| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.3001968503937008, | |
| "eval_steps": 153, | |
| "global_step": 305, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.000984251968503937, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.0, | |
| "loss": 10.4072, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.001968503937007874, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.0, | |
| "loss": 11.0865, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.002952755905511811, | |
| "grad_norm": 60.69786071777344, | |
| "learning_rate": 9.940357852882705e-10, | |
| "loss": 9.5114, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.003937007874015748, | |
| "grad_norm": 59.147647857666016, | |
| "learning_rate": 1.988071570576541e-09, | |
| "loss": 9.9584, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.004921259842519685, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.988071570576541e-09, | |
| "loss": 10.068, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.005905511811023622, | |
| "grad_norm": 65.82320404052734, | |
| "learning_rate": 2.9821073558648116e-09, | |
| "loss": 11.0224, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.006889763779527559, | |
| "grad_norm": 59.096107482910156, | |
| "learning_rate": 3.976143141153082e-09, | |
| "loss": 9.7703, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.007874015748031496, | |
| "grad_norm": 61.43330764770508, | |
| "learning_rate": 4.970178926441353e-09, | |
| "loss": 10.5005, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.008858267716535433, | |
| "grad_norm": 61.25212860107422, | |
| "learning_rate": 5.964214711729623e-09, | |
| "loss": 10.1987, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.00984251968503937, | |
| "grad_norm": 61.6600341796875, | |
| "learning_rate": 6.9582504970178946e-09, | |
| "loss": 10.0277, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.010826771653543307, | |
| "grad_norm": 61.71075439453125, | |
| "learning_rate": 7.952286282306164e-09, | |
| "loss": 10.6965, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.011811023622047244, | |
| "grad_norm": 59.34035873413086, | |
| "learning_rate": 8.946322067594435e-09, | |
| "loss": 10.0609, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.012795275590551181, | |
| "grad_norm": 64.14221954345703, | |
| "learning_rate": 9.940357852882705e-09, | |
| "loss": 11.6214, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.013779527559055118, | |
| "grad_norm": 57.682823181152344, | |
| "learning_rate": 1.0934393638170978e-08, | |
| "loss": 9.4053, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.014763779527559055, | |
| "grad_norm": 62.256858825683594, | |
| "learning_rate": 1.1928429423459246e-08, | |
| "loss": 10.4014, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.015748031496062992, | |
| "grad_norm": 64.56417083740234, | |
| "learning_rate": 1.2922465208747517e-08, | |
| "loss": 10.4119, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.01673228346456693, | |
| "grad_norm": 56.612098693847656, | |
| "learning_rate": 1.3916500994035789e-08, | |
| "loss": 9.4658, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.017716535433070866, | |
| "grad_norm": 60.02314758300781, | |
| "learning_rate": 1.4910536779324056e-08, | |
| "loss": 9.2169, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.018700787401574805, | |
| "grad_norm": 66.38670349121094, | |
| "learning_rate": 1.590457256461233e-08, | |
| "loss": 11.2337, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.01968503937007874, | |
| "grad_norm": 64.9875259399414, | |
| "learning_rate": 1.68986083499006e-08, | |
| "loss": 11.0572, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02066929133858268, | |
| "grad_norm": 59.36448287963867, | |
| "learning_rate": 1.789264413518887e-08, | |
| "loss": 11.0452, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.021653543307086614, | |
| "grad_norm": 59.18537902832031, | |
| "learning_rate": 1.888667992047714e-08, | |
| "loss": 10.31, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.022637795275590553, | |
| "grad_norm": 54.053524017333984, | |
| "learning_rate": 1.988071570576541e-08, | |
| "loss": 9.1395, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.023622047244094488, | |
| "grad_norm": 49.979618072509766, | |
| "learning_rate": 2.087475149105368e-08, | |
| "loss": 8.4201, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.024606299212598427, | |
| "grad_norm": 51.654518127441406, | |
| "learning_rate": 2.1868787276341955e-08, | |
| "loss": 8.6036, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.025590551181102362, | |
| "grad_norm": 63.55083465576172, | |
| "learning_rate": 2.2862823061630224e-08, | |
| "loss": 11.7579, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0265748031496063, | |
| "grad_norm": 59.30263137817383, | |
| "learning_rate": 2.3856858846918493e-08, | |
| "loss": 10.1307, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.027559055118110236, | |
| "grad_norm": 50.75270462036133, | |
| "learning_rate": 2.4850894632206765e-08, | |
| "loss": 9.2915, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.028543307086614175, | |
| "grad_norm": 51.7747917175293, | |
| "learning_rate": 2.5844930417495034e-08, | |
| "loss": 9.0208, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.02952755905511811, | |
| "grad_norm": 46.81666564941406, | |
| "learning_rate": 2.6838966202783303e-08, | |
| "loss": 8.6867, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03051181102362205, | |
| "grad_norm": 44.82905578613281, | |
| "learning_rate": 2.7833001988071578e-08, | |
| "loss": 8.0925, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.031496062992125984, | |
| "grad_norm": 47.148406982421875, | |
| "learning_rate": 2.8827037773359847e-08, | |
| "loss": 8.6617, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.03248031496062992, | |
| "grad_norm": 47.153053283691406, | |
| "learning_rate": 2.982107355864811e-08, | |
| "loss": 8.3374, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.03346456692913386, | |
| "grad_norm": 45.4912223815918, | |
| "learning_rate": 3.081510934393639e-08, | |
| "loss": 7.8566, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0344488188976378, | |
| "grad_norm": 51.92588806152344, | |
| "learning_rate": 3.180914512922466e-08, | |
| "loss": 9.0698, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.03543307086614173, | |
| "grad_norm": 43.18679428100586, | |
| "learning_rate": 3.280318091451293e-08, | |
| "loss": 7.7727, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.03641732283464567, | |
| "grad_norm": 43.12812805175781, | |
| "learning_rate": 3.37972166998012e-08, | |
| "loss": 7.6128, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.03740157480314961, | |
| "grad_norm": 44.7684211730957, | |
| "learning_rate": 3.479125248508947e-08, | |
| "loss": 7.8762, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.038385826771653545, | |
| "grad_norm": 36.34556198120117, | |
| "learning_rate": 3.578528827037774e-08, | |
| "loss": 7.5191, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.03937007874015748, | |
| "grad_norm": 39.9330940246582, | |
| "learning_rate": 3.6779324055666005e-08, | |
| "loss": 7.5638, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.040354330708661415, | |
| "grad_norm": 39.49993133544922, | |
| "learning_rate": 3.777335984095428e-08, | |
| "loss": 7.1878, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.04133858267716536, | |
| "grad_norm": 34.661251068115234, | |
| "learning_rate": 3.8767395626242556e-08, | |
| "loss": 6.8878, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.04232283464566929, | |
| "grad_norm": 40.655540466308594, | |
| "learning_rate": 3.976143141153082e-08, | |
| "loss": 7.5775, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.04330708661417323, | |
| "grad_norm": 36.12670135498047, | |
| "learning_rate": 4.0755467196819094e-08, | |
| "loss": 7.1076, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.04429133858267716, | |
| "grad_norm": 34.178226470947266, | |
| "learning_rate": 4.174950298210736e-08, | |
| "loss": 6.5589, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.045275590551181105, | |
| "grad_norm": 36.6577033996582, | |
| "learning_rate": 4.274353876739563e-08, | |
| "loss": 7.4456, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.04625984251968504, | |
| "grad_norm": 32.548709869384766, | |
| "learning_rate": 4.373757455268391e-08, | |
| "loss": 6.8233, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.047244094488188976, | |
| "grad_norm": 34.000553131103516, | |
| "learning_rate": 4.4731610337972176e-08, | |
| "loss": 6.7633, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.04822834645669291, | |
| "grad_norm": 32.247859954833984, | |
| "learning_rate": 4.572564612326045e-08, | |
| "loss": 6.6024, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.04921259842519685, | |
| "grad_norm": 28.947786331176758, | |
| "learning_rate": 4.6719681908548713e-08, | |
| "loss": 6.2778, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05019685039370079, | |
| "grad_norm": 30.279062271118164, | |
| "learning_rate": 4.7713717693836986e-08, | |
| "loss": 6.1026, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.051181102362204724, | |
| "grad_norm": 31.13785171508789, | |
| "learning_rate": 4.870775347912525e-08, | |
| "loss": 6.632, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.05216535433070866, | |
| "grad_norm": 29.648237228393555, | |
| "learning_rate": 4.970178926441353e-08, | |
| "loss": 6.6962, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.0531496062992126, | |
| "grad_norm": 28.224645614624023, | |
| "learning_rate": 5.06958250497018e-08, | |
| "loss": 5.8514, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.054133858267716536, | |
| "grad_norm": 28.693328857421875, | |
| "learning_rate": 5.168986083499007e-08, | |
| "loss": 5.9951, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.05511811023622047, | |
| "grad_norm": 24.777812957763672, | |
| "learning_rate": 5.268389662027834e-08, | |
| "loss": 5.4554, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.05610236220472441, | |
| "grad_norm": 26.11226463317871, | |
| "learning_rate": 5.3677932405566605e-08, | |
| "loss": 6.0147, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.05708661417322835, | |
| "grad_norm": 24.698375701904297, | |
| "learning_rate": 5.467196819085488e-08, | |
| "loss": 5.215, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.058070866141732284, | |
| "grad_norm": 26.616317749023438, | |
| "learning_rate": 5.5666003976143156e-08, | |
| "loss": 6.4525, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.05905511811023622, | |
| "grad_norm": 26.09321403503418, | |
| "learning_rate": 5.666003976143142e-08, | |
| "loss": 5.4048, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.060039370078740155, | |
| "grad_norm": 19.83713150024414, | |
| "learning_rate": 5.7654075546719694e-08, | |
| "loss": 5.0424, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.0610236220472441, | |
| "grad_norm": 26.56923484802246, | |
| "learning_rate": 5.864811133200796e-08, | |
| "loss": 6.2646, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.06200787401574803, | |
| "grad_norm": 23.580089569091797, | |
| "learning_rate": 5.964214711729623e-08, | |
| "loss": 5.0847, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.06299212598425197, | |
| "grad_norm": 23.453126907348633, | |
| "learning_rate": 6.06361829025845e-08, | |
| "loss": 5.4415, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.0639763779527559, | |
| "grad_norm": 21.229190826416016, | |
| "learning_rate": 6.163021868787278e-08, | |
| "loss": 5.2469, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.06496062992125984, | |
| "grad_norm": 19.477190017700195, | |
| "learning_rate": 6.262425447316104e-08, | |
| "loss": 5.1378, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.06594488188976377, | |
| "grad_norm": 19.40647315979004, | |
| "learning_rate": 6.361829025844931e-08, | |
| "loss": 5.1636, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.06692913385826772, | |
| "grad_norm": 22.20977210998535, | |
| "learning_rate": 6.461232604373759e-08, | |
| "loss": 5.5596, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.06791338582677166, | |
| "grad_norm": 19.186826705932617, | |
| "learning_rate": 6.560636182902586e-08, | |
| "loss": 4.9508, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.0688976377952756, | |
| "grad_norm": 20.190908432006836, | |
| "learning_rate": 6.660039761431412e-08, | |
| "loss": 5.2355, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06988188976377953, | |
| "grad_norm": 18.122196197509766, | |
| "learning_rate": 6.75944333996024e-08, | |
| "loss": 4.7359, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.07086614173228346, | |
| "grad_norm": 17.524765014648438, | |
| "learning_rate": 6.858846918489067e-08, | |
| "loss": 4.8947, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.0718503937007874, | |
| "grad_norm": 18.821767807006836, | |
| "learning_rate": 6.958250497017893e-08, | |
| "loss": 4.6269, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.07283464566929133, | |
| "grad_norm": 18.19922637939453, | |
| "learning_rate": 7.057654075546721e-08, | |
| "loss": 4.6072, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.07381889763779527, | |
| "grad_norm": 16.908899307250977, | |
| "learning_rate": 7.157057654075548e-08, | |
| "loss": 4.9125, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.07480314960629922, | |
| "grad_norm": 19.90263557434082, | |
| "learning_rate": 7.256461232604374e-08, | |
| "loss": 4.5856, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.07578740157480315, | |
| "grad_norm": 17.92584800720215, | |
| "learning_rate": 7.355864811133201e-08, | |
| "loss": 4.7879, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.07677165354330709, | |
| "grad_norm": 16.29261589050293, | |
| "learning_rate": 7.455268389662029e-08, | |
| "loss": 4.5348, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.07775590551181102, | |
| "grad_norm": 16.3350887298584, | |
| "learning_rate": 7.554671968190855e-08, | |
| "loss": 4.3554, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.07874015748031496, | |
| "grad_norm": 14.408184051513672, | |
| "learning_rate": 7.654075546719683e-08, | |
| "loss": 4.2984, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0797244094488189, | |
| "grad_norm": 16.71326446533203, | |
| "learning_rate": 7.753479125248511e-08, | |
| "loss": 4.5505, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.08070866141732283, | |
| "grad_norm": 14.62590217590332, | |
| "learning_rate": 7.852882703777338e-08, | |
| "loss": 4.5325, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.08169291338582677, | |
| "grad_norm": 17.189268112182617, | |
| "learning_rate": 7.952286282306164e-08, | |
| "loss": 4.2725, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.08267716535433071, | |
| "grad_norm": 16.960248947143555, | |
| "learning_rate": 8.051689860834992e-08, | |
| "loss": 4.3054, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.08366141732283465, | |
| "grad_norm": 15.114398956298828, | |
| "learning_rate": 8.151093439363819e-08, | |
| "loss": 4.5536, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.08464566929133858, | |
| "grad_norm": 16.153371810913086, | |
| "learning_rate": 8.250497017892645e-08, | |
| "loss": 4.0265, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.08562992125984252, | |
| "grad_norm": 15.731820106506348, | |
| "learning_rate": 8.349900596421472e-08, | |
| "loss": 4.7453, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.08661417322834646, | |
| "grad_norm": 14.69382381439209, | |
| "learning_rate": 8.4493041749503e-08, | |
| "loss": 4.071, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.08759842519685039, | |
| "grad_norm": 13.735575675964355, | |
| "learning_rate": 8.548707753479126e-08, | |
| "loss": 4.1582, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.08858267716535433, | |
| "grad_norm": 16.017065048217773, | |
| "learning_rate": 8.648111332007953e-08, | |
| "loss": 4.1131, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.08956692913385826, | |
| "grad_norm": 17.237276077270508, | |
| "learning_rate": 8.747514910536782e-08, | |
| "loss": 3.6582, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.09055118110236221, | |
| "grad_norm": 15.59334945678711, | |
| "learning_rate": 8.846918489065609e-08, | |
| "loss": 4.143, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.09153543307086615, | |
| "grad_norm": 14.918270111083984, | |
| "learning_rate": 8.946322067594435e-08, | |
| "loss": 4.2273, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.09251968503937008, | |
| "grad_norm": 14.899909019470215, | |
| "learning_rate": 9.045725646123262e-08, | |
| "loss": 3.9321, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.09350393700787402, | |
| "grad_norm": 18.112892150878906, | |
| "learning_rate": 9.14512922465209e-08, | |
| "loss": 4.2061, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.09448818897637795, | |
| "grad_norm": 15.854629516601562, | |
| "learning_rate": 9.244532803180916e-08, | |
| "loss": 4.1042, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.09547244094488189, | |
| "grad_norm": 16.44801139831543, | |
| "learning_rate": 9.343936381709743e-08, | |
| "loss": 3.9513, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.09645669291338582, | |
| "grad_norm": 14.854127883911133, | |
| "learning_rate": 9.44333996023857e-08, | |
| "loss": 3.8627, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.09744094488188976, | |
| "grad_norm": 17.02035903930664, | |
| "learning_rate": 9.542743538767397e-08, | |
| "loss": 4.3613, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.0984251968503937, | |
| "grad_norm": 15.354338645935059, | |
| "learning_rate": 9.642147117296224e-08, | |
| "loss": 3.8513, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09940944881889764, | |
| "grad_norm": 16.35565757751465, | |
| "learning_rate": 9.74155069582505e-08, | |
| "loss": 3.5866, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.10039370078740158, | |
| "grad_norm": 16.66194725036621, | |
| "learning_rate": 9.840954274353878e-08, | |
| "loss": 3.5239, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.10137795275590551, | |
| "grad_norm": 15.875446319580078, | |
| "learning_rate": 9.940357852882706e-08, | |
| "loss": 3.5921, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.10236220472440945, | |
| "grad_norm": 14.344114303588867, | |
| "learning_rate": 1.0039761431411533e-07, | |
| "loss": 3.5962, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.10334645669291338, | |
| "grad_norm": 18.503963470458984, | |
| "learning_rate": 1.013916500994036e-07, | |
| "loss": 4.0001, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.10433070866141732, | |
| "grad_norm": 16.944435119628906, | |
| "learning_rate": 1.0238568588469187e-07, | |
| "loss": 4.1374, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.10531496062992125, | |
| "grad_norm": 16.46833038330078, | |
| "learning_rate": 1.0337972166998014e-07, | |
| "loss": 3.9049, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.1062992125984252, | |
| "grad_norm": 14.921700477600098, | |
| "learning_rate": 1.043737574552684e-07, | |
| "loss": 3.2511, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.10728346456692914, | |
| "grad_norm": 15.574972152709961, | |
| "learning_rate": 1.0536779324055668e-07, | |
| "loss": 3.2479, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.10826771653543307, | |
| "grad_norm": 16.810884475708008, | |
| "learning_rate": 1.0636182902584495e-07, | |
| "loss": 3.6414, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.10925196850393701, | |
| "grad_norm": 17.074661254882812, | |
| "learning_rate": 1.0735586481113321e-07, | |
| "loss": 3.6429, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.11023622047244094, | |
| "grad_norm": 18.52947235107422, | |
| "learning_rate": 1.0834990059642149e-07, | |
| "loss": 3.423, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.11122047244094488, | |
| "grad_norm": 18.681869506835938, | |
| "learning_rate": 1.0934393638170976e-07, | |
| "loss": 3.4967, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.11220472440944881, | |
| "grad_norm": 20.385292053222656, | |
| "learning_rate": 1.1033797216699802e-07, | |
| "loss": 3.7649, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.11318897637795275, | |
| "grad_norm": 18.912538528442383, | |
| "learning_rate": 1.1133200795228631e-07, | |
| "loss": 3.2845, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.1141732283464567, | |
| "grad_norm": 17.856229782104492, | |
| "learning_rate": 1.1232604373757458e-07, | |
| "loss": 3.356, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.11515748031496063, | |
| "grad_norm": 17.08562469482422, | |
| "learning_rate": 1.1332007952286284e-07, | |
| "loss": 3.2086, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.11614173228346457, | |
| "grad_norm": 17.54237937927246, | |
| "learning_rate": 1.1431411530815111e-07, | |
| "loss": 3.5561, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.1171259842519685, | |
| "grad_norm": 19.936498641967773, | |
| "learning_rate": 1.1530815109343939e-07, | |
| "loss": 3.7353, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.11811023622047244, | |
| "grad_norm": 17.135496139526367, | |
| "learning_rate": 1.1630218687872765e-07, | |
| "loss": 3.403, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.11909448818897637, | |
| "grad_norm": 17.260093688964844, | |
| "learning_rate": 1.1729622266401592e-07, | |
| "loss": 3.1009, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.12007874015748031, | |
| "grad_norm": 17.075611114501953, | |
| "learning_rate": 1.182902584493042e-07, | |
| "loss": 3.2139, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.12106299212598425, | |
| "grad_norm": 23.433874130249023, | |
| "learning_rate": 1.1928429423459245e-07, | |
| "loss": 3.3339, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.1220472440944882, | |
| "grad_norm": 18.25501251220703, | |
| "learning_rate": 1.2027833001988073e-07, | |
| "loss": 2.9464, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.12303149606299213, | |
| "grad_norm": 18.079578399658203, | |
| "learning_rate": 1.21272365805169e-07, | |
| "loss": 3.3366, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.12401574803149606, | |
| "grad_norm": 16.392736434936523, | |
| "learning_rate": 1.222664015904573e-07, | |
| "loss": 3.0618, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 15.782499313354492, | |
| "learning_rate": 1.2326043737574557e-07, | |
| "loss": 3.0092, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.12598425196850394, | |
| "grad_norm": 14.74819278717041, | |
| "learning_rate": 1.2425447316103382e-07, | |
| "loss": 2.7152, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.12696850393700787, | |
| "grad_norm": 17.743946075439453, | |
| "learning_rate": 1.2524850894632207e-07, | |
| "loss": 2.9423, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.1279527559055118, | |
| "grad_norm": 15.759135246276855, | |
| "learning_rate": 1.2624254473161035e-07, | |
| "loss": 2.6569, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.12893700787401574, | |
| "grad_norm": 18.54253387451172, | |
| "learning_rate": 1.2723658051689863e-07, | |
| "loss": 2.8469, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.12992125984251968, | |
| "grad_norm": 18.318775177001953, | |
| "learning_rate": 1.282306163021869e-07, | |
| "loss": 2.9089, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.1309055118110236, | |
| "grad_norm": 17.751266479492188, | |
| "learning_rate": 1.2922465208747519e-07, | |
| "loss": 2.5809, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.13188976377952755, | |
| "grad_norm": 21.29975128173828, | |
| "learning_rate": 1.3021868787276344e-07, | |
| "loss": 2.6987, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.1328740157480315, | |
| "grad_norm": 19.9519100189209, | |
| "learning_rate": 1.3121272365805172e-07, | |
| "loss": 3.2518, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.13385826771653545, | |
| "grad_norm": 19.99887466430664, | |
| "learning_rate": 1.3220675944333997e-07, | |
| "loss": 2.9145, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.13484251968503938, | |
| "grad_norm": 17.0355167388916, | |
| "learning_rate": 1.3320079522862825e-07, | |
| "loss": 2.4809, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.13582677165354332, | |
| "grad_norm": 18.628467559814453, | |
| "learning_rate": 1.3419483101391653e-07, | |
| "loss": 2.8264, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.13681102362204725, | |
| "grad_norm": 19.511394500732422, | |
| "learning_rate": 1.351888667992048e-07, | |
| "loss": 2.5724, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.1377952755905512, | |
| "grad_norm": 21.264480590820312, | |
| "learning_rate": 1.3618290258449306e-07, | |
| "loss": 2.6949, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.13877952755905512, | |
| "grad_norm": 19.76134490966797, | |
| "learning_rate": 1.3717693836978134e-07, | |
| "loss": 2.6925, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.13976377952755906, | |
| "grad_norm": 20.930673599243164, | |
| "learning_rate": 1.381709741550696e-07, | |
| "loss": 2.9311, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.140748031496063, | |
| "grad_norm": 21.966018676757812, | |
| "learning_rate": 1.3916500994035787e-07, | |
| "loss": 2.5667, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.14173228346456693, | |
| "grad_norm": 21.916505813598633, | |
| "learning_rate": 1.4015904572564615e-07, | |
| "loss": 3.2471, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.14271653543307086, | |
| "grad_norm": 20.081771850585938, | |
| "learning_rate": 1.4115308151093443e-07, | |
| "loss": 2.2441, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.1437007874015748, | |
| "grad_norm": 22.893489837646484, | |
| "learning_rate": 1.421471172962227e-07, | |
| "loss": 2.75, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.14468503937007873, | |
| "grad_norm": 23.95358657836914, | |
| "learning_rate": 1.4314115308151096e-07, | |
| "loss": 2.9669, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.14566929133858267, | |
| "grad_norm": 21.101062774658203, | |
| "learning_rate": 1.4413518886679924e-07, | |
| "loss": 2.736, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.1466535433070866, | |
| "grad_norm": 25.240341186523438, | |
| "learning_rate": 1.451292246520875e-07, | |
| "loss": 3.104, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.14763779527559054, | |
| "grad_norm": 18.358688354492188, | |
| "learning_rate": 1.4612326043737577e-07, | |
| "loss": 2.2175, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1486220472440945, | |
| "grad_norm": 21.986661911010742, | |
| "learning_rate": 1.4711729622266402e-07, | |
| "loss": 2.7415, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.14960629921259844, | |
| "grad_norm": 20.64093017578125, | |
| "learning_rate": 1.4811133200795232e-07, | |
| "loss": 1.8707, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.15059055118110237, | |
| "grad_norm": 20.602142333984375, | |
| "learning_rate": 1.4910536779324058e-07, | |
| "loss": 2.5961, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.15059055118110237, | |
| "eval_Qnli-dev_cosine_accuracy": 0.634765625, | |
| "eval_Qnli-dev_cosine_accuracy_threshold": 0.8508153557777405, | |
| "eval_Qnli-dev_cosine_ap": 0.6461335447626624, | |
| "eval_Qnli-dev_cosine_f1": 0.6505636070853462, | |
| "eval_Qnli-dev_cosine_f1_threshold": 0.7770615816116333, | |
| "eval_Qnli-dev_cosine_precision": 0.5246753246753246, | |
| "eval_Qnli-dev_cosine_recall": 0.8559322033898306, | |
| "eval_Qnli-dev_dot_accuracy": 0.634765625, | |
| "eval_Qnli-dev_dot_accuracy_threshold": 653.7443237304688, | |
| "eval_Qnli-dev_dot_ap": 0.6461682282377894, | |
| "eval_Qnli-dev_dot_f1": 0.6505636070853462, | |
| "eval_Qnli-dev_dot_f1_threshold": 597.0731811523438, | |
| "eval_Qnli-dev_dot_precision": 0.5246753246753246, | |
| "eval_Qnli-dev_dot_recall": 0.8559322033898306, | |
| "eval_Qnli-dev_euclidean_accuracy": 0.634765625, | |
| "eval_Qnli-dev_euclidean_accuracy_threshold": 15.141305923461914, | |
| "eval_Qnli-dev_euclidean_ap": 0.6461382925406688, | |
| "eval_Qnli-dev_euclidean_f1": 0.6505636070853462, | |
| "eval_Qnli-dev_euclidean_f1_threshold": 18.50943946838379, | |
| "eval_Qnli-dev_euclidean_precision": 0.5246753246753246, | |
| "eval_Qnli-dev_euclidean_recall": 0.8559322033898306, | |
| "eval_Qnli-dev_manhattan_accuracy": 0.6328125, | |
| "eval_Qnli-dev_manhattan_accuracy_threshold": 331.46282958984375, | |
| "eval_Qnli-dev_manhattan_ap": 0.6431949026371255, | |
| "eval_Qnli-dev_manhattan_f1": 0.6501650165016502, | |
| "eval_Qnli-dev_manhattan_f1_threshold": 404.6050109863281, | |
| "eval_Qnli-dev_manhattan_precision": 0.5324324324324324, | |
| "eval_Qnli-dev_manhattan_recall": 0.8347457627118644, | |
| "eval_Qnli-dev_max_accuracy": 0.634765625, | |
| "eval_Qnli-dev_max_accuracy_threshold": 653.7443237304688, | |
| "eval_Qnli-dev_max_ap": 0.6461682282377894, | |
| "eval_Qnli-dev_max_f1": 0.6505636070853462, | |
| "eval_Qnli-dev_max_f1_threshold": 597.0731811523438, | |
| "eval_Qnli-dev_max_precision": 0.5324324324324324, | |
| "eval_Qnli-dev_max_recall": 0.8559322033898306, | |
| "eval_allNLI-dev_cosine_accuracy": 0.67578125, | |
| "eval_allNLI-dev_cosine_accuracy_threshold": 0.9452645182609558, | |
| "eval_allNLI-dev_cosine_ap": 0.4264736612515921, | |
| "eval_allNLI-dev_cosine_f1": 0.512, | |
| "eval_allNLI-dev_cosine_f1_threshold": 0.8565204739570618, | |
| "eval_allNLI-dev_cosine_precision": 0.39143730886850153, | |
| "eval_allNLI-dev_cosine_recall": 0.7398843930635838, | |
| "eval_allNLI-dev_dot_accuracy": 0.67578125, | |
| "eval_allNLI-dev_dot_accuracy_threshold": 726.30615234375, | |
| "eval_allNLI-dev_dot_ap": 0.42647535250956575, | |
| "eval_allNLI-dev_dot_f1": 0.512, | |
| "eval_allNLI-dev_dot_f1_threshold": 658.1103515625, | |
| "eval_allNLI-dev_dot_precision": 0.39143730886850153, | |
| "eval_allNLI-dev_dot_recall": 0.7398843930635838, | |
| "eval_allNLI-dev_euclidean_accuracy": 0.67578125, | |
| "eval_allNLI-dev_euclidean_accuracy_threshold": 9.171283721923828, | |
| "eval_allNLI-dev_euclidean_ap": 0.4264736612515921, | |
| "eval_allNLI-dev_euclidean_f1": 0.512, | |
| "eval_allNLI-dev_euclidean_f1_threshold": 14.84876823425293, | |
| "eval_allNLI-dev_euclidean_precision": 0.39143730886850153, | |
| "eval_allNLI-dev_euclidean_recall": 0.7398843930635838, | |
| "eval_allNLI-dev_manhattan_accuracy": 0.67578125, | |
| "eval_allNLI-dev_manhattan_accuracy_threshold": 201.49061584472656, | |
| "eval_allNLI-dev_manhattan_ap": 0.4252213828672732, | |
| "eval_allNLI-dev_manhattan_f1": 0.5107692307692308, | |
| "eval_allNLI-dev_manhattan_f1_threshold": 417.52728271484375, | |
| "eval_allNLI-dev_manhattan_precision": 0.3480083857442348, | |
| "eval_allNLI-dev_manhattan_recall": 0.9595375722543352, | |
| "eval_allNLI-dev_max_accuracy": 0.67578125, | |
| "eval_allNLI-dev_max_accuracy_threshold": 726.30615234375, | |
| "eval_allNLI-dev_max_ap": 0.42647535250956575, | |
| "eval_allNLI-dev_max_f1": 0.512, | |
| "eval_allNLI-dev_max_f1_threshold": 658.1103515625, | |
| "eval_allNLI-dev_max_precision": 0.39143730886850153, | |
| "eval_allNLI-dev_max_recall": 0.9595375722543352, | |
| "eval_loss": 2.2652623653411865, | |
| "eval_runtime": 50.7627, | |
| "eval_samples_per_second": 32.78, | |
| "eval_sequential_score": 0.6461682282377894, | |
| "eval_steps_per_second": 0.138, | |
| "eval_sts-test_pearson_cosine": 0.2749904272806095, | |
| "eval_sts-test_pearson_dot": 0.27496363262371837, | |
| "eval_sts-test_pearson_euclidean": 0.2934483033082174, | |
| "eval_sts-test_pearson_manhattan": 0.2923996087310511, | |
| "eval_sts-test_pearson_max": 0.2934483033082174, | |
| "eval_sts-test_spearman_cosine": 0.31159390381099095, | |
| "eval_sts-test_spearman_dot": 0.31138581044552094, | |
| "eval_sts-test_spearman_euclidean": 0.3115817314678925, | |
| "eval_sts-test_spearman_manhattan": 0.3095556181083969, | |
| "eval_sts-test_spearman_max": 0.31159390381099095, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.1515748031496063, | |
| "grad_norm": 22.330442428588867, | |
| "learning_rate": 1.5009940357852886e-07, | |
| "loss": 3.1149, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.15255905511811024, | |
| "grad_norm": 23.656953811645508, | |
| "learning_rate": 1.510934393638171e-07, | |
| "loss": 2.2976, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.15354330708661418, | |
| "grad_norm": 20.271608352661133, | |
| "learning_rate": 1.5208747514910539e-07, | |
| "loss": 2.4436, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.1545275590551181, | |
| "grad_norm": 25.410293579101562, | |
| "learning_rate": 1.5308151093439367e-07, | |
| "loss": 2.8826, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.15551181102362205, | |
| "grad_norm": 23.772783279418945, | |
| "learning_rate": 1.5407554671968192e-07, | |
| "loss": 2.3664, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.15649606299212598, | |
| "grad_norm": 23.44937515258789, | |
| "learning_rate": 1.5506958250497022e-07, | |
| "loss": 2.2485, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.15748031496062992, | |
| "grad_norm": 23.024261474609375, | |
| "learning_rate": 1.5606361829025848e-07, | |
| "loss": 2.5167, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.15846456692913385, | |
| "grad_norm": 20.63090705871582, | |
| "learning_rate": 1.5705765407554675e-07, | |
| "loss": 1.7183, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.1594488188976378, | |
| "grad_norm": 21.573190689086914, | |
| "learning_rate": 1.58051689860835e-07, | |
| "loss": 2.1003, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.16043307086614172, | |
| "grad_norm": 23.774974822998047, | |
| "learning_rate": 1.5904572564612329e-07, | |
| "loss": 2.5785, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.16141732283464566, | |
| "grad_norm": 27.151123046875, | |
| "learning_rate": 1.6003976143141154e-07, | |
| "loss": 2.8789, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.1624015748031496, | |
| "grad_norm": 23.50958824157715, | |
| "learning_rate": 1.6103379721669984e-07, | |
| "loss": 2.3425, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.16338582677165353, | |
| "grad_norm": 23.77661895751953, | |
| "learning_rate": 1.620278330019881e-07, | |
| "loss": 2.0966, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.1643700787401575, | |
| "grad_norm": 22.070526123046875, | |
| "learning_rate": 1.6302186878727637e-07, | |
| "loss": 2.1126, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.16535433070866143, | |
| "grad_norm": 22.653602600097656, | |
| "learning_rate": 1.6401590457256465e-07, | |
| "loss": 2.1824, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.16633858267716536, | |
| "grad_norm": 21.470808029174805, | |
| "learning_rate": 1.650099403578529e-07, | |
| "loss": 2.2009, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.1673228346456693, | |
| "grad_norm": 25.822694778442383, | |
| "learning_rate": 1.6600397614314118e-07, | |
| "loss": 2.3796, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.16830708661417323, | |
| "grad_norm": 22.609458923339844, | |
| "learning_rate": 1.6699801192842944e-07, | |
| "loss": 2.3096, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.16929133858267717, | |
| "grad_norm": 24.10075569152832, | |
| "learning_rate": 1.6799204771371774e-07, | |
| "loss": 2.7897, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.1702755905511811, | |
| "grad_norm": 22.21641731262207, | |
| "learning_rate": 1.68986083499006e-07, | |
| "loss": 2.2097, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.17125984251968504, | |
| "grad_norm": 17.717933654785156, | |
| "learning_rate": 1.6998011928429427e-07, | |
| "loss": 1.7508, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.17224409448818898, | |
| "grad_norm": 22.352798461914062, | |
| "learning_rate": 1.7097415506958253e-07, | |
| "loss": 2.353, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.1732283464566929, | |
| "grad_norm": 23.421472549438477, | |
| "learning_rate": 1.719681908548708e-07, | |
| "loss": 2.4276, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.17421259842519685, | |
| "grad_norm": 20.41706657409668, | |
| "learning_rate": 1.7296222664015906e-07, | |
| "loss": 2.1016, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.17519685039370078, | |
| "grad_norm": 19.39253807067871, | |
| "learning_rate": 1.7395626242544734e-07, | |
| "loss": 1.8461, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.17618110236220472, | |
| "grad_norm": 19.994935989379883, | |
| "learning_rate": 1.7495029821073564e-07, | |
| "loss": 1.8104, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.17716535433070865, | |
| "grad_norm": 24.119571685791016, | |
| "learning_rate": 1.759443339960239e-07, | |
| "loss": 2.6023, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.1781496062992126, | |
| "grad_norm": 28.111419677734375, | |
| "learning_rate": 1.7693836978131217e-07, | |
| "loss": 2.5261, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.17913385826771652, | |
| "grad_norm": 26.244142532348633, | |
| "learning_rate": 1.7793240556660042e-07, | |
| "loss": 2.1053, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.18011811023622049, | |
| "grad_norm": 24.29507827758789, | |
| "learning_rate": 1.789264413518887e-07, | |
| "loss": 1.9712, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.18110236220472442, | |
| "grad_norm": 26.753236770629883, | |
| "learning_rate": 1.7992047713717695e-07, | |
| "loss": 2.4693, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.18208661417322836, | |
| "grad_norm": 23.031953811645508, | |
| "learning_rate": 1.8091451292246523e-07, | |
| "loss": 2.1119, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.1830708661417323, | |
| "grad_norm": 24.926044464111328, | |
| "learning_rate": 1.819085487077535e-07, | |
| "loss": 2.4797, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.18405511811023623, | |
| "grad_norm": 19.047605514526367, | |
| "learning_rate": 1.829025844930418e-07, | |
| "loss": 2.1587, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.18503937007874016, | |
| "grad_norm": 26.785459518432617, | |
| "learning_rate": 1.8389662027833004e-07, | |
| "loss": 1.9578, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.1860236220472441, | |
| "grad_norm": 22.257556915283203, | |
| "learning_rate": 1.8489065606361832e-07, | |
| "loss": 2.1368, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.18700787401574803, | |
| "grad_norm": 24.006427764892578, | |
| "learning_rate": 1.8588469184890657e-07, | |
| "loss": 2.4212, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.18799212598425197, | |
| "grad_norm": 22.14805793762207, | |
| "learning_rate": 1.8687872763419485e-07, | |
| "loss": 1.9591, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.1889763779527559, | |
| "grad_norm": 19.438581466674805, | |
| "learning_rate": 1.8787276341948313e-07, | |
| "loss": 1.5816, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.18996062992125984, | |
| "grad_norm": 19.473068237304688, | |
| "learning_rate": 1.888667992047714e-07, | |
| "loss": 1.4029, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.19094488188976377, | |
| "grad_norm": 22.895261764526367, | |
| "learning_rate": 1.898608349900597e-07, | |
| "loss": 1.9385, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.1919291338582677, | |
| "grad_norm": 22.117504119873047, | |
| "learning_rate": 1.9085487077534794e-07, | |
| "loss": 1.5596, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.19291338582677164, | |
| "grad_norm": 25.059682846069336, | |
| "learning_rate": 1.9184890656063622e-07, | |
| "loss": 1.6663, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.19389763779527558, | |
| "grad_norm": 31.338993072509766, | |
| "learning_rate": 1.9284294234592447e-07, | |
| "loss": 2.0026, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.19488188976377951, | |
| "grad_norm": 25.27398681640625, | |
| "learning_rate": 1.9383697813121275e-07, | |
| "loss": 2.0046, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.19586614173228348, | |
| "grad_norm": 22.62946128845215, | |
| "learning_rate": 1.94831013916501e-07, | |
| "loss": 1.5016, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.1968503937007874, | |
| "grad_norm": 26.856422424316406, | |
| "learning_rate": 1.958250497017893e-07, | |
| "loss": 2.184, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.19783464566929135, | |
| "grad_norm": 27.561872482299805, | |
| "learning_rate": 1.9681908548707756e-07, | |
| "loss": 2.3442, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.19881889763779528, | |
| "grad_norm": 29.985654830932617, | |
| "learning_rate": 1.9781312127236584e-07, | |
| "loss": 2.6981, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.19980314960629922, | |
| "grad_norm": 22.171070098876953, | |
| "learning_rate": 1.9880715705765412e-07, | |
| "loss": 2.5481, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.20078740157480315, | |
| "grad_norm": 26.214073181152344, | |
| "learning_rate": 1.9980119284294237e-07, | |
| "loss": 2.9798, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.2017716535433071, | |
| "grad_norm": 22.690643310546875, | |
| "learning_rate": 2.0079522862823065e-07, | |
| "loss": 2.287, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.20275590551181102, | |
| "grad_norm": 21.528181076049805, | |
| "learning_rate": 2.017892644135189e-07, | |
| "loss": 1.9393, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.20374015748031496, | |
| "grad_norm": 24.61495018005371, | |
| "learning_rate": 2.027833001988072e-07, | |
| "loss": 2.892, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.2047244094488189, | |
| "grad_norm": 21.661781311035156, | |
| "learning_rate": 2.0377733598409546e-07, | |
| "loss": 2.26, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.20570866141732283, | |
| "grad_norm": 21.17841339111328, | |
| "learning_rate": 2.0477137176938374e-07, | |
| "loss": 2.5911, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.20669291338582677, | |
| "grad_norm": 22.297616958618164, | |
| "learning_rate": 2.05765407554672e-07, | |
| "loss": 2.1239, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2076771653543307, | |
| "grad_norm": 18.447376251220703, | |
| "learning_rate": 2.0675944333996027e-07, | |
| "loss": 2.0683, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.20866141732283464, | |
| "grad_norm": 19.692792892456055, | |
| "learning_rate": 2.0775347912524852e-07, | |
| "loss": 1.768, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.20964566929133857, | |
| "grad_norm": 24.793012619018555, | |
| "learning_rate": 2.087475149105368e-07, | |
| "loss": 2.5468, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.2106299212598425, | |
| "grad_norm": 21.60909652709961, | |
| "learning_rate": 2.097415506958251e-07, | |
| "loss": 1.8956, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.21161417322834647, | |
| "grad_norm": 22.114286422729492, | |
| "learning_rate": 2.1073558648111336e-07, | |
| "loss": 2.044, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.2125984251968504, | |
| "grad_norm": 19.02837562561035, | |
| "learning_rate": 2.1172962226640164e-07, | |
| "loss": 1.5721, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.21358267716535434, | |
| "grad_norm": 19.785751342773438, | |
| "learning_rate": 2.127236580516899e-07, | |
| "loss": 1.6278, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.21456692913385828, | |
| "grad_norm": 21.3282470703125, | |
| "learning_rate": 2.1371769383697817e-07, | |
| "loss": 1.7754, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.2155511811023622, | |
| "grad_norm": 25.80916404724121, | |
| "learning_rate": 2.1471172962226642e-07, | |
| "loss": 1.8594, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.21653543307086615, | |
| "grad_norm": 21.912315368652344, | |
| "learning_rate": 2.1570576540755473e-07, | |
| "loss": 1.8309, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.21751968503937008, | |
| "grad_norm": 24.051366806030273, | |
| "learning_rate": 2.1669980119284298e-07, | |
| "loss": 2.0619, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.21850393700787402, | |
| "grad_norm": 24.29237174987793, | |
| "learning_rate": 2.1769383697813126e-07, | |
| "loss": 2.3335, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.21948818897637795, | |
| "grad_norm": 25.850160598754883, | |
| "learning_rate": 2.186878727634195e-07, | |
| "loss": 2.023, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.2204724409448819, | |
| "grad_norm": 27.208112716674805, | |
| "learning_rate": 2.196819085487078e-07, | |
| "loss": 2.1975, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.22145669291338582, | |
| "grad_norm": 22.276878356933594, | |
| "learning_rate": 2.2067594433399604e-07, | |
| "loss": 1.9228, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.22244094488188976, | |
| "grad_norm": 30.213895797729492, | |
| "learning_rate": 2.2166998011928432e-07, | |
| "loss": 2.3565, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.2234251968503937, | |
| "grad_norm": 22.16749382019043, | |
| "learning_rate": 2.2266401590457263e-07, | |
| "loss": 1.896, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.22440944881889763, | |
| "grad_norm": 21.131744384765625, | |
| "learning_rate": 2.2365805168986088e-07, | |
| "loss": 2.0912, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.22539370078740156, | |
| "grad_norm": 25.1036376953125, | |
| "learning_rate": 2.2465208747514916e-07, | |
| "loss": 2.7703, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.2263779527559055, | |
| "grad_norm": 25.316814422607422, | |
| "learning_rate": 2.256461232604374e-07, | |
| "loss": 1.6988, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.22736220472440946, | |
| "grad_norm": 25.363996505737305, | |
| "learning_rate": 2.266401590457257e-07, | |
| "loss": 2.0406, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.2283464566929134, | |
| "grad_norm": 21.906835556030273, | |
| "learning_rate": 2.2763419483101394e-07, | |
| "loss": 1.9288, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.22933070866141733, | |
| "grad_norm": 21.407150268554688, | |
| "learning_rate": 2.2862823061630222e-07, | |
| "loss": 2.0457, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.23031496062992127, | |
| "grad_norm": 21.2374210357666, | |
| "learning_rate": 2.296222664015905e-07, | |
| "loss": 1.7061, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.2312992125984252, | |
| "grad_norm": 20.94179344177246, | |
| "learning_rate": 2.3061630218687878e-07, | |
| "loss": 1.6244, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.23228346456692914, | |
| "grad_norm": 21.845712661743164, | |
| "learning_rate": 2.3161033797216703e-07, | |
| "loss": 2.0241, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.23326771653543307, | |
| "grad_norm": 19.496191024780273, | |
| "learning_rate": 2.326043737574553e-07, | |
| "loss": 1.567, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.234251968503937, | |
| "grad_norm": 21.819353103637695, | |
| "learning_rate": 2.335984095427436e-07, | |
| "loss": 1.8084, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.23523622047244094, | |
| "grad_norm": 27.17051124572754, | |
| "learning_rate": 2.3459244532803184e-07, | |
| "loss": 2.4363, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.23622047244094488, | |
| "grad_norm": 24.850723266601562, | |
| "learning_rate": 2.3558648111332012e-07, | |
| "loss": 1.7532, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2372047244094488, | |
| "grad_norm": 24.120052337646484, | |
| "learning_rate": 2.365805168986084e-07, | |
| "loss": 2.0797, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.23818897637795275, | |
| "grad_norm": 23.708179473876953, | |
| "learning_rate": 2.3757455268389668e-07, | |
| "loss": 1.9562, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.23917322834645668, | |
| "grad_norm": 20.58576774597168, | |
| "learning_rate": 2.385685884691849e-07, | |
| "loss": 1.6751, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.24015748031496062, | |
| "grad_norm": 25.161970138549805, | |
| "learning_rate": 2.395626242544732e-07, | |
| "loss": 2.0265, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.24114173228346455, | |
| "grad_norm": 22.079387664794922, | |
| "learning_rate": 2.4055666003976146e-07, | |
| "loss": 1.6065, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.2421259842519685, | |
| "grad_norm": 20.125717163085938, | |
| "learning_rate": 2.4155069582504976e-07, | |
| "loss": 1.7439, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.24311023622047245, | |
| "grad_norm": 22.855445861816406, | |
| "learning_rate": 2.42544731610338e-07, | |
| "loss": 2.0237, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.2440944881889764, | |
| "grad_norm": 22.16312026977539, | |
| "learning_rate": 2.4353876739562627e-07, | |
| "loss": 1.6128, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.24507874015748032, | |
| "grad_norm": 21.942440032958984, | |
| "learning_rate": 2.445328031809146e-07, | |
| "loss": 1.6581, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.24606299212598426, | |
| "grad_norm": 24.027729034423828, | |
| "learning_rate": 2.4552683896620283e-07, | |
| "loss": 2.1538, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2470472440944882, | |
| "grad_norm": 23.1455078125, | |
| "learning_rate": 2.4652087475149113e-07, | |
| "loss": 2.049, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.24803149606299213, | |
| "grad_norm": 22.017520904541016, | |
| "learning_rate": 2.475149105367794e-07, | |
| "loss": 1.2573, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.24901574803149606, | |
| "grad_norm": 18.747554779052734, | |
| "learning_rate": 2.4850894632206764e-07, | |
| "loss": 1.5619, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 18.277063369750977, | |
| "learning_rate": 2.495029821073559e-07, | |
| "loss": 1.2611, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.25098425196850394, | |
| "grad_norm": 19.781038284301758, | |
| "learning_rate": 2.5049701789264414e-07, | |
| "loss": 1.3443, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.25196850393700787, | |
| "grad_norm": 21.605199813842773, | |
| "learning_rate": 2.5149105367793245e-07, | |
| "loss": 1.3436, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.2529527559055118, | |
| "grad_norm": 29.64748764038086, | |
| "learning_rate": 2.524850894632207e-07, | |
| "loss": 2.8117, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.25393700787401574, | |
| "grad_norm": 19.245962142944336, | |
| "learning_rate": 2.53479125248509e-07, | |
| "loss": 1.7563, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.2549212598425197, | |
| "grad_norm": 18.752405166625977, | |
| "learning_rate": 2.5447316103379726e-07, | |
| "loss": 1.3148, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.2559055118110236, | |
| "grad_norm": 24.133914947509766, | |
| "learning_rate": 2.554671968190855e-07, | |
| "loss": 2.0278, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.25688976377952755, | |
| "grad_norm": 19.490673065185547, | |
| "learning_rate": 2.564612326043738e-07, | |
| "loss": 1.2403, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.2578740157480315, | |
| "grad_norm": 20.14169692993164, | |
| "learning_rate": 2.5745526838966207e-07, | |
| "loss": 1.588, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.2588582677165354, | |
| "grad_norm": 26.948959350585938, | |
| "learning_rate": 2.5844930417495037e-07, | |
| "loss": 2.0071, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.25984251968503935, | |
| "grad_norm": 22.330224990844727, | |
| "learning_rate": 2.5944333996023857e-07, | |
| "loss": 1.5312, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.2608267716535433, | |
| "grad_norm": 21.379629135131836, | |
| "learning_rate": 2.604373757455269e-07, | |
| "loss": 1.8641, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.2618110236220472, | |
| "grad_norm": 20.703140258789062, | |
| "learning_rate": 2.614314115308152e-07, | |
| "loss": 1.2933, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.26279527559055116, | |
| "grad_norm": 22.9117488861084, | |
| "learning_rate": 2.6242544731610343e-07, | |
| "loss": 1.6262, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.2637795275590551, | |
| "grad_norm": 23.842002868652344, | |
| "learning_rate": 2.634194831013917e-07, | |
| "loss": 1.721, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.26476377952755903, | |
| "grad_norm": 20.449384689331055, | |
| "learning_rate": 2.6441351888667994e-07, | |
| "loss": 1.4713, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.265748031496063, | |
| "grad_norm": 25.885969161987305, | |
| "learning_rate": 2.6540755467196824e-07, | |
| "loss": 1.4625, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.26673228346456695, | |
| "grad_norm": 24.52666473388672, | |
| "learning_rate": 2.664015904572565e-07, | |
| "loss": 1.7254, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.2677165354330709, | |
| "grad_norm": 23.4957275390625, | |
| "learning_rate": 2.6739562624254475e-07, | |
| "loss": 1.5108, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.2687007874015748, | |
| "grad_norm": 23.828855514526367, | |
| "learning_rate": 2.6838966202783305e-07, | |
| "loss": 2.1126, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.26968503937007876, | |
| "grad_norm": 21.05967903137207, | |
| "learning_rate": 2.693836978131213e-07, | |
| "loss": 1.3967, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.2706692913385827, | |
| "grad_norm": 24.555776596069336, | |
| "learning_rate": 2.703777335984096e-07, | |
| "loss": 1.7067, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.27165354330708663, | |
| "grad_norm": 22.135860443115234, | |
| "learning_rate": 2.7137176938369786e-07, | |
| "loss": 1.4847, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.27263779527559057, | |
| "grad_norm": 22.105913162231445, | |
| "learning_rate": 2.723658051689861e-07, | |
| "loss": 1.6515, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.2736220472440945, | |
| "grad_norm": 16.617225646972656, | |
| "learning_rate": 2.7335984095427437e-07, | |
| "loss": 0.9367, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.27460629921259844, | |
| "grad_norm": 26.01727867126465, | |
| "learning_rate": 2.743538767395627e-07, | |
| "loss": 2.0267, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.2755905511811024, | |
| "grad_norm": 22.522462844848633, | |
| "learning_rate": 2.75347912524851e-07, | |
| "loss": 1.5023, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.2765748031496063, | |
| "grad_norm": 20.646358489990234, | |
| "learning_rate": 2.763419483101392e-07, | |
| "loss": 1.1248, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.27755905511811024, | |
| "grad_norm": 23.3087158203125, | |
| "learning_rate": 2.773359840954275e-07, | |
| "loss": 1.6224, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.2785433070866142, | |
| "grad_norm": 24.115968704223633, | |
| "learning_rate": 2.7833001988071574e-07, | |
| "loss": 1.7969, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.2795275590551181, | |
| "grad_norm": 27.229188919067383, | |
| "learning_rate": 2.7932405566600404e-07, | |
| "loss": 2.2498, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.28051181102362205, | |
| "grad_norm": 25.797216415405273, | |
| "learning_rate": 2.803180914512923e-07, | |
| "loss": 1.7477, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.281496062992126, | |
| "grad_norm": 23.666858673095703, | |
| "learning_rate": 2.8131212723658055e-07, | |
| "loss": 1.6261, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.2824803149606299, | |
| "grad_norm": 26.826387405395508, | |
| "learning_rate": 2.8230616302186885e-07, | |
| "loss": 2.0911, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.28346456692913385, | |
| "grad_norm": 23.288511276245117, | |
| "learning_rate": 2.833001988071571e-07, | |
| "loss": 1.9519, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.2844488188976378, | |
| "grad_norm": 19.76810073852539, | |
| "learning_rate": 2.842942345924454e-07, | |
| "loss": 1.3132, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.2854330708661417, | |
| "grad_norm": 24.65778923034668, | |
| "learning_rate": 2.852882703777336e-07, | |
| "loss": 2.3292, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.28641732283464566, | |
| "grad_norm": 19.230430603027344, | |
| "learning_rate": 2.862823061630219e-07, | |
| "loss": 1.3781, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.2874015748031496, | |
| "grad_norm": 20.665143966674805, | |
| "learning_rate": 2.8727634194831017e-07, | |
| "loss": 1.5753, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.28838582677165353, | |
| "grad_norm": 18.584993362426758, | |
| "learning_rate": 2.8827037773359847e-07, | |
| "loss": 1.4158, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.28937007874015747, | |
| "grad_norm": 20.731081008911133, | |
| "learning_rate": 2.892644135188867e-07, | |
| "loss": 2.1661, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.2903543307086614, | |
| "grad_norm": 20.006755828857422, | |
| "learning_rate": 2.90258449304175e-07, | |
| "loss": 1.4928, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.29133858267716534, | |
| "grad_norm": 20.01453971862793, | |
| "learning_rate": 2.912524850894633e-07, | |
| "loss": 2.2825, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.29232283464566927, | |
| "grad_norm": 18.137697219848633, | |
| "learning_rate": 2.9224652087475153e-07, | |
| "loss": 1.7261, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.2933070866141732, | |
| "grad_norm": 20.046411514282227, | |
| "learning_rate": 2.9324055666003984e-07, | |
| "loss": 1.8635, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.29429133858267714, | |
| "grad_norm": 15.859444618225098, | |
| "learning_rate": 2.9423459244532804e-07, | |
| "loss": 0.974, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.2952755905511811, | |
| "grad_norm": 17.976015090942383, | |
| "learning_rate": 2.9522862823061634e-07, | |
| "loss": 1.53, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.296259842519685, | |
| "grad_norm": 20.52704429626465, | |
| "learning_rate": 2.9622266401590465e-07, | |
| "loss": 1.5985, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.297244094488189, | |
| "grad_norm": 19.45134925842285, | |
| "learning_rate": 2.972166998011929e-07, | |
| "loss": 1.2169, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.29822834645669294, | |
| "grad_norm": 22.01141357421875, | |
| "learning_rate": 2.9821073558648115e-07, | |
| "loss": 1.771, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.2992125984251969, | |
| "grad_norm": 17.591617584228516, | |
| "learning_rate": 2.992047713717694e-07, | |
| "loss": 1.4506, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.3001968503937008, | |
| "grad_norm": 25.382244110107422, | |
| "learning_rate": 3.001988071570577e-07, | |
| "loss": 1.9496, | |
| "step": 305 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 3048, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 305, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |