| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9975062344139651, | |
| "eval_steps": 500, | |
| "global_step": 300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0033250207813798837, | |
| "grad_norm": 1.1512730121612549, | |
| "learning_rate": 2e-05, | |
| "loss": 2.3647, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006650041562759767, | |
| "grad_norm": 1.1141036748886108, | |
| "learning_rate": 4e-05, | |
| "loss": 2.2794, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00997506234413965, | |
| "grad_norm": 0.9610893726348877, | |
| "learning_rate": 6e-05, | |
| "loss": 2.2596, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.013300083125519535, | |
| "grad_norm": 1.1339858770370483, | |
| "learning_rate": 8e-05, | |
| "loss": 2.3787, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01662510390689942, | |
| "grad_norm": 0.8878076672554016, | |
| "learning_rate": 0.0001, | |
| "loss": 2.3961, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0199501246882793, | |
| "grad_norm": 0.829910397529602, | |
| "learning_rate": 0.00012, | |
| "loss": 2.1948, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.023275145469659187, | |
| "grad_norm": 0.9420105814933777, | |
| "learning_rate": 0.00014, | |
| "loss": 2.4329, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.02660016625103907, | |
| "grad_norm": 0.8519226908683777, | |
| "learning_rate": 0.00016, | |
| "loss": 2.3078, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.029925187032418952, | |
| "grad_norm": 0.7587653994560242, | |
| "learning_rate": 0.00018, | |
| "loss": 1.9353, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03325020781379884, | |
| "grad_norm": 0.9927352666854858, | |
| "learning_rate": 0.0002, | |
| "loss": 1.9429, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03657522859517872, | |
| "grad_norm": 0.9643892049789429, | |
| "learning_rate": 0.00019999413227831132, | |
| "loss": 2.0925, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0399002493765586, | |
| "grad_norm": 0.941749632358551, | |
| "learning_rate": 0.00019997652980184843, | |
| "loss": 1.8099, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.043225270157938485, | |
| "grad_norm": 0.5177962779998779, | |
| "learning_rate": 0.00019994719463633997, | |
| "loss": 1.6693, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.046550290939318374, | |
| "grad_norm": 0.56168133020401, | |
| "learning_rate": 0.0001999061302243977, | |
| "loss": 1.9593, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.04987531172069826, | |
| "grad_norm": 0.5392152070999146, | |
| "learning_rate": 0.00019985334138511237, | |
| "loss": 1.6836, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.05320033250207814, | |
| "grad_norm": 0.5796711444854736, | |
| "learning_rate": 0.00019978883431348845, | |
| "loss": 1.7744, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.05652535328345802, | |
| "grad_norm": 0.5629785060882568, | |
| "learning_rate": 0.0001997126165797167, | |
| "loss": 2.0442, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.059850374064837904, | |
| "grad_norm": 0.48991289734840393, | |
| "learning_rate": 0.00019962469712828614, | |
| "loss": 1.679, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.06317539484621779, | |
| "grad_norm": 0.47867172956466675, | |
| "learning_rate": 0.0001995250862769342, | |
| "loss": 1.6641, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.06650041562759768, | |
| "grad_norm": 0.49752330780029297, | |
| "learning_rate": 0.00019941379571543596, | |
| "loss": 1.5331, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06982543640897755, | |
| "grad_norm": 0.49927300214767456, | |
| "learning_rate": 0.00019929083850423225, | |
| "loss": 1.5704, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.07315045719035744, | |
| "grad_norm": 0.5634847283363342, | |
| "learning_rate": 0.00019915622907289694, | |
| "loss": 1.9051, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.07647547797173733, | |
| "grad_norm": 0.5214512944221497, | |
| "learning_rate": 0.00019900998321844367, | |
| "loss": 1.756, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0798004987531172, | |
| "grad_norm": 0.46316221356391907, | |
| "learning_rate": 0.00019885211810347184, | |
| "loss": 1.6153, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0831255195344971, | |
| "grad_norm": 0.45869576930999756, | |
| "learning_rate": 0.00019868265225415265, | |
| "loss": 1.8899, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.08645054031587697, | |
| "grad_norm": 0.4824669063091278, | |
| "learning_rate": 0.00019850160555805486, | |
| "loss": 1.8861, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.08977556109725686, | |
| "grad_norm": 0.509224534034729, | |
| "learning_rate": 0.000198308999261811, | |
| "loss": 1.8507, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.09310058187863675, | |
| "grad_norm": 0.4441746771335602, | |
| "learning_rate": 0.00019810485596862392, | |
| "loss": 1.7326, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.09642560266001662, | |
| "grad_norm": 0.4595758318901062, | |
| "learning_rate": 0.00019788919963561422, | |
| "loss": 1.8283, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.09975062344139651, | |
| "grad_norm": 0.5222824215888977, | |
| "learning_rate": 0.00019766205557100868, | |
| "loss": 1.5678, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.10307564422277639, | |
| "grad_norm": 0.43890196084976196, | |
| "learning_rate": 0.00019742345043117045, | |
| "loss": 1.5899, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.10640066500415628, | |
| "grad_norm": 0.4542831778526306, | |
| "learning_rate": 0.00019717341221747056, | |
| "loss": 1.6733, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.10972568578553615, | |
| "grad_norm": 0.43134549260139465, | |
| "learning_rate": 0.00019691197027300205, | |
| "loss": 1.7386, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.11305070656691604, | |
| "grad_norm": 0.44071701169013977, | |
| "learning_rate": 0.00019663915527913625, | |
| "loss": 1.7685, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.11637572734829593, | |
| "grad_norm": 0.4880881607532501, | |
| "learning_rate": 0.0001963549992519223, | |
| "loss": 1.8461, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.11970074812967581, | |
| "grad_norm": 0.40884578227996826, | |
| "learning_rate": 0.00019605953553832988, | |
| "loss": 1.5538, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.1230257689110557, | |
| "grad_norm": 0.39413318037986755, | |
| "learning_rate": 0.00019575279881233577, | |
| "loss": 1.4222, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.12635078969243557, | |
| "grad_norm": 0.44478997588157654, | |
| "learning_rate": 0.00019543482507085482, | |
| "loss": 1.7247, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.12967581047381546, | |
| "grad_norm": 0.4295913875102997, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 1.5788, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.13300083125519535, | |
| "grad_norm": 0.47360050678253174, | |
| "learning_rate": 0.00019476531711828027, | |
| "loss": 1.7429, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.13632585203657524, | |
| "grad_norm": 0.483909547328949, | |
| "learning_rate": 0.00019441386147691335, | |
| "loss": 1.6674, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.1396508728179551, | |
| "grad_norm": 0.47071558237075806, | |
| "learning_rate": 0.0001940513259502924, | |
| "loss": 1.8229, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.142975893599335, | |
| "grad_norm": 0.43929168581962585, | |
| "learning_rate": 0.0001936777530835689, | |
| "loss": 1.6562, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.14630091438071488, | |
| "grad_norm": 0.4329998791217804, | |
| "learning_rate": 0.0001932931867171751, | |
| "loss": 1.5274, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.14962593516209477, | |
| "grad_norm": 0.44375908374786377, | |
| "learning_rate": 0.00019289767198167916, | |
| "loss": 1.7084, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.15295095594347466, | |
| "grad_norm": 0.48119276762008667, | |
| "learning_rate": 0.0001924912552924889, | |
| "loss": 1.7645, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.15627597672485452, | |
| "grad_norm": 0.4040566384792328, | |
| "learning_rate": 0.00019207398434440478, | |
| "loss": 1.5925, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1596009975062344, | |
| "grad_norm": 0.4708506464958191, | |
| "learning_rate": 0.00019164590810602262, | |
| "loss": 1.8461, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.1629260182876143, | |
| "grad_norm": 0.431772381067276, | |
| "learning_rate": 0.000191207076813987, | |
| "loss": 1.5356, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.1662510390689942, | |
| "grad_norm": 0.4952054023742676, | |
| "learning_rate": 0.00019075754196709572, | |
| "loss": 1.8034, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.16957605985037408, | |
| "grad_norm": 0.43522897362709045, | |
| "learning_rate": 0.00019029735632025618, | |
| "loss": 1.6717, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.17290108063175394, | |
| "grad_norm": 0.46861544251441956, | |
| "learning_rate": 0.00018982657387829445, | |
| "loss": 1.766, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.17622610141313383, | |
| "grad_norm": 0.44363775849342346, | |
| "learning_rate": 0.00018934524988961738, | |
| "loss": 1.5169, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.17955112219451372, | |
| "grad_norm": 0.41366782784461975, | |
| "learning_rate": 0.00018885344083972914, | |
| "loss": 1.6495, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.1828761429758936, | |
| "grad_norm": 0.4273390769958496, | |
| "learning_rate": 0.0001883512044446023, | |
| "loss": 1.5952, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1862011637572735, | |
| "grad_norm": 0.4389772117137909, | |
| "learning_rate": 0.00018783859964390464, | |
| "loss": 1.7003, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.18952618453865336, | |
| "grad_norm": 0.480125367641449, | |
| "learning_rate": 0.0001873156865940823, | |
| "loss": 1.6503, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.19285120532003325, | |
| "grad_norm": 0.48973348736763, | |
| "learning_rate": 0.00018678252666130013, | |
| "loss": 1.737, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.19617622610141314, | |
| "grad_norm": 0.4558335840702057, | |
| "learning_rate": 0.0001862391824142402, | |
| "loss": 1.571, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.19950124688279303, | |
| "grad_norm": 0.45777326822280884, | |
| "learning_rate": 0.00018568571761675893, | |
| "loss": 1.6462, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2028262676641729, | |
| "grad_norm": 0.4185212254524231, | |
| "learning_rate": 0.00018512219722040425, | |
| "loss": 1.5729, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.20615128844555278, | |
| "grad_norm": 0.4137243330478668, | |
| "learning_rate": 0.0001845486873567932, | |
| "loss": 1.675, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.20947630922693267, | |
| "grad_norm": 0.42468297481536865, | |
| "learning_rate": 0.00018396525532985108, | |
| "loss": 1.4519, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.21280133000831256, | |
| "grad_norm": 0.46751776337623596, | |
| "learning_rate": 0.00018337196960791302, | |
| "loss": 1.7264, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.21612635078969245, | |
| "grad_norm": 0.47722429037094116, | |
| "learning_rate": 0.00018276889981568906, | |
| "loss": 1.5392, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.2194513715710723, | |
| "grad_norm": 0.4753107726573944, | |
| "learning_rate": 0.00018215611672609317, | |
| "loss": 1.5328, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.2227763923524522, | |
| "grad_norm": 0.4401816129684448, | |
| "learning_rate": 0.00018153369225193782, | |
| "loss": 1.4793, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.22610141313383209, | |
| "grad_norm": 0.4473712146282196, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 1.5596, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.22942643391521197, | |
| "grad_norm": 0.45505204796791077, | |
| "learning_rate": 0.00018026021244992287, | |
| "loss": 1.7437, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.23275145469659186, | |
| "grad_norm": 0.44190192222595215, | |
| "learning_rate": 0.00017960930657056438, | |
| "loss": 1.7401, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.23607647547797173, | |
| "grad_norm": 0.501592218875885, | |
| "learning_rate": 0.0001789490581861102, | |
| "loss": 1.7464, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.23940149625935161, | |
| "grad_norm": 0.43836328387260437, | |
| "learning_rate": 0.00017827954477963557, | |
| "loss": 1.7451, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.2427265170407315, | |
| "grad_norm": 0.611949622631073, | |
| "learning_rate": 0.0001776008449215073, | |
| "loss": 1.6921, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.2460515378221114, | |
| "grad_norm": 0.46015432476997375, | |
| "learning_rate": 0.0001769130382601629, | |
| "loss": 1.7985, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.24937655860349128, | |
| "grad_norm": 0.44316309690475464, | |
| "learning_rate": 0.00017621620551276366, | |
| "loss": 1.7806, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.25270157938487114, | |
| "grad_norm": 0.4749353229999542, | |
| "learning_rate": 0.00017551042845572208, | |
| "loss": 1.7349, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.25602660016625106, | |
| "grad_norm": 0.4712280333042145, | |
| "learning_rate": 0.00017479578991510506, | |
| "loss": 1.4129, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.2593516209476309, | |
| "grad_norm": 0.44466859102249146, | |
| "learning_rate": 0.00017407237375691392, | |
| "loss": 1.6819, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.2626766417290108, | |
| "grad_norm": 0.42531418800354004, | |
| "learning_rate": 0.00017334026487724225, | |
| "loss": 1.6154, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.2660016625103907, | |
| "grad_norm": 0.4512370228767395, | |
| "learning_rate": 0.0001725995491923131, | |
| "loss": 1.6736, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.26932668329177056, | |
| "grad_norm": 0.4131537079811096, | |
| "learning_rate": 0.00017185031362839626, | |
| "loss": 1.5468, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.2726517040731505, | |
| "grad_norm": 0.47616103291511536, | |
| "learning_rate": 0.00017109264611160708, | |
| "loss": 1.523, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.27597672485453034, | |
| "grad_norm": 0.4459686279296875, | |
| "learning_rate": 0.000170326635557588, | |
| "loss": 1.8612, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.2793017456359102, | |
| "grad_norm": 0.4500899612903595, | |
| "learning_rate": 0.00016955237186107387, | |
| "loss": 1.643, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.2826267664172901, | |
| "grad_norm": 0.44385287165641785, | |
| "learning_rate": 0.00016876994588534234, | |
| "loss": 1.3833, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.28595178719867, | |
| "grad_norm": 0.4063577950000763, | |
| "learning_rate": 0.0001679794494515508, | |
| "loss": 1.3494, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.2892768079800499, | |
| "grad_norm": 0.43013447523117065, | |
| "learning_rate": 0.00016718097532796063, | |
| "loss": 1.5205, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.29260182876142976, | |
| "grad_norm": 0.46770158410072327, | |
| "learning_rate": 0.00016637461721905045, | |
| "loss": 1.6897, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.2959268495428096, | |
| "grad_norm": 0.4841039478778839, | |
| "learning_rate": 0.00016556046975451963, | |
| "loss": 1.5793, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.29925187032418954, | |
| "grad_norm": 0.48426705598831177, | |
| "learning_rate": 0.00016473862847818277, | |
| "loss": 1.6988, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3025768911055694, | |
| "grad_norm": 0.5768110752105713, | |
| "learning_rate": 0.0001639091898367576, | |
| "loss": 1.7846, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.3059019118869493, | |
| "grad_norm": 0.446196049451828, | |
| "learning_rate": 0.00016307225116854622, | |
| "loss": 1.7882, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.3092269326683292, | |
| "grad_norm": 0.4034564793109894, | |
| "learning_rate": 0.00016222791069201207, | |
| "loss": 1.6616, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.31255195344970904, | |
| "grad_norm": 0.424376517534256, | |
| "learning_rate": 0.00016137626749425377, | |
| "loss": 1.5353, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.31587697423108896, | |
| "grad_norm": 0.45510077476501465, | |
| "learning_rate": 0.00016051742151937655, | |
| "loss": 1.7947, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.3192019950124688, | |
| "grad_norm": 0.4815070331096649, | |
| "learning_rate": 0.00015965147355676343, | |
| "loss": 1.581, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.32252701579384874, | |
| "grad_norm": 0.4505084156990051, | |
| "learning_rate": 0.00015877852522924732, | |
| "loss": 1.6186, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3258520365752286, | |
| "grad_norm": 0.4437141418457031, | |
| "learning_rate": 0.0001578986789811849, | |
| "loss": 1.6509, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.32917705735660846, | |
| "grad_norm": 0.4133874475955963, | |
| "learning_rate": 0.00015701203806643433, | |
| "loss": 1.7992, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.3325020781379884, | |
| "grad_norm": 0.4500593841075897, | |
| "learning_rate": 0.00015611870653623825, | |
| "loss": 1.6654, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.33582709891936824, | |
| "grad_norm": 0.4359726309776306, | |
| "learning_rate": 0.00015521878922701246, | |
| "loss": 1.6461, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.33915211970074816, | |
| "grad_norm": 0.40108025074005127, | |
| "learning_rate": 0.00015431239174804328, | |
| "loss": 1.5237, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.342477140482128, | |
| "grad_norm": 0.43869125843048096, | |
| "learning_rate": 0.00015339962046909364, | |
| "loss": 1.6909, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.3458021612635079, | |
| "grad_norm": 0.42006051540374756, | |
| "learning_rate": 0.00015248058250792008, | |
| "loss": 1.5046, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.3491271820448878, | |
| "grad_norm": 0.38756394386291504, | |
| "learning_rate": 0.00015155538571770218, | |
| "loss": 1.3747, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.35245220282626766, | |
| "grad_norm": 0.47784286737442017, | |
| "learning_rate": 0.0001506241386743854, | |
| "loss": 1.673, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.3557772236076476, | |
| "grad_norm": 0.4587322175502777, | |
| "learning_rate": 0.00014968695066393923, | |
| "loss": 1.7987, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.35910224438902744, | |
| "grad_norm": 0.42091092467308044, | |
| "learning_rate": 0.00014874393166953192, | |
| "loss": 1.5309, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.3624272651704073, | |
| "grad_norm": 0.47224530577659607, | |
| "learning_rate": 0.00014779519235862365, | |
| "loss": 1.7268, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.3657522859517872, | |
| "grad_norm": 0.44596192240715027, | |
| "learning_rate": 0.00014684084406997903, | |
| "loss": 1.7108, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3690773067331671, | |
| "grad_norm": 0.4590005874633789, | |
| "learning_rate": 0.0001458809988006011, | |
| "loss": 1.638, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.372402327514547, | |
| "grad_norm": 0.43627721071243286, | |
| "learning_rate": 0.00014491576919258792, | |
| "loss": 1.6721, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.37572734829592686, | |
| "grad_norm": 0.41456034779548645, | |
| "learning_rate": 0.00014394526851991364, | |
| "loss": 1.6863, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.3790523690773067, | |
| "grad_norm": 0.4247894883155823, | |
| "learning_rate": 0.0001429696106751352, | |
| "loss": 1.5659, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.38237738985868663, | |
| "grad_norm": 0.4657272696495056, | |
| "learning_rate": 0.00014198891015602646, | |
| "loss": 1.4086, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.3857024106400665, | |
| "grad_norm": 0.4860394597053528, | |
| "learning_rate": 0.0001410032820521416, | |
| "loss": 1.4603, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.38902743142144636, | |
| "grad_norm": 0.41849544644355774, | |
| "learning_rate": 0.00014001284203130868, | |
| "loss": 1.3991, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3923524522028263, | |
| "grad_norm": 0.4544629752635956, | |
| "learning_rate": 0.00013901770632605547, | |
| "loss": 1.8028, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.39567747298420614, | |
| "grad_norm": 0.5051787495613098, | |
| "learning_rate": 0.0001380179917199692, | |
| "loss": 1.8854, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.39900249376558605, | |
| "grad_norm": 0.41150030493736267, | |
| "learning_rate": 0.00013701381553399145, | |
| "loss": 1.6686, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4023275145469659, | |
| "grad_norm": 0.4593510925769806, | |
| "learning_rate": 0.0001360052956126499, | |
| "loss": 1.5844, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4056525353283458, | |
| "grad_norm": 0.42087090015411377, | |
| "learning_rate": 0.00013499255031022885, | |
| "loss": 1.4865, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4089775561097257, | |
| "grad_norm": 0.4708739221096039, | |
| "learning_rate": 0.00013397569847687984, | |
| "loss": 1.7089, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.41230257689110555, | |
| "grad_norm": 0.4878352880477905, | |
| "learning_rate": 0.00013295485944467405, | |
| "loss": 1.8006, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.41562759767248547, | |
| "grad_norm": 0.43254002928733826, | |
| "learning_rate": 0.000131930153013598, | |
| "loss": 1.6949, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.41895261845386533, | |
| "grad_norm": 0.47519850730895996, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 1.7601, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.4222776392352452, | |
| "grad_norm": 0.4135800898075104, | |
| "learning_rate": 0.00012986961940995138, | |
| "loss": 1.5955, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.4256026600166251, | |
| "grad_norm": 0.46267929673194885, | |
| "learning_rate": 0.0001288340340501351, | |
| "loss": 1.8398, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.428927680798005, | |
| "grad_norm": 0.43891721963882446, | |
| "learning_rate": 0.00012779506488857945, | |
| "loss": 1.4741, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.4322527015793849, | |
| "grad_norm": 0.4456429183483124, | |
| "learning_rate": 0.00012675283385292212, | |
| "loss": 1.7454, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.43557772236076475, | |
| "grad_norm": 0.4604743719100952, | |
| "learning_rate": 0.00012570746325359607, | |
| "loss": 1.8192, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.4389027431421446, | |
| "grad_norm": 0.46728062629699707, | |
| "learning_rate": 0.00012465907576947622, | |
| "loss": 1.7551, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.44222776392352453, | |
| "grad_norm": 0.436298668384552, | |
| "learning_rate": 0.000123607794433482, | |
| "loss": 1.6592, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.4455527847049044, | |
| "grad_norm": 0.39828214049339294, | |
| "learning_rate": 0.00012255374261813944, | |
| "loss": 1.4603, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.4488778054862843, | |
| "grad_norm": 0.4469813406467438, | |
| "learning_rate": 0.00012149704402110243, | |
| "loss": 1.6449, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.45220282626766417, | |
| "grad_norm": 0.4820503294467926, | |
| "learning_rate": 0.0001204378226506365, | |
| "loss": 1.8473, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.45552784704904403, | |
| "grad_norm": 0.49072131514549255, | |
| "learning_rate": 0.00011937620281106585, | |
| "loss": 1.6843, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.45885286783042395, | |
| "grad_norm": 0.48773854970932007, | |
| "learning_rate": 0.00011831230908818563, | |
| "loss": 1.625, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.4621778886118038, | |
| "grad_norm": 0.4438723623752594, | |
| "learning_rate": 0.00011724626633464127, | |
| "loss": 1.7558, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.46550290939318373, | |
| "grad_norm": 0.4389275014400482, | |
| "learning_rate": 0.0001161781996552765, | |
| "loss": 1.4621, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.4688279301745636, | |
| "grad_norm": 0.4611305296421051, | |
| "learning_rate": 0.00011510823439245169, | |
| "loss": 1.59, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.47215295095594345, | |
| "grad_norm": 0.43601059913635254, | |
| "learning_rate": 0.00011403649611133444, | |
| "loss": 1.7462, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.47547797173732337, | |
| "grad_norm": 0.41201236844062805, | |
| "learning_rate": 0.00011296311058516389, | |
| "loss": 1.5341, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.47880299251870323, | |
| "grad_norm": 0.46523982286453247, | |
| "learning_rate": 0.00011188820378049065, | |
| "loss": 1.6327, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.48212801330008315, | |
| "grad_norm": 0.42490893602371216, | |
| "learning_rate": 0.00011081190184239419, | |
| "loss": 1.6178, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.485453034081463, | |
| "grad_norm": 0.42238375544548035, | |
| "learning_rate": 0.00010973433107967902, | |
| "loss": 1.534, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.48877805486284287, | |
| "grad_norm": 0.48569226264953613, | |
| "learning_rate": 0.00010865561795005177, | |
| "loss": 1.5332, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.4921030756442228, | |
| "grad_norm": 0.4933275878429413, | |
| "learning_rate": 0.00010757588904528106, | |
| "loss": 1.5928, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.49542809642560265, | |
| "grad_norm": 0.4781058728694916, | |
| "learning_rate": 0.00010649527107634108, | |
| "loss": 1.6578, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.49875311720698257, | |
| "grad_norm": 0.4651820659637451, | |
| "learning_rate": 0.00010541389085854176, | |
| "loss": 1.6884, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5020781379883624, | |
| "grad_norm": 0.4429711103439331, | |
| "learning_rate": 0.00010433187529664623, | |
| "loss": 1.6723, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.5054031587697423, | |
| "grad_norm": 0.4521614611148834, | |
| "learning_rate": 0.00010324935136997806, | |
| "loss": 1.6269, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.5087281795511222, | |
| "grad_norm": 0.4930736720561981, | |
| "learning_rate": 0.00010216644611751975, | |
| "loss": 1.7933, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.5120532003325021, | |
| "grad_norm": 0.4855606257915497, | |
| "learning_rate": 0.000101083286623004, | |
| "loss": 1.6702, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.515378221113882, | |
| "grad_norm": 0.4960128366947174, | |
| "learning_rate": 0.0001, | |
| "loss": 1.7428, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5187032418952618, | |
| "grad_norm": 0.42107459902763367, | |
| "learning_rate": 9.891671337699602e-05, | |
| "loss": 1.6235, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.5220282626766417, | |
| "grad_norm": 0.4479861855506897, | |
| "learning_rate": 9.783355388248027e-05, | |
| "loss": 1.5158, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.5253532834580216, | |
| "grad_norm": 0.4954458177089691, | |
| "learning_rate": 9.675064863002196e-05, | |
| "loss": 1.6743, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.5286783042394015, | |
| "grad_norm": 0.5591014623641968, | |
| "learning_rate": 9.56681247033538e-05, | |
| "loss": 1.9691, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.5320033250207814, | |
| "grad_norm": 0.46626871824264526, | |
| "learning_rate": 9.458610914145826e-05, | |
| "loss": 1.5621, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5353283458021613, | |
| "grad_norm": 0.4377134144306183, | |
| "learning_rate": 9.350472892365892e-05, | |
| "loss": 1.5524, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.5386533665835411, | |
| "grad_norm": 0.3984418511390686, | |
| "learning_rate": 9.242411095471897e-05, | |
| "loss": 1.6454, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.541978387364921, | |
| "grad_norm": 0.42802637815475464, | |
| "learning_rate": 9.134438204994824e-05, | |
| "loss": 1.4036, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.545303408146301, | |
| "grad_norm": 0.4567003846168518, | |
| "learning_rate": 9.026566892032105e-05, | |
| "loss": 1.6606, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.5486284289276808, | |
| "grad_norm": 0.45452797412872314, | |
| "learning_rate": 8.918809815760585e-05, | |
| "loss": 1.8219, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5519534497090607, | |
| "grad_norm": 0.4367886781692505, | |
| "learning_rate": 8.811179621950936e-05, | |
| "loss": 1.5962, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.5552784704904405, | |
| "grad_norm": 0.4670146703720093, | |
| "learning_rate": 8.703688941483616e-05, | |
| "loss": 1.6382, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.5586034912718204, | |
| "grad_norm": 0.5069778561592102, | |
| "learning_rate": 8.596350388866558e-05, | |
| "loss": 1.7067, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.5619285120532004, | |
| "grad_norm": 0.4080033302307129, | |
| "learning_rate": 8.489176560754834e-05, | |
| "loss": 1.4192, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.5652535328345802, | |
| "grad_norm": 0.491526335477829, | |
| "learning_rate": 8.382180034472353e-05, | |
| "loss": 1.8687, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5685785536159601, | |
| "grad_norm": 0.5429246425628662, | |
| "learning_rate": 8.275373366535877e-05, | |
| "loss": 1.776, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.57190357439734, | |
| "grad_norm": 0.4131667912006378, | |
| "learning_rate": 8.168769091181438e-05, | |
| "loss": 1.3345, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.5752285951787198, | |
| "grad_norm": 0.5055519342422485, | |
| "learning_rate": 8.062379718893417e-05, | |
| "loss": 1.7716, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.5785536159600998, | |
| "grad_norm": 0.4675292670726776, | |
| "learning_rate": 7.956217734936353e-05, | |
| "loss": 1.5941, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.5818786367414797, | |
| "grad_norm": 0.5096448659896851, | |
| "learning_rate": 7.85029559788976e-05, | |
| "loss": 1.9376, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.5852036575228595, | |
| "grad_norm": 0.4687637686729431, | |
| "learning_rate": 7.744625738186059e-05, | |
| "loss": 1.7242, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.5885286783042394, | |
| "grad_norm": 0.437148779630661, | |
| "learning_rate": 7.639220556651799e-05, | |
| "loss": 1.4993, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.5918536990856192, | |
| "grad_norm": 0.44125625491142273, | |
| "learning_rate": 7.534092423052381e-05, | |
| "loss": 1.5076, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.5951787198669992, | |
| "grad_norm": 0.4794883131980896, | |
| "learning_rate": 7.42925367464039e-05, | |
| "loss": 1.6401, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.5985037406483791, | |
| "grad_norm": 0.42347967624664307, | |
| "learning_rate": 7.324716614707793e-05, | |
| "loss": 1.444, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6018287614297589, | |
| "grad_norm": 0.4843563437461853, | |
| "learning_rate": 7.220493511142059e-05, | |
| "loss": 1.7117, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.6051537822111388, | |
| "grad_norm": 0.48885542154312134, | |
| "learning_rate": 7.116596594986494e-05, | |
| "loss": 1.6799, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.6084788029925187, | |
| "grad_norm": 0.48835498094558716, | |
| "learning_rate": 7.013038059004866e-05, | |
| "loss": 1.7308, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.6118038237738986, | |
| "grad_norm": 0.38506001234054565, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 1.5766, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.6151288445552785, | |
| "grad_norm": 0.5520392656326294, | |
| "learning_rate": 6.806984698640202e-05, | |
| "loss": 1.5418, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6184538653366584, | |
| "grad_norm": 0.4401935935020447, | |
| "learning_rate": 6.704514055532597e-05, | |
| "loss": 1.7715, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.6217788861180382, | |
| "grad_norm": 0.4164566695690155, | |
| "learning_rate": 6.602430152312017e-05, | |
| "loss": 1.4711, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.6251039068994181, | |
| "grad_norm": 0.4750818610191345, | |
| "learning_rate": 6.500744968977116e-05, | |
| "loss": 1.374, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.628428927680798, | |
| "grad_norm": 0.5478043556213379, | |
| "learning_rate": 6.399470438735014e-05, | |
| "loss": 1.7294, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.6317539484621779, | |
| "grad_norm": 0.4560893476009369, | |
| "learning_rate": 6.298618446600856e-05, | |
| "loss": 1.8216, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6350789692435578, | |
| "grad_norm": 0.49942511320114136, | |
| "learning_rate": 6.19820082800308e-05, | |
| "loss": 1.6108, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.6384039900249376, | |
| "grad_norm": 0.3901759088039398, | |
| "learning_rate": 6.0982293673944544e-05, | |
| "loss": 1.4635, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.6417290108063175, | |
| "grad_norm": 0.45033466815948486, | |
| "learning_rate": 5.9987157968691344e-05, | |
| "loss": 1.5153, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.6450540315876975, | |
| "grad_norm": 0.44514134526252747, | |
| "learning_rate": 5.899671794785839e-05, | |
| "loss": 1.6015, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.6483790523690773, | |
| "grad_norm": 0.42773956060409546, | |
| "learning_rate": 5.801108984397354e-05, | |
| "loss": 1.6624, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.6517040731504572, | |
| "grad_norm": 0.42323529720306396, | |
| "learning_rate": 5.703038932486484e-05, | |
| "loss": 1.642, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.6550290939318371, | |
| "grad_norm": 0.4852340519428253, | |
| "learning_rate": 5.605473148008638e-05, | |
| "loss": 1.5533, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.6583541147132169, | |
| "grad_norm": 0.46353092789649963, | |
| "learning_rate": 5.5084230807412126e-05, | |
| "loss": 1.5137, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.6616791354945969, | |
| "grad_norm": 0.5486162304878235, | |
| "learning_rate": 5.411900119939895e-05, | |
| "loss": 1.5682, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.6650041562759768, | |
| "grad_norm": 0.4136289656162262, | |
| "learning_rate": 5.3159155930021e-05, | |
| "loss": 1.5902, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6683291770573566, | |
| "grad_norm": 0.457292765378952, | |
| "learning_rate": 5.2204807641376354e-05, | |
| "loss": 1.6669, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.6716541978387365, | |
| "grad_norm": 0.4368407726287842, | |
| "learning_rate": 5.12560683304681e-05, | |
| "loss": 1.7747, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.6749792186201163, | |
| "grad_norm": 0.4596605598926544, | |
| "learning_rate": 5.03130493360608e-05, | |
| "loss": 1.5868, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.6783042394014963, | |
| "grad_norm": 0.437491238117218, | |
| "learning_rate": 4.9375861325614606e-05, | |
| "loss": 1.7614, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.6816292601828762, | |
| "grad_norm": 0.47249388694763184, | |
| "learning_rate": 4.844461428229782e-05, | |
| "loss": 1.582, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.684954280964256, | |
| "grad_norm": 0.44100067019462585, | |
| "learning_rate": 4.751941749207995e-05, | |
| "loss": 1.6814, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.6882793017456359, | |
| "grad_norm": 0.5000886917114258, | |
| "learning_rate": 4.660037953090639e-05, | |
| "loss": 1.6634, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.6916043225270158, | |
| "grad_norm": 0.4667086899280548, | |
| "learning_rate": 4.5687608251956714e-05, | |
| "loss": 1.7767, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.6949293433083957, | |
| "grad_norm": 0.4677750766277313, | |
| "learning_rate": 4.4781210772987514e-05, | |
| "loss": 1.785, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.6982543640897756, | |
| "grad_norm": 0.40729814767837524, | |
| "learning_rate": 4.388129346376178e-05, | |
| "loss": 1.5742, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7015793848711555, | |
| "grad_norm": 0.4622965157032013, | |
| "learning_rate": 4.298796193356566e-05, | |
| "loss": 1.755, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.7049044056525353, | |
| "grad_norm": 0.42128920555114746, | |
| "learning_rate": 4.210132101881516e-05, | |
| "loss": 1.359, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.7082294264339152, | |
| "grad_norm": 0.4670293927192688, | |
| "learning_rate": 4.12214747707527e-05, | |
| "loss": 1.8743, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.7115544472152951, | |
| "grad_norm": 0.474398136138916, | |
| "learning_rate": 4.034852644323661e-05, | |
| "loss": 1.6977, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.714879467996675, | |
| "grad_norm": 0.5026089549064636, | |
| "learning_rate": 3.948257848062351e-05, | |
| "loss": 1.566, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7182044887780549, | |
| "grad_norm": 0.40603697299957275, | |
| "learning_rate": 3.862373250574626e-05, | |
| "loss": 1.3894, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.7215295095594347, | |
| "grad_norm": 0.4771779179573059, | |
| "learning_rate": 3.7772089307987936e-05, | |
| "loss": 1.6296, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.7248545303408146, | |
| "grad_norm": 0.44347891211509705, | |
| "learning_rate": 3.6927748831453836e-05, | |
| "loss": 1.6663, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.7281795511221946, | |
| "grad_norm": 0.435149610042572, | |
| "learning_rate": 3.609081016324243e-05, | |
| "loss": 1.6662, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.7315045719035744, | |
| "grad_norm": 0.453782856464386, | |
| "learning_rate": 3.5261371521817244e-05, | |
| "loss": 1.7286, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7348295926849543, | |
| "grad_norm": 0.42496827244758606, | |
| "learning_rate": 3.44395302454804e-05, | |
| "loss": 1.7376, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.7381546134663342, | |
| "grad_norm": 0.45447835326194763, | |
| "learning_rate": 3.3625382780949574e-05, | |
| "loss": 1.5055, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.741479634247714, | |
| "grad_norm": 0.5035948157310486, | |
| "learning_rate": 3.28190246720394e-05, | |
| "loss": 1.866, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.744804655029094, | |
| "grad_norm": 0.47680604457855225, | |
| "learning_rate": 3.202055054844921e-05, | |
| "loss": 1.9692, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.7481296758104738, | |
| "grad_norm": 0.45373663306236267, | |
| "learning_rate": 3.123005411465766e-05, | |
| "loss": 1.6879, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.7514546965918537, | |
| "grad_norm": 0.49925628304481506, | |
| "learning_rate": 3.0447628138926156e-05, | |
| "loss": 1.5313, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.7547797173732336, | |
| "grad_norm": 0.4820810556411743, | |
| "learning_rate": 2.9673364442412e-05, | |
| "loss": 1.6259, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.7581047381546134, | |
| "grad_norm": 0.5111257433891296, | |
| "learning_rate": 2.890735388839295e-05, | |
| "loss": 1.6068, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.7614297589359933, | |
| "grad_norm": 0.3893967568874359, | |
| "learning_rate": 2.8149686371603767e-05, | |
| "loss": 1.5461, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.7647547797173733, | |
| "grad_norm": 0.42585450410842896, | |
| "learning_rate": 2.7400450807686938e-05, | |
| "loss": 1.4092, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7680798004987531, | |
| "grad_norm": 0.5068459510803223, | |
| "learning_rate": 2.665973512275778e-05, | |
| "loss": 1.8426, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.771404821280133, | |
| "grad_norm": 0.44372087717056274, | |
| "learning_rate": 2.59276262430861e-05, | |
| "loss": 1.5669, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.7747298420615129, | |
| "grad_norm": 0.4483433663845062, | |
| "learning_rate": 2.520421008489494e-05, | |
| "loss": 1.508, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.7780548628428927, | |
| "grad_norm": 0.4225240647792816, | |
| "learning_rate": 2.4489571544277945e-05, | |
| "loss": 1.4963, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.7813798836242727, | |
| "grad_norm": 0.4540765583515167, | |
| "learning_rate": 2.3783794487236365e-05, | |
| "loss": 1.7699, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.7847049044056525, | |
| "grad_norm": 0.5303469896316528, | |
| "learning_rate": 2.308696173983711e-05, | |
| "loss": 1.7887, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.7880299251870324, | |
| "grad_norm": 0.4368319809436798, | |
| "learning_rate": 2.2399155078492694e-05, | |
| "loss": 1.6762, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.7913549459684123, | |
| "grad_norm": 0.41934987902641296, | |
| "learning_rate": 2.1720455220364444e-05, | |
| "loss": 1.6372, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.7946799667497921, | |
| "grad_norm": 0.4291558861732483, | |
| "learning_rate": 2.1050941813889836e-05, | |
| "loss": 1.6668, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.7980049875311721, | |
| "grad_norm": 0.414044052362442, | |
| "learning_rate": 2.0390693429435627e-05, | |
| "loss": 1.6885, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.801330008312552, | |
| "grad_norm": 0.4342755377292633, | |
| "learning_rate": 1.9739787550077116e-05, | |
| "loss": 1.5082, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.8046550290939318, | |
| "grad_norm": 0.45343807339668274, | |
| "learning_rate": 1.9098300562505266e-05, | |
| "loss": 1.5635, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.8079800498753117, | |
| "grad_norm": 0.4498422145843506, | |
| "learning_rate": 1.8466307748062205e-05, | |
| "loss": 1.6047, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.8113050706566916, | |
| "grad_norm": 0.4087926149368286, | |
| "learning_rate": 1.784388327390687e-05, | |
| "loss": 1.3402, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.8146300914380715, | |
| "grad_norm": 0.42908143997192383, | |
| "learning_rate": 1.7231100184310956e-05, | |
| "loss": 1.5664, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.8179551122194514, | |
| "grad_norm": 0.4820065200328827, | |
| "learning_rate": 1.6628030392087e-05, | |
| "loss": 1.7218, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.8212801330008312, | |
| "grad_norm": 0.4803646206855774, | |
| "learning_rate": 1.6034744670148972e-05, | |
| "loss": 1.837, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.8246051537822111, | |
| "grad_norm": 0.4350035786628723, | |
| "learning_rate": 1.5451312643206827e-05, | |
| "loss": 1.5924, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.827930174563591, | |
| "grad_norm": 0.49933725595474243, | |
| "learning_rate": 1.4877802779595762e-05, | |
| "loss": 1.6023, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.8312551953449709, | |
| "grad_norm": 0.44506213068962097, | |
| "learning_rate": 1.4314282383241096e-05, | |
| "loss": 1.4533, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8345802161263508, | |
| "grad_norm": 0.46771377325057983, | |
| "learning_rate": 1.376081758575981e-05, | |
| "loss": 1.7391, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.8379052369077307, | |
| "grad_norm": 0.44328737258911133, | |
| "learning_rate": 1.3217473338699859e-05, | |
| "loss": 1.6868, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.8412302576891105, | |
| "grad_norm": 0.4481683373451233, | |
| "learning_rate": 1.2684313405917703e-05, | |
| "loss": 1.4394, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.8445552784704904, | |
| "grad_norm": 0.452848881483078, | |
| "learning_rate": 1.2161400356095375e-05, | |
| "loss": 1.6657, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.8478802992518704, | |
| "grad_norm": 0.42388778924942017, | |
| "learning_rate": 1.1648795555397719e-05, | |
| "loss": 1.459, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.8512053200332502, | |
| "grad_norm": 0.43063634634017944, | |
| "learning_rate": 1.1146559160270875e-05, | |
| "loss": 1.6652, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.8545303408146301, | |
| "grad_norm": 0.40587228536605835, | |
| "learning_rate": 1.0654750110382628e-05, | |
| "loss": 1.5131, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.85785536159601, | |
| "grad_norm": 0.4573078751564026, | |
| "learning_rate": 1.0173426121705576e-05, | |
| "loss": 1.6047, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.8611803823773898, | |
| "grad_norm": 0.4255686104297638, | |
| "learning_rate": 9.702643679743817e-06, | |
| "loss": 1.493, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.8645054031587698, | |
| "grad_norm": 0.48064589500427246, | |
| "learning_rate": 9.242458032904311e-06, | |
| "loss": 1.6691, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8678304239401496, | |
| "grad_norm": 0.4468303620815277, | |
| "learning_rate": 8.792923186013024e-06, | |
| "loss": 1.5707, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.8711554447215295, | |
| "grad_norm": 0.4417254328727722, | |
| "learning_rate": 8.354091893977401e-06, | |
| "loss": 1.5591, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.8744804655029094, | |
| "grad_norm": 0.42065221071243286, | |
| "learning_rate": 7.926015655595254e-06, | |
| "loss": 1.5657, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.8778054862842892, | |
| "grad_norm": 0.3902848958969116, | |
| "learning_rate": 7.508744707511117e-06, | |
| "loss": 1.5445, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.8811305070656692, | |
| "grad_norm": 0.41993579268455505, | |
| "learning_rate": 7.102328018320858e-06, | |
| "loss": 1.4065, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.8844555278470491, | |
| "grad_norm": 0.4170606732368469, | |
| "learning_rate": 6.70681328282492e-06, | |
| "loss": 1.5117, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.8877805486284289, | |
| "grad_norm": 0.513680636882782, | |
| "learning_rate": 6.322246916431107e-06, | |
| "loss": 1.9662, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.8911055694098088, | |
| "grad_norm": 0.43288302421569824, | |
| "learning_rate": 5.948674049707603e-06, | |
| "loss": 1.6208, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.8944305901911886, | |
| "grad_norm": 0.38253968954086304, | |
| "learning_rate": 5.58613852308667e-06, | |
| "loss": 1.5302, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.8977556109725686, | |
| "grad_norm": 0.4266990125179291, | |
| "learning_rate": 5.2346828817197655e-06, | |
| "loss": 1.6815, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9010806317539485, | |
| "grad_norm": 0.561107873916626, | |
| "learning_rate": 4.8943483704846475e-06, | |
| "loss": 1.6608, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.9044056525353283, | |
| "grad_norm": 0.41741943359375, | |
| "learning_rate": 4.565174929145188e-06, | |
| "loss": 1.2898, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.9077306733167082, | |
| "grad_norm": 0.39722350239753723, | |
| "learning_rate": 4.247201187664218e-06, | |
| "loss": 1.585, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.9110556940980881, | |
| "grad_norm": 0.41877254843711853, | |
| "learning_rate": 3.940464461670135e-06, | |
| "loss": 1.605, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.914380714879468, | |
| "grad_norm": 0.5125119090080261, | |
| "learning_rate": 3.6450007480777093e-06, | |
| "loss": 1.5922, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9177057356608479, | |
| "grad_norm": 0.43189626932144165, | |
| "learning_rate": 3.360844720863765e-06, | |
| "loss": 1.559, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.9210307564422278, | |
| "grad_norm": 0.44040048122406006, | |
| "learning_rate": 3.0880297269979653e-06, | |
| "loss": 1.67, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.9243557772236076, | |
| "grad_norm": 0.5034830570220947, | |
| "learning_rate": 2.826587782529444e-06, | |
| "loss": 1.9225, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.9276807980049875, | |
| "grad_norm": 0.4373987019062042, | |
| "learning_rate": 2.576549568829578e-06, | |
| "loss": 1.7428, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.9310058187863675, | |
| "grad_norm": 0.45258763432502747, | |
| "learning_rate": 2.3379444289913342e-06, | |
| "loss": 1.5951, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9343308395677473, | |
| "grad_norm": 0.4411347210407257, | |
| "learning_rate": 2.110800364385812e-06, | |
| "loss": 1.5906, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.9376558603491272, | |
| "grad_norm": 0.4530499577522278, | |
| "learning_rate": 1.8951440313760837e-06, | |
| "loss": 1.4591, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.940980881130507, | |
| "grad_norm": 0.484295129776001, | |
| "learning_rate": 1.6910007381890081e-06, | |
| "loss": 1.6808, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.9443059019118869, | |
| "grad_norm": 0.43871116638183594, | |
| "learning_rate": 1.4983944419451613e-06, | |
| "loss": 1.5378, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.9476309226932669, | |
| "grad_norm": 0.4450673460960388, | |
| "learning_rate": 1.317347745847386e-06, | |
| "loss": 1.6353, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.9509559434746467, | |
| "grad_norm": 0.4276074171066284, | |
| "learning_rate": 1.1478818965281911e-06, | |
| "loss": 1.5403, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.9542809642560266, | |
| "grad_norm": 0.46902570128440857, | |
| "learning_rate": 9.900167815563465e-07, | |
| "loss": 1.5077, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.9576059850374065, | |
| "grad_norm": 0.42395344376564026, | |
| "learning_rate": 8.437709271030603e-07, | |
| "loss": 1.4276, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.9609310058187863, | |
| "grad_norm": 0.45644354820251465, | |
| "learning_rate": 7.091614957677517e-07, | |
| "loss": 1.6846, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.9642560266001663, | |
| "grad_norm": 0.5007442831993103, | |
| "learning_rate": 5.862042845640403e-07, | |
| "loss": 1.8228, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9675810473815462, | |
| "grad_norm": 0.4374752640724182, | |
| "learning_rate": 4.7491372306580627e-07, | |
| "loss": 1.6439, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.970906068162926, | |
| "grad_norm": 0.42748919129371643, | |
| "learning_rate": 3.7530287171387843e-07, | |
| "loss": 1.5968, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.9742310889443059, | |
| "grad_norm": 0.3897336721420288, | |
| "learning_rate": 2.873834202833159e-07, | |
| "loss": 1.5657, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.9775561097256857, | |
| "grad_norm": 0.5000090599060059, | |
| "learning_rate": 2.1116568651156076e-07, | |
| "loss": 1.6331, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.9808811305070657, | |
| "grad_norm": 0.4116688668727875, | |
| "learning_rate": 1.4665861488761813e-07, | |
| "loss": 1.349, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.9842061512884456, | |
| "grad_norm": 0.45309168100357056, | |
| "learning_rate": 9.386977560232879e-08, | |
| "loss": 1.5287, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.9875311720698254, | |
| "grad_norm": 0.4364264905452728, | |
| "learning_rate": 5.2805363660046734e-08, | |
| "loss": 1.5738, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.9908561928512053, | |
| "grad_norm": 0.4273874759674072, | |
| "learning_rate": 2.347019815158724e-08, | |
| "loss": 1.585, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.9941812136325852, | |
| "grad_norm": 0.42349839210510254, | |
| "learning_rate": 5.867721688690431e-09, | |
| "loss": 1.6562, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.9975062344139651, | |
| "grad_norm": 0.46075183153152466, | |
| "learning_rate": 0.0, | |
| "loss": 1.592, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9975062344139651, | |
| "eval_loss": 1.6082537174224854, | |
| "eval_runtime": 15.2561, | |
| "eval_samples_per_second": 33.233, | |
| "eval_steps_per_second": 4.195, | |
| "step": 300 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 300, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "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": 1990397395992576.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |