "step": 30020 |
| }, |
| { |
| "epoch": 9.58, |
| "grad_norm": 4.928623187344225, |
| "learning_rate": 1e-05, |
| "loss": 0.0568, |
| "step": 30030 |
| }, |
| { |
| "epoch": 9.59, |
| "grad_norm": 5.566008072381535, |
| "learning_rate": 1e-05, |
| "loss": 0.0738, |
| "step": 30040 |
| }, |
| { |
| "epoch": 9.59, |
| "grad_norm": 4.049037517574465, |
| "learning_rate": 1e-05, |
| "loss": 0.0528, |
| "step": 30050 |
| }, |
| { |
| "epoch": 9.59, |
| "grad_norm": 5.148600504817122, |
| "learning_rate": 1e-05, |
| "loss": 0.0827, |
| "step": 30060 |
| }, |
| { |
| "epoch": 9.59, |
| "grad_norm": 9.136406027541556, |
| "learning_rate": 1e-05, |
| "loss": 0.0674, |
| "step": 30070 |
| }, |
| { |
| "epoch": 9.6, |
| "grad_norm": 6.991685015988132, |
| "learning_rate": 1e-05, |
| "loss": 0.07, |
| "step": 30080 |
| }, |
| { |
| "epoch": 9.6, |
| "grad_norm": 8.878554371189487, |
| "learning_rate": 1e-05, |
| "loss": 0.0825, |
| "step": 30090 |
| }, |
| { |
| "epoch": 9.6, |
| "grad_norm": 5.175500434599069, |
| "learning_rate": 1e-05, |
| "loss": 0.0506, |
| "step": 30100 |
| }, |
| { |
| "epoch": 9.61, |
| "grad_norm": 7.0763912814453125, |
| "learning_rate": 1e-05, |
| "loss": 0.0718, |
| "step": 30110 |
| }, |
| { |
| "epoch": 9.61, |
| "grad_norm": 3.146063352011234, |
| "learning_rate": 1e-05, |
| "loss": 0.0625, |
| "step": 30120 |
| }, |
| { |
| "epoch": 9.61, |
| "grad_norm": 3.6766364210538454, |
| "learning_rate": 1e-05, |
| "loss": 0.061, |
| "step": 30130 |
| }, |
| { |
| "epoch": 9.62, |
| "grad_norm": 5.192696967028004, |
| "learning_rate": 1e-05, |
| "loss": 0.0797, |
| "step": 30140 |
| }, |
| { |
| "epoch": 9.62, |
| "grad_norm": 4.45869691501134, |
| "learning_rate": 1e-05, |
| "loss": 0.0707, |
| "step": 30150 |
| }, |
| { |
| "epoch": 9.62, |
| "grad_norm": 5.036529614148935, |
| "learning_rate": 1e-05, |
| "loss": 0.0757, |
| "step": 30160 |
| }, |
| { |
| "epoch": 9.63, |
| "grad_norm": 4.885757900703045, |
| "learning_rate": 1e-05, |
| "loss": 0.0566, |
| "step": 30170 |
| }, |
| { |
| "epoch": 9.63, |
| "grad_norm": 3.3293940314108275, |
| "learning_rate": 1e-05, |
| "loss": 0.0706, |
| "step": 30180 |
| }, |
| { |
| "epoch": 9.63, |
| "grad_norm": 5.541475832911146, |
| "learning_rate": 1e-05, |
| "loss": 0.0698, |
| "step": 30190 |
| }, |
| { |
| "epoch": 9.64, |
| "grad_norm": 3.9009159137471148, |
| "learning_rate": 1e-05, |
| "loss": 0.0564, |
| "step": 30200 |
| }, |
| { |
| "epoch": 9.64, |
| "grad_norm": 4.159710812787411, |
| "learning_rate": 1e-05, |
| "loss": 0.0668, |
| "step": 30210 |
| }, |
| { |
| "epoch": 9.64, |
| "grad_norm": 3.6968872070110965, |
| "learning_rate": 1e-05, |
| "loss": 0.045, |
| "step": 30220 |
| }, |
| { |
| "epoch": 9.65, |
| "grad_norm": 5.7436000202509705, |
| "learning_rate": 1e-05, |
| "loss": 0.0715, |
| "step": 30230 |
| }, |
| { |
| "epoch": 9.65, |
| "grad_norm": 4.0987697430292505, |
| "learning_rate": 1e-05, |
| "loss": 0.0798, |
| "step": 30240 |
| }, |
| { |
| "epoch": 9.65, |
| "grad_norm": 5.015846699125768, |
| "learning_rate": 1e-05, |
| "loss": 0.067, |
| "step": 30250 |
| }, |
| { |
| "epoch": 9.66, |
| "grad_norm": 5.287732181005577, |
| "learning_rate": 1e-05, |
| "loss": 0.0636, |
| "step": 30260 |
| }, |
| { |
| "epoch": 9.66, |
| "grad_norm": 3.761687059754235, |
| "learning_rate": 1e-05, |
| "loss": 0.069, |
| "step": 30270 |
| }, |
| { |
| "epoch": 9.66, |
| "grad_norm": 3.842365046397892, |
| "learning_rate": 1e-05, |
| "loss": 0.0608, |
| "step": 30280 |
| }, |
| { |
| "epoch": 9.66, |
| "grad_norm": 3.629631241072155, |
| "learning_rate": 1e-05, |
| "loss": 0.0629, |
| "step": 30290 |
| }, |
| { |
| "epoch": 9.67, |
| "grad_norm": 7.003718342257617, |
| "learning_rate": 1e-05, |
| "loss": 0.0808, |
| "step": 30300 |
| }, |
| { |
| "epoch": 9.67, |
| "grad_norm": 8.140324621827466, |
| "learning_rate": 1e-05, |
| "loss": 0.0725, |
| "step": 30310 |
| }, |
| { |
| "epoch": 9.67, |
| "grad_norm": 4.8129529863665415, |
| "learning_rate": 1e-05, |
| "loss": 0.0553, |
| "step": 30320 |
| }, |
| { |
| "epoch": 9.68, |
| "grad_norm": 4.115225800522035, |
| "learning_rate": 1e-05, |
| "loss": 0.0605, |
| "step": 30330 |
| }, |
| { |
| "epoch": 9.68, |
| "grad_norm": 3.8523645397732347, |
| "learning_rate": 1e-05, |
| "loss": 0.0623, |
| "step": 30340 |
| }, |
| { |
| "epoch": 9.68, |
| "grad_norm": 6.665762394930085, |
| "learning_rate": 1e-05, |
| "loss": 0.0842, |
| "step": 30350 |
| }, |
| { |
| "epoch": 9.69, |
| "grad_norm": 6.607382904758711, |
| "learning_rate": 1e-05, |
| "loss": 0.0656, |
| "step": 30360 |
| }, |
| { |
| "epoch": 9.69, |
| "grad_norm": 4.817524454542726, |
| "learning_rate": 1e-05, |
| "loss": 0.0567, |
| "step": 30370 |
| }, |
| { |
| "epoch": 9.69, |
| "grad_norm": 6.701097017124148, |
| "learning_rate": 1e-05, |
| "loss": 0.0619, |
| "step": 30380 |
| }, |
| { |
| "epoch": 9.7, |
| "grad_norm": 5.9421506637326225, |
| "learning_rate": 1e-05, |
| "loss": 0.0594, |
| "step": 30390 |
| }, |
| { |
| "epoch": 9.7, |
| "grad_norm": 5.412033603651291, |
| "learning_rate": 1e-05, |
| "loss": 0.0724, |
| "step": 30400 |
| }, |
| { |
| "epoch": 9.7, |
| "grad_norm": 4.380425495051845, |
| "learning_rate": 1e-05, |
| "loss": 0.0541, |
| "step": 30410 |
| }, |
| { |
| "epoch": 9.71, |
| "grad_norm": 7.810256513810762, |
| "learning_rate": 1e-05, |
| "loss": 0.063, |
| "step": 30420 |
| }, |
| { |
| "epoch": 9.71, |
| "grad_norm": 5.580414801599444, |
| "learning_rate": 1e-05, |
| "loss": 0.0729, |
| "step": 30430 |
| }, |
| { |
| "epoch": 9.71, |
| "grad_norm": 8.066792609542599, |
| "learning_rate": 1e-05, |
| "loss": 0.0879, |
| "step": 30440 |
| }, |
| { |
| "epoch": 9.72, |
| "grad_norm": 3.8317608441403297, |
| "learning_rate": 1e-05, |
| "loss": 0.0572, |
| "step": 30450 |
| }, |
| { |
| "epoch": 9.72, |
| "grad_norm": 6.7911838601716745, |
| "learning_rate": 1e-05, |
| "loss": 0.0667, |
| "step": 30460 |
| }, |
| { |
| "epoch": 9.72, |
| "grad_norm": 6.103768432264369, |
| "learning_rate": 1e-05, |
| "loss": 0.063, |
| "step": 30470 |
| }, |
| { |
| "epoch": 9.73, |
| "grad_norm": 6.342514443928019, |
| "learning_rate": 1e-05, |
| "loss": 0.0767, |
| "step": 30480 |
| }, |
| { |
| "epoch": 9.73, |
| "grad_norm": 5.927926466718715, |
| "learning_rate": 1e-05, |
| "loss": 0.0802, |
| "step": 30490 |
| }, |
| { |
| "epoch": 9.73, |
| "grad_norm": 4.658370591198322, |
| "learning_rate": 1e-05, |
| "loss": 0.0599, |
| "step": 30500 |
| }, |
| { |
| "epoch": 9.73, |
| "eval_loss": 1.1611328125, |
| "eval_runtime": 54.8204, |
| "eval_samples_per_second": 46.224, |
| "eval_steps_per_second": 2.9, |
| "step": 30500 |
| }, |
| { |
| "epoch": 9.74, |
| "grad_norm": 7.332643592070997, |
| "learning_rate": 1e-05, |
| "loss": 0.0685, |
| "step": 30510 |
| }, |
| { |
| "epoch": 9.74, |
| "grad_norm": 5.160399460818401, |
| "learning_rate": 1e-05, |
| "loss": 0.0697, |
| "step": 30520 |
| }, |
| { |
| "epoch": 9.74, |
| "grad_norm": 3.710116582719881, |
| "learning_rate": 1e-05, |
| "loss": 0.0673, |
| "step": 30530 |
| }, |
| { |
| "epoch": 9.74, |
| "grad_norm": 4.907090619230188, |
| "learning_rate": 1e-05, |
| "loss": 0.0611, |
| "step": 30540 |
| }, |
| { |
| "epoch": 9.75, |
| "grad_norm": 4.859917220992294, |
| "learning_rate": 1e-05, |
| "loss": 0.0651, |
| "step": 30550 |
| }, |
| { |
| "epoch": 9.75, |
| "grad_norm": 6.103691559964321, |
| "learning_rate": 1e-05, |
| "loss": 0.0677, |
| "step": 30560 |
| }, |
| { |
| "epoch": 9.75, |
| "grad_norm": 1.466014993994252, |
| "learning_rate": 1e-05, |
| "loss": 0.0678, |
| "step": 30570 |
| }, |
| { |
| "epoch": 9.76, |
| "grad_norm": 3.8354580146402166, |
| "learning_rate": 1e-05, |
| "loss": 0.0782, |
| "step": 30580 |
| }, |
| { |
| "epoch": 9.76, |
| "grad_norm": 8.162617620347998, |
| "learning_rate": 1e-05, |
| "loss": 0.0733, |
| "step": 30590 |
| }, |
| { |
| "epoch": 9.76, |
| "grad_norm": 4.875112092123278, |
| "learning_rate": 1e-05, |
| "loss": 0.0576, |
| "step": 30600 |
| }, |
| { |
| "epoch": 9.77, |
| "grad_norm": 3.1821200277155754, |
| "learning_rate": 1e-05, |
| "loss": 0.0914, |
| "step": 30610 |
| }, |
| { |
| "epoch": 9.77, |
| "grad_norm": 6.791485212589531, |
| "learning_rate": 1e-05, |
| "loss": 0.0882, |
| "step": 30620 |
| }, |
| { |
| "epoch": 9.77, |
| "grad_norm": 5.304251696476169, |
| "learning_rate": 1e-05, |
| "loss": 0.0592, |
| "step": 30630 |
| }, |
| { |
| "epoch": 9.78, |
| "grad_norm": 5.048095744450733, |
| "learning_rate": 1e-05, |
| "loss": 0.0798, |
| "step": 30640 |
| }, |
| { |
| "epoch": 9.78, |
| "grad_norm": 3.4721491199851386, |
| "learning_rate": 1e-05, |
| "loss": 0.0493, |
| "step": 30650 |
| }, |
| { |
| "epoch": 9.78, |
| "grad_norm": 5.740638742030708, |
| "learning_rate": 1e-05, |
| "loss": 0.0733, |
| "step": 30660 |
| }, |
| { |
| "epoch": 9.79, |
| "grad_norm": 7.6805629289471735, |
| "learning_rate": 1e-05, |
| "loss": 0.0856, |
| "step": 30670 |
| }, |
| { |
| "epoch": 9.79, |
| "grad_norm": 5.014306961809476, |
| "learning_rate": 1e-05, |
| "loss": 0.0666, |
| "step": 30680 |
| }, |
| { |
| "epoch": 9.79, |
| "grad_norm": 4.750590237536855, |
| "learning_rate": 1e-05, |
| "loss": 0.0624, |
| "step": 30690 |
| }, |
| { |
| "epoch": 9.8, |
| "grad_norm": 3.997394070536901, |
| "learning_rate": 1e-05, |
| "loss": 0.0669, |
| "step": 30700 |
| }, |
| { |
| "epoch": 9.8, |
| "grad_norm": 5.52582779978935, |
| "learning_rate": 1e-05, |
| "loss": 0.0746, |
| "step": 30710 |
| }, |
| { |
| "epoch": 9.8, |
| "grad_norm": 4.515718980490992, |
| "learning_rate": 1e-05, |
| "loss": 0.0594, |
| "step": 30720 |
| }, |
| { |
| "epoch": 9.81, |
| "grad_norm": 3.7796042147319513, |
| "learning_rate": 1e-05, |
| "loss": 0.0562, |
| "step": 30730 |
| }, |
| { |
| "epoch": 9.81, |
| "grad_norm": 4.07920080734996, |
| "learning_rate": 1e-05, |
| "loss": 0.0836, |
| "step": 30740 |
| }, |
| { |
| "epoch": 9.81, |
| "grad_norm": 6.670093164436732, |
| "learning_rate": 1e-05, |
| "loss": 0.0653, |
| "step": 30750 |
| }, |
| { |
| "epoch": 9.81, |
| "grad_norm": 4.981019998598724, |
| "learning_rate": 1e-05, |
| "loss": 0.0987, |
| "step": 30760 |
| }, |
| { |
| "epoch": 9.82, |
| "grad_norm": 7.849385824292466, |
| "learning_rate": 1e-05, |
| "loss": 0.0613, |
| "step": 30770 |
| }, |
| { |
| "epoch": 9.82, |
| "grad_norm": 2.8811152561149918, |
| "learning_rate": 1e-05, |
| "loss": 0.0682, |
| "step": 30780 |
| }, |
| { |
| "epoch": 9.82, |
| "grad_norm": 4.579545877922619, |
| "learning_rate": 1e-05, |
| "loss": 0.0681, |
| "step": 30790 |
| }, |
| { |
| "epoch": 9.83, |
| "grad_norm": 3.465986921229618, |
| "learning_rate": 1e-05, |
| "loss": 0.0601, |
| "step": 30800 |
| }, |
| { |
| "epoch": 9.83, |
| "grad_norm": 4.027008427480628, |
| "learning_rate": 1e-05, |
| "loss": 0.0454, |
| "step": 30810 |
| }, |
| { |
| "epoch": 9.83, |
| "grad_norm": 14.224516988106199, |
| "learning_rate": 1e-05, |
| "loss": 0.0876, |
| "step": 30820 |
| }, |
| { |
| "epoch": 9.84, |
| "grad_norm": 2.822692880991985, |
| "learning_rate": 1e-05, |
| "loss": 0.0622, |
| "step": 30830 |
| }, |
| { |
| "epoch": 9.84, |
| "grad_norm": 3.835570525661628, |
| "learning_rate": 1e-05, |
| "loss": 0.1056, |
| "step": 30840 |
| }, |
| { |
| "epoch": 9.84, |
| "grad_norm": 4.606370056326912, |
| "learning_rate": 1e-05, |
| "loss": 0.0796, |
| "step": 30850 |
| }, |
| { |
| "epoch": 9.85, |
| "grad_norm": 4.110384856905129, |
| "learning_rate": 1e-05, |
| "loss": 0.0726, |
| "step": 30860 |
| }, |
| { |
| "epoch": 9.85, |
| "grad_norm": 4.448705574964993, |
| "learning_rate": 1e-05, |
| "loss": 0.0566, |
| "step": 30870 |
| }, |
| { |
| "epoch": 9.85, |
| "grad_norm": 3.713667401525456, |
| "learning_rate": 1e-05, |
| "loss": 0.0906, |
| "step": 30880 |
| }, |
| { |
| "epoch": 9.86, |
| "grad_norm": 4.3230686881057165, |
| "learning_rate": 1e-05, |
| "loss": 0.064, |
| "step": 30890 |
| }, |
| { |
| "epoch": 9.86, |
| "grad_norm": 4.142137474353707, |
| "learning_rate": 1e-05, |
| "loss": 0.0589, |
| "step": 30900 |
| }, |
| { |
| "epoch": 9.86, |
| "grad_norm": 6.858437580538484, |
| "learning_rate": 1e-05, |
| "loss": 0.0639, |
| "step": 30910 |
| }, |
| { |
| "epoch": 9.87, |
| "grad_norm": 7.547713776094693, |
| "learning_rate": 1e-05, |
| "loss": 0.0875, |
| "step": 30920 |
| }, |
| { |
| "epoch": 9.87, |
| "grad_norm": 3.6660528247147, |
| "learning_rate": 1e-05, |
| "loss": 0.0655, |
| "step": 30930 |
| }, |
| { |
| "epoch": 9.87, |
| "grad_norm": 5.3176842021737105, |
| "learning_rate": 1e-05, |
| "loss": 0.0769, |
| "step": 30940 |
| }, |
| { |
| "epoch": 9.88, |
| "grad_norm": 5.891930716216652, |
| "learning_rate": 1e-05, |
| "loss": 0.0739, |
| "step": 30950 |
| }, |
| { |
| "epoch": 9.88, |
| "grad_norm": 5.820771597350766, |
| "learning_rate": 1e-05, |
| "loss": 0.0637, |
| "step": 30960 |
| }, |
| { |
| "epoch": 9.88, |
| "grad_norm": 4.612291350982465, |
| "learning_rate": 1e-05, |
| "loss": 0.0739, |
| "step": 30970 |
| }, |
| { |
| "epoch": 9.89, |
| "grad_norm": 5.275478301578427, |
| "learning_rate": 1e-05, |
| "loss": 0.0812, |
| "step": 30980 |
| }, |
| { |
| "epoch": 9.89, |
| "grad_norm": 4.166337928519378, |
| "learning_rate": 1e-05, |
| "loss": 0.0729, |
| "step": 30990 |
| }, |
| { |
| "epoch": 9.89, |
| "grad_norm": 5.792111313398325, |
| "learning_rate": 1e-05, |
| "loss": 0.0639, |
| "step": 31000 |
| }, |
| { |
| "epoch": 9.89, |
| "eval_loss": 1.193359375, |
| "eval_runtime": 54.7781, |
| "eval_samples_per_second": 46.259, |
| "eval_steps_per_second": 2.903, |
| "step": 31000 |
| }, |
| { |
| "epoch": 9.89, |
| "grad_norm": 6.64303552297236, |
| "learning_rate": 1e-05, |
| "loss": 0.079, |
| "step": 31010 |
| }, |
| { |
| "epoch": 9.9, |
| "grad_norm": 5.670213860678456, |
| "learning_rate": 1e-05, |
| "loss": 0.0963, |
| "step": 31020 |
| }, |
| { |
| "epoch": 9.9, |
| "grad_norm": 6.4348319737595, |
| "learning_rate": 1e-05, |
| "loss": 0.0494, |
| "step": 31030 |
| }, |
| { |
| "epoch": 9.9, |
| "grad_norm": 6.370019500558849, |
| "learning_rate": 1e-05, |
| "loss": 0.0668, |
| "step": 31040 |
| }, |
| { |
| "epoch": 9.91, |
| "grad_norm": 7.620370929330536, |
| "learning_rate": 1e-05, |
| "loss": 0.0755, |
| "step": 31050 |
| }, |
| { |
| "epoch": 9.91, |
| "grad_norm": 6.345090827742712, |
| "learning_rate": 1e-05, |
| "loss": 0.0721, |
| "step": 31060 |
| }, |
| { |
| "epoch": 9.91, |
| "grad_norm": 6.096195610512417, |
| "learning_rate": 1e-05, |
| "loss": 0.084, |
| "step": 31070 |
| }, |
| { |
| "epoch": 9.92, |
| "grad_norm": 3.85450381617197, |
| "learning_rate": 1e-05, |
| "loss": 0.0785, |
| "step": 31080 |
| }, |
| { |
| "epoch": 9.92, |
| "grad_norm": 7.480631933598696, |
| "learning_rate": 1e-05, |
| "loss": 0.0674, |
| "step": 31090 |
| }, |
| { |
| "epoch": 9.92, |
| "grad_norm": 3.22093541294873, |
| "learning_rate": 1e-05, |
| "loss": 0.068, |
| "step": 31100 |
| }, |
| { |
| "epoch": 9.93, |
| "grad_norm": 2.94032184525962, |
| "learning_rate": 1e-05, |
| "loss": 0.0667, |
| "step": 31110 |
| }, |
| { |
| "epoch": 9.93, |
| "grad_norm": 5.705423471980753, |
| "learning_rate": 1e-05, |
| "loss": 0.0622, |
| "step": 31120 |
| }, |
| { |
| "epoch": 9.93, |
| "grad_norm": 4.311531068615841, |
| "learning_rate": 1e-05, |
| "loss": 0.0724, |
| "step": 31130 |
| }, |
| { |
| "epoch": 9.94, |
| "grad_norm": 6.239663999754633, |
| "learning_rate": 1e-05, |
| "loss": 0.0807, |
| "step": 31140 |
| }, |
| { |
| "epoch": 9.94, |
| "grad_norm": 4.235887545885939, |
| "learning_rate": 1e-05, |
| "loss": 0.0674, |
| "step": 31150 |
| }, |
| { |
| "epoch": 9.94, |
| "grad_norm": 3.0696324961628587, |
| "learning_rate": 1e-05, |
| "loss": 0.0653, |
| "step": 31160 |
| }, |
| { |
| "epoch": 9.95, |
| "grad_norm": 2.705111264463072, |
| "learning_rate": 1e-05, |
| "loss": 0.0626, |
| "step": 31170 |
| }, |
| { |
| "epoch": 9.95, |
| "grad_norm": 4.069364646039541, |
| "learning_rate": 1e-05, |
| "loss": 0.0641, |
| "step": 31180 |
| }, |
| { |
| "epoch": 9.95, |
| "grad_norm": 3.679832633815107, |
| "learning_rate": 1e-05, |
| "loss": 0.0593, |
| "step": 31190 |
| }, |
| { |
| "epoch": 9.96, |
| "grad_norm": 5.393208025846824, |
| "learning_rate": 1e-05, |
| "loss": 0.0792, |
| "step": 31200 |
| }, |
| { |
| "epoch": 9.96, |
| "grad_norm": 5.198409592897835, |
| "learning_rate": 1e-05, |
| "loss": 0.0566, |
| "step": 31210 |
| }, |
| { |
| "epoch": 9.96, |
| "grad_norm": 3.796063622957473, |
| "learning_rate": 1e-05, |
| "loss": 0.0587, |
| "step": 31220 |
| }, |
| { |
| "epoch": 9.96, |
| "grad_norm": 5.390923444806887, |
| "learning_rate": 1e-05, |
| "loss": 0.0622, |
| "step": 31230 |
| }, |
| { |
| "epoch": 9.97, |
| "grad_norm": 5.123158007575118, |
| "learning_rate": 1e-05, |
| "loss": 0.0775, |
| "step": 31240 |
| }, |
| { |
| "epoch": 9.97, |
| "grad_norm": 3.406997738581322, |
| "learning_rate": 1e-05, |
| "loss": 0.0666, |
| "step": 31250 |
| }, |
| { |
| "epoch": 9.97, |
| "grad_norm": 7.463542830863555, |
| "learning_rate": 1e-05, |
| "loss": 0.0752, |
| "step": 31260 |
| }, |
| { |
| "epoch": 9.98, |
| "grad_norm": 6.584443236197947, |
| "learning_rate": 1e-05, |
| "loss": 0.0678, |
| "step": 31270 |
| }, |
| { |
| "epoch": 9.98, |
| "grad_norm": 5.569079847232272, |
| "learning_rate": 1e-05, |
| "loss": 0.0641, |
| "step": 31280 |
| }, |
| { |
| "epoch": 9.98, |
| "grad_norm": 4.795734878886076, |
| "learning_rate": 1e-05, |
| "loss": 0.0941, |
| "step": 31290 |
| }, |
| { |
| "epoch": 9.99, |
| "grad_norm": 2.59574289628697, |
| "learning_rate": 1e-05, |
| "loss": 0.0709, |
| "step": 31300 |
| }, |
| { |
| "epoch": 9.99, |
| "grad_norm": 4.518738198165994, |
| "learning_rate": 1e-05, |
| "loss": 0.0661, |
| "step": 31310 |
| }, |
| { |
| "epoch": 9.99, |
| "grad_norm": 4.606763403760556, |
| "learning_rate": 1e-05, |
| "loss": 0.0719, |
| "step": 31320 |
| }, |
| { |
| "epoch": 10.0, |
| "grad_norm": 6.309546355612505, |
| "learning_rate": 1e-05, |
| "loss": 0.0647, |
| "step": 31330 |
| }, |
| { |
| "epoch": 10.0, |
| "grad_norm": 7.70128827898332, |
| "learning_rate": 1e-05, |
| "loss": 0.0758, |
| "step": 31340 |
| }, |
| { |
| "epoch": 10.0, |
| "step": 31340, |
| "total_flos": 913810558271488.0, |
| "train_loss": 0.20468773061900514, |
| "train_runtime": 109923.8226, |
| "train_samples_per_second": 4.561, |
| "train_steps_per_second": 0.285 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 31340, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 10, |
| "save_steps": 500, |
| "total_flos": 913810558271488.0, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
