| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9878213802435725, | |
| "eval_steps": 500, | |
| "global_step": 276, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010825439783491205, | |
| "grad_norm": 5.807889292253254, | |
| "learning_rate": 4.04061e-06, | |
| "loss": 0.8567, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02165087956698241, | |
| "grad_norm": 5.811690427529652, | |
| "learning_rate": 8.08122e-06, | |
| "loss": 0.8604, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03247631935047361, | |
| "grad_norm": 4.3700725808471175, | |
| "learning_rate": 1.212183e-05, | |
| "loss": 0.8206, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.04330175913396482, | |
| "grad_norm": 2.118091273430897, | |
| "learning_rate": 1.616244e-05, | |
| "loss": 0.7738, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05412719891745602, | |
| "grad_norm": 5.4648275516158655, | |
| "learning_rate": 2.020305e-05, | |
| "loss": 0.7833, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06495263870094722, | |
| "grad_norm": 8.553109003392324, | |
| "learning_rate": 2.424366e-05, | |
| "loss": 0.7883, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07577807848443843, | |
| "grad_norm": 6.503055944847511, | |
| "learning_rate": 2.828427e-05, | |
| "loss": 0.7834, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.08660351826792964, | |
| "grad_norm": 3.8336486548329196, | |
| "learning_rate": 3.232488e-05, | |
| "loss": 0.7479, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.09742895805142084, | |
| "grad_norm": 2.684849997410019, | |
| "learning_rate": 3.636549e-05, | |
| "loss": 0.6909, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.10825439783491204, | |
| "grad_norm": 2.161503695468669, | |
| "learning_rate": 4.04061e-05, | |
| "loss": 0.6719, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11907983761840325, | |
| "grad_norm": 1.884181921509628, | |
| "learning_rate": 4.444671e-05, | |
| "loss": 0.6462, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.12990527740189445, | |
| "grad_norm": 1.9443412615480906, | |
| "learning_rate": 4.848732e-05, | |
| "loss": 0.6351, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.14073071718538566, | |
| "grad_norm": 1.541881119258122, | |
| "learning_rate": 5.2527930000000004e-05, | |
| "loss": 0.6228, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.15155615696887687, | |
| "grad_norm": 1.476390312725872, | |
| "learning_rate": 5.656854e-05, | |
| "loss": 0.6311, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.16238159675236807, | |
| "grad_norm": 2.344926743941767, | |
| "learning_rate": 6.060915e-05, | |
| "loss": 0.6178, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.17320703653585928, | |
| "grad_norm": 1.4098664598368982, | |
| "learning_rate": 6.464976e-05, | |
| "loss": 0.604, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.18403247631935046, | |
| "grad_norm": 2.1013992594295536, | |
| "learning_rate": 6.869037e-05, | |
| "loss": 0.6064, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.19485791610284167, | |
| "grad_norm": 3.729407760648478, | |
| "learning_rate": 7.273098e-05, | |
| "loss": 0.5883, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.20568335588633288, | |
| "grad_norm": 1.9287898977179496, | |
| "learning_rate": 7.677159e-05, | |
| "loss": 0.5875, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2165087956698241, | |
| "grad_norm": 2.141344286692683, | |
| "learning_rate": 8.08122e-05, | |
| "loss": 0.5996, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2273342354533153, | |
| "grad_norm": 1.718091978310923, | |
| "learning_rate": 8.485281e-05, | |
| "loss": 0.5792, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2381596752368065, | |
| "grad_norm": 1.5391109254273139, | |
| "learning_rate": 8.889342e-05, | |
| "loss": 0.5806, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2489851150202977, | |
| "grad_norm": 1.5474160910291503, | |
| "learning_rate": 9.293403e-05, | |
| "loss": 0.573, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2598105548037889, | |
| "grad_norm": 2.015119305794227, | |
| "learning_rate": 9.697464e-05, | |
| "loss": 0.5753, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2706359945872801, | |
| "grad_norm": 1.2610566881703422, | |
| "learning_rate": 0.00010101525000000001, | |
| "loss": 0.564, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2814614343707713, | |
| "grad_norm": 2.2902330489952347, | |
| "learning_rate": 0.00010505586000000001, | |
| "loss": 0.5529, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2922868741542625, | |
| "grad_norm": 1.3601476225980274, | |
| "learning_rate": 0.00010909647, | |
| "loss": 0.5578, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.30311231393775373, | |
| "grad_norm": 2.5182389116157196, | |
| "learning_rate": 0.00011313708, | |
| "loss": 0.5725, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.31393775372124494, | |
| "grad_norm": 2.1290153329265697, | |
| "learning_rate": 0.00011313254125743567, | |
| "loss": 0.5751, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.32476319350473615, | |
| "grad_norm": 1.3037114858673693, | |
| "learning_rate": 0.00011311892575806915, | |
| "loss": 0.5513, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.33558863328822736, | |
| "grad_norm": 1.950503866929347, | |
| "learning_rate": 0.00011309623568676317, | |
| "loss": 0.5561, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.34641407307171856, | |
| "grad_norm": 1.3015061613988743, | |
| "learning_rate": 0.00011306447468456599, | |
| "loss": 0.5524, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3572395128552097, | |
| "grad_norm": 2.844778505243191, | |
| "learning_rate": 0.0001130236478481272, | |
| "loss": 0.56, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3680649526387009, | |
| "grad_norm": 1.6940649560718253, | |
| "learning_rate": 0.00011297376172887978, | |
| "loss": 0.5715, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.37889039242219213, | |
| "grad_norm": 2.296456387550305, | |
| "learning_rate": 0.00011291482433198894, | |
| "loss": 0.5509, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.38971583220568334, | |
| "grad_norm": 2.2983132666819706, | |
| "learning_rate": 0.00011284684511506735, | |
| "loss": 0.5394, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.40054127198917455, | |
| "grad_norm": 1.4966185446845217, | |
| "learning_rate": 0.00011276983498665771, | |
| "loss": 0.5478, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.41136671177266576, | |
| "grad_norm": 2.0647499571970616, | |
| "learning_rate": 0.00011268380630448204, | |
| "loss": 0.535, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.42219215155615697, | |
| "grad_norm": 12.286624512326911, | |
| "learning_rate": 0.00011258877287345884, | |
| "loss": 0.6342, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.4330175913396482, | |
| "grad_norm": 142.759088551662, | |
| "learning_rate": 0.00011248474994348775, | |
| "loss": 0.9216, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4438430311231394, | |
| "grad_norm": 4.27084515384108, | |
| "learning_rate": 0.00011237175420700238, | |
| "loss": 0.6257, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.4546684709066306, | |
| "grad_norm": 1.3276672657350452, | |
| "learning_rate": 0.00011224980379629178, | |
| "loss": 0.5676, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4654939106901218, | |
| "grad_norm": 2.5925858366379186, | |
| "learning_rate": 0.0001121189182805907, | |
| "loss": 0.5571, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.476319350473613, | |
| "grad_norm": 1.8224855694265998, | |
| "learning_rate": 0.00011197911866293939, | |
| "loss": 0.5752, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.4871447902571042, | |
| "grad_norm": 1.9153712099066185, | |
| "learning_rate": 0.00011183042737681327, | |
| "loss": 0.5675, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4979702300405954, | |
| "grad_norm": 1.4984797297805004, | |
| "learning_rate": 0.00011167286828252297, | |
| "loss": 0.5509, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.5087956698240866, | |
| "grad_norm": 1.3000230619923165, | |
| "learning_rate": 0.00011150646666338564, | |
| "loss": 0.5437, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.5196211096075778, | |
| "grad_norm": 1.1273300703794096, | |
| "learning_rate": 0.00011133124922166763, | |
| "loss": 0.5353, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.530446549391069, | |
| "grad_norm": 0.9939902978944836, | |
| "learning_rate": 0.00011114724407429968, | |
| "loss": 0.532, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5412719891745602, | |
| "grad_norm": 1.4227558376445275, | |
| "learning_rate": 0.00011095448074836505, | |
| "loss": 0.5445, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5520974289580515, | |
| "grad_norm": 1.6500800441263705, | |
| "learning_rate": 0.00011075299017636123, | |
| "loss": 0.5331, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.5629228687415426, | |
| "grad_norm": 0.8867959751888957, | |
| "learning_rate": 0.00011054280469123641, | |
| "loss": 0.5243, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5737483085250338, | |
| "grad_norm": 1.4279560361603063, | |
| "learning_rate": 0.00011032395802120092, | |
| "loss": 0.5243, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.584573748308525, | |
| "grad_norm": 0.956318053267928, | |
| "learning_rate": 0.00011009648528431498, | |
| "loss": 0.5345, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5953991880920162, | |
| "grad_norm": 1.163599419520797, | |
| "learning_rate": 0.00010986042298285322, | |
| "loss": 0.5178, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6062246278755075, | |
| "grad_norm": 1.0528942367951604, | |
| "learning_rate": 0.00010961580899744742, | |
| "loss": 0.5246, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.6170500676589986, | |
| "grad_norm": 1.152436232083363, | |
| "learning_rate": 0.00010936268258100764, | |
| "loss": 0.5218, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.6278755074424899, | |
| "grad_norm": 1.288579266241482, | |
| "learning_rate": 0.00010910108435242351, | |
| "loss": 0.5221, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.638700947225981, | |
| "grad_norm": 1.148211758545437, | |
| "learning_rate": 0.00010883105629004603, | |
| "loss": 0.5022, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6495263870094723, | |
| "grad_norm": 0.8500064313704729, | |
| "learning_rate": 0.00010855264172495148, | |
| "loss": 0.5053, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6603518267929634, | |
| "grad_norm": 1.526146143997704, | |
| "learning_rate": 0.00010826588533398805, | |
| "loss": 0.5204, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.6711772665764547, | |
| "grad_norm": 0.8393491341796541, | |
| "learning_rate": 0.00010797083313260666, | |
| "loss": 0.5043, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6820027063599459, | |
| "grad_norm": 1.484485217333713, | |
| "learning_rate": 0.0001076675324674768, | |
| "loss": 0.5159, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6928281461434371, | |
| "grad_norm": 1.0584547196728273, | |
| "learning_rate": 0.00010735603200888909, | |
| "loss": 0.5075, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.7036535859269283, | |
| "grad_norm": 1.3549586079445122, | |
| "learning_rate": 0.00010703638174294492, | |
| "loss": 0.5008, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7144790257104194, | |
| "grad_norm": 0.6668671224564001, | |
| "learning_rate": 0.00010670863296353553, | |
| "loss": 0.5041, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.7253044654939107, | |
| "grad_norm": 1.2494202811591621, | |
| "learning_rate": 0.00010637283826411074, | |
| "loss": 0.5023, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.7361299052774019, | |
| "grad_norm": 0.9215229237806866, | |
| "learning_rate": 0.00010602905152923944, | |
| "loss": 0.5114, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.7469553450608931, | |
| "grad_norm": 1.0110141606592546, | |
| "learning_rate": 0.00010567732792596276, | |
| "loss": 0.5007, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7577807848443843, | |
| "grad_norm": 0.8129099507326153, | |
| "learning_rate": 0.00010531772389494152, | |
| "loss": 0.4829, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7686062246278755, | |
| "grad_norm": 0.8894581968902109, | |
| "learning_rate": 0.0001049502971413992, | |
| "loss": 0.4843, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.7794316644113667, | |
| "grad_norm": 1.0766490523242207, | |
| "learning_rate": 0.00010457510662586213, | |
| "loss": 0.4952, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.790257104194858, | |
| "grad_norm": 0.9096596609582451, | |
| "learning_rate": 0.00010419221255469808, | |
| "loss": 0.4991, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.8010825439783491, | |
| "grad_norm": 0.6679397460269166, | |
| "learning_rate": 0.00010380167637045515, | |
| "loss": 0.4934, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.8119079837618404, | |
| "grad_norm": 0.5550572181240487, | |
| "learning_rate": 0.00010340356074200197, | |
| "loss": 0.488, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.8227334235453315, | |
| "grad_norm": 0.5738282924820368, | |
| "learning_rate": 0.00010299792955447147, | |
| "loss": 0.4813, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.8335588633288228, | |
| "grad_norm": 0.6379614597485335, | |
| "learning_rate": 0.00010258484789900923, | |
| "loss": 0.4793, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.8443843031123139, | |
| "grad_norm": 1.1998908169361215, | |
| "learning_rate": 0.00010216438206232844, | |
| "loss": 0.4868, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.8552097428958051, | |
| "grad_norm": 0.957795674849162, | |
| "learning_rate": 0.00010173659951607282, | |
| "loss": 0.4896, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8660351826792964, | |
| "grad_norm": 1.0879462318136461, | |
| "learning_rate": 0.00010130156890598967, | |
| "loss": 0.484, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8768606224627875, | |
| "grad_norm": 1.0365480887213832, | |
| "learning_rate": 0.00010085936004091438, | |
| "loss": 0.491, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.8876860622462788, | |
| "grad_norm": 0.8798367719337482, | |
| "learning_rate": 0.0001004100438815681, | |
| "loss": 0.4779, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.8985115020297699, | |
| "grad_norm": 1.0902126566790828, | |
| "learning_rate": 9.995369252917098e-05, | |
| "loss": 0.4851, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.9093369418132612, | |
| "grad_norm": 1.3512814806916393, | |
| "learning_rate": 9.949037921387205e-05, | |
| "loss": 0.4854, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.9201623815967523, | |
| "grad_norm": 0.7496109448453503, | |
| "learning_rate": 9.902017828299812e-05, | |
| "loss": 0.477, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.9309878213802436, | |
| "grad_norm": 0.8284390728674584, | |
| "learning_rate": 9.854316518912328e-05, | |
| "loss": 0.4856, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.9418132611637348, | |
| "grad_norm": 0.9111112551516629, | |
| "learning_rate": 9.805941647796121e-05, | |
| "loss": 0.4813, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.952638700947226, | |
| "grad_norm": 0.9417845675499721, | |
| "learning_rate": 9.756900977608196e-05, | |
| "loss": 0.4655, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.9634641407307172, | |
| "grad_norm": 1.2726325932306266, | |
| "learning_rate": 9.707202377845526e-05, | |
| "loss": 0.4848, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.9742895805142084, | |
| "grad_norm": 0.8519456352856999, | |
| "learning_rate": 9.656853823582246e-05, | |
| "loss": 0.4783, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.9851150202976996, | |
| "grad_norm": 1.2137981567024454, | |
| "learning_rate": 9.605863394189905e-05, | |
| "loss": 0.4808, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.9959404600811907, | |
| "grad_norm": 0.8493809236019717, | |
| "learning_rate": 9.554239272040974e-05, | |
| "loss": 0.4794, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.006765899864682, | |
| "grad_norm": 1.1076546554118754, | |
| "learning_rate": 9.501989741195838e-05, | |
| "loss": 0.7428, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.0175913396481733, | |
| "grad_norm": 1.6305614136858817, | |
| "learning_rate": 9.44912318607345e-05, | |
| "loss": 0.4696, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.0284167794316643, | |
| "grad_norm": 0.7416433203255349, | |
| "learning_rate": 9.395648090105913e-05, | |
| "loss": 0.4548, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.0392422192151556, | |
| "grad_norm": 1.3812445546043028, | |
| "learning_rate": 9.341573034377135e-05, | |
| "loss": 0.4597, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0500676589986468, | |
| "grad_norm": 0.9119824983143324, | |
| "learning_rate": 9.28690669624584e-05, | |
| "loss": 0.4469, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.060893098782138, | |
| "grad_norm": 1.1446563492957806, | |
| "learning_rate": 9.231657847953129e-05, | |
| "loss": 0.4579, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.0717185385656292, | |
| "grad_norm": 0.7486659375486545, | |
| "learning_rate": 9.175835355214798e-05, | |
| "loss": 0.4525, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.0825439783491204, | |
| "grad_norm": 0.751379457270728, | |
| "learning_rate": 9.119448175798677e-05, | |
| "loss": 0.4465, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0933694181326117, | |
| "grad_norm": 0.6717441683857258, | |
| "learning_rate": 9.062505358087179e-05, | |
| "loss": 0.449, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.104194857916103, | |
| "grad_norm": 0.7286206911922811, | |
| "learning_rate": 9.005016039625325e-05, | |
| "loss": 0.4427, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.115020297699594, | |
| "grad_norm": 0.6470798687068924, | |
| "learning_rate": 8.94698944565445e-05, | |
| "loss": 0.4495, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.1258457374830853, | |
| "grad_norm": 0.6480605622628216, | |
| "learning_rate": 8.888434887631836e-05, | |
| "loss": 0.4389, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.1366711772665765, | |
| "grad_norm": 0.4296944063679475, | |
| "learning_rate": 8.829361761736521e-05, | |
| "loss": 0.4423, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.1474966170500678, | |
| "grad_norm": 0.6099046152220434, | |
| "learning_rate": 8.769779547361498e-05, | |
| "loss": 0.4474, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.1583220568335588, | |
| "grad_norm": 0.4360950041119209, | |
| "learning_rate": 8.709697805592578e-05, | |
| "loss": 0.4404, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.16914749661705, | |
| "grad_norm": 0.5017314757670758, | |
| "learning_rate": 8.649126177674119e-05, | |
| "loss": 0.436, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.1799729364005414, | |
| "grad_norm": 0.49957087979350345, | |
| "learning_rate": 8.588074383461926e-05, | |
| "loss": 0.441, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.1907983761840324, | |
| "grad_norm": 0.492100303371075, | |
| "learning_rate": 8.526552219863502e-05, | |
| "loss": 0.446, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.2016238159675237, | |
| "grad_norm": 0.5107451733829671, | |
| "learning_rate": 8.464569559265956e-05, | |
| "loss": 0.4437, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.212449255751015, | |
| "grad_norm": 0.5195883033342021, | |
| "learning_rate": 8.402136347951797e-05, | |
| "loss": 0.4383, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.2232746955345062, | |
| "grad_norm": 0.484203520160995, | |
| "learning_rate": 8.339262604502857e-05, | |
| "loss": 0.4345, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.2341001353179972, | |
| "grad_norm": 0.4052565892354387, | |
| "learning_rate": 8.275958418192629e-05, | |
| "loss": 0.4406, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.2449255751014885, | |
| "grad_norm": 0.44019644743443515, | |
| "learning_rate": 8.212233947367252e-05, | |
| "loss": 0.4469, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.2557510148849798, | |
| "grad_norm": 0.5974100344045092, | |
| "learning_rate": 8.148099417815411e-05, | |
| "loss": 0.4383, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.266576454668471, | |
| "grad_norm": 0.645910343994611, | |
| "learning_rate": 8.083565121127416e-05, | |
| "loss": 0.4365, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.277401894451962, | |
| "grad_norm": 0.7963176183442402, | |
| "learning_rate": 8.01864141304373e-05, | |
| "loss": 0.4373, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.2882273342354533, | |
| "grad_norm": 1.1462155724648733, | |
| "learning_rate": 7.953338711793185e-05, | |
| "loss": 0.4423, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.2990527740189446, | |
| "grad_norm": 0.8620914093895499, | |
| "learning_rate": 7.887667496421197e-05, | |
| "loss": 0.4335, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.3098782138024356, | |
| "grad_norm": 0.8473383837505596, | |
| "learning_rate": 7.821638305108189e-05, | |
| "loss": 0.4398, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.320703653585927, | |
| "grad_norm": 1.0572991577428552, | |
| "learning_rate": 7.755261733478551e-05, | |
| "loss": 0.4367, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.3315290933694182, | |
| "grad_norm": 1.140899053275173, | |
| "learning_rate": 7.688548432900373e-05, | |
| "loss": 0.4394, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.3423545331529092, | |
| "grad_norm": 0.5348253131421844, | |
| "learning_rate": 7.621509108776232e-05, | |
| "loss": 0.4347, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.3531799729364005, | |
| "grad_norm": 0.4277663692986505, | |
| "learning_rate": 7.554154518825313e-05, | |
| "loss": 0.4353, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.3640054127198917, | |
| "grad_norm": 0.8923758173820789, | |
| "learning_rate": 7.486495471357116e-05, | |
| "loss": 0.4353, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.374830852503383, | |
| "grad_norm": 1.0914091592056432, | |
| "learning_rate": 7.418542823537082e-05, | |
| "loss": 0.4293, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.3856562922868743, | |
| "grad_norm": 0.493002517578744, | |
| "learning_rate": 7.350307479644334e-05, | |
| "loss": 0.4289, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.3964817320703653, | |
| "grad_norm": 0.47827880244934046, | |
| "learning_rate": 7.281800389321898e-05, | |
| "loss": 0.4289, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.4073071718538566, | |
| "grad_norm": 0.9376108017080121, | |
| "learning_rate": 7.21303254581962e-05, | |
| "loss": 0.4358, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.4181326116373478, | |
| "grad_norm": 0.9428376309570359, | |
| "learning_rate": 7.144014984230094e-05, | |
| "loss": 0.4362, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.4289580514208389, | |
| "grad_norm": 0.5757201367588043, | |
| "learning_rate": 7.074758779717872e-05, | |
| "loss": 0.4369, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.4397834912043301, | |
| "grad_norm": 0.6048455591912312, | |
| "learning_rate": 7.005275045742245e-05, | |
| "loss": 0.4295, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.4506089309878214, | |
| "grad_norm": 0.523998052532321, | |
| "learning_rate": 6.935574932273883e-05, | |
| "loss": 0.4315, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.4614343707713127, | |
| "grad_norm": 0.5397427917080534, | |
| "learning_rate": 6.865669624005612e-05, | |
| "loss": 0.4329, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.472259810554804, | |
| "grad_norm": 0.5235978832989827, | |
| "learning_rate": 6.795570338557618e-05, | |
| "loss": 0.4313, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.483085250338295, | |
| "grad_norm": 0.3129416083173963, | |
| "learning_rate": 6.72528832467737e-05, | |
| "loss": 0.4279, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.4939106901217862, | |
| "grad_norm": 0.4970063830078409, | |
| "learning_rate": 6.65483486043455e-05, | |
| "loss": 0.4341, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.5047361299052775, | |
| "grad_norm": 0.47909790019765314, | |
| "learning_rate": 6.584221251411267e-05, | |
| "loss": 0.433, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.5155615696887685, | |
| "grad_norm": 0.3226892723695655, | |
| "learning_rate": 6.513458828887876e-05, | |
| "loss": 0.4319, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.5263870094722598, | |
| "grad_norm": 0.4440079453972877, | |
| "learning_rate": 6.442558948024646e-05, | |
| "loss": 0.429, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.537212449255751, | |
| "grad_norm": 0.5152686044377337, | |
| "learning_rate": 6.371532986039628e-05, | |
| "loss": 0.4293, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.548037889039242, | |
| "grad_norm": 0.43779509661289856, | |
| "learning_rate": 6.30039234038295e-05, | |
| "loss": 0.436, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.5588633288227334, | |
| "grad_norm": 0.31984409816781834, | |
| "learning_rate": 6.229148426907894e-05, | |
| "loss": 0.4274, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.5696887686062246, | |
| "grad_norm": 0.27406976333777083, | |
| "learning_rate": 6.157812678039005e-05, | |
| "loss": 0.4254, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.5805142083897157, | |
| "grad_norm": 0.28606524914172654, | |
| "learning_rate": 6.086396540937541e-05, | |
| "loss": 0.4254, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.5913396481732072, | |
| "grad_norm": 0.31699705967037795, | |
| "learning_rate": 6.014911475664562e-05, | |
| "loss": 0.4222, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.6021650879566982, | |
| "grad_norm": 0.27781981741999395, | |
| "learning_rate": 5.9433689533419476e-05, | |
| "loss": 0.4411, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.6129905277401895, | |
| "grad_norm": 0.31382913986251637, | |
| "learning_rate": 5.871780454311643e-05, | |
| "loss": 0.4235, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.6238159675236807, | |
| "grad_norm": 0.31850123273660286, | |
| "learning_rate": 5.8001574662934144e-05, | |
| "loss": 0.4303, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.6346414073071718, | |
| "grad_norm": 0.24056778596700748, | |
| "learning_rate": 5.7285114825414433e-05, | |
| "loss": 0.4311, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.645466847090663, | |
| "grad_norm": 0.2586338153936464, | |
| "learning_rate": 5.656854e-05, | |
| "loss": 0.4247, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.6562922868741543, | |
| "grad_norm": 0.23655283610366792, | |
| "learning_rate": 5.5851965174585584e-05, | |
| "loss": 0.4368, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.6671177266576453, | |
| "grad_norm": 0.2592063875823944, | |
| "learning_rate": 5.5135505337065866e-05, | |
| "loss": 0.4294, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.6779431664411368, | |
| "grad_norm": 0.25226779767266616, | |
| "learning_rate": 5.441927545688358e-05, | |
| "loss": 0.4225, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.6887686062246279, | |
| "grad_norm": 0.216123495205421, | |
| "learning_rate": 5.3703390466580534e-05, | |
| "loss": 0.4302, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.699594046008119, | |
| "grad_norm": 0.2822002037132782, | |
| "learning_rate": 5.2987965243354394e-05, | |
| "loss": 0.4171, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.7104194857916104, | |
| "grad_norm": 0.21868337726749193, | |
| "learning_rate": 5.22731145906246e-05, | |
| "loss": 0.4278, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.7212449255751014, | |
| "grad_norm": 0.2416291692003587, | |
| "learning_rate": 5.1558953219609965e-05, | |
| "loss": 0.4344, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.7320703653585927, | |
| "grad_norm": 0.21904695751514086, | |
| "learning_rate": 5.084559573092107e-05, | |
| "loss": 0.4288, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.742895805142084, | |
| "grad_norm": 0.23360874499091466, | |
| "learning_rate": 5.013315659617052e-05, | |
| "loss": 0.4258, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.753721244925575, | |
| "grad_norm": 0.220076679488733, | |
| "learning_rate": 4.9421750139603754e-05, | |
| "loss": 0.4311, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.7645466847090663, | |
| "grad_norm": 0.2620382815091359, | |
| "learning_rate": 4.871149051975356e-05, | |
| "loss": 0.4244, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.7753721244925575, | |
| "grad_norm": 0.20533381271916243, | |
| "learning_rate": 4.8002491711121275e-05, | |
| "loss": 0.4214, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.7861975642760486, | |
| "grad_norm": 0.21959248207061696, | |
| "learning_rate": 4.729486748588734e-05, | |
| "loss": 0.4286, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.79702300405954, | |
| "grad_norm": 0.24544077165398928, | |
| "learning_rate": 4.6588731395654526e-05, | |
| "loss": 0.4214, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.8078484438430311, | |
| "grad_norm": 0.21079099929633802, | |
| "learning_rate": 4.588419675322632e-05, | |
| "loss": 0.4271, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.8186738836265224, | |
| "grad_norm": 0.27737463865816947, | |
| "learning_rate": 4.5181376614423824e-05, | |
| "loss": 0.4242, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.8294993234100136, | |
| "grad_norm": 0.18815438760526704, | |
| "learning_rate": 4.448038375994388e-05, | |
| "loss": 0.4321, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.8403247631935047, | |
| "grad_norm": 0.22645725922608995, | |
| "learning_rate": 4.3781330677261174e-05, | |
| "loss": 0.4231, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.851150202976996, | |
| "grad_norm": 0.2511332563566454, | |
| "learning_rate": 4.308432954257754e-05, | |
| "loss": 0.4193, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.8619756427604872, | |
| "grad_norm": 0.2133202149757832, | |
| "learning_rate": 4.238949220282127e-05, | |
| "loss": 0.4206, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.8728010825439783, | |
| "grad_norm": 0.22435313580130933, | |
| "learning_rate": 4.169693015769905e-05, | |
| "loss": 0.4144, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.8836265223274695, | |
| "grad_norm": 0.2702204492113976, | |
| "learning_rate": 4.1006754541803794e-05, | |
| "loss": 0.4271, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.8944519621109608, | |
| "grad_norm": 0.21199239148820068, | |
| "learning_rate": 4.031907610678103e-05, | |
| "loss": 0.4214, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.9052774018944518, | |
| "grad_norm": 0.23353872965172265, | |
| "learning_rate": 3.963400520355668e-05, | |
| "loss": 0.4162, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.9161028416779433, | |
| "grad_norm": 0.26040652564822436, | |
| "learning_rate": 3.89516517646292e-05, | |
| "loss": 0.4184, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.9269282814614344, | |
| "grad_norm": 0.22200003840805208, | |
| "learning_rate": 3.827212528642884e-05, | |
| "loss": 0.4276, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.9377537212449256, | |
| "grad_norm": 0.23150185026637812, | |
| "learning_rate": 3.7595534811746885e-05, | |
| "loss": 0.4218, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.9485791610284169, | |
| "grad_norm": 0.21285367773990543, | |
| "learning_rate": 3.6921988912237676e-05, | |
| "loss": 0.424, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.959404600811908, | |
| "grad_norm": 0.17015460687966813, | |
| "learning_rate": 3.6251595670996275e-05, | |
| "loss": 0.4224, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.9702300405953992, | |
| "grad_norm": 0.18773663319664743, | |
| "learning_rate": 3.5584462665214514e-05, | |
| "loss": 0.4186, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.9810554803788905, | |
| "grad_norm": 0.19181378149826905, | |
| "learning_rate": 3.492069694891813e-05, | |
| "loss": 0.427, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.9918809201623815, | |
| "grad_norm": 0.21051925665273966, | |
| "learning_rate": 3.4260405035788045e-05, | |
| "loss": 0.4274, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.002706359945873, | |
| "grad_norm": 0.3073354258820752, | |
| "learning_rate": 3.360369288206815e-05, | |
| "loss": 0.6681, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.013531799729364, | |
| "grad_norm": 0.25064046553309194, | |
| "learning_rate": 3.295066586956273e-05, | |
| "loss": 0.3897, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.024357239512855, | |
| "grad_norm": 0.24338261440618886, | |
| "learning_rate": 3.230142878872587e-05, | |
| "loss": 0.3883, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.0351826792963466, | |
| "grad_norm": 0.22284014657928528, | |
| "learning_rate": 3.1656085821845896e-05, | |
| "loss": 0.3893, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.0460081190798376, | |
| "grad_norm": 0.2194057958448553, | |
| "learning_rate": 3.101474052632748e-05, | |
| "loss": 0.3827, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.0568335588633286, | |
| "grad_norm": 0.23698144898082665, | |
| "learning_rate": 3.0377495818073712e-05, | |
| "loss": 0.3878, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.06765899864682, | |
| "grad_norm": 0.2543712215613022, | |
| "learning_rate": 2.974445395497144e-05, | |
| "loss": 0.3924, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.078484438430311, | |
| "grad_norm": 0.25600423720764554, | |
| "learning_rate": 2.9115716520482054e-05, | |
| "loss": 0.3967, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.089309878213802, | |
| "grad_norm": 0.2511300968768311, | |
| "learning_rate": 2.849138440734045e-05, | |
| "loss": 0.3885, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.1001353179972937, | |
| "grad_norm": 0.2253675731410667, | |
| "learning_rate": 2.787155780136501e-05, | |
| "loss": 0.3805, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.1109607577807847, | |
| "grad_norm": 0.23257374074012813, | |
| "learning_rate": 2.725633616538076e-05, | |
| "loss": 0.3901, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.121786197564276, | |
| "grad_norm": 0.20754444795906105, | |
| "learning_rate": 2.6645818223258827e-05, | |
| "loss": 0.3889, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.1326116373477673, | |
| "grad_norm": 0.23974692697842115, | |
| "learning_rate": 2.6040101944074254e-05, | |
| "loss": 0.3836, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.1434370771312583, | |
| "grad_norm": 0.1865329505582591, | |
| "learning_rate": 2.5439284526385022e-05, | |
| "loss": 0.3831, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.15426251691475, | |
| "grad_norm": 0.20901891546099988, | |
| "learning_rate": 2.48434623826348e-05, | |
| "loss": 0.3897, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.165087956698241, | |
| "grad_norm": 0.17039892814308769, | |
| "learning_rate": 2.4252731123681627e-05, | |
| "loss": 0.3921, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.175913396481732, | |
| "grad_norm": 0.18978843204311083, | |
| "learning_rate": 2.366718554345549e-05, | |
| "loss": 0.3869, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.1867388362652234, | |
| "grad_norm": 0.13803495552773962, | |
| "learning_rate": 2.3086919603746738e-05, | |
| "loss": 0.3842, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.1975642760487144, | |
| "grad_norm": 0.19506478963791687, | |
| "learning_rate": 2.25120264191282e-05, | |
| "loss": 0.3909, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.208389715832206, | |
| "grad_norm": 0.13474482771600654, | |
| "learning_rate": 2.194259824201323e-05, | |
| "loss": 0.3886, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.219215155615697, | |
| "grad_norm": 0.14807125927777573, | |
| "learning_rate": 2.1378726447852015e-05, | |
| "loss": 0.3867, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.230040595399188, | |
| "grad_norm": 0.15365440578968242, | |
| "learning_rate": 2.082050152046872e-05, | |
| "loss": 0.3881, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.2408660351826795, | |
| "grad_norm": 0.14558829971823375, | |
| "learning_rate": 2.0268013037541607e-05, | |
| "loss": 0.3914, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.2516914749661705, | |
| "grad_norm": 0.1371560184262318, | |
| "learning_rate": 1.9721349656228674e-05, | |
| "loss": 0.3823, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.2625169147496615, | |
| "grad_norm": 0.11747332446934883, | |
| "learning_rate": 1.918059909894089e-05, | |
| "loss": 0.387, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.273342354533153, | |
| "grad_norm": 0.13170177080479875, | |
| "learning_rate": 1.8645848139265496e-05, | |
| "loss": 0.3828, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.284167794316644, | |
| "grad_norm": 0.12950576152608145, | |
| "learning_rate": 1.8117182588041633e-05, | |
| "loss": 0.3843, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.2949932341001356, | |
| "grad_norm": 0.1262147300774054, | |
| "learning_rate": 1.759468727959026e-05, | |
| "loss": 0.3848, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.3058186738836266, | |
| "grad_norm": 0.13499342219785854, | |
| "learning_rate": 1.707844605810096e-05, | |
| "loss": 0.3922, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.3166441136671176, | |
| "grad_norm": 0.12019629775786407, | |
| "learning_rate": 1.6568541764177544e-05, | |
| "loss": 0.3868, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.3274695534506087, | |
| "grad_norm": 0.123057853560049, | |
| "learning_rate": 1.6065056221544746e-05, | |
| "loss": 0.3875, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.3382949932341, | |
| "grad_norm": 0.1276153162222321, | |
| "learning_rate": 1.556807022391805e-05, | |
| "loss": 0.3932, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.349120433017591, | |
| "grad_norm": 0.11893041854841553, | |
| "learning_rate": 1.5077663522038803e-05, | |
| "loss": 0.3866, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.3599458728010827, | |
| "grad_norm": 0.12645634242130177, | |
| "learning_rate": 1.4593914810876749e-05, | |
| "loss": 0.3941, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.3707713125845737, | |
| "grad_norm": 0.1325594756769902, | |
| "learning_rate": 1.4116901717001894e-05, | |
| "loss": 0.3915, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.381596752368065, | |
| "grad_norm": 0.11427081831052392, | |
| "learning_rate": 1.3646700786127952e-05, | |
| "loss": 0.3799, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.3924221921515563, | |
| "grad_norm": 0.1182894319259809, | |
| "learning_rate": 1.3183387470829026e-05, | |
| "loss": 0.3865, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.4032476319350473, | |
| "grad_norm": 0.13812990571742856, | |
| "learning_rate": 1.2727036118431905e-05, | |
| "loss": 0.3916, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.414073071718539, | |
| "grad_norm": 0.1240184216152659, | |
| "learning_rate": 1.2277719959085634e-05, | |
| "loss": 0.392, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.42489851150203, | |
| "grad_norm": 0.12146813464559397, | |
| "learning_rate": 1.1835511094010329e-05, | |
| "loss": 0.389, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.435723951285521, | |
| "grad_norm": 0.12612143081575952, | |
| "learning_rate": 1.1400480483927205e-05, | |
| "loss": 0.3869, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.4465493910690124, | |
| "grad_norm": 0.14119782771540126, | |
| "learning_rate": 1.0972697937671591e-05, | |
| "loss": 0.3898, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.4573748308525034, | |
| "grad_norm": 0.11466559847267338, | |
| "learning_rate": 1.0552232100990782e-05, | |
| "loss": 0.3833, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.4682002706359945, | |
| "grad_norm": 0.11857984272947328, | |
| "learning_rate": 1.0139150445528555e-05, | |
| "loss": 0.3853, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.479025710419486, | |
| "grad_norm": 0.11379457603595537, | |
| "learning_rate": 9.733519257998048e-06, | |
| "loss": 0.3821, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.489851150202977, | |
| "grad_norm": 0.12172192596025545, | |
| "learning_rate": 9.335403629544873e-06, | |
| "loss": 0.3833, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.500676589986468, | |
| "grad_norm": 0.11773376245986623, | |
| "learning_rate": 8.94486744530191e-06, | |
| "loss": 0.393, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.5115020297699595, | |
| "grad_norm": 0.10285756232686279, | |
| "learning_rate": 8.561973374137876e-06, | |
| "loss": 0.3897, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.5223274695534506, | |
| "grad_norm": 0.11531820935775133, | |
| "learning_rate": 8.186782858600797e-06, | |
| "loss": 0.3849, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.533152909336942, | |
| "grad_norm": 0.10364864354320348, | |
| "learning_rate": 7.819356105058482e-06, | |
| "loss": 0.3886, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.543978349120433, | |
| "grad_norm": 0.09908743652461002, | |
| "learning_rate": 7.459752074037232e-06, | |
| "loss": 0.3803, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.554803788903924, | |
| "grad_norm": 0.0955884820706306, | |
| "learning_rate": 7.1080284707605624e-06, | |
| "loss": 0.3804, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.565629228687415, | |
| "grad_norm": 0.09846787056135109, | |
| "learning_rate": 6.764241735889261e-06, | |
| "loss": 0.3866, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.5764546684709067, | |
| "grad_norm": 0.09799859422334903, | |
| "learning_rate": 6.428447036464477e-06, | |
| "loss": 0.3858, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.5872801082543977, | |
| "grad_norm": 0.09679034120618978, | |
| "learning_rate": 6.100698257055085e-06, | |
| "loss": 0.3834, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.598105548037889, | |
| "grad_norm": 0.10093400603955019, | |
| "learning_rate": 5.781047991110923e-06, | |
| "loss": 0.3896, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.60893098782138, | |
| "grad_norm": 0.10104638095796786, | |
| "learning_rate": 5.46954753252318e-06, | |
| "loss": 0.3924, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.6197564276048713, | |
| "grad_norm": 0.10038405717214048, | |
| "learning_rate": 5.166246867393346e-06, | |
| "loss": 0.3918, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.6305818673883627, | |
| "grad_norm": 0.09904219658805112, | |
| "learning_rate": 4.87119466601194e-06, | |
| "loss": 0.3835, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.641407307171854, | |
| "grad_norm": 0.11316612551154176, | |
| "learning_rate": 4.58443827504852e-06, | |
| "loss": 0.3821, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.6522327469553453, | |
| "grad_norm": 0.097473479720162, | |
| "learning_rate": 4.306023709953976e-06, | |
| "loss": 0.3833, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.6630581867388363, | |
| "grad_norm": 0.0952854432600168, | |
| "learning_rate": 4.035995647576508e-06, | |
| "loss": 0.3834, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.6738836265223274, | |
| "grad_norm": 0.09299918262798558, | |
| "learning_rate": 3.774397418992369e-06, | |
| "loss": 0.3889, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.6847090663058184, | |
| "grad_norm": 0.09888399612743574, | |
| "learning_rate": 3.521271002552602e-06, | |
| "loss": 0.3851, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.69553450608931, | |
| "grad_norm": 0.10298924594419286, | |
| "learning_rate": 3.2766570171467893e-06, | |
| "loss": 0.3806, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.706359945872801, | |
| "grad_norm": 0.10085730184506674, | |
| "learning_rate": 3.040594715685033e-06, | |
| "loss": 0.3918, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.7171853856562924, | |
| "grad_norm": 0.09724136823771373, | |
| "learning_rate": 2.8131219787990694e-06, | |
| "loss": 0.3876, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.7280108254397835, | |
| "grad_norm": 0.08743227467952154, | |
| "learning_rate": 2.5942753087635825e-06, | |
| "loss": 0.3913, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.7388362652232745, | |
| "grad_norm": 0.09212384514082983, | |
| "learning_rate": 2.3840898236387664e-06, | |
| "loss": 0.3903, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.749661705006766, | |
| "grad_norm": 0.08767500659026299, | |
| "learning_rate": 2.1825992516349626e-06, | |
| "loss": 0.3837, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.760487144790257, | |
| "grad_norm": 0.08408432260309959, | |
| "learning_rate": 1.9898359257003126e-06, | |
| "loss": 0.3797, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.7713125845737485, | |
| "grad_norm": 0.08602719858789641, | |
| "learning_rate": 1.8058307783323744e-06, | |
| "loss": 0.379, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.7821380243572396, | |
| "grad_norm": 0.08198855685273775, | |
| "learning_rate": 1.630613336614365e-06, | |
| "loss": 0.3913, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.7929634641407306, | |
| "grad_norm": 0.08559891153184518, | |
| "learning_rate": 1.464211717477035e-06, | |
| "loss": 0.3848, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.803788903924222, | |
| "grad_norm": 0.08076564462446145, | |
| "learning_rate": 1.30665262318675e-06, | |
| "loss": 0.3833, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.814614343707713, | |
| "grad_norm": 0.07997674076899126, | |
| "learning_rate": 1.1579613370606156e-06, | |
| "loss": 0.3893, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.825439783491204, | |
| "grad_norm": 0.08211459263332578, | |
| "learning_rate": 1.0181617194093145e-06, | |
| "loss": 0.3787, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.8362652232746957, | |
| "grad_norm": 0.08012043175321967, | |
| "learning_rate": 8.872762037082338e-07, | |
| "loss": 0.3841, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.8470906630581867, | |
| "grad_norm": 0.08182261519883058, | |
| "learning_rate": 7.653257929976282e-07, | |
| "loss": 0.3882, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.8579161028416777, | |
| "grad_norm": 0.08367217916154837, | |
| "learning_rate": 6.523300565122583e-07, | |
| "loss": 0.3838, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.8687415426251692, | |
| "grad_norm": 0.07915591706058425, | |
| "learning_rate": 5.483071265411561e-07, | |
| "loss": 0.3881, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.8795669824086603, | |
| "grad_norm": 0.0834396132172137, | |
| "learning_rate": 4.5327369551796154e-07, | |
| "loss": 0.3861, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.8903924221921518, | |
| "grad_norm": 0.08040680444066323, | |
| "learning_rate": 3.6724501334229996e-07, | |
| "loss": 0.3824, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.901217861975643, | |
| "grad_norm": 0.07960262290429217, | |
| "learning_rate": 2.9023488493264326e-07, | |
| "loss": 0.3831, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.912043301759134, | |
| "grad_norm": 0.0813772697299331, | |
| "learning_rate": 2.2225566801107246e-07, | |
| "loss": 0.3852, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.9228687415426253, | |
| "grad_norm": 0.0799088400195742, | |
| "learning_rate": 1.6331827112021966e-07, | |
| "loss": 0.3873, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.9336941813261164, | |
| "grad_norm": 0.0792761094231235, | |
| "learning_rate": 1.134321518728097e-07, | |
| "loss": 0.384, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.944519621109608, | |
| "grad_norm": 0.08041405015290681, | |
| "learning_rate": 7.260531543401518e-08, | |
| "loss": 0.39, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.955345060893099, | |
| "grad_norm": 0.08011333717359025, | |
| "learning_rate": 4.0844313236838236e-08, | |
| "loss": 0.387, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.96617050067659, | |
| "grad_norm": 0.08677763388970985, | |
| "learning_rate": 1.815424193085831e-08, | |
| "loss": 0.3861, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.976995940460081, | |
| "grad_norm": 0.0781257700674137, | |
| "learning_rate": 4.538742564345954e-09, | |
| "loss": 0.3934, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.9878213802435725, | |
| "grad_norm": 0.07617532903691134, | |
| "learning_rate": 0.0, | |
| "loss": 0.381, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.9878213802435725, | |
| "step": 276, | |
| "total_flos": 6.629121904995205e+18, | |
| "train_loss": 0.46416003235440323, | |
| "train_runtime": 27521.3535, | |
| "train_samples_per_second": 5.155, | |
| "train_steps_per_second": 0.01 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 276, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.629121904995205e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |